Microsoft Word - Creep Behaviour of Reformer Tubes

Sheffield Hallam University

School of Engineering

Creep Behavior and Life Time Estimation of Reformer Tube (G-NiCr28W)

Prepared by: EL-Sayed Ibrahim

Supervised by: Dr: Syed Hassan

Dedication

To my wife and my children Reem and Mohab, and

my company ANSDK

ACKNOWLEDGEMENT

I would like to express my deep thanks to my supervisor Dr Syed Hassan for his great support and guidance throughout the project. I would like also to thank Dr Rachel Clington for her assistance especially with the preparation of the machine for the test.

I am also grateful to Mr. Wilknson, the technical manager of the school, for his always support. I’d like to express my thanks to Brain (workshop) for his efforts to solve the problems I had with the creep-testing machine.

2

Table of Contents Preface

5

1. Introduction

6

2. Characteristics of High Temperature Applications

10

3. Creep Curve

13

4. Creep Characteristics 4.1 Time

17

4.2 Temperature

17

4.3 Stress

18

4.4 Microstructure

18

5. Fundamentals Mechanisms Responsible for Creep 5.1 Diffusion Creep Process

19 21

5.1.1 Lattice Diffusion Creep

24

5.1.2 Grain Boundary Diffusion Creep

26

5.2 Dislocation Creep

28

5.3 Grain Boundary Sliding and Supperplasticity

30

6. Micro-Mechanical Deformation Mechanisms During Creep

33

6.1 Intragranular Creep Deformation 6.1.1 Crystallographic Slip

33

6.1.2 Sub grain Formation

34

6.2 Intergranular Creep Deformation 6.2.1 Grain Boundary Sliding

35

6.2.2 Creep Cavity Nucleation

36

6.2.3 Fold Formation and Grain Boundary Migration

37

7. Creep Life Assessment 7.1 Current Approaches of Creep Life Assessment 7.1.a Operational Condition Monitoring Approach

39 40

7.1.b Methods Based on Post-Service Examination Sampling and Calculation 7.1.1 Creep Rupture Testing 7.1.1.1 Larson Miller Method

40 41 41

3

7.1.1.2 Manson and Hafered Method

42

7.1.1.3 Orr-Shery-Dorn Method

43

7.1.2 Strain Measurement Methods

44

7.2 Mechanistic Models of Creep Damage Process 7.2.1 Reduction in Load Bearing Section

44

7.2.2 Micro structural Degradation

45

7.2.3 Cavitation Damage

45

7.2.4 Effect of Environment

46

7.3 Phenomenological Models

46

8. Material Specifications 8.1 Chemical Composition

49

8.2 Mechanical Properties at Room Temperature

49

8.3 Physical Properties

50

9. Experimental Procedures 9.1 Test Pieces

51

9.2 Impact Test

52

9.3 Hardness Test

54

9.4 Tensile Test

56

9.5 Accelerated Creep Rupture Test

58

9.5.1 Test Technique

58

9.5.2 Test Procedure

59

9.6 Microstructure Examination

75

10. Analysis of The Tests Result

78

11. Conclusion

85

12. Future Plan and Recommendation

86

13. References 14. Appendixes 14.1 List of Figures

89

14.2 List of Tables

93

14.3 Recorded Test Results

94

4

Preface For the industrial components operating at elevated temperatures and subjected to steady load conditions, there is a need to have enough information about the creep behavior, creep mechanism controls that behavior, and to understand the failure mechanism. Meanwhile there should be a method for assessing the remaining lifetime of those components. This project aims at studying the creep behavior of tubes of alloys GNiCr28W operating in high temperature furnace (10000C), classifying the creep mechanism and establish a criterion for remaining lifetime.

5

1. Introduction Alexandria National Iron and Steel Co. (ANSDK) is one of the leading companies in the Middle East for producing reinforced steel. The company consists of four main plants •

Direct reduction plant,

•

Steel making plant,

•

Continuous casting plant, and

•

Rolling plant.

Direct reduction plant is one of the Midrex plants, which uses Midrex technology to produce highly metallized iron. This iron is called sponge iron and used as raw material in the Steel making furnaces with small ratio of scrape. Midrex technology is based on using hydrogen (H2) & carbon monoxide (CO), as reducing gases to reduce the iron oxide pellets (FeO) into sponge iron (DRI), direct reduced iron, according to the following reactions H2 + FeO = Fe + H2O CO + FeO = Fe + CO2 Typical figure of ANSDK Midrex plant is illustrated in figure (1.1) Feed C onveyor

Natural Gas

C O : 20% C o 2 :2 0 % C H 4 :8 % P ro c e ss g a s c o m p resso r

H 2 0 :1 7 %

U p p e r s lid e g a te

H 2 : 35%

P ro c e ss g a a s

C O : 35%

T op gas s c ru b b e r

S h a ft F u rn a c e

C O 2 : 3% H2 R e c u p e ra to r

: 55%

R e fo r m e d g a s

C O +FeO = Fe+C O 2 H 2+ FeO =Fe+ H 2O

C H 4 : 2% H 2O : 5% C o o lin g g a s Feed M ix gas

R e fo rm e r

L o w e r s lid e g a te

T y p ic a l M id r e x P la n t ANSDK

C o o lin g

C o o lin g

G as

G as

co m p resso r

S cru b b e r

D is c h a rg e C o n v e y o r

Feeder

SD R P

Figure (1.1) ANSDK Midrex plant configuration 6

These reducing gases are generated in catalytic tube placed in the so-called reformer. The reformer contains 468 nickel based alloy tubes. Endothermic reactions take place inside the tube to produce the reducing gases as follows: CH4 + CO2 = 2CO + 2 H2 CH4 + H2O = CO + 3 H2 The reformer is provided with Burners to produce the heat required to achieve the two reactions. Figure (1.2) indicates the condition at which the tubes operate

Refractory

~ 930oC

1100 oC

Textile bellow

Top canister

Welding line Reformer

Catalyst

Welding line

Bottom canister

Textile bellow

Gas inlet 450oC

1020oC

Figure (1.2) Operating condition of tubes

7

The tubes operate at very high temperature @ 1050 oC. The suspension mechanism of the tube, as indicated in figure (1.3) counter-acts the effect of internal components and allow the tube’s own weight to be the only factor affecting the expanding of the tube under that high temperature. The tubes are being replaced every 5 years as a preventive action to avoid any disaster that might take place, in case of tubes failure because tubes contain highly flammable and explosive media. There is no criteria available indicating whether the tubes deserve to change or not or whether the tubes can be used for further period. However there is a recommendation from the supplier to limit the elongation of tube to 280mm, without any explanation why exactly that figure. This limit is taking into consideration when replacing the tube, but it is not the main reason for replacement. Tubes themselves are costly as well as the replacement tasks. As an example, in 1990 the tube price only (excludes the inside accessories) was $ 12000 / tube. More over to replace the tube the plant has to stop for about one month, which simply means production loss. To rough calculate the total cost of tubes replacement we have to mention that at the time of Direct Reduction Plant stoppage more scrape is used in steel making furnace. Scrape price is more than DRI price by about $50 / ton. So by simple calculation based on the production of DRI is 2400 ton /day (72000 ton/month production loss) the total cost increase is 72000* 50 = $3,600,000 Nowadays the market became very competitive so saving the $9,216,000 (Tube prices+ cost increase regardless the cost of replacement) even for only one year will affect the competitive position of the company.

8

This study aims at investigating the creep behavior of the tube under the used temperatures and its own weight. This investigation will be through conducting accelerated creep rupture. Establishing of criteria for tubes replacement will be obtaining from the resulted creep behavior curve from the test. The purpose of this is to save money spending without actually need for that spending and consequently increasing the profitability of the company and strengthened its competitive position. In addition to that other tests such as tensile test impact test, hardness test will also be conducted to have some properties of the tube alloy at room and elevated temperatures.

Reformer structure

Pulley

Tension W ire

Counter weight ~600 Kg

Suspension mechanism

Figure (1.3) Suspension mechanism of tube

9

2. Characteristics of High Temperature Applications The technological developments since the early 20th century has required materials that resist very high temperatures. Applications of these developments lie mainly in the following areas:1 1. Gas turbine (stationary and on craft), whose blades operate at temperatures of 800- 900 K. The burner and after burner sections operate at higher temperatures, 1300 – 1400 K. 2. Nuclear reactors where pressure vessels and piping operate at 560 – 750 K. Reactor skirts operate at 850- 950 K. 3. Chemical and petrochemical industries where some steam reformers operate at 1250- 1350 K. Materials used in these applications are subjected to a working temperatures range (0.4~0.65) Tm, where Tm is the melting point of the material. Of course these extreme working conditions affect the materials life and cause degradation, which can be classified as follows:

Material Degradation at High Temperatures Mechanical Degradation 1.

Chemical Degradation

Mechanical degradation. In spite of initially resist the applied loads, the material undergoes anelastic deformation; its dimensions change with time.

2.

Chemical degradation. This is due to the reaction of the material with the chemical environment and to the diffusion of external elements into the

10

materials. Chlorination (which affects the properties of super alloys used in jet turbines) and internal oxidation are examples of chemical degradation. Our concern here is the mechanical degradation. The thermo-mechanical degradation or time dependant deformation of a material is known as creep. Micro-structural changes take place at elevated temperatures; lead to materials failure at a stress that is much lower than its ultimate tensile strength measured at ambient temperature. A great number of failures can be attributed either to creep or to a combination of creep and fatigue. Creep is characterized by a slow flow of the material, which behaves as if it were viscous. If a component of a structure is subjected to a constant tensile load, the decrease in cross-sectional area (due to the increase in length resulting from creep) generates an increase in stress; when the stress reaches the value at which failure occurs statically (ultimate tensile stress), failure occurs. Elastic, plastic, visco-elastic, and visco-plastic deformations can all be included in the creep process, depending on the material and the characteristic time of the deformation. However, creep deformation is always treated as plastic deformation because the failures associated with creep are similar to those due to yield in plastic deformation of the materials. There are various mechanisms of creep in materials at elevated temperatures and thus there are different creep models. The temperature range, in Kelvin, for which creep is important in metals and ceramics, is 0.4 Tm
∃c as a consequence of the relationship

9a>9b> 9c. In figure (3.2) the rupture times tar, trb, and trc increase with decreasing stress. The strains


are called instantaneous strains and correspond to the strains at the instant of loading. 0

In figure (3.3) the engineering stress was kept constant and the temperature was varied. Since the tests are conducted in tension, the stress rises as the length of the test piece increases, because of the reduction in area. The dotted lines in Figures (3.2) and (3.3) represent the constant stress curves. Initially they are identical, because
e = 0.As the specimen increases in length, the stress increases and so does the creep rate, at a constant load. The failure times under constant stress and constant load can be drastically different. The curves shown in Figure (3.3) have been expressed mathematically as

Figure (3.3) Creep strain vs. time at constant engineering stress and different temperatures Reference: M.Andre et al, Mechanical Behavior of Materials


t =
0 +
[ 1 – exp(-mt)] +
∃st

(3.1)1

15

The minimum creep rate is usually represented by


∃s = (AGb/ kT)D0 exp(-QC/RT)(b/d)p(9/G)n

(3.2)1

Where A is the dimensionless constant, D0 is the frequency factor, G is the shear modulus, b is the Burgers vector, k is Boltzmann’s constant, T is the absolute temperature, 9 is the applied stress, d is the grain size, P is the inverse grain size exponent, n is the stress exponent, QC is the appropriate activation energy, and R is the gas constant. This equation is known as the Mukherjee-Bird-Dorn equation. The activation energy for diffusion is often equal to the activation energy for creep (QC = QD). The diffusion coefficient is

D = D0 exp (-QD/RT)

(3.3 )1


∃s = AGbD/ kT)(b/d)p(9/G)n

(3.4)1

And

Essentially, equations 3.2 and 3.4 express the steady –state creep rate as a function of the applied stress, temperature, and grain size.

16

4. Creep Characteristics Creep characteristics depend on several factors such as time, temperature, Stress, and the microstructure.

4.1 Time. A time scale is always involved in a creep test. As a comparison, for most engineering materials tested at low temperatures, measured tensile properties are relatively dependent of the test time, regardless of whether it is 5 minutes or 5 hour. The main reason of that is the involvement of the thermally activated time dependent processes. A single dominant thermally activated process usually controls the overall creep rate during creep deformation. For example, if the controlling process is diffusional, the creep rate is called diffusion controlled.1

4.2 Temperature Normally the creep processes involve mechanisms at the atomic scale. The mobility of atoms and vacancies increases rapidly with temperature so they can diffuse through the lattice of the material along the direction of the hydrostatic stress gradient, which is called self-diffusion. The self-diffusion of atoms can or vacancies helps dislocations (line crystal defect) climb (a motion of dislocation toward the direction perpendicular to its slip plane). At low temperatures, becomes less diffusion- controlled. Diffusion can occur, but is limited in local porous areas, like grain boundaries and phase interfaces, which is called grin boundary diffusion.3 Generally creep becomes of engineering importance at T>0.5Tm. This is just empirical guideline based on the observations that above 0.5 Tm, creep is most likely to be governed by mechanisms that depend on self-diffusion.

17

4.3 Stress Stress level and stress state greatly affect the creep rate. As explained in figure (3.2) shows the effect of different stress at constant temperature. Measurements of creep property are classified into creep test and creep rupture test according to the stress level. •

Creep test Normally the test is carried out at low stresses to determine the steady state creep rate. The total strain is often less than 0.5%

•

Creep rupture test It is similar to creep test except that high load is applied to precipitate failure of the material. The total strain can be as high as 50%.

4.4 Micro-structure Microstructure of the material plays an important part of determining the creep behavior of the material. Grain size affects the creep rate in all the three stages. Precipitations and impurity particles initiate creep cavities. Porosity due to sintering, particularly in ceramics, is another microstructure effect.10

18

5. Fundamental Mechanisms Responsible For Creep The models for creep deformation are based on the observations of different micro structural mechanisms and are developed to represent


∃s. Usually
∃s is established

as the most basic creep parameter for modeling purposes. creep stress and temperature. Therefore, for obtaining


∃s is strongly influenced by


∃s, various models for creep

deformations are developed according to stress and temperature ranges. For each material, a creep deformation mechanism map can be plotted to identify the dominant deformation mechanism for different stress- temperature combinations. Figure (5.1) is a simplified deformation mechanism map in stress –temperature space. Stresses have been normalised by the shear modulus, and the temperatures have been normalised to the homologenous scale. According to the map, various creep models can be related to steady-state creep.3

Figure (5.1) Simplified creep deformation mechanisms map Reference: IEEE Vol.42, NO.3, 1993

19

In fact the understanding of creep can be divided into two periods: before and after 1954. In that year, Orr et al. introduced the concept that the activation energy for creep and diffusion are the same for an appreciable number of metals (more than 25). The activation energy for diffusion is connected to the diffusion coefficient by Equation (3.3). For temperatures below 0.5 Tm, half the melting point of the material, in kelvins, the activation energy for creep tends to be lower than that for self-diffusion, because diffusion takes place preferentially along dislocations (pipe diffusion), instead of in bulk. Figure (5.2) shows the variation in QC /QD for some metals and ceramics. The activation energy for diffusion through dislocations is considerably lower than that for bulk diffusion.

Figure (5.2) Ratio between activation energy for secondary creep and activation energy for bulk diffusion as a function of temperature Reference: Marc Andre et al, Mechanical Behavior of Materials For the temperature range T> 0.5 Tm, the mechanisms responsible for creep can be conveniently described as a function of the applied stress. The creep mechanisms can be divided into two major groups: boundary mechanisms, in which grain boundaries and, therefore, grain size, play a major role; and lattice mechanisms, which occur independently of grain boundaries. In Equation 3.2, the exponent p = 0 for lattice mechanisms, and p > 1 for boundary mechanisms. Figure (5.3) summaries the models of creep deformations

20

Models for Creep Deformations

Dislocation Creep

Dislocation Glide

Dislocation Climb

Diffusion Creep

Lattice diffusion creep (Coble)

Grain boundary diffusion creep

Figure (5.3) Models of creep deformation

5.1 Diffusion Creep Process When the high temperature creep properties of materials are described using a power law as


∃s = (AGb/ kT)D0 exp(-QC/RT)(b/d)p(9/G)n

The activation energy for creep QC is often reported to be close to the activation energy for the lattice self-diffusion. Since the diffusion is clearly important in determining creep behavior, it is perhaps relevant to summarize the principles governing diffusion in metals and alloys.1 Diffusion normally occurs due to presence of vacancies within the crystal lattice. As illustrated in figure (5.4) an atom can move to the site of an adjacent vacancy, provided only that the atom has sufficient thermal energy to jump from its original site. At any temperature, the average thermal energy of an atom is 3kT, where k is Boltzmann’s constant (1.38 * 10-23 Jatom-1K-1). However, the thermal energy is not uniformly distributed among the atoms in the crystal. As atoms vibrate about their mean position, with frequency − (typically ~ 1013 per second), the atoms collide repeatedly with their neighborus so energy is transferred continually from one to another. At any instant, any

21

single atom therefore has more or less energy than the average value, i.e. it is vibrating more or less violently about its mean position than the other atoms in the crystal.2 Let the energy, which the atom requires in order to move into the adjacent vacant lattice site be q. The probability, p, of any atom having an energy that is equal to or greater than q, at any instant, is given as

p = exp-(q/kT)

(5.1)2

Figure (5.4) Schematic representation of diffusion as a result of vacancy movement, showing the energy associated with the different atom locations. Reference. R.W.Evans et al, Introduction to Creep Even if an atom has sufficient energy to jump from its original lattice site, it can only move if a vacancy exists on the adjacent site to allow the move to take place. The equilibrium concentration of vacancies in a lattice at temperature, T, is given by the expression

N  q^  = exp−   N  kT  v

(5.2)2

L

where Nv is the number of vacancies out of a total number of lattice sites, NL, and q' is the thermal energy needed to displace an atom from a lattice site to create a vacancy. Thus,

22

the probability of an atom having sufficient energy to jump successfully is given by equation (5.1) while the probability of this atom being adjacent to a vacancy, so that the jump can take place, is given by equation (5.2). The probability of two independent processes occurring simultaneously is the product of the individual probabilities, so the frequency with which atom movements occur () can be expressed as

 = exp-(q/kT). exp-(q^I kT) therefore

 = exp -[(q + q^)/kT ]

(5.3)2 (5.4)2

The rate of diffusion in a solid depends directly on the frequency with which individual atom movements occur, so that

D∝  ∝ exp-( q* /kT) *

With q =q+q

(5.5)2

^

D is called the diffusion coefficient, which has units of m2s-1. However, q* represents a very small quantity of thermal energy (defined in units of joules per atom) and it is then more usual to rewrite equation (5.5) as

D =D0exp-(Q/RT)

(5.6)2

Where D0 is a constant and R is the universal gas constant (8.31 Jmol-1K-1) given as R =NAk NA is Avogadro's number, which is the number of atoms per mol (6.022 x 1023 mol-1). The parameter, Q, is then termed the activation energy for diffusion (with units of Jmol1

), given as

Q =NAq*

(5.7)2

23

Vacancy movement is the dominant diffusion mechanism in most metals and alloys, and the movement of vacancies in a pure metal, say copper, is called self-diffusion, i.e. the movement of copper atoms in the copper lattice. The activation energy for this process is called the activation energy for self-diffusion, QSD. Yet while the activation energy for creep is frequently found to be equal to that for self-diffusion at high creep temperatures, QC values significantly less than QSD are often reported at creep temperatures of around 0.4 to 0.6 Tm, Diffusion creep tends to occur for 9/G [10-4. (This value depends, to a certain extent, on the metal.) Two mechanisms are considered important in the region of diffusion creep.

5.1.1 Lattice Diffusion Creep Lattice diffusion creep represents the creep process controlled by the diffusion of atoms or vacancies under low applied stress and is often called Nabbaro- Herring creep. Usually at 9/G 0.5Tm Nabarro and Herring proposed the mechanism shown schematically in Figure (5.5). It involves the flux of vacancies inside the grain. The vacancies move in such a way as to produce an increase in length of the grain along the direction of applied (tensile) stress. Hence, the vacancies move from the top and bottom region in the figure to the lateral regions of the grain. The boundaries perpendicular (or close to perpendicular) to the loading direction are distended and are sources of vacancies. The boundaries close to parallel to the loading direction act as sinks.1

24

Nabarro and Herring developed a mathematical expression connecting the vacancy flux to the strain rate. They started by supposing that the “source” boundaries had a concentration of vacancies equal to C0 + C and sink boundaries a concentration C0. They assumed that

C = C0 9/ kT, Where 9 is the applied stress and C0 the equilibrium vacancy concentration

C0 = NV/NL = exp –( q^/kT)

(5.8)2

Figure (5.5) Flow of vacancies according to (a) Nabarro-Herring and (b) coble mechanisms, resulting in increase in the length of the specimen Reference: Marc Andre et al, Mechanical Behavior of Materials The flux of vacancies is therefore given by

J = k` De(C/x) = k``De (C/d)

(4.8)1

25

Where x is the diffusion distance, which is a direct function of the grain size (approximately equal to d/2), De is the lattice diffusion coefficient, d is the grain size `

``

``

`

diameter, k and k are proportionality constants. k = 2 k . The strain rate is related to the increase in grain size d in the direction of the applied stress:


∃ = 1/d (dd/dt). The change in grain length, dd/dt, can be obtained from the shift of vacancies, each having volume : dd/dt = J. thus the following equation can be obtained for creep rate


∃NH = k``De C0 9/ d2 kT,

(5.9)1

Expressing this equation in the format of equation 1.3 (making  = 0.7b3)


∃NH = ANH(GbDe/ kT)(b/d)2(9/G)

(5.10)1

ANH is typically equal to 10 – 15. The equation for the steady state creep rate is derived according to diffusion dynamics in poly-crystals is


∃s = 14(a/d)2.a.9.(De/kT)

(5.11)3

Where a is the atomic size

5.1.2

Grain Boundary Diffusion Creep At 9/G 10-3, dislocation gliding dominates over atomic diffusion effect Table (5.1) summarize the n and Qc values associated with different creep mechanisms, together with the general stress and temperature conditions under which each process is usually considered to become dominant.

29

Creep Process

Temp.

Stress

n value

Qc value

High temperature dislocation

>~0.7Tm

Intermediate/high

>3

~QSD

Low temperature dislocation

~0.4 to~0.7Tm

Intermediate/high

>3

Qcore***

Low

~1

QSD

Low

~1

QGB

High temp. diffusional creep >~0.7Tm (Nabarro-Herring) Low temp. diffusional creep ~0.4 to (Coble creep)

~0.7Tm

*** Qcore is the activation energy for self-diffusion along dislocation cores (Qcore=QGB) Table (5.1) Values of n and QC associated with dislocation and diffusional creep (pure metal) Reference: R.W.Evans et al, Introduction to Creep

5.3 Grain-Boundary Sliding And Supperplasticity Grain-boundary sliding usually does not play an important role during primary or secondary creep. However, in tertiary creep it does contribute to the initiation and propagation of intercrystalline cracks. in fact it has some contribution to creep in primary. Because of that this creep mechanism is not shown in figure (5.1), though temperature and stress conditions are similar to that of diffusion creep. There is a good relation between the activation energy of creep deformation ( mainly due to grain boundary sliding ) and of the lattice self-diffusion. Therefore, the sliding is considered as diffusion controlled process, even though the creep deformation is caused by grain boundary sliding. Another deformation process to which it contributes significantly is super plasticity; it is thought that most of the deformation in super plastic forming takes place

30

by grain-boundary sliding. The rate of diffusion-controlled grain boundary sliding can be expressed as:


∃s=C.9.exp(-Q/kT)

(5.16)3

Modifications of equation (5.16) are required if impurity or 2nd phase particles are existed in the grain boundaries and if both lattice diffusion and grain boundary diffusion are involved. The stress dependence of the power law in equation (5.16) is the same as the mechanism of diffusion creep (n =1). The accommodating processes control the grain-boundary sliding rate where the sliding surface deviates from a perfect plane. One can readily see that we cannot have a perfect plane defined by the boundaries between different grains; we cannot look separately at the sliding between two grains having a common interface. The requirements of strain compatibility are such that we have to model the interface as sinusoidal, as is depicted in Figure (5.7). The applied stress a can produce sliding only if it is coupled with diffusional flow that transports material (or vacancies) over a maximum distance of A, the wavelength of the irregularities. Figure (5.7-b) shows the same effect in a polycrystalline aggregate. The individual grain boundaries are translated by a combination of sliding and diffusional flow under the influence of the applied stress.

31

Element E

τa

Flow of matter

τa Mode 2 Mode 1

Figure (5.7) (a) Steady state grain- boundary sliding with diffusional accommodations (b) Same as in (a), in an idealized polycrystal; the dotted lines show the flow of vacancies Reference: Marc Andre et al, Mechanical Behavior of Materials

The manner in which the individual grains move and change their relative positions by sliding and diffusional accommodation is shown in Figure (5.8). The sliding of grains under the influence of 9 coupled with minor changes in shape, makes possible the sequence a-b-c, which results in a strain of 0.55; the unique features of this mechanism is that the sequence is accomplished with relatively little strain within the grains. +σ

Sliding displacement

Relative Translation of grains

+σ

Figure (5.8) Grain boundary sliding assisted by diffusion in AshbyVerrall’s model. Reference: Marc Andre et al, Mechanical Behavior of Materials

32

6. Micro-mechanical deformation mechanisms during creep (Creep rupture mechanisms)

Intragranular creep deformation

Crystallograp hic slip

Sub-grain formation deformation

Intergranular creep deformation

Grain boundary sliding

Creep cavity nucleation

Fold formation and grain boundary migration

The micro structural deformation mechanisms, which are responsible for creep rupture at elevated temperature, can be classified into two types: •

Intragranular creep deformation

•

Intergranular creep deformation.

The difference between the two deformations is that, intragranular involves creep deformation observed within single crystals and individual grains of polycrystals, while the intergranular occurring in grain boundaries of polycrystals. The mechanisms of deformation under each category are explained herein:

6.1 Intragranular Creep Deformation 6.1.1 Crystallographic Slip Plastic deformation occurs due to slip of dislocations by gliding on certain preferred slip plans. At elevated temperatures, several dislocations slip mechanisms can exist:

33

•

Single slip: Dislocations glide in one slip system. Each slip system in crystalline materials consists of a certain slip plane and a certain direction.

•

Multiple slips: Dislocations glide in several slip systems.

•

Cross slip: Dislocations change their slip systems during glide

Figure (6.1) Schematic Representation of slip bands on the surface of polycrystalline specimen Reference: IEEE Vol.42, NO. 3,1993 Diffusion of atoms at elevated temperatures facilitates dislocations movement and makes it more active. Creep deformation by slip can be directly observed by slip bands on polished surfaces of creep specimen. Figure (6.1) shows schematically the slip bands in a polycrystalline specimen. Each slip band is caused by hundreds of dislocations gliding out of the specimen surface. The slip bands spacing are usually sensitive to the stress and temperature during creep test. As the stress decreases or the temperature increases, the slip band spacing increases for most materials. At high temperatures they become wavy, discontinuous, and coarse.3

6.1.2 Sub-grain Formation The second deformation under the category of intragranular deformation is the sub-grain formation. Sub-grains usually form during the primary stage of creep. Local

34

bending occurs in single crystal or individual grains of polycrystals, because of the in homogeneity of creep deformation. Local bending further causes dislocations of one sign to line up in the way shown in figure (6.2) .Due to interaction force between dislocations, they arrange themselves by cross-slip into low-angle sub-grain boundaries. At high temperatures, dislocations climb also helps to form sub grain boundaries. The density of dislocations increases during primary creep to a level that remains constant during steady state creep; thus sub grain boundaries usually form in the primary stage of creep. Similar to the slip-band spacing, the size of sub-grain depends on creep stress and temperature. Small sub grains are produced at high stresses or low temperatures.5

Figure (6.2) Schematic of Dislocations Lining Up to Form Sub grain Boundaries Reference: IEEE Vol.42.NO.3, 1993

6.2 Intergranular Creep Deformation This type of deformation includes, grain boundary sliding, creep cavity nucleation, fold formation, and grain boundary migration

6.2.1 Grain boundary sliding

35

Grain size plays an important part in creep deformation. At low temperatures, the strength of the material increases as the grain size decreases, because grain boundaries can stop dislocation glide. However at elevated temperatures increases the grain boundaries behave differently, has a negative effect on the strength of the material. Sliding is one of the grain boundary displacement modes, i.e., grains shear relative to each other at grain boundaries. The strain introduced by grain boundary sliding can range from a few percent up to 90% of the total strain depending on the material and testing conditions. To understand grain boundary sliding, straight line can be scratched on the polished specimen surface after creep test at elevated temperatures, grain boundary sliding is shown by the shear separation of the lines where they cross the grain boundaries. Figure (6.3) explains the shearing mechanism. Grain boundary sliding is often discontinuous with time and is not uniform along grin boundaries

Figure (6.3) Schematic of grain boundary sliding, showing a scratched line Reference: IEEE Vol.42.NO.3, 1993 Large stress concentrations can be produced due to grain boundary sliding, which can be released by self- diffusion.3

6.2.2 Creep Cavity nucleation

36

In this mechanism, voids are formed at grain boundaries where the self- diffusion rate is not high enough to release the stress concentrations. Figure (6.4) shows the type of voids that can be produced. These voids rapidly weaken the material and can lead to creep rupture.

Figure (6.4) Schematic of Creep Cavity Nucleation Reference: IEEE Vol.42, NO.3, 1993

6.2.3 Fold formation and grain boundary migration The other two mechanisms for releasing stress concentration caused by grain boundary sliding are fold formation and grain boundary migration. If stress relaxation prevents creep cavity nucleation, then strains continue along the grain boundary, and fold forms by lattice slip and propagate from a triple junction of grain boundaries towards the interior of the grain (figure (6.5-a). The propagation direction is mainly parallel to the grain boundary along which sliding occurs. Grain boundary migration refers to

37

movement of the grain boundary normal to itself by short-range diffusion of point defects near the grain boundary. Grain boundaries become wavy to release the stress induced during grain boundary sliding (figure (6.5-b) Grain boundary migration is often observed during high temperature creep

(a)

(b)

Figure (6.5) Schematic of: a) Fold formation and b) Grain Boundary Migration Reference: IEEE Vol.42, NO.3, 1993

38

7. Creep Life Assessment There is a need to assess the remaining life of industrial components, which are subjected to steady load and high temperatures environment. Particularly when the components have been in service for long time. Lifetime of a component is a function of stress and temperature. The design stress is generally the relevant rupture stress factored to allow for material variability, some variation in plant operating conditions, corrosion, and weldments in structure. Many plants operate safely beyond their design life. Thus two distinct parts of service life can be defined.4 a) The original design life which can typically be 100000hr (11.4 years); b) the safe economic life. Due to the time dependent nature of materials properties at high temperatures and the fact that ultimate failure is thus implicit, consideration must be given to a ‘beyond design’ end – of- life criterion. Since the time for growth of a crack can be very short, life extension is only safe within the time scale of crack initiation unless defect growth is being monitored.8 Figure (7.1) illustrates the concept of life extension.

7.1 Current Approaches of Creep Life Assessment Two distinct approaches are currently used to assess the remaining life of a component. 1. First approach involves the acquisition and monitoring of operational parameters, the use of standard material data, and life fraction rule; 2. the second approach is based on post- service examination and testing which require direct access to the component for sampling and measurement.

39

Figure (7.1) Life extension of high temperature equipment Reference: Journal of strain analysis Vol.29 NO 3, 1994

7.1 Approach (a); operational condition monitoring Operating temperatures and pressures are monitored and recorded. A combination of the operational parameters, standard creep rupture data and inverse design stress calculations which, in association with simple damage summation rules, produce a preliminary estimate of the remaining life. Some assumptions are included in the estimation of remaining life, which lead to uncertainties of the results. Therefore this approach does not provide a realistic basis on which to base life extension or planning for component replacement.4

7.1 Approach (b); methods base on post-service examination sampling and calculation These techniques are more accurate than the operational monitoring approach since they don not rely on standard materials data. These methods position the component materials within the standard data scattered either:

40

•

by measuring its properties and hence providing a refined prediction based on stress and temperature records and damage summation rules, or

•

Direct assessment of the extend of damage experienced by the component as a result of actual service exposure.

Both destructive and non-destructive methods can be used. Component type, location of critical area, and economic factors control the method used, either destructive or nondestructive. The main methods of the second approach are: 1. Creep and rupture testing; 2. Methods based on assessment of micro structural degradation and cavitations damage; 3. Methods based on component strain measurement

7.1.1 Creep and rupture testing Several extrapolation methods have been developed to predict creep life based on tests conducted over a shorter period. There are more than 30 methods can be used in this concern. However, three most common methods are the Larson –Milller, MansonHaferd, and Sherby- Dorn.1 7.1.1.1 Larson- Miller Method: This method simply correlates the temperature T (in kelvins) with the time to failure tr, at a constant engineering stress 9. The Larson – Miller equation has the form

Τ(log t + C ) = m r

(7.1 )1

Where C is a constant that depends on the alloy, m is a parameter that depends on the stress, and rupture time. According to that equation, if C is known for a particular alloy,

41

m can be easily found in a single test. So one can find the rupture time at any temperature, as long as the same engineering stress is applied. Figure (7.2) graphically represents the equation

Figure (7.2) Relationship between time to rupture and temperature at three levels of engineering stress 9a, 9b, and 9c using Larson-Miller equation Reference: Marc Andre et al, Mechanical Behavior of Materials

7.1.1.2 Manson and Hafered Method After Larson and Miller proposed their parameter, Manson and Hafered proposed the second method of correlating the creep test data to the life of the component. Manson and Hafered apply the following equation:

log t r − log t a T −Ta

(7.2)1

Where Ta and log ta is the intercept of the family of lines. Figure (7.3) represents the Manson and Hafered equation, graphically. In that figure three stresses are shown, leading to three lines with different slops mc>mb >ma. The times tr and ta are usually expressed in hours.

42

Figure (7.3) Relationship between time rupture and temperature at three levels of stress 9a, 9b, and 9c using Manson-Haferd parameter, (9a> 9b> 9c) Reference: Marc Andre et al, Mechanical Behavior of Materials 7.1.1.3 Orr-Sherby- Dorn Method The equation representing this method is

ln tr −

Q =m kT

(7.3)1

Where Q is the activation energy of diffusion (or creep), m is the Sherby- Dorn parameter, and tr is the time to rupture. Figure (7.4) shows the graphical representation of this parameter. It differs form Larson-Miller parameter in that the isostress lines are parallel

Figure (7.4) Relationship between time to rupture and temperature at three levels of stress, 9a, 9b, and 9c using Sherby-Dorn Parameter Reference: Marc Andre et al, Mechanical Behavior of Materials

43

7.1.2 Strain Measurement Methods. In order to determine the remaining life from strain measurement, models are required which relate the strain or strain rate to the life fraction consumed. Generally and re-calling for the creep failure mechanisms, the processes that lead to failure at elevated temperature can be categorized as follows: 1. Creep strain accumulation with no significant decrease in creep strength relative to the original value 2. Structural degradation, Crystallographic Slip, Sub-grain Formation, Gain boundary sliding, giving rise to a continuous reduction in creep strength; 3. Creep cavitations; 4. Environmental attack In many circumstances, some of these processes occur simultaneously. The prevalence of any one will be determined by the initial microstructure, purity, stress, temperature, and component size.4

7.2 Mechanistic models of Creep Damage Process: These models develop equation relating creep strain or creep strain rate to the damage feature.

7.2.1 Reduction in load bearing section This process is only likely to be significant at very high temperatures in solidsolution strengthened alloys. Failure- time relationships have been derived by Hoff and predict infinite strain at failure. More realistically, necking limits the failure strain

44

7.2.2 Micro structural degradation High temperature creep resistance arises from the finely dispersed precipitates. Creep rate is primarily dependent on the interparticle spacing and the associated mechanisms by which dislocations overcome the particles. Under these conditions, it is possible to relate the creep rate,
., to the structural state via a modified Norton low


. =A1(9-90)n f (T)

(7.4 )4

Where A1 and n are material parameters, and T is temperature. 90 is regarded as a threshold stress having a value close to the Orwan stress, i.e.

90 ~αµb / λ Where µ is the shear modulus, b is the burgers vector, and

α is a particle – dislocation

interaction coefficient. By specifying a failure criteria e.g. critical level of strain, equation (7.4) provides a means of relating strain or strain rate measured to component lifetime.

7.2.3 Cavitation damage Creep cavities nucleate and grow predominantly on grain boundaries oriented normal to the maximum principle stress. Numerous controlling mechanisms have been suggested to creep cavitation- namely vacancy flow, power –law growth, and constrained cavity growth. Power- law growth and diffusion control are most likely at high strain rate or low cavity spacing. Creep rate is described as


.= B9n/(1-Ac)n

(7.5)4

Where Ac is the fraction of cavitating grain boundaries and B and n are material constants. Same here, a failure criterion must be specified

45

7.2.4 Effect of Environment Creep in aggressive environments, including air, normally leads to changes in rupture life and ductility. The low alloy ferritic steel, for example, form oxides, which spell during exposure, reducing the load bearing section and accelerating, creep. The kinetics of creep and oxidation in these materials and such that metal loss measured after creep fracture increases with decreasing temperature in iso-stress test. Consequently if the oxide is non-load bearing it is relatively simple task to predict the creep curve as a function of stress and temperature using the appropriate oxidation data. At service temperatures the oxide is often strongly adherent and is likely to bear some of the load because its inherent creep resistance is grater generally than the metal. In this case, the rate of attack is controlled no longer by inward diffusion of oxygen but by the frequency of surface cracking. The extent of oxidation is thus dependent on strain rate as well as time. Under these circumstances the rate of oxidation becomes linear with time, even though the oxidation kinetics are parabolic, and the specimen size is important in determining life expectancy.

7.3 Phenomenological models From an engineering viewpoint, the mechanics of creep damage, tertiary creep, and fracture have developed as a continuum mechanical approach. Damage is described by a generalized scalar parameter, ω , which varies from zero at the start of life to unity at failure. Both damage and strain are assumed to increase as functions of stress, σ , temperature, T, and the current state of the damage, so that for uniaxial tension the creep rate and damage rate are expressed as


= A9 9n/(1 - ω )n

(7.6)4

46

• η ω = Bσ v (1 − ω )

(7.7)4

Where A, B, n, v, and η are temperature- dependent constants. These equations can be integrated for constant stress conditions to give

ε = 1 − (1 − t t R )1 Λ εR •

ε = (ε s t R )(1 − t t R ) ω = 1 − (t − t t R )

(1 Λ )

(Λ−1)

(7.8)4

(7.9)4 (7.10)4

Where

εR

= Rupture ductility

εs

= Minimum creep rate% rupture life ( Monkman-Grant Constant)

Λ= ε R ε s In conjunction with an appropriate failure criterion (i.e., a critical level of strain equations (7.9)- (7.10) thus provide a means of relating strain or strain rate measured on component life fraction consumed. The relationship between the strain rate and the remaining life fraction is given for various ε s and ε R values in figure (7.5).

47

Figure (7.5) predicted relation between strain rate at time t and remaining life fraction for ε s and ε R . (Strain rate obtained by direct measurement or by post-exposure testing Reference: Journal of Strain Analysis Vol.29, NO.3, 1994

Note that remaining life is insensitive to

ε R through out most of the creep life, so

that ε s is the only material parameter required. Further refinement of the approach is possible if ω can also be measured. ω is analogous to Ac in equation (7.5). Ac is the number fraction of cavitated boundaries, which thus provides a practical measure of the degree of damage. The model can be further refined by generalization for mutiaxial stress rupture criterion in ε s .

48

8. Material Specifications As mentioned in the introduction part the tubes are used in very high temperature applications, so the material should resist high temperature oxidation as well as mechanical degradation. The typical specifications are summarized below.

8.1Chemical Composition: NW-

Approxi. Value in %.

Designation

no

acc.to

Trade-

DIN mark

17006

Centracero Type

2.4879 C

Si

Mn

Cr

Ni

W

%

%

%

%

%

%

0.45

1.5

1.5

28

48

5

G-NiCr28W

CA-4879

Table (8.1) Chemical composition of G-NiCr28W Reference: Supplier, Centracero, catalogue and DIN standard

8.2 Mechanical Properties at Room Temperature Tensile Strength

Elongation

N/mm2

Lo = 5d%

400

4

Table (8.2) Mechanical properties at room temperature of G-NiCr28W Reference: Supplier Catalogue

49

8.3 Physical Properties Density Shrinkage Thermal g/cm3

%

Specific

Conductivity heat W/moC

Thermal

expansion Highest

between 20oC

application

20oCJ/goC 10-6m/mmoC

temp. in air

400oC 800oC 1000oC 8.2

2.5

11.0

0.502

14.5

16

17

1150

Table (8.3) Physical properties of G-NiCr28W Reference: Supplier Catalogue

50

9. Experimental procedure Because of the alloy in question is a special one there was no enough data available on reference books or papers. So in addition to the accelerated creep test other tests such as impact test, tensile test, hardness test and micro-structural examination have been conducted to get the properties of that alloy at room temperature as well as at elevated temperature

9.1Test piece Test piece varied according to the conducted test. However all test pieces have been cut from longitudinal strip from 5 years used tube. Test piece's dimensions were adopted according to British Standard except for creep test as explained in creep test procedure (9.5.2.10.2). Figure (9.1) illustrates the various test pieces

6

(b)

(a)

R=12 1 19

6

6

(c) 100

Figure (9.1) specimen configuration for tests, (a) Charpy impact test, (b) accelerated creep test, (c) tensile test. 51

It is worth mentioning that it was not easy to manufacture the required specimen as special tools were used to machine them.

9.2 Impact test Fracture toughness of a material represents its resistance to crake propagation. Creep mechanisms, in someway, are similar to crack propagation. When small cavities link together forming internal crake. Propagation of that crack will cause rupture of the material, if reached certain value. From this point, impact test was conducted to assess the resistance of the alloy for crack propagation, particularly at elevated temperature. The test was performed as follows 1. Machining the test piece as illustrated in figure (9.1), the notch is located in the center of the test specimen. 2. The test piece was supported horizontally at two points 3. The test piece received an impact from a pendulum of specific weight on the side opposite that of the notch, figure (9.3) 4. Electric furnace was used to heat the specimen to the test temperature N.B. about 5-15 seconds time consumed to take out the specimen from the furnace and put it in the test position. The material was examined at different temperatures, from ambient to 8000C. Results are shown in figure (9.2)

52

Charpy Impact test 10

Series1

Energy Cv (J)

8 6

Expon. (Series 1)

4 2 0 0

200

400

600

800

1000

Temperature (C)

Figure (9.2) Result of Charpy impact test, conducted at different temperatures

Figure (9.3) Charpy impact testing machine and test specimen

53

9.3 Hardness test Hardness is one of the most important properties of metals. The reason is because it relates to several other properties of metal, such as strength, brittleness, and ductility. Therefore by measuring the hardness we are indirectly measuring the strength, the brittleness, and the ductility of that metal. Vickers hardness test has been conducted on two different samples. First sample was fresh samples and the second was unbroken specimen in accelerated creep test, which lasted for 55 hours, as will be explained later. Four steps were involved in Vickers hardness test Figure (9.4)

Figure (9.4) Diagram illustrates operation of Vickers hardness tester

54

1. The test piece was placed on the anvil of the tester, below a hardened steel penterator with a diamond point 2. The square diamond penetrator was slowly brought into contact with the test piece and penetrated until the contact pressure between the penetrator and test piece reached 50 kilograms 3. The penetrator was automatically retracted and the test piece showed a small, pyramidal shaped hole. 4. Test piece then removed and a microscope was used to measure the diagonal of the dent. 5. Standard conversion table then used to convert the ocular reading into Vickers hardness number. Figures (9.5) & (9.6) illustrate the test results

Hardness (DPH)

Hardness Measurement (Fresh Samples) 285 265

Series1

245 225 0

10

20

Linear (Series1)

Samples

Figure (9.5) Vickers hardness test of fresh samples

55

Hardness (DPH)

Hardness Measurment (2) 250 200

Series1

150 100 0

5

10

Point

Figure (9.6) Vickers hardness test of specimen, which subjected to accelerated creep test for 55 hours

9.4 Tensile Test For more study off the material properties at room temperature especially tensile strength, tensile test has been conducted to 10 test pieces. Result of the test is illustrated in figure (9.7).

9.4.1 Test condition •

Test procedure were according to British Standard BS 18:1987

•

Test temperature ~190C

•

Load cell Magnification X 0.4

•

Extension Magnification 25/1

56

2528 N

2752 N 2432N 1760 N

1

2

3

4

2816 N

2880N 2272 N

5

7 6

Figure (9.7) result of tensile test of some test pieces

57

Maximum loads, at which tensile strength is calculated, are as follows; Sample

1

Max. load 2816

2

3

4

5

6

7

8

9

10

2272

2880

2528

2752

2432

1760

2352

2720

2720

1.08

1.92

1.52

1.52

1.36

2.2

2.32

2.12

1.76

(N) Elongation 2.4

Table (9.1) tensile test result

9.5 Accelerated Creep Test The main object of this project is to study the creep behavior of the material used in steam and CO2 reformers so creep rupture test had to be conducted. Due to time availability, accelerated creep rupture test was the only way to achieve the aims of the project.

9.5.1 Test Technique: The used technique is one of the most widely ones for providing the engineering data which involves the tensile creep rupture test machine. The machine consists of a dead weight applied to lower system acting on the specimen. To heat the test piece a resistance furnace is used. Test temperature, load, rupture time, and rupture elongation are the parameters measured. Elongation measurements are carried out using extensometer with two legs attached to the two ends of the gauge length of the test piece. So the relative displacement of the legs is transferred out of the furnace to two transducers that record the displacement in volt. Calibration of the two transducers was done to correlate the reading into displacement in mm, as discussed in the test procedure.

58

9.5.2 Test procedure Figure (9.8) illustrates all the test procedures (1) Setting up the machine

(2) Fixing and calibrating Extensometer

(3) Design of new Grips

(4) Selecting the material of the new grips

(5) Furnace set up and specimen heating up

(6) Loading the specimen

(7.3) Selecting new material for fixing bolts

(7.4) Replacing screws of the extensometer

(7) Encountering of problem (1)

(8) Furnace re-set up and specimen heating up

Next page

(7.1) Modify specimen’s hole and fixing bolt sizes (7.2) Re- estimation of test load

59

(9) Loading the specimen (10.2) Re-modification of the specimen dimension

(10.3) Re-calculation of the test load

(10) Encountering of problem (2)

(10.1) Reforming the grips

(11) Furnace re-setting and heating up new specimen

(12) Successful completion of the test

(13) Revision of the test load

(14) test of new specimen

(15) Successful completion of the second test

Figure (9.8) Procedures of accelerated creep rupture test

60

9.5.2.1 Setting up the machine After hardly finding the machine accessories the following actions were necessary to prepare the machine for test, refer to figure (9.9) of machine general assembly.

Upper arm

M/C Base

Figure (9.9) Machine configuration of accelerated creep test 1. Leveling of machine upper arm through the counter weight, using sprit level 2. Leveling of machine base through the tuning the adjustable lower fixing bolts 3. Examine the function of micro-switches which resulted in replacing the microswitch of the clock counter

61

9.5.2.2 Fixing and calibrating the extensometer Two transducers were attached to both ends of the specimen, through two legs as indicated in figure (9.10). To convert the voltage reading recorded by the extensometer, calibration, using slip gauges, has to be done.

Electric Furnace Elongation recording unit

Furnace control unit

Pin recorder

Extensometer

Dead load (test load)

Clock counter Figure (9.10) Main parts of testing machine 62

Calibration results are illustrated in figure (9.11). Applying regression function to the calibration graph resulted in the following equation that represents the relation between elongation and voltage reading Y = 5.214034 - 0.49642 X Where Y = Elongation X = Extensometer reading in volt Calibration of Extensometer

10.5 10 9.5 9 8.5 8 7.5 7 6.5 6 5.5 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0 -1 -1 -1 -9 -9 -8 -8 -7 -7 -6 -6 -5 -5 -4 -4 -3 -3 -2 -2 -1 -1 -0 0 0. 1 1. 2 2. 3 3. 4 4. 5 5. 6 6. 7 7. 8 8. 9 9. 10 10 11 11 1 1 0 .5 .5 .5 .5 .5 .5 .5 .5 .5 .5 5 5 5 5 5 5 5 5 5 5 .5 .5

Elongation

Y = 5.214034 - 0.49642 X

Mv

Figure (9.11) Calibration of Extensometer

63

9.5.2.3 Design of New Grips The original grips of machine were not suitable to handle 12mm thickness test test piece. The slot size was less than 12mm so that the specimen could not get in the grip 10 6

Threads

Figure (9.12) Original design of grips To facilitate fabrication of the suitable grip at workshop a new simpler grip design had been decided as indicated in figure (9.13) 14 6

Original grip

Modified grip

Figure (9.13) configuration of the new grip design

64

9.5.2.4 Selection of material of the new grip Unfortunately machine manual was not available to find out the original material, which can resist temperature of 10000C. Accordingly suitable material had to be selected for this purpose. Three different materials were suggested, based on similar chemical composition and mechanical properties to the test piece. These alloys are widely used at high temperature applications (Reference Nickel Development Institute) •

35 Ni-20Cr.bal Fe alloy (ASTM B344),

•

Incoloy 825 (ASTM B 425) Chemical composition of this alloy is as follows: Ni

Cr

Fe

Mo

Cu

Ti

Al

C

42

21.5

30

3

2.2

0.9

0.1

0.03

Table (9.2) Chemical composition of Incoloy 825 Reference: NiDI (Nickel Development Institute, publication No.297 •

Inconel alloy 617 Chemical composition of this alloy is as follows: Ni

Cr

Fe

Mo

Co

Ti

Al

Si

C

52

22

1.5

9

12.5

0.3

1.2

0.5

0.07

Table (9.3) Chemical composition of Inconel 617 Reference: NiDI (Nickel Development Institute, publication No.297 Incoloy 825 was the only available material in the market at that time so it had been used for manufacturing of the new grips 9.5.2.5 Furnace set up and Specimen heading up Three thermocouples were located at three zones of the furnace, lower, middle, and top zones. An extra thermocouple has been directly attached to the test piece to

65

measure its actual temperature. The attached thermocouple to the test piece was connected to pin recorder and the temperature was continuously plotted on chart. Heating the test piece usually took 24 hours to reach the test temperature (1000 0C) in addition to one hour at least at the test temperature. 9.5.2.6 Loading the Specimen Based on the actual practical stress applied on the tube (o.7366 N/mm2), test load was decided to be same, as a trail to get some information about creep behavior. Accordingly the test load was 0.7336* 72 (cross sectional area of the test piece) = 53 N. As the load ration is 1:25, which means that 1 Newton load on the loading plate equal to 25 Newton load on the test piece so the load on the loading plate had to be 2.12 N The first suggested load was too small to figure out the creep behavior of the material. For that reason the load configuration of the first trail was as indicated in figure (9.14)

Loading configuration of first test piece 50

Load (N)

40 30 20

Encountering of problem (1)

10 0 0

100

200

300

400

500

Time (Hrs)

Figure (9.14) loading configuration of the first trail test

66

9.5.2.7 Encountering of problem (1) After increasing the load to (45) the loading plate touched the machine base and hence the clock counter stopped. By checking the test piece condition after cooling down the furnace the following have been observed. 1. Rupture of the test piece fixing bolts upper and lower ones 2. Fusion of the ruptured part of fixing bolts with the test piece holes (welded in the holes) 3. Fusion of extensometer screws 4. No effect on the test piece By analyzing this problem we can conclude the following 1. Rupture of the fixing bolts is due to the cross sectional area of the bolts was less than that of the test piece, cross sectional area of test piece was 72mm2 while the bolt’s was 28.26 mm2 2. Fusion of bolts and screws can be explained ad the material was not suitable for 10000C applications

• Action Taken 1. Dismantle the test piece with grips 2. Drilling the fused bolts, as hammering and knocking did not succeeded to take out the bolts from the test piece. Drilling operation led to distortion in holes size the matter, which led to increasing the hole size of extensometer screws.

67

Counter measures of Problem (1) 9.5.2.7.1. Modify Specimen’s hole and fixing bolt sizes To avoid repetition of bolt rupture due to smaller cross sectional area than that of test piece fixing hole of test piece and fixing bolt sizes were increased from 6mm to 10mm. Taking into consideration that this increasing will not affect the fixing area of the test piece as the cress sectional area at the point of fixation was still bigger than the cross sectional area of the test piece itself.

9.5.2.7.2 Re-estimation of the Test Load The only benefit we got from the first trail test is the approximation of the load at which problems can take place. To save time and to accelerate the test, the load that would break the test piece within 5 days was calculated; using the Larson Miller parameter as follows, refer to figure (9.15)

Larson Miller Parameter [P=T (15+logt)10-

Figure (9.15) Larson Miller Parameter of G-NiCr28W Reference: Supplier catalogue 68

P= T (K) (15 + log t)10-3 P =1273 (15 + 2.079) 10-3 P = 21.74 From the graph the stress would be approximately 27 N/mm2 so the load would be 1944 N and according to the machine ratio 77.76 N would be loaded on the loading plate.

9.5.2.7.3 Selecting of New Material for Fixing Bolts As mentioned in the fusion analysis of the fixing bolt a suitable material for 10000C application had to be selected. Based on the re-estimation of the test load two materials were proposed, in order, Stainless steel (SS) 310 and SS309. By comparing the creep rate and rupture time of both materials at the test load SS 310 was selected as it has better creep resistance (longer rupture time). Figures (9.16) illustrates the properties of SS310 and SS309 at different stress and temperatures

69

Figure (9.16) Properties of SS309 and 310 at elevated temperatures. Reference: NiDI Publication No.2980 For further simplification and facilitate machining, fixation was changed from bolts to pins

9.5.2.7.4 Replacing screws of the extensometer Screws of the extensometer had to be replaced by another ones that can resist high temperature applications. There was no stock of stainless steel screws at machine shop and only SS316 screws were only available in the market. SS316 has inferior properties at elevated temperature compared with SS310. However it has been used because the 70

screws were no subjected to any stress, their main functions are to hold the extensometer legs at its position.

9.5.2.8 Furnace Re-set up and Test piece Heating up Light up the furnace re-started after taking all the counter measures related to problem (1). Heating up regime was the same as the first trial test.

9.5.2.9 Loading the Test piece During heating up of the test piece just only 7 Newton load (10% of the test load) was loaded on the loading plate for keeping the machine parts straight. 77.76 Newton was loaded after heating up and socking of the test piece

9.5.2.10 Encountering of Problem (2) After 55.2 hours problem No (2) was encountered which is illustrated in figure (9.17) and summarized as follows: 1. Shearing of the new fixing pins 2. Deformation of the grips

71

Shear of pins

Grip deformation

Figure (9.17) configuration of problem (2)

However the 55.2 hours test reveled some information about the creep behavior of the material, which is illustrated in figure (9.18)

Creep behaviour of G-NiCr28W(sample#1)

True Strain (%)

0.07 0.06 0.05 0.04 0.03 0.02 0.01 0 0.0

10.0

20.0

30.0

40.0

50.0

60.0

Time (Hrs)

Figure (9.18) Creep behavior of sample #1 (unbroken sample)

72

• Counter measures of problem (2) It was concluded that the material is quiet strong so to go through the test successfully permanent counter measures should be taken as follows:

9.5.2.10.1 Reforming the grips It was not so easy to order new material for the grip in addition to the long time consuming in fabrication of that grips so as a trouble shouting the grips had heated up the reformed again to its original configuration as much as possible.

9.5.2.10.2 Re-modification of the test piece dimensions In order to avoid shearing of the fixing pins cross sectional area of the test piece had to be reduced so much to be much weaker than that of the fixing bin. The cross sectional area had been reduced from 72mm2 to 24mm2. Figure (9.19) shows the final configuration of the test piece 10φ

Figure (9.19) Final modified dimension of test piece 9.5.2.10.3 Re-calculation of the test load Based on the final dimension of the test piece, the test load had to be recalculated. Referring to the previously mentioned procedure of calculating the test load (part 9.5.2.7.2) The final test load was 648 Newton (26 Newton according to machine ratio). 73

9.5.2.11

Furnace setting and test piece heating up

Same procedure of heating up was applied to the 2nd test piece. (3 Newton were load on the loading plate during heating of the test piece.

9.5.2.12

Successful Completion of the test

After 63 hours the test of the 2nd piece has successfully completed. The creep behavior pattern is illustrated in figure (9.20)

Creep Behaviour of G-NiCr28W(sample#2)

0.14

Primary creep

Tertiary creep

True Strain (%)

0.12 Secondary creep

0.1 0.08

Test piece rupture

0.06 0.04 0.02

Time (Hrs)

0 0.0

10.0

20.0

30.0

40.0

50.0

60.0

70.0

Figure (9.20) Pattern of creep behavior based on complete successful test, actual test temperature is 990oC 9.5.2.13

Revision of the Test Load

After successful completion of the test of the 2nd test piece, test load of the 3rd test piece was revised from 650 Newton to 600 Newton to prolongate the duration of the test and obtain reliable data about the creep behavior.

74

9.5.2.14

Test of the 3rd Test piece

Same procedure of furnace setting, heating up, and loading of the previous test pieces have been applied to the third one

9.5.2.15

Successful Completion of the Third Test

Revision of the test load has resulted in prolongation of the test duration to 189.4 hours. The pattern of creep behavior is illustrated in figure (9.21).

Creep Behaviour of G-NiCr28W (sample #3) 0.15 0.14 0.13

Primary Creep

0.12

True Strain (%)

0.11

Tertiary Creep Secondary Creep

0.1 0.09 0.08 0.07 0.06

Material Rupture

0.05 0.04 0.03 0.02 0.01 0 0

10

20

30

40

50

60

70

80

90

100

110

120

130

140

150

160

170

180

190

Time (Hrs)

Figure (9.21) Pattern of creep behavior of the 3rd test piece, actual test temperature is 9900C

9.6 Microstructure Examination Microstructure of three different samples was examined to find out the effect of temperature and stress. The three samples are: 1. Fresh sample, sample from tube, 2. Sample from the test piece, which lasted for 55 Hrs without broken

75

3. Sample from the broken test piece in a complete successful creep test of 63 hours.

Procedure of the test is as follows: 1. Sectioning of the samples, using Sparactron machine 2. Polishing of the samples to 1 µm diameter paste finish 3. No etching has been used. 4. Optical photomicroscope was used with two different magnifications, X120 and X240

MAG X 120

MAG X 240

Figure (9.22) Microstructure examination of fresh samples

76

MAG X120

MAG X240

Figure (9.23) Microstructure examination of unbroken sample after 55 hrs on test

MAG X120

MAG X240

Figure (9.24) Microstructure examination of broken samples after 63 hrs on test

77

10. Analysis of the Test Results In this section the results of the various test will be analyised as much as possible. However the accelerated creep test will be highlighted, as it is the main aim of this project.

10.1 Charpy Impact Test From figure (9.2) we can conclude the following: a. Transition temperature, at which great change in the energy absorption takes place, is not clear within the conducted temperature range. b. There is sensible energy increase at temperature above 250oC. However we cannot say it is the transient temperature. c. Generally it is clear that energy absorption increases with temperature increase, which means that toughness of the alloy increases with increasing the temperature. Such improvement of toughness is an indication of suitability of such alloy for high temperature applications

10.2 Hardness Test Processing of the results, indicated in figures (9.5) of fresh samples, by considering acceptable limits of results based on X ±σ where σ is the standard deviation, has reveled the following 1. Ignoring hardness values less than 245 DHP and higher than 272 DHP 2. Average value of hardness is 258 DHP Same concept was applies for the samples that taken from the unbroken test piece subjected to 55 Hrs accelerated creep test figure (9.6). Average value of that samples is 205 DHP.

78

It is clear that hardness decreases after subjecting the test piece to heat for sometimes. Decreasing the hardness increases the toughness of the alloy and hence the resistance of the alloy fore crack propagation. This is another prove of the suitability of such alloy for high temperature applications and underlines the result of charpy impact test, in the way that when the alloy subjects to high temperature its hardness decreases and accordingly the strength. Decreasing the strength improve the toughness.

10.3 Tensile Test The only property needed from that test was the tensile strength of the alloy in question. Results of the test as indicated in table (9.1) shows some variation of maximum load to break the test piece. By ignoring the shaded values, which are the lowest and highest values of either load or extension we obtain an average value of max. Load as

2628 Newton. Thus Tensile strength = Max Load/ original cross sectional area

= 2628/6 = 438 N/mm2 This figure is around the specification of that alloy which mentioned in table (8.2)

10.4 Accelerated Creep Test; The result of the accelerated creep test will be analyised from the following viewpoints: •

Creep Curve Pattern

•

Creep rate

79

•

Effect of temperature on creep behavior

•

Effect of stress on creep behavior

•

Creep deformation mechanism

•

Life time assessment

10.4.1 Creep Curve Pattern The resulted creep curves, figure (9.20) and (9.21 match completely with the standard creep curve illustrated in figure (3.1). Each of the curves of the completed tests has three distinct stages that characterize any standard creep curve; Stage (1), primary creep, which includes elastic and plastic deformations continued up to strain of 0.01 and 0.02, in figure (9.20) and (9.21) respectively. The difference may be explained due to different applied load and stress. Primary creep lasted from few hours at higher stress and for 27 hours at the decreased stress test. Figure (10.1) illustrate the primary creep in both tests Creep Behaviour of G-NiCr28W(sample#2)

Creep Behaviour of G-NiCr28W (sample #3)

0.01

Primary creep

Primary creep

True Strain (%)

True Strain (%)

0.02

0.01

Time (Hrs)

0 0.0

1.0

2.0

3.0

4.0

5.0

0 0

10

20

30

40

Time (Hrs)

(a)

(b)

Figure (10.1) primary creep pattern; (a) higher stress test (27 N/mm2), (b) decreased stress test (25N/mm2)

80

10.4.2 Creep rate Stage (2) in both curves of figure (9.20) and (9.21) lasted for 44.3 and 95.5 hours respectively. The slop of the secondary creep represents the creep rate, which is as follows:


∃s = 0.0012% /Hr for the higher stress test (27 N/mm2), figure (9.20)
∃s = 0.000494% /Hr for the decreased stress test (25 N/mm2), figure (9.21) ∃

n

This result matches with the equation (3.2) and (3.4) in which
s α σ

10.4.3 Effect of Temperature on creep behavior In fact there was no chance to examine the effect of different temperature on the pattern of creep behavior. The test has been conduct at only one temperature 980 0C ± 10 deg .

10.4.4 Effect of stress on creep behavior By comparing the curves of figures (9.20) and (9.21) we can see clearly that when the load decreased from 650 Newton (27 N/mm2) to 600 Newton (25 N/mm2) the creep rate of secondary creep decreased to more than a half of its value and hence the secondary creep duration increased to more than a double. Meanwhile the fracture time was three times of the higher stress test.

10.4.5 Creep mechanism To classify the creep mechanism according to the simplified creep deformation map, figure (5.1) we have to determine the shear modulus of the alloy. Shear modulus of the main components is illustrated in table (10.1)

81

Element

Ni

Cr

W

Mn

Iron

C

Si

Shear

76

115.4

160.6

58.05***

81.6

---

---

modulus (G) GPa *** Calculated from the equation G =

E 2(1 + ν )

Where G: shear modulus E: Young’s modulus

ν : Poisson’s ratio Table (10.1) shear modulus of alloy main components Reference: Richard W.Hertzberg, Deformation and Fracture Mechanics of Engineering Materials According to the chemical composition of the alloy in question, the calculated shear modulus is 90.38 GPa. So to classify the mechanism two ratios have to be calculated,

σ

27 −3 *10 = 2.88*10-4 for the higher stress test G 90.38

σ

=

25 −3 *10 = 2.77*10-4 for the decreased stress test G 90.38 =

The second ratio is

T

T

= m

990 = 0.43 2300

Tm is estimated to 23000C, based on the stated operating temperature of the alloy specification, 11500C, and the chemical composition relative to the melting point of the iron and stainless steel. We have to emphasis once again there is no enough published information about this alloy.

82

It is now possible to classify the creep deformation mechanism of this alloy. By investigating the deformation mechanism map, the creep mechanism controlling the behavior of creep is the grain boundary diffusion creep (Coble creep), where T