AWS PWE_The Practical Welding Engineer

The Practical Welding Engineer BY J. Crawford Lochhead and Ken Rodgers Brown and Root McDermott Fabricators, Ltd., Inverness, Scotland.

American Welding Society 550 N.W. LeJeune Rd. Miami, FL 331 26

International Standard Book Number: 0-87171-620-8 American Welding Society, 550 N.W. LeJeune Road, Miami, FL 33126 O 2000 by American Welding Society. All rights reserved. Text edited by Tim Heston. Printed in the United States of America

The American Welding Society is not responsible for any statement made or opinion expressed herein. Data and informationdeveloped by the authors are for informational purposes only and are not intended for use without independent, substantiating investigation on the part of potential users.

Table of Contents Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

v

Chapter 1: Contracts and Role of the Welding Engineer ................. .i Commercial Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Dealing with Specifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1 7

Chapter 2: Selection of Welding Processes, Equipment. and Consumables 13 Welding Process Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Equipment and Consumable Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 3: Weld Procedure Qualification

........................

Assessing Weld ProcedureRequirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Routine Mechanical Tesis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SimpleChecks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fracture Mechanics Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Test Failures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 4: Production Welding Control

...............................

Defect Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Welder Training and Qualification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Supervision . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Useful Aids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Consumable Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Production Weld Test Pieces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43 43 47

50 51 58 60 67 67 72

.................................

83 83 84 89 90 92 94 99

....................

101 102 106 114 120 122

................................................. .................................................

Chapter 6:Practical Problem Solving

25 25 30 36 37 39

....................

Chapter 5: Estimating and Reducing Welding Costs Estimating Welding Costs Reducing Welding Costs

13 18

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

WhatisaProblern? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Chevron Cracking in Submerged Arc Welds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Low Toughness in Selt-Shielded Flux Cored Arc Welds . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cast-to-CastVariability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . MagneticArcBlow . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elimination of Postweld Heat Treatment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Fitness for Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 7: Common Defects and Remedial Actions

Cracks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Profile Defects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Volumetric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Incomplete Fusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Some Additional Informationon SolidificationCracking . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Chapter 8: Oxyfuel Cutting, Arc Air, and Electrode Gouging .............125 OxyiuelCuiiing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125

Air Arc GouginglCuting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Electrode GougingKutting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

129 130

Appendix I: Recommended Reading ................................. 133 Appendix II: Useful Tables, Formulas, and Diagrams . . . . . . . . . . . . . . . . . . .135 Index ............................................................ 149 iii

When we, the authors, decided to write this book, we had a definite aim in mind - to present a “practical” approach to the application of welding theories.

Over recent years universities and colleges have recognized the previous lack of attention paid to the welding fraternity and subsequently greatly improved teaching capabilities and lecture contents. As a result, the modem engineer is well versed in basic metallurgical behavior; he is aware of the application of electronic wizardry to modem equipment; fracture mechanics is not just an obscure theory but a practical everyday tool; and, modem materials and consumables have apparently eliminated many of the problems of the past. What the modem welding engineer lacks is the knowledge of how to apply this knowledge in a practical sense. What we have attempted to write is basically a distillation of almost 60 years (between the two of us) of hard-gained realism in heavy engineering fabrication. The basis of the book is therefore an assumption that the reader is already knowledgeable of basic welding and metallurgical theory. He is most likely a metallurgist, materials science or mechanical engineering graduate who, during his or her university career has stumbled, or been fortuitously directed, into the welding field. It is obviously a biased view, but in the opinion of the authors, welding is one of the most exciting fields available to a young graduate. It is both vibrant and dynamic with new avenues to be explored becoming available on a regular basis. Synergic gas metal arc welding and inverter power sources, electron and laser welding, magnetic-impelled arc butt-joint welding (MIAB), robotic welding, and diffusion bonding are careers in themselves. It is difficult to identify another discipline where the range of possibilities are as diverse, broad, and exciting, and where the potentials for exploration and discovery stretch enticingly into the future. However, enough of such esoteric digressions. This book was not written from that approach. It is intended to present the inexperienced welding engineer with some “sage” advice on some of the pitfalls awaiting in the hard commercial world that awaits. Be under no illusions; it is not sufficient to be the best theoretical welding engineer in your company. You must know how to apply that knowledge in an almost “street-wise’’ manner.

Welding is regarded by many employers as a “black art.” Some of this reputation has been due to welding engineers camouflaging their inadequacies, or uncertainties, with professional jargon. Telling one’s employer that the problem is one of “cracking initiated in a highly tensile stressed region of hard martensite or body centered cubic microstructure of poor crack resistance surrounded by material of similar sensitivity to crack propagation into which atomic hydrogen has diffused, and that until the diffusion rate is beneficially altered the problem will persist,” is not clear. Telling him that you have identified the problem to be “one of delayed hydrogen cracking and that increasing the preheat temperature by 25°C will resolve it” will undoubtedly raise your standing in the company - unless you have an enlightened employer who asks you why you didn’t recognize that a higher preheat was necessary in the first place. The book is entitled “The Practical Welding Engineer.” We hope you find it to be practical. We also hope that, although you may not totally or even partially agree with its contents, you find it readable and interesting. Good Reading

J. C. Lochhead and K. J. Rodgers

Acknowledgments The authors would like to thank the following personnel for their assistance in the execution of this work. T. Clement and M. Dorricott, Managing Directors, Brown & Root Highlands Fabricators Ltd. D. J. Wright, Managing Director, Brown and Root McDermott Fabricators, Ltd. I. G. Hamilton, Consultant (for general advice). Dr. W. Welland, for assistance with run-outístub length information. Mrs. Patricia Vass and Claire Lochhead, for general secretarial assistance. All other suppliers of photographs, tables, suggestions, etc. The authors would also like to thank Training Publications, Ltd., Watford, England, for permission to use data and Figures 8.1-8.9 and 8.11-8.13 extracted from Module Manual F10 of the General Welding and Cutting for Engineering Craftsmen manual. Permission is not transferable.

vi

Contracts and the Role of the Welding Engineer This may appear to be a strange starting point for a book intended to assist a welding engineer. However, it must be appreciated at an early stage that, as is common with most disciplines, decisions based on technical judgments must be tempered with economic awareness. In general, there can be several possible solutions (and hence several possible costs) for any one problem. The principle behind every commercial venture is to make a profit, and the welding engineer must always remember that what leaves the factory gates is what pays his wages. It may leave in a timely manner, and it may be of the finest quality; but it also must be profitable. Commercial awareness usually is presented as an unessential part of the welding engineer’s discipline. This thinking is misguided because in most fabrications welding plays a primary role of cost containment. If it is not right, either technically or commercially, the company’s profitability will suffer. This is an aspect that still is not sufficiently recognized by many companies and engineers. This chapter will deal with two aspects in some detail - commercial awareness, and dealing with specifications.

1.1 Commercial Awareness This section is not intended to be a detailed study of the commercial management of a project. It is intended simply to make you, the welding engineer, aware and appreciative of the key links and actions in the chain of events that will ensure your company is fully compensated for everything it does for a client -or, conversely, receives everything it is paying for as a client. The following subjects will be discussed: 1. What is commercial awareness? 2. Making a profit. 3. The key elements of a contract. 4. Ensuring the company is fully compensated (or receives a full service). 5. Variations and claims. In all of these elements there are fundamental points applicable to the welding engineer, regardless of the size of the company in which he operates. They may not be instantly recognizable under the descriptions given. However, they will exist in some form, and the welding engineer should play a leading role in all these aspects.

2

The Practical Welding Engineer

1.1.1 What is Commercial Awareness? In simple terms, commercial awareness is the need for everyone to carry out their work in such a way that the company makes a profit. This means that estimates for welding should be constructed on the basis of sound judgments and well-defined logic, everything should be done right the first time and completed in the most cost-effective and economic manner, and everything possible should be done to maximize revenue and reduce expenditure. These objectives can be achieved only if the welding engineer is fully aware of his role and of the cost and planning parameters that control his functions.

1.1.2 Making a Profit Profit is the lifeblood of any company. The essential ingredients that will ensure a company makes a profit are a good cost and price estimate, a good plan, an ability to manage both people and work efficiently, quality (get it right the first time), safety (bad practices cost money), cost-effective execution of all work, and maximizing revenue (i.e., ensuring that the company is paid in full for everything it does).

1.1.3 Key Elements of a Contract The seven key elements of a contract are 1. the tender (i.e., the bid), 2. the plan,

3. the scope of work, 4. purchasing, 5 . subcontracting, 6 . measurement and evaluation of the work, and 7. contractual obligations.

Contracts and the Role of the Weldno Engineer

3

On first impression, the welding engineer may perceive that few of these aspects are applicable to him. This is erroneous. In fact, the welding engineer should have a fundamental role in every phase of the contract from the preparation of a tender to the fulfillment of the last contractual obligation; and greater emphasis on this role should be undertaken by the conscientious engineer. The seven key elements presented above will now be described briefly.

The Tender The key elements of a tender (i.e., the bid) that form the criteria against which the job will be measured are specifications, drawings, scope of work, procedures, resources, methods, and price. The tender describes the criteria and assumptions upon which the work is priced and planned, and it establishes the base from which all changes will be measured. Therefore, it is of paramount importance to define clearly the data and assumptions used in compiling the price and plan. In addition, it must be made clear that if the assumptions are wrong, or if they are not acceptable to the client, then there will be an effect on the price, or the delivery date, or both. All factors and calculations used in compiling the price and plan must be clearly recorded and retained throughout the life of the contract. Remember, they will form the basis for any cost adjustments resulting from changes. The Plan The plan describes how, when, and where the work will be carried out, as well as the resources to be used. There are many instances when the time allowed by a client for the tender period is very short, and the information relating to the scope of work and deliverables is incomplete. This combination of factors complicates the development of a comprehensive plan. Nevertheless, the aim should be to develop an accurate plan that represents the way the work is intended to be carried out. The plan is the base from which the effect of all changes will be measured, and this includes selfinduced changes. The Scope of Work In an ideal situation, the work would be executed strictly in accordance with the original plan and cost estimate. In the real world, however, this rarely happens -usu-

4

The Practical Welding Engineer

ally because the work is insufficiently defined at the time of the tender. It is important that the people who are responsible for executing the work are fully aware of how the work was planned and costed, so they can operate within their parameters or can identify and notify change to the same. The identification and notification of changes is the most important link in the chain of events that leads to payment for the effects of changes.

Purchasing Cost-effective purchasing is a key factor in successfully executing a contract. At the tender stage, delivery dates and prices for all required materials should be obtained. After the contract is awarded, it is important that materials are procured in accordance with the needs of the production department - that is, in accordance with the plan and within the quoted prices. Additionally, if items such as new welding machines or consumables are necessary for the job, sufficient notice should be given by the welding engineer to the relevant departments to obtain adequate quotations. Any relevant purchase lead-times also must be included in the plan. Subcontracting Regardless of the size of the subcontract. the rules are the same. The subcontract must o clearly define the scope of work, o specify the dates for deliverables to the subcontractor, o agree to a schedule for completion, and o specify the services to be provided (if any) to the subcontractor. Subsequent changes in specifications given to the subcontractor should be minimized. If this is unavoidable, any effects must be properly monitored. It is the responsibility of the welding engineer to ensure that all necessary approvals of the subcontractors’ welding procedures, etc., are made on time; otherwise, claims for consequential delays are likely to appear on his desk. Measurement and Evaluation of the Work There are a number of ways of measuring the work, but the two most common are lump-sum pricing with a schedule of rates, in which only variations are measured; and lump-sum pricing based on a bill of quantities, and a schedule of rates, in which all of the work is measured. The work is measured from the drawings, and all changes that flow through drawings should be picked up in that measurement. Of course, the increased work resulting from a change to drawings would be picked up in a subsequent re-measure and valued at the schedule rates, and the effect of the increase on the schedule would warrant a claim for extending the duration of the contract.

Contracts and the Role of the Welding Engineer

5

Changes initiated by means other than drawings are the subject of variation orders, for example, changes in specification, changes in timing, and changes in design after work has been completed. Generally, such changes would be measured as an effect on the cost of labor, equipment, and facilities and would be priced accordingly - not on the basis of the schedule of unit rates.

Contractual Obligations The major contractual obligations that affect the performance of the work are: execution of the work in accordance with drawings and specifications; execution of the work in accordance with the schedule, unless it can be proven that this has been prevented by factors beyond the company?s control; provision that work is free from defects (noting that, even where work has been inspected and/or certified, the manufacturer is liable for any defects that may be found subsequently; and, while a contractual obligation extends through to the end of the maintenance period, a common-law and/or moral obligation extends far beyond that date); appreciation that approval of drawings, method statements, weld procedures, etc., do not relieve the company from contractual obligations; appreciation that inspectors and certifications by certifying authorities do not relieve the company from contractual obligations; and knowledge that, in cases where the client causes disruption or delay to the progress of the work, the contractor has an obligation to minimize the effect of the same, provided such mitigation does not add to its cost.

1.1.4 Ensuring the Company Is Fully Compensated The welding engineer can make a significant contribution toward ensuring identification of the company?s full entitlement. The re-measurement of quantities of work and the monetary evaluation of variations issued by a client are generally straightforward. The difficulties arise with changes that affect the progress of the work,

6

The Practical Welding Engineer

the cumulative effect on the schedule of a number of changes that, individually, may have little of no effect, and the introduction of changes late into the schedule. There is no easy method for identifying or quantifying the above types of changes. However, there are two basic rules that assist in carrying out this identification and qualification: Each employee must be fully aware of, and be fully conversant with, their individual scope of work, its budget and schedule, and how their work fits into the overall plan. When a change occurs to that scope of work andor schedule, whatever the cause, then the individual concerned must immediately notify the project manager of change and ensure that its effects are quantified. In the evaluation of schedule and cost effect of all changes, the following actions will make the task simpler and more productive: Identify the change as early as possible; notify relevant personnel and/or the client; quantify the schedule and cost effects as soon as possible and within a prescribed time; keep the client informed of the effects; and, request the client’s instructions on recovery measures.

1.1.5 Variations and Claims The quality of the presentation of a variation request, or claim, can have an important bearing on the amount the contractor will be paid. A sloppy presentation will indicate either lack of knowledge on the subject or lack of confidence in any entitlement to be paid, and it will be treated accordingly by the client. Good presentation will maximize the payment. The presentation should be well prepared and built up systematically from the contract base, and it should clearly detail all effects of the change. All backup documentation should be clearly referenced and attached to the variation request. It will be much easier to achieve a high-quality presentation if all involved parties pay attention to the actions previously described. While there is often the temptation to take shortcuts on the preparation of variations, this is usually counterproductive. By good preparation and good presentation, the welding engineer will help the client to pay his company its full entitlement -and on some occasions, perhaps more.

Contracts and the Role of the Welding Engineer

7

Three main factors therefore emerge, all essential when dealing with commercial aspects: Keep good and explicit records, be vigilant, and think profit. The foregoing was a general summation of the relevant commercial aspects in which a company welding engineer should be involved during a project. However, there is one very important function in particular that deeply involves this individual - dealing with specifications. Section 1.2 will discuss this aspect in detail. Many other facets also relevant to commercial success - welding costs, choice of equipment and consumables, assessing procedure requirements etc. - are dealt with in subsequent chapters.

1.2 Dealing with Specifications International codes and specifications often vary with respect to the degree of legal influence they carry. Similar variation exists internationally in the administration of such codes and practices. In some countries there is an inspectorate -that is, a board of inspectors - that makes rulings on the interpretation of the code, approves the design, and carries out physical inspections during construction. In other countries (the U.K., for example) there is no government-approved inspectorate; instead, an independent authority is generally appointed by the purchaser to inspect on their behalf. In such a disparate legal and political environment, the only safe procedure is to work according to the code specified. However, there is no logical reason why specifications and codes related to welding fabrication should be exempt from rational and critical scrutiny, with the intent of obtaining cost reductions. Of course, the importance of welding to the overall integrity and reliability of a fabricated component must not be understated; but, by the same token, the specified requirements for materials and for finished weldments should not be regarded as sacrosanct edicts carved in stone. This awareness is especially pertinent when considering a client’s individual specifications that supplement a national code. Such additional requirements usually come about in one of two ways: from individuals who choose to incorporate certain objectives through personal experience and prejudice; and from a committee seeking to achieve the highest common denominator acceptable to all (i.e., the most rigid interpretation). The cost implications of the second approach are usually severe. One natural consequence of supplemental contract specifications is that, more often than not, they tend to place overly heavy emphasis on “how-to” rather than simply specifying what is required. In other words, they are not performance driven. If a given material is sufficient to achieve the desired results, then the welding engineer should be allowed to use it, whether it is alloy steel or chewing gum. Ultimately, such an approach could result in a welding specification comprised of just two tables: One

8

The Practical Welding Engineer

specifying the base material and weld metal properties; the other specifying any nondestructive examination requirements. Nevertheless, great care must be taken in assessing the implications of any contract specification out of the ordinary. Particularly important is the stage of negotiation at which this assessment is carried out - i.e., has a contract actually been placed, or is it still at the bid stage? If the latter, then mitigating the apprehension of the client must be the foremost consideration. Sound judgment must be used in deciding which contract specifications will have serious cost implications and which are merely advantageous to avoid, but not serious enough to jeopardize a contract award. Two convenient means can be utilized in exercising this determination. These may be labeled Exceptions to the Specijìcation and Clarifications to the Specijìcation, and they can be easily written directly into the tender. Two other possibilities exist, but these will be explained in more detail later.

Exceptions to Specification The Exceptions category should be avoided if possible, or at least restricted to those few major items where the specification demands are virtually impossible to achieve economically. The reasons for making such exceptions must be clearly identified. A common example would be a requirement to maintain preheat until a certain percentage of the weld volume has been completed. A simple illustration of this would be rolling a tubular section in the manufacture of a pressure vessel or offshore rig. It is very common for the rolling contractor to tack and root weld the longitudinal joint of the rolled cylinder when it is still in the rolls, then to transfer it later to a welding station. Maintenance of preheat throughout this process is not practicable, and abandoning this requirement can be justified based on the success of past practice. Indeed, the argument of successful past practice is a very persuasive one and should be used whenever possible. Clarifications to Specification Clar$cations to the Specification can be a subtle method of identifying what are really exceptions. These are basically in-house or preferred interpretations of sections of the specification that are unclear or ambiguous. Obviously, the interpretation most practical for the welding engineer will be preferred; but, on occasion, it is advisable for the engineer also to consider foregoing the preferred interpretation and applying the less-convenient one. In the latter instance, when a significant cost can be attributed directly to the client’s preferred or anticipated interpretation, then it should be noted specifically in the tender. If the client’s perceived benefit does not outweigh the additional cost, then a reversal of opinion will likely be forthcoming. As mentioned previously, there are two other useful tactics that fall outside of the above classifications. One is to include a passing general statement in the tender that would leave an open door for future compromises on the requirements of the contract.

Contracts and the Role of the Welding Engineer

9

No client likes to see pages of alteration to his specification, especially if much of it is relatively minor; but a convenient phrase, such as, “there are in addition a number of items on which we would welcome discussion,” can tentatively gloss over an indefinite number of exceptions and clarifications. Further discussion is often delayed until after the contract award; or, alternatively, such discussion can be deferred until the post-contract period and slowly advanced to the client under the guise of engineering queries. Small modifications in the specification to avoid changes in production activities or welding practices can be swept up rather informally by this approach without irritating the client. In addition, exceptions, clarifications, etc. - although they are common practice can reflect negatively upon the client; and it may be worthwhile, especially in pre-tender negotiations, to offer options. Although usually designed to suit the fabricator, these options also should convey to the client that acceptance of such will be advantageous to him either technically, economically, or otherwise. Consequently, these should be presented in a logical and structured fashion with client benefits clearly stressed.

Monitoring Production There is a very common pitfall of which the welding engineer must be ever mindful when dealing with specifications. It is the assumption that his interpretation of a client’s specification, if it is against the company’s practice, will be applied in production when a tender becomes a contract. Ideally, the welding engineer’s responsibilities with respect to specifications will be defined loosely enough to permit his feedback throughout the company’s departmental structure. Generally, it is better (and safer) for the company to allow this sort of follow-through on a contract, rather than assume that it will be covered by some other department. Of course, the responsibility of the welding engineer principally will be with those points in the specification dealing directly with welding activities. However, there can be instances outside of the engineer’s day-to-day responsibilities in which other departments rely on his guidance. If, for example, the engineer is aware of recent changes in welder qualification requirements, it is his obligation to convey this to others, regardless of departmental responsibilities, to ensure that the contract is executed correctly. In every industrial setting, engineers face process-control problem areas, and the welding engineer is no exception. Therefore, all specifications should be compared to the last contract and examined for changes. Never assume that the specification is identical just because the client is the same. Likewise, never assume that different clients will interpret the same specification in a similar manner. Examples of such potential problem areas are: Material Weldability -Is the steel identical to that supplied for the last contract, or should new weldability tests be carried out?

1O

The Practical Welding Engineer

D = Max. depth relative to the surface, typically 1.O, 1.5 or 2.0 mm. S = Max. space (center to center) between indentations through heat-affected zone (HAZ), typically 0.5 mm or 0.75 mm (may vary with location in survey). 1. The higher the value of S the fewer the indentations made and the less risk of encountering a “hard spot.” 2. The value of D will affect different welds in different ways depending on the weld interface shape. 3. Generally higher loads provide an averaging effect and decease the risk of reporting “hard spots.” 4. Some surveys ask for additional impressions (shown as dots above) following the weld interface. This type of survey will increase the risk of reporting high values due to the increase in the number of impressions adjacent to the maximum hardness zone.

FIGURE 1.1 -ASSESSING HARDNESS SURVEY REQUIREMENTS FOR STEEL WELDMENTS Different manufacturers can supply to the same specification using different routes, resulting in wide weldability differences. Hardness Surveys -Are the test locations and test loads similar to those previously used? Small changes to these details can change the values obtained. Some typical survey requirements are illustrated in Figure 1.1. Impact Tests - Are the acceptance values and test locations the same? Are the test temperatures specified the same? There are numerous other examples, and the welding engineer should, at the very least, draw up a mental checklist of such potential pitfalls. Having identified the differences, what should be done about them? One option would be simply to identify them as exceptions or clarifications, as shown previous-

Contructs and the Role of the Welding Engineer

11

ly; but, obviously, it would be better if they were not. A preferable option, if it were possible, would be to carry out in-house testing to ascertain'the effects of the change on the cost and time of production. Possible testing methods might include simple repeat hardness surveys, or bead-on-plate trials to examine effect of preheathardness levels. These need not be extensive or expensive, but the results can reaffirm confidence in accepting a specification. A final word of caution is extended here regarding the interpretation of suppliers' typical data (consumable or weldability data, and the like), and the relevance of this data to specification requirements. Do not assume these values are minimum or even average values; in fact, they are more likely to represent typical good results from tests carried out under ideal conditions. In cases where such typical data are close to your minimum specified requirements, take great care to avoid assuming responsibility for aspects of a specification that may prove to be technically unachievable. Such assumptions may lead your company to penalties for failing to attain specified requirements, with all the commercial implications such failures carry.

Selection of Welding Processes, Equipment and Consumables In this chapter it has been assumed that the welding engineer has a basic theoretical knowledge of the various welding processes. There are many worthwhile books available on this subject (see recommended reading), so no attempt will be made here to provide detailed information on welding processes. However, as a memory aid, Table 2.1 lists the main processes likely to be encountered. Some of the advantageddisadvantages pertaining to each are also identified.

2.1 Welding Process Selection The “ideal” welding process is that which achieves the minimum specification requirements at the minimum cost; and, although the selection of a process for a given welding application is seldom scientific or precise, it always requires careful judgement. Moreover, the approach to process selection should be sufficiently thorough to ensure balanced judgment. There are several aspects to be considered, and a careful assessment of each in turn should be undertaken by the welding engineer in close association with production personnel. The main factors to be considered are shown in Table 2.2. These factors address quality (a contractual obligation) in conjunction with resources and cost (both related to profitable operation). The correct process choice, therefore, is the best compromise between resources and cost, which also satisfies quality. Each of these aspects will now be discussed in more detail, but a summary of the selection method is given in Figure 2.1.

Specification Requirements The fabrication specification is the first and most important step in selecting a process. At this stage the engineer must establish what is required -in terms of joint type, mechanical properties, nondestructive examination (NDE), etc. - not only for the particular joint in question, but also for the overall effect of welding on tolerances, where these could influence the approach to a particular fabrication problem. Clearly, the specified requirements represent a fixed point in the process selection exercise, and, unlike the many other factors concerned, a compromise is not acceptable in terms of the minimum quality demanded by the specification. Therefore, it is the duty of the welding engineer to ensure the process, or processes, accepted at this initial stage are capable of meeting all specification requirements. A list of typical points for consideration at this stage is given below. These at least should be questioned mentally and assessed by the welding engineer prior to his decision.

14

The Practical Welding EIIQ¡neel

-roe m rom

m mz $ m

m m z

m

N

$ m

- m

TABLE 2.1 -WELDING PROCESSES

Selection of Welding Processes, Equipment and Consumables I

"pil:U

I-

!

J

FIGURE 2.1 - PROCESS SELECTION METHOD

J

15

16

The Practical Welding Engineer

Mechanical properties: tensile strength, impact toughness, higMow temperature properties, etc. NDE perjormance: visual only vs. volumetric; technique specified, acceptance levels, etc. Special features: dimensional tolerances, surface finish, etc. Weldability (i.e., special material requirements): ferrous vs. nonferrous, dissimilar or reactive metal, etc. Limited selection per speci3cation: Does the specification limit process choice directly? They often do. Consumable availability: choice limited by availability of suitable consumables

Practical Constraints Within this category are found the many and varied aspects of a fabrication method that can influence the choice of welding process. It is therefore necessary to establish the overall manufacturing sequence ahead of, or at least in parallel with, any decision on welding methods. For example, the initial selection stage may have identified three processes - shielded metal arc welding (SMAW), flux cored arc welding (FCAW), and submerged arc welding (SAW) - as suitable for a simple fillet weid. Yet, it quickly becomes evident that SAW is not suitable if the component happens to be fabricated in a sequence that places this fillet in, say, the 3G position. The meclianical properties inherent in certain combinations of processes and consumables for various welding positions also must be considered at this stage. For instance, if low-temperature impact properties are not important, then a particular self-shielded FCAW consumable could be used for 3G uphill welding, whereas if impact properties are critical [ 11, downhill welding or even another process may be required. Other factors such as accessibility, fitup, type and standard of weld preparation, etc. -can all influence the suitability of the welding process chosen. Similarly, other environmental features such as indoor (shop) vs. outdoor (site or field) fabrication have a major influence on process choice, particularly with respect to the suitability of gas shielded processes.

FACTOR

GOVERNED BY

Quality

Specification

Resources

Practical constraints Functional constraints

cost

Economic factors

TABLE 2.2 -WELDING

PROCESS SELECTION

Selection of Welding Processes, Equipment and Consumables

17

Functional Constraints Unlike the previous considerations, this group contains a number of intangible factors as well as tangible and straightforward problems. The more easily recognizable areas to be considered are availability of equipment; availability of personnel and skills; availability of services such as gas, power, water, air, etc.; and, availability of shop space. Each of the above items will influence the choice of welding process - either directly via the total unsuitability of available resources, or indirectly via the additional cost of providing suitable resources. As such, these aspects are dealt with relatively easily during the selection of a welding process. More difficult is the assessment of the sometimes-less-tangible constraints imposed on the selection decision, such as utilization of personnel (i.e., if there are a number of skilled welders from another project available on a part-time basis, economic factors may demand the use of such personnel), capacity of individual work stations (i.e., there may be existing production bottlenecks to be avoided), and overall time savings (Le., there is little point in welding a component faster unless the total production time is reduced as a result).

Economic Factors If all other factors are equal, the final choice of welding process should be made on the basis of production costs. An assessment of costs, however, involves many interrelated factors, some of which already have been mentioned. It is important to consider costs on the basis offinal cost, not on the basis of individual process costs in isolation. Thus, if a group of skilled shielded metal arc welders were available for an average of 10 hours per week (surplus to the requirements of another project), then it may be worthwhile to utilize SMAW for a particular application rather than the nominally more productive FCAW or SAW. Similarly, it may prove more economic to choose a less productive welding process to achieve some other desired feature (e.g., surface finish), where the additional time spent welding the component can benefit overall production costs by reducing machining or dressing operations later. Careful consideration should also be given to the merits of mechanization or automation; since, despite the major productivity benefits, the potential payback is highly dependent upon the degree of utilization in the plant. As a result, what may be a good investment in a production line environment (high utiliza-

78

The Praciical Welding Engineer

tion) .may prove excessively expensive in a mixed fabrication shop (low utilization), despite any improvement in the welding time for the item in question.

2.2 Equipment and Consumable Evaluation General Principles The evaluation of new equipment, or altemative consumables, can sometimes form a significant part of the welding engineer’s function, although this obviously depends on the type of business in which the engineer is employed and, in some cases, only if sufficient time is available. Nevertheless the importance of a good evaluation system should be recognized by all. As a starting point, the following questions should be posed: Why is the proposed evaluation being carried out? What are the key points of interest? If the answer to either of the above cannot be identified positively, then it is likely that the proposed evaluation is either premature or unnecessary, and of little benefit to you. It is very important to identify in advance the main factors of interest and not allow good salesmanship by your supplier to lead you into receiving a demonstration of only the best features of the equipment or consumable. These are of little value unless they are also what you require. Another point worth remembering is that by the time your evaluation is complete and your “technical” choice has been made, it may then be too late to obtain the best commercial deal with your supplier. It is therefore a good general practice to obtain quotations or pricing information at an early stage, particularly in situations where competitive products are being assessed. For both consumables and equipment, there are two general reasons leading to a need for assessing new or alternative products, namely, alteration of existing practice, e.g., replacement plant or consumables, and introduction of new practices, e.g., replacement of SMAW by semiautomatic welding. Each of the above require a different treatment. In the first case, where there will be no change in working practices, the comparison to be made should be straightforward. Here, existing equipment and consumables will form a benchmark against which the performance of the new product can be measured. It is still important, however, to approach the evaluation methodically. For this reason a checklist, or score sheet of some form, can introduce a degree of objectivity. This aspect will be discussed in more detail later. In the second case, the evaluation can be twofold in that the equipment and consumables are not only being evaluated against competitive products, but also against existing practice in terms of productivity, NDE performance, etc. This situation can lead to problems, and it is better to keep both of these aspects separate. Although this may be difficult, it is important to avoid situations where a product is being condemned on the basis of a requirement related to an existing practice, which

Selection of Welding Processes, Equbment and Consumables

19

may not be relevant if the overall working practices are changed. There is no doubting the fact that the availability of capable welding equipment and consumables will affect the decision-making process in relation to changing working practices. However, unless only one specific consumable or piece of equipment is potentially suitable, the process decision can be made based on generic information. Having made the decision in principle to change working practice, then the equipment or consumable assessment can be carried out against clearly defined target requirements.

Equipment Assessment As mentioned above, it is worthwhile to establish a checklist against which both your requirements and equipment performance can be judged. This will differ, obviously, for different types of equipment; nevertheless, the following lists are offered as examples dealing with two distinct applications. Power Source Checklist Type of current (AC or DC). o Polarity (electrode positive or negative). Pulsing facilities (peak current range, background current range, frequency range, synergic capability). 8 Programmability (e.g., preset facilities). Process capability (Shielded metal arc, submerged arc, gas metal arc [GMA], flux cored arc, and gas tungsten arc welding [GTAW]). 8 Interchangability with existing plant (e.g. spares). Power input requirements (power limitations, single-phase, threephase, type and availability of fuel for generator engine). Energy consumption (i.e. efficiency). Duty cycle. Ancillary equipment required (wire feeders, high frequency units, etc.). Availability, cost, and ease of servicing.

. . .

Orbital Gas Tungsten Arc WeldinP Unit Checklist Type of head (direct pipe mounting vs. track mounting). Power source and programmer (pulsing mechanisms, programming systems, level and number of programming steps possible for given current, voltages, wire speed, travel speed). Pipe size capacity. Ability for interchange of heads. Head facilities (wire positioning facility, wire drive on head, external arc-length or arc-voltage control, gas lens, water-cooling facilities, electrical and thermal protection, general ruggedness). Head access limitations.

20

The Practical Welding Engineer

Length of interconnects. Number of passes possible on continuous operation. Headtrack clamping methods (Le., automatic vs. manual centering, arc voltageílength monitoring mechanisms, etc.). Previous industrial experience. Availability, cost, and ease of servicing. Availability of machining facilities for weld preparation. Necessity for orbital welding (possible options such as rotation of component, etc.). The above examples are intended to illustrate the advisability of an objective approach to equipment assessment and purchase; they should not be regarded as ideal checklists. The ideal checklist is the one outlining your requirements in detail.

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Consumable Assessment The selection and assessment of consumables depends very much on the application range in view. For instance, there is little value in assessing the positional welding capability of a filler metal if the intended use is exclusively for flat-welding-position fillets. Obviously, there is a need to match the assessment to the application. Having established the target application(s), the assessment of any consumable provides two main areas for evaluation, namely, operability, and weld properties. Each of the above features is examined differently -that is, “operability” is a judgment affected by the welder’s ability and bias, whereas “weld properties” normally will present a well defined target that may or may not be achieved. The only complication regarding weld properties is that these are influenced by the detailed weld procedure used. It is recommended, therefore, that you incorporate the recommendations of the consumable manufacturer regarding specific techniques in any evaluation involving a property assessment. If these recommendations are impractical, or limiting (but necessary), then this factor in itself could eliminate a consumable from further consideration. Operability, however, is of equal importance; there is much to be said for a product that has “welder appeal.” Ease of use normally will translate into fewer defects and better productivity, so operability should be an important consideration in any evaluation. Given that operability can be a subjective assessment, it is worthwhile to establish a score sheet covering the various aspects of operability that should be addressed. An example of such a score sheet is shown in Figure 2.2. This is a particularly useful tool when evaluating manual-process consumables. Another consideration is to hear reactions from several welders, because opinions often vary. In terms of general approach, the first action would be to identify a number of consumables that meet the mechanical and chemical analysis requirements of the weld “on paper.” Having estab-

Selection of Welding Processes, Equipment and Consumables

21

CONSUMABLE ASSESSMENT SHEET

Welding Current: DC o ACO Special Tests: Welder: Date:

Electrode: Power Source: Joint Prep: Welding Position:

Amp:

EVALUATION OF WELDING CHARACTERISTICS Score*

Comment

Arc Action: Striking/Re-Striking

Weld Root Stability Fill & Cap Pass Stability

o o

0

Slag Action: Control Removal Fume Emission Coating Stability

o o O o

Deposit: Shape/Profile

o

Spatter

O

Total:

o

General Comments:

*Scale: 10 = Excellent

FIGURE 2.2

8-9 = Above Average

6-7 = Average

û-5 = Below Average

-SAMPLE SCORE SHEET FOR CONSUMABLE ASSESSMENT

22

The Practical Welding Engineer

lished such a list, samples can be obtained and used for simple operability tests. These should be designed to suit your intended application (e&, for SMAW on a fully positional pipe weld using a butt joint, a simple test involving the filling of a shallow groove in a 5G- or 6G-positioned pipe would suffice). The best two or three products can then be assessed further on the basis of full weld procedure tests to establish required properties. The “operability” factor obviously can mean different things for different processes; examples of what should be considered for shielded metal arc welding are given below: Deposition efficiency. Coating type (basic, rutile, iron powder, etc. -choices may be limited by specification). Electrode application range (current and polarity, positional limitations per available resources and applications, etc.). Electrode operability (factors to be considered and “scored” include arc action [strikinghestriking, root stability, and the stability of the cap pass]; slag action [control, removal, fume emission, coating stability, etc.]; and deposit [shape and spatter]). An example of an evaluation code that incorporates many of these features in greater detail is shown in Figure 2.3. For processes employing a bare wire electrode, there is seldom a need for an “operability” type of assessment on the wire consumable, since these usually are ordered according to an analysis specification. Other processes, especially those that involve a flux, can be treated in a fashion similar to the SMAW scenario described above. For all welding processes, including SMAW, a further consideration in many industries is the level and type of consumable-handling practices required to meet and maintain low weld-metal hydrogen values. As this can have cost implications and affect the preheat levels required, it is a factor that also must be considered before the final choice of a consumable.

References [i] Rodgers, K. J., and Lochhead, J. C. 1987. “Self shielded flux cored arc welding - the route to good toughness.” Welding Journal 66(7): 49-59.

Selection of Welding Processes, Equipment and Consumables

23

EVALUATION CODE FOR TEST WELDING

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

SLAG REMOVAL 1. Slag very difficult to remove. 2. Slag difficult to remove 3. Slag cover is whole and remains on bead but can be removed with normal de-slagging method for the type of electrode, ¡.e., wire brushing, use of chipping hammer, etc. 4. Slag cover remains on bead but is loosened up by cross cracking and is easy to remove. 5. Slag is self-releasing. Auxiliary Code SS Large areas of slag remain on bead after de-slagging. S Small areas of slag remain on bead after de-slagging. Sp Slag particles 'fly o f f during cooling. h addition to 4 if the slag loosens in one piece with light de-slagging. + used when comparing two electrodes where the difference between them is not great enough to shift from one main code to another. SPATTER 1. More spatter than normal for the type of electrode. 2. Normal spatter. 3. Less spatter than normal for the type of electrode. Note: The above may be augmented by a "+"to differentiate small differences between two electrodes.

ARC STABILITY 1. Less stable than normal for the type of electrode. 2. Normal stability 3. More stable than normal for the type of electrode. Note: The above may be augmented by "+s''if there is a tendency for the arc to extinguish, or "+n" if there is a tendency for the electrode to "stick or "íreeze."

OVERHEATING Any overheatingtendency is shown by indicating approximately how many mm of the electrode remains at the point when overheating effects are noticed. WELD BEAD APPEARANCE Two numbers are used here. The first describes bead shape in a V-Joint as follows: 1 Convex (high peaks) 2 Convex (very high peaks) 3. Flat 4. Concave A second number is used to describe bead surface smoothness (¡.e., solidification ripple pattern) as follows: i.Ripples coarser than normal for the electrode type. 2. Normal ripple pattern. 3. Ripples finer than normal for the electrode type. Note: A n additional "+"may be added to differentiate between two relatively close electrodes.

COATING BRITTLENESS The electrode is bent over a 150-mm-diameter pipe, and a scale of 1-5 is used to describe the effect on the coating. 1 =very brittle 5 =very ductile RE-STRIKING For those electrode types where this property is of interest, restriking is tried 5, 10, and 30 seconds after the arc is extinguished. Welding time before the arc is extinguished is about 10 seconds. If the electrode re-strikes then the appropriate box is marked with X. COMMENTS Any special observations are noted here, e.g., porosis: slag removal on root side, if electrode gives unusually much or little fume, if the coating breaks off around the arc, if the slag characteristics change during a test series run, if the arc column is stable in the joint, etc.

FIGURE 2.3 -SAMPLE EVALUATION CODE

Chapter 3

Weld Procedure Qualification A major part of any welding engineer’s job is the assessment, initiation, qualification, and reporting of weld procedure tests, and the engineer’s performance in this area has considerable financial implications. Significant cost penalties can result if he should fail to identify completely the specified requirement, choose consumables that prove inadequate for the function intended, or fall short of completing the proposed weld procedure qualifications within the production program requirements. The following sections discuss various finite stages to be observed during the welding procedure qualification process.

3.1 Assessing Weld Procedure Requirements --``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

During the bidding or pre-contract stage, drawings and specifications must be examined carefully to assess the number of tests that will be required, taking into consideration the thickness ranges, the material groupings, the heat treatment conditions, and the welding positions. If there is sufficient time, this initial assessment should be circulated among managers in other appropriate disciplines -such as planning, quality assurance, and, especially, production - for comment and feedback. Cognizance should be taken of any restricted-access conditions or equipment limitations; and, where necessary, alternative procedures should be proposed. Insomuch as an initial procedure-requirement estimate is seldom sufficient to accommodate client alterations, changes in fabrication methods, and other unforeseen factors, it is a good rule of thumb to overestimate by 10 percent when establishing budget requirements. Of course, this “contingency multiplier” could be increased or reduced depending on the engineer’s level of confidence in, or familiarity with, the type of work being bid. Having established the initial procedure test requirements, the engineer preparing the bid should determine whether any of the proposed procedures can be considered suitable for acceptance by virtue of being “prequalified.” Confusion can arise between the casual use of the terms “prequalified” and “previously qualified.” A prequalified welding procedure specification is defined in ANSIIAWS A3.0-94 - Standard Welding Terms and Definitions as “a welding procedure that complies with the stipulated conditions of a particular code or specification and is therefore acceptable for use under that code or specification without a requirement for qualification testing.” (author’s emphasis). In some cases, prequalification may relate to the use of code-approved procedures (e.g., AWS Dl.l), but it can equally relate to situations where previously qualified procedures (satisfying all current requirements) are the only allowable means of pre-

26

The Practical Welding Engineer

qualification. This assumes, naturally, that the relevant national or client specification permits prequalification, and that the proposed material is sufficiently similar to that on which the previous tests were performed. Even so, the engineer might consider testing a limited number of specimens to reassure both himself and the client that the materials are worthy of prequalification. Qualification is a significant factor in the cost of most fabrications; therefore, one must take advantage of prequalification whenever permissible. This is why engineers often will specify a desired procedure in terms of more than one process, each of which is prequalified separately; or, they will combine the results of several procedures into a single “hybrid” procedure, which can then be offered (with supporting data) for consideration by the client as being prequalified. At this point, one could deliberate the extent to which the engineer may apply the strategy of substituting specified procedures with prequalified procedures. For instance, it may be that certain specified procedures differ from existing qualified procedures in only minor details -e g , number of specimens, location of hardness tests, etc. Should the prequalified procedure be discounted? - not necessarily! In the interest of cost reduction, many clients will accept such procedures, especially if the rest are qualified as originally specified. However, the engineer must have enough familiarity with the client to win his confidence, as this action presumes a good deal of faith in the engineer’s judgement. Offering alternatives is an easy way to avoid cost-inflating specification details, particularly when they impact procedure qualification requirements. The engineer can always offer a small amount of additional testing once the bid is accepted. This can be a useful tactic in persuading the client to accept his recommendations. Finally, when the information at hand is inadequate to fully establish welding procedure requirements, the welding engineer must be prepared to recognize this during the bidding or pre-contract stage. Two strategies are available to the engineer in this event. First, he can assume, from background knowledge and experience, what type and number of procedure tests are likely to be required; then, these can be listed and identified to the prospective client as the total number upon which the price has been established. Second, an average price per individual test plate can be calculated; this figure can then be inserted into the bid document, leaving the final price subject to change. Most clients favor the former method, not surprisingly, as they prefer to have at least some knowledge of what the ultimate figure will be.

Planning a Test Program At this stage, the number of prequalified procedures should be removed from the pre-contract list of procedures to be tested, and the welding engineer should subsequently engage other departments, as necessary, in the preparation of a qualification test program. Priorities must be established as early as possible so that the required procedure will be qualified, reported, and accepted by the client as far in advance of the production starting date as possible. Seldom will a program run 100-percent smoothly; so, a time cushion should be included to allow for possible rewelding due

Weld Procedure Qualification

6

b P

O

O I-

r I-

?i 3 O

c

O

w

J

2

2

w

93

E !i m

!i

m

o-

t

>

d

-1

w

r

z

-I

Y

-I

W

Ed U

2

N

Y

Y

zm 3

Y

m

s

M

cl 3

3

m

N

27

28

The Pracfical Welding Engineer

to nondestructive examination (NDE) or mechanical test failures. Regardless of the number of times a procedure has performed satisfactorily in the past, statistical laws guarantee that there will be a failed result eventually, and Murphy’s Law guarantees this result will occur at a critical time. Figure 3.1 illustrates a typical weld procedure summary sheet identifying most of the relevant points mentioned in this section.

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Assessing Test Material Costs The quantity of material required must be considered carefully, since additional costs can result from underestimation as well as overestimation. A modest overestimate, however, is preferable to an underestimate that results in embarrassing program delays. The foremost requirement is to provide sufficient test material for conducting all required mechanical tests plus an allowance for retests. The importance of this extra allowance should not be discounted, as there are few experiences more frustrating than having to rerun entire procedure tests for lack of a few extra millimeters in the original test piece. In estimating the amount of weld required for mechanical test purposes, it is necessary not only to list the number of tests to be taken (making an allowance for retests), but also to identify the amount of material required per individual test piece. Also allow for the wastage of material due to machining or cutting. This issue is best discussed in advance with the testing facility performing the mechanical tests: the testing facility can often provide useful guidance on overall material requirements for individual weld procedure tests. As a simple illustration, consider the following cases: A pipe butt joint weld procedure qualification on small-bore pipe (say, 1 in. [25 mm] or less) - here, several individual butt joint welds may be required to obtain the tests needed for one weld procedure qualification. A thick plate (say, 2 in. [50 mm] or greater) butt joint weld involving Charpy impact testing at several locations. In this case, impact specimens for weld root, mid-thickness and the cap pass subsurface usually can be machined from a single through-thickness slice at a particular location: hence, the total length of weld required may be less than for some thinner plates. The importance of having some spare procedure test material should not be ignored: but the cost of providing redundant test samples must be taken into account as well, since the cost of a procedure test program can quickly escalate. Remember that the largest single expense item in a welding test program is often not the material, but labor. If all procedures in a weld procedure qualification program were based on manual welding processes (e.g., shielded metal arc welding [SMAW]), any major over-allowance on the amount of weld required could prove very costly. Conversely, for automatic and mechanized welding (e.g., submerged arc welding [SAW]) the cost of welding a 6-ft-long (2-m) test plate may not be significantly higher than welding a 3-ft (1-m) test plate; and, in this case, a provision for additional test material would be relatively inexpensive. In all cases, a common-sense approach should prevail. A

Weld Procedure Qualification

29

10- or 20-percent allowance for potential retests should be adequate for all but the most problematic of procedure tests.

Making the Test Welds Once the procedure test requirements are established, the material identified and sourced, and the test program finalized, then actual welding of test samples can begin. The foremost consideration here is that qualification of procedures should simulate the actual site conditions as closely as possible, using production equipment and production welders whenever and wherever circumstances permit. There is no benefit in qualifying a test piece under ideal conditions, using specially trained welders with better-than-normal equipment. This will only invite trouble later, in production - when welds are rejected during inspection or, worse, when they fail during mechanical testing. Welding parameters used during procedure qualification should be logged for every pass, so that the subsequent welding procedure instruction or operating sheet may be derived based on realistic working ranges for the main parameters. The importance of simulating production conditions may seem obvious, but it is not always observed in practice. Consider, for example, the use of a temper bead technique to obtain acceptable heat-affected zone (HAZ) hardness levels in alloy steels. While this can be achieved under strictly controlled conditions (such as during a procedure test), it is often impossible to guarantee or even measure -in a production environment,unless CONTROLLED SEPUENCE FOR CAP PASS : LAST BEAD MUST NOT BE ADJACENT TO BASE PLATE, BUT NO OTHER CONTROL IS EXERCISED. extensive provisions are made to supervise the operation. If specifically required for particular applications, any cap-pass sequence method to be used should be clearly stated and agreed upon with the client. Figure 3.2 illustrates the difference between a true temper-bead technique and a simple, controlled cap-pass JEMPER BEAD: LAST BEADS MUST NOT BE ADJACENT TO BASE PLATE; ANO. A sequence. The latter method, DEFINED OVERLAP BETWEEN BEADS IS SPECIFIED, AS ESTABLISHED BY EXPERalthough sometimes effecIMENT. (DIAGRAM ABOVE REFERS TO DIMENSION NOTOVERMPPED). tive in reducing HAZ hardnesses, cannot be relied FIGURE 3.2 -COMPARISON BETWEEN CONTROLLED CAP-PASS SEQUENCE AND TEMPER BEAD TECHNIQUE

upon in this capacity unless subjected to the level of control inherent in m e temperbead techniques. On completion of its welding, and prior to being machined for test purposes, the test plate should be subjected to the same NDE, heat treatment, and other postweld. operations planned for the production welds. If the weld fails at this stage (i.e., after NDE), any further action should be confirmed between the welding engineer and the client. It may be still possible to utilize the test plate if the defects found were welderinduced and unlikely to affect mechanical performance of the joint. Otherwise, a new procedure test may be required. In this case, however, the cause of the original NDE failure should be considered; and, if appropriate, the procedure should be changed prior to rewelding.

3.2 Routine Mechanical Tests The extent of mechanical testing during procedure qualification will depend on the particular application, the appropriate national standards, client specifications, etc. This section is intended to provide an overview of mechanical testing, its relevance, and control. No attempt will be made to discuss specific standards or to provide detailed test methodology. Rather, the more common weld procedure test requirements will be examined, and a number of simple checks will be recommended for use by the welding engineer in assessing both test-house capability and test results. A key point is that all unusual results should be queried (if only mentally), as it is from such results that most experience is gained. Such queries often can lead to a potential production problem being identified at an early stage, and consequently prevented.

Macro-Examination The purpose of a macro-specimen is twofold: to provide an overall view of the metallographic appearance of a weld, and to provide a cross section that can be examined for weld defects, etc. This specimen can be either a section that samples the weld in a typical or pre-specified location, or a section taken to investigate some particular problem or aspect of the weldment. Given the considerable amount of information that can be gained from simple macro-examination of a weld, one must question why the humble “macro” is so often underrated. With a detailed knowledge of the welding process, one can gain from the macro-specimen a means of establishing whether or not the weld was completed within the stated parameters. An example of such a use is given in Chapter 4. In addition, a simple bead count and bead placement check can quickly establish the accuracy of the written weld record for the procedure test in question. In production tests, placing a limit on the total number of beads, or the bead count per unit length of the weld interface, can help ensure that production welds are comparable to procedure test welds.

Weld Procedure Qualification

31

Aside from the merits of macro-examination, remember that this method is intended to be performed with minimal, if any, optical magnification (commonly xlCLx20 maximum), and care should be taken in assessing any apparent defects discovered at higher magnifications. Also, if a specification requires examination at a particular magnification, use it.

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Hardness Survey A hardness survey, normally performed on the specimen taken for macro-examination, is a common requirement of weld procedure tests. The function served by the hardness survey will vary according to the application. Probably the most widely known application relates to identifying the maximum HAZ hardness in structural/pipe steels, as this is regarded as a good indicator of the risk of encountering HAZ hydrogen cracking. In other applications, such as hardfacing, the hardness survey is important with respect to wear resistance; and, in this case, minimum hardness criteria will be specified. In terms of testing technique, the three main factors to be aware of are choice of load (commonly, Vickers Diamond 10-kg load is specified), calibration of equipment, and accuracy of placement for indentations. The last-mentioned factor is particularly important with respect to many steel fabrication specifications in which keeping below the maximum HAZ hardness is the main objective. Here, indentations are required to be within, say, 0.5 mm of the weld interface and positioned at intervals of 0.5 mm on a traverse through the HAZ. This requires accurate placement of the indentations; and, as this can have a marked influence on the results obtained, the indentation location should always be checked in such cases, with particular attention paid to any unusually low hardness values reported. For most structural steel applications, a macro-specimen employing a Nital (i.e., 10- to 20-percent Nitric acid in Methanol) etch is a long-established and normally acceptable practice for delineating the weld zones. However, some specifications call for the use of dendritic etches using, for example, a saturated solution of picric acid with a wetting agent (SASPA-NANSA), which delineate the fusion boundary more clearly, and also assist in locating the hardness indentation. The relative appearance of both types of etches on similar steel weld samples is shown in Figures 3.3(a) and 3.3(b), respectively. The use of such special etches should be governed by need rather than routine, since they require a considerably better standard of preparation (typically polished to a 1-micron finish) and therefore involve more time and cost. In addition, the SASPA-NANSAetch has been found to be unsuccessful when examining some self-shielded flux cored arc welds. On some materials, such as certain stainless steels and nickel alloys, the sample preparation can influence the result obtained in a hardness survey due to the formation of a work-hardened surface layer. Awareness of this probability should govern any assessment-of unusually high hardness values reported in these materials. Also, on these materials, avoid severe preparation methods such as heavy grinding or milling. It is best to prepare samples by progressive, light surface-grinding passes

32

The Practical Welding Engineer

directly on the as-cut surface of the specimen. Although time-consuming,this method is usually successful; if not, the use of other methods, such as electrolytic techniques, may be worthwhile considering. On some types of welds, particularly single-pass welds in steel, it is possible to get significant hardness variations along the length of the test piece; this is probably due to slight variations in heat input, heat buildup, etc. In such cases, higher hardness values are often found near the start of the weld. In these circumstances, use additional sampling for information purposes - particularly in a situation where the maximum hardness levels reported are approaching any specified limits.

Micro-Examination Micro-examination is rarely a requirement in weld procedure tests, except in cases where it can influence the serviceability of a component, and where discontinuities must be detected prior to their effect being noticed via other, more routine tests. Micro-examination procedures may include ferrite testing - that is, measuring ferrite levels, which are known to affect solidification cracking, sigmatization potential, and corrosion properties in some stainless steels. Corrosion resistance testing is an alternative, nonoptical form of micro-examination. As with macro-examination, microexamination should be carried out only after the specimen has been correctly prepared, and always at an appropriate magnification. The information that a micro-examination can provide for the welding engineer is more likely to be worthwhile in situations such as failure investigations, investigations of poor mechanical test performance, etc. - where a detailed metallurgical assessment of the weld and HAZ is often invaluable.

Nital Etch (HAZ Region)

x 500 (0)

SASPA-NANSA (HAZ Region) x 500

(b)

-

FIGURE 3.3 ETCHES OF SIMILAR STEEL WELD METALS

Weld Procedure Qualification 33

Tensile Testing The type of tensile test specimens used are variable and are normally governed by the application of a national standard or client specification. Within the scope of weld procedure testing, these fall into two main categories, namely, all-weld tensile tests (those in which only the weld metal is tested), and transverse or cross-weld tensile tests (those in which the complete weld cross section, including adjacent base material, is tested). The significance of the tensile test is readily apparent, inasmuch as the information generated has a specific design relevance to the strength of a component or structure. By pointing out this relevance, it is sometimes possible to have results that are slightly outside of specification accepted -presuming, of course, that one checks with the design or structural engineer responsible. Often, the tensile test performance is predictable, and any sudden departure from expected results is worthy of investigation. For instance, an unusually high or low result could indicate a problem with material, specimen location, specimen identification, etc.; such factors should be checked before retesting. The specimen location within a weld can influence tensile values obtained as a result of dilution effects on the weld metal analysis. This is demonstrated in Figure 3.4, which shows the effect of specimen typeAocation on results obtained in a typical structural-steel weld. In the case illustrated in diagram (a), the all-weld tensile result is shown to be affected by its “through thickness” location. This is associated with small, compositionaldifferences between the sample close to the root (more dilution) and the sample close to the final layer of the weld (less dilution). Diagram (b) shows a similar example taken from an actual procedure test. Here, because of the limited capacity of tensile testing equipment, the initial transverse tensile test was carried out as a series of overlapping specimens (an acceptable practice). The results obtained were marginally outside of the specified minimum ultimate tensile strength (UTS) and therefore deemed unacceptable by the client. Then, it was noted that previous all-weld tests performed on the same weld were acceptable, and that the transverse sample taken toward the root side of the weld was also acceptable. For the retest of this weld, it was decided to have a full-section tensile test performed at a different test establishment - where machine capacity was not a factor, and a fully acceptable retest could be obtained. This example is worth remembering, particularly when, as in this case, it is known that the weld metal strength is marginal. In general, the use of a full-size specimen should be beneficial in such situations. Another test result warranting caution would be any unexpected increase in the yield stress or yield stressAJTS ratio. Again, this could be indicative of a material problem or simply an error in calculation; but, it could be the result of incidental cold work due to improper handling of the test material. An example of the effect of previous cold work, or pre-straining, is shown in Table 3.1.

34

The Practical Welding Engineer

Yet another notable tensile-test feature in steel weldments would be the appearance of “fish eyes” on the fracture surface, as shown in Figure 3.5. Attributed to the presence of hydrogen, these are sometimes noted on welds made with cellulosic/rutile SMAW electrodes. They also occur occasionally on some self-shielded flux cored arc welds, particularly if tested in as-welded conditions. They should not be considered as defects; but, if noted on welds made with a low-hydrogen process, they are worthy of investigation (e.g., verifying that correct consumables were used).

Bend Testing Bend tests are essentially qualitative in nature, and so they do not generate data of direct relevance in engineering terms. The bend test is, however, a widely specified (and cheap) test -both as a part of weld procedure qualification and, more often, as a requirement in welder qualification tests. Although crude, the bend test is good at

DATA FROM A RANGE OF AS-WELDED SELF-SHIELDED FCAW TEST PLATES Description FL = FINAL LAVER R =WELD ROOT

Yield Stress Nlmm’ 405-455 444-485

UitimaeTensile Strength Nimm’ 485555 52û-551’

(a) Varlobillly due Io All-Weld Tensile Lowtion

DATA FROM A HEAVY SECTION SELF-SHIELDED FCA WELD (POSTWELD HEATTREATED) SPECIFIED MIN. UTS = 450 Wmm’ Description FL E FINAL LAVER M = MID R =WELD ROOT F e FULL SECTION

UltimateTensile Strength Nimm’ 439,442 442,445 483,483 478,402,482

(b) Varlablllty due IO Transverse Tensile Dimensions

FIGURE 3.4 - EFFECT OF TENSILE TEST SPECIMEN TYPE AND LOCATION

Weld Procedure Qualification

35

highlighting the presence of weld defects, and it gives a general indication of weld metal ductility. One situation where more care is required is that involving a weld metal significantly overmatched or under-matched in tensile strength compared to the base material, and similarly for transition welds involving quite different materials. Such FIGURE 3.5 EXAMPLE OF "FISH EYES" situations can have the effect of concentrating the strain in the lower strength zone, resulting in a failure not necessarily indicative of poor practice. In the former case (weld metal), the preferred solution would be to select a weld consumable more closely matched in tensility. In the latter case, a slight offset of the fulcrum position can sometimes provide a more representative test.

-

Impact Testing (e.g., Charpy Tests) The impact test, particularly the Charpy V-notch test, is widely used in materials where toughness properties are important. In structural steels, the test is usually carried out at a specified temperature, often -40°C or F; and it provides useful (although only comparative) data on the fracture toughness of the material tested. Unlike the crack tip opening displacement (CTOD) test, which is conducted to provide a direct measurement of fracture toughness, the impact test does not provide a value of direct engineering significance; rather, it produces a relatively cheap and simple way by which materials can be compared against each other over a range of temperatures. After impact testing, and by comparison against historical data, the material can be judged as safe, or otherwise, in terms of fracture toughness.

~

~

CONDITION

YIELD STRESS Nlnnm2

UTS Nlnnm'

YSAJTS

As Received (AR)

363

536

0.68

+ 2% Pre-Strain AR + 5% Pre-Strain AR + 10% Pre-Strain

406

553

0.73

500

569

0.88

583

608

0.96

AR

TABLE 3.1-

EFFECT OF PREVIOUS COLD WORK

36

The Practical Welding Engineer

In weld procedure tests on steels, it is normal practice to test both the weld metal and the heat-affected zone. In the latter case, the positioning of the notch is important; and, close attention must be paid to this point, as moving the notch by as little as 0.5 mm can often have a dramatic effect on the results obtained. Therefore, the notch locations should be checked by etching individual specimens to ensure that the correct locations have been taken. A similar procedure should be adopted prior to notching Charpy specimens to ensure correct notch location. Notch profile and test temperature also must be closely controlled. Despite its simplicity, the Charpy test is one that requires close attention to detail in order to achieve reliable results. Otherwise, the unpredictability associated with impact testing of welds (particularly H u s ) will be so chronic, it will leave the welding engineer seeking divine intervention.

3.3 Simple Checks Any test performed is of little value if the competence of the testing facility (whether in-house or independent) is questioned. The welding engineer may sometimes be in a situation where a review or witnessing of weld procedure tests is required, possibly at a subcontractor’s premises. In such a situation, the simple checks mentioned in Table 3.2 can be useful for establishing a good level of confidence in the tests being undertaken.

Subject

Check

Equipment Calibration

Verify that all pieces of equipment are uniquely identified and traceable to current calibration certificates.

Test Piece Identification

Verify how incoming test pieces are identified, and that identification is maintained during machining.

Recording of Results

Verify that all relevant data are recorded, and are previous data retrievable?

Tensile Tests

Spot check dimensions, particularly those relevant to cross section.

Impact Tests

Spot check notch profile and review methods used by test house. Check machine zero and specimen alignment. Also check bath temperature (where applicable) just before and/or during testing.

Micro/Macro-Examination

Check that a representative sample has been taken. Verify that macro corresponds to weld records, and check opposite (unprepared) face for obvious defects.

Hardness Survey

Check indentation locations. Also check load used.

Results

Query any unusual results (see previous text).

TABLE 3.2 - SIMPLE CHECKS

Weld Procedure Qualification

37

The checks presented in Table 3.2 are not intended to provide the requirements for a comprehensive quality or technical audit of a testing establishment; rather, they are provided so the welding engineer may conduct checks at an individual level, easily and informally. Any grossly unacceptable practice highlighted by such checks would, however, warrant a much more detailed assessment under formal guidelines.

3.4 Fracture Mechanics Test The tests noted thus far in this chapter form a basis for routine weld procedure qualification testing in most industrial fields and have been the norm for many years. However, in some situations (e.g., nuclear industry, offshore structural fabrication, pressure vessel fabrication, etc.) there is an increasing demand for more data on fracture toughness properties - enough so that full consideration of fracture safety can be built into the design of a structure at an early stage. The Charpy impact test, as already discussed, is an excellent ?comparator? in terms of fracture toughness; however, this test does not provide data of direct engineering relevance in terms useful to the designer. For data that can be used in such a manner, the crack tip opening displacement (CTOD) test must be carried out (usually at the design minimum temperature). In extreme cases, fullscale load-to-fracture testing or wide-plate fracture toughness testing may be required. The CTOD test is, however, the test most widely applied to welds. This test is fully

W

1

Standard Dimension

Speclmen

Subsldy Specimen

WIDTH THICKNESS NOTCH THICKNESS EFFECTWE NOTCH LENGTH EFFECTIVE CRACK LENGTH

W B = 0.5W

B-W

N

m a

FIGURE 3.6 - CRACK TIP OPENING DISPLACEMENT SPECIMEN (REFER TO STANDARD BS 7448 FOR DETAILS)

38

The Practical Welding Engineer

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

described in various national standard, and a specimen form is shown diagrammatically in Figure 3.6. Normally, the CTOD test is performed on the full section thickness of the weld. The test can be applied either to the weld metal using a notch placed at the centerline, or to the HAZ at a preselected location. The most commonly specified location for HAZ testing in steels is the coarse-grained HAZ adjacent to the weld interface. Remember, however, that this idea assumes such location represents the lowest toughness zone. A specific feature of this type of test when applied to HAZ testing is the criticality of accurately placing notch and fatigue cracks, since an error of just 0.25 mm can make a very significant difference to the values obtained. For this reason, HAZ-CTOD data must be supported by metallurgical examination reports on the broken specimens to confirm that the fatigue crack tip has indeed sampled the microstructural zones targeted. A good explanation of such examinations is now provided in various standards [3]. The necessity for accurate notch placement influences the overall approach to such a test program; and, while the testing facility technician must inevitably play a major role in the success of targeting specific microstructural areas, his chance of success is greatly affected by the standard of weld supplied for the test. Two forms of CTOD testing are relatively common, namely, through thickness notch specimen, and surface notch specimen. When testing the through thickness notch specimen, commonly carried out on a single-bevel butt joint weld, it is important that the weld interface be kept reasonably straight so the notch can sample as many areas as possible in the specified microstructure. This often means that additional precautions must be taken during welding, such as controlling wire-to-wall position in submerged arc welding to ensure that the weld interface remains straight. However, some might argue that, even with extra precautions, this method may not produce a .test representative of production conditions. When a fully representative sample is demanded, the surface notch approach can be taken; but, this method can be expected to produce a high number of microstructurally invalid test pieces (often in excess of 50 percent), which can become prohibitively expensive. Another approach is described in other literature [l, 21, based on “searching” for the zone of minimum toughness. The methods above, however, are those normally specified. Another use of the CTOD test is with respect to weldability testing for the qualification of material supply routes. This is now a fairly common requirement for offshore structural fabrication activities, obligating the steel supplier to provide fracture toughness data for all thickness ranges and heat input ranges to be applied during fabrication. Often, by presenting such data, the fabricator can avoid extensive CTOD testing as part of the weld procedure requirements. However, when reviewing such information (supplied, for example, by the steelmaker), ask the following questions: Is the data recent and does it reflect current steel chemistry and production routes?

Weld Procedure Bualifcation

39

How independent was the data? Were the welds performed by a steelmaker or by a fabricator? Are they representative of fabrication practices? Are all results reported? (Beware of data reporting only averages, as this can hide poor minimum values.) Any assessment the welding engineer makes regarding the overall acceptability of a material must take into account the above factors, as well as purely technical aspects regardless of whether the data is viewed from the specifier’s or the fabricator’s viewpoint. Fracture toughness testing of this type remains the exception rather than the rule, and it will not be required in the majority of weld procedure qualifications undertaken. Even so, the welding engineer should make himself aware of the potential for such tests. The ability of the CTOD test to provide information of direct relevance to the designer can sometimes be advantageous to the welding engineer faced with, say, procedure testing, or production-stage Charpy impact test failures. In such situations, resorting to a fracture toughness test can sometimes satisfy the client that the weld is “fit for purpose.” Another use of CTOD testing is to justify as-welded fabrication. For instance, by demonstrating good as-welded fracture toughness, the avoidance of expensive postweld heat treatment is sometimes possible (see the section on CTOD, titled Fracture Toughness Justification, in Chapter 6, page 98).

3.5 Test Failures During procedure testing, it is almost inevitable that the welding engineer will be faced with test failures. Whether these are NDE rejections or mechanical test failures, such failures immediately raise several questions. For example: Can the cause of the failure be identified? What impact, if any, will the failure have on production programs? Can the procedure test be “salvaged” via retests and/or negotiation with the client? Is a complete rethink of the proposed welding procedure required? In a well organized operation, the answer to the second question above should be known in advance, and the amount of time available to the welding engineer prior to a production requirement will obviously affect the way in which a failed procedure test should be approached. For example, if the production need is not immediate, then there may be time to fully assess the reason for the failure and take the required actions in due course. However, if there is little time to spare (or, indeed, the procedure is already late), then the welding engineer can expect little praise for providing an “ideal” solution to the problem in a week or two. A solution in this case is required immediately. In a time-sensitive situation, the engineer must act quickly to obtain a qualified procedure in the shortest possible time. This may not be the best or most productive weld

40

The Practical Welding Engineer

procedure, but a better solution can always be adopted later. In this type of situation, it is usually advisable to generate options. For instance, if your instinct tells you that it is possible to convince your client of a procedure’s fitness for purpose, then by all means pursue this course of action. In the meantime, however, rerun the procedure with a different weld preparation, consumable, or whatever is suspected to be the source of the initial problem. Delays to production are far more costly than an extra weld procedure test. So, do not waste time waiting for the answer to your first option; it may be negative. When presented with a test failure, it is important to establish the cause of the failure as soon as possible - or, at least, to rule out all non-causal factors. The cause may be attributable to human error, equipment malfunction, a metallurgical problem, or simply an unsuitable procedure. If the problem is traceable to the equipment used or to the welder (e.g., porosity related to an equipment malfunction or slag inclusions), then it is usually possible to get the procedure accepted on the basis of mechanical properties alone - possibly with the proviso of satisfactory NDE performance on the first production weld. Such occurrences should not be regarded as indicative of poor weld procedures, provided of course that the slag inclusions were not related to some adverse geometrical feature or access problem that made the weld unusually difficult to accomplish. The engineer’s reaction to failed mechanical tests should be governed to some extent by previous experience. If the procedure test was utilizing previously proven technology with respect to the consumables, then the f i s t thing to check is the source and quality of the materials and consumables. At this stage, it is also worth checking whether the same batches, casts, etc., were used in production -especially if serious doubts are arising as to their acceptability. Finally, it is necessary for the engineer to examine clearly the nature of the failure to eliminate the possibility of simple errors such as incorrectly located specimens, inaccuracy in notch location (impact tests), etc. Even if such a problem is found, the fact remains that a failed result was obtained, and this cannot be ignored. Nevertheless, close examination is required to establish where the problem lies, both technically and contractually; because, if the failure is related to HAZ or base material, then it may be your client’s problem (e.g., if the material was free issued or from a contractually specified supplier). This in itself does not solve the technical problem, and it does not absolve the welding engineer from his responsibility to solve the problem; but it may affect who pays the cost of rerunning weld procedures and, more importantly, of delays in production. Contractual responsibilities must, therefore, be borne in mind. A simple decision tree is shown in Figure 3.7 to illustrate the various points noted and actions advised. Note that Figure 3.7 is not intended to provide a fully comprehensive list of questions. The engineer must consider additional questions as necessary.

Weld Procedure Qualification

r

I

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

I

FIGURE 3.7 - FAILED TEST PLATE DECISION TREE

41

42

The Practical Welding Engineer

References [i] Rodgers, K. J. 1988. Heat-Affected Zones -A Fabricator’s Viewpoint. OMAEA 88, Paper 903. [2] Private communication of original idea, Tad Boniszewski.

[3] American Petroleum Institute standard RP-2Z, Preproduction Qualificationsfor Steel Plates for OfSshore Structures, 3rd Edition, 1998. Washington D.C.: American Petroleum Institute.

Chapter 4

Production Welding Control Production welding control is undoubtedly the sharp end of the welding engineer’s involvement in any production environment. It cannot be emphasized enough that for manual and semiautomatic applications, and to a lesser extent for machine welding processes, the welder is the most important factor in controlling weld quality. No matter how good the welding procedure appears on paper, or how advanced the consumable or equipment may be, if the welder is not properly instructed, or controlled, the chances of a poor weld resulting from his work increase dramatically. For example, if the welder feels what he is doing is incorrect, or if the fume from the electrode is causing problems, or if the equipment seems awkward or wrong, then the welder will prove that your procedure led to poor quality. To help the welder produce good quality welds, there are several factors that must be continually monitored and controlled. These are listed below in the order of importance to a welding engineer, then subsequently examined in detail. 1.

2. 3. 4. 5. 6.

Defect analysis Welder training/qualification Supervision Useful aids Consumable control Production tests

4.1 Defect Analysis This may appear to be a peculiar placing for this topic. However, unless the welding engineer is aware of what the problems actually are, he can make little progress in rectifying them. Many companies benefit from the availability of a defect analysis system. The usual format is one in which percent defect is expressed as defect length divided by weld length. This may be further divided into linear or volumetric defects. The use of weld length as an overall measure is simple to apply but lacks the effect of volume. Consequently it has no absolute meaning, so take care in interpreting such data - especially when a large amount of fillet welding is included in the overall total. In this instance, a problem with a high defect rate on full penetration butt joint welds could be easily masked.

44

The Practical Welding Engineer

Should the engineer be fortunate enough to be able to relate percent defect to individual weld procedure (again, a format greatly desired and becoming increasingly common), then a strict analytical routine should be applied to determine the cause(s) behind the defect levels being achieved. The defects will, in general, be induced by one or more of the following: geometry, equipment, consumable, procedure, or operator. A preliminary determinant can be the type of weld. Fillet weld defects are most likely to be linear, caused by one or any combination of the following: cracking, undercut, poor surface profile or overlap. With butt joints, one must ascertain if the defect is linear or volumetric in order to help pinpoint the reason. Presuming the defect can be identified, the causes listed in Table 4.1 can be examined. However, there are a number of other reasons for defects that should also be examined in conjunction with those listed.

Geometry Related The welding engineer must ask this basic question: Was the weld preparation suitable for the particular application? That automatically raises further questions. Did the welder have sufficient access or vision? Was the bevel angle too steep for adequate fusion? Was the root opening too tighvwide? Was the nose too thick or too thin? If the weld preparation was such that gouging was specified prior to second-side welding, was the backgouge too shallow? Was there too small a radius at the weld root? Do not assume, for instance, that if a backgouge depth of 8-10 mm (minimum) is called for, this will always be what is needed. In reality this range will, more often than not, need to be extended usually upward to, say, 15 mm. An examination of the weld procedure preparation and careful consideration of the location of the reported discontinuity often gives clues. If geometry is thought to be the basic cause of the discontinuities, then the necessary remedial action can be taken. This may be re-preparation, relocation of the workpiece to increase welder accesshision, or even use of simple depth gauges and profiles to ensure more accurate backgouging.

Production Welding Control

45

Welder Related The main welder-induced faults are listed in Table 4.2. Other causes more difficult to pinpoint are also possible. These include misuse of the procedure where, for example, a welder may utilize a larger gauge of electrode at an early stage where the weld preparation is unsuitable. (Indeed, he may simply be new to the process or particular technique.) Another aspect not to be overlooked is familiarity with the particular consumable itself. In this case, communication, trust, and rapport with the welder is of

Welder or Shop Floor Problem

Procedure Problem

Porosity

Dirty base metal or consumable Wrong arc length or poor technique Excessive moisture in weld area

Incorrectchoice of consumable or gas shielding Wrong preheat, Contaminatedbase metal

Tungsten inclusions

improper Inter-run cleaning Current too highhungstentoo small Poor technique Joint positioned incorrectly Use of damaged electrodes

Joint access too restrictive, preparation problem

Inadequate Fusion

Welding outside procedure incorrect eiectrodeitorchposition Weld pool running ahead of arc

Insufficientheat input Poor joint design Wrong choice of gas shielding

Inadequate Joint Penetration

Welding outside procedure Incorrectelectrodes Positioning or poor technique Poor joint fitup Incorrect backgouge

Wrong joint geometry Insufficientheat input Incorrectelectrode diameter specified Incorrectbackgougespecified

Weld metal Cracking

Poor fitup incorrect consumable handling Welding outside procedure Poor technique - crater cracking Balanced welding and/or backstep or block welding required

Joint rigidity not allowed for in choice of preheat Consumable choice and dilution effects Contamination from base material

HAZ Cracking

incorrect preheat used Incorrectconsumable used Welding outside procedure Contamination due to poor cleaning

Incorrectconsumable specified Wrong preheat specified Known base material problems not properly catered for in weld procedure

TABLE 4.1 TYPICAL DEFECTS AND CAUSES

46

The Practical Welding Engineer

overwhelming importance. Many a welding engineer has spent an inordinate amount of time investigating “red herrings” given to him by a welder with something to hide. Changes in manual or semiautomatic electrode wire from one supplier to another, or even within the same manufacturer’s range, can cause problems. A welder can become accustomed to the handling characteristics of one particular consumable; this can be at variance with another manufacturer’s assumed “equivalent” electrode.

Material Related Base materials being welded are less common sources for discontinuities, but nevertheless warrant investigation. The usual assumption is that what is supplied is correct regarding type, condition, microstructure, properties, etc. This is not always the case. Plate, forging, casting, and piping manufacturers have been known to produce out-of-specification products. Cracking can result from material having higher carbon contents than specified. Discontinuities located by ultrasonic examination have been traced to a large grain size that, according to the specification, should not be there. High carbon piano wire has even been erroneously supplied as C-Mn wire for submerged arc welding with disastrous results for the weld metal. Another material-related phenomenon is that of magnetism. During working of carbon and low-alloy steels (by gouging, grinding, or at an even earlier stage of material manufacture), the component to be welded can gain residual magnetism. If high enough, this can manifest itself in the form of “arc blow.” This problem may also have other sources related to the welding current itself. It is usually encountered when using direct current (DC) and can result in incomplete fusion, porosity, and excessive spatter (see Chapter 6). Overview The overwhelming conclusion in defect analysis is that the welding engineer must have, if possible, no preconceived ideas. The problem should be approached with an open mind, not accepting the approved or obvious without question. The variability of discontinuities and the many reasons for them require the welding engineer to investigate each instance comprehensively so that the actual reason for the discontinuity may be ascertained. Only then can the proper remedial actions be implemented. From a practical and managerial position, it always provides satisfaction to reach a definitive conclusion, but do not forget that this is not always possible. No true engineer should be afraid to state that the reasons for a problem are not completely understood. Indeed, the solution may be a combination of factors that will never be satisfactorily explained.

Production Welding Control

47

4.2 Welder Training and Qualification The training andor qualification of welders plays a vital role in the success of any fabrication business. Ultimately, the quality of the weld depends on the welder - an obvious but often forgotten fact. The various national codes for the qualification of welders provide a means by which a welder’s ability to reach a nominal quality level can be assessed. The ability of a welder to pass any national code requirement is, however, no guarantee that he meets your requirements. Often the best, and possibly the only, good reason for requiring a welder to undertake qualification to a national code is that this is a specification requirement. The welding engineer and supervisory staff must identify whether or not this requirement is the only test required, and, if not, what other practical training or workmanship testing is necessary. National codes and standards must be observed, but unless it is known that these effectively cover your needs, they should always be regarded as a minimum. Along with recognizing any need for additional practical training, the whole question of welder education, or instruction, should be addressed thoroughly. Good communication between the welding engineer and the welder can play an important role in achieving an optimum quality level. In too many cases, communication is restricted to the welding procedure only. Recognize that the weld procedure is only a limited set of welding instructions to be followed by the welder who, unfortunately, is often given little, or no, say in what is specified. Many problems with weld procedures could be overcome simply by discussing the proposed parameters with a skilled welder whose opinion can be invaluable in many practical circumstances. No critical weld should be undertaken unless there has been discussion with the welder to ensure that he understands the quality requirements and has had the opportunity to comment on the procedure. In cases where the welder’s advice is ignored, he will invariably prove the engineer wrong, resulting in a high defect rate. The welder is more likely (unconsciously, at least) to respect an engineer who values his expertise; the welder will soon recognize the value of mutual respect. Consequently the welder will be more likely to accept changes to previously standard practices that the engineer may be forced to introduce to meet specification requirements. Also remember that while the weld procedure documentation required by national codes is necessary, it may be useful and appropriate in some cases to provide welders with simplified procedure data in the form of a readable pocket-sized card. This would contain the minimum of data required and only basic working parameters. The advantage here is that the welder will always have the procedure information at hand - a key to quality assurance. Easy reference should also reduce the likelihood of working outside procedure parameters. Additional instructions given to welders can take many forms, from a casual conversation or an informal seminar, to the provision of written instructions to supplement weld procedures. Regardless of the method, the important feature should be the provision of all relevant information to the welder in a format he can understand and accept. There is little point in presenting a welder with a highly technical explanation

48

The Practical Welding Engineer

of heat flow, critical cooling rates, microstructural effects, etc., when all that you wish to ensure is correct preheat. The following statement is suggested as an adequate explanation for this specific requirement: “The faster a weld cools, the harder and more brittle the metal will become and the more likely it is to crack. Preheat is used to slow down the rate of cooling. The preheat stated on the weld procedure has been chosen to suit the material and it is important that this preheat is correctly applied.”

Type of Discontinuity

Main Causes

Porosity

Poor welding technique Incorrect setting Lack of cleaning Electrodes not dried

Slag Inclusions

Poor welding technique InsuffMent interpass cleaning

Tungsten Inclusions

Welding current too high Electrode contamination Poor welding technique Current too low Welding speed too high

Incomplete Fusion

Incomplete Penetration

Poor arc control Current too low Welding speed too high Root opening too narrow

Excess Penetration

Poor arc control Current too high Welding speed too low Root opening too wide

Undercut

Poor welding technique Current too high

Underfill

Insufficient weld layers deposited

Arc Strikes

Poor welding technique

Crater Cracking

Poor welding technique Incorrect termination of the welding arc and/or shielding gas

TABLE 4.2 WELDER INDUCED DEFECTS AND CAUSES

Production Weldin0 Control

49

In some circumstances, even just a simple, highlighted statement can suffice, such as “Preheat prevents cracking.” Some argue that the welder does not need to know, for instance, why a particular preheat is used; in strict terms this may be true. However, it is probably also true that more attention will be paid to detailed requirements where an appreciation of the need for such requirements exists; this should benefit quality levels overall. Similarly, a useful approach to reducing defect levels is to ensure that welders and other shop floor personnel are fully aware of the consequences of their particular operations - for example, how the standard of fitup or machining can influence weld discontinuities, or the importance of pre-weld cleaning on porosity, or interpass cleaning on avoiding slag inclusions. This introduces the principle of self-inspection whereby the welder is the first person with the opportunity to judge the visual acceptability of a weld. By empowering the welder to make this judgement, the company should reap benefits, establishing a principle of pride in workmanship. Actual samples illustrating these points to the welder provide an excellent communication technique. Given a piece of hardware demonstrating, for example, that welding in the uphill position instead of downhill can produce large beads and hence poor notch impact values, will make the point in a more memorable and meaningful fashion than mere words on a weld procedure. The welder should be made aware of the main discontinuities encountered in welding and, specifically, any discontinuities known to be prevalent in the particular components or material being welded. A typical list of such discontinuities can be produced for general reference, and those shown in Table 4.2 are offered as an example (together with the causes relevant to the welder). The main causes of discontinuities are in many standard textbooks, but these should be augmented by any specific knowledge from past experience. No matter how good a textbook seems, such books cannot be expected to cover all situations. It may be.that for a particular procedure or process, the parameters are particularly critical. A good example is the self-shielded flux cored arc welding of offshore structural steels. Here, the requirement for good, low-temperature toughness properties effectively restrict the type of weld procedure that can be used, despite that “defect-free” welds can be produced over a fairly wide parameter band [i].In such situations, it is even more important that the welder be well informed to ensure that he does not unwittingly “improve” production by increasing deposition rate, thereby causing problems with weld metal toughness. In just such a case, strict control over weld travel speed was required; this was monitored via relationships established with bead width (Le., controlled bead width = controlled travel speed = good toughness). An example of how strict this requirement had to be is given in the section of this chapter discussing production tests (see Production Weld Test Pieces, page 60).

50

The Practical Welding Engineer

4.3 Supervision The supervision of welded fabrication encompasses many aspects of shop floor management, Within the concept of this book, however, only the technical supervision of welded fabrication is relevant. Company structures of course vary considerably, and although ideally all personnel should be committed to producing the maximum output at an acceptable quality level, sometimes the interests of production management and technical management will appear to be in conflict. The interchange of ideas between the production supervisory staff and the welding engineer can assist in minimizing such conflict. It does this by ensuring that, as in the case of welders, supervisors are fully aware of the requirements of the weld procedure and their associated specifications. Supervisors must also be aware of the consequences of not adhering to procedure and specification. Production supervision by its nature, of course, will always tend to be more concerned with maximizing output. While technical instructions and specification rules should not be ignored or broken, you can assume they will be given their most liberal interpretation. This will also apply to the welding procedure and any general welding instructions that the welding engineer may issue. For this reason, it is important that all such requirements are both justifiable and extremely clear. The relationship between the welding engineer and the welding supervisor is crucial. The engineer may, in some situations, be totally reliant on the supervisor for the implementation of specific requirements. At the same time, the supervisor represents a first-hand source of information on problems occurring on the shop floor, thus providing an opportunity for early correction. The next section in this chapter on “Useful Aids” provides guidance on some methods that can enhance shop floor control. Knowledge of such gadgetry to supervisors is worthwhile in itself. As a general rule, individuals will be less likely to ignore requirements when a means of checking them (and the knowledge that checks are carried out) is available. Thus, if a welder knows that spot checks on weld parameters are carried out he is more likely to ensure he will be working within procedure limits. Similarly, the availability (and visibility of use) of contact pyrometers, or a range of temperature-indicator crayons, should significantly improve the application and maintenance of preheat levels. Good supervision, however, need not and should not involve constant checking of such detail. If the communication of ideas between the engineer, the supervisor, and the welder is functioning properly, then less time will be needed for such routine checks and more time will be available for addressing real problems. It is becoming more common in certain industries to allow a suitably trained supervisor to perform formal visual acceptance of completed welds. The main criteria here are the needs for appropriate training and qualification. Companies operating such practices have quickly realized benefits from less waiting time on the shop floor and greater pride in workmanship. Empowering the work force to be responsible for quality rather than trying to “inspect quality ” can be shown to increase both productivity and quality [2].

Production Welding Control

51

4.4 Useful aids There are various methods by which production welding control can be enhanced with simple tools and gauges, often at relatively low cost.

Standard Workmanship FIGURE 4.1 Examples STANDARD WORKMANSHIP EXAMPLE Exhibiting a typical component, or section of a component, in the workplace that demonstrates the required weld quality in terms of, say, fillet size or surface finish can sometimes be extremely helpful. Although this would not be appropriate or necessary in most situations, it can prove a worthwhile exercise in applications where production is regularly affected by disputes regarding quality. Such an example is shown in Figure 4.1. --``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

-

Weld Replicas As an alternative to the above, and especially useful on large projects involving many work fronts, there are methods available by which accurate epoxy replicas of weld surfaces can be produced. These lightweight samples are then available to both fabricator and client inspection personnel as necessary. One method for producing these replicas is described below. At the procedure qualification stage, the fabricator and client should select a sample from the procedure test that will form the basis of either a ?typical? or a ?worst acceptable? weld profile or surface finish. A block containing this portion, ideally about 6 in. (150 m)long (in weld direction) and trimmed to provide about 1-2 in. (25-50 mm) of base material adjacent to the weld, would be removed. This block would then be placed in a suitable container and a silicon rubber compound cast around it. After curing, this component would be separated from the block to leave a silicon rubber mold, or ?negative,? of the weld sample. As FIGURE 4.2 - WELD REPLICAS many replicas as required could then be

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

weldment has been carried out to achieve a certain profile.

Arc Monitoring' While there are some sophisticated arc monitoring equipment packages available, care must be taken to balance the expense of providing a measurement and the need for, and usefulness of, the data provided. If the tool required is intended as a means for the welding engineer, or supervisor, to ensure that the welders are working within the specified procedure tolerances, then simple hand-held meters or tong testers may be all that is needed. If, on the other hand, a detailed record with printout or graphical display is required, then the equipment needed will be considerably more expensive and probably less flexible. Where detailed monitoring of this nature is required by specification (for example, in some nuclear applications), a monitoring system can be included in a purpose-built power sourcekontrol unit. Also, a number of suitcaseor briefcase-sized portable monitoring packages are capable of printing out current, voltage, and wire feed speeds, as well as providing such ancillary functions as temperature monitoring and heat input calculations (Figure 4.4). The availability of such Portable equip- FIGURE 4.4 - PORTABLE ARC MONITORING PACKAGE ment, although not recom-

Production Welding Control

53

mended for full-time recording of data, does provide the welding engineer with something very useful. Now he has a method by which clients can be convinced (by demonstration) of details about which they may otherwise have been reluctant to accept, e.g., the consistency in operation of a particular piece of equipment, the ability of welders to work to strict tolerances or weld parameters, etc. Remember, however, that any such equipment should be calibrated and that the methods employed to calibrate production equipment and measuring equipment should be the same, or at least the differences should be understood. A good example of this last point refers to a well-known “suitcase” measuring piece of equipment that records the true mean AC current. This, however, gives an 11 percent lower value than the RMS (root mean squared) current value displayed in most standard welding plant. It is obvious that in the wrong hands such equipment identifying an apparent “error” could lead to disaster. The importance of understanding what is being measured should not be underestimated. Also note where the monitoring unit is connected, as this may not necessarily be the same position as the metered reading on the welding equipment (especially in relation to welding voltages where voltage drops can affect readings). Once again, care is required in evaluating any data produced.

Gauges Gauges come in many forms, but fall essentially into two categories: those used by inspection personnel and others for actual measurement, and those used by the welder to measure progress, check workmanship, etc. The first category is to a large extent self-explanatory and will not be discussed in detail here. These would include items ranging from accurate dimensional survey tools to simple goho-go gauges. The gauges of more immediate interest here are those used on the shop floor by the welder. These should ideally have the following characteristics: easy to use, inexpensive, and no more accurate than necessary. The types of gauges that fall into this category would be simple fillet-size gauges, backgouge depth and profile gauges, a steel ruler, root opening gauge, torch flow meters, etc. A point worth noting is that it seems that some of the best and simplest gauges in this group have been those supplied as promotional aids by consumable suppliers. It may at first appear that some of the items mentioned (for example, a ruler) are unworthy of note. However, the situation where welders are asked to produce a certain size or length of fillet, and are then left to judge this by eye, is probably rather

~

54

The Practical Welding Engineer

common. In this case, the welder is likely to err on the “safe side” by producing more weld (and hence more cost) than required. It is thus to everyone’s benefit if the welder and other shop floor personnel are provided with the ancillary tools necessary for the work at hand. This does not mean that every welder should be issued an expensive gauge that measures everything (usually slowly); it is usually better to only FIGURE 4.5 GOOD GAUGES AT A SMALL COST use a tool made for a specific required task. Such gauges can usually be made with even limited machining facilities; and, although best made in metal, in some situations a wooden gauge would suffice (particularly if intended only for short-term use). A few typical examples are shown in Figure 4.5.As can be seen, the unit cost of some of the items shown is minimal.

-

Use of Stub Lengths The use of electrode stub lengths to assist in shop floor monitoring of the shielded metal arc welding (SMAW) process is a somewhat arbitrary, yet very useful technique. It is a simple process for which the rules can be quickly derived during the qualification of procedure tests. Study of SMAW procedure qualification test data will show that there is a very strong inverse relationship between current and melt-off time, and that these two will compensate for each other almost perfectly for any given electrode size (for example, if heat input was calculated using an arbitrary constant voltage, then a constant power per unit length of electrode will result). Also, any increase in arc voltage will result from an increased arc length leading to greater heat losses by radiation, so that the changes in arc voltage do not themselves contribute to changes in melt-off rate. The consequence of this conclusion is that the only parameter that needs to be measured to detect heat input variations is the pass length.

Production Welding Control

55

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

The Significance of Heat Input Welding procedure tests, especially on high-strength, high-toughness steels used for offshore constructions, generally have two main technical objectives: to demonstrate the weldment has adequate toughness, and that the heat-affected zone (HAZ) has acceptable hardness. Tensile and bend tests are also included in destructive testing requirements, but these are relatively unaffected by variations in heat input. Since hardness in the HAZ is governed by cooling rate, the two key factors in the procedure test are the preheat or interpass temperature and the heat input. A low heat input will increase HAZ hardness but will rarely have any other adverse effect on the mechanical properties of the weldment. The minimum acceptable heat input for the level of preheat selected is, therefore, a key parameter and, for SMA welding, is reflected in the maximum acceptable electrode pass length. Weld microstructure, and therefore toughness, is also governed by cooling rate; but in this case, the adverse changes tend to result from slow cooling. Procedure tests can again be used to determine the maximum acceptable heat input for the preheat selected, which defines the minimum acceptable SMAW electrode pass length. Establishing Pass Length Limits [3] Historically, it has been common for SMA welding procedure tests to be run almost independently of the proposed welding procedure specification (WPS). The welder was presented with the test joint in the prescribed position and told to weld it. If all of the tests passed, then everyone was happy; and, although the way in which the test weld was made was often scrupulously recorded, no one seriously believed that all other welders making production welds to that procedure number would actually weld exactly the same way. This approach has a variety of unfortunate consequences. Since the heat input may vary significantly from one pass to another, no one can be quite sure of the properties developed by any given heat input, so it is not clear how essential variables limiting changes in heat input (whether directly, or in terms of the number of passes to complete a given weld groove area) should be implemented. And if someone does establish a technically justifiable way of defining pass length ranges, any determined inspector will be able to find welders depositing passes outside these limits. A change of approach was recently introduced on some offshore fabrications in the United Kingdom where the welding procedure test was used to establish and justify pass length limits, rather than being some token point of reference that bears no obvious relationship to the limits applied. To achieve this, it is essential to explain to the welder performing the test that this is not intended to represent a way a production weld would be made. Consequently, test welders should be told that the entire plate has to be completed using pass lengths within 10 percent of the figure defined in the WPS; if it is not, it will be discarded and a new one will be welded by someone more skilled.

56

The Practical Welding Engineer

The choice of pass length selected by the welding engineer becomes critically important, since a whole series of tests would normally be run using the same parameters. The “minimum” pass length is chosen to represent a value slightly shorter than any production welder is likely to use. Again, because of the number of test plates, it is essential that the weld metal be suitable for this heat input, so it is important to establish this fact -by changing consumable if necessary -before any major testing program begins. Similarly, the “maximum” pass length is chosen to represent a value slightly longer than any production welder is likely to use. Since the property at risk here is the HAZ hardness level, the critical parameter is the preheat level for this pass length. To avoid undue difficulties for the test welder, the selection of pass length limits is combined with the selection of welding position, so that the long pass length tests are performed 2G and the short pass length tests 3G. It is assumed that all welds in either 1G or 4G (or indeed, 5G or 350 6G) will use pass lengths within these limits, and so there is no justification for separate tests in these posi300 tions. This approach has been condensed in some specifications into a 250 requirement simply to qualify standard procedures in ‘‘,‘ the 2G and 3G positions to 200 cover all other positions. One concession to Diameter normal production welding 150 is permitted - variable stub lengths. Again, asking the welder to work to a con1O0 stant 50-mm stub length could put an unnecessary additional strain on his concentration, so the stub ends 50 should instead be collected for each pass. Knowing the total length of the pass, the number of electrodes used, 50 1O0 150 their original lengths, and Stub Length (mm) the total length of the stub ends, it is possible to calcuFIGURE 4.6 - MONITORING CHART FOR A TYPICAL late the length of electrode 80166 TYPE 4-MM-DIAMETER ELECTRODE used per unit length of weld.

‘.

l-7

Production Welding Control

57

From this figure, the pass length to be expected with a nominal 50-mm stub is established. Applying the 10 percent tolerance to this “full pass” value also avoids the risk of a single minor aberration producing a rejected plate. (Note that in monitoring procedure tests, reporting the results relating to only one electrode in each pass is now common practice.) If the welding engineer has done his job correctly, all of the mechanical tests will pass, and these limits on pass length will have been validated. More importantly for production welding, all of the production welders will naturally work within these limits. This welding procedure test will normally provide data for more than one diameter of electrode, and a series of graphdelectrode size can be produced.

Production Weld Monitoring by Pass Length If the pass length is the only parameter to be measured, the supervisor/inspector can choose his welder after the arc has been struck, and can then approach the welder on completion of the electrode. Only two measurements are necessary: the length of the pass, and the length of the stub end remaining. Because the pass length limits have been established on the basis of length of electrode per unit length of weld, it is possible to construct monitoring charts that define these limits for any length of stub. Different charts, one from each electrode diameter, can be developed, as shown in Figures 4.6 and 4.7. In these diagrams, the dotted lines represent the “normal” range that would result from the parameters quoted on a WPS. The bands labeled “Caution Welder” cover the 10 percent tolerance permitted on these limits to both the welder

350

300

250

E E

Y

c

200

C

al

1 c

3

150 3

K

1O0

50

50

1O0

150

Stub Length (mm)

FIGURE 4.7 - COMBINED MONITORING CHARTS FOR TYPICAL 8016G TYPE 3.25-MM AND 5.0-MM DIAMETER ELECTRODES

58

The Prucficu/ Welding Engineer

being assessed and procedure welder. Although welding within these bands will produce acceptable results, any welder found welding within the “Caution” zone should be appropriately advised, and the welding engineer should also be warned that his chosen limits may prove to be non-conservative. If a production welder is discovered welding within the “Quarantine” zone, then the resultant weld will require either removal or qualified acceptance subject to a review by the welding engineerklient, etc. This is because the properties of these more extreme welding parameters have not been established. Figure 4.7 shows that two electrode sizes can be conveniently displayed on a single chart if the limits do not overlap. This approach has the added advantage that, if the welder is aware of the supervisorhspector’s presence and breaks his arc early to avoid monitoring, the weld length he has already completed can still be assessed. This may sound as if the object of the exercise is to catch welders doing the wrong thing, but it is not. If the welding engineer has chosen his test parameters correctly, all welders will be working within them as a result of their natural techniques. The production monitoring exercise then becomes a far more convincing way of demonstrating to the client that the procedure test results really do represent the properties of production welds.

Limitations of Approach It is unlikely that a welder will use a single electrode diameter, either in production or in a procedure test, so separate limits (and separate charts) will need to be established for each size in use. The welding engineer may be ill-advised enough to believe that he should specify different pass-length limits in different parts of the weld; this causes complications. This is definitely not recommended in the root pass, since pass length here is heavily dependent on fitup. As such there is likely to be significant variation - even for a single welding position. This is not considered a major drawback as the root pass of a single-sided weld is unlikely to affect the mechanical properties measured in the procedure test and a root pass in a double-sided weld is likely to be removed by gouging.

4.5 Consumable Control To achieve good welds, the consumables used must be both correctly issued (identifiable and traceable) and in the correct condition (clean, baked, free of rust, etc., as applicable). The responsibility for consumable control will lie predominately with production supervision and welders themselves. However, the welding engineer should always ensure that the controls being exercised are sufficient and that the supervisors and welders are made fully aware of the importance of such control. Identification and traceability of consumables, of course, are not a major problem while the consumables are in the suppliers’ packaging. It is when smaller quantities are issued to the shop floor that problems may arise. Packaging may be damaged and identification lost, especially if any unused consumable is not returned to correct stor--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Production Welding Control

59

age immediately. The only acceptable rule in such situations is: i f i n doubt, throw it out.

Depending on the industry involved, the traceability of consumables may or may not be a significant requirement. Where traceability is required, then the stores system must be designed to accommodate this, with all issues and returns of consumables requiring close control. It is also important that, where required, all necessary consumable certification is checked prior to issue of materials to the shop floor. Most consumable manufacturers will have an adequate system for identifying consumables, such as the typical systems that follow: Submerged arcfluxes: Every bag or container marked with type and batch. Solid wires and flux cored wires: External packaging and individual coils of cored wire (spooled) marked with type and batch. Shielded metal arc welding (SMAw) electrodes: External packaging marked with type and batch. Individual electrodes should be marked with type as a minimum. Solid wire (straight lengths) for gas tungsten arc welding (GTAV: External packaging marked with type and batch. Some manufacturers also roll mark individual wire lengths with batch number.

If due care is taken, the use of the wrong consumable should be easily avoided. Mix-ups do occur, however, and this could lead to major problems in a mixed fabrication area. Use a color coding system as an additional precaution by paint-marking the ends of electrodes or wires in a readily identifiable manner. This ideally should not be required, but conditions are seldom ideal. This and other small expenditures may save considerably more if serious problems are avoided. Equally important is the condition of the consumable. It is only through correct storage and treatment that consumables can be delivered to the welder in the correct condition. The manner of handling the consumables will obviously depend on the type of product in question. For solid wire products, the main requirement is simply to keep the product clean and dry to avoid contamination with dust, rusting, etc. More control is required on flux-containing products such as submerged arc fluxes, manual electrodes and flux cored wires. While some products require only storage under clean, dry conditions, there are others, such as basic low-hydrogen electrodes and fluxes, that require additional treatments to ensure the product reaches the welder in its correct low-hydrogen con-

60

The Practico1 Welding Engineer

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

dition. For electrodes, this would normally involve a regime of electrode baking (typically 300-35O0C), electrode holding (typically 120-15OoC), and direct issue to welders in heated quivers (approximately 70°C). Most importantly, recognize that once a practice has been established, it must be strictly enforced; again, some form of audit on the operation of the system should be regularly carried out. In some situations, it is also sensible to carry out hydrogen determinations on product samples taken from shop floor level on a random basis. The introduction (c. 1990) of EMR (extra moisture resistant) fluxes, electrodes, and packaging is a fairly recent, but now well established, development in terms of low-hydrogen consumables. For SMAW electrodes, this offers the possibility of issuing electrodes directly to the welder in the supplier’s packaging with a guarantee of achieving very low weld metal hydrogen (-4mL/lOO g) values over fairly prolonged shop floor periods, typically 10 hours. Note that any electrodes returned to the issue store after the 10 hours must be baked and handled as per the regime noted above to ensure weld metal hydrogen values are kept low. This is achievable through the use of improved binders, and also through the introduction of specialized packaging systems. These systems are designed to eliminate moisture pickup, thereby maintaining the electrode in the condition it was manufactured. Such systems can be based on either vacuum packing or the use of atmospheric control within the packaging. The treatment of submerged arc fluxes, particularly the fully basic agglomerated types, requires special attention. This is necessitated by the often rapid deterioration, in terms of moisture pickup, which can occur if conditions are not correct. These types of flux are usually either used straight from the newly opened packaging, preheated, or baked before use. In each case, problems can occur. If used straight from bag or tin, establish that the manufacturer’s packaging is both intact and designed to deliver the product in a usable condition. If preheated or baked, the control over this operation must ensure that both the temperatures used and the times at temperature are adequately monitored. Otherwise, it is possible to increase the moisture content of a flux while in the oven if the flux is handled incorrectly. It is beneficial to ensure that the consumable supplier has been made aware of and has approved the handling techniques employed; also, have this recorded for the purpose of informing your clients.

4.6 Production Weld Test Pieces It may be surprising to find this section relegated to that of least importance in terms of production welding control. This is because such tests are “after the event,” whereas the preceding notes refer to actions that should help ensure that production tests do not become a problem area - or, indeed, justify their avoidance. In many industries, it will be a specification requirement to provide production test pieces for mechanical testing. The frequency of such tests is usually based on either time (e.g., one test per day/week) or by production quantities (e.g., one per 50 m of

Production Welding Control

61

weld, per vessel). The principal objective of production tests is usually to demonstrate that specified mechanical properties are being consistently achieved in the production environment. In either case, the most important points for the engineer are that the test is correctly identified, traceable to a known production quantity or unit, and correctly carried out. The last point may seem strange considering that by definition a production test should be fully representative of production performance. However, while achieving this is simple in the case of a longitudinal joint weld where the production test piece can be an extension of the actual production component, this cannot always be assumed in cases where a separate test piece must be set up. Here, two extremes are possible, both of which are wrong: The test piece is given a level of detailed attention far beyond normal production and as such is not representative. The test piece represents a “nuisance” to production. As a result it is not completed correctly, leading to potential failures, again not representative of production. A production test encompasses both a check on the welder and on the welding consumables. However, periodically such tests can also highlight problems with base materials that previously may have been missed. An example will be given later. It is important to be able to identify the production quantity against which the production test can be referenced. This should be straightforward, but it has been known for production departments to produce a ‘run’ of test pieces. Beware; if one of these fails, the acceptability of a much larger production quantity may be cast in doubt.

Dealing with Production Weld Test Failures The failure of a production weld test is (we hope) a rare occurrence, but one which will place the welding engineer in the limelight - or, more likely, at the whipping post. The first reactions from production could be something like: Why did you choose that consumable? Why did you change the weld parameters? What are you going to do about the problem? There are many more typical comments usually in more descriptive, colorful language, but probably the only valid question is: What can be done? The first thing to establish is the nature of the failure (e.g., welding discontinuities, mechanical properties, etc.) and identify and arrange associated production welds to be quarantined pending investigation. All relevant information must be generated quickly, and it is important that a strategy is established regarding the approach both

62

The Practical Welding Engineer

to the technical investigation and to satisfy the client as to the acceptability of the overall production run. Identification of the problem may not always be straightforward, and a few examples of problems that have occurred in the past are given later in this section. Satisfying the client organization can depend upon their perception of the problem, the urgency of their need for the component in question, and the relationship and trust previously established with the client on prior work. Obviously, there will be occasions where a production test failure must result in materials being scrapped - but this should always be a last resort. The following are typical cases of production test failures highlighting three different causes.

Example 1: Performance-Related Failure In this example, a semiautomatic self-shielded flux cored arc welding procedure had been subjected to a routine production test and failed due to low weld metal impact properties. The procedure used was well known to be sensitive to parameter variations. A minor complication was that the particular batch of consumable used had been reported as giving wire feed problems during the period preceding the failure. The questions to be addressed are, therefore, the following: Was the wire batch satisfactory and was the wire feeding problem in any way related? Was the equipment operating properly? Was the production test welded within production parameters? What could be done to convince the client of the acceptability of associated production components? It was quickly established by trial that both items 1 and 2 were noncontributory. The wire feeding problem had been related to tests on new equipment, and the equipment used for the production test had been fully checked and found to be working satisfactorily.This was important since, although any problem found may have helped to explain the problems occumng, it would also call into question all similar production equipment. The indications were, therefore, that the problem was related to the actual weld procedure used. Given that the weld exhibited no immediately obvious indications of poor practice, and the usual assurances from production personnel were obtained, it seemed unlikely that positive proof of any poor

Production Welding Control

practice could be obtained. However, despite that at first glance the macroscopic examination of this weld appeared normal, close examination and comparison with a previous “good” weld yielded interesting results. It was noted that the average layer of thickness was higher in the weld giving the low impacts; this corresponded to a reduced amount of inter-layer refinement. By using a simple technique, the macroscopic examination can also provide confirmation of incorrect practice via a calculated actual deposition or fill rate and the corresponding theoretical rate. From Macroscopic Examination (see Figure 4.8)

1.

By direct estimate, estimate area “A,” Le., area of original weld preparation excluding root face and reinforcement

(DI. 2.

3. 4.

Convert unit “A” to a volume per unit length of weld (cmVm). Count number of passes corresponding to area “A.” Calculate average volume of weld deposited per bead per unit weld length (cmVm).

FIGURE 4.8 - MACROSCOPIC EXAMINATION

From Manufacturers and/or Previous Data

5. 6.

Predict deposition rate at nominal current used (kgíh). Convert this to a volume deposition rate via density relationship (cmVmin). Based on specified procedure travel speed, calculate range of arc time per bead per unit weld length (min/m).

63

64

The Practical Welding Engineer

7.

From the above, calculate predicted range of volume of weld deposited per bead per unit weld length (cmVm).

Now, if the weld from which the macroscopic examination was sampled was completed within parameters given in the weld procedure, the value derived in Step 4 should fall within the range predicted by Step 7. Obviously the above technique is not precise, but where a large “error” is found, it would provide a reasonable pointer to the problem encountered. In the particular case involved here, such a technique was used to establish that the travel speed was about 20 percent below the minimum of the range specified. The fault, therefore, lay with the welder in this case, but it remained a problem to convince the client of the acceptability of the associated product. This was done by performing crack tip opening displacement (CTOD) tests on the failed production test plate and demonstrating that, even in this condition, the weld metal was fit for purpose on a fracture toughness basis.

Example 2: Material-Related Failure This case relates to a major base material problem discovered purely by chance due to a production weld test. On testing the particular weld involved between a pipe and flange, it was found that the flange HAZ impact properties were very poor. Immediate investigation of the problem showed it was unrelated to welding, but caused by the incorrect heat treatment of externally supplied (and certified) flange materials. Further investigation of other flanges highlighted this to be a widespread problem. Resolution of this problem involved on-site metallography on previously welded spool pieces with many subsequent complete rejections, or repeat heat treatments required at considerable expense and inconvenience. Although not a feature that could be easily controlled or predicted by the fabricator or welding engineer, this example serves as a useful warning: Do not place too much trust in cert$cation, especially i f problems become apparent.

Example 3: Consumable-Related Problems In the previous pages of this chapter, an emphasis was placed on checking consumable certification prior to issue to the shop floor. This, however, does not guarantee elimination of consumable-relat-

ed problems. An example of this is the submerged arc welding of offshore structural steels using a basic agglomerated flux. There is evidence available that for a given wire chemistry, different process applications can give significant differences in results. It has been known that residual elements can play a major role in the optimization of weld metal toughness. Among these elements, nitrogen is detrimental and, for that reason, has been limited to around 100ppm or less in consumable specifications. However, whereas using a wire with 100 ppm nitrogen has yielded acceptable results when using conventional single or tandem submerged arc procedures with and without iron powder additions, the same wire has resulted in poor impact properties when employed with a narrow-groove welding technique. The mechanisms involved in such a case are complex and this example is included merely as a warning. When employing any new practices, it should never be assumed that all the “old” rules apply. Be ready for surprises.

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Production Welding Control 6 5

These examples highlight typical production failures. They also reinforce the fact that such results do not always reflect poor production practice, as only the first such example proved to be the case. Having investigated a problem, the client must be fully satisfied as to the acceptability of the product to avoid scrapping valuable production components. Some investigative approaches have already been examined. In addition, the following can be considered: 1.

2. 3. 4.

5.

6.

Assess the production batch on a “fitness-for-purpose” basis (eg., in offshore construction, the design may permit a different impact or crack tip opening displacement (CTOD) test temperature for components below the waterline; fabrication specifications often specify the more stringent above-waterline conditions as a general requirement). Can the properties be recovered by heat treatment? Can the property requirements be relaxed? (See 1 above) Can additional test pieces be produced to demonstrate a potential “one off’ effect in relation to failure? Often previous data can be used to support such an argument. Can test pieces be taken from an actual production component to satisfy the acceptability of that production run? In the case of impact test failures, can fracture toughness tests be utilized? (See 1 above.)

This list represents common approaches. However, every problem tends to have its own individual solution that may be a combination of approaches covering more than

66

The Practical Welding Engineer

one of the above. In addition, the cost element must always be considered. It may be cheaper to scrap a number of components rather than to recover them. While this will not always be an acceptable option, it must always be considered. This is especially true where any additional required testing is likely to be extensive.

References [i] Rodgers, K.J. and Lochhead, J. C. 1987. Self-shielded flux cored arc welding the route to good toughness. Welding Journal 66(7): 49-59. [2] Lochhead, J. C., and Rodgers, K. J. 1997. The Welding Paradigm. London: International Conference on Joining and Welding for the Oil and Gas Industry, The Welding InstituteDBC U. K. Conferences, Ltd. [3] Based on an original methodology by Dr. W. Welland.

Chapter 5

Estimating and Reducing Welding Costs I

l

There are many reasons for requiring an estimate or evaluation of welding costs, e.g., in bidding for work, or the evaluation of new or alternative methods. It is also clear that to reduce welding costs you must fist identify and understand them.

5.1 Estimating Welding Costs There are several approaches to the estimation of welding costs, some specific to the company methods involved, some making a detailed assessment, others only a rough estimate. Regardless of the particular methods used, remember that the accuracy of the result is dependent on the accuracy of the input data. Where a rough estimate is all that is necessary, it may be possible to utilize published non-company data; but where a more accurate estimate is required, some measurement of achievable, or previously achieved, performance particular to your company is essential. In all cases, the following factors should be considered in the most convenient unit: labor cost, including overhead (unit cost/unit time); operating factor (OF), where OF = arc timehotal time; joint completion rate, i.e., unit weldunit time; consumable cost per unit weld (deposited); and total weld quantity. From these, the total weld cost can be obtained as follows: total weld cost =

labor cost x total weld quantity + total consumable cost operating factor x joint completion rate

Labor Costs Labor cost is the unit cost per unit time (e.g., $/hour) for a welder. It should include all the overhead costs associated with the operation as determined by the normal accountancy practices of the organization involved.

68

The Practical Welding Engineer

Operating Factor The operating factor is the ratio between the arc time and the total time spent by a welder in completing a joint. This factor is crucial to the costing exercise, since any change in the factor used has a proportional effect on the costs predicted. By definition, the operating factor will always be less than 1.O, since a value of 1.O would imply continuous welding. The operating factor is higher on automated or semiautomated processes and lower on manual processes. The range [ 11 of typical operating factors for various processes is shown in Figure 5.1. When considering operating factors, understand that different fabrication shops may achieve vastly different operating factors for essentially similar work. In Figure 5.1, a higher operating factor indicates a well-organized andor better-equipped shop. Site construction applications will, by their nature, produce lower operating factors than, say, an equivalent (in welding terms) shop application. For example, the shop welder should spend considerably less time getting to the workstation; fixturing may be better, more automatic welding equipment may be available, and handling facilities are normally better. Joint Completion Rate This factor can be expressed in a number of ways, the choice of which may best be made by the individual fabricator. Any of the following data can be used, provided the same unit quantity is used when estimating total job content. Deposition rate (lbhour, kghour, etc.) - The quantity of weld metal deposited per unit arc time. Volume $11 rate (in.3/hour,cm3/hour, etc.) - The volume of weld metal deposited per unit Mechanization Raises Operating Factor arc time. Linear completion rate (inhour, ftlhour, d o u r , etc.) The length of weld completed per unit arc time. This method is most suited to single-pass welds such as simple fillets, because other weld types such as butt joints are influenced O IO 20 30 40 50 80 70 80 DO 100 by joint thickness.

FIGURE 5.1

- EFFECT OF MECHANIZATION ON OPERATOR FACTOR ~-

From Welding Handbwk,Vol. 1, ûth Ed., American Welding Society, Miami, Fla.

‘Ost

The consumable cost should be calculated to the same unit quantity as used in the above example, i.e.,

Estimating and Reducing Welding Costs

69

consumable cost per unit weld weight, volume, or length. There are essentially two components of consumable cost to consider: 1. cost of consumable as purchased (e.g., per unit weight of electrode), and 2. cost of consumable as deposited ( e g , per unit weight of deposit). The first of these two items is identifiable from purchase invoice data, but the second must either be estimated by trials, or, more commonly, by using a deposition efficiency factor from the consumable supplier. The deposition efficiency factor is defined as follows and gives a measure of spatter loss, slag loss, stub losses, etc.: deposition efficiency factor =

weight of weld metal deposited í unit time weight of consumable used unit time

Typical values [i]for the various arc welding processes are given in Table 5.1, but it is usually better to obtain specific product data from the supplier. For gas shielded processes, the cost of supplying the shielding gas should be considered separately; the same goes for the cost of the flux for the submerged arc process. For the submerged arc process, the amount of flux consumed to produce the welding slag is typically near the same weight as the electrode consumed. However, this 1:1 ratio is never likely to be achieved in actual flux usage records, due to spillage losses, etc. As an approximate guide, a ratio of between 1.5:l and 2: 1 should be used for estimating purposes (assumFiller Metal Form and Process

Deposition Efficiency (%)*

Covered electrodes SMAW - 14 in. long SMAW - 18 in. long SMAW - 28 in. long

55 to 65 60 to 70 65 to 75

Bare solid wire SAW GMAW

95 to 99 90 to 97

Flux cored electrodes FCAW

80 to 90

* Includes slub Ias. SMAW = Shielded Metal Arc Welding; SAW = Submerged Arc Welding; GMAW = Gas Metal Arc Welding; FCAW = Flux Cored Arc Welding

-

TABLE 5.1 DEPOSITION EFFICIENCY FOR WELDING PROCESSES AND FILLER METALS From Table 8.10, Welding Handbook, Vol. I, 8th Ed., American Welding Society, Miami, Fla.

70

The Practical Welding Engineer

ing a fairly good flux recycling regime is in operation). If no flux recycling is carried out, the flux consumption is more likely to be between 3: 1 and 4: 1.

Total Weld Quantity This is the total amount of weld for which the cost estimate is being made. This can be calculated in various forms as listed below. The unit used to measure total weld quantity should have the same basis as that used for the joint completion rate.

Weld Volume - Calculate the total weld volume by measuring the weld length and multiplying the cross-sectional area of the joint preparation, making allowances for reinforcement, root openings, backgouging, etc., where applicable. We2d Weight -As above, converting to the weld weight by multiplying by the density of the weld metal. Weld Length - Measured from drawings (see note under linear completion rates made under joint completion rates). Related Costs None of the above items specifically deals with related costs, such as equipment investment costs, costs of repairs and rework, etc. These must always be considered, and it may depend on the individual fabricator’s accountancy practices as to how these costs are identified within the above general methods. Equipment costs can be built into the labor cost as part of the overhead cost; an allowance for repairs, etc., may also be dealt with in this way. Another method used is the combination of the operating factor, repair cost, etc., into an overall joint completion rate based on welder hours (not arc time). This combined rate would be based on similar previously measured work. The important point to note is that, regardless of how the estimate is made, recognition of the various points indicated must be included in some form.

ExampIe Estimate the welding cost for a 1-m-long, 50-rnm-thick mild steel test plate using a single-V (45-degree groove angle) joint preparation. All welding to be completed using the shielded metal arc welding (SMAW) process.

Item 1 Item 2 -

Labor cost: $30/hour Operating factor: 0.35 (measured from previous work)

Estimating and RedUChQ Welding Costs

Item 3 Item 4 -

Item 5 -

71

Joint completion rate: 1.7 kghour (from manufacturer’s data) Consumable cost: $lO/kg (from purchase data); deposition efficiency = 0.65 (from consumable manufacturer’s data) Total weld quantity (see estimated weld deposit weight below)

weld length = 1,000 mm (1) estimated cross section of weld = [area of weld body]+ [area of weld reinforcement]

[

= t x t x tan:]

+3t

[

= (50)2x (tan;)]

+ [(3)x (50)]

= 1,186 mm2

(2)

where t = plate thickness and x = groove angle. The allowance for reinforcement assumes a “nominal” bead height of 3 mm and a final layer width equal to plate thickness - accurate enough for this type of estimate. estimated weld volume = [Eq. 11x [Eq. 21 = 1,000 mmx 1,186 m2 = 1,186 cm3 densiîy of steel = 7.85 g I cm3 estimated weld deposit weight = [Eq. 31 x [Eq. 41 = 9,3 10 g or 9.3 1 kg

(3) (4)

Therefore, for the weld weight of 9.310 kg above, the consumable cost is 9.31 kgx $15.381 kg= $143.19

total weld cost =

labor cost x total weld quantity + total consumable cost operating factor x joint completion rate - $30/hx9.31 kg -

0.35x1.7 kgl h

+

= $469.41 $143.19

= $612.60

72

The Practical Welding Engineer

5.2 Reducing Welding Costs There are many ways of reducing welding costs. These fall into three main categories that, for the purpose of this section, will be termed direct, indirect, and design. A direct method is one that is either controlled by the welding procedure or is closely associated with the weld procedure. An indirect method is any other method influencing the overall welding performance and, hence, cost. A design method is simply one through which the designer can introduce useful cost savings using intelligent design. A few examples of methods falling into these definitions are listed below. Direct Methods (i.e,, via the weld procedure) Change from a manual process to a semiautomatic or automatic process. Introduce mechanization, robotics, etc. Change type of consumable used, e.g., to a higher recovery version. Use high deposition techniques, e.g., multi-wire submerged arc andor addition of metal powders to submerged arc welds. Use of reduced angle or narrow-groove weld joint preparations. Reduce defect levels. Indirect Methods fie., via services, etc.) Improve fixturing. Improve work environment. Improve consumable issue practice. Better utilization of facilities ( e g , availability of lifting equipment). Train personnel (refer to Chapter 4). Design Methods (i.e., at the design stage or with designer’s consent) Joint type - make optimum choice in terms of service requirements and costs. Build method and welding access - ensure that practical constraints are recognized.

All three methods are capable of yielding considerable cost reductions. The direct methods are more often those associated with the welding engineer whose detailed metallurgical and welding process knowledge is needed to evaluate these features. Compare these to the indirect methods where there is often little need for specialized welding knowledge; many of the problems can be more closely identified with pro-

Estimating and Reducing Welding Costs

73

duction engineering. For all indirect methods, there is essentially one common aim to maximize the arc time recovered from a welder or welding station. It would be a mistake to ignore the indirect route to reduced welding costs based on a possible lack of detailed knowledge. Even if the welding engineer is not capable of providing a detailed solution, he should at least be capable of highlighting the need for a solution. The best solution will normally be achieved by close cooperation between production engineers and welding engineers, both contributing their own specialized knowledge. Remember that while the indirect methods are too often ignored, they are capable of producing significant benefits. As shown in the previous section, any change that improves the operating factor produces a proportional change in costs. Cost reduction by changes to the detailed design generally improves with, and results from, previous experience. Familiarity with production methods and shop floor practices is essential to appreciate how relatively minor changes in design can have major effects on costs. However, discussions to achieve this aim must take place as early as possible, preferably during preproduction activities. Design engineers require time to check calculations and change drawings (and overcome prejudices). Time may also be required to convince clients and inspecting authorities as to the merit of such proposed changes. Some of the methods noted above will be expanded slightly, highlighting some of the potential pitfalls of cost reduction exercises rather than providing specific recommendations.

5.2.1 Direct Methods Change from a Manual Process to a Semiautomatic or Automatic Processes A change of this type would normally be expected to yield significant benefits. These benefits would come as a result of two main features: There is often (but not always) an associated increase in deposition rate with a change in process. There is usually a significant increase in the operating factor associated with increased mechanization. The above factors, although usually very persuasive regarding the potential for change, must not, however, be taken in isolation. There are other points to consider before assuming that any predicted benefits are indeed achievable. For example, if the process change proposed forms only a minor part of the overall workload or relates

74

The Practical Welding Engineer

only to sporadic business, then the additional costs in buying equipment and qualifying procedures and personnel, etc., may not be justifiable. The degree of automation available will be a major factor regarding the potential for improvement, as discussed in more detail below. The introduction of semiautomatic methods, while achieving a less dramatic increase in the operating factor (refer to Figure 5.1), is normally within the reach of all fabricators given that the capital expenditure required is usually modest in comparison with more extensive mechanization or automation proposals.

Introduction of Robotics In many respects, this is a natural extension of the above discussion, and there is again a need for a careful approach to such proposed changes. The potential gains identified by comparing the operating factors such as those given in Table 5.1 are only achievable if a higher operating factor can indeed be realized. There would be little point in committing considerable amounts of capital to the introduction of mechanized SAW (e.g., column and boom or tractor machines) to replace manual welding if, for the particular application, this involved an additional setup time, thus vastly reducing any other benefits achieved. In general, the more organized and consistent the throughput of work, the more scope there is to achieve benefits via increased mechanization of automation. The simple diagram in Figure 5.2 illustrates this idea. The intended message is simple. Utilize the potential of increased automation to its full extent, but always judge each case in detail (on its own merits), since there are undoubtedly benefits in retaining what may seem to be dated manual practices in some situations. Change of Consumable As a welding engineer, you will be continually offered alternative consumables as direct replacements within existing practices. These may be offered on the basis of reduced cost, better properties, higher productivity, etc. Guidance has been given in Chapter 2 on how to assess such offers. The most important factor when considering a change is that the welding engineer knows exactly what he requires and does not

Production Line Repetitive

Robotic -Automatic

FIGURE 5.2

Type of Work

-Mechanized

-SemiauIomatic

-

Manual

- DEGREE OF MECHANIZATION

Production Welding Control

75

allow a short-term price gain, or the availability of enhanced (but possibly over-specified) properties, to dictate the change. The main factor to remember is that reduced consumable costs, and any gain obtained through reduced consumable costs, could be quickly erased by higher defect rates, increased weld cleaning, procedure requalification costs, etc. The main reason for changing consumables within existing practices are listed below. Problems with an existing consumable (either technical or supply problems). Improved productivity (via better finish, improved deposition rate, etc.) - A particularly good example of this is that the requirements for consumables to be used for fillet welding should not necessarily be confused with those chosen for butt joints. Weld bead contour and toe profiles are often the most significant features with respect to fillet welds; here, rutile consumables may be preferred to basic low-hydrogen kinds. In many cases, a higher deposition rate electrode or process may also be worthy of consideration. Improved properties required, e.g., to meet increased specification demands. Consumable cost - As with all cost-related questions, all aspects of a proposed consumable change should be considered carefully prior to making a change. Only overall quality and production costs should be of importance in the final analysis.

The Use of High Deposition Rate Techniques The previous sections referred to the use of mechanization and automation mainly as a means of achieving higher operating factors and hence improved productivity. Productivity improvements can also’be achieved by utilizing a higher deposition rate process, causing little or no change to the operating factor. For example, in submerged arc welding, it may be possible to introduce a double- or triple-wire method to replace a single-wire application and, in doing so, reap considerable productivity improvements. However, simply multiplying by the number of wires used would not give an accurate view of the productivity improvements achievable. There is often some trade-off in terms of operating factor or procedure restrictions that would not be reflected in such a simple assessment. Remaining with submerged arc welding, it is also possible to utilize equipment delivering a metered quantity of metal powder to the weld. By doing so, it is possible to achieve a higher deposition rate with no significant increase in heat input or major change to working practice (Le., almost “consumable cost” weld metal). This method has been widely used for offshore structural fabrication [2,3] but obviously depends on the availability of suitable metal powder consumables. Similarly, the use of flux or metal cored submerged arc wires and process options, such as long stickout welding, can enhance the deposition rates of submerged arc welding [4,5].

76

The Practical Welding Engineer

Other high-deposition-rate methods worthy of consideration include electroslag welding (both in its conventional form for welding in a butt joint and as a strip cladding technique), submerged arc strip cladding methods, hot wire gas tungsten arc welding (GTAW), etc. Plate Thickness (percentages)

Single-V-groove weld, 50 degrees groove angle Single-V-groove weld, 45 degrees groove angle Double-V-groove weld, 50 degrees groove angle Narrow-groove weld (1 9 mm) (e.g., SAW) Narrow-groove weld (1 O mm) (e.g., GTAW, GMAW)

TABLE 5.2

25 mm 1O0 89 50 124

76

75mm 150mm 100 100 89 89 50 50 50 26 27

14

- COMPARATIVE WELD VOLUMES

Reduced Weld Volume This mainly applies to butt joint welding applications and introduces the concept of narrow-groove welding. Narrow-groove techniques, however, are not the only method to reduce weld volumes. A change in joint preparation to a reduced bevel or groove angle, an increased root face or a change from a single-V to a double-V preparation can all have significant effects on weld volume. Table 5.2 shows the comparative weld volumes of a number of different joint preparations at three thickness levels. In each case tabulated, the comparison is made with a single-V-groove, 50 degrees groove angle preparation and should be regarded as a rough approximation only. Clearly, major reductions in weld volumes are achievable by using a narrow-groove joint preparation on thicker sections, and, in addition, the 11 percent reduction in volume for a very simple 5 degrees reduction in weld groove angle is highlighted. This last point identifies the benefits of good workmanship in all applications showing that accurate bevel cutting to the lower end of a tolerance range can effect a useful savings with little or no additional expenditure. Although the measured volumes indicate apparently massive benefits for narrow-groove processes on thick sections, these should be treated with some caution since weld volumes should not be examined in isolation. The respective deposition rates and equipment costs must also be considered before any final judgment. For example, if narrow-groove submerged arc welding (SAW-NG) (single-wire, approx. 6-8 kghour deposition rate) is compared with tandem wire submerged arc welding with iron powder additions on a conventional V preparation (deposition rate approx. 20-22 kghour), then the material thickness at

I

Estimating and Reducing Welding Costs

77

which a break-even point is reached in terms of welding time between 80-100 mm (assuming single unit operation). Therefore, unless the introduction of such a process also provides some other overriding benefit, it would not be economical to employ SAW-NG in the above comparison unless the thicknesses involved were in excess of 100 m.Obviously, the thicker the section the more favorable the SAW-NG route would become. In addition, the use of narrow-groove practices often involves major capital expenditure on both welding equipment and machining facilities. Again, the emphasis must be placed on thoroughly examining each particular case. In fact, SAW-NG equipment is in use where one welder operates three welding stations simultaneously - a feature dramatically reducing this break-even point above. Hence, there are undoubtedly many situations where narrow-groove processes provide an appropriate avenue to reduced costs; in some situations, however, such a change would result in few, if any, benefits and could in fact be uneconomical when overall costs are considered.

Reducing Defect Levels Areduction in defect levels is an obvious route to reduce welding costs and is mentioned here merely to reinforce this point. Rectification and rework cost several times more per unit weld volume than the original work, and even a small reduction in defect levels can achieve useful savings [ 6 ] .This topic is discussed in more detail in Chapter 4.

5.2.2 Indirect Methods As noted earlier, an indirect method of reducing welding costs is one not directly related to, or applied via, the welding procedure. Indirect methods could vary from items totally removed from the workstation, e.g., improved coffee-making facilities (reducing welder downtime), to items crucial to the working practice but not specific to welding, e.g., availability of cranes. In fact, anything that will enable or encourage the welder to produce more arc time will have beneficial effect on the operating factor (and hence costs), provided of course that the expense of making this improvement is not prohibitive. A few of these methods are discussed below.

Improved Fixturing The availability of jigs and fixtures, their suitability for the job, and their ease of use, among other things, all have major bearing on welding efficiency. All time the welder spends setting up a workpiece, or assisting in such operations, could be regarded as lost production. Obviously there will always be some time spent on such operations, but the aim should be to minimize this whenever possible. Such problems need

78

The Practical Welding Engineer

not involve complex situations. For example, in a general fabrication shop where the product varies, there will be little opportunity for employing standard jigs or automatic clamping devices. There may, however, be scope for standardizing on a range of standard strongbacks, pre-cut run-odoff blocks, and the like. While these may not be obvious ways of improving operating factor, they can prove to be beneficial in some instances. Improvement is gained through a better planned operation with such apparently nonproductive operations carried out either by less costly manpower or, alternatively, by the welder during periods of “slack” time. The benefit comes by reducing the time needed by the skilled operative while “productive” work is in progress. Other equipment worth considering include positioners, turning rolls, indeed, anything that reduces time spent on repositioning workpieces. Another simple example can be found in the attachment of electrical preheating bands. These are often attached using magnetic clamps - a very sensible approach in many situations. However, consider a very large component that requires preheat to be maintained over a wide area for a prolonged period. Here, clamps would necessitate a considerable equipment investment due to the numbers involved. A simple solution could be the use of nails (2-3 mm diameter) attached by a capacitor discharge method and bent to hold the heat-

application. The greater the stanFIGURE 5.3 dardization and the higher the A NAIL HOLDS A HEATING PAD IN PLACE volume of work, then the greater the need for such an approach. Such methods vary from standard (manually fixed) jigs, through pneumatically operated jigs, to combined jigs and welding equipment that essentially automate the whole operation. An example of the last-mentioned would be an automatic machine for the production of I-beams. Here, although the weld procedure may be similar to that used on, for example, a simple tractor unit, the overall welding operation will be much more efficient due to the higher operating factor obtained. Again, however, the benefit must be weighed against the capital cost of such equipment, and this would only be viable for major producers of such beams.

Esfimofing and Reducing Welding Costs

79

Working Environment There are two ways of looking at working environment: 1.

2.

meeting regulatory requirements regarding health and safety; and maximizing the “comfort” of the work force.

While there can be no disputing health and safety requirements, all too often little regard is placed on matters beyond these requirements, Le., the comfort-level of the work force. Surveys [7] indicate that when welders have been asked to rank the most negative aspect of their work, the list is typically as follows: 1. 2.

3. 4. 5.

Fume/smoke Dust Monotony Heat (generally from high air temperature, not arc radiation) Physical strain.

Any competent welding engineer should appreciate the factors listed and make changes to reduce their effect. This can often be achieved with little cost and minimum alteration to working practices. Obviously the statement “maximize the comfort of the work force” must be taken in the context of a working environment, since real comfort should be reserved for the home. However, simple steps like providing a chair for operations that can be performed from a sitting position would reduce operator fatigue and should produce benefits. There are some foremen who would be horrified by the above suggestion, but if that same chair is not provided, the welder will most likely waste time arranging the work and/or adjacent planks and scaffolding to improve his lot. Similarly, in many welding operations, the use of preheat can result in inhospitable welding environments, e.g., working in a confined space within a preheated vessel. Here, any additional ventilation and/or insulation will help. In an extreme case where the tolerable time the welder can spend at a given location is, say, only 5 min, then even an increase of a minute of two represents a significant improvement. The working environment should not therefore be regarded purely as a set of safety regulations. It should go further than that; all ways of encouraging the welder to safely spend more time doing the job he is paid for should be explored. Often, necessq improvements come at little cost, while also (though infrequently recognized) improving the operating factor. Reject rates may also be reduced, thus giving additional cost savings.

80

The Practical Welding Engineer

The availability of tools will, for the purpose of this discussion, be included in working environment. If items such as grinders, hammers, etc., are not readily available and close to the work, then considerable time can be lost in either fetching such items from a store or waiting for someone else to finish using them. Restricted tool issue can be a false economy and should be addressed carefully. Often, the hidden cost of lost production can outweigh the visible cost of tool purchase.

Consumable Issue Practice The consumable issue practices used in the fabrication industry will vary from site to site depending on the type of work involved, the standards involved, the level of traceability required, etc. Again it may not be an obvious area for improved operability factor, but in a situation where many welders are involved and a large quantity of correctly processed electrodes must be issued and recorded at the start of a shift, then any reduction in the time required to achieve this can affect the operability factor significantly. Obviously it is better to employ a single clerk to service many welders than to have welders spend additional time on this activity. In this particular area, there is a relatively recent introduction offering the potential for significant improvement. This involves the introduction of EMR (extra moisture-resistant) electrodes and packaging systems that, by obviating the requirement to bake electrodes (basic types), can significantly simplify consumable handling and issue practices, leading to savings. Package bar codes offer the potential for computer-based recording of issues, thus enhancing both stock control and traceability. The objective of all consumable handling methods should be to get consumables in the correct condition to the welder at his place of work with the minimum of fuss and lost time. Anything that assists in the above should improve the operability factor and help to reduce welding costs. Every 10 minutes spent by the welder on consumable issue represents approximately 2 percent of a welder’s standard working day. A few welders will even spend 30 minutes or more doing this per shift - the consumable storage area is a good place to talk.

5.2.3 Design Methods Joint Type It is not the intention here to discuss the detailed design of welded joints, but rather to highlight how this aspect can influence welding costs, both favorably and adversely. A good example of this is the design of fillet welds. Unfortunately, the design engineer will often specify a specific minimum fillet size for no reason other than past practice. A change from an 8-mm-leg-length fillet to a 10-mm-leg-lengthfillet may not be crucial or even required in design terms, but it can have significant effects for the welding engineer in most situations. It is extremely difficult to consistently

Estimating and Reducing Welding Costs

81

achieve a 10-mm leg length in a single pass (even using large-gauge electrodes), so a multipass procedure will be needed. This will often require a minimum three-pass technique to maintain an acceptable weld profile. In addition, the end result will probably be a vastly overcompensated 10-mm leg length, requiring additional interpass cleaning. Discussion with the designer to ensure he is aware of such costs can prove very beneficial. A separate issue regarding fillet welds, but connected to the above, is the often general requirement of compensating for a root opening resulting from a poor fïtup. In general fabrication, a zero root opening cannot be guaranteed unless machined surfaces are specified or a fixture is utilized. For example, code requirements often require that a 2-mm root opening must be compensated for with a 2-mm addition to the leg length. Thus, a minimum 6-mm specified leg length becomes an 8-mm leg length in practice. Similarly specifying an 8-mm leg length can lead to the need for multipass welding as an actual 10-mm leg may be required. The design engineer must

Tested in shear Tested in tension

3 mm Root ODening* 252 : 272

339 : 384

Zero Root ODening* 240 : 227 300 : 341

*Maximum load to failure (kN)

Note:Shielded metal arc fillet weld: Self-shielded flux cored arc fillet weld (using E701 8-6and E61T8-K6consumables, respectively). The results of these tests indicate that, despite similar (uncompensated) leg lengths, the strengths of the open root fillets were higher than those from a closed root in this example.

-

TABLE 5.3 WELD STRENGTHS OF 100-MM FILLET WELD WITH 6-MM LEG LENGTH. again be made aware of this point and asked to pay particular attention to the specification of fillet welds around these sizes. From the welding engineer’s viewpoint, it is sometimes permissible to use weld penetration to compensate for leg length. It also may be possible in some circumstances to demonstrate that adequate fillet weld strength has been achieved without weld leg length compensationby performing a simple test program. The results shown in Table 5.3 were obtained in a simple test of this type based on a 100-mm length of fillet weld in each case. This is a particularly useful exercise where the amount of fillet welding is high and a change from single- to multipass welding would have major cost implications. Other important aspects of weld joint design are the specification of partial penetration welds rather than full penetration welds and designing for “buildability,” e.g.,

82

The Practico/ Welding Engineer

taking access and fabrication sequences into consideration. Although many of the aspects may not be under the direct control of the welding engineer, it is important that you, as a welding engineer, make every effort to inform the designer of such problems and of the advantages obtained by considering all of these factors.

Build Methomelding Access As mentioned previously, your familiarity with your company?s procedures, techniques, equipment, and skill availability can lead to productive design changes. Design can (and should) be altered to suit manufacturing methods provided you are in early communication with the design engineer. The design engineer generally would not be familiar with such details and consequently will (or should) rely on your guidance and assistance in assessing any such situations. Intimate knowledge of these resources and a willingness to challenge historical or conventional wisdom to benefit your company?s profitability are the best requisites pertaining to design changes. The above examples represent only a few of the areas where welding costs can be reduced. The intended message of this chapter as a whole is to encourage a broad view of welding costs, not simply of deposition rates, etc., which can often be misleading. The only cost that is important is the overall cost, which can be affected in many different ways. Therefore, as a welding engineer, you must always be conscious of costs and the cost implications of your decisions. Always try to look at the overall operation costs - not just at welding procedures in isolation.

References [i] Connor, L. P. (ed.). 1987. Welding Handbook. Vol. 1, 8th Ed. American Welding Society, Miami, Na. [2] Rodgers, K. J. and Lochhead, J. C. 1987. Submerged arc welding metal powder additions, productivity and properties. Welding Journal 66 (10): 21-27. [3] Fraser, R., et al. 1982. High deposition rate submerged arc welding for critical applications. Int. Con$ Offshore Welded Structures. London, U.K. [4] Lochhead, J. C., and Rodgers, K. J. 1997. The Welding Paradigm. London: International Conference on Joining and Welding for the Oil and Gas Industry, The Welding InstituteABC U.K. Conferences, Ltd. [ 5 ] Lochhead, J. C., and Bews, R. O. 1998. The use of mechanised and latest cored wire technology in the construction of a 32,000-ton production jack up. International Conference, Exploiting Advances in Arc Welding Technology. [6] Lochhead, J. C., and Rodgers, K. J. 1999. Weld Defects - Considering the Big Picture. Welding Journal 78( 10): 49-54. 161 Sundin, J. 1990. Work environment for welders. Svetsen. Special issue.

Practical Problem Solving Problem solving is perhaps the welding engineer’s most interesting task. Here, he is called upon to be the “Columbo” of his company. The solution of welding problems requires a combination of theoretical knowledge, practical experience, and, on occasion, intuition. Problem solving can be both most rewarding and extremely frustrating.

6.1 What is a Problem? A typical dictionary definition states that a problem is “a matter which is difficult to deal with or solve” or “is a question set for solution.” Welding problems certainly fit the above general definitions, but numerous alternative views pop up from time to time. A good reason for a vacation. Time to find a scapegoat. Time to change the specification. While any of the above (or other similar views) may well come to mind, it is much more productive and interesting to regard welding problems as opportunities. An opportunity to learn. An opportunity to improve practices. An opportunity to display your worth to your employer. The last item above leads to the question: Would your employer need a welding engineer if no welding problems were likely? Welding problems, in this sense, are the lifeblood of the welding engineer’s profession, and although your main function should always be aimed at preventing problems, there will always be a need for prompt and efficient reaction to welding problems as they arise. This chapter outlines five specific examples of welding-related problems and also introduces a general “fitness for purpose” concept that offers an alternative solution to many problems. It is hoped these will provide an insight into problem-solving in general, as well as providing some specific guidance on the topics discussed.

84

The Practical Welding Engineer

The following will be examined: chevron cracking in submerged arc welds, low toughness in self-shielded flux cored arc welds (FCAW-S), magnetic arc blow. postweld heat treatment (PWHT) avoidance, and cast-to-cast variability,

6.2 Chevron Cracking in Submerged Arc Welds Chevron cracking in submerged arc welds (SAW) in carbon, medium- and lowalloy steels is a kind of weld metal cracking generally recognized as being a form of hydrogen cracking [i],although other mechanisms have been proposed [2]. The name chevron cracking relates to the typical appearance of this kind of crack - roughly Vshape (further information on this point is given in Chapter 7). While regarded as a problem mainly in submerged arc weld metals, it is also possible to encounter this general form of discontinuity in shielded metal arc welding (SMAW). Chevron cracking, once found, is often difficult to deal with because of its several potential causes. However, treating this discontinuity simply as weld metal hydrogen cracking identifies the main causal features (similar to heat-affected zone [HAZ] hydrogen cracking). Presence of hydrogen. Presence of stress, i.e., restraint. Time. Note that the technique employed for the nondestructive examination (NDE) of welds - particularly of submerged arc welds - should be designed with chevron cracking in mind. Due to the orientation of chevron cracks in submerged arc welds, they could be missed unless an ultrasonic testing (UT) scan, designed for their detection, is incorporated. Chevron cracking has mainly been associated with welds in materiais more than 50 mm thick, but cases are known in thicknesses as low as 22 mm. Additionally, chevron cracking is a delayed form of cracking, normally not appearing until 12 hours after cessation of welding. Allow at least 24 (or, more commonly, 48) hours prior to the required nondestructive examination.

Detection The ultrasonic technique for detection of chevron cracking involves the use of a 45-degree probe, scanning longitudinally along the surface of submerged arc welds or, alternatively, along the finished surface of manual or semiautomatic welds. As it is

Practical Problem Solving

85

most unlikely that chevron cracking will occur in processes other than SAW, the requirement for additional NDE is normally ignored for these procedures. This scan is carried out as a supplement to other scans, specifically to locate discontinuities lying in the transverse plane of the weld. The 45-degree probe is utilized because chevron cracks present a reflecting face within the range of 30-50 degrees to normal. Conventional scans from either side of the weld surface will not resolve such discontinuities for two reasons: 1. A transverse planar discontinuity will not present a reflecting surface of sufficient magnitude to produce a signal response. 2. The major dimensions of such discontinuities lie in a different plane to the ultrasonic beam and will not reflect a signal response to the transducer, even if the probe is angled in such a way that a proportion of the sound energy impinges upon this dimension.

Hydrogen The presence of hydrogen cannot be treated in precise quantitative terms since there is no specific value that will result in chevron cracking. The problem of chevron cracking is usually associated with basic agglomerated submerged arc fluxes that, due to their mode of manufacture, are hygroscopic and will increase in moisture content unless stored and handled properly. This, therefore, presents one main area of investigation if chevron cracking is found, i.e., consumable handling practices. One difficulty is that, because you will be investigating a problem that occurred some days (or even weeks) previously, it may not be possible to assess the state of consumable handling at the time of welding with any confidence. Here, it is necessary to rely on your judgment of the normal practices involved. Are consumables routinely abused? Are they always rigorously applied? Are they adequate? These questions identify the extremes possible and can often provide useful guidance. Regardless of what handling practice was actually specified, the following points are worth considering: Is the chevron cracking an isolated occurrence or have many different welds been affected over a period of time? If the latter, then the chance of the problem being consumable related (and perhaps “batch” related) is greater. If the cracking is an isolated occurrence, then although the consumable may still be a causal factor, this will be difficult to establish; other factors may well be the cause. Is the flux used straight from the bag or is it preheated or baked? If preheated or baked, the procedure must be carried out carefully to proven and established practices. It is possible to achieve an even higher moisture content in a flux after heating if the baking

86

The Practical Welding Engineer

ovens are not well designed and properly used. Baking ovens should have heating elements passing through the charge, not just enveloping the charge. The time in the oven should also be carefully controlled to ensure all flux charged is brought to the temperature specified. Remember that flux is a good insulating medium and does not allow rapid heat flow. If flux is used from the bag and a problem occurs, check the as-supplied moisture level. It helps if you can relate the figures obtained to previous data since the “acceptable value” will vary from application to application. An ideal moisture value is zero, and for guidance, a O. 1 percent moisture level could be considered high and possibly problematic in some situations. Flux handling and flux condition are possibly the most likely potential sources of a chevron cracking problem, but do not assume it is the only cause. In a high-volume fabrication area using the same flux in a number of differing situations, it is difficult to blame the flux if only a limited proportion of the work is affected, Le., if the problem occurs only on one thickness or at one welding station. Although the flux could still be the main contributory factor, other aspects should be examined, such as preheat or local flux storage problems (e.g., roof leaks). Flux recycling can also result in a local (or, indeed, general) pattern. A number of commercially available units use compressed air for transporting the flux within a recycling system. Should this air become contaminated in any way, either by oil or water, then potential carry-over into the flux is likely. In terms of the effect of hydrogen, another significant aspect is preheat. Again, this is via the effect of preheat not only on weld microstructure (via the cooling rate), but more importantly on the rate at which hydrogen will diffuse from the weld. Production personnel will seldom admit to using low preheat, but rest assured it can be a common occurrence. Remember that in SAW preheat plays an important role with respect to weld metal cracking; in some cases weld metal cracking - not HAZ requirements - will dictate preheat levels. Experience shows that attempts to reduce preheat levels for SAW have resulted in chevron cracking. This doesn’t mean established levels cannot be reduced. It only indicates a need to take extreme caution.

Restraint As with the effect of hydrogen, the degree of restraint necessary for chevron cracking to occur has never been accurately established. With chevron cracking, most cracks are often located within the top one-third of the thickness. This would indicate that tensile loading is a very relevant feature; it also indicates that, in general, the greater the thickness, the greater the chance of chevron cracking. There is very little that the welding engineer can do to alter these particular parameters. However, the following example illustrates that, occasionally, beneficial changes to existing practices can be arranged.

Practical Problem Solving

f

87

. Weld location

u FIGURE 6.1 -ANNULAR STIFFENER WELD This particular example, illustrating the combined effect of restraint and preheat, relates to an occasion when radially oriented chevron cracking was experienced in joints between cylindrical cans and external annular plate stiffeners, as shown in Figure 6.1. In the initial manufacturing procedure, the cylindrical shell was preheated from the outside, but the stiffener plate was not directly heated. The resultant stress pattern was analyzed as follows: The can was modeled as an infinitely long, thin elastic cylinder with an axisymmetric radial line load acting outward on the cylinder at the weld interface. The external stiffener was modeled as an annular plate with a radial load acting on the inner radius. The thermal description of the setup is that after the two components had been welded together, the cylinder cooled by a stated amount relative to the disc. If the two components had not been welded, this would have resulted in a radial gap at the weld interface, calculable from the free contraction of the cylinder. The radial force required, therefore, was that which was sufficient to close the gap by outward deformation of the can and inward deformation of the plate. The circumferential stress in the weld was found from analysis of the disc. In essence, this indicated that when the can preheat was withdrawn, after welding or at some intermediate stage, the shell would tend to shrink in

88

Tße Practical Welding Engineer

diameter during cooling, but the external plate would remain at the original diameter. The resulting strain discontinuity across the weld would develop radial stress in the joint leading to a circumferential stress along the weld interface. The calculated radial and circumferential stresses were shown to be significant in relation to the yield strength of the material. From the analysis, it was obvious that any temperature differential at the beginning of the cooling phase from the preheat temperature would lead to significant stress. This would be alleviated if the external stiffener was preheated to match the can temperature. In addition, if the stiffener preheat was removed first, then compressive strain would be produced in the weld joint during cooling. A revised preheating method based on the above analysis was developed and no further cases of chevron cracking occurred on this particular item.

Summary To summarize (and oversimplify), regarding the problem of chevron cracking, the following actions are advised: Examine the ultrasonic examination report and agree with the NDE operator on the location of the representative and typical discontinuity signals. If in any doubt, take a sample for macroscopic examination check that you are indeed dealing with chevron cracks. Check the condition of the flux - has it been storeaandled properly? Take samples and analyze for moisture. Is the problem associated with one batch of flux or one storage site? If so, isolate the source immediately pending further investigation. If the flux is heated, ensure the thoroughness and correctness of the practice used. Check the preheat utilized. Was specified preheat adequate? Compare with past practices. Contact the consumable manufacturer. If the consumables appear correctly supplied and utilized, increase the preheat for subsequent work. Check that correct welding procedure and consumables were indeed used. If SMAW consumables were used, obtain samples of the same batch number and carry out hydrogen tests. Check consumable handling procedure. As a general rule: Reduce hydrogen reduce the problem.

-

Practical Problem Solvino

89

6.3 low Toughness in Self-Shielded Flux Cored Arc Welds This example relates to a specific problem faced by the offshore fabrication industry in the early 1980s. With increasing demands on low temperature toughness properties (typically 36 J at 4 0 ° C or F), it soon became apparent that the existing practice of using self-shielded flux cored (E61-T8-K6 type) I I wires in the uphill mode was extremely inconsistent when examining impact data at -4O"CF. The details of one major investigation into this problem are presented elsewhere [3]; for the purpose of this discussion, only the salient features will be highlighted. The macroscopic specimen shown in Figure 6.2 shows a satisfactory weld made with the consumable in the vertical position using an uphill technique with a weave. This technique is perfectly satisfactory in most situations and would probably be regarded as the preferred method of apply- FIGURE 6.2 - A SELF-SHIELDED FCAW UPHILL WELD ing such consumables. Unfortunately the low temperature toughness of this weld, as shown in Table 6.1, was not sufficient to satisfy the needs of the industry. Given the particular benefits of the process in terms of adaptability and productivity, other procedural methods were examined in effort to retain the technique's use. Two main possibilities existed given that the lower toughness could be related to the relatively high proportion of unrefined weld metal resulting from an uphill weave technique. Test Temp. (OC)

-40

TABLE 6.1

Heat Treatment Condition

As-welded

Position

cap-pass weld, centerline Mid-weld, centerline Root weld, centerline

Absorbed Energy (Jouies) 29, 14, 1 18: Ave 54 92,114, 18: Ave 75 8, 14, 12: Ave 11

- CHARPY V-NOTCH IMPACT TESTS ON AN UPHILL FCAW-S PLATE

90

The Practical Welding Engineer

Use stringer beads - downhill, no weave. Use a wide weave - uphill with controlled layer thickness. Both methods are successful in improving toughness. With production welding, however, only one of the above techniques could be easily controlled; this, in many ways, was the key to the problem. If the wide-weave method was introduced into a production situation, it would be very difficult to control given that a wide-weave bead can be produced with very thick layers using a block weave; this would be contrary to what is required (i.e., thin layers with a high degree of refinement). On the other hand, a stringer-bead technique is by its nature self-limiting and results in a weld metal microstructure providing more consistent and better impact properties. The need for the strict control resulting from a stringer bead technique was necessary in this case. It is worth noting that weld metal refinement is often a factor when investigating low toughness results with other processes and consumables.

6.4 Cast-to-Cast Variability The generally recognized use of the term “cast-to-cast variability” relates to the variable penetration behavior exhibited (Figure 6.3) by different casts of nominally identical materiais under a constant set of welding conditions - that is, the variability is due to the material. It is a problem most commonly associated with stainless steels, although not necessarily limited to them. Cast-to-cast variability is one of the few welding-related problems the welding engineer has little control over, especially at the point of occurrence. It is a problem that has been widely recognized during the last 20 to 30 years [4] and about which many wide and varied hypotheses have been promoted. It is not the intention of this book to discuss the merits of the various proposed causes other than the general cornent that in the authors’ experience, the only one to date that could be confirmed [5] at a practical level was that proposed by Heiple and Roper [6]. They related this variability to convection currents in the weld pool generated by the effect of temperature on surface tension. This effect was influenced by FIGURE 6.3 the presence of surface-active CAST-TO-CAST VARIABILITY

Practical Problem Solving

91

solutes, such as oxygen and sulfur, the levels of which were important. In the particular example referenced above [ 5 ] , it was found that casts of 316-type stainless steel with a sulfur content of 0.001-0.002 percent (Figure 6.3, left) behaved drastically different from one with a sulfur content of 0.010 percent (Figure 6.3, right). The transition in behavior between the two levels was predicted as being around the 0.005 percent level.

The following general observations are relevant: Cast-to-cast variability should be recognized as a potential problem, especially with stainless steels. The problem is usually more prevalent (or more likely to be recognized) in automatic welding applications and, in particular, automatic gas tungsten arc welding (GTAW) equipment. Other than by either cast-matching or by allowing procedure variations, little can be done in a practical sense to alleviate castto-cast problems. In other words, do not expect a cure - concentrate on managing the problem. It is important, however, to correctly diagnose cast-to-cast variability -often high on the list of abused excuses. If cast-to-cast problems are thought to be a problem, consider the following checks: Perform bead-on-pipe/plate tests using standardized welding conditions over a number of casts of material. If no difference in penetration or width-to-depth ratio is noted, then cast variability is not the problem. Repeat the above exercise on a single cast of material. If differences in penetration are noted, cast-to-cast variability is not the problem (or not the only problem). In considering the possibility that the problem under investigation may be cast-tocast variability, you must also consider other causes of variable penetration, such as variable power output, variable shielding gas supply, variable arc length, and inconsistent tungsten sharpening.

92

The Practical Welding Engineer

The above and other factors must be ruled out prior to diagnosing cast- to-cast variability, if for no other reason than that most of the other causes are curable. Having established that cast-to-cast variability exists, consider the following possible problem management methods: Match components to be welded within a cast or within a group of casts of similar behavior. Agree to a relaxation of procedure tolerances to allow some variation in parameters. Note that in extreme cases exhibiting a high width-to-depth ratio (i.e., very low penetration), increasing the current level can often have little effect on penetration depth. Gather as much evidence of the problem as practicable to present to your client. Only by doing so can you expect the cooperation that you need. If the methods noted in the first two items prove unsuccessful, consider more drastic changes to the weld procedure, such as shielding gas, pulsing, etc., although in severe cases the likelihood of success is limited. Consider the use of activated fluxes. These have been shown to influence the degree of penetration during gas tungsten arc welding. If all else fails, consider either re-sourcing the materials giving the problem, or, if feasible, carrying out a manual weld to replace an automatic weld. If either option is used, there will be obvious and significant commercial implications; these options should be used only as a last resort.

6.5 Magnetic Arc Blow In many cases, this problem may not even be brought to the welding engineer’s attention. Unless it is severe, the welder may persevere with the problem until, either by luck or by creating a bridge within the weld preparation, the phenomenon decreases to a more acceptable level. The phenomenon of magnetic arc blow occurs [ll, 121 when a welding process involving an electric arc (generally direct cur-

Icurrent

FIGURE 6.4

-ARC BLOW

Practical Problem Solving

93

rent) is carried out in the presence of a magnetic field and disruption or distortion of the arc results. The disruption can be a consequence of the magnetic field produced by the arc itself or by the interaction with any magnetism persisting in the steel. The magnetic field causes the arc to be deflected, as shown in Figure 6.4, and behave in an often unpredictable and erratic manner. In some very severe cases, the arc may be completely extinguished. Typical discontinuities resulting from such arc instabilities include slag inclusions, porosity, and incomplete fusion. Magnetism, of a level to cause arc blow when welding, can result from two separate sources that may be additive. The steel, as supplied, may possess its own residual magnetism; also, the welding current will induce a magnetic field surrounding the component during welding. Considering the former source - residual magnetism - this may have a number of possible origins. The steel has solidified in a magnetic field at the steel mill. The component has been lifted with a magnetic clamp. Magnetic particle inspection has been performed. The component has been stored near a magnetic field or left in a north/south direction sufficiently long for magnetism to build up from the earth’s magnetic field. The latter point may be particularly relevant to pipelines. The component has been exposed to a magnetic field during manufacture, e.g., from welding. The second source of magnetism leading to arc blow is where it arises during welding. Here, no magnetic field can be measured on an unwelded section, but during the welding process, the current causes the resultant magnetic field. This effect will increase with higher currents and can be influenced by the shape of the component and earthing arrangements. As a guide, there are few problems with low magnetic fields of 20 gauss or less. Between 20-40 gauss, arc instability can be observed, whereas fields greater that 40 gauss can create definite arc blow. These values assume a “facility to measure,” but this is often not available. A simple test uses iron filings that, if attracted, indicate a magnetic field and, therefore, a potential for trouble. A severe collection of filings is obviously an indication of a severe problem. The shape and depth of the weld preparation influences the magnetic effect on the arc. It will be more pronounced in deep and narrow preparations; root runs will also be more affected until “a bridging” affects, minimizes, or alters the magnetic effect. Having identified arc blow as the problem, the welding engineer has a number of options in order to eliminate, or at least reduce, the problem to acceptable limits. These are indicated below in order of severity, available resources, and expense.

The Practical Welding Engineer

94

1. 2.

3.

4.

5.

Use an AC power source in preference to DC. Alter the electrode angle or use a backstep welding sequence as shown in Figure 6.5. Reposition the current cable and earth return points or use two workpiece leads connected between the power source and different positions on the component. Demagnetize the component by using an alternating magnetic field generated by an alternating current (AC): Utilize the largest available AC power source, wrap the welding cable around the workpiece, apply a high current for a short time (say, 30 seconds), then reduce the current to zero. This should be repeated, increasing the number of wrapping turns and reducing current levels for each further sequence. Use a proprietary demagnetizing unit comprised of a gaussmeter feeding all relevant data to a power unit into which is connected a heavy-duty demagnetizing cable, arranged on the workpiece.

6.6 Elimination of Postweld Heat Treatment Postweld heat treatments (PWHT) to reduce residual stress levels have been a common practice in many industries for years. A suitable heat treatment operation can sometimes have additional benefits, such as a reduction in peak hardness values, an improvement in weld metal properties, or a lowering of A any adverse effect of a welding process on the mechanical properties of the heat-affected zone (HAZ). I However beneficial these effects may be, there are situaWeld tions where PWHT can, and Progression Direction’ should, be avoided wherever possible, especially where the practicality of performing the operation is virtually insurmountable. Note that the use of heat treatment avoidance techniques are dictated by an overall assessment of the situation; for example, should stress corrosion be an influFIGURE 6.5 - BACKSTEP WELDING TECHNIQUE encing factor, then a heat treat-

I

Practical Problem Solving

95

ment to provide stress relief and to reduce maximum hardness levels may be a compulsory requirement.

Normalizing of Electroslag Welds In most cases, electroslag welds are refined by a normalizing treatment (see Figure 6.6[a] and 6.5[b] illustrating welds in the as-welded and normalized conditions, respectively). This treatment enables the weld to be ultrasonically examined to a greater sensitivity; it also greatly improves the notch ductility of the weld metal. However, the welding engineer, when examining the potential techniques available for the welding of thick sections, should not dism i s s electroslag welding simply because of this normalizing treatment. Extremely fast, this technique is ideally suited for the welding of thick sections. Unless a problem occurs during the welding operation (e.g., wire feed problems or loss of slag pool), the weld integrity is virtually guaranteed. The welding engineer’s main question should be: What are the property requirements? If notch ductility is not an essential feature, then normalizing of the electroslag weld is probably not required. Such a principle has been adopted in the conAs-Welded Condition struction of winding barrels for the turbine indusîry. (A) Conversely, should the fabrication require enhanced notch ductility values (at, say,--lO to -2O”C), a double normalizing treatment can be examined. This generally causes further improvement of the impact properties of the weld. Buttering Buttering is “a surfacing variation in which one or more layers of weld metal are deposited on the groove face of one member that is to be welded to a dissimilar base metal. The buttering provides a suitable transition Normalized Condition weld deposit for subsequent completion of the butt joint.” [7] (BI The use of buttering to avoid postweld heat treatment in steel structures relates to the abil- FIGURE 6.6 ELECTROSLAGWELDS

-

96

Tite Practical Welding Engineer

ity to avoid high HAZ hardness in the subsequent butt joint. This is achieved via the use of a buttering consumable that will not transform in the final HAZ to produce martensitic constituents. The HAZ from the buttering operation must itself be controlled by PWHT, but this can be done piece meal, and ail final assembly welds can be carried out and left as welded. The technique can also be applied to situations where, for example, one side of a transition weld was in a material that could suffer deterioration during any subsequent PWHT. For example, certain stainless steels containing high femte levels could, under some circumstances, “sigmatize” with consequent serious impairment to properties. The use of previously stress-relieved, buttered weld preparations is perhaps the most common technique to avoid subsequent PWHT. A number of weld metal types may be used for the buttering material. One type frequently applied is the InconelB family of consumables. Here, the buttering layer is welded to the fabrication requiring PWHT at a thickness sufficient to contain a new Component Requiring weld preparation and subPostweld Heat sequent HAZ. The butTreatment tered section is nondestructively examined, heat treated, then prepared for the final butt joint welds, which are not subject to Weld Buildup Buttering PWHT (Figure 6.7 illustrates the sequence of events). Inspect, then Postweld Heat Treat The requirement to examine volumetrically the buttering before PWHT may itself pose problems. If this is done by radiography, considerable difficulty can be I experienced in the subseI quent interpretation of the films caused by the Xrays scattering at the edge of the preparation. One solution to this problem is to make a complete butt ( Final As-Welded joint weld between pairs Assembly Weld of components using the buttering consumable, FIGURE 6.7 -WORKING WITH BUTTERED SECTIONS nondestnictively examine,

kL

I

Practical Problem Solving

97

repair if necessary, and finally PWHT the subassembly. The butt joint weld can then be parted and a new preparation made for the final assembly weld. Figure 6.8 illustrates this sequence. Buttering with ferritic electrodes can also be a potential solution to overcome the problem of high HAZ hardness. In this instance, the component could be of a hardenable alloy or high-carbon steel; the buttering deposit would use low-hydrogen electrodes of a composition giving the desired levels of strength within the final butt joint after PWHT, but not producing a high hardness in the final HAZ. The buttered component would be subjected to PWHT, then the closing weld would be carried out using the same low-carbon ferritic electrodes. To overcome any potential discontinuity problems in the buttered layer prior to the PWHT, sufficient buttering must again be applied to allow, in this case, for an ultrasonic test.

Temper Bead The temper bead technique is I I a fairly well known method for Components Requiring Postweld Heat Treatment controlling HAZ properties. It can sometimes mean that PWHT, when conducted to achieve hardness criteria, can be avoided. Only a brief summary Set Up for Butt Joint Weld of the method is given here (see also Figure 3.2, page 29). It is essential that the HAZ L - 1 caused by the first weld bead is Make Butt Joint Weld Using Buttering Consumable tempered by the subsequent bead. Trials must be conducted Inspect by the welding engineer on his particular consumables and welding practices to ascertain the degree of bead overlap Postweld Heat Treat Butt Joint Welds required and the tolerances for this overlap. Generally, this tolerance is not great. So, the control Split Butt to Produce Buttered Components required during production welding to ensure correct overlap is such that the technique should only be used as a last resort, not on a large scale. It is Final As-Welded Assembly also important not to confuse this technique with a simple cap- FIGURE 6.8 - USING A BUTTERING CONSUMABLE

JJ-

JJ-

JJ-

I I

98

The Practical Welding Engineer

pass sequence control that, while offering scope for reduced HAZ hardnesses, cannot be relied upon unless strictly controlled and therefore definable as a true temper bead. The use noted above relates mainly to reducing HAZ hardness levels, but the principle can also be applied in butt joint welds to produce a maximum tempering effect throughout the weld [SI by combining controlled layer grinding and/or bead placement. This approach eliminates the need for PWHT and has particular relevance to inservice repairs where PWHT may be impractical.

Vibratory Stress Relief A technique to avoid PWHT - seldom occurring to the welding engineer - is vibratory stress relief (VSR), or vibrational conditioning [9, 101. VSR uses high force exciters to induce localized plastic flow into a component at room temperature. Although the nominal applied strains are elastic, local regions of high residual stress are elevated above the yield strength of the material, and the consequential plastic flow causes redistribution and concomitant reduction in these internal stresses. In practice, a vibrator is attached to the structure, energized, then scanned very slowly from zero through the entire range of the vibrator. The structure is monitored during the operation. When a resonant frequency is indicated, the vibration continues at that level for several thousand cycles. After each dwell period, the frequency is increased until some other natural resonant level is noted. Two or three repeats are generally required to achieve satisfactory stress redistribution. Take care, however, when considering the use of this technique. Where metallurgical improvements are necessary (e.g., relating to brittle fracture or stress corrosion cracking), then thermal treatments will most likely be necessary. Nevertheless, a general improvement in dimensional stability is usually observed, and, for this reason, the process is particularly applicable to large welded components or castings that require machining to very close tolerances. Fracture Toughness Justification The use of crack tip opening displacement (CTOD) data to justify the avoidance of stress relief is increasingly apparent in the offshore fabrication industry, relating to the production of jacket and deck structures for the oil industry. The basis of this permitted route to as-welded fabrication lies in the ability to demonstrate that the welded structure has sufficient fracture toughness in the “as-welded‘’ condition to accommodate the design loading and residual stresses locked into the structure during fabrication. It is normally required that such demonstration of good “as-welded’’ fracture toughness is demonstrated both for the weld metal and for the HAZ at the thickest section employed in the fabrication.

Practical Problem Solving

99

6.6 Fitness for Purpose The majority of modem fabrication codes, such as those from the American Welding Society (AWS) or American Society of Mechanical Engineers (ASME), specify nondestructive examination acceptance levels. These codes, defining acceptable weld flaw defect sizes, have been derived from a combination of good engineering and working practices, and experience. They are therefore arbitrary and do not relate to the joint’s fitness-for-purpose condition; in other words, these codes do not tell whether a specific location of a particularly sized and particular kind of defect affects the service integrity of a welded joint. For this reason, an alternative based on the “fitness-for-purpose” approach, which involves a fracture mechanics assessment, is now used. Essentially, the approach provides a basis for stating that a weld defect may be acceptable and, therefore, does not need to be removed, provided that conditions causing failure are not attained within the design and service life of the component. Fracture mechanics assessment, which has been incorporated into some recent standards (e.g., British Standard 7910 [replacing PD 64931, available from the American National Standards Institute), can now be calculated on computer. Commercially available packages, such as Crackwise from The Welding Institute, now can calculate critical flaw dimensions with varied geometries, such as surfacebreaking, embedded or through thickness. Fracture mechanics assessment is now accepted engineering practice and has been utilized in offshore construction, power generation, pipelines, pressure vessels, bridges, and in many other structures. It has been used to assess the significance of defects, define life extension and change-of-service applications, and determine if postweld heat treatment is required. (See section on fracture toughness justification, page 98.) The practical welding engineer should be aware of the potential behind fracture mechanics and be prepared to utilize it wherever possible - either proactively (for example, to encourage designers to relax existing requirements), or reactively (as a solution to an existing defective condition or problem). Removing defects from welds is expensive; postweld heat treatment is also expensive; therefore, time spent using a computer program -possibly eliminating PWHT and possible defects -is time well spent.

1O0

The Pmctical Welding Engineer

References [ i ] Wright, V. S . 1978. Chevron cracking in submerged arc welds. The Welding Institute Int. Con$ Trends in Steels and Consumables for Welding, Nov., 581-602. Paper 38. [2] Tuliani, S . S . 1976. A metallographic study of chevron cracks in submerged arc weld metals. Welding Research Znt. 6 (6): 1 9 4 6 . [3] Rodgers, K. J. and Lochhead J. C. 1987. Self-shielded flux cored arc welding -the route to good toughness. Welding Journal 66 (7):49-59. [4] Lucas, W. and Eardley, J. A. 1981. Effect of cast to cast material variations in TIG welding literature review. Welding Inst. Report 168. [5] Rodgers, K. J. 1983. A study of penetration variability using mechanized TIG welding. The Welding Institute Int. Conf. Effects of Residual Impurity and Microalloying Elements on Weldability and Weld Properties. Paper 2: 2-1 to 2-8. [6] Heiple, C. R. and Roper, J. R. 1982. Mechanism for minor element effect on GTA fusion zone geometry. Welding Journal 61 (4): 97-s to 102-s. [7] Metals Handbook, 9th Ed.: Vol. 6, Glossary 3. Materials Park, Ohio: ASM International, 1983. [SI Albeny, P. J. 1981. Simple test reveals level of two layer refinement. Welding and Metal Fabricator 49 (9): 543-547. [9] Parlane, A. J. A. 1977. Vibrational stress relief. The Welding Institute Research Bulletin. pp, 339-342. [lo] Claxton, R. A., and Saunders, G. G. 1976. Vibratory stress relief. Metallurgist and Mat. Technol. 8(12): 651-656. [ i 11 Blakely, P. 1988. Magnetic arc blow -causes, effects and cures. Metal Constr. 20(2):58-6 1. U21 Anon. 1990. What a blow. Welding Inst. Oct., p. 7.

Chapter 7

Common Defects and Remedial Actions It is both a boon and a bane to the practical welding engineer that a relatively large number of defect types can be observed in a welded structure. No matter how good the engineer’s procedure may be, there are too many variables, e g , equipment, consumables, and, most of all, “the A - Incomplete groove-face fusion; B - lamellar tearing; welder,” to conC - poor profile; D - slag inclusions; E - undercut struct the perfect fabrication. The FIGURE 7.1 -WELD DEFECTS most that the competent welding engineer can hope to achieve is to minimize the occurrence of such defects and, once discovered, rapidly diagnose and correct them with remedial actions. This chapter identifies the more common types of defects that can occur in a welded steel structure, how they may be recognized, and typical remedial actions. The list is not all inclusive, but it has been many years since any new phenomena have been discovered (e.g., lamellar tearing and chevron cracking). The problems that follow are not new but are generally the old favorites, albeit recycled under a new guise. Figure 7.1 is an exceptional illustration of five individual defect types, the nature of which are described below. It is an example of what poor technique, poor material choice, and poor control can produce. The faults illustrated in Figure 7.1 are identified as follows: A. Incomplete groove-face fusion (see section 7.4)

102

The Practical Welding Engineer

B. Lamellar tearhg (see section 7.1.2) C. Poor profile (see section 7.2.8) D. Slag inclusions (see section 7.3.5) E. Undercut (see section 7.2.11) The defects discussed are identified within four main headings: cracks, profile, volumetric, and incomplete fusion. To assist the welding engineer in dealing with these defects, they have been defined, their causes identified, and remedial actions proposed. Whenever possible macro/micro-photographs of physical samples, or sketches thereof, have been used to illustrate the individual defect.

7.1 Cracks Five types of cracking can be found in steel weldments. Hydrogen cracks - chevron. Heat-affected zone hydrogen cracks. Lamellar tearing. Reheat cracks. Solidification cracks (including crater cracks). The f i s t was discussed in the previous chapter. The next four are discussed below.

xl

7.1.1 Heat-Afîected Zone Hydrogen Cracks Characteristics Heat-affected zone (HAZ) hydrogen cracks (toe or underbead cracks) are discontinuities originating in a heat-affected zone due to high internal stresses combined with a susceptible microstructure and the presence of hydrogen. This is shown in Figure 7.1.1.

x 400

FIGURE 7.1.1

- HYDROGEN TOE CRACK

Identification of Common Defects and Remedial Actions

103

Causes Hydrogen is diffused into a hardened heat-affected zone.'The level at which cracking can occur is influenced by the following factors: Increasing section thickness Too high residual stresses. Increasing carbon equivalent. Too low a heat input. Insufficient preheat. Poor consumable handling (i.e., contaminated or not dried). Remedial Actions The foremost activity is to reduce the hydrogen level retained within the weld metal by doing the following: Use low-hydrogen consumables (properly handled). Remove contaminants. Increase hydrogen diffusion time by increasing preheat level or soak time, or both. Increase heat input. Postweld heat treat.

7.1.2 lamellar Tearing Characteristics These are discontinuities caused by the progressive cracking, under tensile loading, of inclusions within the base metal. The inclusions are approximately parallel to the plate surface and not generally associated with the heat-affected zone, although in some cases the defect may initiate from a HAZ toe crack. Figure 7.1 (page 101) illustrates this type of defect. Origins Presence of thin layers of nonmetallic inclusions parallel to the plate surface. Thermally induced strain causes through-thickness stresses that result in these inclusions linking up.

104

The Practico1 Welding Engineer

Remedial Action Use of steel grades (usually designated Z quality) with high throughthickness ductility (>35 percent reduction of area). Redesign of weld joint to reduce through-thickness strain. Longitudinal Section with Cracks Circled (x 1)

(a)

7.1.3 Reheat Cracking Characteristics This defect can be located after a postweld heat treatment of a lowalloy weldment, e.g., 2.25 percent Cr, 1 percent Mo. The cracking is transverse to the weld metal and of an intergranular nature with respect to the prior austenite grain structure, Intergranular Cracking (x 400) where these prior austenite grain (Etch SASPAINANSAINITAL) boundaries are extremely "decorat(b) ed" with segregates. Figure 7.1.3(a) FIGURE 7.1.3 - REHEAT CRACKING illustrates the general nature of the cracking, which is relatively small in size. Figure 7.1.3(b) shows the intergranular nature of the cracking and decoration of the prior austenite grain boundaries. Origins Cracking of this nature is generally associated with relatively high levels of residual elements in the weld metal. Such residuals, and levels likely to cause cracking, are phosphorus (0.025 percent), copper (0.25 percent), tin (0.30 percent), and arsenic (0.55 percent).

Remedial Action Removal of the defective weld metal and tighter control of the analysis of the welding consumables.

Identification of Common Defects and Remedial Actions

105

7.1.4 Solidification (Centerline) Cracks Characteristics These are cracks created when the weld material, while still hot, yields plastically due to high internal stresses. This usually occurs along the centerline. Figure 7.1.4(a) shows a typical example occumng on the fist side of a doublesided weld that has propagated from the unfused root face at the weld root up through the center plane of the columnar grains of the as-deposited bead. Figure 7.1.4(b) clearly illustrates that the cracks had originated at the weld root and propagated along the center plane of the as-deposited columnar bead. Figure 7.1.4(c) illustrates the inter-columnar nature of the crack. This was etched in the SASPANANSA (a saturated solution of picric acid with a wetting agent) etchant to reveal the solidification pattern rather than the transformation structure. (See Figure 3.3, page 32).

approx. x 6.5

x 60

(b)

(cl

FIGURE 7.1.4 - SOLIDIFICATION CRACK

106

The Practical Welding Engineer

Origins High bead depth to width ratios (>2: 1). High carbon, sulfur (hot shortness) and phosphorus contents. Decreasing Mn/S ratios. Contaminants. Inadequate filling of craters at the end of the weld runs (in this instant, the defect is more commonly known as crater cracks, although these can propagate to zones outside of the filled crater). For additional information on solidification cracking, see section 7.5 (page 122).

Remedial Actions Identify the pertinent cause, then apply the appropriate remedial action. For example, adjust parameters to obtain better depth-to-width ratios (aim for l:l), use lower residual element base materials andor consumables, clean joint faces, use backstepping technique to eliminate craters, or use slope-out device.

7.2 Profile Defects As the name indicates, these are imperfections in which the weldment has failed to meet some predetermined physical dimension or surface acceptability. Eleven such imperfections can be identified. 1. Arc strikes (7.2.1). 2. Excess distortion (7.2.2). 3. Excess reinforcement (7.2.3). 4. Incomplete root penetration (7.2.4). 5. Misalignment (7.2.5). 6. Overlap (7.2.6). 7. Overpenetration (7.2.7). 8. Poor profile (7.2.8). 9. Root concavity (7.2.9). 10. Spatter (7.2.10). 11. Undercut (7.2.11).

Identification of Common Defects and Remedial Actions

107

7.2.1 Arc Strikes Characteristics Arc strikes are caused by arcing between the electrode, the electrode holder, or the workpiece lead clamp, and the workpiece. The results are areas of fused metal with associated heat-affected zones that may or may not contain cracks. Note that if the arc strike has been caused by a copper contact tip or cable arcing onto steel, then serious contamination defects can result (see section 7.3.6, page 119). Origins Poor or missing insulation on electrode holder or torch. Loose current return (workpiece lead) clamp. Poor access to work. Careless practice.

Remedial Actions Correct insulation. Maintain current return clamping system. Improve access. Train welders.

7.2.2 Excess Distortion Characteristics Excess distortion is an imperfection whose occurrence can seldom be predicted unless the work is of a repetitive nature. It occurs when welded members are physically outside acceptable predetermined dimensions relative to one another, usually at the end of the welding operation. It is also an imperfection that can be recognized to be occurring during the welding operation; steps can be taken at this intermediate stage to minimize or eliminate this problem. With respect to this imperfection, recommendationsfollow on how to minimize or rectify the situation: Methods to Minimize Distortion Restrain the welded joint either by jigging or the use of strongbacks. Preset the plates to counter movement.

108

The Practical Welding Engineer

Use balanced welding (equalize weld runs on either side of the weld joint). Use special welding sequences - Backstepping. - Intermittent backstepping. Minimize (where possible) number of weld runs. Keep weld sizes to a minimum - do not over-weld. Ensure good fitup; avoid large root openings and misalignment.

Rectification of Distortion Use of force -Force usually is permitted (without the simultaneous application of heat) provided that a preset level of strain is not exceeded. Use of heat treatment -If used at an intermediate fabrication stage, this reduces the peak stress levels (i.e., produces more even disbibution of stress) and hence reduces distortion caused by these stresses. Use of heat line bending - This is permitted in some cases, but often only where the method has been qualified by destructive testing. The workpiece temperature is usually controlled to a preset maximum using temperature-indicating crayons or infrared pyrometers. Particular attention needs to be taken with steels, such as quench and tempered steels, where the temperature reached may have a significant effect on properties.

7.2.3 Excess Reinforcement Characteristics Most welding codes, whether of national or client origin, will specify a maximum cap-pass height for a weldment, e.g., 3 mm. This allows both an aesthetically pleas-

1 FIGURE 7.2.3 - EXCESS REINFORCEMENT

EXCESS

Identification of Common Defects and Remedial Actions

1O9

ing and a better designed transition between the base metals being welded (a sudden joining and stress-raising “hump” is avoided). In addition, this restriction assists in producing more economic weldments by eliminating additional man-hours adding unnecessary weld metal. The difference between an acceptable cap-pass height and excess reinforcement is sketched in Figure 7.2.3.

7.2.4 Incomplete Root Penetration Characteristics Incomplete penetration of weld metal into the root of a joint. Figure 7.2.4 illustrates its features. Origins Current too lowhigh. Irregular wire feed. Preparation too narrow. Too large electrode for joint preparation. FIGURE 7.2.4 Root face too thick. INCOMPLETE ROOT PENETRATION Root opening too small. Wrong polarity. Mismatched joint. Incorrect electrode angle for joint configuration. Arc length too high. Poor techniques. Stickout too long. Insufficient cleaning on second side.

Remedial Action The necessary corrective action(s) follows the identification of the cause, e.g., correction of parameters, re-preparation of joint configuration, etc.

7.2.5 Misalignment Characteristics This profile imperfection occurs when the abutting members of a weld joint relative to one another are outside a specified maximum permissible level. Figure 7.2.5 illustrates this imperfection.

1 1o

The Practical Welding Engineer

FIGURE 7.2.5

- MISALIGNMENT

Origins Poor fitup/assembly. Out-of-roundness/flatness of the base metal.

Remedial Action Prevent by using better preparation, fabrication, and assembly techniques. Pre-survey of base metal. Rectify via localized dressinglmachiningif permissible.

7.2.6 Overlap Characteristics Overlap is basically a lack of surface fusion at the toe, or root, of the weld. It is caused by the weld metal fíowing onto the base material surface without fusing to it. Figure 7.2.6 illustrates this imperfection. Origins Poor manipulative technique. Excessive weaving. Too low arc energy. Too low travel speed. Incorrect positioning of workpiece. Remedial Actions Identification of the pertinent cause and subsequent correction.

FIGURE 7.2.6

-OVERLAP

Identification of Common Defects and Remedial Actions

111

7.2.7 Overpenetration Characteristics The penetration of the root bead is excessive, resulting in unacceptable protrusion. Figure 7.2.7 illustrates this feature.

FIGURE 7.2.7 - OVERPENETRATION Origins Excessive root opening. Welding current too high. Travel speed too low. Poor welder technique. Root nose too thin.

Remedial Actions Use of correct parameters. Improvement of welding technique. Improvement of joint configuration. Can sometimes be reduced by changing the welding position.

7.2.8 Poor Profile In most instances, poor profile has a bumpy or ragged appearance; the weld surface does not flow in an aesthetically pleasing manner. This feature is evident in Figure 7.1.

Origins Poor technique.

1 12

The Practical Welding Engineer

Poor positional behavior of electrode. Incorrect parameters, particularly current/voltage/speed.

Remedial Actions Improve welder training. Consumable with better positional characteristics. Optimize parameters.

7.2.9 Root Concavity Also known as underwashing, this imperfection occurs as a shallow groove at the root of a weld in a butt joint. Figure 7.2.9 illustrates this feature.

Origins Excess back-purging gas pressure. Effect of gravity on a ?wide? root bead. Can be influenced by weld preparation, especially with automatic gas tungsten arc welds.

Remedial Actions Restriction of back-purging gas pressure. Optimize parameters to improve weld root shape. Change from V to J preparation (automatic gas tungsten arc welds).

7.2.10 Spatier Characteristics Spatter is defined as small droplets of weld metal thrown clear of the weld pool that may or may not be fused to the adjacent base metal. While not a ?defect? in the sense of affecting weld integrity, spatter produces a poor appearance and increases subse-

FIGURE 7.2.9

-ROOT CONCAVITY

Identification of Common Defects and Remedial Actions

1 13

quent cleaning costs - particularly on items requiring finishing treatments such as painting, plating etc.

Origins Arc length too long. Current range too high. Incorrect polarity. Magnetic arc blow. Contaminated, damp, or poor-operability consumables. Incorrect electrode angle. Poor gas shielding.

Remedial Actions Use of correct parameters. Elimination of arc blow. Improvement of technique. Use better consumables and/or improve consumable storage practice.

7.2.1 1 Undercut Characteristics This defect is an irregular groove usually found at the weld toe in the base metal or in previously deposited weld metal. An example of the latter is illustrated in Figure 7.1, and typical forms of this defect are sketched in Figure 7.2.11.

Origins Excessive weaving. Too high current, travel speed, or electrode size. Incorrect electrode angle. Incorrect shielding gas.

Remedial Actions Identification of pertinent cause, then corrective action, e.g., correct parameters, better manipulative techniques, etc.

I 14

The Ploctical Welding Engineer

FIGURE 7.2.1 1

- UNDERCUT

7.3 Volumetric Characteristics Seven types of defects have been identified within this category. These imperfections are usually not surface breaking. With respect to copper inclusionskontamination (7.3.6), this can also be manifested as a crack defect, but it is included here as this is not always the case. 1. Crater pipes (7.3.1). 2. Restart porosity (7.3.2). 3. Uniformísurface porosity (7.3.3). 4, Elongated Porosity (7.3.4). 5. Slag inclusions (linear and isolated) (7.3.5). 6. Copper inclusionskontamination (7.3.6). 7. Tungsten inclusions (7.3.7).

7.3.1 Crater Pipes Characteristics These are depressions due to shrinkage at the end of a weld run where the heat source was removed. Figure 7.3.1. illustrates this feature. Origins Combination of interrupted deoxidation reactions and the liquid-tosolid volume change. Often associated with porosity.

Identification of Common Defects and Remedial Actions

1 15

FIGURE 7.3.1 - CRATER PIPES Remedial Actions Improve weld termination techniques. Use slope-idslope-out current decay devices. Use runoff blocks.

7.3.2 Restart Porosity Characteristics This is a very localized defect normally occurring in manual or automatic arc welding at the start of the weld run. It happens as a result of delay in the establishment of suitable fluxing and shielding reactions at the start of the weld run due to nonequilibnum temperature conditions. Figure 7.3.2 illustrates this feature as it would appear on a radiograph. Origins Ineffective filling of weld craters. Poor technique.

FIGURE 7.3.2 - RESTART POROSITY Remedial Actions Use run-odrunoff blocks to contain the defect. Pay greater attention to starthestart manipulative techniques.

1 16

The Practical Welding Engineer

7.3.3UnifornVSurface Porosity Characteristics These are voids, or pores, distributed fairly uniformly throughout a weld run. They are generally equiaxed and result from gases formed during reactions in the weld pool being trapped as the weld metal solidifies. In surface porosity, these pores break the weld metal surface. The extent of porosity is generally defined by the number of pores noted per 10 cm of a radiograph (weld only). Extensive > 100 t Scattered < 100 t to > 25 t Sparse < 25 t to 3 t Very sparse < 3 t where t is the weld thickness in cm.

Figure 7.3.3 illustrates the pattern of extensive uniform porosity as it would appear on a radiograph.

Origins Porosity can generally result from one or more of the following. The more obvious causes marked (*) should be eliminated first. Damp flux or electrode coating. * Contaminated surfaces. * Welding current too lowhigh. * Insufficient flux/gas coverage. * Drafty conditions. * Damaged electrode coating. Loss of, or contaminated, gas shielding. Gas flow too high. Water leakage (in water-cooled unit). Contaminated consumables.

Identification of Common Defects and Remedial Actions

7 17

Arc too long. Incorrect weaving technique. Incorrect or insufficient deoxidant in consumable or base metal. Excessive travel speed. Excessive sulfur in consumable or base metal (surface porosity).

Remedial Actions The necessary corrective action(s) follows the identification of the cause, e.g., degreasing, correcting of gas shield, replacement or drying of consumables, improvement of technique, etc.

7.3.4 Elongated Porosity Characteristics Elongated porosity (or tunneling) defects are elongated, or tubular, voids with circular cross sections typically running along the axis of the weld. They are formed by gas entrapment that occurs as the weld metal solidifies; they may appear as a single entity or in groups. Figure 7.3.4 below illustrates the cross section of such elongated porosity (length around, 7 cm) in a narrow-groove gas tungsten arc weldment. Elongated porosity can sometimes appear in an extensive chevron pattern. Origins Incorrect welding variables, particularly travel speed. Surface contaminants. Joint geometry (e.g., opening between the vertical member of a T-joint welded from both sides).

Remedial Actions Ensure pre-weld cleanliness. Eliminate susceptible joint configurations. Correct travel speed. FIGURE 7.3.4

- ELONGATED POROSITY

1 18

The Practical Welding Engineer

7.3.5 Slag Inclusions (linear and Isolated) Characteristics These are slag particles, or other foreign matter, trapped during welding. The irregular nature of the defect differentiates it from a gas pore when examined radiographically. Slag inclusions can often be associated with, and have related causes, to incomplete fusion defects. A linear slag inclusion is generally considered a more serious defect than isolated inclusions, which are often disregarded within acceptance criteria unless multiple. Figure 7.1 (page 101) illustrates typical slag defects.

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

Origins Slag inclusions generally result mainly from one or more of the following causes. Those marked (*) should be eliminated first. Poor manipulative technique causing loss of slag control (shielded metal arc welding, self-shielded flux cored arc welding).* Inadequate cleaning between runs (shielded metal arc welding, submerged arc welding, self-shielded flux cored arc welding, gas shielded flux cored arc welding).* Electrode too large (shielded metal arc welding).* Presence of mill scale andor rust.* Slag flooding in front of arc caused by work position (shielded metal arc welding, submerged arc welding, self-shielded flux cored arc welding). Travei speed too low (shielded metal arc welding). Arc too long (shielded metal arc welding). Variation in welding speed (shielded metal arc welding). Welding over irregular profile (shielded metal arc welding, submerged arc welding). Voltage too low (submerged arc welding). Poor bead positioning (shielded metal arc welding, submerged arc welding). Poor joint configuration (shielded metal arc welding, submerged arc welding). Remedial Actions The necessary corrective action is dependent on the identification of the cause, e.g., better inter-run cleaning, better manipulative technique or positioning, etc.

Identiíication of Common Defects and Remedial Actions

1 19

7.3.6 Copper Inclusions/Contamination Characteristics This particular defect occurs when copper has been accidentally introduced to the weld pool. The resultant effect is weld metal cracking or penetration of the copper into the grain boundaries of the steel. Figure 7.3.6 illustrates copper contamination due to the brazing of anode attachments directly onto a C-Mn pipe. Origins touching9and FIGURE 7.3.6 - COPPER CONTAMINATION (X 5) especially arcing, of the contact tip on the weld preparation groove faces. Loss of, or the melting of, copper contact tips. Transfer by abrasion andor arcing from clamps, cable etc.

Remedial Actions Avoid contactlarcing. Complete removal of contaminated area. This must be excavated to a depth ensuring removal of any zone affected by grain boundary penetration. In severe cases, metallurgical examination by local etching may be required to establish this.

7.3.7 Tungsten Inclusions Characteristics Tungsten inclusions occur when tungsten has been accidentally introduced into the weld pool. As this can only result from the use of the gas tungsten arc welding process - inherently free from slag inclusions - assume that any inclusions found are tungsten residue. In addition, tungsten is a denser material than steel and most other commonly welded materials and thus shows as light areas on radiographs, often having an angular shape.

120

The Practical Welding Engineer

Origins Poor technique allowing electrode to touch. Incorrect polarity. Disintegration of electrode during welding. Using thoriated electrode for AC. Current too high for electrode diameter.

Remedial Action Avoid accidental contact between tungsten electrode and pool. Use correct polarity, grade, and size of electrode to suit application. Tungsten inclusions are often disregarded if small and are treated similarly to porosity in terms of defect acceptance.

7.4 Incomplete Fusion Characteristics This type of defect, although planar in nature, has been separated from the “cracking” category as this defect generally relates to procedural or technique problems, Le., it is not of a metallurgical nature. Incomplete fusion occurs in these typical locations: between adjacent runs in a multipass weld (incomplete inter-run fusion); between the weld and base metal and at either (or both) side(s) of the joint configuration (incomplete groove-face fusion); and at the root of the weld configuration (incomplete weld root fusion). Figure 7.4 illustrates these defects.

Origins Poor manipulative techniques. Contamination of weld surface. Arc length too short. Travel speed too fast. Current too low. Incorrect electrode angle. Incorrect inductance setting. Incorrect work position resulting in molten metal flooding ahead of arc. Incorrect weld preparation. In addition to the above, the following are particular to the root condition:

/dentifieation of Common Defects and Remedial Actions

Incomplete Inter-Run Fusion

Incomplete Groove-Face Fusion (See Figure 7.1)

121

Incomplete Weld Root Fusion

FIGURE 7.4 - INCOMPLETE FUSION Too large an electrode diameter. Excessive root face. Undersized root opening.

Remedial Actions Identification of cause(s) and application of specific corrective action(s) improvement of parameters, joint setup, etc.

122

The Prucficul Welding Engineer

7.5 Some Additional Information on Solidification Cracking The tendency for weld metal solidification cracking is critically dependent on the weld metal composition. Several probability formulae have been included in the literature. For example,

c (s+P + S A 5 + HCS =

Ng0)

x 1000

3 M n + C r + 2 (Mo+V)

UCS = 230C* + 190s + 75P + 45Nb - 12.3Si - 5.4Mn - 1 [2] (* C values below 0.08 percent should be taken as equal to 0.08 percent) where HCS (hot cracking susceptibility) and UCS (units of crack susceptibility) are indices of crack susceptibility. The effect of C levels and M d S ratios is illustrated graphically in Figure 7.5 [3].

0.10

0.12

0.14

0.16

Carbon Content

FIGURE 7.5 - EFFECT OF C LEVEL AND Mn/S RATIO

Identification of Common Defects and Remedial Actions

123

References [il Bonomo, E 1972. Sulfur-induced solidification cracking in low alloy steel weld metal deposited from basic low hydrogen electrodes. Welding Research International 2(4): 1-28. [2] Bailey, N., and Jones, S.B. 1978. Solidification cracking of ferritic steels during submerged arc welding. Welding Journal 57(8) 217-s to 231-s. [31 Lancaster, J. E 1992. Handbook of Structural Welding, Cambridge, U.K.: Abington, p. 207.

Chapter 8

Oxyfuel Cutting, Arc Air, and Electrode Gouging The practical welding engineer should have at least a working knowledge of three other processes: oxyfuel (oxygen fuel gas) cutting, arc air gouging/cutting, and electrode gougingkutting. They are obviously the opposite of welding or joining processes. However, they are crucial to the welding industiy, often going hand-in-hand with a welding process. Despite their close association with welding, they are processes that the welding engineer tends to learn by accident. The following is intended to assist the welding engineer to gain a working knowledge of the processes, and to be able to offer practical advice regarding their application.

8.1 Oxyfuel Cutting This process, popularly known as burning or flame cutting, is widely used to cut straight lines and shapes, and to produce a variety of edge profiles on plates, pipes, and sections. Figure 8.1 illustrates the components of two typical torches. The process operates via the removal of metal by a chemical reaction between oxygen and hot material. The preheat flame is used to raise the surface temperature of the metal to the temperature at which this chemical reaction can take place. The heat from the resultant reaction melts the material, which is blown from the cut by the oxygen jet. By moving the torch across the workpiece, a continuous cutting action can be achieved. The cutting responses of the process is very dependent on the material being sectioned. It is very good in mild, low-carbon, and low-alloy steels. If used to cut stainless steels, it is necessary to add a flux, or metal powder, to the cutting oxygen stream; even then, a relatively poor quality cut is achieved. The process is not at all suitable for any nonferrous material. Two terms used in the process, kerf and drug, are illustrated in Figure 8.2; their relative significance in the process is outlined below.

8.1.1 Kerf Kerf, defined as the width of the cut, is a function of the oxygen jet dimensions, type of tip, speed of cutting, and the flow rates of both the cutting oxygen and preheat gases. Because of these factors, the width of the kerf increases as the material thickness being cut increases.

126

The Practical Welding Engineer

CUTTING nD

TIP NUT,

~ E E - T U B E DESIGN

/CUTTING OXIGEN

FIGURE 8.1 - TYPICAL PREMIXING-TYPE CUTTING TORCH (LEFT), AND TYPICAL TIP MIX CUTTING TORCH (RIGHT). From Welding Handbook, Vol. 2 , American Welding Society. Miami, Fla.

Oxyfuel Cutting, Arc Air and Electrode Gouging 127

8.1.2 Drag The drug is the distance of the lag between the most distant part of the cutting stream and the position nearest to the torch tip. Zero drag occurs when the oxygen stream enters and exits the kerf along the axis of the tip. As the drag width increases, or moves into reverse, there is generally a loss in quality.

FIGURE 8.2 - KERF AND DRAG

8.1.3 Problem Solving

From Welding Handbook, Vol. 2, American Welding Society, Miami, Fia.

The following sketches, together with the relevant conditions and descriptions, are intended to assist the welding engineer in assessing the quality of any cuts produced by the oxyfuel process when used in a machine mode. Recommendations to rectify the various faults follow.

Correct Cutting In a correct cut (Figure 8.3), the top of the cut is sharp and clean, and the drag lines are almost invisible, producing a smooth side. Oxide is easily removed, the cut is square, and the bottom edge is clean and sharply defined. Drag lines should be vertical for profiles. A small amount of drag is allowed on straight cuts. Cutting Speed Too Slow Because of melting, the top edge has become rounded (Figure 8.4). Gouging is pronounced at the bottom edge, which is also rough. Scale on the cut face is difficult to remove. To rectify, traverse at recommended speed; increase the oxygen pressure.

FIGURE 8.3

-CORRECT CUTTING

FIGURE 8.4 - CUTTING SPEED TOO SLOW

Top and bottom

FIGURE 8.5 - CUTTING SPEED TOO FAST Figures 8.3-8.5 from Module Manual FIO of the General Welding and Cutting for Engineering Craffsmen. Training Publications, Ltd., Watford. England. Reprinted with permission.

128

The Practical Welding Engineer

Cutting Speed Too Fast The top edge may not be sharp; there is a possibility of beading (Figure 8.5). To rectify, slow down the traverse to the recommended speed; leave the oxygen pressure as set.

above Work Excessive rounding and melting of the top edge (Figure 8.6). Undercut has been caused by the oxygen stream opening out. To rectify, adjust the standoff distance between the nozzle and the plate. Preheat Flames Too Close to Work Heavily beaded and rounded top edge, otherwise of good appearance (Figure 8.7). To rectify, correct the standoff distance by raising the nozzle to the recommended height.

FIGURE 8.6 - PREHEAT FLAMES TOO HIGH

?

I

FIGURE 8.7 - PREHEAT FLAMES TOO CLOSE Rounded edge,

slag

FIGURE 8.8 - PREHEAT FLAMES TOO LARGE

Preheat Flames Too Large Due to excessive heat, the preheat flame has caused the top edge to melt and become FIGURE 8.9 CUTTING OXYGEN rounded (Figure 8.8). The kerf PRESSURE TOO HIGH tapers from just below the top Figures 8.6-8.9 from Module Manual FIO of the General Welding and edge to the bottom of the cut Cutting for Engineering Craftsmen. Training Publications, Ltd., Watlord. England. Reprinted with permission. face. To rectify, set a preheat flame as recommended; use the correct nozzle at the recommended gas pressures.

-

Cutting Oxygen Pressure Too High The edge has a regular bead. The kerf is wider at the top with undercutting just beneath it (Figure 8.9). To rectify, set the oxygen at the recommended pressure. Note

Oxyfuel Cutting, Arc Air and Electrode Gouging

129

that on thinner steel, high oxygen pressure can cause a taper cut likely to give the impression that the cutting machine is set incorrectly in relation to the plate.

process is dependent upon a number of variables, including the following.

Correct Gouging In a correct gouge (Figure 8.11), the groove is of uniform width and depth; free of oxide and scale, both in the groove and surrounding plate; and

\ I /

130

The Practical Welding Engineer

Gouging Nozzle Held at Incorrect Angle A groove of varying width and depth caused by holding the gouging nozzle at the wrong angle (Figure 8.12): Too steep an angle increases depth and removes too much metal. Too shallow an angle gives a superficial gouge. Incorrect Nozzle Alignment A shallow groove with heavy oxide deposits (Figure 8.13) is caused by failing to present the gouging nozzle axially in line with the direction of gouging, due to working too quickly, using incorrect gas pressures, and using incorrect nozzle.

8.3 Electrode Gouging/Culting Special-purpose manual metal arc electrodes are available for the arc grooving, cutting, and piercing of ferrous and nonferrous metals. Metal-arc cutting with such electrodes occurs as a result of melting and removal of metal along a desired line of travel using an electric arc struck between the workpiece and a special covered electrode. The electrode coating is specifically designed to concentrate a forceful and penetrating arc; stabilize the arc and prevent its extinction; and blow the molten metal and dross away with a positive jet of gases. The above criteria are carefully balanced to enable the operator to maintain a high degree of control. The physical properties of the coating ensure it decomposes more slowly than the melting rate of the core wire, which results in the formation of a I deep cup 3-5 mm deep at the FIGURE 8.12 GOUGING NOZZLE HELD tip of the electrode. This AT WRONG ANGLE ensures the Operation Of the From Module Manual FtO of the General Welding end Cutting for Engineering Craftsmen, Training Publications. Ltd., Watford. England. arc within that space without Reprinted with permission.

'

-

Oxyfuel Cutting, Arc Air and Electrode Gouging 13 1

causing any short circuits, even when the electrode is inserted into holes during piercing, or in tight openings and grooves. The insulating properties of the coating prevent side arcing. For situations with difficult access, a cutting and gouging flux-covered electrode has distinct advantages over the carbon-arc and oxygen-cutting processes because it can be manipulated in very confined spaces.

Cutting Metal sheet and plate up to I 10 mm thick can be metal-arc FIGURE 8.13 - INCORRECT NOZZLE ALIGNMENT cut with ease* The From Module Manual F10 of the General Weldmg and Cutting for should be held at a shallow Engineering Craftsmen. Training Publications, Ltd , Watford, England Reprinted with permission angle of about 15 degrees to the surface of the plate, as shown in Figure 8.14. With thicker plates, an up-and-down motion should be made in the direction of thickness so that the molten metal and slag may run clear of the cut, as shown in Figure 8.15.

FIGURE 8.14 -CUTTING

132

The Practical Welding Engineer

Direction of Travel

-

FIGURE 8.15 CUTTING THICKER PLATES GroovingIGouging The force of the arc removes the molten metal by pushing it out in front of the groove while a forward-and-backwardmotion is applied, as shown in Figure 8.16. Where possible, the workpiece should be inclined so that the molten metal can run free under the force of gravity. Such electrodes can be used to gouge out any faulty weld metal deposit without the need for special cutting and grinding tools. Metal-arc cutting is widely used for cutting holes into piping for subsequent welding of branches and connections. The process is particularly effective in cutting, gouging, and piercing metals and alloys that are difficult to machine, such as armor steel, air and deep hardening steels, stainless steels, cast irons and hard, or work hardening, deposits.

f -\

Direction of Travel

FIGURE 8.16

’ -GOUGING

Recommended Reading Journals Welding Journal, Miami, Ha.: American Welding Society. Welding and Metal Fabrication, Redhill, U.K.: Argus Business Media, Ltd. Welding in the World, International Institute of Welding, New York, N.Y.: Pergamon Press.

Books Anderson, T. L. Fracture Mechanics: Fundementals & Applications. 2nd. Ed. Boca Raton, Na.: CRC Press, Inc., 1995.

ASM Handbook, Vol. 6, Materials Park, Ohio: ASM International, 1993. Boniszewski, T., Self Shielded Arc, cambridge, U.K.: Abington Publishing, 1992. Castro, R., and deladenet, J. J., Welding Metallurgy of Stainless and Heat-Resisting Steels, New York, N.Y.: Cambridge University Press, 1975. (Out of Print) Colangelo, V.J. and Heiser EA., Analysis of Metallurgical Failures, 2nd. Ed. New York, N.Y.: John Wiley and Sons, 1987. Davies, A.C., The Science and Practice of Welding, 10th Ed. 2 vols. New York, N.Y.: Cambridge University Press, 1993. Gray, T. G. E , and Spence, J, Rational Welding Design, London, U.K.: Butterworths 1982. Kou, &Welding Metallurgy, New York, N.Y.: John Wiley and Sons, 1987. Lancaster, J. E, Handbook of Structural Welding, Cambridge, U.K.: Abington Publishing, 1992.

Handbook of Case Histories in Failure Analysis, 2 vols. Materials Park, Ohio: ASM International, 1992-93. Pickering, E B., Physical Metallurgy and the Design of Steels, 4th Ed. New York, N.Y.: Applied Science Publishers, 1996.

The Procedure Handbook of Arc Welding, 13th Ed., Cleveland, Ohio: The Lincoln Electric Company, 1995.

Welding Handbook, Vol. 2: Welding Processes, 8th Ed. Miami, Fla.: The American Welding Society, 1991.

Welding Handbook, Vol. 5: Engineering Costs, Quality, and Safety, 7th Ed., Miami, Fia.: American Welding Society, 1984.

Welding Handbook, Vol. 1: Welding Technology, 8th Ed. Miami, Fla.: American Welding Society, 1987.

Appendix II

Useful Tables, Formulas, and Diagrams A. Useful Tables

Page

1. Hardness Equivalent 2. Stress Conversion 3. Temperature Conversion

136 137 148

B. Formulas l a & b. Carbon Equivalents 2. Electrode Basicity 3. Electrode Consumption 4. Heat Input 5. Thickness vs. Yield Stress

140 140 140 141 141

C. Diagrams 1. Iron Carbon 2. Nelson Curves 3. Schaeffler and DeLong Diagrams 4. WRC-1992 Diagram 5. Electrode Classification

142 143 144 145 146

736

The Placiical Welding Engineer

A. Useful Tables 1. Hardness Equivalent ~~

Brinell Brinell ,Hardness Dia Impression Number HB mrn

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

2.50 2.55 2.60 2.65 2.70 2.75 2.80 2.85 2.90 2.95 3.00 3.05 3.10 3.15 3.20 3.25 3.30 3.35 3.40 3.45 3.50 3.55 3.60 3.65 3.70 3.75 3.80 3.85 3.90 3.95 4.00 4.05 4.10 4.15 4.20 4.30 4.40 4.50 4.60 4.70 4.80 4.90 5.00 5.10 5.20 5.30 5.40 5.50 5.60 5.70 5.80

(601) (578) (555) (534) (51 4) (495) (477) );!;( 429 415 40 1 388 375 363 352 34 1 33 1 32 1 31 1 302 293 285 277 269 262 255 248 241 235 229 223 217 212 207 197 187 179 170 163 156 149 143 137 131 126 121 116 111 107 103

Vickers Hardness Number HV

Rockwell C Scale Hardness Number

640 615 591 569 547 528 508 491 474 455 440 425 41 O 396 383 372 360 350 339 328 319 309 301 292 284 276 269 261 253 247 241 235 228 223 218 208 197 189 179 172 165 157 150 144 138 133 127 122 117 113 108

57 56 54.5 53.5 52 51 49.5 48.5 47 45.5 44.5 43 42 40.5 39 38 36.5 35.5 34.5 33 32 31 30 29 27.5 26.5 25.5 24 23 22 20.5

-

-

-

-

-

-

-

-

Equiv. R

Equiv. R

Equiv. Rm

Ton%?

k g f k n f

Wmm2

-

-

-

-

1o1 98 95 92 88 85 82 80 77 75 73 71 68 66 64 63 61 59 58 56 55 53 51 50 49 48 46 45 43 41 39 36 35 34 32 31 31 30 29 28 27 26 25 24

-

160 155 150 145 139 134 129 126 121 118 114 111 107 104 101 99 96 93 91 89 87 84 81 79 77 76 73 71 68 65 62 57 55 54 51 49 49 47 46 44 43 41 39 38

-

-

-

1569 1520 1471 1422 1363 1314 1265 1236 1187 1157 1118 1089 1049 1020 990 97 1 94 1 912 892 873 853 824 794 775 755 745 716 696 667 637 608 559 539 530 500 481 481 461 45 1 431 422 402 382 373

The figures in parenthesis require a "modified" Brinell test, ¡.e., a tungsten carbide ball is required where the Brinell hardness value exceeds 450, HB to HV and HV to HRC conversions are based on E.140, by the American Society for Testing and Materials.

Appendix II - Useful Tables and Diagrams

2. Stress Conversion tonvln.'

kgVmnP

Wmnf

IMn.'

tonMn.'

kgVmnf

Illrmm'

IMn:

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

1.6 3.2 4.7 6.3 7.9 9.5 11.0 12.6 14.2 15.7 17.3 18.9 20.5 22.0 23.6 25.2 26.8 28.3 29.9 31.5 33.1 34.6 36.2 37.8 39.4 40.9 42.5 44.1 45.7 47.2 48.8 50.4 52.0 53.5 55.1 56.7 58.3 59.8 61.4 63.0 64.6 66.1 67.7 69.3 70.9 72.4 74.0 75.6 77.2

15.4 30.9 46.3 61.8 77.2 92.7 108.1 123.6 139.0 154.4 169.9 185.3 200.8 216.2 231.7 247.1 262.6 278.0 293.4 308.9 324.3 339.8 355.2 370.7 386.1 401.6 417.0 432.4 447.9 463.3 478.8 494.2 509.7 525.1 540.5 556.0 571.4 586.9 602.3 617.8 633.2 648.7 664.1 679.5 695.0 710.4 725.9 741.3 756.8

2240 4480 6720 8960 11200 13440 15680 17920 20160 22400 24640 26880 29120 31360 33600 35840 38080 40320 42560 44800 47040 49280 51520 53760 56000 58240 60480 62720 64960 67200 69440 7 1680 73920 76160 78400 80640 82880 85120 87360 89600 91840 94080 96320 98560 100800 103040 105280 107520 109760

50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 1O0

78.7 80.3 81.9 83.5 85.0 86.6 88.2 89.8 91.3 92.9 94.5 96.1 97.6 99.2 100.8 102.4 103.9 105.5 107.1 108.7 110.2 111.8 115.0 116.5 118.1 119.7 121.3 122.8 124.4 126.0 127.6 129.1 130.7 132.3 133.9 135.4 137.0 138.6 140.2 141.7 143.3 144.9 146.5 148.0 149.6 151.2 152.8 154.3 155.9 157.5

772.2 787.7 803.1 818.5 834.0 849.4 864.9 880.3 895.7 91 1.2 926.7 942.1 957.5 973.0 988.4 1004 1019 1034 1050 1066 1081 1097 1127 1143 1158 1174 1189 1205 1220 1236 1251 1266 1282 1297 1313 1328 1344 1359 1375 1390 1405 1421 1436 1452 1467 1483 1498 1514 1529 1544

112000 114240 116480 118720 120960 123200 125440 127680 129920 132160 134400 136640 138880 141120 143360 145600 147840 150080 152320 154560 156800 159040 163520 165760 168000 170240 172480 174720 176960 179200 181440 183680 185920 188160 190400 192640 194880 197120 199360 201600 203840 206080 208320 2 10560 2 12800 2 15040 2 17280 2 19520 22 1760 224000

For 101 or greater, add 100 measurements to number adding up to desired measurement (e.g., for 111, add measurements for 100 and l l ) .

137

138

The Practical Welding Engineer

3. Temperature Conversion Table Numbers in the center (bold) column are those to be converted. Refer to the left column (under "'Cy')to convert to Celsius, the right (under to convert to Fahrenheit. "C = %("F-32),"F = %"C+ 32. "OF')

c

b

-17.8 -1 7.2 -16.7 -16.1 -15.6 -1 5 -14.4 -1 3.9 -1 3.3 -12.8 -12.2 -1 1.7 -11.1 -1 0.6 -1 o -9.4 -8.9 -8.3 -7.8 -7.2 -6.7 -6.1 -5.6 -5 -4.4 -3.9 -3.3 -2.8 -2.2 -1.7 -1.1 -0.6

O 0.6 1.1 1.7 2.2 2.8 3.3 3.9 4.4 5 5.6 6.1 6.7 7.2 7.8 8.3 8.9 9.4 10 10.6 11.1 11.7 12.2 12.8 13.3

O 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 50

32 33.8 35.6 37.4 39.2 41 42.8 44.6 46.4 48.2 50 51.8 53.6 55.4 57.2 59 60.8 62.6 64.4 66.2 68 69.8 71.6 73.4 75.2 77 78.8 80.6 82.4 84.2 86 87.8 89.6 91.4 93.2 95 96.8 98.6 100.4 102.2 1 04 105.8 107.6 109.4 111.2 113 114.8 116.6 118.4 120.2 122 123.8 125.6 127.4 129.2 131 132.8

i3.9 14.4 15 15.6 16.1 16.7 17.2 17.8 18.3 18.9 19.4 20 20.6 21.1 21.7 22.2 22.8 23.3 23.9 24.4 25 25.6 26.1 26.7 27.2 27.8 28.3 28.9 29.4 30 30.6 31.1 31.7 32.2 32.8 33.3 33.9 34.4 35 35.6 36.1 36.7 37.2 37.8 43.3 48.9 54.4 60 65.6 71.1 76.7 82.2 87.8 93.3 98.9 104.4 110

57 58 59 00 61 62 03 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94

95 96 97 98 99 1O0 110 120 130 140 150 160 170 180 190 200 210 220 230

i34.6 136.4 138.2 140 141.8 143.6 145.4 147.2 149 150.8 152.6 154.4 156.2 158 159.8 161.6 163.4 165.2 167 168.8 170.6 172.4 174.2 176 177.8 179.6 181.4 183.2 185 186.8 188.6 190.4 192.2 194 195.8 197.6 199.4 201.2 203 204.8 206.6 208.4 21 0.2 212 230 248 266 284 302 320 338 356 374 392 41 O 428 446

1 1 5.6 121.1 126.7 132.2 137.8 143.3 148.9 154.4 160 165.6 171.1 176.7 182.2 187.8 193.3 198.9 204.4 210 21 5.6 221,l 226.7 232.2 237.8 243.3 248.9 254.4 260 265.6 271.1 276.7 282.2 287.8 293.3 298.9 304.4 31 O 31 5.6 321.1 326.7 332.2 337.8 343.3 348.9 354.4 360 365.6 371.1 376.7 382.2 387.8 393.3 398.9 404.4 41 O 41 5.6 421.1 426.7

240 250 260 270 280 290 300 31O 320 330 340 350 360 370 380 390 400 410 420 430 440 450 460 470 480 490 500 51O 520 530 540 550 560 570 580 590 600 01O 620 630 640 050 660 670 680 690 700 710 120 730 740 750 760 770 780 790 800

464 482 500 51 8 536 554 572 590 608 626 644 662 680 698 716 734 752 770 788 806 824 842 860 878 896 914 932 950 968 986 1004 1022 1040 1058 1076 1094 1112 1130 1148 1166 1184 1202 1220 1238 1256 1274 1292 1310 1328 1346 1364 1382 1400 1418 1436 1454 1472

Appendix II - Useful Tables and Diagrams 139

3. Temperature Conversion Table (Cont.) 'C

432.2 437.8 443.3 448.9 454.4 460 465.6 471.1 476.7 482.2 487.8 493.3 498.9 504.4 51 O 51 5.6 521.1 526.7 532.2 537.8 543.3 548.9 554.4 560 565.6 571.1 576.7 582.2 587.8 593.3 598.9 604.4 61O 615.6 621.1 626.7 632.2 637.8 643.3 648.9 654.4 660 665.6 671.1 676.7 682.2 687.8 693.3 698.9 704.4 710 715.6 721.1 726.7 732.2 737.8 743.3 748.9 754.4 760 765.6

81o 820 830 840 850 860 870 880 890 900 910 920 930 940 950 960 970 980 990 1O00 1010 1020 1030 1040 1050 1060 1070 1080 1O90 1100 1110 1120 1130 1140 1150 1160 1170 1180 1190 1200 1210 1220 1230 1240 1250 1260 1270 1280 1290 1300 1310 1320 1330 1340 1350 1360 1370 1380 1390 1400 1410

'F

'C

1490 1508 1526 1544 1562 1580 1598 1616 1634 1652 1670 1688 1706 1724 1742 1760 1778 1796 1814 1832 1850 1868 1886 1904 1922 1940 1958 1976 1994 2012 2030 2048 2066 2084 2102 21 20 2138 2156 2174 21 92 2210 2228 2246 2264 2282 2300 231 8 2336 2354 2372 2390 2408 2426 2444 2462 2480 2498 251 6 2534 2552 2570

771.1 776.7 782.2 787.8 793.3 798.9 804.4 81O 815.6 821.1 826.7 832.2 837.8 843.3 848.9 854.4 860 865.6 871.1 876.7 882.2 887.8 893.3 898.9 904.4 910 915.6 921.1 926.7 932.2 937.8 943.3 948.9 954.4 960 965.6 971.1 976.7 982.2 987.8 993.3 998.9 1004.4 1010 1 O1 5.6 1021.1 1026.7 1032.2 1037.8 1043.3 1048.9 1054.4 1060 1065.6 1071.1 1076.7 1082.2 1087.8 1093.3 1098.9 1104.4

1420 1430 1440 1450 1460 1470 1480 1490 1500 1510 1520 1530 1540 1550 1560 1570 1580 1590 1600 1610 1620 1630 1640 1650 1660 1670 1680 1690 1700 1710 1720 1730 1740 1750 1760 1770 1780 1790 1800 1810 1820 1830 1840 1850 1860 1870 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 1980 1990 2000 2010 2020

'F

'C

2588 2606 2624 2642 2660 2678 2696 271 4 2732 2750 2768 2786 2804 2822 2840 2858 2876 2894 291 2 2930 2948 2966 2984 3002 3020 3038 3056 3074 3092 3110 3128 31 46 3164 31 82 3200 3218 3236 3254 3272 3290 3308 3326 3344 3362 3380 3398 3416 3434 3452 3470 3488 3506 3524 3542 3560 3578 3596 3614 3632 3650 3668

fil0 1115.6 1121.1 1126.7 1132.2 1137.8 1143.3 1148.9 1154.4 1160 1165.6 1171.1 1176.7 1182.2 1187.8 1193.3 1198.9 1204.4 1210 1215.6 1221.1 1226.7 1232.2 1237.8 1243.3 1248.9 1254.4 1260 1265.6 1271.1 1276.7 1282.2 1287.8 1293.3 1298.9 1304.4 1310 1315.6 1321.1 1326.7 1332.2 1337.8 1343.3 1348.9 1354.4 1360 1365.6 1371.1 1376.7 1382.2 1387.8 1393.3 1398.9 1404.4 1410 1415.6 1421.1 1426.7 1432.2 1437.8 1443.3

2030 2040 2050 2060 2070 2080 2090 2100 2110 2120 2130 2140 2150 2160 2170 2180 2190 2200 2210 2220 2230 2240 2250 2260 2270 2280 2290 2300 2310 2320 2330 2340 2350 2360 2370 2380 2390 2400 2410 2420 2430 2440 2450 2460 2470 2480 2490 2500 2510 2520 2530 2540 2550 2560 2570 2580 2590 2600 2610 2620 2630

'F

3686 3704 3722 3740 3758 3776 3794 381 2 3830 3848 3866 3884 3902 3920 3938 3956 3974 3992 4010 4028 4046 4064 4082 4100 41 18 4136 4154 4172 4190 4208 4226 4244 4262 4280 4298 4316 4334 4352 4370 4388 4406 4424 4442 4460 4478 4496 4514 4532 4550 4568 4586 4604 4622 4640 4658 4676 4694 4712 4730 4748 4766

140

The Practical Welding Engineer

B. Formulas la. Carbon Equivalent C.E. Purpose: This value is calculated to estimate the susceptibility of steel to cold cracking in the HAZ. There are several equations proposed for calculating the carbon equivalents. Two are identified below. CE=C+-+Mn 6

CE = C+

C r + M o + V +-Ni+Cu%(BS5135)U.K.andEurope 15

5

Mn Si Ni Cr Mo V + -+ -+ ++% (WES 3001) Japan 4 14 5 6

24

40

lb. Carbon Equivalent for Cracking Susceptibility Pcm Purpose: As for CE, but in combination with other factors, this is used to predict preheat temperatures. Si Mn Cu Ni Pcm=C+-+-+-+-+-+-+-+SB 30 20 20 60

Cr

Mo

V

20

15

10

% (WES 3002)

2. Electrode Basicity Index B (Bas. Index) =

Ca0 + MgO + Ba0 + SrO + K,O + Li,O + CaF, SiO, + 0.5(Alzo, + Tio, + ZrO,)

+

(MnO+ Feo) 2

3. Electrode Consumption Formulas W=

DxAxL Efficiency

where W = Weight of electrode/welding wire required (kg) D = Density of weld metal (kg/m3).Approximate metal density of a steel (0.06% C and 0.4% Mn) is 7780 kg/m3at 20°C. A = Cross-sectional area of joint to be filled (mZ) L = Length of joint (m) Efficiency = Efficiency factor for various welding processes used, i.e.,l.O = 100 percent efficient.

Appendix II - Useful Tables and Diagrams

14 1

Typical efficiency factors are shielded metal arc welding = 0.65, gas tungsten arc welding = 1.00, gas metal arc welding = 0.95, submerged arc welding = 1.00 metal cored wires = 0.95 . (see also Chapter 5, Table 5.1, page 69)

4. Heat Input Formula ExIx6O J= v x 1000 where, J = Heat input in kJ/in. (or kJ/mm) E = Arc voltage in volts I = Welding current in amperes V = Arc speed, in./min (&min)

5. Thickness vs. Yield Stress If one grade of steel is replaced with another of a higher yield stress, the change in plate thickness achieved is expressed by the following relationship:

where, T, = thickness of higher yield stress plate material Ti = thickness of original plate material Rz = minimum yield stress of higher yield stress material Ri = minimum yield stress of original plate material

142

The Practical Welding Engineer

C. Diagrams 1. Iron Carbon Equilibrium Diagram Fe-Fe3C SYSTEM OC

2

3

4

5

CARBON CONTENT, wt. %

2. Nelson Cume Diagram To improve the resistance of steel to hydrogen attack, the two main alloying elements most commonly added are molybdenum and chromium. The most widely held theory regarding hydrogen attack is that atomic hydrogen dissolved in the steel reacts with iron carbides to form methane. If sufficient pressure of methane is generated, fissuring of the steel can result. Safe operating limits of temperature and pressure for CrMo steels have been established, and these are indicated on the Nelson curve diagram. If the operating temperature and hydrogen partial pressure fall below or to the left of the line for the alloy, then freedom from hydrogen damage is expected. At high temperatures indicated by the broken lines, surface decarburization results.

Appendix Il - Useful Tables and Djagrams

143

Hydrogen partial pressure, MPa abs

3.45 O F

6.90

10.34

13.79

17.24

20.7 800 OC

1500 1400

700 6.OCr-0.5Mo steel

_------I 600

I

~

2.OCr-0.5Mo steel

400

Carbon steel

400 300 500

1000

1500

2000

2500

3000

Hydrogen partial pressure, Ib/in.2 abc Safe Operating Limits of Temperature and Pressure for Cr-Mo Steels (After Nelson) Diagram as per American Petroleum Institute standard 941, Steels for Hydrogen Services at Elevated Temperatures and Pressures in Petroleum Refineries and Petrochemical Plants, 5th Edition, January 1997. Reprinted courtesy of the American Petroleum Institute.

NELSON CURVES

3. Schaeffler, DeLong, and WRC-1992 The Schaeffler and DeLong diagrams both relate the microstructural constitution of chromium-nickel based stainless steels to composition. Chromium, molybdenum, niobium and silicon are grouped as ferrite formers, while nickel, carbon and manganese are grouped as those elements that promote austenite formation. In the case of the DeLong diagram, nitrogen is included in the latter category. The Schaeffler Diagram is most commonly used to predict the approximate microstructure and hence resistance to hot cracking in manual metal-arc weld metal. The diagram can be applied to mixed and dissimilar welding. To apply the diagram, it is necessary to know the composition of the undiluted weld metal, the composition of the base metal(s), and the dilution and the proportion of base metal in the final weld.

~~

144

~

The Practical Welding Engineer

APPROXIMATE BOUNDARY OF AUSTENITE REGION FOR

.15c

003% MIN FOR ALL STEELS WRoUGHT MATERIALS

36

EXEFTWHERENOTED

SILIWN:

/

\

03W MIN THROUGHO.10 FOR HK; Cr STEELS. EQUALTO CASAUSTENITE

O

4

8

12

16

20

24

28

32

36

40

Creq = XCr +%Mo+ (1.5 x %Si) + í0.5x %I%)

THE SCHAEFFLER DIAGRAM FERRITE NUMBERíFN)

Appendix II - Useful Tables and Diagrams 145

18

20

22

24

26

28

30

Creq=%Cr +%Mo+ (0.7x%Nb)

THE WRC-1992 DIAGRAM

The following may be used as a guide to the dilutions that can occur with shielded metal arc welding (SMAW): 30 f40% Root run or square butt joint with root opening ,MAW{ Single run fillet or normal cladding 20 f30% Gas tungsten arc welding dilution varies from 30 percent for normal butt joint, and fillet welds up to 100 percent for autogenous root runs. Gas metal arc welds usually give 2 0 4 5 percent dilution, while submerged arc gives 30-50 percent. Fill passes of multi-run welds can range from O to 45 percent, depending upon the process and the exact position of the run. The DeLong diagram was developed as a result of the growing use of the gas-shielded welding processes. These are more prone to variable nitrogen pickup by the weld metal than the shielded metal arc welding process. The diagram is shown with a larger scale focus upon that area in which the majority of austenitic stainless weld metals lie. It is used specifically to predict the ferrite content of weld metals in which the nitrogen has been established by analysis. It is applicable to the majority of welding processes; note, however, that nitrogen content can vary with welding conditions and gas shielding efficiency.

146

The Practical Welding Engineer

More recently, the Welding Research Council’s diagram (WRC- 1992) has been developed and is generally considered more accurate, especially for grades such as duplex stainless steels. Section III of the American Society of Mechanical Engineers (ASME) Code (NB2433) allows the use of both the DeLong and WRC-1992 diagrams to predict ferrite content, preferably expressed as the Ferrite Number (FN), which differs from the previously used Ferrite Percent only above about 8 FN. Where nitrogen is not actually measured, the code permits the use of the following assumed values: shielded metal arc, gas tungsten arc, and submerged arc welding: 0.06 percent gas metal arc and gas shielded flux cored welding: 0.08 percent Self-shielded flux cored welding: 0.12 percent

4. Electrode Classification Mild Steel Electrodes The method of classifying electrodes is based on the use of a four-digit number, preceded by the letter “E’ for “electrode.” The fist two digits designate the minimum tensile strength of the weld metal (in 1000 lb/in.*) in the as-welded condition. The third digit indicates the position in which the electrode is capable of making satisfactory welds. The fourth digit indicates the current to be used and the type of flux coating. For example, the classification of E7018 electrodes is derived as follows: E7018 = Metal arc welding electrode Em18 = Weld metal UTS (ultimate tensile strength) 70,000 1b/h2 E7018 = Usable in all positions E7018 = Basic-type coating with iron powder AC or DC The detail of the classification is shown below. First and second digits E 6ûxx: As-welded deposit, UTS 60,000 lb/in.’ minimum, for E 6010, E 6011, E 6012, E 6013, E 6020, E 6027 UTS E 7ûxx: As-welded deposit, UTS 70,000 lb/in.’ min for E 7014, 7015, 7016,7018, E 7024 and E 7028

Appendix Il - Useful Tables and Diagrams 147

Third and fourth digits

The third and fourth digits indicate positional usability and flux coating types. Em1 O: High-cellulose coating bonded with sodium silicate; deeply penetrating, forceful, spray-type arc; thin, friable slag; all-positional; DC electrode positive only. E m l l : Similar to ExxlO, but bonded with potassium silicate to permit use on AC or DC electrode positive only. Exxl2: High rutile coating, bonded with sodium silicate; quiet arc, medium penetration; all-positional; AC or DC electrode negative. Eml3: Coating similar to Exxl2, but with addition of easily ionized materials and bonded with potassium silicate to give steady arc on low voltage supply; slag is fluid and easily removed; all-positional; AC or DC electrode negative. E m 1 4 Coating similar to Exxl2 and Exxl3 types with addition of medium quantity of iron powder; all-positional; AC or DC. E m 1 5 Lime-fluoride coating (basic, low-hydrogen) type, bonded with sodium silicate; all-positional; for welding high-tensile steels; DC electrode positive only. E m 1 6 Similar coating.to Exxl5, but bonded with potassium silicate; AC or DC electrode positive. E m 1 8 Coating similar to ExxlS and Exxló, but with addition of iron powder; all-positional; AC or DC. E m 2 0 High iron oxide coating bonded with sodium silicate; for welding in flat or horizontalhertical (HV) positions; good X-ray quality; AC or DC.

Index

access to joints, 82 air-jet efficiency, in cutting, 129 all-weld tensile tests, 33 annular stiffener weld, 87-88 arc air gougingkutting, 129-130 arc blow, 92-94, 113 arc monitoring, 52-53 arc stability, 23 arc strikes, 107 arc time, cost of, 68 automated processes, cost of, 68 automatic processes, cost of, 73-74 automation, degree of, 73-74 backgouging, 44 backstep welding technique, 94, 106, 108 baking ovens, 86 bare wire electrode bead appearance, 23 bead contours, 75 bend testing, 34-35 bevel angle, 76 bid. See tender block weave, 90 BS 7910,99 budgets, estimating, 25 burning. See oxyfuel cutting buttering, 95-97 cap-pass sequence control, 97-98 carbon equivalent formulae, 140 cast-to-cast variability, 90-92 centerline cracks, 105-106 Charpy V-notch test, 35-36 chevron cracking, in SAW, 84-88

claims, 1 codes, 108. See also specifications for fracture toughness, 99 communication, engineer and welder, 47-49 about procedure tests, 55 compensation, 1, 5-6 conflict of interest, 50 consumables. See also electrodes availability of, 16 changing, for costs, 74-75 coating brittleness, 23 color coding, 59 control of, 58-60 COSt Of, 67,68-70,75 evaluation of, 18-22 in defect analysis, 44 issue of, 80 low hydrogen, storage, 59-60 metal powder, 75 nitrogen in, 65 operability of, 20 organization of, 59-60 problems with, 64 properties of, 20 storage of, 58-60 traceability of, 58 contracts, 1, 2-5 obligations of, 2, 5 planning for, 3 purchasing, 2 , 4 subcontracting, 2, 4 tender, 2, 3 copper inclusions, 114, 119 corrosion resistance testing, 32

149

I 50

The Practicaí Weiúing Engineer

crack tip opening displacement (CTOD), 35, 3739,64,65,98-99 cracking susceptibility formula, 140 cracks, types of, 102-106 crater pipes, i 14-115 craters, 115 cross-weid tensile tests, 33 cutting oxygen pressure, 128 D1.l, 25 defect analysis, 43-46 defects. See weld defects DeLong diagram, 144 deposition efficiency factor, 69 deposition rate, in calculating costs, 68 discontinuities. See also weld defects chevron cracking, 84-88 communicating to welder, 49 from arc blow, 93 hydrogen cracking, 31 in micro-examination, 32 transverse planar, 85 distortion, excess, 107-108 drag, 127 electrode basicity index, 140 electrode gouging/cutting, 130-132 electrodes. See also consumables angle, 94 bare wire, 22 cellulostic/rutile, 34, 75 classification of, 146-147 coating, 23 diameter, charts for, 56, 57 ferritic,buttering with, 97 low-hydrogen, 75 re-striking, 23 SMAW rods, storage of, 59 solid GTAW wire (straight lengths), 59 stub lengths, 54 electroslag welding (ESW), 76 normalizing, 95 equipment assessment, 19-20 etches, dendritic and nital, 31.32 ferrite testing, 32 filler metal. See consumables fish eyes, 34,35 fitness for purpose, 99 fixturing, 77-78 flame cutting. See oxyfiel cutting flux cored arc welding (FCAW), 16,31,34, 49, 84, 118 low toughness in, 89-90 test procedures, 62-63 flux recycling, 86 flux-covered electrode, gouging with, 131 fracture mechanics. See fitness for purpose

fracture mechanics tests, 37-39 gas shielded processes, cost of, 69-70 gas tungsten arc welding (GTAW), 19, 145 cast-to-cast variability in, 91 gauges, 53-54 geometry defects, 44 groove angle, 76 grooving. See electrode gouging hardness equivalent table, 136 hardness survey, 31-32 hardness surveys, 10 heat input formula, 141 heat input, significance of, 55 heat line bending, 108 heat treatment, 108 heat-affected zone (HAZ) hardness of, 29, 31, 55, 56, 96 hydrogen cracking in, 31 in CTOD, 38-39 in impact testing, 36 in micro-examination, 32 in temper bead technique, 29-30 heating pad. See fixturing high deposition rate techniques, 75-76 hot cracking susceptibility (HCS), 122 hot-wire gas tungsten arc welding (GTAW), 76 hydrogen cracking, 31 hydrogen cracks, in HAZ, 102-103 hydrogen, in chevron cracking, 85-86 impact tests, 10, 35-36, 65 incomplete fusion, 46, 120-121 incomplete root penetration, 109 international specifications, 7 iron carbon equilibrium diagram, 142 joint completion rate, 68 joint type, costs, 72, 80-82 kerf. 125 labor, cost of, 67 lamellar tearing, 103-104 linear completion rate, 68 low-alloy steels, cutting, 125 low-carbon steel, cutting, 125 macro-examination, 30-31 macroscopic examination, FCAW, 63 manual processes, cost of, 68,73 material mechanical properties, 16-18 material weldability, 9 mechanical tests, 30-36 bend testing, 34-35 hardness survey, 31-32 impact testing, 10, 35-36

Index 151 macro-examination, 30-3 1 micro-examination, 32 tensile testing, 33-34 mechanized processes, cost of. See automated processes micro-examination, 32 mild steel, cutting, 125 misalignment, 109-110 M d S ratios, 106, 122 moisture level, in flux, 86 monitoring production, 9-1 1 by pass length, 57-58 multiwire welding, 75 narrow-groove welding, 65, 76-77 nelson curve diagram, 142-143 nickel alloys, 31 nital etch, 31, 32 nondestructive examination (NDE), 13, 16, 18, 28,30,84, 99 nozzle, cutting, 130 offshore fabrication,justifying pass lengths in, 55 overheating, 23 overpenetration, 111 oxyfuel cutting, 125-129 partial penetration welds, costs, 81-82 pass length (SMAW), 55-58 pipe butt joint weld procedure, 28 poor profile, 111-112 porosity, 46 porosity, elongated, 117 porosity, restart, 115 porosity, uniformísurface, 116-117 positioners. See faturing postweld heat treatment (PWHT) elimination of, 94-99 power sources, 19 selecting to avoid arc blow, 94 preheat, 49 preheat flames, in cutting, 128 preheating bands, 78 prequalification procedures, 25-26 material costs, 28-29 production monitoring, 9-11 by pass length, 57-58 production time, 17 profile defects, 106-113 profit, 1, 2 purchasing, 2 , 4 qualification test program, 26,28 reheat cracking, 104 reinforcement, excess, 108-109 replicas, weld, 51-52 residual magnetism, arc blow, 93

rods, storage of, 59 root concavity, 112 SASPA-NANSA, 31,32 schaeffler diagram, 144 shielded m e G arc welding (SMAW), 16, 17, 18, 22, 28, 54,55, 118, 145 costs of, 70-71 procedure tests for, 55 single-pass welds, 32 slag inclusions, 118 slag removal, 23 solidification cracks, 105-106, 122-123 spatter, 23,46, 112-113 specifications, 7-1 1 clarifications to, 8 exceptions to, 8 international codes, 7,47 national codes, documentation of, 47 stainless steel alloys, 31 Standard Welding Terms and Definitions, 25 stress conversion, 137 stringer bead technique, 90 stub lengths, 54 variable, 56-57 subcontracting, 2 , 4 submerged arc strip cladding, 76 submerged arc welding (SAW), 16, 28, 38.75, 118 fluxes, storage of, 59 narrow-groove, 77 without iron powder additions, 65 suppliers, data from, 11 temper bead technique, 97-98 temperature conversion table, 138-139 tender, 2, 3 tensile loading, 86 tensile testing, 33-34 fish-eyes in, 34 test failures, 39-42 test plates, 26, 30 test programs, 26-28 test welds, 29-30, 31 material costs, 28-29 techniques, 3 1 yield stress, in test welds, 33 toe profiles, 75 tool issue, restricted, 80 total weld cost, equation, 67 transverse tensile test, 33 tungsten inclusions, 119-120 tunneling. See porosiq, elongated turning rolls. See f i t u r i n g ultimate tensile strength (UTS), 33 ultrasonic testing, buttering for, 97 ultrasonic testing, of SAW, 84-85

The Plactlcal Welding €ngineer

152

undercut, 113-114 units of crack susceptibility (UCS), 122 variation request, 6 vibratory stress relief, 98 volume fill rate, in calculating costs, 68 volumetric defects, 114-120

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---

weave technique, FCAW, 89-90 weid bead appearance, 23 weid porosity, 46 weid cracks, types, 102-106 weld defects. See also discontinuities analysis of, 43-46 arc strikes, 107 copper inclusions, 114, 119 crater pipes, 114-115 excess distortion, 107-108 geometry-related, 44 hydrogen cracks, in HAZ, 102-103 incomplete fusion, 46, 120-121 incomplete root penetration, 109 lamellar tearing, 103-104 material-related, 46 misalignment, 109-110 overpenetration, 111 poor profile, 111-112 porosity, elongated, 117 porosity, restart, 115 porosity, uniform/surface, 116-117 profile, 106-113 reheat cracking, 104 reinforcement, excess, 108-109 root concavity, 112 slag inclusions, 118 solidification cracks, 105-106 spatter, 46, 112-113 tungsten inclusions, 119-120 undercut, 113-114 volurnehic defects, 114-120 weld cracks, types of, 102-103 weld metal cracking, 86 welder-related, 45-46 weid failure, material related, 64 weld geometry defects, 44 weld iength, 70 weld metal cracking, 86 weid microstructure, 55 weid procedure requirements, 25-30 prequalification procedures, 25-26 test programs, 26-28 weld procedures, 27 pipe butt joint weid qualification, 28 weld replicas, 51-52 weid test failures, dealing with, 61-66 weid test pieces, 60-66 weld volume, 70, 76-77 weld weight, 70

weldability, 9, 16-18 welder access, 44 welders communicating with, 47-49 supervision of,50 training and qualification of, 47-49 welding costs, estimating, 67-71 welding costs, reducing, 72-82 welding procedure specification (WF'S), 55,57 welding processes constraints of, 16.17 consumable availability, 16 economic factors, 17-18 electroslag welding (ESW), 76 normalizing, 95 flux cored arc welding (FCAW), 16, 31, 34, 49,84, 118 low toughness in, 89-90 gas tungsten arc welding (GTAW), 19, 145 cast-to-cast variability in, 91 hot-wire gas tungsten arc welding (GTAW), 76 material weldability, 16-18 pass length (SMAW), 55-57 production time with, 17 selection of, 13-18 shielded metal arc welding (SMAW), 16, 17, 18, 22.28, 54.55, 118, 145 Costs of, 70-71 procedure tests for, $5 submerged arc welding (SAW), 16,28,38, 75, 118 fluxes, storage of, 59 narrow-groove, 77 without iron powder additions, 65 wire, storage of, 59 working environment, 79-80 workmanship example, 5 1 WRC-1992 diagram, 145 yield stress, 33 yield stress formula, 141

--``,,`````,,,`,`,`,,,```,```,`-`-`,,`,,`,`,,`---