ISF Aachen Welding Technology Part II

1. Weldability of Metals

1. Weldability of Metals

4

DIN 8580 and DIN 8595 classify welding into production technique main group 4 "Joining“, group 3.6 "Joining by welding“, Figure 1.1.

Production Techniques DIN 8580

Main group 2 Deforming

Main group 1 Forming

Group 4.1 Assembling

Group 4.2 Filling

Group 4.3 Pressing

Main group 3 Separating

Group 4.4 Joining by forming

Main group 4 Joining DIN 8593

Group 4.5 Joining by deforming

Main group 5 Plating

Group 4.6 Joining by welding

Sub-group 4.6.1 Pressure welding

Main group 6 Changing material characteristics

Group 4.7 Joining by soldering

Group 4.8 Bonding

Sub-group 4.6.2 Fusion welding

© ISF 2002

br-er01-01-E.cdr

Classification of Production Techniques to DIN 8580

Figure 1.1

Material Welding suitability

Weldability of a component is determined by three outer features according to DIN 8528, Part 1. This also indicates whether a

in De g saf sig ety n

ility sib os g p ture ldin fac We anu M

Figure 1.2.

Weldability of a component

We ld

given joining job can be done by welding,

br-er01-02-E.cdr

© ISF 2002

Influencing Factors on Weldability to DIN 8528 Part 1

Figure 1.2

1. Weldability of Metals

5

Material influence on weldability, i.e. welding suitability, can be detailed for a better

understanding

in

three subdefinitions, Figure 1.3. The chemical composition of a material and also its metallurgical

properties

are

mainly set during its production, Figure 1.4. They have a very strong influence on the

Figure 1.3

physical characteristics of the material. Process steps on steel manufacturing, shown in Figure 1.4, are the essential steps on the way to a processible and usable material. During manufacture, the requested chemical composition (e.g. by alloying) and metallurgiBlast furnace: Reduction of ore to raw iron Intake of C, S, and P

Top-blow (BOF)-, bottom blow (OBM)-, stirrerconverter

Converter: Removal of C and P through oxygen and CaO

cal properties (e.g. type of teeming) of the steel are obtained. Another modification of the material behaviour takes place during subsequent treatment, where the raw material is rolled to processible

Injection of solid material or feeding cored wires

Ladle treatment: Alloying and vacuum degassing (removal of N2, H2, CO/CO2) Ladle treatment electrically heated

semi-finished goods, e.g. like strips, plates, bars, profiles, etc.. With the rolling process, material-typical

transformation

processes,

hardening and precipitation processes are used to adjust an optimised material characContinuous casting: casting of billets, blooms, slabs br-er01-04-E.cdr

© ISF 2002

Important Process Steps During Steel Production

Figure 1.4

teristics (see chapter 2).

1. Weldability of Metals

6

A survey from quality point of view about the influence of the most important alloy elements to some mechanical and metallurgical properties is shown in Figure 1.5.

C

Si

Mn

P

S

O

Cr

Ni

Al

Tensile strength

+

+

+

+

(-)

+

+

+

+

Hardness

+

+

+

+

+

+

+

Charpy-V-toughness

-

-

+

-

-

(-)

++

+

+

-

-

-

--

Hot cracking Creep resistance

+(-400°C)

(+)

Critical cooling rate

-

-

-

Formation of seggregations

+

++

++

Formation of inclusions

++ (+)

(-)

++

+

+ + with Mn with S

+ Increase of property ++ Strong increase of property

+

--

+ with Al

+

Decrease of property Strong decrease of property © ISF 2002

br-er01-05-E.cdr

Influence of Alloy Elements on Some Steel Properties

Figure 1.5

Figure 1.6 depicts the decisive importance of the carbon content to suitability of fusion welding of mild steels. A guide number of flawless fusion weldability is a carbon content of C < 0,22 %. with

C-content (%) (Melt analysis)

Fusion weldability

S185 (St 33) [EN 10 025]

unlimited (up to 0,30)

Not guaranteed, however mostly no problem with low C-content

S250GT (St 34), S235JR (St 37), S275JR (St 42) [EN 10 025] L235GT (St 35), L275GT (St 45) [Steels for tubing EN 10 208] P235GH (H I), P265GH (H II), P285NH (H III) [Steels for pressure vessel construction EN10 028] C10 (C 10), C15 (C 15), C22 (C 22) [Case hardening and tempering steels EN 10 083]

up to 0,21

up to 0,22% C: good weldable (exception: plate thickness 2 6 with D £ 50 12 with 50 < D £ 168,3 ³ L S + 60 ³ 25

mens are used. Figure 10.1 shows both standard specimen shapes for that test. A specimen is ruptured by a test machine while the actual force and the elongation of the

1

d1

d

S

S

) for pressure welding and beam welding, L S = 0. 2 ) for some other metallic materials (e.g.aluminium, copper and their alloys) __ L c ³ L S +100 may be required

r

is typical for this test, Figure 10.2.

L0 = measurement length (L0 = k ÖS0 with k = 5,65) Lt = total length S0 = initial cross-section within test length

br-er10-01.cdr

ment values, tension σ and strain ε are calculated. If σ is plotted over ε, the drawn diagram

LO LC Lt d = specimen diameter d1 = head diameter depending on clamping device LC = test length = L0 + d/2 r = 2 mm

specimen is measured. With these measure-

Normally, if a steel with a bcc lattice structure © ISF 2002

Flat and Round Tensile Test Specimen to EN 895, EN 876, and EN 10 002

is tested, a curve with a clear yield point is obtained (upper picture). Steels with a fcc lattice structure show a curve without yield

Figure 10.1

point. The most important characteristic values

s

which are determined by this test are: yield stress ReL, tensile strength Rm, and elongation

Rm ReH Rel sf

A. To determine the deformability of a weld, a e

ALud Ag

bending test to DIN EN 910 is used, Figure

A

10.3. In this test, the specimen is put onto two

s

supporting rollers and a former is pressed

Rm RP0,2 RP0,01 sf

through between the rollers. The distance of the supporting rollers is Lf = d + 3a (former diameter + three times specimen thickness). e

0,2 % 0,01 % Ag

is observed. If a surface crack develops, the

A br-er10-02.cdr

© ISF 2002

Stress-Strain Diagram With and Without Distinct Yield Point

Figure 10.2

The backside of the specimen (tension side) test will be stopped and the angle to which the specimen could be bent is measured. The

10. Testing of Welded Joints

127

test result is the bending angle and the diameter of the used former. A bending angle of 180° is reached, if the specimen is pressed through the supporting rollers without development of a crack. In Figure 10.3 specimen shapes of this test are shown. Depending on the direction the weld is bent, one distinguishes (from top to bottom) transverse, side, and longitudinal bending specimen. The tension side of all three specimen types is machined to eliminate any influences through Specimen

on

the

notch

test

effects.

thickness

of

transverse and longitudinal specimens

is

thickness.

Side

the

plate

bending

specimens are normally only used with very thick plates, here the specimen thickness is fixed at 10 mm. A

determination

of

the

toughness of a material or

Figure 10.3

welded joint is carried out with the notched bar impact test. A cuboid specimen with a V-notch is placed on a support and then hit by a pendulum ram of the impact testing machine (with very tough materials, the specimen will be bent and drawn through the supports). The used energy is measured.

Figure

10.4

represents sample shape, notch

shape

(Iso-V-

specimen), and a schematic presentation of test results.

Figure 10.4

10. Testing of Welded Joints

128 Three specimens are tested at each test tem-

b

Designation

VWS a/b

Dicke

a

RL

VWS a/b (fusion weld)

Fusion line/bonding zone

perature, and the average values as well as b

Weld centre

Designation

RL

the range of scatter are entered on the impact

a

Dicke

b

b

energy-temperature diagram (AV-T curve). VWT 0/b

VHT 0/b

This graph is divided into an area of high im-

a

b

b

pact energy values, a transition range, and an VHT a/b a

area of low values. A transition temperature is

VWT 0/b

b

a

b

VWT a/b

VHT a/b

b

b

VWT a/b

drop of toughness values. When the tempera-

a RL

RL

ture falls below this transition temperature, a

VHT a/b

transition of tough to brittle fracture behaviour

a RL

a RL

V = Charpy-V notch W = notch in weld metal; reference line is centre line of weld H = notch in heat affected zone; reference line is fusion line or bonding zone (notch should be in heat affected zone) S = notched area parallel to surface T = notch through thickness a = distance of notch centre from reference line (if a is on centre line of weld, a = 0 and should be marked) b = distance between top side of welded joint and nearest surface of the specimen (if b is on the weld surface, then b = 0 and should be marked) br-er10-05.cdr

assigned to the transition range, i.e. the rapid

takes place. As this steep drop mostly extends across a certain area, a precise assignment of transi© ISF 2002

Position of Charpy-V Impact Test Specimen in Welded Joints to EN 875

tion temperature cannot be carried out. Following DIN 50 115, three definitions of the transition temperature are useful, i.e. to fix TÜ

Figure 10.5

to:

1.) a temperature where the level of impact values is half of the level of the high range, 2.) a temperature, where the fracture area of the specimen shows still 50% of tough fracture behaviour 3.) a temperature with an impact energy value of 27 J. Figure 10.5 illustrates a specimen position and notch position related to the weld according to DIN EN 875. By modifying the notch position, the impact energy of the individual areas like HAZ, fusion line, weld metal, and base metal can be determined in a relatively accurate way. Figure 10.6 presents the influence of various alloy elements on the AV-T - curve. Three basically different influences can be seen. Increasing manganese contents increase the impact values in the area of the high level and move the transition temperature to lower values. The values of the low levels remain unchanged, thus the steepness of the drop becomes clearer with increasing Mn-content. Carbon acts exactly in the opposite way. An increasing carbon content increases the transition temperature and lowers the values of the high level, the steel becomes more brittle. Nickel decreases slightly the values of the high level, but increases the

10. Testing of Welded Joints

129 values of the low level with increasing con-

specimen position: core longitudinal

J

tent. Starting with a certain Nickel content

specimen shape: ISO V

(depends also from other alloy elements), a

300 2% Mn

steep drop does not happen, even at lowest

1% Mn

200

0,5% Mn

temperature the steel shows a tough fracture

Charpy impact energy AV

100

behaviour. 0% Mn

27 200

In Figure 10.7, the AV-T – curves of some

J 100

27

13% Ni 8,5% 5% 3,5%

2% Ni

commonly used steels are collected. These

0% Ni

curves are marked with points for impact en-

200

ergy values of AV = 27 J as well as with points

0,1% C

J

where the level of impact energy has fallen to

100 0,4% C

half of the high level. It can clearly be seen

0,8% C

27 -150

-100

-50 0 Temperature

50

°C 100 © ISF 2002

br-er10-06.cdr

Influence of Mn, Ni, and C on the Av-T-Curve

that mild steels have the lowest impact energy values together with the highest transition temperature. The development of finegrain structural steels resulted in a clear im-

Figure 10.6

provement of impact energy values and in

addition, the application of such steels could be extended to a considerably lower temperature range. With the example of the steels St E 355 and St E 690 it is clearly visible that an increase of strength goes mostly hand in hand with a decrease of the impact energy

level.

provement

Another showed

imthe

application of a thermomechanical

treatment

(con-

trolled rolling during heat treatment). The application of this treatment resulted in an increase of strength and

Figure 10.7

10. Testing of Welded Joints

130

impact energy values together with a parallel saving of alloy elements. To make a comparison, the AV-T - curve of the cryogenic and high alloyed steel X8Ni9 was plotted onto the diagram. The material is tested under very high P

C

growth and fracture mechanisms.

1,2h ± 0,25

there are no reliable findings about crack

0,55h ± 0,25

C

test speed in the impact energy test, thus

P a

b

CT - specimen

L h 1,25h ± 0,13

Figure 10.8 shows two commonly used

specimen height h = 2b ± 0,25 specimen width b total crack length a = (0,50 ± 0,05)h test load P

specimen shapes for a fracture mechanics

a

h

test to determine crack initiation and crack growth. The lower figure to the right shows a

2,1h

2,1h

b

S

possibility how to observe a crack propagation in a compact tensile specimen. During

SENB -specimen 3PB

specimen width b

bearing distance S = 4h

sample height h = 2b ± 0,05

total crack length a = (0,50 ± 0,05)h

F,U

crack initiation

U F

the test, a current I flows through the speci-

UE,aE U

men, and the tension drop above the notch is

UO V

measured. V

br-er10-08.cdr

© ISF 2002

Fracture Mechanics Test Sample Shape and Evaluation

As soon as a crack propagates through the material, the current conveying cross section Figure 10.8

decreases, resulting in an increased voltage

drop. Below to the left a measurement graph of such a test is shown. If the force F is plotted across the widening V, the drawn curve does not indicate precisely the crack initiation. Analogous to the stress-

F

F

strain diagram, a decrease of force is caused by a reduction of

the stressed

h

cross-section. If the voltage drop is plotted over the

d

force, then the start of

d

d1

2

crack initiation can be determined with suitable accuracy,

and

the

crack

br-er-10-09.cdr

Hardness Testing to Brinell and Vickers

propagation can be observed.

Figure 10.9

10. Testing of Welded Joints

131

Another typical characteristic of material behaviour is the hardness of the workpiece. Figure 10.9 shows hardness test methods to Brinell (standardised to DIN 50 351) and Vickers (DIN 50 133). When testing to Brinell, a steel ball is pressed with a known load to the surface of the tested workpiece. The diameter of the resulting impression is measured and is a magnitude of hardness. The hardness value is calculated from test load, ball diameter, and diameter of rim of the impression (you find the formulas in the standards). The hardness information contains in addition to the hardness magnitude the ball diameter in mm, applied load in kp and time of influence of the test load in s. This information is not required for a ball diameter of 10 mm, a test load of 3000 kp (29420 N), and a time of influence of 10 to 15 s. This hardness test method may be used only 3 6

2

7

10

3

6

7

7

0

8,9

reference level for measurement

10

3 10

specimen surface

6

130

30 0

hardness scale

hardness scale

100

6

4 5 3 8

130 30 0

specimen surface

0,200 mm

Instead of a ball, a diamond pyramid is

1

3

100 0

Hardness testing to Vickers is analogous. This method is standardised to DIN 50133.

4 5 3 8

0,200 mm

Hardness Number).

0,200 mm

1

0,200 mm

on soft materials up to 450 BHN (Brinell

8,9

reference level for measurement

7 10

pressed into the workpiece. The lengths of the two diagonals of the impression are

Terms

Abbreviation

ball diameter = 1,5875 mm ( 1/16 inch)

-

cone angle = 120°

2

-

radius of curvature of cone tip = 0,200 mm

3

F0

test preload

4

F1

test load

5

F

total test load = F0 + F1

6

t0

penetration depth in mm under test preload F0. This defines the reference level for measurement of tb.

The impressions of the test body are always

7

t1

total penetrationn depth in mm under test load F1

8

tb

resulting penetration depth in mm, measured after release of F1 to F0

geometrically similar, so that the hardness

9

e

resulting penetration depth, expressed in units of 0,002 mm: tb / 0,002

10

HRC HRA

measured and the hardness value is calculated from their average and the test load.

1

value is normally independent from the size of the test load. In practice, there is a hard-

Rockwell hardness = 100 - e

HRB HRF

e =

Rockwell hardness = 130 - e

br-er10-10.cdr

© ISF 2002

Hardness Test to Rockwell

ness increase under a lower test load because of an increase of the elastic part of the deformation.

Figure 10.10

Hardness testing to Vickers is almost universally applicable. It covers the entire range of materials (from 3 VHN for lead up to 1500 VHN for hard metal). In addition, a hardness test can be carried out in the micro-range or with thin layers. Figure 10.10 illustrates a hardness test to Rockwell. In DIN 50103 are various methods standardised which are based on the same principle.

10. Testing of Welded Joints

132

With this method, the penetration depth of a penetrator is measured. At first, the penetrator is put on the workpiece by application of a pre-test load. The purpose is to get a firm contact between workpiece and penetrator and to compensate for possible play of the device. Then the test load is applied in a shock-free way (at least four times the pre-force) and held for a certain time. Afterwards it is released to reach minor load. The remaining penetration depth is characteristic for the hardness. If the display instrument is suitably scaled, the hardness value can be read-out directly. All hardness test methods to Rockwell use a ball (diameter 1.5875 mm, equiv. to 1/16 Inch) or a diamond sphero-conical penetrator (cone angle 120°) as the penetrating body. There are differences in size of pre- and test load, so different test methods are scaled for different hardness ranges. The most commonly used scale methods are Rockwell B and C. The most considerable advantage of these test methods compared with Vickers and Brinell are the low time duration and a possible fully-automatic measurement value recognition. The disadvantage is the reduced accuracy in contrast to the other methods. Measured hardness numbers are only comparable under identical conditions and with the same test method. A comparison of hardness values which were determined with different methods can only be carried out for similar materials. A conversion of hardness values of different methods can be carried out piston

for steel and cast steel according to a table in DIN 50150. A relation of hardness and tensile strength is also given in that table. All the hardness test methods described above require a coupon which must be taken from the

reference bar

workpiece and whose hardness is then determined in a test machine. If a workpiece on-site is to be tested, a dynamical hardness test

specimen

method will be applied. The advantage of these methods is that measurements can be taken

br-er10-11.cdr

on completed constructions with handheld

© ISF 2002

Poldi - Hammer

Figure 10.11

10. Testing of Welded Joints

133

units in any position. Figure 10.11 illustrates a hardness test using a Poldi-Hammer. With this (out of date) method, the measurement is carried out by a comparison of the workpiece hardness with a calibration piece. For this purpose a calibration bar of exactly determined hardness is inserted into the unit, which is held by a spring force play-free between a piston and a penetrator (steel ball, 10 mm diameter). The unit is put on the workpiece to be tested. By a hammerblow to the piston, the penetrator penetrates the workpiece and the calibration pin simultaneously. The size of both impressions is measured and with the known hardness of the calibration bar the hardness of the workpiece can be determined. However, there are many sources of errors with this method which may influence the test result, e.g. an inclined resting of the unit on the surface or a hammerblow which is not in line with the device axis. The major source of errors is the measurement of the ball impression on the workpiece. On one hand, the edge of the impression is often unsharp because of the great ball diameter, on the other hand the measurement of the impression using magnifying glasses is subjected to serious errors. Figure 10.12 shows a modern measurement method which works with ultrasound and combines a high flexibility with easy handling and high accuracy. Here a test tip is pressed manually against a workpiece. If a defined test load is passed, a spring mechanism inside the test tip is triggered and the measurement starts. Test force

The measurement principle is based on a measurement of damping characteristics in 5 kp

5.0

the steel. The measurement tip is excited to

kp

emit ultrasonic oscillations by a piezoelectric

4.0

crystal. The test tip (diamond pyramid) pene3.0

trates the workpiece under the test pressure 2.0

caused by the spring force. With increasing Federweg

penetration depth the damping of the ultrasonic oscillation changes and consequently the frequency. This change is measured by the device. The damping of the ultrasonic os- little work on surface preparation of specimens (test force 5 kp) - Data Logger for storage of several thousands of measurement points - interfaces for connection of computers or printers - for hardness testing on site in confined locations

br-er10-12.cdr

© ISF 2002

cillation depends directly on penetration depth thus being a measure for material hardness. The display can be calibrated for all commonly used measurement methods, a meas-

Figure 10.12

10. Testing of Welded Joints

134

urement is carried out quickly and easily. Measurements can also be carried out in confined

pulsation range (compression)

Application

Dye penetrant method

σm = σa

σm > σ a

crack is free, surface is clean

σm < σa

compression -

+ tension

Description σm = 0

σ m < σa

σm = σa

σ m > σa

spaces. This measurement method is not yet standardised.

time

crack and surface with penetrant liquid cleaned surface, dye penetrant liquid in crack

pulsation range (tension)

alternating range

all materials with surface cracks

surface with developer shows the crack by coloring

Wöhler line Magnetic particle testing

II

A workpiece is placed between the poles of a magnet or solenoid. Defective parts disturb the power flux. Iron particles are collected.

I

III

σD

Stress σ

failure line

Surface cracks and cracks up to 4 mm below surface. However: Only magnetizable materials and only for cracks perpendicular to power lines

0 1 10 102 103 104 105 106 Fatigue strength (endurance) number lg N

107

I area of overload with material damage II area of overload without material damage III area of load below fatigue strength limit

br-er10-13.cdr

© ISF 2002

br-er10-14.cdr

© ISF 2002

Fatigue Strength Testing

Figure 10.13

Figure 10.14

To test a workpiece under oscillating stress, the fatigue test is standardised in DIN 50100. Mostly a fatigue strength is determined by the Wöhler procedure. Here some specimens (normally 6 to 10) are exposed to an oscillating stress and the number of endured oscillations until rupture is determined (endurance number, number of cycles to failure). Depending on where the specimen is to be stressed in the range of pulsating tensile stresses, alternating stresses, or pulsating compressive stresses, the mean stress (or sub stress) of a specimen group is kept constant and the stress amplitude (or upper stress) is varied from specimen to specimen, Figure 10.13. In this way, the stress amplitude can be determined with a given medium stress (prestress) which can persist for infinite time without damage (in the test: 107 times). Test results are presented in fatigue strength diagrams (see also DIN 50 100). As an example the extended Wöhler diagram is shown in Figure 10.13. The upper line, the Wöhler line, indicates after how many cycles the specimen ruptures under tension amplitude σa. The

10. Testing of Welded Joints

135

Application

Description X-ray or isotope radiation penetrate a workpiece. The thicker the workpiece, the weaker the radiation reaching the underside.

W ire diameter

Mainly for defects with orientation in radiation direction.

Tolerated deviation

mm 3,2 2,5 2 1,6 1,25 1 0,8 0,63 0,5 0,4 0,32 0,25 0,2 0,16 0,125 0,1

¬

-

W ire number

mm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

± 0,03

± 0,02

± 0,01

± 0,005

° Abbreviation

®

W ire number to Table 1

FE 1/7

1 to 7

FE 6/12 FE 10/16 CU 1/7

¬ radiation source -

¯

CU 10/16 AL 1/7 AL 6/12

workpiece

® film (displayed in distance from workpiece) ¯ defect in radiation direction; difficult to identify (flank lack of fusion) ° defect in radiation direction; easy to identify br-er10-15.cdr

AL 10/16

© ISF 2002

50

6 to 12

50 or 25

10 to 16

50 or 25

1 to 7

CU 6/12

W ire length mm

6 to 12 10 to 16

Material groups to be tested

mild steel

iron materials

copper

copper, zink, tin and its alloys

aluminium

aluminium and its alloys

50 50 50 or 25

1 to 7

50

6 to 12

50

10 to 16

W ire material

50 or 25

br-er10-16.cdr

© ISF 2002

Determination of Picture Quality Number to DIN 54105

Non-Destructive Test Methods Radiographic Testing

Figure 10.16

Figure 10.15

damage line indicates analogously, when a Description US-head generates high-frequency sound waves, which are transferred via oil coupling to the workpiece. Sound waves are reflected on interfaces (echo).

Application Mainly for defects with an orientation transverse to sound input direction.

damage to the material starts in form of cracks. Below this line, a material damage does not occur.

Ã

Test

À

methods

described

above

require

specimens taken out of the workpiece and a

Á

partly very accurate sample preparation. A testing of completed welded constructions is

Â

impossible, because this would require a deÄ À sound head Á oil coupling  workpiece à defect Ä ultrasonic test device Å radiation pulse Æ defect echo ³ backwall echo

Å

Æ

© ISF 2002

Non-Destructive Test Methods Ultrasonic Testing II

Figure 10.17

why various non-destructive test methods were developed, which are not used to determine technological properties but test the

³

br-er10-17.cdr

struction of the workpiece. This is the reason

workpiece for defects. Figure 10.14 shows

10. Testing of Welded Joints

136

two methods to test a workpiece for surface defects. Figure 10.15 illustrates the principle of radiographic testing which allows to identify also defects in the middle of a weld. The size of the minimum detectable defects depends greatly on the intensity of radiation, which must be adapted to the thickness of the workpiece to be radiated. As the film with documented defects does not permit an estimation of the plate thickness, a scale bar must be shown for estimation of the defect size. For that purpose, a plastic template is put on the workpiece before radiation which contains metal wires with different thickness and incorporated metallic marks, Figure 10.16. The size of the thinnest recognisable wire indicates the size of the smallest visible defect. Radiation

Figure 10.18

testing provides information about the defect position in the plate plane, but not about the position within the thickness depth. A clear advantage is the good documentation ability of defects.

Figure 10.18

An information about the depth of the defect is provided by testing the workpiece with ultrasound. The principle is shown in Figures

10.17

and

10.18

(principle of a sonar). The

display

of

original

br-er10-19.cdr

pulse, backwall and defect

Ultrasonic Testing of Fillet Welds

echo is carried out with an oscilloscope.

Figure 10.19

10. Testing of Welded Joints

137

This method provides not only a perpendicular sound test, but also inaccessible regions can be tested with the use of so called angle testing heads, Figure 10.19.

Pores between 10 and 20 mm depth provide an unbroken echo sequence across the entire display starting from 10mm. The backwall echo sequence of 30 mm is not yet visible.

30

Wall thickness is below 40 mm. The roughness provides smaller and wider echos.

Echo sequence of 20 mm depth. The backwall is completely screened.

The perpendicular crack penetrating the material does not provide a display because the reflecting surface (tip of crack) is too small.

40 The oblique and rough defect from 20 to 30 mm provides a wide echo of 20 to 30 mm. Starting with SKW 4, an unbroken echo sequence follows. The inclination of the reflector is recornised by a change of the 1st echo when shifting the test head.

The oblique backwall reflects the soundwaves against the crack. this is the reason why an ‘impossible’ depth of 65 mm is displayed.

Echo sequence of 10 mm depth. The reflector in 30 mm depth is completely screened.

br-er10-20.cdr

© ISF 2002

br-er10-21.cdr

Defect Identification with Ultrasound

© ISF 2002

Defect Identification With Ultrasound

Figure .10.20

Figure 10.21

Figures 10.20 and 10.21 show macro section

schematically

the

display of various defects on an oscilloscope. A cor-

base material

50 µ

ferrite + perlite

coarse grain zone

bainite

rect interpretation of all the signals requires great experience,

2,5 mm

fine grain zone

ferrite + perlite

fusion line Steel: S355N (T StE 355) weld metal

bainite

because

the

shape of the displayed signals is often not so clear.

cast structure

br-er10-22.cdr

Metallographic Examination of a Weld

Figure 10.22 illustrates the potential of metallographic

Figure 10.22

examination. Grinding and

10. Testing of Welded Joints

138 etching with an acid makes the microstructure visible. The reason is that depending on structure and orientation, the individual grains react very differently to the acid attack thus 100

25

Fe

% Fe

% Cr

macrosection, i.e. without magnification, gives

Cr 20

60 40

15

20

10 % Ni

reflecting the light in a different way. The

80

a complete survey about the weld and fusion line, size of the HAZ, and sequence of solidification. Under adequate magnification, these

0 10

areas can still not be distinguished precisely,

8

Ni

however, an assessment of the developed

6 4

5

microstructure is possible.

2 0

0 200

mm

100

0

An assessment of the distribution of alloy

100

Distance from fusion line br-er10-23.cdr

elements across the welded joint can be car-

© ISF 2002

Micro-Analysis of the Transition Zone Base Material - Strip Cladding

ried out by the electron beam micro-analysis. An example of such an analysis is shown in

Figure 10.23

Figure 10.23. If a solid body is exposed to a

focused electron beam of high energy, its atoms are excited to radiate X-rays. There is a simple relation between the wave length of this radiation and the atomic number of the chemical elements. As the intensity of the radiation depends on the concentration of the elements, the chemical composition of the solid body can be concluded from a survey of the emitted

X-ray

qualitatively

and

spectrum quantita-

tively. A detection limit is

50

50

50

20 20

1. weld

about 0.01 mass % with this

50

20 20

2. weld 0 10

method. Microstructure areas of a minimum diameter

weld

of about 5 µm can be ana-

axis of bending former

weld

Agents: - electrolytic copper in the form of chips (min. 50 g/l test solution) - 100 ml H2SO4 diluted with 1 l water and then . 110 g CuSO 5 H2O are added

lysed. If the electron beam is

Test: The specimens remain for 15 h in the boiling test solution. Then the specimens are bent across a former up to an angle of 90° and finally examined for grain failure under a 6 to 10 times magnification.

moved across the specimen (or the specimen under the br-er-10-24.cdr

beam), the element distribu-

Strauß - Test

tion along a line across the Figure 10.24

axis of bending former

10. Testing of Welded Joints

139

solid body can be determined. Figure 10.23 presents the distribution of Ni, Cr, and Fe in the transition zone of an austenitic plating in a ferritic base metal. The upper part shows the related microsection which belongs to the analysed part. This microanalysis was carried out along a straight line between two impressions of a Vickers hardness test. The impressions are also used as a mark to identify precisely the area to be analysed. The so called Strauß test is 12

standardised in DIN 50 914. it serves to determine

80

web

the resistance of a weld

measurement points

tack welds

against intergranular corro-

base plate weld1

40

40

20

sion. Figure 10.24 shows the specimen shape which

a

a a

20

aa

a

a

12

weld2

is normally used for that

120

80

aa

test. In addition, some debr-er-10-25.cdr

Test of Crack Susceptibility of Welding Filler Materials to DIN 50129

tails of the test method are explained.

Figure 10.25

Figure 10.25 presents a specimen shape for testing the crack susceptibility of welding consumables. For this test, weld number 1 is welded first. The 2. weld is welded not later than 20 s in reversed direction after completion of the first weld. Throat thickness of weld 2 must be 20% below of weld 1. After cooling down, the beads are examined for cracks. If tensioning bolt hexagon nut min. M12 DIN 934

guidance plates

a tensioning plate specimen base body

cracks are found in weld 1, the test is void. If weld 1 is free from cracks, weld 2 is examined for crack with magnifying glasses. Then weld 1 is machined off and weld 2 is cracked by bend-

br-er-10-26.cdr

Tensioning Specimen for Crack Susceptibility Test

Figure 10.26

ing the weld from the root. Test results record any

10. Testing of Welded Joints

140

surface and root cracks together with information about position, orientation, number, and length. The welding consumable is regarded as 'non-crack-susceptible' if the welds of this test are free from cracks. Figure 10.26 presents two proposals for self-stressing specimens for plate tests regarding their hot crack tendency. Such tests are not yet standardised to DIN.

thermo couple electrode

cross-section

groove shape 60°

60°

welding direction

weld metal support plate

Wd./2 H

Wd.

2

implant

Hc

Wd./2

2 load temperature in °C

specimen shape

load in N

Tmax start

end crater

150

crack coefficient

C=

c

x 100 (in %)

800 500

1

2 3

4 5

sections 60 anchor weld

80 test weld

150 100 60 anchor weld

br-er10-27.cdr

t8/5

© ISF 2002

rupture time

br-er10-28.cdr

Tekken Test

Figure 10.27

time in s

© ISF 2002

Implant Test

Figure 10.28

There are various tests to examine a cold crack tendency of welded joints. The most important ones are the self-stressing Tekken test and the Implant test where the stress comes from an external source. In the Tekken test which is standardised in Japan, two plates are coupled with anchor joints at the ends as a step in joint preparation see Figure 10.27. Then a test bead is welded along the centre line. After storing the specimen for 48 hours, it is examined for surface cracks. For a more precise examination, various transverse sections are planned. The value to be determined is the minimum working temperature at which cracks no longer occur. The specimen shape simulates the conditions during welding of a root pass.

10. Testing of Welded Joints

141

The most commonly used cold crack test is the Implant test, Figure 10.28. A cylindrical body (Implant) is inserted into the bore hole of a support plate and fixed by a surface bead. After the bead has cooled down to 150°C the implant is exposed to a constant load. The time is measured until a rupture or a crack occurs (depending on test criterion 'rupture' or 'crack'). Varying the load provides the possibility to determine the stress which can be born for 16 hours without appearance of a crack or rupture. If a stress is specified to be of the size of the yield point as a requirement, a preheat temperature can be determined by varying the working temperature to the point at which cracks no longer appear. As explained in chapter 'cold cracks' the hydrogen content plays an important role for cold crack development. Figure 10.29 shows results of trials where the cold crack behaviour was examined using the Tekken and Implant test. Variables of these tests were hydrogen content of the weld metal and preheat temperature. The variation of the hydrogen content of the weld metal was carried out by different exposure to humidity (or rebaking) of the used stick electrodes. Based on the hydrogen content, the preheat temperature was increased test by test. Consequently, the curves of Figure 10.29 represent the limit curves for the related test. Specimens above these heat input: 12 kJ/cm basic coated stick electrode plate and support plate thickness: 38 mm

°C

cracks, below these curves

°C Implant-Test

150

Tekken-Test

100

50

cracks are present. Evi-

150

Rcr = Rp0,2 = 358 N/mm²

Preheat temperature

Preheat temperature

curves remain free from

fractured starting cracks crack-free

20

dent for both graphs is that with

100

temperature 50

starting cracks crack-free

20 0

10

20

30

ml/ 40 100 g

increased

0

10

Diffusible hydrogen content br-er-10-29.cdr

Test Result Comparison of Implant and Tekken Test

20

30

ml/ 40 100 g

preheat

considerably

higher hydrogen contents are tolerated without any crack

development

be-

cause of the much better hydrogen effusion.

Figure 10.29

If both graphs are compared it becomes obvious that the tests produce slightly different findings, i.e. with identical hydrogen content, the determined preheat temperatures required for the avoidance of cracking, differ by about 20°C.

10. Testing of Welded Joints

142

Figure 10.30 illustrates a method to measure the diffusible hydrogen content in welds which is standardised in DIN 8572. Figure a) shows the burette filled with mercury before a specimen is inserted. The coupons are inserted into the opened burette and drawn with a magnet through the mercury to the capillary side (density of steel is lower than that of mercury, coupons surface). Then the burette is closed and evacuated. The hydrogen, which effuses of the coupons but does not diffuse through the mercury, collects in the capillary. The samples remain in the evacuated burette 72 hours for degassing. To determine the hydrogen volume the burette is ventilated and the coupons are removed from the capillary side. The volume of the effused hydrogen can be read out from the capillary; the height difference of the two mercury menisci, the air pressure, and the temperature provide the data to calculate the

norm

volume

to pump hydrogen under reduced pressure

under

VT

air pressure B

evacuated

standard

conditions.

This

capillary side

volume and the coupons

M

meniskus1

weight are used to calculate,

meniskus2 mercury

coupons

as measured value, the hydrogen volume in ml/100 g weld metal. This is the most

a) starting condition

b) during degassing

c) ventilated after degassing

br-er-10-30.cdr

commonly used method to determine

the

Burettes for Determination of Diffusible Hydrogen Content

hydrogen

content in welded joints.

Figure 10.30

2. TTA / TTT - Diagrams

2. TTA / TTT – Diagrams

9 An essential feature of low alloyed ferrous materials is g -Iron face-centered

a -Iron body-centered

the crystallographic transformation

of

the

body-

centred cubic lattice which is stable at room temperature (α-iron, ferritic structure) to the face-centred cubic lattice (γ-iron, aus-

Lattice constant 0.364 nm at 900 °C

Lattice constant 0.286 nm at room temperature

tenitic

© ISF 2002

br-eI-02-01.cdr

structure),

2.1.

Body- and Face-Centered Lattice Structures

The

Figure

temperature,

where this transformation

Figure 2.1

occurs, is not constant but depends on factors like

alloy content, crystalline structure, tensional status, heating and cooling rate, dwell times, etc.. In order to be able to the

basic

processes it is necessary to

S

have a look at the basic

TsA T1

processes occuring in an binary

system.

Figure 2.2 shows the state

1

T2

S+ a 3 2

4 5

Temperature T

idealized

L1

L1

Li So

TsB

Temperature T

understand

a - ss

of a binary system with a

b

complete solubility in the liquid and solid state. If the melting of the L1 alloy

A (Ni)

c2

c0

c3

Concentration c

c4

B (Cu)

Time t

© ISF 2002

br-eI-02-02.cdr

Binary System With Complete Solubility in Liquid and Solid Phase

is cooling down, the first crystals of the composition

c1

Figure 2.2

c1 are formed with reaching the temperature T1. These crystals are depicted as mixed crystal α, since they consist of a compound of the components A (80%) and of B (20%). Further, a melting with the composition c0 is present at the temperature T1. With dropping temperature, the remaining melt is en-

2. TTA / TTT – Diagrams

10

riched with component B, following the course of line Li (liquidus line, up to point 4). In parallel, always new and B richer α-mixed crystals are forming along the connection line So (solidus line, points 1, 2, 5). The distribution of the components A and B in the solidified structure is homogeneous since concentration differences of the precipitated mixed crystals are balanced by diffusion processes. The other basic case of complete solubility of two components in the liquid state and of complete insolubility in the solid state shows Figure 2.3 If two components are completely insoluble in the solid state, no mixed crystal will be formed of A and B. The two liquidus lines Li cut in point e which is also designated as the eutectic point. The isotherm Te is the eutectic line. If an alloy of free composition solidifies according to Figure 2.3, the eutectic line must be cut. This is the temperature (Te) of the eutectic transformation: S → A+B (T = Te = const.). This means that the melt at a constant temperature Te dissociates in A and B. If an alloy of the composition L2 solidifies, a purely eutectic structure results. On account of the eutectic reaction, the temperature of the alloy remains constant up to the completed transformation (critical point) (Figure 2.2). Eutectic structures are normally fine-grained and show a characteristic orientation between the constituents. The alloy L1 will consist of a compound of alloy A and eutectic alloy E in the solid state. You can find further inL1

L2

L1

L2

TsA

tion behaviour in relevant

S 1

Temperature T

2

TsB

So Te

Li

Li

S+A

S+B

specialist literature.

Temperature T

2’

formation on transforma-

The definite use of the

3 4

principles occurs in the A+E

E

B+E

iron-iron carbide diagram. A br-eI-02-03.cdr

c1

ce Concentration c

B

Transformation behaviour

Time t © ISF 2002

of carbon containing iron Binary System With Complete Solubility in Liquid Phase and Complete Unsolubility in Solid Phase

Figure 2.3

in the equilibrium condition is described by the

2. TTA / TTT – Diagrams

11

stable phase diagram iron-graphite (Fe-C). In addition to the stable system Fe-C which is specific for an equilibrium-close cooling, there is a metastable phase diagram iron cementite (Fe-Fe3C). During a slow cooling, carbon precipitates as graphite in accord with the stable system Fe-C, while during accelerated cooling, what corresponds to technical conditions, carbon precipitates as cementite in agreement with the metastable system (Fe-Fe3C). Per definition, iron carbide is designated as a structure constituent with cementite although its stoichiometric composition is identical (Fe3C). By definition, cementite and graphite can be present in steel together or the cementite can decompose to iron and graphite during heat treatment of carbon rich alloys. However, it is fundamentally valid that the formation of cementite is encouraged with increasing cooling rate and decreasing carbon content. In a double diagram, the stable melt + d - solid solution

system is shown by a d-

dashed, the metastable by

solid sol.

d -+g-

melt + austenite

austenite + graphite austenite + cementite

diagram is limited by the

austenite + ferrite

formation of cementite with

ferrite

The

stoichiometry

strict of

the

stable equilibrium metastable equilibrium

Mass % of Carbon © ISF 2002

br-eI-02-04.cdr

Stable and Metastable Iron-Carbon-Diagram

formed carbide phase can be read off at the top X-

ferrite + graphite ferrite + cementite

perlite

a carbon content of 6,67 mass%.

ledeburite

phase

melt + cementite

austenite

Temperature °C

metastable

Fe3C (cementite)

solid sol.

a solid line, Figure 2.4. The

melt + graphite

melt

Figure 2.4

coordinate of the molar carbon content. In accordance with the carbon content of Fe3C, cementite is formed at a molar content of 25%. The solid solutions in the phase fields are designated by Greek characters. According to convention, the transition points of pure iron are marked with the character A - arrêt (stop point) and distinguished by subjacent indexes. If the transition points are determined by cooling curves, the character r = refroidissement is additionally used. Heat-up curves get the supplement c - chauffage. Important transition points of the commercially more important metastable phase diagram are:

-

1536 °C: solidification temperature (melting point) δ-iron,

-

1392 °C: A4- point γ- iron,

2. TTA / TTT – Diagrams -

12

911 °C: A3- point non-magnetic α- iron,

with carbon containing iron: -

723 °C: A1- point (perlite point).

The corners of the phase fields are designated by continuous roman capital letters. As mentioned before, the system iron-iron carbide is a more important phase diagram for technical use and also for welding techniques. The binary system iron-graphite can be stabilized by an addition of silicon so that a precipitation of graphite also occurs with increased solidification velocity. Especially iron cast materials solidify due to their increased silicon contents according to the stable system. In the following, the most important terms and transformations should be explained more closely as a case of the metastable system. The transformation mechanisms explained in the previous sections can be found in the binary system iron-iron carbide almost without exception. There is an eutectic transformation in point C, a peritectic one in point I, and an eutectoidic transformation in point S. With a temperature of 1147°C and a carbon concentration of 4.3 mass%, the eutectic phase called Ledeburite precipitates from cementite with 6,67% C and saturated γ-solid solutions with 2,06% C. Alloys with less than 4,3 mass% C coming from primary austenite and Ledeburite are called hypoeutectic, with more than 4,3 mass% C coming from primary austenite and Ledeburite are called hypereutectic.

If an alloy solidifies with less than 0,51 mass percent of carbon, a δ-solid solution is formed below the solidus line A-B (δ-ferrite). In accordance with the peritectic transformation at 1493°C, melt (0,51% C) and δ-ferrite (0,10% C) decompose to a γ-solid solution (austenite).

The transformation of the γ-solid solution takes place at lower temperatures. From γ-iron with C-contents below 0.8% (hypoeutectoidic alloys), a low-carbon α-iron (pre-eutectoidic ferrite) and a fine-lamellar solid solution (perlite) precipitate with falling temperature, which consists of α-solid solution and cementite. With carbon contents above 0,8% (hypereutectoidic alloys) secondary cementite and perlite are formed out of austenite. Below 723°C, tertiary cementite precipitates out of the α-iron because of falling carbon solubility.

2. TTA / TTT – Diagrams

13

The most important distinguished feature of the three described phases is their lattice structure. α- and δ-phases are cubic body-centered (CBC lattice) and γ-phase is cubic facecentered (CFC lattice), Figure 2.1. Different carbon solubility of solid solutions also results from lattice structures. The three above mentioned phases dissolve carbon interstitially, i.e. carbon is embedded between the iron atoms. Therefore, this types of solid solutions are also named interstitial solid solution. Although the cubic face-centred lattice of austenite has a higher packing density than the cubic body-centred lattice, the void is bigger to disperse the carbon atom. Hence, an about 100 times higher carbon solubility of austenite (max. 2,06% C) in comparison with the ferritic phase (max. 0,02% C for α-iron) is the result. However, diffusion speed in γ-iron is always at least 100 times slower than in α-iron because of the tighter packing of the γ-lattice.

Although α- and δ-iron show the same lattice structure and properties, there is also a difference between these phases. While γ-iron develops of a direct decomposition of the melt (S → δ), α-iron forms in the solid phase through an eutectoidic transformation of austenite (γ → α + Fe3C). For the transformation of non- and low-alloyed steels, is the transformation of δferrite of lower importance, although this δ-phase has a special importance for weldability of high alloyed steels. Unalloyed steels used in industry are multi-component systems of iron and carbon with alloying elements as manganese, chromium, nickel and silicon. Principally the equilibrium diagram Fe-C applies also to such multi-component systems. Figure 2.5 shows a

Ac3

schematic cut through the Ac1e

three phase system Fe-M-C. During precipitation, mixed carbides of the general composition M3C develop. © ISF 2002

br-eI-02-05.cdr

In contrast to the binary Description of the Terms Ac1b, Ac1e, Ac3

Figure 2.5

system Fe-C, is the three

2. TTA / TTT – Diagrams

14

phase system Fe-M-C characterised by a temperature interval in the three-phase field α + γ + M3C. The beginning of the transformation of α + M3C to γ is marked by Aclb, the end by Acle. The indices b and e mean the beginning and the end of transformation. The described equilibrium

°C

diagrams apply only to low heating and cooling rates. However, higher heating and cooling rates are present during welding, consequently other structure

s

types develop in the heat

© ISF 2002

br-eI-02-06.cdr

affected zone (HAZ) and in

TTA Diagram for Isothermal Austenitization

the weld metal. The structure transformations during

Figure 2.6

heating and cooling are described by transformation diagrams, where a temperature change is not carried out close to the equilibrium, but ASTM4; L=80µm

at different heating and/or cooling rates.

ASTM11; L=7µm

A

representation

processes

during

of

the

transformation

isothermal

austenitizing

shows Figure 2.6. This figure must be read 20µm

20µm

exclusively along the time axis! It can be recognised

that

several

transformations

during isothermal austenitizing occur with e.g. 800°C.

Inhomogeneous

austenite

means

Temperature

both, low carbon containing austenite is formed in areas, where ferrite was present before transformation, and carbon-rich austenite is formed in areas during transformation, where carbon was present before Time br-er02-07.cdr

© ISF 2002

TTA-Diagram for Continuous Warming

Figure 2.7

transformation. During sufficiently long annealing times, the concentration differences are balanced by diffusion, the border to a ho-

2. TTA / TTT – Diagrams

15

mogeneous austenite is passed. A growing of the austenite grain size (to ASTM and/or in µm) can here simultaneously be observed with longer annealing times.

The influence of heating rate on austenitizing is shown in Figure 2.7. This diagram must only be read along the sloping lines of the same heating rate. For better readability, a time pattern was added to the pattern of the heating curves. To elucidate the grain coarsening during austenitizing, two microstructure photographs are shown, both with different grain size classes to ASTM. Figure 2.8 shows the relation between the TTA and the Fe-C diagram. It's obvi-

Ac3

ous that the Fe-C diagram is only valid for infinite long

Ac1e Ac1b

dwell times and that the TTA diagram applies only for one individual alloy. Figure 2.9 shows the dif© ISF 2002

br-eI-02-08.cdr

ferent

Dependence Between TTA-Diagram and the Fe-M-C System

time-temperature

passes during austenitizing and

Figure 2.8

subsequent

cooling

down. The heating period is comAc3

posed of a continuous and

continuous

an isothermal section. Ac1e

Ac1b

During cooling down, two

isothermal

different ways of heat control can be distinguished: 1. : During continuous temperature Heating and Cooling Behaviour With Several Heat Treatments

Figure 2.9

control

a

© ISF 2002

br-eI-02-09.cdr

cooling is carried out with a constant cooling rate out of

2. TTA / TTT – Diagrams

16

the area of the homogeneous and stable austenite down to room temperature. 2. : During isothermal temperature control a quenching out of the area of the austenite is carried out into the area of the metastable austenite (and/or into the area of martensite), followed by an isothermal holding until all transformation processes are completed. After transformation will be cooled down to room temperature. Figure

2.10

shows

the

time-temperature diagram of a isothermal transformation of the mild steel Ck 45. Read such diagrams only along the time-axis! Below the Ac1b line in this figure, there is the area of the metastable austenite, marked © ISF 2002

br-eI-02-10.cdr

Isothermal TTT-Diagram of Steel C45E (Ck 45)

with

an

A.

The

areas

marked with F, P, B, und M represent areas where fer-

Figure 2.10

rite, perlite, Bainite and martensite are formed. The

lines which limit the area to the left mark the beginning of the formation of the respective structure. The lines which limit the area to the right mark the completion of the formation of the respective structure. Because the ferrite formation is followed by the perlite formation, the completion of the ferrite formation is not determined, but the start of the perlite formation. Transformations to ferrite and perlite, which are diffusion controlled, take place with elevated temperatures, as diffusion is easier. Such structures have a lower hardness and strength, but an increased toughness.

Diffusion is impeded under lower temperature, resulting in formation of bainitic and martensitic structures with hardness and strength values which are much higher than those of ferrite and perlite. The proportion of the formed martensite does not depend on time. During quenching to holding temperature, the corresponding share of martensite is spontanically formed. The present rest austenite transforms to Bainite with sufficient holding time. The right

2. TTA / TTT – Diagrams

17

detail of the figure shows the present structure components after completed transformation and the resulting hardness at room temperature. Figure 2.11 depicts the graphic representation of the TTT diagram, which is more important for welding techniques. This is the TTT diagram for continuous cooling of the steel Ck 15. The diagram must be read along the drawn cooling passes. The lines, which are limiting the individual areas, also depict the beginning and the end of the respective transformation. Close to the cooling curves, the amount of the formed structure is indicated in per cent, at the end of each curve, there is the hardness value of the structure at room temperature. Figure 2.12 shows the TTT diagram of an alloyed steel containing

approximately

the same content of carbon

27 19 40

as the steel Ck 15. Here you can see that all transformation 370

are

strongly postponed in rela-

170

235 220

processes

tion to the mild steel. A

Time © ISF 2002

br-eI-02-11.cdr

Continuous TTT-Diagram of Steel C15E (Ck 15)

completely

martensitic

transformation

is

carried

out up to a cooling time of

Figure 2.11

about 1.5 seconds, comC 0,13

Chemical composition %

Si 0,31

Al P S 0,023 0,009 0,010

Mn 0,51

1000 °C 900

Cr 1,5

Mo 0,06

Ni 1,55

V < 0,01

austenitizing temperature 870°C (dwell time 10 min) heated in 3 min

Ck 15. In addition, the

Ac3

800

completely diffusion con-

Ac1

Temperature

700 F

47

25

10 22

55 67

75 75 25 25

75

75 75

A+C

600

trolled transformation proc-

P

5

esses of the perlite area

25

500 MS

B

23

60

400

pared with 0.4 seconds of

72

55

M

37

are postponed to clearly 30

300

22 9

longer times.

2

200 100 417

400

396

314 304 287 268 251 224 192

167

152

151

0 10

-1

br-eI-02-12.cdr

10

0

10

1

10

2

10

3

104

Time

Continuous TTT-Diagram of Steel 15 CrNi 6

Figure 2.12

10

5

s

106

The hypereutectoid steel C

© ISF 2002

100

behaves

completely

different, Figure 2.13. With this carbon content, a pre-

2. TTA / TTT – Diagrams

18 eutectoid ferrite formation cannot still be car-

C Si Mn P S Cr Cu Mo Ni V 1,03 0,17 0,22 0,014 0,012 0,07 0,14 0,01 0,10 traces

Chemical composition % 1000 °C 900

austenitizing temperature 790°C dwell time 10 min, heated in 3 min

ried out (see also Figure 2.3). The term of the figures 2.9 to 2.11 "austenitiz-

800 AC1e

Temperature

700

A+C

100

100

100

100

100

100

100 AC1b

100

600 P

500

where the workpiece transforms to an austen-

2 15

400

180

300 200

M

RA»30 914 901 817 366

351

283

236

214

215

177

0 1000 °C 900

austenitizing temperature 860°C dwell time 10 min, heated in 3 min

AC1e

Temperature

700

C A

100

P

100

100

100

100

100

100

AC1b

that only martensite is formed from the austenite, provided that the cooling rate is suffi-

100

100

500

the AC3 temperature, where above it there is only pure austenite. In addition you can see

800

600

itic microstructure in the course of a heat treatment. Don’t mix up this temperature with

MS

100

ing temperature“ means the temperature,

5

194

400

ciently high, a formation of

any other

300

200

microstructure is completely depressed. With

MS RA»40

100 M

876 887 867 496 457 442

0 10-1

100

101

102

347

289

103

246

227

104

Time

br-er02-13.cdr

200

s

105

© ISF 2002

Continuous TTT-Diagram of Steel C100U (C 100 W1)

this type of transformation, the steel gains the highest hardness and strength, but loses its toughness, it embrittles. The slowest cooling rate where such a transformation happens, is

Figure 2.13

called critical cooling rate.

Ar1

Perlite

100%

Low number of nuclei due to melting, high temperature, long dwell time, coarse austenite grain, C-increase up to 0,9%, Mn, Ni, Mo, Cr

Cr, V, Mo

900°C 1300°C

°C 800

High number of nuclei, low hardening temperature, C-increase above 0,9%

Cr, V, Mo

1000

Temperature

Ar3

C, Cr, Mn, Ni, Mo, high temperature, ferrite precipitation in perlite

A F 600

P B

MS 400

Bainite

M

Ms Martensite C, Mn, Cr, Ni, Mo, V, high hardening temperature, preprecipitation in bainite

Co, Al, deformation of austenite, low hardening temperature

200 100

Structure distribution

Temperature

Low hardening temperature (special carbides), austenite above bainite

% 75

M

M

B

B

50 25 0 10-1

Transition time br-er02-14.cdr

10

102

s

103

Cooling time (A3 to 500°C) © ISF 2002

br-er02-15.cdr

Influence of Alloy Elements on Transformation Behaviour of Steels

Figure 2.14

1

© ISF 2002

Temperature Influence on Transformation Behaviour of Steels

Figure 2.15

2. TTA / TTT – Diagrams

19

Figure 2.14 shows schematically how the TTT diagram is modified by the chemical composition of the steel. The influence of an increased austenitizing temperature on transformation behaviour shows Figure 2.15. Due to the higher hardening temperature, the grain size of the austenite is higher (see Figure 2.6 and 2.7). This grain growth leads to Max. temperature 1350 °C

S355J2G3 (St 52-3) C 0,16

Chemical composition %

Si 0,47

Welding heat cycle

Mn P S Al N Cr 1,24 0,029 0,029 0,024 0,0085 0,10

Cu 0,17

Ni 0,06

900 °C 800

sion lengths which must be passed during the trans-

700

formation. As a result, the

48

Temperature

an extension of the diffu-

600 500

"noses" in the TTT diagram

B

75

55

400

222

are shifted to longer times.

215

300

The lower part of the figure

200 449

420

400

363 334 324 270

253

251 249 243

shows the proportion of

100

0

1

2

4

6

8 10

20 Time

br-eI-02-16.cdr

40

60 80 100

200

s

400

© ISF 2002

formed

martensite

and

Bainite depending on cool-

Welding TTT-Diagram of Steel S355J2G3 (St 52-3)

ing time. You can see that

Figure 2.16

with

higher

austenitizing

temperature the start of Max. temperature 1350 °C

15 Mo 3 C 0,16

Chemical composition %

Si 0,30

Bainite formation together

Welding heat cycle

Mn P S Mo 0,68 0,012 0,038 0,29

with the drop of the mart-

900 °C 800

ensite proportion is clearly

Ac3=861°C Ac1=727°C

700

Temperature

F

7

1

32 8 4

19

53

45

32

17

shifted to longer times.

P

600 500

99

B

MS 14

74

83

60

77

38

As Bainite formation is not

15

87

so much impeded by the

95

400

208

M

200

178

300

coarse austenite grain as

200 440

HV30

431

338

285

255

234

224

210

with the completely diffu-

100

0

1

2

4

6

8 10

20 Time

40

60 80 100

s

400

© ISF 2002

br-eI-02-17.cdr

Welding TTT - Diagram of Steel 15Mo3 (15 Mo 3)

Figure 2.17

200

sion controlled processes of ferrite and perlite formation, the maximum Bainite proportion

is

increased

from about 45 to 75%.

2. TTA / TTT – Diagrams

20

Due to the strong influence of the austenitizing temperature to the transformation behaviour of steel, the welding technique uses special diagrams, the so called Welding-TTT-diagrams. They are recorded following the welding temperature cycle with both, higher austenitizing temperatures (basically between 950° and 1350°C) and shorter austenitizing times. You find two examples in Figures 2.16 and 2.17. Figure 2.18 proves that the

2 %C 1

iron-carbon diagram was

0,45 0,5

developed as an equilib-

1000 0

°C 800

1000

rium diagram for infinite

°C

long cooling time and that

P 600 400

B

MS

M

400

Temperature

Temperature

800 F

600

200 200 0 10-1

10

0

10

1

2

10 Time

10

3

s

10

4

0 © ISF 2002

br-eI-02-18.cdr

Relation Between TTT-Diagram and Iron-Carbon-Diagram

Figure 2.18

¥

a TTT diagram applies always oy for one alloy.

3. Residual Stresses

3. Residual Stresses

22 The emergence of residual stresses can be of very different nature, see three tension

examples in Figure 3.1. Figure

grinding disk

3.2

details

the

causes of origin. In a protension

pressure

pressure

duced workpiece, material-

weld

, production-, and wearcaused residual stresses are overlaying in such a © ISF 2002

br-eI-03-01e.cdr

way that a certain condition

Various Reasons of Residual Stress Development

of residual stresses is cre-

Figure 3.1

ated. Such a workpiece shows in service more or

less residual stresses, and it will never be stress-free! Figure 3.3 defines residual stresses of 1., 2., and 3. type. This grading is independent from the origin of the residual stresses. It is rather based on the three-dimensional extension of the stress conditions. Based on this definition, FigAnalysis of Residual Stress Development

ure 3.4 shows a typical distrirelevant material

bution of residual stresses. Residual

stresses,

which

build-up around dislocations

wear

production

e.g. polyphase systems, non-metallic inclusions, grid defects

and other lattice imperfections

mechanical

thermal

chemical

e.g. partial-plastic deformation of notched bars or close to inclusions, fatigue strain

e.g. thermal residual stresses due to operational temperatur fields

e.g. H-diffusion under electro-chemical corrosion

(σIII), superimpose within a grain causing stresses of the 2

nd

type and if spreading

forming

deforming

separating

joining

plating

e.g. thermal residual stresses

residual stresses due to inhomogenuous deformationanisotropy

residual stresses due to machining

residual stresses due to welding

layer residual stresses

changing material characteristics induction hardening, case hardening, nitriding

around several grains, bring © ISF 2002

br-eI-03-02e.cdr

st

out residual stresses of the 1

Development of Residual Stresses

type. The

formation

of

residual

stresses in a transition-free

Figure 3.2

3. Residual Stresses

23

steel cylinder is shown in Figures 3.5. and 3.6. During water quenching of the homogeneous heated cylinder, the edge of the cylinder cools down faster than the core. Not before 100 seconds have elapsed is the temperature across the cylinder's cross section again

s III

tension s

General Definition of the Term ‘Residual Stresses’

Residual stresses of the I. type are almost homogenuous across larger material areas (several grains). Internal forces related to residual stresses of I. type are in an equilibrium with view to any cross-sectional plane throughout the complete body. In addition, the internal torques related to the residual stresses with reference to each axis disappear. When interfering with force and torque equilibrium of bodies under residual stresses of the I. type, macroscopic dimension changes always develop.

s II +

sI

0

x

-

y

Residual stresses of the II. type are almost homogenuous across small material areas (one grain or grain area). Internal forces and torques related to residual stresses of the II. type are in an equilibrium across a sufficient number of grains. When interfering with this equilibrium, macroscopic dimension changes may develop.

x 0

grain boundaries

Residual stresses of the III. type are inhomogenuous across smallest material areas (some atomic distances). Internal forces and torques related to residual stresses of the III. type are in an equilibrium across small areas (sufficiently large part of a grain). When interfering with this equilibrium, macroscopic dimension changes do not develop.

sIII

= residual stresses between several grains = residual stresses in a single grain = residual stresses in a point

< <

sI sII sIII

+

<

br-er03-03e.cdr

sE = s I + sII

© ISF 2002 br-er03-04e.cdr

Definition of Residual Stresses

© ISF 2002

Definition of Residual Stresses of I., II., and III. Type

Figure 3.3

Figure 3.4

homogeneous. The left part of 1000 °C 900

Figure 3.5 shows the T-t°C

urement points in the cylinder. of quenching on the stress condition in the cylinder. At

Temperature

Figure 3.6 shows the results

1

750

2

3

35 mm diameter water cooling 500

250

MS

1 edge 2 50 % radius 3 core

1s

5s

15 s

800

1000

Temperature

curve of three different meas-

0s 10 s

700

20 s

600

25 s

500 35 s 400 45 s

300

53 s 200

the beginning of cooling, the

68 s

0 -2 10

10-1

cylinder edge starts shrinking

10-0

101 102 Cooling time

103

s

104

100 280 s 0 17,5

7

14 10,5

faster than the core (upper

7

0 3,5

3,5

10,5

Radius © ISF 2002

br-eI-03-05e.cdr

figure). Through the stabilising

Temperature in a Cylinder During Water Cooling

effect of the cylinder core, Figure 3.5

mm 17,5

3. Residual Stresses

24

tensile stress builds up at the edge areas while the core is exposed to pressure stress. Resulting volume differences between core and edge are balanced by elastic and plastic deformations. When cooling is completed, edge and core are on the same temperature level, the plastically stretched edge now supports the unstressed core, so that pressurestresses are present in the edge areas and tensile residual stresses in the core.

300

tension pressure

N/mm²

E

200

tension

Stresses in the central rod

Volume differences between edge and core at start of cooling

tension pressure

tension

Compensation of volume differences by plastic deformation and stresses at start of cooling

pressure

D

100

0

A

C

-100

-200

B'

tension

B

pressure

br-er03-06e.cdr

Compensation of volume differences by plastic deformation and stresses at end of cooling

-300 0 © ISF 2002

400

°C

600

br-er03-07e.cdr

© ISF 2002

Residual Stress Development by Warming the Central Rod

Volume Changes During Cooling

Figure 3.6

200

Temperature of the central rod

Figure 3.7

These changes are principally shown once again in Figure 3.7 with the 3-rod model. A warming of the middle rod causes at first an elastic expansion of the outer rods, the inner rod is exposed to pressure stress (line A-B). Along the line B-C the rod is plastically deformed, because pressure stresses have exceeded the yielding point. At point C, the cooling of the rod starts, it is exposed to tensile stress due to shrinking. Along the line D-E the rod is plastically deformed due to the influence of the counter members beeing in tension. At the point E the system has cooled down to its initial temperature. This point represents the remaining residual stress condition of this construction. If heating is stopped before point C is reached and cooled down to the initial temperature, then stress increase in the centre rod will be in parallel

3. Residual Stresses

25

with the elastic areas. Starting with point B, the same residual stress condition is present as in a case of heating up to a temperature above 600°C. Figure 3.8 divides the development of residual stresses in welded seams in three different mechanisms. Shrinking stresses: these are stresses formed through uniform cooling of the seam. Caused by expansion restriction of the colder areas at the edge of the weld and base material , tensile stresses develop along and crosswise to the seam. Quenching stresses: If cooling is not homogenous, the surface of the weld cools down faster than the core areas. If the high-temperature limit of elasticity is exceeded due to buildup stress differences, pressure stresses will be present at the weld surface after cooling. In contrast, the core shows tensile stresses in cold condition (see also Figure 3.6). Transition stresses: Transitions in the ferrite and perlite stage cause normally only residual stresses, because within this temperature range the yield strength of the steel is so low that generated stresses can be undone by plastic deformations. This is not the case with transitions in the Bainite and martensite stage. A transition of the austenite causes an increase in volume (transition cfc in cbc, the cfc lattice has a higher density, additional volume increase through lat+y

tice deformation). In the case of a homoge-

-x

nous transition, the weld will consequently unfold pressure stresses. If the transition of +x

the edge areas happens earlier than the transition of the slower cooling core, plastic de-

-y 2. Quenching stresses

1. Shrinking stresses

+x

-x

+s +y

formations of the core area may be present similar to quenching (see above: quenching

-x

-s -y

+x

stresses). In this case, the weld surface will 3. Transformation stresses

show tensile stresses after cooling. Generally these mechanisms cannot be separated accurately from each other, thus

4. Overlap options of case 1., 2. and 3.

+s +y

+s +y

inhomogenuous transformation

-x

+x

-x

+x

the residual stress condition of a weld will represent an overlap of the cases as shown in the 4th figure. This overlap of the different

homogenuous transformation

-s -y br-er03-08.cdr

-s -y © ISF 2002

Stress Distributions and Superpositions Perpendicular to Welded Joint

mechanisms makes a forecast of the remaining residual stress condition difficult. Figure 3.8

3. Residual Stresses

26 Figure 3.9 shows the building-up of residual

Temperature distribution

Seam

Stress distribution sX

ogy to the 3-rod model of Figure 3.7. This fig-

1. cut A-A DT ~ 0

x

stresses crosswise to a welded seam in anal-

stress-free

ure considers only shrinking residual stresses. Before application of welding heat, the seam

A

A

2. cutt B-B

area is stress-free (cut A-A). At the weldpool tension

weldpool B

the highest temperature of the welding cycle

B

area of plastic deformations

C

pressure

C

can be found (cut B-B), metal is liquid. At this point, there are no residual stresses, because

3. cut C-C

molten metal cannot transmit forces at the D

D

M

weldpool. Areas close to the joint expand through welding heat but are supported by

M' 4. cut D-D

residual stresses

areas which are not so close to the seam.

DT = 0

Thus, areas close to the joint show compresbr-er03-09e.cdr

© ISF 2002

Formation of Residual Stresses Caused by Welding Heat

sion stress, areas away from the joint tensile stress. In cut C-C the already solidified weld metal starts to shrink and is supported by

Figure 3.9

areas close to the seam, the weld metal shows tensile stresses, the adjacent areas compression stresses. In cut D-D is the temperature completely balanced, a residual stress condition is recognised as shown in the lower right figure. 31 15 mm 15 mm

material S235JR (St 37)

103 a a

Figure 3.10 shows how much residual stresses are influenced by constraining ef-

1.

a = 100 mm

s = 800 N/mm²

fects of adjacent material. The resulting

2.

a = 150 mm

s = 530 N/mm²

stress in the presented case is calculated

3.

a = 200 mm

s = 400 N/mm²

according to Hooke:

4.

a = 250 mm

s = 300 N/mm²

σ= ε·E

5.

a = 300 mm

s = 270 N/mm²

br-er03-10e.cdr

© ISF 2002

Shrinking Stresses in a Firmly Clamped Plate

Elongation ε is calculated as ∆ l/a (∆ l is the length change due to shrinking). With conFigure 3.10

3. Residual Stresses

27

stant joint volume will shrinking and ∆ l always have the same value. Thus the elongation ε depends only on the value a. The smaller the a is chosen, the higher are the resulting stresses. Effects of transition on cooling can be estimated from Figure 3.11. Here curves of temperature- and length-changes of ferritic and austenitic steels are drawn. It is clear that a ferritic lattice has a higher volume than an austenitic lattice at the same temperature. A steel which transforms from austenite to one of the ferrite types increases its volume at the critical point. This sudden rise in volume can be up to 3% in the case of martensite formation.

Longitudinal expansion Dl

welding sample 300 x 10 x 30 (70,140) groove angle 60°, depth 4,5 mm

firm clamping

force sensor

el el

thermo couples

links

ste

nit ic

ste

ic

t rri fe

ste

au

to calculator 1000

N

°C

600

800

14

m tra ild ns ste fo el rm w at ith io n

800

200

Temperature

Force

elektrode 400

600

heat affected zone

400 force

0

Temperature [°C]

200 temperature

-200

0 -1 10

100

101

102

103

104

s

105

Time br-er03-11e.cdr

© ISF 2002

br-er03-12e.cdr

Force Measurement During Cooling of a Weld

Longitudinal Expansion of Various Steels

Figure 3.11

© ISF 2002

Figure 3.12

To record the effects of this behaviour on the stress condition of the weld, sample welds are carried out in the test device outlined in Figure 3.12. Thermo couples measure the T-t – curve at the weld seam, a force sensor records the force which tries to bend the samples. The lower picture shows the results of such a test. The temperature behaviour at the fusionline as well as the force necessary to hold the sample over the time is plotted.

3. Residual Stresses

28

In the temperature range above 600°C the force sensor registers a tensile force which is caused by the shrinking of the austenite. Between 600 and 400°C a large drop in force can be seen, which is caused by the transition of the austenite. The repeated increase of the force is based on further shrinking of the ferrite. With the help of TTT diagrams of base material and welding

steel

austenitic

S690QL (StE 70)

consumable,

consumable electrode

austenitic

austenitic

surface weld

surface weld

the

transition

temperatures and/or tempera-

sample shape (V-groove, 60°)

ture areas for the individual

type of welding

zones of the welded joint can

S690QL (StE 70) high-strength

surface weld

position of the HAZ

temperature it can be clearly

residual stress distribution sL

0

pressure

data and with the course of

tension

be determined. With these

determined in which part of

© ISF 2002

br-eI-03-13e.cdr

the curve the force drop is

Influence of Material Combination on Residual Stress Distribution in a Weld

caused by the transition of the welding consumable and in

Figure 3.13

which part by transition in the heat affected 5°42'

2°8'

1°51'

zone (HAZ). These results can be used to determine the longitudinal residual stresses transversal to the joint, as shown in Figure 3.13. During

140

welding of austenitic transition-free materials

Angle change

% 100

only tensile residual stresses are caused in

80 60

the welded area according to Figure 3.8. If an

40 20

austenitic electrode is welded to a StE 70, transitions occur in the area of the heat af-

f = 1°

f = 3°

f = 7°

fected zone which lead to a decrease of ten-

f = 13°

sile stresses. If a high-strength electrode which has a martensitic transition, is welded a=5

a=7

a=9

br-er03-14e.cdr

© ISF 2002

Influence of Welding Sequence on Angle Distortion

Figure 3.14

a = 12,5

to a StE 70, then there will be pressure residual stresses in the weld metal and tensile residual stresses in the HAZ.

3. Residual Stresses

29

If parts to be welded are not fixed, the shrinking of the weld will cause an angular distortion of the workpieces, Figure 3.14 . If the workpieces can shrink unrestricted in this way, the remaining residual stresses will be much lower than in case with firm clamping. Methods to determine residual stresses can be divided into destructive, non-destructive, and conditionally destructive methods. The borehole and ring core method can be considered plan

section

as conditionally destructive, Figures 3.15 and 3.16.

a WSG

In both cases, present residual stresses are released

c

through partial material removal and the resulting deformations

are

measured by wire

b

then

workpiece

strain © ISF 2002

br-eI-03-15e.cdr

gauges. An essential advan-

Residual Stress Determination Using Bore Hole Procedure

tage of the borehole method is the very small material

Figure 3.15

removal, the diameter of the borehole is only 1 to 5 mm, the bore depth is 1- to 2-times the borehole diameter. The disadvantage here is that only surface elongations can be measured, thus the results are limited residual stresses in the surface area of the workpiece. With the ring core method, a crown milling cutter is used to mill a ring groove around a three-axes wire strain gauge. The core is released from the force effects and stress-relieved. At the time when the resilience of the core is measured, the detection of the residual stress distribution

Figure 3.16

3. Residual Stresses

30

across the depth is also possible. Both methods are limited in their suitability for measuring welding residual stresses, because steep strain gradients in the HAZ may cause wrong measurements. The table in Figure 3.17 shows a survey of measurement methods for residual stresses and what causes residual stresses to be picked-up when using one of the respective methods.

Figure 3.17

assumption of stress distribution

Figure 3.18 shows a sur-

measured variable

cutting in layers

vey of the completely destructive

procedures

f

biaxial

f

any

y

0

of

residual stresses

bending deflection f curves reduced curves

sy sz tzy

tear f

partial residual stress relief by Dsz

z

x

cutting-in

residual stress recognition.

f

uniaxial locally different linear, tensile residual stresses on top, down pressure stresses

drilling eT

e45 eL

slitting 0.46f

tripleaxial independent of smple length sL, sT, sR

uniaxial linear symmetrically with reference to rod axis

length change eL circumference change eT

tear f

sL sT sR

partial residual stress relief by Dsz

© ISF 2002

br-eI-03-18e.cdr

Destructive Methods for Determination of Residual Stresses

Figure 3.18

4. Classification of Steels, Welding of Mild Steels

4. Classification of Steels, Welding of Mild Steels

32

In the European Standard DIN EN 10020 (July 2000), the designations

Definition of the term “steel” Steel is a material with a mass fraction if iron which is higher than of every other element, ist carbon content is, in general, lower than 2% and steel contains, moreover, also other elements. A limited number of chromium steels might contain a carbon content which is higher than 2%, but, however, 2% is the common boundary between steel and cast iron [DIN EN 10020 (07.00)].

(main symbols) for the classification of steels are standardised. Figure 4.1 shows the definition of the term „steel“ and the classification of the steel

Classification in accordance with the chemical composition: l

unalloyed steels

l

stainless steels

l

other, alloyed steels

grades

quality classes.

- unalloyed quality steels - unalloyed special steels

· stainless steels · other, alloyed steels

accordance

with

their

chemical composition and the main

Classification in accordance with the main quality class: · unalloyed steels

in

- alloyed quality steels - alloyed special steels

br-er05-01.cdr

© ISF 2004

Definition for the classification of steels

Figure 4.1

In accordance with the chemical composition the steel grades are classified into unalloyed, stainless and other alloyed steels. The mass fractions of the individual elements in unalloyed steels do not achieve the limit values which are indicated in Figure 4.2. Stainless steels are grades of steel with a mass fraction of chromium of at least 10,5 % and a maximum of 1,2 % of carbon. Other alloyed steels are steel grades which do not comply with the definition of stainless steels and where one alloying element exceeds the limit value indicated in Figure 4.2. Figure 4.2

4. Classification of Steels, Welding of Mild Steels

33

As far as the main quality classes are concerned, the steels are classified in accordance with their main characteristics and main application properties into unalloyed, stainless and other alloyed steels. As regards unalloyed steels a distinction is made between unalloyed quality steels and unalloyed high-grade steels. Regarding unalloyed quality steels, prevailing demands apply, for example, to the toughness, the grain size and / or the forming properties. Unalloyed high-grade steels are characterised by a higher degree of purity than unalloyed quality steels, particularly with regard to non-metal inclusions. A more precise setting of the chemical composition and special diligence during the manufacturing and monitoring process guarantee better properties. In most cases these steels are intended for tempering and surface hardening. Stainless steels have a chromium mass fraction of at least 10,5 % and maximally 1,2 % of carbon. They are further classified in accordance with the nickel content and the main characteristics (corrosion resistance, heat resistance and creep resistance). Other alloyed steels are classified into alloyed quality steels and alloyed high-grade steels. Special demands are put on the alloyed quality steels, as, for example, to toughness, grain size and / or forming properties. Those steels are generally not intended for tempering or surface hardening. The alloyed high-grade steels comprise steel grades which have improved properties through precise setting of their chemical composition and also through special manufacturing and control conditions.

4. Classification of Steels, Welding of Mild Steels

34

The European Standard DIN EN 10027-1 (September 1992) stipulates the rules for the designation of the steels by means of code letters and identification numbers. The code letters and identification numbers give information about the main application field, about the mechanical or physical properties or about the composition. The code designations of the steels are divided into two groups. The code designations of the first group refer to the application and to the mechanical or physical properties of the steels. The code designations of the second group refer to the chemical composition of the steels. l S = Steels for structural steel engineering e.g. S235JR, S355J0

According to the utilization of the

l P = Steels for pressure vessel construction e.g. P265GH, P355M

steel and also to the mechanical or

l L = Steels for pipeline construction e.g. L360A, L360QB

physical properties, the steel grades of the first group are designated with

l E = Engineering steels e.g. E295, E360

different main symbols (Fig. 4.3).

l B = Reinforcing steels e.g. B500A, B500B l Y = Prestressing steels e.g. Y1770C, Y1230H l R = Steels for rails (or formed as rails) e.g. R350GHT l H = Cold rolled flat-rolled steels with higher-strength drawing quality e.g. H400LA l D = Flat products made of soft steels for cold reforming e.g. DD14, DC04 l T = Black plate and tin plate and strips and also specially chromium-plated plate and strip e.g. TH550, TS550 l M = Magnetic steel sheet and strip e.g. M400-50A, M660-50D br-er05-03.cdr

© ISF 2004

Classification of steels in accordance with their designated use

Figure 4.3

4. Classification of Steels, Welding of Mild Steels

35

An example of the code designation structure with reference to the usage and the mechanical or physical properties for “steels in structural steel engineering“ is explained in Figure 4.4.

Figure 4.4

4. Classification of Steels, Welding of Mild Steels

36

For designating special features of the steel or the steel product, additional symbols are added to the code designation. A distinction is made between symbols for special demands, symbols for the type of coating and symbols for the treatment condition. These additional symbols are stipulated in the ECISS-note IC 10 and depicted in Figures 4.5 and 4.6.

Symbol1)2)

Coating

+A + AR + AS + AZ + CE + Cu + IC + OC +S + SE +T + TE +Z + ZA + ZE + ZF + ZN

hot dipped aluminium, cladded by rolling coated with Al-Si alloy coated with Al-Tn alloy (>50% Al) electrolytically chromium-plated copper-coated inorganically coated organically coated hot-galvanised electrolytically galvanised upgraded by hot dipping with a lead-tin alloy electrolytically coated with a lead-tin alloy hot-galvised coated with Al-Zn alloy (>50% Zn) electrolytically galvanised diffusion-annealed zinc coatings (galvannealed, with diffused Fe) nickel-zinc coating (electrolytically) 1 2

) The symbols are separated from the preceding symbols by plus-signs (+) ) In order to avoid mix-ups with other symbols, the figure S may precede,

for example +SA © ISF 2004

br-er-05-05.cdr

Symbols for the coating type

Figure 4.5

Symbol1)2)

treatment condition

+A + AC +C

softened annealed for the production of globular carbides work-hardened (e.g., by rolling and drawing), also a distinguishing mark for cold-rolled narrow strips) cold-rolled to a minimum tensile strength of nnn MPa/mm² cold-rolled thermoformed/cold formed slightly cold-drawn or slightly rerolled (skin passed) quenched or hardened treatment for capacity for cold shearing solution annealed untreated

+ Cnnn + CR + HC + LC +Q +S + ST +U

1

) The symbols are separated from the preceding symbols by plus-signs (+) ) In order to avoid mix-ups with other symbols, the figure T may precede,

2

for example +TA © ISF 2004

br-er-05-06.cdr

Symbols for the treatment condition

Figure 4.6

4. Classification of Steels, Welding of Mild Steels

37

Figure 4.7 shows an example of the novel designation of a steel for structural steel engineering which had formerly been labelled St37-2.

The steel St37-2 (DIN 17100) is, according to the new standard (DIN EN 10027-1), designated as follows:

S235 J 2 G3 further property (RR = normalised)

Steel for structural steel engineering

ReH ³ 235 MPa/mm2

test temperature 20°C impact energy ³ 27 J

S = steels for structural steel engineering P = steels for pressure vessel construction L = steels for pipeline construction E = engineering steels B = reinforcing steels © ISF 2002

br-er-05-07.cdr

Steel designation in accordance with DIN EN 10027-1

Figure 4.7

Steel Stahl S355J0 (St 52-3) S500N (StE500) P295NH (HIV) S355J2G1W (WTSt510-3) S355G3S (EH36) Steel Stahl

C

Si

Mn

P

S

Cr

Al

Cu

N

Mo

Ni

Nb

V

£0,20

£0,55

£1,60

0,040

0,040

/

/

/

£0,009

/

/

/

/

0,1 - 0,6 1 - 1,7

0,035

0,030

0,30

0,020

0,20

0,020

0,1

1

0,05

0,22

0,21 £0,26

£0,35

£0,05

£ 0,05

/

/

/

/

/

/

/

/

£0,15

£0,50 0,5 - 1,3 0,035

0,035

0,40 0,80

/

0,25 0,5

/

£0,30

£0,65

/

0,02 0,12

£ 0,18

£0,1 0,7 - 1,5 £0,05 0,35

£ 0,05

/

/

/

/

/

/

/

/

³0,6

Tensile strength Zugfestigkeit RmRm [N/mm²]

yield point ReeHH Streckgrenze [N/mm²]

elongation after fracture Bruchdehnung A A [%]

impact energy AVV Kerbschlagarbeit [J] -20°C

0°C S355J2G3 (St 52-3) S500N (StE500) P295NH (HIV) S355J2G1W (WTSt510-3) S355G3S (EH36)

510-680

355

20-22

27 31-47

610-780

500

16

460-550

285

>18

510-610

355

22

400-490

355

>22

27 21-39

49 (bei +20°C)

76 (bei -10°C) © ISF 2004

br-er-05-08.cdr

Chemical composition and mechanical parameters of different steel sorts

Figure 4.8

Figure 4.8 depicts the chemical composition and the mechanical parameters of different steel grades. The figure explains the influence of the chemical composition on the mechanical properties.

4. Classification of Steels, Welding of Mild Steels

38

The steel S355J2G2 represents the basic type of structural steels which are nowadays commonly used. Apart from a slightly increased Si content for desoxidisation it this an unalloyed steel. S500N is a typical fine-grained structural steel. A very fine-grained microstructure with improved tensile strength values is provided by the addition of carbide forming elements like Cr and Mo as well as by grain-refining elements like Nb and V. The boiler steel P295NH is a heat-resistant steel which is applied up to a temperature of 400°C. This steel shows a relatively low strength but very good toughness values which are caused by the increased Mn content of 0,6%. S355J2G1W is a weather-resistant structural steel with mechanical properties similar to S355J2G2. By adding Cr, Cu and Ni, formed oxide layers stick firmly to the workpiece surface. This oxide layer prevents further corrosion of the steel. S355G3S belongs to the group of shipbuilding steels with properties similar to those of usual structural steels. Due to special quality requirements of the classification companies (in this case: impact energy) these steels are summarised under a special group.

4. Classification of Steels, Welding of Mild Steels

39

The steel grades are classified into four subgroups according to the chemical composition (Fig. 4.9): ● Unalloyed steels (except free-cutting steels) with a Mn content of < 1 % ● Unalloyed steels with a medium Mn content > 1 %, unalloyed free-cutting steels and alloyed steels (except high-speed steels) with individual alloying element contents of less than 5 percent in weight ● Alloyed steels (except high-speed steels), if, at least for one alloying element the content is ≥ 5 percent in weight ● High-speed steels

The unalloyed steels with Mn contents of < 1% are labelled with the code letter C and a number which complies with the hundredfold of the mean value which is stipulated for the carbon content. Unalloyed steels with a medium Mn content > 1 %,

unalloyed free-

cutting steels and alloyed steels (individual alloying element contents < 5 %) are labelled with a number which also complies with a hundredfold of the mean value which is stipulated for the carbon content, the chemical symbols for the alloying elements, ordered according to the decreasing contents of the alloying Figure 4.9

elements and numbers, which in the sequence of the designating alloying

elements give reference about their content. The individual numbers stand for the medium content of the respective alloying element, the content had been multiplied

4. Classification of Steels, Welding of Mild Steels

40

by the factor as indicated in Fig. 4.9 / Table 4.1 and rounded up to the next whole number. The alloyed steels are labelled with the code letter X, a number which again complies with the hundredfold of the mean value of the range stipulated for the carbon content, the chemical symbols of the alloying elements, ordered according to decreasing contents of the elements and numbers which in sequence of the designating alloying elements refer to their content. High-speed steels are designated with the code letter HS and numbers which, in the following sequence, indicate the contents of elements:: tungsten (W), molybdenum (Mo), vanadium (V) and cobalt (Co).

The European Standard DIN EN 10027-2 (September 1992) specifies a numbering system for the designation of steel grades, which is also called material number system.. The structure of the material number is as follows: 1.

XX

XX (XX) Sequential number The digits inside the brackets are intended for possible future demands. Steel group number (see Fig. 4.10) Material main group number (1=steel)

4. Classification of Steels, Welding of Mild Steels Figure 4.10 specifies the material numbers for the material main group „steel“.

Figure 4.10

41

4. Classification of Steels, Welding of Mild Steels

42

The influence of the austenite grain size on the transformation behaviour has been explained in Chapter 2. Figure 4.11 shows the dependence between grain size of the austenite which develops during the welding cycle, the distance from the fusion line and the energy-per-unit length from the welding method. The higher the energy-peruntil

length,

the

bigger the austenite grains in the

13

HAZ and the width

Austenite grain size index according to DIN 50601

Energy-per-unit length in kJ/cm

11 9

12

18

of

36

the

HAZ

in-

9

creases.

Such

7

coarsened austenite grain decreases

5

the critical cooling 3 0

0,2

0,4 0,6 Distance of the fusion line

0,8

mm

1,0 © ISF 2004

br-er-05-11.cdr

Influence of the energy-per-unit length on the austenite grain size

time, thus increasing the tendency of the steel to harden.

Figure 4.11

With fine-grained structural steels it is tried to suppress the grain growth with alloying elements. Favourable are nitride and carbide forming alloys. They develop precipitations which suppress undesired grain growth. There is, however, a limitation due to the solubility of these precipitations, starting with a certain temperature, as shown in Figure 4.12. Steel 1 does not contain any precipitations and shows therefore a continuous grain growth related to temperature. Steel 2 contains AIN precipitations which are stable up to a temperature of approx. 1100°C, thus preventing a growth of the austenite grain.

4. Classification of Steels, Welding of Mild Steels

43

With

Grain size index according to DIN 50601

mm 1 8 6

Medium fibre length

4

2

10 8 6

-1

4

2

10-2 8 6 10

higher

temperatures,

-4

precipitations dissolve and cannot

-2

suppress a grain growth any more.

0

Steel 3 contains mainly titanium car-

2

bonitrides of a much lower grain-

4

refining effect than that of AIN. Steel 4

6

is a combination of the most effective properties of steels nos. 2 and 3.

8 Steel 1 Steel 2 Steel 3 Steel 4

10

-3

12 900

1000

1100 1200 Austenitization temperature

1300

The importance of grain refinement for the mechanical properties of a

°C

1400

steel is shown in Figure 4.13. Pro-

Steel

%C

% Mn

% Al

%N

% Ti

1

0,21

1,16

0,004

0,010

/

2

0,17

1,35

0,047

0,017

/

3

0,18

1,43

0,004

0,024

0,067

4

0,19

1,34

0,060

0,018

0,140

br-er05-12.cdr

vided the temperature keeps constant, the yield strength of a steel increases with decreasing grain size.

© ISF 2004

This influence on the yield point Rel is

Austenite grain size as a function of the austenitization temperature

specified

in

Rel = σ i + K ⋅

Figure 4.12

According

to

1 d

propor-

tional to the root of the medium grain

N/mm² 800 Yield point or 0,2 boundary

the yield point is

Temperature in °C:

700

-193 -185

600

-170 -155

-100

σi

300

-40

stands for the inter-

200

diameter d.

-180

500 400

+20 0

nal friction stress of

1

2

3

4

5 6 -1/2 Grain size d

7

grain

for

is

a

mm-1/2

10

Connection between yield point and grain size

boundary

resistance K measure

The

8

© ISF 2004

br-er-05-13.cdr

material.

Hall-Petch-law:

900

law, the increase of inversely

the

the

above-mentioned

the

these

Figure 4.13

the

influence of the grain size on the forming mechanisms. Apart from this increase of the yield point, grain refinement also results in improved toughness values. As far as

4. Classification of Steels, Welding of Mild Steels

44

structural steels are concerned, this means the improvement of the mechanical properties without any further alloying. Modern fine-grained structural steels show improved mechanical properties with, at the same time, decreased content of alloying elements. As a consequence of this chemical composition the carbon equivalent decreases, the weldability is improved and processing of the steel is easier. The major advanSteel type Stahlsorte

S235JR (St37-2)

S355J2G3 (St52-3)

S690Q (StE690)

S890Q (StE890)

S960Q (StE960)

Verhältnis Ratio S235JR - S960Q

N/mm2

215

345

690

890

960

1:5

Plate thickness Blechdicke

mm

50

31

14,4

11

10

5:1

Yield point Streckgrenze Weld cross-section Nahtquerschnitt

mm2

870

370

100

60

50

17 : 1

Welding wire Øø1.2 Schweißdraht 1.2

mm

SG2

SG3

NiMoCr

X 90

X 96

-

Welding wire costs Schweißdrahtkosten

Ratio Verhältnis

1

1

2,4

3,2

3,3

1 : 3,3

Steel costs Stahlkosten

Ratio Verhältnis

1

1,2

1,9

2,3

2,4

1 : 2,4

Weld metal costs Schweißgutkosten

Ratio Verhältnis

5,3

2,3

1,5

1,16

1

5,3 : 1

Special weld costs Spez. Schweißnahtkosten

Ratio Verhältnis

12

5,1

1,8

1,18

1

12 : 1

Costs ratio inclusive base Kostenverhältnis inklusive materials Grundwerkstoffe

Randbedingungen: Boundary condition:

tages of microalloyed

fine-grained

structural steels in comparison

with

conventional structural

5:1

steels

shown

Schweißverfahren = MAG welding process = MAG

in

are

Figure

Deposition rate = 3 kg=welding wire/h, weld /shape X -60° X - 60° Abschmelzleistung 3 kg Schweißdraht h, Nahtform

4.14. Due to the

Costs labour and equipment == 60 30€/h Lohn-ofund Maschinenkosten DM / h Special costs = weld filler materials + welding Spez. weld Schweißnahtkosten = Schweißzusatzwerkstoffe + Schweißen

considerably better

Berechnungsgrundlage =szul = Re / 1.5 Calculation base = szul = Re/1.5 © ISF 2004

br-er-05-14.cdr

mechanical proper-

Influence of the steel selection on the producing costs of welded structures

ties of the finegrained

Figure 4.14

structural

steel in comparison with unalloyed structural steel, substantial savings of material are possible. This leads also to reduced joint cross-sections and, in total, to lower costs when making welded steel constructions. Based

on

steels

the

alloyed

unalloyed

classification Figure

4.2,

of Fig-

low-alloyed mild steel

higher-carbon steel Hardening Underbead cracking

ure 4.15 divides the steels with regard

rimmed steel

to their problematic

cutting of segregation zones

processes

during

welding. When it

killed steel duplex killed steel

cold brittleness (coarse-grained recrystallization after critical treatment) stress corrosion cracking safety from brittle fracture

comes to unalloyed

high-alloyed

hardening corrosion tool steels special properties are resistant steels achieved, for example: Hardening, special properties heat resistance, are achieved tempering resistant, high-pressure hydrogen resistance, toughness at low temperatures, surface treeatment condition, etc. ferritic

pearlitic-martensitic

austenitic

grain increase in the weld interfaces

hardening embrittlement formation of chromium carbide

grain desintegration stress corrosion cracking hot cracks (sigma phase embrittlement)

Post-weld treatment for highest corrosion resistance © ISF 2004

br-er-05-15.cdr

steels, only ingot

Classification of steels with respect to problems during welding

Figure 4.15

4. Classification of Steels, Welding of Mild Steels

45

casts, rimmed and semi-killed steels are causing problems. “Killing” means the removal of oxygen from the steel bath. Figure 4.16 shows cross-sections of ingot blocks with different oxygen contents. Rimming steels with increased oxygen content show, from the outside to the inside, three different zones after solidification: 1.: a pronounced, very pure outer envelope, 2.: a typical blowhole formation (not critical, blowholes are forged together during rolling), 3.: in the centre

a

segregated

clearly zone

where unfavourable elements like sulphur and phosphorus are enriched.

0,025 0,012

During rolling, such

0,003

fully killed steel

semi-killed steel

zones are stretched

rimmed steel

along the complete

Figures: mass content of oxygen in % © ISF 2004

br-er-05-16.cdr

length of the rolling

Ingot cross-sections after different casting methods

profile. Figure 4.16

Figure 4.17 shows important points to be observed during welding such steels. Due to their enrichment with alloy elements, the segregation zones are more transformation-inert than the outer

envelope

a

b

and are inclined to hardening.

In

addition, they are sensitive

to

cracking,

as,

hotin

B

these zones, the

D

C E

elements phosphorus are

and

sulphur

© ISF 2004

br-er-05-17.cdr

enriched.

Example of unfavourable (a) and favourable (b) welds

Figure 4.17

4. Classification of Steels, Welding of Mild Steels

46

Therefore, “ touching” such segregation zones during welding must be avoided by all means. In the case of lowalloy

steels,

the

Microstructures

Average Brinell Hardness (Approximately)

Ferrite

80

Austenite

250

Perlite (granular)

200

welding

Perlite (lamellar)

300

observed.

Sorbite

350

Troostite

400

Cementite

600 - 650

ness

Martensite

400 - 900

various microstruc-

problem

of

HAZ

hardening

during must

be

Figure

4.18 shows hardof

tures. The highest

© ISF 2004

Br-er-05-18.cdr

values

hardness

Hardness of Several Microstructures

values

can be found with Figure 4.18

martensite

and

cementite. Hardness values of cementite are of minor importance for unalloyed and low-alloy steels because its proportion in these steels remains low due to the low Ccontent. However, hardening because of martensite formation is of greatest importance as the martensite proportion in the microstructure depends mainly on the cooling time. Figure 4.19 shows the essential influHV

HRC

strength, calculated at max. hardness N/mm2

root cracking presumable

400

41

1290

70

root cracking possible

400 - 350

41 - 36

1290 - 1125

70 - 60

no root cracking

350

36

1125

60

sufficient operational safety without heat treatment

280

28

900

30

maximum hardness

ence of the martensite

content

in

the HAZ on the crack formation of welded

joints.

Hardening through martensite

forma-

with maximum martensite content %

If too much martensite develops in the heat affected zone during welding (below or next to the weld), a very hard zone will be formed which shows often cracks.

tion is not to be © ISF 2004

Br-er-05-19.cdr

expected with pure

Influence of Martensite Content

carbon steels up to about

0,22%,

Figure 4.19

4. Classification of Steels, Welding of Mild Steels

47

because the critical cooling rate with these low C-contents is so high that it normally won’t be reached within the welding cycle. In general, such steels can be welded without special problems (e.g., S. 235). In addition to carIIW

C - Äqu. = C +

Mn Cr + Mo + V Cu + Ni + + 6 5 15

Stout

C - Äqu. = C +

Mo Ni Cu Mn Cr + Mn + + + 6 10 20 40

Ito and Bessyo

PCM = C +

Mannesmann

C - Äqu.PLS = C +

Hoesch

C - Äqu. = C +

C ET

Thyssen

bon, all other alloy elements are important

Si Mn + Cu + Cr Ni Mo V + + + + + 5B 30 20 60 15 10

site

formation

in

the welding cycle,

Si + Mn + Cu + Cr + Ni + Mo + V 20

as they have sub-

Mn + Mo Cr + Cu Ni = C+ + + 10 20 40

stantial

PLS = pipeline steels

it

comes to marten-

Si Mn + Cu Cr Ni Mo V + + + + + 25 16 20 60 40 15

C-Äqu.= carbon equivalent (%)

when

influence

on the transforma-

PCM = cracking parameters (%) © ISF 2002

Br-er-05-20.cdr

tion behaviour of Definition of C - Equivalent

steels

Figure 4.20

(see

Fig. 2.12 ). It is not appropriate just

to take the carbon content as a measure for the hardening tendency of such steels. To estimate the weldability, several authors developed formulas for calculating the so-called carbon equivalent, which include the contribution of the other alloy elements to hardening tendency, (Fig. 4.20). As these approximation formulas are empirically determined as

for

0,35

Tp ==750 CET - 150- 150 Tp 750 CET

delta Tp HD HD0,35 - 100 delta Tp= 62 = 62 - 100 80

200

the

delta Tp [°C]

and

100

250

hardening tendency

Tp [° C]

150

100

d = 30 mm d = 30 mm HD HD = 4= 4 1 kJ/mm Q = Q1=kJ/mm

0 0,2

tions

like

0,3

0,4

CET = =0,33 % CET 0,33 % = 30mm mm d =d30 kJ/mm Q =Q1= 1kJ/mm

0 0

0,5

5

60

heat

10

15

20

25

Wasserstoffgehalt Hydrogen contentHD of des theSchweißgutes weld metal [%]

Kohlenstoffäquivalent CET [%] Carbon aquivalent

plate

40

delta TpTp = 160 tanhtanh (d/35) (d/35) - 110 - 110 delta = 160

thickness,

40

20

50

the general condi-

60

delta Tp CETCET - 32)-Q32) - 53Q CET + 32 delta Tp= (53 = (53 - 53 CET + 32 20

50

CET = 0,4 %

CET = 0,2 %

CET = 0,2 %

CET = 0,2 %

CET = 0,4 %

CET = 0,2 %

0

delta Tp [°C]

input, etc., are also

delta Tp [°C]

40

30

-20

-40

20 -60

of importance, the

10

CET 0,4 CET ==0,4 %% HD = 2 2 HD QQ== 11kJ/mm kJ/mm

0

carbon

equivalent

cannot be a com-

0

20

40

60

80

100

-80

d =d50 = 50mm mm =8 HDHD =8

-100 0

0,5

Tp =697 CET + 160 tanh (d/35) + 62 HD

mon limit value for the weldability. For the determina- Figure 4.21

1,5

2

2,5

3

3,5

4

4,5

Wärmeeinbringen Heat input Q [kJ/mm]

Blechdicke d [mm] Plate thickness

br-er05-21.cdr

1

0,35

+ (53 CET - 32) Q - 328

Source: Quelle: DIN EN 1011-2

Calculation of the preheating temperatures

© ISF 2005

5

4. Classification of Steels, Welding of Mild Steels

48

tion of the preheating temperature Tp, the formula as shown in Figure 4.21 is used. The effects of the chemical composition which is marked by the carbon equivalent CET, the plate thickness d, the hydrogen content of the weld metal HD and the heat input Q are considered. The essential factor to martensite forma-

Temperature T

Tmax

tion in the welding cycle is the cooling

°C

time. As a measure 800

of cooling time, the DT

time of cooling from

500

800 to 500°C (t8/5) is

t8/5

defined (Fig. 4.22). t800

t500

s

The

Time t

temperature

© ISF 2004

br-er-05-22.cdr

range was selected

Definition of t8/5

in such a way that it covered the most

Figure 4.22

important structural transformations and that the time can be easily transferred to the TTT diagrams. Figure 4.23

shows 2000

measured

time-

temperature

distri-

°C

ity of a weld. Peak values

and

dwell

times depend obvi-

Temperature T

butions in the vicin-

B

1500

A

A

of

B 500 C

the

0 0

measurement

10mm

1000

ously on the location

and

50

100

150

200

are clearly strongly determined by the conduction Figure 4.23

conditions.

250

s

300

Time t © ISF 2004

br-er-05-23.cdr

heat

C

Temperature-time curves in the adjacence of a weld

4. Classification of Steels, Welding of Mild Steels

49

With the use of thinner plates with complete heating of the cross-section during welding, the heat conductivity is only carried out in parallel to the plate surface, this is the two-dimensional heat dissipation. With thicker plates, e.g. during welding of a blind bead, heat dissipation can also be carried out in direction of plate thickness, heat dissipation is three-dimensional. These two cases

3 - dimensional:

K3 t8 / 5 =

universal formula:

ö h U ×I æ 1 1 ÷ × ×ç 2 × p × l v çè 500 - T0 800 - T0 ø÷

are covered by the

) Uv× I × æçç 5001- T

formulas given in

(

extended formula For low-alloyed steel:

t8 / 5 = 0,67 - 5 ×10 - 4 T0 ×

è

-

0

ö 1 ÷ ×h ¢ × N 3 800 - T0 ø÷

Figure 4.24, which K2

2 - dimensional: t8 / 5 =

universal formula:

extended formula For low-alloyed steel:

provide a method

2 2 2 ö ù ö æ h2 1 1 æ U × I ö 1 éæç ÷ ú ÷ -ç ×ç ÷ × ×ê 4 × p × l × r × c è v ø d 2 êçè 500 - T0 ÷ø çè 800 - T0 ÷ø ú ë û

of calculating the

2 2 2 ö æ ö ù 2 1 1 æ U × I ö 1 éæç ÷ -ç ÷ ú ×h ¢ × N 2 t8 / 5 = 0,043 - 4,3 ×10 -5 T0 × ç ÷ × 2 ×ê è v ø d ëêçè 500 - T0 ÷ø çè 800 - T0 ÷ø ûú

(

formula for the transition thickness of low-alloyed steel:

)

dü =

0,043 - 4,3 ×10 -5 T0 U ×I ×h ¢ × 0,67 - 5 ×10 - 4 T0 v

cooling time t8/5 of

ö æ 1 1 ÷÷ × çç + è 500 - T0 800 - T0 ø

low-alloyed steels. In the case of a © ISF 2004

br-er-05-24.cdr

three-dimensional

Calculation equation for two- and three-dimensional heat dissipation

heat

dissipation,

t8/5 it independent

Figure 4.24

of plate thickness. In the case of two-dimensional heat dissipation it is clear that t8/5 becomes the shorter the thicker the plate thickness d is. Provided, the cooling times are equal, the plate thickness can be calculated from these relations where a two-dimensional heat dissipation changes to a three-dimensional heat dissipation. Figure 4.25 shows welding methods

the influence of the

TIG-(He)-welding

welding method on

TIG-(Ar)-welding

the heat dissipa-

MIG-(Ar)-welding

tion. With the same

MAG-(CO2)- welding

heat

the

Manual arc welding

is

SA welding

input,

energy

which

0

transferred to the base

material

depends

on

0,1

0,2

0,3

0,4

0,5

0,6

0,7

0,8

0,9

Relative thermal efficiency degree h‘ © ISF 2004

Br-er-05-25.cdr

the

Relative thermal efficiency degree of different welding methods

Figure 4.25

1

4. Classification of Steels, Welding of Mild Steels

50

welding method. This dependence is described by the relative thermal efficiency ŋ’. The influence of the

groove

Type of weld

ge-

2-dimensional heat dissipation

ometry is covered

weld factor 3-dimensional heat dissipation

1

1

0,45 - 0,67

0,67

0,9

0,67

0,9

0,9

by seam factors according

to

Fig. 4.26. Empirically determined, these factors were introduced for an

© ISF 2004

br-er-05-26.cdr

easier calculation.

Weld factors for different weld geometries

For other groove geometries, tests Figure 4.26 to measure the cooling time are recommended.

Fig. 4.27 shows the transition of the two-dimensional to the three-dimensional heat dissipation for two different preheating temperatures in form of a curve according to the equation of Fig. 4.24. Above the curve, t8/5 depends only on the energy input, but not on the plate thickness, heat dissipation is carried out three-dimensionally.

5 cooling time t8/5 [s] 10 15 20

cm

cooling time t8/5 [s] 10 20 30

25

Plate thickness

TA=20°C

40

50

TA=200°C

3 30 40

3-dimensional 2

60 80 100 150

3-dimensional

60 100

1

2-dimensional

2-dimensional

0 0

10

20

30

40

50

0

10

20

30

40

Heat input E.h.Nn [kJ/cm] © ISF 2004

Br-er-05-27.cdr

Transition From Two to Three Dimensional Heat Flow

Figure 4.27

50

4. Classification of Steels, Welding of Mild Steels

51 Fig. 4.28 shows the possible range of

20

heat input depend-

kJ/cm

ing on the elec-

-spray arc

trode diameter. It is

Heat input

12

clear that a rela-

8

tively large working

4 -short arc

3,25 4 5 6 Manual metal arc welding

0,8 1,0 1,2 1,6 MAGC-, MAGMmethod

range is available for

2,5 3,0 4,0 5,0 SA-welding

© ISF 2004

br-er-05-28.cdr

arc

procedures. variation

Heat Inputs of Various Welding Methods

welding of

A the

energy-per-unit

Figure 4.28

length

can

be

carried out by alteration of the welding current, the welding voltage and the welding speed. Fig. 4.29 depicts variations of the heat Stick electrode (mm)

2,5

3,25

4,0

5,0

6,0

input during manual metal arc weld-

Current intensity (A)

90

135

180

235

275

ing. The shorter the fused electrode

Current intensity (A)

75

120

140

190

250

distance, i.e., the shorter the extracted length, the higher the energy-

35

per-unit length. kJ/cm

Energy-per-unit length

25 20

Æ6,0mm x 450mm

15 Æ5,0mm x 450mm

10

Æ4,0mm x 450mm Æ3,25mm x 350mm

5 0

Æ2,5mm x 350mm

0

50 100 150 200 250 300 350 400 450 500 mm 600 run-out length

br-er05-29.cdr

© ISF 2004

Energy-per-unit length as a function of the run-out length

Figure 4.29

4. Classification of Steels, Welding of Mild Steels

52

In order to minimize calculation efforts in practice, the specified relations were transferred into nomograms from which permissible welding parameters can be read out, provided some additional data are available. Fig. 4.30 shows diagrams for twodimensional heat dissipation, where a dependence between energy-per-unit length, cooling time and preheating temperature is given, depending on the plate thickness. .

50 40 30

T0 200°C 150°C 100°C

20

20°C

Cooling time t8/5 in s

10

d = 7,5 mm

7 50 40 30

T0 200°C 150°C 100°C

20

20°C

10

d = 10 mm

7 50 40 30

T0 200°C 150°C 100°C

20

20°C

10

d = 15 mm

7 50 40 30

T0 200°C 150°C 100°C

20 transition to 3-dimensional heat flow

10

20°C d = 20 mm

7 5 br-er05-30.cdr

6

7 8 9 10

15 20

30

kJ/cm 50

Heat input E

© ISF 2004

Dependence of E, t8/5 and d During SA - Welding

Figure 4.30

If a fine-grained structural steel is to be welded, the steel manufacturer presets a certain interval of cooling times, where the steel characteristics are not too negatively affected. The user lays down the plate thickness and, through the selection of a welding method, a specified range of heat input E. Based on the data E and t8/5 the diagram provides the required preheating temperature for welding the respective plate thickness.

4. Classification of Steels, Welding of Mild Steels

With the transition to thicker plates,

Transition thickness dÜ

50 mm 40

the diagrams in Fig. 4.31 apply. The

aera of 3-dimensional heat flow

30

T0

20 15

10 9 8 7

0 °C °C 20 °C 2 50 00 1 C ° 1 50 °C 20

upper part of the figure determines whether a two-dimensional or a threedimensional heat dissipation is pre-

area of 2-dimensional heat flow

sent. For the three-dimensional heat dissipation, the lower diagram applies

5

6

7 8 9 10

15 20

30

kJ/cm 50

where the same information can be

Heat input E 50 s 40

determined,

Cooling time t8/5

independent

of

thickness, as with Fig. 4.30.

30

20 15

25 T

0

0

°C

20

0

°C

15

0

°C 10

10 9 8 7

53

5

6

7 8 9 10

0

°C 20

15 20

°C

30

Heat input E

br-er05-31.cdr

kJ/cm 50 © ISF 2004

Dependence of E, T0, t8/5 And dÜ

Figure 4.31

The

relation

be-

tween current and

35 V

voltage for MAG

gas composition: C1 100% CO2 M21 82% Ar + 18% CO2 M23 92% Ar + 8% O2

C1 M21

30

in Fig. 4.32

and

the used shielding gas is one of the

Welding voltage

M23

welding is shown

25

20

15

parameters. Welding

voltage

mixed arc

contact tube distance ~15mm

150

and

welding current, or

3,5 br-er-05-32.cdr

wire feed speed,

4,5

spray arc

contact tube distance ~19mm

200

250 Welding current

A

300

5,5

7,0 Wire feed

9,0

10,5

8,0

m/min © ISF 2004

Dependence of Current And Voltage During MAG-Welding, Solid Wire, Æ1.2 mm

determine the type of arc.

short arc

Figure 4.32

plate

4. Classification of Steels, Welding of Mild Steels

54

The diagram in Fig. 4.33 demonh'UP = 1 h'MAG = 0,85 dU max = 32 mm dU min = 15 mm

F3 = 0,67 F2 = 0,67

t8/5 max = 30 s t8/5 min = 6 s

Emax = 66 kJ/cm Emin = 14 kJ/cm

ness, heat input E and cooling time

60 fillet welds T0= 150 °C

kJ/cm

30s

70

t8/5

kJ/cm

temperature of T0 = 150°C. If d and

59

50

53

20s

41

35 30

15s

heat input can be determined with the

Heat input E MAG - weldind

47

40

35

help of this diagram. The kinks of the curves mark the transition between

29

25 10s

20

two-dimensional

and

three-

23

dimensional heat dissipation.

18

15

for fillet welds at a preheating

t8/5 are given, the acceptable range of

25s

toughness affection

45

Heat input E SA - welding

strates the dependence of plate thick-

6s

12

10 cracking tendency

5 0

0

5

10

15

20 25 30 Plate thickness

6 mm

0 40 © ISF 2004

br-er05-33.cdr

Permissible E-Range During SA - And MAG - Welding

Figure 4.33

Fig. 4.34 shows the same dependF3 = 0,9 F2 = 0,9

t8/5 max = 30 s t8/5 min = 6 s

Emax = 49 kJ/cm Emin = 10 kJ/cm

60

70 butt welds T0= 150 °C

kJ/cm

kJ/cm

50

59 toughness affection

45

30s

40

53 47

25s

35 30

20s

25

41 35 29

15s

20

23

15

10s

18

10

6s

12

cracking tendency

5 0

Heat input E MAG - welding

preparation.

h'UP = 1 h'MAG = 0,85 dU max = 34 mm dU min = 15 mm

Heat input E SA - welding

ence for butt welds with V groove

0

br-er05-34.cdr

5

10

15

20 25 30 Plate thickness

mm

6 0 40 © ISF 2004

Permissible E-Range During SA - And MAG - Welding

Figure 4.34

4. Classification of Steels, Welding of Mild Steels

55

The curve family in Fig. 4.35 shows the dependence of the heat input from the welding speed as well as the acceptable working range. The parameters of the curves 1 to 8 in the table curve

25 kJ/cm

2

3

4

5

6

7

8

V

29

27

24

22

20

19

18

17

A

300 275 250 225 200 175 150 125

from Figures 4.32

5.5 4.5 3.5 3.0

and 4.34 and apply

vZ(m/min) 10.5 9.0 8.0 7.0

20

1 2

Heat input E

have been taken

1

wor

king

3

15

rang

4

only

related

conditions like wire

6 7

10

for

e

5 8

diameter,

5

wire

feed,

0 10

15

20

25

30 35 40 45 Welding speed vS

50 cm/min 60

welding

voltage, etc.

MAG/ M21 (82% Ar, 18% CO) © ISF 2004

br-er-05-35.cdr

E as a Function of Welding Speed, Solid Wire, Æ1.2mm

Figure 4.35

shows

Sheet

Nr. 0916). In this example, a plate thickness of 15 mm and a cooling

time

t8/5

be-

1

2

3

4

5

6

7

8

V

29

27

24

22

20

19

18

17

59

A

300 275 250 225 200 175 150 125

toughness affection

45

53 30s

40

47 25s

35 30

20s

25

10s

15 10

6s

cracking tendency

5 0

41 35 29

15s

20

0

5

10

15

20 25 30 Plate thickness

mm

curve

kJ/cm

23 18

16 12 13 6 0 40

vZ(m/min) 10.5 9.0 8.0 7.0

5.5 4.5 3.5 3.0

25 kJ/cm 20

1 2

heat input E

Reference

SA - welding

(according to DVS-

70 butt welds T0= 150 °C

50

Heat input E

for such diagrams

60 kJ/cm

MAG - welding

a reading example

Heat input E

Figure 4.36

16 15 13

work

ing

3 4

rang

e

5 6 7

10

8

5

33

0 10

15

20

25

41

30 35 40 45 Welding speed vS

50 cm/min 60

© ISF 2004

br-er-05-36.cdr

Determination of Welding Speed for MAG - Welding

tween 10 and 20 s are given. In this case, the maximum

Figure 4.36

cooling time for MAG welding is 15 s. A solid wire with a diameter of 1.2 mm at 29V and 300A is used. The left diagram provides heat input values between 13 and 16 kJ/cm, based on the given data. Using these values, the acceptable range of welding speeds can be taken from the diagram on the right.

4. Classification of Steels, Welding of Mild Steels

Fig. 4.37 presents a simplification of

56

800 °C

the determination of the microstruc-

700

tural composition and cooling time subject to peak temperatures which

Temperature

F

occur in the welding cycle. In the

line. The point of intersection of the

500 400

M Peak temperature 1000°C 1400°C

200 HV30=400

300

200

1400

Peak temperature

the point of heat input at the lower

P B

300

lower diagram, the point of the plate thickness at the top line is linked with

600

°C

B+M

M

F+B

1000 Arc3 800

Arc1

linking line with the middle scale 600

represents the cooling time t8/5 .

middle diagram in which transition

1

plate thickness 40

If the peak temperature of the welding cycle is known, one can read from the

30

two-dimensional

10 25

three-dimensional

1

20

s

100

15

10 9 8

7

6

1000

t8/5

5 mm 4

300 200 100

2 3

5

10

20

50 100 200 400 s 1000 0

100 °C

200

t8/5

preheating temperature

energy-per-unit length 6

field the final microstructures are

F+P

1200

8

10

20

30

40

50 kJ/cm 70

bie5-37.cdr

formed. The advantage of the determination of microstructures compared

© ISF 2004

Peak temperature/cooling time – diagram for the determination of t8/5 and the structure

with the upper TTT diagram is that Figure 4.37 a TTT diagram applies only for exactly one peak temperature, other peak temperatures are disregarded. The disadvantage of the PTCT diagram (peak temperature cooling time diagram) is the very expensive determination, therefore, due to the measurement efforts a systematic application of this concept to all common steel types is subject to failure.

5. Welding of High-Alloy Steels, Corrosion

5. Welding of High-Alloy Steels, Corrosion

58

Basically stainless steels are characterised by a chromium content of at least 12%. Figure 5.1 shows a classification of

corrosion

corrosion-resistant steels

resistant

steels. They can be sin-

stainless steels

gled out as heat- and scale-resistant

scale- and heat-resistant steels

and

stainless steels, depend-

perlitic martensitic

semi-ferritic

ferritic

X40Cr13

X10Cr13

X8Cr13

ferritic-austenitic

austenitic

ing on service temperaX20CrNiSi25-4

ture. Stainless steels are used at room temperature conditions and for water-

non-stabilized

stabilized

(austenite with delta-ferrite) X12CrNi18-8

(austenite without delta-ferrite) X8CrNiNb16-13 © ISF 2002

br-er-06-01e.cdr

based media, whilst heatClassification of Corrosion-Resistant Steels

and scale-resistant steels are applied in elevated

Figure 5.1

temperatures and gaseous media. Depending on their microstructure, the alloys can be divided into perlitic-martensitic, ferritic, and austenitic steels. Perlitic-martensitic steels have a high strength and a high wear resistance, they are used e.g. as knife steels. Ferritic and corrosion resistant steels are mainly used as plates for household appliances and other decorative purposes. The most important group are austenitic steels, which can be used for very many applications and which are corrosion resistant against most media. They have a very high low temperature impact resistance. Based on the simple Fe-C T

T A4

T d

phase diagram (left figure), d

Figure 5.2 shows the ef-

A4

A4 g

g

A3

g

a(d)

fects

of

two

different

A3

A3 a

groups of alloying elements

a

on the equilibrium diagram. Alloy elements in %

Alloy elements in % Chromium Vanadium Molybdenum Aluminium Silicon

Alloy elements in %

Ferrite

Nickel Manganese Cobalt

developers

with

chromium as the most important element cause a © ISF 2002

br-er-06-02e.cdr

Modifications to the Fe-C Diagram by Alloy Elements

Figure 5.2

strong reduction of the aus-

5. Welding of High-Alloy Steels, Corrosion

59

tenite area, partly with downward equilibrium line according to Figure 5.2 (central figure). With a certain content of the related element, there is a transformation-free, purely ferritic steel. An opposite effect provide austenite developers. In addition to carbon, the most typical member of this group is nickel. Element

Steel type, no.

Effect

Carbon l l l Chromium l

All types l l l

Increases the strength, supports development of precipitants which reduce corrosion resistance, increasing C content reduces critical cooling rate

All types l

Works as ferrite developer, increases oxidation- and corrosion-resistance

Nickel l l

All types

Works as austenite developer, increases toughness at low temperature, grain-refining

Oxygen l

Special types l

Works as strong austenite developer (20 to 30 times stronger than Nickel)

Niobium l

1.4511,1.4550, 1.4580 u.a.

Binds carbon and decreases tendency to intergranular corrosion

Increases austenite stabilization, reduces hot crack tendency by formation of manganese sulphide Improves creep- and corrosion-resistance Molybdenum 1.4401,1.4404, l 1.4435 and others. against reducing media, acts as ferrite l developer l 1.4005, 1.4104, Phosphorus, 1.4305 Improve machinability, lower weldability, selenium, or l reduce slightly corrosion resistance l sulphur l

Silicon l

l

Titanium l l

All types l l

the austenite area to Figure 5.2 (right figure) and form a purely austenitic and transforma-

All types l l

Manganese l l

Austenite developers cause an extension of

tion-free steel. The table in Figure 5.3 summarises the effects of some selected elements on high alloy steels.

Improves scale resistance, acts as ferrite developer, all types are alloyed with small contents for desoxidation

Aluminium l

1.4510, 1.4541, Binds carbon, decreases tendency to 1.4571 and others intergranular corrosion, acts as a grain refiner l and as ferrite developer Type 17-7 PH Works as strong ferrite developer, mainly l used as heat ageing additive

Copper l l l

Type 17-7 PH, 1.4505, 1.4506 l l

Improves corrosion resistance against certain media, decreases tendency to stress corrosion cracking, improves ageing

br-er06-03e.cdr

© ISF 2002

Effects of Some Elements in Cr-Ni Steel

Figure 5.3

The binary system Fe-Cr in Figure 5.4 shows the influence of chromium on the iron lattice. Starting with about 12% Cr, there is no more transformation into the cubic face-centred lattice, the steel solidifies purely as ferritic. In the temperature range between 800 and 500°C this system contains the intermetallic σ-phase, which decomposes in the lower temperature range into a low-chromium α-

Figure 5.4

solid solution and a chromium-rich α’-solid solution. Both, the development of the σ-phase and of the unary α-α’-decomposition cause a

5. Welding of High-Alloy Steels, Corrosion

60

strong embrittlement. With higher alloy steels, the diffusion speed is greatly reduced, therefore both processes require a relatively long dwell time. In case of technical cooling, such embrittlement processes are suppressed by an increased cooling speed. Nickel is a strong austenite developer, see Figure 5.5 Nickel and iron develop in this system under elevated temperature a complete series of face-centred cubic solid solutions. Also in 1600 °C 1400

d

Fe Ni3

the binary system Fe-Ni S+d

S+g

decomposition

d+g

in the lower temperature

1200

range take place.

g

Temperature

processes

1000

Along two cuts through the

800

ternary system Fe-Cr-Ni, 600 a

Figure 5.6 shows the most

a+g

400

important

Fe Ni3 200

phases

which

develop in high alloy steels.

0 Fe

20

10

30

50

40

60

70

80

90 % Ni

Nickel

br-er-06-05e.cdr

© ISF 2002

A solidifying alloy with 20%

Binary System Fe - Ni

Cr and 10% Ni (left figure) forms at first δ-ferrite. δ-

Figure 5.5

ferrite is, analogous to the 60 % Fe

70 % Fe 1600 °C

1600 °C

S

1500

S+g

S+d

1400

S+d+g

1400

from the melt solidifying

S+g

S+d

body-centred

1300

1200 g

d+g

cubic

solid

solution. However α-ferrite

1200 d

g

d+g

d

1100

1100

is developed by transfor-

1000

1000

mation of the austenite, but

900

900

800

800

700

d+s

d+ g+ s

Temperature

1300

Fe-C diagram, the primary

S

1500

S+d+g

d+g+s

is of the same structure g+s

d+s

g+s

from the crystallographic

700

0

5

10

15

30

25

20

15

20 % Ni 10 % Cr

0

5

10

15

20

40

35

30

25

20

% Ni

15

% Cr

point of view, see Figure

© ISF 2002

br-er-06-06e.cdr

Sections of the Ternary System Fe-Cr-Ni

Figure 5.6

25

5.4.

5. Welding of High-Alloy Steels, Corrosion

61

During an ongoing cooling, the binary area ferrite + austenite passes through and a transformation into austenite takes place. If the coolls

ing is close to the equilibrium, a partial transst ee rri tic

ee st

takes place in the temperature range below

st

st en

en

iti c

si tic

Au

Au

800°C. Primary ferritic solidifying alloys show

4.

3.

2.

iti cfe

ls

ls st ee

ls st ee

M ar te n

rri tic Fe 1.

formation of austenite into the brittle α-phase

C

£ 0.1

0.1 1.2

£ 0.1

£ 0.1

Si

max. 1.0

max. 1.0

max. 1.0

max. 1.0

Mn

max. 1.0

max. 1.5

max. 2.0

max. 2.0

Cr

15 18

12 18

17 26

24 28

Mo

up to 2.0

up to 1.2

up to 5.0

up to 2.0

Ni

£ 1.0

£ 2.5

7 26

4 7.5

a reduced tendency to hot cracking, because δ-ferrite can absorb hot-crack promoting elements like S and P. However primary austenitic solidifying alloys show, starting at a certain

up to 2.2

Cu Nb

+

+

Ti

+

+

Al

+

alloy content, no transformations during cool+

ing (14% Ni, 16% Cr, left figure). Primary austenitic solidifying alloys are much more susceptible to hot cracking than primary fer-

+

V

+ indicates that the alloy elements can be added in a defined content to achieve various characteristics

+

N +

S

ritic solidifying alloys, a transformation into the

+

br-er06-07e.cdr

σ-phase normally does not take place with

© ISF 2002

Typical Alloy Content of High-Alloy Steels

these alloys. Figure 5.7 shows some typical compositions

Figure 5.7

of certain groups of high alloy steels.

The diagram of Strauß and Maurer in Figure 5.8 shows the influence on the microstructure formation of steels with a C-content of 0,2%. The classification of high-alloy steels in Figure 5.1 is based on this dia-

28

gram. If a steel only con-

% 24

tains C, Cr and Ni, the austenite

Nickel

20

lowest austenite corner will

16

be at 18% Cr and 6% Ni.

12

And also other elements

8

austen

4

ensite

martensite / troostite / sorbite ferrite / perlite

0

ite / ma rt

0

2

4

austenite / ferrite

austenite

/ martens

ite / ferrite

martensite / ferrite 6

8

10

12 14 Chromium

16

18

20

22 © ISF 2002

br-er-06-08e.cdr

Maurer - Diagram

24 % 26

than Ni and Cr work as an austenite or ferrite developer.

The

these

elements

is

of de-

scribed by the so-called chromium

Figure 5.8

influence

and

nickel

5. Welding of High-Alloy Steels, Corrosion

62

equivalents. The Schaeffler diagram reflects additional alloy elements, Figure 5.9. It represents molten weld metal of high alloy steels and determines the developed microstructures after cooling down from very high temperatures. The diagram was always prepared considering identical cooling conditions, the influence of different cooling speeds is here disregarded. The areas 1 to 4 in this diagram limit the chemical compositions of steels, where specific defects may occur during welding. Depending on the composition, purely ferritic chromium steels have a tendency to embrittlement by martensite and therefore to hot cracking (area 2) or to embrittlement due to strong

Nickel-equivalent = %Ni + 30x%C + 0,5x%Mn

grain growth (area 1). A cause for this strong grain growth during welding is the greatly increased diffusion speed in the ferrite compared with austenite. After reaching

a

temperature,

diffusion-start Figure

5.10

30 28 26

0%

24

austenite

t rri Fe 5%

%

10

22 20

20

A+F

16

%

40%

18

A +M

14

80 %

12 10 8

100%

2

martensite F + M

00

2

6 4

4

6

A+M+F M+F ferrite 8

10

12 14 16

18

20 22 24

26 28

30 32

34

36 38

40

Chromium-equivalent = %Cr + %Mo + 1,5x%Si + 0,5x%Nb

shows that ferritic steels have

a

hardening crack susceptibility (preheating to 400°C!) hot cracking susceptibility above 1250°C

considerably

grain growth above 1150°C © ISF 2002

br-er-06-09e.cdr

stronger grain growth than

Schaeffler Diagram With Border Lines of Weld Metal Properties to Bystram

austenites. Therefore high alloyed ferritic steels are to

sigma embrittlement between 500-900°C

Figure 5.9

be considered as of limited weldability.

6000 m²

The area 3 marks a possible

5000

embrittlement of the material due to the development of σ-phase. As explained in 5.6, this risk occurs with increased increased

ferrite

contents,

chromium

grain size

4000

3000

2000

1000 ferritic steel

con-

tents, and sufficiently slow

austenitic steel

0

200

400

600

800

1000

°C

temperature

cooling speed.

br-er-06-10e.cdr

© ISF 2002

Grain Size as a Function of Temperature

Figure 5.10

1200

5. Welding of High-Alloy Steels, Corrosion

63

Finally, area 4 marks the strongly increased tendency to hot cracking in the austenite. Reason is, that critical elements responsible for hot cracking like e.g. sulphur and phosphorous have only very limited solubility in the austenite. During welding, they enrich the melt residue, promoting hot crack formation (see also chapter 9 - Welding Defects). There is a Z-shaped area in the centre of the diagram which does not belong to any other endangered area. This area of chemical composition represents the minimum risk of welding defects, therefore such a composition should be adjusted in the weld metal. Especially when welding austenitic steels one tries to aim at a low content of δ-ferrite, because it has a much greater solubility of S and P, thus minimising the risk of hot cracking. The Schaeffler diagram is not only used for determining the microstructure with known chemical composition. It is also possible to estimate the developing microstructures when welding different materials with or without filler metal. Figures 5.11 and 5.12 show two examples for a determination of the weld metal microstructures of so-called 'black and white' joints.

28 28 24 10

A

9 8

40

3 ² : ·=1:1

80

20%

1

3

A+F 100 %

A+M+F

²

M+F

+

F

·

F

Nickel-equivalent

· M

12

20

20

A+M 16

40 M

12

· 20% 123

A+M

² : ·=1:1 +

8 4

4

8

12

16

20

24

28

32

36

S235JR (St 37)

·

Welding consumable

0

F

4

8

12 16 20 24 Chromium-equivalent

28

32

· X8Cr17 (W.-Nr. 1.4510) 21% Cr, 14% Ni, 3% Mo

²

S235JR (St 37)

·

Welding consumable

9

Weld metal under 30 % dilution (= base metal amount)

br-er06-11e.cdr

© ISF 2002

· X10CrNiTi18-9 (W.-No. 1.4541) 21% Cr, 14% Ni, 3% Mo

Weld metal under 30 % dilution (= base metal amount)

br-er06-12e.cdr

© ISF 2002

Application Example of Schaeffler - Diagram

Application Example of Schaeffler - Diagram

Figure 5.11

100 %

A+M+F

0

Chromium-equivalent ²

A+F

M+F

F 0

80 ·

²

0

9

10

9

Nickel-equivalent

20

16

4

24 A

20

Figure 5.12

36

5. Welding of High-Alloy Steels, Corrosion

64

The ferrite content can only be measured with a relatively large dispersal, therefore DeLong proposed to base a measurement procedure on standardized specimens. Such a system makes it possible to measure comparable values which don't have to match the real ferrite content. Based on these measurement values, the ferrite content is no longer given in percentage, but steels are grouped by ferrite numbers. In addition to ferrite numbers, DeLong proposed a reworked Schaeffler diagram where the ferrite number can be determined by the chemical composition, Figure 5.13. Moreover, DeLong has considered the influence of nitrogen as a strong austenite developer (effects are comparable with influence of carbon). Later on, nitrogen was included into the nickel-equivalent of the Schaeffler diagram. Nickel-equivalent = %Ni + 30 x %C + 30 x %N + 0,5 x %Mn

21 20

te rri fe

19

nu

austenite

18

16 15 14 13 12 11 10 16

of high alloy steels is their 4 6

d re su ea ym all .-% tic vol e n in ag s m nt 0% ly te er con 2% rm fo rrite 4% Sc e f ha effl 6% % er6 au 7, 2% ste nite 9, 7% , -m art 10 ,3% en site 12 ,8% -lin 13 e

17

The most important feature

r be m 0 2

corrosion resistance start-

8 10 12 14 16 18

ing with a Cr content of 12%. In addition to the problems during welding described by the Schaeffler

austenite + ferrite

diagram, these steels can 17

18

26

25 19 20 21 22 23 24 Chromium-equivalent = %Cr + %Mo + 1,5 x %Si + 0,5 x %Nb

27

© ISF 2002

br-er-06-13e.cdr

be negatively affected with view to their corrosion re-

De Long Diagram

sistance caused by the Figure 5.13

welding process.

Figure

air O

5.14 shows schematically

2Fe+++O+H2O ® 2Fe++++2OH-

the processes of electro-

OHFe+++

lytic

corrosion

under

water

a

drop of water on a piece of

O2

OH H2O

iron. In such a system a

2Fe++

cathode anode

4e-

potential difference is a

2Fe ® 2Fe+++4e-

precondition for the development of a local element

Fe(OH)3

O2+2H2O+4e ® 4OH -

iron

-

© ISF 2002

br-er-06-14e.cdr

consisting of an anode and

Corrosion Under a Drop of Water

a cathode. To develop Figure 5.14

5. Welding of High-Alloy Steels, Corrosion

65

such a local element, a different orientation of grains in the steel is sufficient. If a potential difference under a drop of water is present, the chemically less noble part reacts as an anode, i.e. iron is oxidised here and is dissolved as Fe2+-ion together with an electron emission. Caused by oxygen access through the air, a further oxidation to Fe3+ takes place. The cathodic, chemically nobler area develops OH- ions, absorbing oxygen and the electrons. Fe3+and OH--ions compose into the water-insoluble Fe(OH)3 which deposits as rust on the surface (note: the processes here described should serve as a principal explanation of electrochemical corrosion mechanisms, they are, at best, a fraction of all possible reactions). If the steel is passivated by chromium, the corrosion protection is provided by the development of a very thin chromium oxide layer which separates the material from the corrosive medium. Mechanical surface damages of this layer are completely cured in a very short time.

passive layer

active dissolution

passive layer

gap tensile stress

active dissolution of the crack base pitting corrosion passive layer

stress corrosion cracking passive layer activly dissolved grain boundary chromium depleted zones

active dissolution of the gap crevice corrosion

grain boundary carbides intergranular corrosion

incorrect

br-er06-15e.cdr

Figure 5.15

© ISF 2002

br-er06-16e.cdr

correct

© ISF 2002

Figure 5.16

The examples in Figure 5.15 are more critical, since a complete recovery of the passive layer is not possible from various reasons.

5. Welding of High-Alloy Steels, Corrosion

66 If crevice corrosion is present, corrosion products built up in the root of the gap and oxygen has no access to restore the passive layer. Thus narrow gaps where the corrosive medium can accumulate are to be avoided by introducing a suitable design, Figure 5.16.

br-er-06-17e.cdr

Pitting Corrosion of a Steel Storage Container

With pitting corrosion, the

Figure 5.17

chemical composition of the attacking medium causes a

local break-up of the passive layer. Especially salts, preferably Cl—ions, show this behaviour. This local attack causes a dissolution of the material on the damaged points, a depression develops. Corrosion products accumulate in this depression, and the access of oxygen to the bottom of the hole is obstructed. However, oxygen is required to develop the passive layer, therefore this layer cannot be completely cured and pitting occurs, Figure 5.17. Stress-corrosion cracking occurs when the material displaces under stress and the passive layer tears, Figure 5.18. Now the unprotected area is subjected to corrosion, metal is dissolved and the passive layer redevelops (figures 13). The repeated displace1

2

3

4

5

6

ment

and

repassivation

causes a crack propagation. 7

8

9

offset;

passive layer;

10

11

metal surface;

dislocation

12

Stress

cracking

corrosion

takes

mainly

place in chloride solutions. The crack propagation is transglobular, i.e. it does

br-er-06-18e.cdr

Model of Crack Propagation Through Stress Corrosion Cracking

Figure 5.18

not

follow

boundaries.

the

grain

5. Welding of High-Alloy Steels, Corrosion

67

Figure 5.19 shows the expansion-rate dependence of stress corrosion cracking. With very low expansion-rates, a curing of the passive layer is fast enough to arrest the crack. With very high expansion-rates, the failure of the specimen originates from a ductile fracture. In the intermediate range, the material damage is due to stress corrosion cracking. Figure 5.20 shows an example of crack propagation at transglobular stress corrosion cracking. A crack propagation speed is between 0,05 to 1 mm/h for steels with 18 - 20% Cr and 8 20% Ni. With view to welding it is important to know that already residual welding stresses

Sensitivity to stress corrosion cracking

may release stress corrosion cracking.

complete cover layer

tough fracture

T=RT

SpRK

e·2

e·1 Elongation speed e

br-er06-19E.cdr

·

© ISF 2002

br-er06-20e.cdr

Transgranular Stress Corrosion Cracking

Influence of Elongation Speed on Sensitivity to Stress Corrosion Cracking

Figure 5.19

© ISF 2002

Figure 5.20

The most important problem in the field of welding is intergranular corrosion (IC). It is caused by precipitation of chromium carbides on grain boundaries. Although a high solubility of carbon in the austenite can be expected, see Fe-C diagram, the carbon content in high alloyed Cr-Ni steels is limited to approximately 0,02% at room temperature, Figure 5.21.

5. Welding of High-Alloy Steels, Corrosion

68 The reason is the very high affinity of chromium to carbon, which causes the precipita-

to Bain and Aborn

Heat treatment temperature

1200

tion of chromium carbides Cr23C6 on grain

°C 1100

boundaries, Figure 5.22. Due to these precipitations, the austenite grid is depleted of

1000

chromium content along the grain boundaries A

900

and the Cr content drops below the parting limit. The diffusion speed of chromium in aus-

800

tenite is considerably lower than that of car700

bon, therefore the chromium reduction cannot

600 0

0.05

0.1 0.15 0.2 Carbon content

0.25 % 0,3

be compensated by late diffusion. In the depleted areas along the grain boundaries (line 2 in Figure 5.22) the steel has become susceptible to corrosion.

br-er06-21e.cdr

© ISF 2002

Carbon Solubility of Austenitic Cr - Ni Steels

Only after the steel has been subjected to sufficiently long heat treatment, chromium will

Figure 5.21

diffuse to the grain boundary and increase the

C concentration along the 1 - homogenuous starting condition 2 - start of carbide formation 3 - start of concentration balance 4 - regeneration of resistance limit

grain boundary (line 3 in Figure 5.22). In this way, the corrosion

resis-

tance can be restored (line 4 in Figure 5.22). Figure 5.23 explains why the IC is also described as intergranular

2 4

Chromium content of austenite

complete

1

resistance limit 3

disintegration. br-er-06-22e.cdr

Distance from grain boundary

Due to dissolution of deSensibility of a Cr - Steel

pleted areas along the grain boundary, complete grains break-out of the steel.

Figure 5.22

© ISF 2002

5. Welding of High-Alloy Steels, Corrosion

69

The precipitation and repassivation

mechanisms

described in Figure 5.22 are covered by intergranular corrosion diagrams according to Figure 5.24. Above a certain temperature carbon remains dissolved in the austenite © ISF 2002

br-er-06-23e.cdr

(see also Figure 5.21).

Grain Disintegration

Below this temperature, a carbon precipitation takes

Figure 5.23

place. As it is a diffusion controlled

process,

the

precipitation occurs after a incubation

time

which depends on temperature (line 1, precipitation characteristic curve). During stoppage at a constant

temperature,

the

3 ¬ Reciprocal of heat treatment temperature 1/T

certain

unsaturated austenite

2

austenite chromium carbide (M23C6) no intergranular disintegration

austenite + chromium caride (M23C6) sensitive to intergranular disintegration

oversaturated austenite

1

parting limit of the steel is Heat treatment time (lgt)

regained by diffusion of chromium.

br-er-06-24e.cdr

1 incubation time 2 regeneration of resistance limit 3 saturation limit for chromium carbide

© ISF 2002

Area of Intergranular Disintegration of Unstabilized Cr - Steels

Figure 5.24

Figure 5.25 depicts characteristic precipitation curves of a ferritic and of an austenitic steel. Due to the highly increased diffusion speed of carbon in ferrite, shifts the curve of carbon precipitation of this steel markedly towards shorter time. Consequently the danger of intergranular corrosion is significantly higher with ferritic steel than with austenite.

5. Welding of High-Alloy Steels, Corrosion

70

As carbon is the element that triggers the intergranular corrosion, the intergranular corrosion diagram is relevantly influenced by the c content, Figure 5.26. By decreasing the carbon content of steel, the start of carbide precipitation and/or the start of intergranular corrosion are shifted towards

lower temperatures

and

longer

quench temperature

times. This fact initiated the development of

precipitation curves for 17% Cr steel

ELC-steels

(Extra-Low-Carbon)

18-8-Cr-Ni steel

Tempering temperature

so-called

where the C content is decreased to less than 0,03% During welding, the considerable influence of

cooling curve

carbon is also important for the selection of the shielding gas, Figure 5.27. The higher the CO2-content

of

the

shielding

gas,

Tempering time

the br-er06-25e.cdr

stronger is its carburising effect. The C-

Precipitation Curves of Various Alloyed Cr Steels

content of the weld metal increases and the steel becomes more susceptible to inter-

© ISF 2002

Figure 5.25

granular corrosion. An often used method to

1000 °C 900

avoid intergranular corro-

800

sion is a stabilisation of the steel by alloy elements like

700

Temperature

0.07%C

0.05%C 0.03%C

niobium and titanium, Fig-

600

ure 5.28. The affinity of

0.025%C

these elements to carbon is

500

significantly

higher

than

that of chromium, therefore 400 1 10 br-er-06-26e.cdr

102

104

103

105

Time

Influence of C-Content on Intergranular Disintegration

s

106 © ISF 2002

carbon is compounded into Nb- and Ti-carbides. Now carbon cannot cause any

Figure 5.26

chromium depletion. The

5. Welding of High-Alloy Steels, Corrosion

71

proportion of these alloy elements depend on the carbon content and is at least 5 times higher with titanium and 10 times higher with niobium than that of carbon. Figure 5.28 shows the effects of a stabilisation in the intergranular corrosion diagram. If both steels are subjected to the same heat treatment (1050°C/W means heating to 1050°C and subsequent water quenching), then the area of intergranular corrosion will shift due to stabilisation to significantly longer times. Only with a much higher heat treatment temperature the intergranular corrosion accelerates again. The cause is the dissolution of titanium carbides at sufficiently high temperature. This carbide dissolution causes problems when welding stabilised steels. During welding, a narrow area of the HAZ is heated above 1300°C, carbides are dissolved. During the subsequent cooling and the high cooling rate, the carbon remains dissolved.

0.058 % C 0.53 % Nb Nb/C = 9

°C 600

0.030 % C 0.51 % Nb Nb/C = 17

0.018 % C 0.57 % Nb Nb/C = 32

M2

550 M1 500

S1

450

Heat treatment temperature

Heat treatment temperature

700

0,5

1

2,5

5

10

50

25

100

250

h

600 550 500

1000

Heat treatment time

Heat treatment temperature

A r [% ]

C O2

O2

S 1

99

/

1

M 1

90

5

5

M 2

82

18

/

br-er06-27e.cdr

h

10000

unstabilized

650

1300°C /W

600

1050°C /W

550 500 450 0,3

© ISF 2002

1000

800 °C 700

1 3 W.-No.:4541

X5CrNiTi18-10

10

30

100 Time

300

1000

h

10000

stabilized

br-er06-28e.cdr

© ISF 2002

Influence of Stabilization on Intergranular Disintegration

Influence of Shielding Gas on Intergranular Disintegration

Figure 5.27

300

X5CrNi18-10

C o m p o sitio n S hie ld ing g a s

1050°C /W

650

450 0,3 1 3 10 30 100 Time W.-No.:4301 (0,06%)

400 0,2

800 °C 700

Figure 5.28

If a subsequent stress relief treatment around 600°C is carried out, carbide precipitations on grain boundaries take place again. Due to the large surplus of chromium compared with niobium or titanium, a partial chromium carbide precipitation takes place, causing again inter-

5. Welding of High-Alloy Steels, Corrosion

72

granular susceptibility. As this susceptibility is limited to very narrow areas along the welded joint, it was called knife-line attack because of its appearance. Figure 5.29. In stabilised steels, the chromium carbide represents an unstable phase, and with a sufficiently long heat treatment to transform to NbC, the steel becomes stable again. The stronger the steel is over-stabilised, the lower is the tendency to knife-line corrosion. Nowadays the importance of Nickel-Base-Alloys increases constantly. They are ideal materials when it comes

to

components

which are exposed to special conditions: high temperature, corrosive attack, low temperature, wear rebr-er-06-29e.cdr

sistance, or combinations

Knife-Line Corrosion

hereof. Figure 5.30 shows one of the possible group-

Figure 5.29

ing of nickel-base-alloys. Materials listed there are selected examples, the total number of available materials is many times higher. Group A consists of nickel alloys. These alloys are Alloy

Chem. composition

Alloy

Nickel 200

Ni 99.6, C 0.08

Duranickel 301 Ni 94.0, Al 4.4, W 0.6

Nickel 212 Nickel 222

Ni 97.0, C 0.05, Mn 2.0 Ni 99.5, Mg 0.075

Incoloy 925 Ni 42.0, Fe 32.0, Cr 21.0, Mo 3.0, W 2.1, Cu 2.2, Al 0.3 Ni-Span-C 902 Y2O3 0.5, Ni 42.5, Fe 49.0, Cr 5.3, W 2.4, Al 0.5

Monel 400

Ni 66.5, Cu 31.5

Monel K-500

Ni 65.5, Cu 29.5, Al 2.7, Fe 1.0, W 0.6

Monel 450

Ni 30.0, Cu 68.0, Fe 0.7, Mn 0.7

Inconel 718

Ni 52.0, Cr 22.0, Mo 9.0, Co 12.5, Fe 1.5, Al 1.2

Ferry Group C

Ni 45.0, Cu 55.0

Inconel X-750 Ni 61.0, Cr 21.5, Mo 9.0, Nb 3.6, Fe 2.5 Nimonic 90 Ni 77.5, Cr 20.0, Fe 1.0, W 0.5, Al 0.3, Y2O3 0.6

Inconel 600

Ni 76.0, Cr 15.5, Fe 8.0

Nimonic 105

Ni 76.0, Cr 19.5, Fe 112.4, Al 1.4

Nimonic 75

Ni 80.0, Cr 19.5

Incoloy 903

Ni 39.0, Fe 34.0, Cr 18.0, Mo 5.2, W 2.3, Al 0.8

Nimonic 86

Ni 64.0, Cr 25.0, Mo 10.0, Ce 0.03

Incoloy 909

Ni 58.0, Cr 19.5, Co 13.5, Mo 4.25, W 3.0, Al 1.4

Incoloy 800

Ni 32.5, Fe 46.0, Cr 21.0, C 0.05

Inco G-3

Ni 38.4, Fe 42.0, Cu 13.0, Nb 4.7, W 1.5, Al 0.03, Si 0.15

Incoloy 825

Ni 42.0, Fe 30.0, Cr 21.5, Mo 3.0, Cu 2.2, Ti 1.0

Inco C-276

Ni 38.4, Fe 42.0, Cu 13.0, Nb 4.7, W 1.5, Al 0.03, Si 0.4

Inco 330

Ni 35.5, Fe 44.0, Cr 18.5, Si 1.1

Group E

Group A

Chem. Composition

characterized by moderate

Group D1

Group B

Group D2

Monel R-405

mechanical strength and high degree of toughness. They can be hardened only by cold working. The alloys are quite gummy in the annealed or hot-worked con-

Ni 66.5, Cu 31.5, Fe 1.2, Mn 1.1, S 0.04

dition,

and

cold-drawn

© ISF 2002

br-er-06-30e.cdr

material is recommended Typical Classification of Ni-Base Alloys

Figure 5.30

for best machinability and smoothest finish.

5. Welding of High-Alloy Steels, Corrosion

73

Group B consists mainly of those nickel-copper alloys that can be hardened only by cold working. The alloys in this group have higher strength and slightly lower toughness than those in Group A. Cold-drawn or cold-drawn and stress-relieved material is recommended for best machinability and smoothest finish. Group C consists largely of nickel-chromium and nickel-iron-chromium alloys. These alloys are quite similar to the austenitic stainless steels. They can be hardened only by cold working and are machined most readily in the cold-drawn or cold-drawn and stress-relieved condition. Group D consists primary of age-hardening alloys. It is divided into two subgroups: D 1 – Alloys in the non-aged condition. D 2 – Aged Group D-1 alloys plus several other alloys in all conditions. The alloys in Group D are characterized by high strength and hardness, particularly when aged. Material which has been solution annealed and quenched or rapidly air cooled is in the softest condition and does machine easily. Because of softness, the non-aged condition is necessary for trouble free drilling, tapping and all threading operations. Heavy machining of the age-hardening alloys is best accomplished when they are in one of the following conditions: 1. Solution annealed 2. Hot worked and quenched or rapidly air cooled Group E contains only one material: MONEL R-405. It was designed for mass production of automatically machined screws. Due to the high number of possible alloys with different properties, only one typical material of group D2 is discussed here: Material No. 2.4669, also known as e.g. Inconel X-750. The aluminium and titanium containing 2.4669 is age-hardening through the combination of these elements with nickel during heat treatment: gamma-primary-phase (γ') develops which is the intermetallic compound Ni3(Al, Ti). During solution heat treatment of X-750 at 1150°C, the number of flaws and dislocations in the crystal is reduced and soluble carbides dissolve. To achieve best results, the material

5. Welding of High-Alloy Steels, Corrosion

74

should be in intensely worked condition before heat treatment to permit a fast and complete recrystallisation. After solution heat treatment, the material should not be cold worked, since this would generate new dislocations and affect negatively the fracture properties. The creep rupture resistance of X-750 is due to an even distribution of the intercrystalline γ' phase. However, fracture properties depend more on the microstructure of the grain boundaries. During an 840°C stabilising heat treatment as part of the triple-heat treatment, the fine γ' phase develops inside the grains and M23C6 precipitates onto the grain boundaries. Adjacent to the grain boundary, there is a γ' depleted zone. During precipitation hardening (700°C/20 h) γ' phase develops in these depleted zones. γ' particles arrest the movement of dislocations, this leads to improved strength and creep resistance properties. During the M23C6 transformation, carbon is stabilised to a high degree without leaving chromium depleted areas along the grain boundaries. This stabilisation improves the resistance of this alloy against the attack of several corrosive media. With a reduction of the precipitation temperature from 730 to 620°C – as required for some special heat treatments – additional γ' phase is precipitated in smaller particles. This enhances the hardening effect and improves strength characteristics. Further metallurgical discussions about X-750, can be taken from literature, especially with view to the influence of heat treatment on fracture properties and corrosion behaviour.

The recommended processes for welding of X-750 are tungsten inert gas, plasma arc, electron beam, resistance, and pressure oxy arc welding. During TIG welding of INCONEL X-750, INCONEL 718 is used as welding consumable. Joint properties are almost 100% of base material at room temperature and about 80% at 700° 820°C. Figure 5.31 shows typical strength properties of a welded plate at a temperature range between -423° and 1500°F (-248 – 820°C). Before welding, X-750 should be in normalised or solution heat treated condition. However, it is possible to weld it in a precipitation hardened condition, but after that neither the seam nor the heat affected zone should be precipitation hardened or used in the temperature range of precipitation hardening, because the base material may crack. If X-750 was precipitation hardened and then welded, and if it is likely that the workpiece is used in the temperature range of precipitation hardening, the weld should be normalised or once again precipitation hardened. In any case it must be noted that heat stresses are minimised during assembly or welding.

5. Welding of High-Alloy Steels, Corrosion

75

X-750 welds should be solution heat treated before a precipitation hardening. Heating-up speed during welding must be from the start fast and even touching the temperature range of precipitation hardening only as briefly as possible. The best way for fast heating-up is to insert the welded workpiece into a preheated furnace. Sometimes a preheating before welding is advantageous – if the component to be welded has a poor accessibility, or the welding is complex, and especially if the assembly proves to be too complicated for a post heat treatment. Two effective welding preparations are: 1. 1550°F/16 h, air cooling 2. 1950°F/1 h, furnace cooling with 25°-100°F/h up to 1200°F, air A repair welding of already fitted parts should be followed by a solution heat treatment (with a fast heating-up through the temperature range of precipitation hardening) and a repeated precipitation hardening. A cleaning of intermediate layers must be carried out to remove the oxide layers which are formed during welding. (A complete isolation of the weld metal using gas shielded processes is hardly possible). If such films are not removed on a regular basis, they can become thick enough to cause material separations together with a reduced strength. Brushing with wire brushes only polishes the surface, the layer surface must be sand-blasted or ground with abrasive material. The frequency of cleaning depends on the mass of the developed oxides. Any sand must be removed before the next layer is welded. X-750 can be joined also by spot-, projection-, seam-, and flash butt welding. The welding equipment must be of adequate performance. X-750 is generally resistance welded in normalized or solution heat treated condition. Figure 5.31

6. Welding of Cast Materials

6. Welding of Cast Materials

77

Figure 6.1 provides a summary of the different cast iron materials.

In

this

connection it is only referred to cast iron, cast steel and malleable

steel,

as

special cast materials,

due

to their

poor weldability, are of no importance in Figure 6.1

welding.

Designation according to the material code (DIN EN 1560)

Figure 6.2 shows the designation of the cast material in accordance with

e.g.: EN-GJ L F – 150

DIN EN 1560. A distinction is made 1 Position 1: Position 2: Position 3: Position 4: Position 5:

EN GJ L F 150

Position 6:

-

2 34

5

between the designation “according to

standardised material cast material graphite structure (lamellar graphite) microstructure (ferritic) mechanical properties (Rm= 150 N/mm2) chemical composition (high alloyed) optionally

the material code” and the designation “according to the material number”. In Figure 6.2, examples of two materials are specified.

Designation according to the material number

e.g.: EN- J L 1271 1 23 Position 1: Position 2: Position 3: Position 4: Position 5: Position 6:

EN J L 1 27 1

-

4,5,6

standardised material cast material graphite structure (lamellar graphite) number for the main characteristic material identification number special requirement

br-er07-02e.cdr

© ISF 2004

Designation of Materials

Figure 6.2

6. Welding of Cast Materials

78

Figure 6.3 depicts a survey of the mechanical properties and the chemical compositions of several customary cast materials. As to its analysis and mechanical properties which are very different from other cast materials, cast steel constitutes an exception to the rule. In Figure 6.4 the stable and the metastable iron-carbon diagram are shown. The differences between the cast material Iron Cast Material

Rp0,2 Rm 2 2 N/mm N/mm EN-GJL-300 EN-GJS-400-15 EN-GJMW-400-12 GS38 EN/ GJL/300 EN -GJS -400 -15 EN -GJMW -400 -12 GS 38

are best explained

Mechanical Properties

250 200 190

300 400 380 380

Chemical Analysis A % 15 12 25

C

Si

% % ˜ 2,8 ˜ 1,4 ˜ 3,7 ˜ 2,2 ˜ 3,2 ˜ 0,5 0,15 0,47

Mn

P

S

% % % ˜ 1,0 < 0,2 < 0,12 ˜ 0,5 ˜ 0,05 ˜ 0,01 ˜ 0,3 < 0,12 ˜ 0,25 0,35 0,045 0,054

spheroidal graphite

has

tween

carbon of

2,8

beand

4,5%. Through the © ISF 2004

Characteristics and Analyses of Cast Materials

with lamellar and

contents

- lamellar graphite cast iron - nodular graphite cast iron - decarburizing annealed malleable cast iron (former : white -heart malleable cast iron) - cast steel

br-er-07-03e.cdr

this way. Cast iron

addition of alloying elements,

above

all Si, these mateFigure 6.3

rials solidify following the stable system, i.e., the carbon is precipitated in

the

form

of

graphite. Malleable cast

iron

shows

similar C-contents, the

solidification

from

the

metal,

molten however,

follows the metastable system. The Figure 6.4

C-contents of cast steel, on the other

6. Welding of Cast Materials

79

hand, comply with those of common structural steels, i.e., they are, as a rule, below 0,8% C. The structure of a normalised cast iron which is composed of ferrite (bright) and pearlite (dark) is shown in Figure 6.5. Since the properties are similar to those of structural steels these materials are weldable, constructional welding is also possible. It is recommended to normalise the cast steel parts before welding. Through this type of heat treatment, on the one hand the transformation of the cast structure is ob-

br-er07-04e.cdr

© ISF 2002

Microstructure of Normally Glowed Cast Steel

tained, the residual stresses inside the workpiece are, on the other hand, reFigure 6.5

duced.

From a C-content in the steel cast of 0,15% up, it is recommended to carry out preheating during welding, the preheating temperature should follow the analysis of the material, the workpiece geometry and the welding method. After welding the cast workpieces are subject to stress-relief annealing. Figure 6.6 shows the structure of cast iron with lamellar graphite (grey cast iron). Apart from their carbon content, br-er07-05e.cdr

© ISF 2002

Microstructure of Lamellar Graphite Cast Iron

Figure 6.6

these materials are characterised by increased contents of S and P which

6. Welding of Cast Materials

80

improves castability. Besides the poor mechanical properties (elongation after fracture of approx. 1%), these chemical properties also impede welding with ordinary means. It is not possible to carry out constructional welding with grey cast iron. Repair welds of grey cast iron are, in contrast, carried out more frequently as damaged cast parts are not easily replaceable. For those repair welds, the cast parts must be preheated (entirely or partly) to temperatures of approx. 650°C. Heating and cooling must be done very slowly as the cast piece may be destroyed already by the thermal stresses. The highly liquid weld metal also constitutes a problem, and thus the molten pool must be supported by a carbon pile. Welding may be carried out with similar filler material (materials of the same composition as the base). If grey cast iron is to be welded without any preheating, the filler material must, as a rule, be dissimilar (of different composition to the base metal). During this type of welding, there are always strong structural changes in the region of the weld which lead to high hardening and high residual stresses. For the minimisation of these structural changes, a highly ductile filler material is applied. The heat input into the base material should be as low as possible. Figure 6.7 depicts the structural constitution of spheroidal graphite cast iron. The graphite spheroidization achieved

by

is the

addition of magnesium and cerium. As, with this type © ISF 2002

br-er-07-06e.cdr

Nodular Graphite Cast Iron

of

graphite,

the

notch actions are Figure 6.7

considerably lesser than this is

the case with grey cast iron, this type of cast iron is characterised by substantially better mechanical parameters with a considerably higher elongation after fracture and improved ductility. For this reason, the risk of material failure caused by weld residual stresses or thermal stresses is considerably reduced for spheroidal graphite

6. Welding of Cast Materials

81 cast iron. Frequently, nickel-based alloys are used as filler material. Problems occur in the HAZ where, besides the ledeburite eutectic alloy system, also Ni-Fe-martensite is frequently formed. Both structures lead to extreme hardening in the HAZ which can

be

removed

only

by

time-

consuming heat treatment.

br-er07-07e.cdr

© ISF 2002

Carburizing Annealed Malleable Cast Iron EN-GJMB-350

Figure 6.8

Figures 6.8 and 6.9 show the structures of Carburized Annealed Malleable Cast Iron (6.8) and of Decarburized Annealed Malleable Cast Iron (6.9). The composition of the malleable cast iron is thus that during solidification, the total of carbon is bound in cemenbr-er07-08e.cdr

tite and precipitated. During a subsequent

Decarburizing Annealed Malleable Cast Iron EN-GJMW-350

annealing process, the iron carbide disintegrates into graphite and iron.

© ISF 2002

Figure 6.9

6. Welding of Cast Materials

82 If annealing is carried out in neutral atmosphere, the structure of Carburized Annealed Malleable Cast Iron develops. Annealing in oxidising at-

Structure core zone : Perlit + (Ferrit) + temper carbon transition zone : Perlit + Ferrit + temper carbon surface zone : Ferrit

mosphere leads to the decarburisation of the workpiece surfaces and Decarburized Annealed Malleable Cast Iron is developed, Figure 6.10. Carburized

Annealed

Malleable

Cast Iron is not weldable. Decarburized Annealed Malleable Cast Iron, in contrast, is weldable.

white-heart malleable cast iron

br-er07-09e.cdr

© ISF 2002

Structure in dependence of the wall thickness

Figure 6.10

You can see in Figure 6.11 that, also with malleable cast iron, hardening in 200

the region of the HAZ occurs. For carrying out constructional welds made of material quality has been developed. Figure 6.11 shows that this material, EN-GJMW-400-12, is characterised by

Hardness after Brinell

malleable cast iron parts, a special

GTW-40

150

GTW-S38

100

considerably less hardening. This ma-

material thickness: 7 mm

terial is weldable without any problems up to a wall thickness of 8 mm.

Testspeciem 50

br-er0-10e.cdr

0

20 mm 10 Distance of center welding seam

30

© ISF 2002

Hardness Process within the Range of the Heat Influence Zone

Figure 6.11

7. Welding of Aluminium Alloys

7. Welding of Aluminium Alloys

84 Figure 7.1 compares basic physical properties

Property

Al

Fe

of steel and aluminium. Side by side with different mechanical behaviour, the following

Atomic weight

[g/Mol]

26.9

55.84

Specific weight

[g/cm³]

2.7

7.87

fcc

bcc

Lattice

differences are important for aluminium weld-

E-module

[N/mm²]

71*10³

210*10³

R pO,2 PO,2

[N/mm²]

ca. 10

ca. 100

R mm

[N/mm²]

ca. 50

ca. 200

spec. Heat capacity

[J/(g*°C)]

0.88

0.53

[°C]

660

1539

[W/(cm*K)]

2.3

0.75

Spec. el. Resistance

[nWm]

28-29

97

Expansion coeff.

[1/°C]

Melting point Heat conductivity

24*10

-6

12*10

Al2O 3

Melting point of oxydes

[°C]

2050

- considerably lower melting point compared with steel - three times higher heat conductivity - considerably lower electrical resistance

-6

FeO Oxydes

ing:

Fe 3O 4

- double expansion coefficient - melting point of Al203 considerably higher

Fe 2O 3

than that of Al; metal and iron oxide melt ap-

1400

proximately at the same temperature.

1600 (1455)

br-er08-01.cdr

© ISF 2002

Basic Properties of Al and Fe

Figure 7.2 compares some mechanical properties of steel with properties of some light metals. The important advantages of light

Figure 7.1

metals compared with steel are especially

shown in the right part of the figure. If a comparison should be based on an identical stiffness, then the aluminium supporting beam has a 1.44 times larger cross-section than the steel beam, however only about 50% of its weight. Figure 7.3 compares qualitatively the stress-strain diagram

of

Aluminium

and

steel. In contrast to steel, aluminium has a fcc (face centred

cubic)-lattice

at

room temperature. This is why there is no distinct yield point as being the case in a bcc (body centred cubic)lattice.

Aluminium

is

br-er-08-02.cdr

Deflexions and Weights of Cantilever Beams Under Load

not

subject to a lattice transFigure 7.2

7. Welding of Aluminium Alloys

85

formation during cooling, thus there is no structure transformation and consequently no danger of hardening in the heat affected zone as with steel.

4 cm 2

low carbon steel

200°C

400

1000 1200

600

800

1500

-2

Steel

-4

Stress

8 cm aluminium 6

100°C 200

4

Al-alloy

2 300 400 500 600 -2 -4 -6 -8 -18

Elongation br-er08-03.cdr

© ISF 2002

-16

-14

br-er08-04.cdr

Comparison of Stress-Elongation Diagrams of Al and Steel

Figure 7.3

-12

-10

-8

-6

-4

-2

0

2

cm

6

© ISF 2002

Isothermal Curves of Steel and Al

Figure 7.4

Figure 7.4 illustrates the effect of the considerably higher heat conductivity on the welding process compared with steel. With aluminium, the temperature gradient around the welding point is considerably smaller than with steel. Although the peak temperature during Al welding is about 900°C below steel, the isothermal curves around the welding point have a clearly larger extension. This is due to the considerably higher heat conductivity of aluminium compared with steel. This special characteristic of Al requires a input heat volume during welding equivalent to steel. Figure 7.5 lists the most important alloy elements and their combinations for industrial use. Due to their behaviour during heat treatment can Al-alloys be divided into the groups hardenable and non-hardenable (naturally hard) alloys.

7. Welding of Aluminium Alloys

86

Al Cu Mg

ing consumables. Al Mg Si

Cu

Aluminium alloys are often welded with conAl Zn Mg

sumable of the same type, however, quite Mg

often over-alloyed consumables are used to

Al Zn Mg Cu

678

Al alloys together with preferably used weld-

hardenable alloys

Figure 7.6 shows typical applications of some

Al

Zn

Al Si Cu

and to improve the mechanical properties of Al Si

the seam.

Si Al Mg

The classification of Al alloys into two groups

Al Mg Mn

Mn

is based on the characteristic that the group Al Mn

of the non-hardenable alloys cannot increase br-er08-05.cdr

the strength through heat treatment, in con-

678

Mg and Zn because of their low boiling point)

non-hardenable alloys

compensate burn-off losses (especially with

© ISF 2002

Classification of Aluminium Alloys

trast to hardenable alloys which have such a potential. The important hardening mechanism for this

Figure 7.5

second group is explained by the figures 7.7 und 7.8. Example: If an alloy containing about 4.2% Cu, which is stable at room temperature, is heat treated at 500°C, then, after a sufficiently long time, there will be only a single phase structure present. All alloy elements were dissolved, Figure 7.8 between point P and Q. When quenched to room Al - alloys Al99,5 AlCuMg1 AlMgSi0,5 AlSi5 AlMg3

AlMg2Mn0,8 AlMn1

Typical use electrical engineering mechanical engineering, food industries architecture, electrical engineering, anodizing quality architecture, anodizing quality architecture, apparatus-, vehicle-, shipbuilding engineering, furniture industry apparatus-, vehicle-, shipbuilding engineering apparatus-, vehicle-engineering, food industry

W elding consumable SG-Al 99,5Ti; SG-Al 99,5

tion, no precipitation will

SG-AlMg4,5Mn

take place. The alloy ele-

SG-AlMg5; SG-AlMg4,5Mn; SG-AlSi5 SG-AlSi5

ments are forced to remain dissolved, the crystal is out

SG-AlMg3; SG-AlMg4,5Mn SG-AlMg5; SG-AlMg3; SG-AlMg4,5Mn

of equilibrium. If such a structure is subjected to an

SG-AlMn1;SG-Al99,5T

age hardening at room or

base material - aluminium percentage of alloy elements without factor

elevated

temperature,

a

© ISF 2002

br-er-08-06.cdr

Use and Welding Consumables of Aluminium Alloys

Figure 7.6

temperature in this condi-

precipitation of a second phase takes place in ac-

7. Welding of Aluminium Alloys

87

cordance with the binary system, the crystal tries to get back into thermodynamical equilibrium. Depending on the level of

stable condition

solution heat treatment

repeated hardening

solidification of alloy elements in solid solution

hardening temperature, the

quenching

regeneration

oversaturated solid solution, metastable condition

precipitation takes place in

warm ageing

cold ageing (RT ageing)

ageing at slightly increased temperature coherent precipitations, cold aged condition

three possible forms: copartly coherent precipitations, warm aged condition

coherent and partly coherent precipitations, transition conditions cold ageing -- warm ageing temperature rise

temperature rise

herent particles (i.e. particles

longer warm ageing partly coherent and incoherent precipitations, softening

from

the

matrix in their chemical composition but having the

longer warm ageing stable incoherent equilibrium phase stable condition © ISF 2002

br-er-08-07.cdr

deviating

Ageing Mechanism

same

lattice

structure),

partly

coherent

particles

(i.e. the lattice structure of the matrix is partly re-

Figure 7.7

tained),

and

incoherent

particles (lattice structure completely different from the matrix), Figure 7.7. Coherent particles formed at room temperature can be transformed into incoherent particles by increase of temperature (i.e. enabling diffusion). The precipitations cause a restriction to the

700 liquid

dislocation movement in the matrix lattice, thus

liquid and solid Q

600

leading to an increase in strength. The finer the

copper containing aluminium solid solution 500

At an increased temperature (heat ageing, Fig-

Temperature

precipitations, the stronger the effect.

P

400

300

ure 7.7) a maximum of second phase has precipitated after elapse of a certain time. Consequently a prolonged stop at this tem-

aluminium solid solution and copper aluminide (Al2Cu)

200

100 copper content of AlCuMg

perature does not lead to an increased strength, but to coarsening of particles due to

0

1

2

3

4

5

mass-%

Copper

diffusion processes and to a decrease in strength (less bigger particles in an extended

br-er08-08.cdr

space).

© ISF 2002

Phase Diagram Al-Cu

Figure 7.8

7

7. Welding of Aluminium Alloys

88 After a very long heat ageing a stable condition is reached again with relatively large precipitations of the second phase in the matrix. In Figure 7.7 is this stable final condition iden-

Q

tical with the starting condition. A deteriorati-

solution heat treatment

500 P

on of mechanical properties only happens

°C

quenching

Temperature

400

during hot ageing, if the ageing time is excessively long.

300

200

heat ageing

The complete process of hardening at room

100

temperature is metallographic also called age age hardening

hardening, at elevated temperature heat age0

2

4

6

8

10

12

h

Time

14

ing. A decrease in strength at too long ageing time is called over-ageing. © ISF 2002

br-er08-09.cdr

Temperature - Time Distribution During Ageing

Figure 7.9 shows a schematic representation of time-temperature curves during hardening

Figure 7.9

Figure

with age hardening and heat ageing.

7.10

shows

the

380

strength increase of AlZnMg The difference between age hardening and heat ageing is here very clear. Due to improved

diffusion

condi-

tions is the strength increase

320 0.2% yield stress s0.2 in N/mm²

1 in dependence of time.

water quenching (~900°C/min) air cooling (~30°C/min)

260 120°C 200 RT 140

80 10-1

in the case of heat ageing much faster than in the case of

age

hardening.

quenched

100

101

10²

10³

Ageing time in h © ISF 2002

br-er-08-10.cdr

Increase of Yield Stress During Ageing of AlZnMg1

The

strength maximum is also reached considerably ear-

Figure 7.10

lier. The curve of hot ageing shows clearly the begin of strength loss when held at a too long stoppage time. This figure shows another specialty of the process of ageing. During ageing, a

7. Welding of Aluminium Alloys

89

second phase is precipitated from a single-phase structure. To initiate this process, the structure must contain nuclei of the second phase. However, a certain time is required to develop such nuclei. Only after formation of nuclei can the increase in strength start. The period up to this point is called incubation time. 500 110

N/mm²

Tensile strength sB

Figure 7.11 shows the effect of the height of ageing temperature level on both, mechanical properties of a hardenable Al-alloy and on in-

135

400

150 180

300

190 205

230

260°C

cubation time. The lower the ageing tempera-

200 110

N/mm² 400 0.2% yield stress s0.2

ture, the higher the resulting values of yield stress and tensile strength. If a low ageing temperature is selected, the ageing time as well as

135

300

150 180 190 205°C

200

the incubation time become extremely long.

230 260 Fracture elongation d2

Figure 7.11 shows that a the maximum yield stress is reached after a period of about one year under a temperature of 110°C. An in-

%

190

180

205

150

135

20 10

0

crease of the ageing temperature shortens the duration of the complete precipitation process

30

110°C 260

230 30 min

10

-2

10

-1

1 day 0

1

10 10 Ageing time

1 week

10

2

1 1 month year

103 h 104

br-er08-11.cdr

© ISF 2002

Influence of Ageing Temperature and -Time on Ageing

by a certain value raised by 1 to a power. On the other hand, such an acceleration of ageing leads to a lowering of the maximum strength.

Figure 7.11

As the lower part of the 400

figure shows, the fracture

N/mm²

elongation

Tensile strength Rm

300

is

counter-

AlMg5

proportional to the strength

AlMg3

values, i.e. the strength

200

increase caused by ageing is accompanied by an em-

100

brittlement of the material.

Al99,5

0 0

30

%

70

Age Hardening of Al Alloys

Figure 7.12

Strain © ISF 2002

br-er-08-12.cdr

7. Welding of Aluminium Alloys

90

Figure 7.12 shows a method of how to increase the strength of non-hardenable alloys. As no precipitations are present to reduce the movement of dislocations, such alloys can only be strengthened by cold working. Figure 7.12 illustrates two essential mechanisms of strength increase of such alloys. On 300

one hand, tensile strength increases with in-

N/mm²

creasing content of alloy elements (solid solu-

250

tion strengthening), on the other hand, this increase is caused by a stronger deformation

Rm or Rp0,2

200

of the lattice. 150

Figure 7.13 shows the effect of the welding process on mechanical properties of a cold-

0,7

100

worked alloy. Due to the heat input during

0,5 50 HV30

0,4

Rp0,2/Rm

0,6

(recovery), in addition, a grain coarsening will

0,3 0,2

0 80

60 40 20 0 20 40 Distance from Seam Centre

welding, the blocked dislocations are released start in the HAZ. This is followed by a strong

60 mm 100

drop in yield point and tensile strength. This

br-er08-13.cdr

strength loss cannot be overcome in the case

© ISF 2002

Non-Hardenable Al Alloy

of a welding process.

Figure 7.13 400

Figure

7.14

illustrates

the

90 days RT

N/mm²

Rm

350

mechanisms in the case of a

21 days RT

hardenable aluminium alloy. welding heat, the precipitations are solution heat treated

Rp0,2

250 90 days RT

Stress

As a consequence of the

1 day RT

300

21 days RT 200 4 mm plates of: AlZnMg1F32 start values: Rp0,2=263N/mm² Rm=363 N/mm² welding method: WIG, both sides, simultaneously welding consumable: S-AlMg5 specimens with machined weld bead

1 day RT

150

and the strength values de100

crease in the weld area. Due to the age hardening, a re-

50 80 br-er-08-14.cdr

strengthening of the alloys

40

20

20 60 0 40 Distance from seam centre

Hardenable Al Alloy

takes place with increasing time.

60

Figure 7.14

80

100

mm

140 © ISF 2002

7. Welding of Aluminium Alloys

91 Figure 7.15 shows another problematic nature of Alwelding. Due to the high thermal expansion of aluminium, high tensions develop during solidification of the weld pool in the course of the welding cycle. If the welded alloy indicates a high melting inter© ISF 2002

br-er-08-15.cdr

val, Hot Cracks in a Al Weld

cracks

may

easily

develop in the weld.

Figure 7.15

A relief can be afforded by preheating of the material, Figure 7.16. With an increasing preheat temperature, the amount of fractured welds decreases. The different behaviour of the three displayed alloys can be explained using the right part of the figure. One can see

100 %

that the manganese content

maximum of this hot crack

2 60 1 40

X X

3

20

susceptibility is likely with

Mg

Cracking susceptibility

hot crack susceptibility. The

Weld cracking tendency

influences significantly the

80

Si

X X

about 1% Mg content (corresponds with alloy 1). With increasing MG content, hot crack

susceptibility

0

100

300

Preheat temperature

400

°C

500 0

1

2

3

%

4

Alloy content 1: AlMgMn 2: AlMg 2,5 3: AlMg 3,5

© ISF 2002

br-er-08-16.cdr

de-

Influence of Preheat Temperature and Magnesium Content

creases strongly (see also alloy 2 and 3, left part).

200

Figure 7.16

To avoid hot cracking, partly very different preheat temperatures are recommended for the alloys. Zschötge proposed a calculation method which compares the heat conductivity conditions of the Al alloy with those of a carbon steel with 0.2% C. The formula is shown in Figure

7. Welding of Aluminium Alloys

melting point pure aluminium

Recommended preheat temperature

600 °C 500 400 300 200

Welding possible without preheating: AlMg5, AlMg7, AlMg4.5Mn, AlZnMg3, AlZnMg1

100

0

mild steel (0.2%C) without preheating

660

lated

temperature of melt start (solidus temperature) preheat temperature heat conductivity

Al Zn Mg Cu 0,5 Al Zn Mg Cu 1,5

in °C in °C in J/cm*s*K

Al Si 5 Al Cu Mg 1 Al R Mg 2 Al Cu Mg 0,5 Al Mn Al Mg 2 Al Cu Mg 2 Al Mg 3 Al Mg 3 Si Al Mg Mn

TS Tvorw. lAl-Leg.

7.17, together with the re-

745 l Al-Leg.; Al 99,98R Al99,9 Al99,8 Al 99,7 Al 99,5 Al 99 Al R Mg0,5 Al Mg Si 0,5 Al Mg Si 0,8 Al Mg Si 1 E Al Mg Si 1 Al Mg 1

TVorw. = TS -

92

calculation

result.

These results are only to be regarded as approximate, the individual application is subject to the information of the manufacturer.

Increasing better weldability © ISF 2002

br-er-08-17.cdr

Figure 7.17

Recommendations for Preheating

Another major problem during Al welding is the strong porosity of the welded joint. It is based on the interplay of several characteristics and hard to suppress. Pores in Al are mostly formed by hydrogen, which is driven out of the weld © ISF 2002

br-er-08-18.cdr

Figure 7.18

Excessive Porosity in a Al Weld

pool during solidification. irregular wire electrode feed

too thick and water containing oxyde layer by too long or open storage in non air-conditioned rooms

Solubility of hydrogen in

humid air (nitrogen, oxygen, water)

aluminium changes abrupt-

nozzle deposits and too steep inclination of the torch cause turbulences

poor current transition

VS

humid air

too thick oxyde layer (condensed water) dirt film (oil, grease)

dissolves many times more just forming crystal at the

H2 H2

festes Schweißgut base material

melt-crystal, i.e. the melt of the hydrogen than the

feuchte Luftpores Poren solid weld metal

ly on the phase transition

same temperature. Grundwerkstoff

© ISF 2002

br-er-08-19.cdr

Ingress of Hydrogen Into the Weld

Figure 7.19

7. Welding of Aluminium Alloys

93

This leads to a surplus of hydrogen in the melt due to the crystallisation during solidification. This surplus precipitates in form of a gas bubble at the solidifying front. As the melting point of Al is very low and Al has a very high heat conductivity, the solidification speed of Al is relatively high. As a result, in the melt ousted gas bubbles have often no chance to rise all the way to the surface. Instead, they are passed by the solidifying front and remain in the weld metal as pores, Figure 7.18. Figure 7.20

To suppress such pore formation it is therefore necessary to minimise the hydrogen content in the melt. Figure 7.19 shows possible sources of hydrogen during MIG welding of Al. Figure 7.20 and 7.21 show the effect of pure thermal expansion during Al welding. The

wedge

flame

large thermal expansion of the aluminium along with the relatively large heat affected zones cause in combination with a parallel gap adjustment a strong distortion of the welded parts. To minimise this distortion, the workpieces must be set at a suitable angle before welding, Figure 7.21. br-er08-21.cdr

© ISF 2002

Examples to Minimise Distortion

Figure 7.21

8. Technical Heat Treatment

8. Technical Heat Treatment

95 When welding a workpiece, not only the weld

6 cm 4

300°C 400°C

6 cm 4

600°C 700°C 800°C 900°C

500°C

2

2

0

0

-2

-2

-4

-4

itself, but also the surrounding base material

600°C 700°C

(HAZ) is influenced by the supplied heat quantity. The temperature-field, which appears around the weld when different welding

-6 -12

-10

-8

-6

-4

-2

0

temperature

-6 -14

500°C 400°C 300°C

2

cm

6

-8

-6

-4

-2

0 cm 2

procedures are used, is shown in Figure 8.1.

°C 1750

Figure 8.2 shows the influence of the material

1250 1000 723°C

properties on the welding process. The de-

oxy-acethylene welding

750 manual metal arc welding

500

termining factors on the process presented in this Figure, like melting temperature and -

250 -60 mm

-40

-20

0

20

40

mm

interval, heat capacity, heat extension etc,

60

distance from weld central line

depend greatly on the chemical composition heat affected zone during oxy-acethylene welding

of the material. Metallurgical properties are

heat affected zone during manual metal arc welding

br-er04-01.cdr

here characterized by e.g. homogeneity,

© ISF 2002

Temperature Distribution of Various Welding Methods

structure and texture, physical properties like heat extension, shear strength, ductility.

Figure 8.1

Structural changes, caused by the heat input

(process 1, 2, 7, and 8), influence directly the mechanical properties of the weld. In addition, the chemical composition of the weld metal and adjacent base material are also influenced by the processes 3 to 6.

1

Heating and melting the welding consumable

Specific heat, melting temperature and interval, melt heat, boiling temperature (metal, coating)

2

Melting parts of base material

Specific heat, melt temperature and interval, heat conductivity, heat expansion coefficient, homogeneity, time

3

Reaction of passing welding consumable with arc atmosphere

Compositionof atmosphere, affinity, pressure, temperature, dissotiation, ionisation, reaction speed

4

Reaction of passed welding consumable with molten base material

Solubility relations, temperature and pressure under influence of heat source, specific weight, weld pool flux

5

Interaction between weld pool and solid base material (possibly weld passes)

Diffusion and position change processes, time, boundary formation, ordered - unordered structure

6

Reaction of metal and flux with atmosphere

Affinity, temperature, pressure, time

7

Solidification of weld pool and slag

Melt heat, cooling conditions, density and porosity of slag, solidification interval

2

8

Cooling of welded joint in solid condition

Phase diagrams (time dependent), heat conductivity, heat coefficient, shear strength, ductility

10

9

Post-weld heat treatment if necessary

Phase diagrams (time dependent), texture by warm deformation, ductility, module of elasticity

10

Sustainable alteration of material properties

Phase diagrams, operating temperature, mechanical and chemical strain, time

Based on the binary system, the formation of the different structure zones is shown in Figure 8.3. So the coarse grain zone occurs in areas 1

of

intensely

elevated

austenitising temperature for

3

6

4

7 8

example. At the same time, hardness peaks appear in

5 9

© ISF 2002

br-eI-04-02.cdr

these greatly

areas

because

reduced

of

Classification of Welding Process Into Individual Mechanisms

critical

cooling rate and the coarse

Figure 8.2

8. Technical Heat Treatment

96

austenite grains. This zone of the weld is the area, where the worst toughness values are found. In Figure 8.4 you can see how much the forma-

hardness peak

hardness sink

1

weld bead

can be influenced.

1500 incomplete melt

°C 1300

width is achieved. Using a three pass tech-

standard transformation

1

ences in the formation of heat affected zones

2

3

4

5

6

3

800

recrystallisation

With the use of different procedures, the differ-

1147

1000 G

incomplete crystallisation

nique, the HAZ is reduced to only 8 mm.

2

1200

ageing blue brittleness

mm thick plate, a HAZ of approximately 30 mm

Temperature

coarse grain

723

4

P 600

5

400

S

6

300 100 0,2

2,06

Applying an electroslag one pass weld of a 200

0,8

zones of unfavourable mechanical properties

Hardness

tion of the individual structure zones and the

1 2 % 3 carbon content

heat affected zone (visible in macro section)

become even clearer as shown in Figure 8.5. These effects can actively be used to the ad-

br-er04-03.cdr

© ISF 2002

Microstructure Zones of a Weld Relation to Binary System

vantage of the material, for example to adjust calculated mechanical properties to one's choice or to remove negative effects of a weld-

Figure 8.3

ing. Particularly with high-strength fine grained steels and high-alloyed materials, which are specifically optimised to achieve special quality, e.g. corrosion resistance against a certain attacking

medium,

this

post-weld heat treatment is of great importance. Figure 8.6 shows areas in the Fe-C diagram of different heat treatment methods. It is clearly visible that the carbon content (and also the content of other alloying elements) has a distinct influence on the Figure 8.4

level of annealing tempera-

8. Technical Heat Treatment

97

tures like e.g. coarse-grain heat treatment or normalising. It can also be seen that the start of martensite formation (MS-line) is shifted to continuously decreasing temperatures with increasing C-content. This is important e.g. for hardening processes (to be explained later).

metastable system iron-carbon (partially) 1600

100

1536 °C

d - solid solution

electron beam welding

A4 1392 cbc atomic lattice

A

1600

melt + d - solid solution

1493°C

H B d - solid solution + austenite N

°C

melt

1400 heat colors melt + austenite

1300 yellow white

1300

1200

diffusion heat treatment

E

2,06

1100 coarse grain heat treatment

1000

40

submerged arc welding pass / capped pass

stress relieving

600 cbc 500 atomic lattice

dark red brown red

300

0,5 5

eutektoidic steel

0 Fe 0

200

hypereutectoidic steel

hypoeutectoidic steel

© ISF 2002

cherry-red

700

dark brown

MS

100

br-er04-05.cdr

light red

800

500

tempering

hardening

200

20

yellow red

900

400

300

gas metal arc welding

yellow

1000

600

recrystallisation heat treatment Q

400

12

light yellow

cm

A

cfc no atomic lattice rm A3 911 G ha alis rde ing austenite nin + austenite austenite + secondary g (g - Mischkristalle) + ferrite cementite (Fe3C) A2 800M 769°C O S K A1 P 723°C soft annealing ferrite700 (a-solid solution) recrystallisation heat treatment

1200 1147 1100

100

0,8 1,5 1 Carbon content in weight %

10 15 20 25 Cementite content in weight %

2

20

30

br-er04-06.cdr

© ISF 2002

Metallurgical Survey of Heat Treatment Methods

Development of Heat Affected Zone of EB, Sub-Arc, and MIG-MAG Welding

Figure 8.5

Figure 8.6

As this diagram does not cover the time influence, only constant stop-tempera°C

tures can be read, predic-

intense heating

austenite

long time several hours

900

possible. Thus the individual

Temperature

cooling-down rates are not

austenite + ferrite

A3 A1

Temperature

tions about heating-up and 700

ferrite + perlite 500

heat treatment methods will be explained by their temperature-time-behaviour

in

300 0,4 0,8 C-Content

%

Time © ISF 2002

br-eI-04-07.cdr

the following.

Coarse Grain Heat Treatment

Figure 8.7

8. Technical Heat Treatment

98

Figure 8.7 shows in the detail to the right a T-t course of coarse grain heat treatment of an alloy containing 0,4 % C. A coarse grain heat treatment is applied to create a grain size as large as possible to improve machining properties. In the case of welding, a coarse grain is unwelcome, although unavoidable as a consequence of the welding cycle. You can learn from Figure 8.7 that there are two methods of coarse grain heat treatment. The first way is to austenite at a temperature close above A3 for a couple of hours followed by a slow cooling process. The second method is very important to the welding process. Here a coarse grain is formed at a temperature far above A3 with relatively short periods. Figure 8.8 shows schemati-

900

mecha-

nisms, they must not be

500 400 300

used as reading off examples.

To

determine

t8/5,

distribution,

MS

200 100 2

hardness values, or microstructure

bainite

structure

martensite

running

A1 perlite

600

ferrite

(Note: the curves explain

e

700

A3

e lin

haviour in a TTT-diagram.

austenite

ferrit

°C

Temperature

cally time-temperature be-

0 0,1

are

3 1

10

br-eI-04-08.cdr

Time

4 10²

5

6 s

1

1: Normalizing 2: Simple hardening 3: Broken hardening 4: Hot dip hardening 5: Bainitic annealing 6: Patenting (isothermal annealing)

10³ © ISF 2002

TTT-Diagram With Heat Treatment Processes

TTT-diagrams always read continuously or isothermally. Mixed types like curves 3 to

Figure 8.8

6 are not allowed for this purpose!). The most important heat treatment methods can be divided into sections of annealing, hardening and tempering, and these single processes can be used individually or combined. The normalising process is shown in Figure 8.9. It is used to achieve a homogeneous ferriteperlite structure. For this purpose, the steel is heat treated approximately 30°C above Ac3 until homogeneous austenite evolves. This condition is the starting point for the following hardening and/or quenching and tempering treatment. In the case of hypereutectoid steels, austenisation takes place above the A1 temperature. Heating-up should be fast to keep the austenite grain as fine as possible (see TTA-diagram, chapter 2). Then air cooling follows, leading normally to a transformation in the ferrite condition (see Figure 8.8, line 1; formation of ferrite and perlite, normalised micro-structure).

8. Technical Heat Treatment

99 To harden a material, austenisation and homogenisation is carried out also at

°C

austenite

transformation and homogenizing of g-solid solution (30-60 min) at 30°C above A3

900

this case one must watch

A3 A1

Temperature

Temperature

austenite + ferrite

30°C above AC3. Also in

700 ferrite + perlite

that the austenite grains

quick heating

remain as small as possi-

500

air cooling

ble. To ensure a complete 300

transformation to marten0,4

0,8 C-Content

Time

%

site, a subsequent quench-

© ISF 2002

br-eI-04-09.cdr

ing

Normalizing

follows

until

the

temperature is far below Figure 8.9

the Ms-temperature, Figure 8.10. The cooling rate dur-

ing quenching must be high enough to cool down from the austenite zone directly into the martensite zone without any further phase transitions (curve 2 in Figure 8.8). Such quenching processes build-up very high thermal stresses which may destroy the workpiece during hardening. Thus there are variations of this process, where perlite formation is suppressed, but due to a smaller temperature gradient thermal stresses remain on an uncritical level (curves 3 and 4 in Figure 8.8). This can be achieved in practice –for example- through stopa

water

quenching

°C

process at a certain temcooling with a milder cooling medium (oil). With longer holding on at elevated tem-

about 30°C above A3

900 austenite + ferrite Temperature

perature and continuing the

austenite

ferrite + perlite

quenching in water

500 start of martensite formation

start of martensite formation

300 0,4

0,8 C-Content

Time

%

© ISF 2002

br-eI-04-10.cdr

Hardening

through in the bainite area (curves 5 and 6).

A1

700

perature level, transformations can also be carried

A3 Temperature

ping

Figure 8.10

8. Technical Heat Treatment

100

Figure 8.11 shows the quenching and tempering procedure. A hardening is followed by another heat treatment below Ac1. During this tempering process, a break down of martensite takes place. Ferrite and cementite are formed. As this change causes a very fine microstructure, this heat treatment leads to very good mechanical properties like austenite

°C

hardening and tempering

ness.

A3 A1

Temperature

austenite + ferrite Temperature

e.g. strength and tough-

about 30°C above A3

900

700 ferrite + perlite

quenching slow cooling

500

Figure 8.12 shows the procedure of soft-annealing.

300 0,4

Time

%

0,8 C-Content

Here we aim to adjust a © ISF 2002

br-eI-04-11.cdr

soft and suitable micro-

Hardening and Tempering

structure Figure 8.11

for

machining.

Such a structure is characterised by mostly globular

formed cementite particles, while the lamellar structure of the perlite is resolved (in Figure 8.12 marked by the circles, to the left: before, to the right: after soft-annealing). For hypoeutectic steels, this spheroidizing of cementite is achieved by a heat treatment close below A1. With these steels, a part of the cementite bonded carbon dissolves during heat treating close below A1, the remaining cementite lamellas transform with time into balls, and the bigger ones grow at the expense of the smaller ones (a transfor-

°C

time dependent on workpiece

mation is carried out because

→

thermodynami-

cally more favourable condition).

Hypereutectic

austenite + ferrite

oscillation annealing + / - 20 degrees around A1

10 to 20°C below A1

A3 A1

Temperature

reduced

900

Temperature

the surface area is strongly

austenite

700 ferrite + perlite

or

500

steels 300

have in addition to the lamel-

0,4

lar structure of the perlite a cementite

network

on

0,8 C-Content

%

Time cementite

the © ISF 2002

br-eI-04-12.cdr

grain boundaries.

Soft Annealing

Figure 8.12

8. Technical Heat Treatment

101

Spheroidizing of cementite is achieved by making use of the transformation processes during oscillating around A1. When exceeding A1 a transformation of ferrite to austenite takes place with a simultaneous solution of a certain amount of carbon according to the binary system Fe C. When the temperature drops below A1 again and is kept about 20°C below until the transformation is completed, a re-precipitation of cementite on existing nuclei takes °C

place. The repetition of this

austenite

900

process leads to a step-

A3 A1

Temperature

Temperature

austenite + ferrite 700 ferrite + perlite

wise spheroidizing of ce-

time dependent on workpiece

between 450 and 650 °C

500

mentite and the frequent transformation

avoids

a

grain coarsening. A soft-

300 0,4

0,8 C-Content

annealed

Time

%

© ISF 2002

br-eI-04-13.cdr

Stress Relieving

microstructure

represents frequently the delivery condition of a material.

Figure 8.13

Figure 8.13 shows the principle of a stress-relieve heat treatment. This heat treatment is used to eliminate dislocations which were caused by welding, deforming, transformation etc. to improve the toughness of a workpiece. Stress-relieving works only if present dislocations are able to move, i.e. plastic structure deformations must be executable in the micro-range. A temperature increase is the commonly used method to Stress releaving

Heat treatment at a temperature below the lower transition point A1 , mostly between 600 and 650°C, with subsequent slow cooling for relief of internal stresses; there is no substantial change of present properties.

Normalising

Heating to a temperature slightly above the upper transition point A3 (hypereutectoidic steels above the lower transition point A1 ), followed by cooling in tranquil atmosphere.

Hardening (quench hardening)

Acooling from a temperature above the transition point A3 or A1 with such a speed that an clear increase of hardness occurs at the surface or across the complete cross-section, normally due to martensite development.

Quenching and tempering

Heat treatment to achieve a high ductility with defined tensile stress by hardening and subsequent tempering (mostly at a higher temperature.

should not cause any other

Solution or quenching heat treatment

Fast cooling of a workpiece. Also fast cooling of austenitic steels from high temperature (mostly above 1000°C) to develop an almost homogenuous micro-structure with high ductility is called 'quenching heat treatment'.

change to properties, so that

Tempering

Heating after previous hardening, cold working or welding to a temperature between room temperature and the lower transformation point A1; stopping at this temperature and subsequent purposeful cooling.

make

such

deformations

possible because the yield strength limit decreases with increasing temperature. A stress-relieve heat treatment

tempering steels are heat © ISF 2002

br-eI-04-14.cdr

treated

below

tempering Type and Purpose of Heat Treatment

temperature. Figure 8.14

8. Technical Heat Treatment

102

Figure 8.14 shows a survey of heat treatments which are important to welding as well as their purposes. Figure 8.15 shows princi-

Types of heat treatments related to welding

heat treatment before welding

combination

accompanying heat treatment

pally the heat treatments in heat treatment after welding (”post-weld heat treatment”)

combination

connection with welding. Heat treatment processes

simple step-hardening welding

annealing

stress releaving

stress releaving

combination

preheating

simple preheating

local preheating

increase of working temperature

preheating of the complete workpiece

pure step hardening welding

constant working temperature isothermal welding

modified step hardening welding

are divided into: before,

annealing hardening quenching and tempering

solution tempering heat treatment

Normally a stress-relieving

postheating (”post weld heat treatment”)

heat treatment of the complete workpiece

during, and after welding. or normalizing heat treat-

local heat treatment

ment

is

applied

before

welding to adjust a proper © ISF 2002

br-eI-04-15.cdr

material condition which for Heat Treatment in Connection With Welding

welding. After welding, almost any possible heat

Figure 8.15

treatment can be carried out. This is only limited by workpiece dimensions/shapes or arising costs. The most important section of the diagram is the kind of heat 800

treatment which accom-panies the welding.

°C 700

The most important processes are explained in 600

Figure 8.16 represents the influence of differ-

Temperature T

the following.

500

400

ent accompanying heat treatments during

300

welding, given within a TTT-diagram. The fast-

200

est cooling is achieved with welding without

TA

MS

(1)

(2)

(3)

100

preheating, with addition of a small share of 0 0

bainite, mainly martensite is formed (curve 1,

1

10

102 Time t

103

104

s

105

tH

analogous to Figure 8.8, hardening). A simple heating before welding without additional stopping time lowers the cooling rate according to

(1): Welding without preheating, (2): Welding with preheating up to 380°C, without stoppage time (3): Welding with preheating up to 380°C and about 10 min. stoppage time TA: Stoppage temperature, tH: Dwell time

br-er04-16.cdr

© ISF 2002

TTT-Diagram for Different Welding Conditions

curve 2. The proportion of martensite is reduced in the forming structure, as well as the Figure 8.16

8. Technical Heat Treatment

103

level of hardening. If the material is hold at a temperature above MS during welding (curve 3), then the martensite formation will be completely suppressed (see Figure 8.8, curve 4 and 5). To explain the temperature-time-behaviours

seam

start

used in the following, Figure 8.17 shows a su-

end

TS

perposition of all individual influences on the A3

the HAZ. As an example, welding with simple preheating is selected. The plate is preheated in a period tV. After re-

Temperature T

materials as well as the resulting T-T-course in

transformation range

A1

TV

moval of the heat source, the cooling of the workpiece starts. When tS is reached, welding

Time t

starts, and its temperature peak overlays the cooling curve of the base material. When the welding is completed, cooling period tA starts. The full line represents the resulting tempera-

tV

tS

TV: Preheat temperature, TS: Melting temperature of material, tV: Preheat time, tS: Welding time, tA: Cooling time (room temperature), MS: Martensite start temperature A3: Upper transformation temperature, A1: Lower transformation temperature

tA Course of resulting temperature in the area of the heat affected zone of the base material. Temperature distribution by preheating, Course of temperature during welding.

ture-time-behaviour of the HAZ. br-er04-17.cdr

Temperature-Time-Distribution During Welding With Preheating

The temperature time course during welding with simple preheating is shown in Figure 8.18.

© ISF 2002

Figure 8.17

During a welding time tS a drop of the working temTemperature T

A3

perature TA occurs. A further air cooling is usually

A1

carried out, however, the TV

cooling rate can also be

TA

reduced by covering with

Time t tV

tS

TV: Preheat temperature, TA: Working temperature, tV: Preheat time, tS: Welding time, tA: Cooling time (room temperature)

heat insulating materials.

tA

Temperature of workpiece, Temperature of weld point

Another variant of welding © ISF 2002

br-eI-04-18.cdr

Welding With Simple Preheating

with preheating is welding at

Figure 8.18

constant

temperature.

working This

is

8. Technical Heat Treatment

104 achieved through further warming during welding to

A3 Temperature T

avoid a drop of the working A1

temperature. In Figure 8.19 is this case (dashed line,

TV TA

MS

TA needs not to be above MS) as well as the special

Time t tS tV

:

tH = 0

TV: Preheat temperature, TA: Working temperature, tV: Preheat time,

case of isothermal welding

tA

tH

illustrated. During isother-

tS: Welding time, tA: Cooling time (room temperature), tH: Dwell time

mal welding, the workpiece © ISF 2002

br-eI-04-19.cdr

is heated up to a working

Welding With Preheating and Stoppage at Working Temperature

temperature

Figure 8.19

above

MS

(start of martensite formation) and is also held there

after welding until a transformation of the austenitised areas has been completed. The aim of isothermal welding is to cool down in accordance with curve 3 in Figure 8.16 and in this way, to suppress martensite formation. 1. Post-heating

Figure 8.20 shows the T-T course during treatment, see Figure 8.15). Such a treatment can be carried out very easy, a gas welding

A3 Temperature T

welding with post-warming (subsequent heat

A1 TN

torch is normally used for a local preheating. Time t

In this way, the toughness properties of some

tS tN

steels can be greatly improved. The lower

tA

2. Pre- and post-heating

sketch shows a combination of pre- and poststeels which have such a strong tendency to

Temperature T

heat treatment. Such a treatment is applied to

A3 A1 TN

TV

TA

hardening that a cracking in spite of a simple Time t

preheating before welding cannot be avoided, if they cool down directly from working temperature. Such materials are heat treated

tV

tS

TV: Preheat temperature, TA: Working temperature, TN: Postheat temperature, tV: Preheating time,

tN

tR

tS: tA: tN: tR:

tA

Welding time, Cooling time (room temperature), Postheat time Stoppage time

br-er04-20.cdr

© ISF 2002

Welding With Pre- and Post-Heating

immediately after welding at a temperature between 600 and 700°C, so that a formation Figure 8.20

8. Technical Heat Treatment

105

of martensite is avoided and welding residual stresses are eliminated simultaneously. Aims of the modified stephardening

Temperature T

THa

not be discussed here, Fig-

A1

ure 8.21. Such treatments are used for transformation-

TAnl

inert materials. The aim of

MS TAnl

the figure is to show how

Time t

complicated a heat treatment

tS tH

tA

tH

tHa

tAnl

tA

can become for a material in

tAb TA: Working temperature, TAnl: Tempering temperature, THä: Hardening temperature,

should

A3

TA

TSt

welding

TSt: Step temperature, tA: Cooling time, tAb: Quenching time,

tAnl: Tempering time, tH: Dwell time, tS: Welding time

Temperature of workpiece, Temperature of weld point

combination with welding.

© ISF 2002

br-eI-04-21.cdr

Modified Step Weld Hardening

Figure 8.22 shows temperature distribution during multi-

Figure 8.21

pass welding. The solid line represents the T-T course of a point in the HAZ in the first pass. The root pass was welded without preheating. Subsequent passes were

1

2

3

4

weld pass heat affected zone

welded without cooling down to a certain temperature. As a result, working temperature in-

}

4 3 weld pass 2 1 observed point

TS

second pass is welded under a preheat temperature which is already above martensite start temperature. The heat which remains in

Temperature T

creases with the number of passes. The A3

TV MS

the workpiece preheats the upper layers of the Time t

weld, the root pass is post-heat treated through

tS tV

tA

the same effect. During welding of the last pass, the preheat temperature has reached such a high level that the critical cooling rate will not be surpassed. A favourable effect of

TV: Preheat temperature, TS: Melting temperature of material, tV: Preheat time, tS: Welding time tA: Cooling time (room temperature), A3: Upper transformation temperature, MS: Martensite start temperature br-er04-22.cdr

multi-pass welding is the warming of the HAZ

Temperature-Time Distribution During Multi-Pass Welding

of each previous pass above recrystallisation temperature with the corresponding crystallisa-

© ISF 2004

Figure 8.22

8. Technical Heat Treatment

106

tion effects in the HAZ. The coarse grain zone with its unfavourable mechanical properties is only present in the HAZ of the last layer. To achieve optimum mechanical values, welding is not carried out to Figure 8.22. As a rule, the same welding conditions should be applied for all passes and prescribed t8/5 – times must be kept, welding of the next pass will not be carried out before the previous pass has cooled down to a certain temperature (keeping the interpass temperature). In addition, the workpiece will not heat up to excessively high temperatures. Figure 8.23 shows a nomogram where working temperature and minimum and maximum heat input for some steels can be interpreted, depending on carbon equivalent and wall thickness. If e.g. the water quenched and tempered fine grain structural steel S690QL of 40 mm wall thickness is welded, the following data can be found: - minimum heat input between 5.5 and 6 kJ/cm - maximum heat input about 22 kJ/cm - preheating to about 160°C - after welding, residual stress relieving between 530 and 600°C. Steels which are placed in the hatched area called soaking

area,

must

be

treated with a hydrogen relieve annealing. Above this area, a stress relieve annealing must be carried out. Below this area, a post-weld heat treatment is not required.

Figure 8.23

9. Welding Defects

9. Welding Defects

108

Figures 9.1 to 9.4 give a rough survey about the classification of welding defects to DIN 8524. This standard does not classify existing welding defects according to their origin but only to their appearance.

undercut, continuous

in the unaffected base metal

unfused longitudinal seam edge

in weld metal

longitudinal crack

in fusion zone

in the HAZ

undercut

in the unaffected base metal in weld metal

transverse crack

end crater with reduction of weld cross section

open end crater

in the HAZ

star shaped crack

nominal

in the unaffected base metal in weld metal in the HAZ

weld reinforcement pore too small throat thickness

globular gas inclusion

nominal

porosity surface defects at a start point

many, mainly evenly distributed pores

start defects nest of pores locally repeated pores weld is too wide

excessive seam width

line of pores pores arranged in a line

burn through

through-going hole in or at the edge of the seam

worm hole elongated gas inclusion in weld direction

br-er09-01.cdr

© ISF 2002

br-er09-02.cdr

Defect Class: Cracks and Cavities

Defect Class: Shape Defects

Figure 9.1

lack of fusion between passes

Figure 9.2

lack of fusion between weld passes or weld beads

root lack of fusion lack of fusion in the area of weld root

flank lack of fusion lack of fusion between weld and base metal

insufficient through weld insufficiently welded cross section

insufficiently welded root one or two longitudinal edges of the groove are unfused © ISF 2002

br-er-09-03.cdr

Defect Class: Lack of Fusion, Insufficient Through-Weld

Figure 9.3

© ISF 2002

9. Welding Defects

109

A distinction of arising defects by their origin is shown in Figure 9.5. The development of the most important welding defects is explained in the following paragraphs. Lack of fusion is defined stringer type inclusions

as unfused area between

slag line

weld metal and base mate-

different shapes and directions

rial or previously welded layer. This happens when

single slag inclusions irregular slag inclusions

the base metal or the previous layer are not completely

pore nest

or

insufficiently

molten. Figure 9.6 explains

locally enriched

© ISF 2002

br-er-09-04.cdr

the influence of welding parameters on the devel-

Defect Class: Solid Inclusions

opment of lack of fusion. In

Figure 9.4

the upper part, arc characteristic lines of MAG welding are shown using CO2

welding joint defects

welding defects due to manufacture

and mixed gas. The weld-

welding defects due to material

ing voltage depends on external weld defects

internal weld defects

hot cracks

cold cracks

cavities with weld metal

welding current and is se-

metallurgical pore formation

lected according to the

spatters and start points

lacks of fusion

solidification cracks

hydrogen cracks

undercuts

slag inclusions

remelt cracks

hardening cracks

seam shape defects

mechanical pore formation

crater formation

lamellar cracks

tension, the welding cur-

precipitation cracks

Welding Defects

Figure 9.5

rent is fixed by the wire © ISF 2002

br-er-09-05.cdr

joint type. With present

feed

speed

(thus

also

melting rate) as shown in the middle part of the figure.

Melting rate (resulting from selected welding parameters) and welding speed define the heat input. As it can be changed within certain limits, melting rate and welding speed do not limit each other, but a working range is created (lower part of the figure). If the heat input is too low, i.e. too high welding speed, a definite melting of flanks cannot be ensured. Due to the

9. Welding Defects

110

poor power, lack of fusion is the result. With too high heat input, i.e. too low welding speed, the weld pool gets too large and starts to flow away in the area in front of the arc. This effect prevents a melting of the base metal. The arc is not directed into the base metal, but onto the weld pool, and flanks are not entirely molten. Thus lack of fusion may occur in such areas.

Welding voltage

welding direction CO2

mixed gas

positive torch angle

neutral

negative torch angle

torch axes

torch axes

correct

false

Welding current

Welding current

correct

false

Wire feed Melting rate

Welding speed

lack of fusion due to too low performance

br-er09-06.cdr

approx. 45° 1...2

wo

ing

ra

ng

e

false

rk

lacks of fusion due to preflow

Melting rate

90°

© ISF 2002

br-er09-07.cdr

Influence of Welding Parameters on Formation of Lack of Fusion

Figure 9.6

correct

© ISF 2002

Influence of Torch Position on Formation of Lack of Fusion

Figure 9.7

Figure 9.7 shows the influence of torch position on the development of weak fusion. The upper part of the figure explains the terms neutral, positive and negative torch angle. Compared with a neutral position, the seam gets wider with a positive inclination together with a slight reduction of penetration depth. A negative inclination leads to narrower beads. The second part of the figure shows the torch orientation transverse to welding direction with multi-pass welding. To avoid weak fusion between layers, the torch orientation is of great importance, as it provides a reliable melting and a proper fusion of the layers. The third figure illustrates the influence of torch orientation during welding of a fillet weld. With a false torch orientation, the perpendicular flank is insufficiently molten, a lack of fusion occurs. When welding an I-groove in two layers, it must be ensured that the plate is com-

9. Welding Defects

111

pletely fused. A false torch orientation may lead to lack of fusion between the layers, as shown in the lower figure. Figure 9.8 shows the influence of the torch orientation during MSG welding of a rotating workpiece. As 12

12

1 2

3

9 Uhr

an example, the upper fi1 2

2

3

9 Uhr

12

1

9 Uhr

gure shows the desired

3

torch orientation for usual 6

6

6

welding speeds. This orientation depends on pa-

br-er09-08.cdr

rameters

like

workpiece

diameter

and

thickness,

© ISF 2002

Influence of Torch Position on Formation of Lacks of Fusion

groove

shape,

melting

rate, and welding speed.

Figure 9.8

The lower figure illustrates variations of torch orientation on seam formation. A torch orientation should be chosen in such a way that a solidification of the melt pool takes place in 12 o'clock position, i.e. the weld pool does not flow in front or behind of the arc. Both may cause lack of fusion. In contrast to faulty fusion, pores in the weld metal due to their globular shape are less critical, provided that their size does not exceed a certain value. Secondly, they must occur isolated and keep a minimum distance from each other. There are two possible mechanisms

to

develop

cavities in the weld metal: the mechanical and the metallurgical pore formation.

Figure

9.9

lists

causes of a mechanical pore formation as well as possibilities to avoid them. To over-weld a cavity (lack Figure 9.9

9. Welding Defects

112

gas/gas developing material

causes

air -nitrogen -hydrogen

too low shielding gas flow through:

avoidance

too low setting leaking lines too small capillary bore hole too low supply pressure for pressure regulator

correct settings search and eliminate leaks correct combination capillary - pressure regulator Pressure of bottles or lines must meet the required supply pressure of the pressure regulator

insufficient gas shield through: open windows, doors, fans etc. insufficient gas flow at start and at completion of welding too large gas nozzle distance excentric wire stick-out false gas nozzle shape false gas nozzle position (with decentralised gas supply)

protect welding point from draught suitable gas pre- and post-flow time reduce distance straighten wire electrode, center contact tube select proper gas nozzle shape for joint type position gas nozzel behind torch - if possible

turbulences through: to high shielding gas flow spatters on gas nozzle or contact tube irregular arc

thermal current - possibly increased by chimmney effects with one-sided welding too high weld pool temperature too high work piece temperature injection effects water

leaking torch (with water-cooled types)

carbonmonoxide

remelting of seggregation zones remelting of rust or scale

reduce gas flow clean gas nozzle and contact tube eliminate wire feed disturbances, increase voltage, if wire electrode splutters, ensure good current transition in contact tube, correct earth connection, remove slag of previously welded layers weld on backing or with root forming gas reduce weld pool size reduce preheat or interpass temperature reduce torch inclination, tighten leaks in gas line, avoid visible gas nozzle slots search and eliminate leaks, dry wire feed hose after ingress of water reduce penetration by decreasing arc power or increasing welding speed clean welding area before welding

br-er09-11.cdr

© ISF 2002

Metallurgical Pore Formation

Figure 9.10

Figure 9.11

of fusion, gaps, overlaps etc.) of a previous layer can be regarded as a typical case of a mechanical pore formation. The welding heat during welding causes a strong expansion of the gasses contained in the cavity and consequently a develop-

a) low crystallisation speed

ment of a gas bubble in the liquid weld metal. If the solidification is carried out so fast that this gas bubble b) high crystallisation speed

cannot raise to the surface © ISF 2002

br-er-09-12.cdr

of the weld pool, the pore

Growth and Brake Away of Gas Cavities at the Phase Border

will be caught in the weld metal.

Figure 9.12

Figure 9.10 shows a X-ray photograph of a pore which developed in this way, as well as a surface and a transverse sec-

9. Welding Defects

113

tion. This pore formation shows its typical pore position at the edge of the joint and at the fusion line of the top layer. Figure 9.11 summarises causes of and measures to avoid a metallurgical pore formation. Reason of this pore formation is the considerably increased solubility of the molten metal compared with the solid state. During solidification, the transition of liquid to solid condition causes a leapwise reduction of gas solubility of the steel. As a result, solved gasses are driven out of the crystal and are enriched as a gas bubble ahead of the solidification front. With a slow growth of the crystallisation front, the bubbles have enough time to raise to the surface of the weld pool, Figure 9.12 upper part. Pores will not be developed. However, a higher solidification speed may lead to a case where gas bubbles are passed by the crystallisation front and are trapped as

Figure 9.13

pores in the weld metal, lower part of the figure. Figure 9.13 shows a X-ray photograph, a surface and a transverse section of a seam with metallurgical pores. The evenly distributed pores across the seam and the accumulation of pores in the upper part of the seam (transverse section) are typical. Figure

9.14

shows

the

ways of ingress of gasses into the weld pool as an example during MAG welding. A pore formation is mainly caused by hydrogen and nitrogen. Oxygen is

Figure 9.14

9. Welding Defects

114

bonded in a harmless way when using universal electrodes which are alloyed with Si and Mn. Figure 9.15 classifies cracks to DIN 8524, part 3. In contrast to part 1 and 2 of this standard, are cracks not only classified by their appearance, but also by their development.

Figure 9.15

9. Welding Defects

115 Figure

1600 °C

TS

allocates

cracks according to their

0011

1200 Temperature

9.16

0010

0012

appearance

0021

0027

800 0020 400

MS

0 1

10

0022 0023 0024 0025

102

103

the

welding heat cycle. Principally there is a distinction 0026 0028

104

105

between the group 0010 106

s

107

(hot

Time

0010 area of hot crack formation 0011 area of solidification crack formation 0012 area of remelting crack formation 0020 area of cold crack formation 0021 area of brittle crack formation 0022 area of shrinking crack formation

during

0023 area of hydrogen crack formation 0024 area of hardening crack formation 0025 area of tearing crack formation 0026 area of ageing crack formation 0027 area of precipitation crack formation 0028 area of lamella crack formation

cracks)

and

0020

(cold cracks).

© ISF 2002

br-er-09-16.cdr

Crack Formation During Steel Welding

Figure 9.16

A model of remelting development and solidification cracks is shown in Figure 9.17. The upT

per part illustrates solidification conditions in a

TmA

simple case of a binary system, under the proC5

vision that a complete concentration balance

TmB A

takes place in the melt ahead of the solidifica-

C0

C’5 B CB

tension

tion front, but no diffusion takes place in the crystalline solid. When a melt of a composition tension

C0 cools down, a crystalline solid is formed

a tension

when the liquidus line is reached. Its concen-

tension

tration can be taken from the solidus line. In the course of the ongoing solidification, the rest tension

of molten metal is enriched with alloy elements b

in accordance with the liquidus line. As defined

tension segregation in base metal

in the beginning, no diffusion of alloy elements

aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa

melt

br-er09-17.cdr

© ISF 2002

Development of Remelting and Solidification Cracks

in the already solidified crystal takes place, thus the crystals are enriched with alloy elements much slower than in a case of the binary

Figure 9.17

system (lower line). As a result, the concentration of the melt exceeds the maximum equilibrium concentration (C5), forming at the end of solidification a very much enriched crystalline solid, whose melting

9. Welding Defects

116

point is considerably lower when compared with the firstly developed crystalline solid. Such concentration differences between first and last solidified crystals are called segregations. This model of segregation development is very much simplified, but it is sufficient to understand the mechanism of hot crack formation. The middle part of the figure shows the formation of solidification cracks. Due to the segregation

b

b

effects

described

above, the melt between t

the crystalline solids at the t

end of solidification has a considerably solidus a: non-preferred bead shape b 1 t

c: non-preferred bead shape

Crystallisation of Various Bead Geometries

temperature.

As

indicated by the black areas, rests of liquid may be

© ISF 2002

br-er-09-18.cdr

decreased

trapped by dendrites. If tensile

stresses

exist

(shrinking stress of the Figure 9.18

welded joint), the liquid areas are not yet able to transfer forces and open up. The lower part of the figure shows the development of remelting cracks. If the base material to be welded contains already some segregations whose melting point is lower than that of the rest of the base metal, then these zones will melt during weld-

Figure 9.19

ing, and the rest of the material remains solid (black areas). If the joint is exposed to tensile stress during solidification, then these areas open up (see above) and cracks occur. A hot cracking tendency of a steel is above all promoted by sulphur and phosphorus, because these elements form with iron very

9. Welding Defects

117

low melting phases (eutectic point Fe-S at 988°C) and these elements segregate intensely. In addition, hot crack tendency increases with increasing melt interval. As shown in Figure 9.18, also the geometry of the groove is important for hot crack tendency. With narrow,

deep

grooves

a

crystallisation takes place of all sides of the bead, entrapping the remaining melt in the bead centre. With

the

shrinking

occurrence stresses,

of hot

cracks may develop. In the case

Figure 9.20

of

flat

beads

as

shown in the middle part of

br-er09-21.cdr

© ISF 2002

Macrosection of a SA-Weld

Figure 9.21

Figure 9.22

9. Welding Defects

118

the figure, the remaining melt solidifies at the surface of the bead. The melt cannot be trapped, hot cracking is not possible. The case in figure c shows no advantage, because a remelting crack may occur in the centre (segregation zone) of the first layer during welding the second layer. The example of a hot crack in the middle of a SA weld is shown in Figure 9.19. This crack developed due to the unsuitable groove geometry. Figure 9.20 shows an example of a remelting crack which started to develop in a segregation zone of the base metal and spread up to the bead centre. structure (hardness)

hydrogen

stresses

The section shown in Figure 9.21 is similar to case c in Figure 9.18. One can clearly see

chemical composition (C-equivalent)

welding consumables humidity on welding edges

residual stresses (yield stress of steels and joints)

that an existing crack develops through the following layers during over-welding. Figure 9.22 classifies cold cracks depending

cooling rate (t8/5)

cooling rate (t8/1)

additional stresses (production conditions)

on their position in the weld metal area. Such a classification does not provide an explanation for the origin of the cracks.

5 mm section plane br-er09-23.cdr

© ISF 2002

Causes of Cold Crack Formation

Figure 9.23

5 mm

0,2 mm

Figure 9.23 shows a summary of the three main causes of cold crack formation and their main influences. As explained in previous

etching: HNO3 5 mm

chapters, the resulting welding microstructure depends on both, the composition of base crack in heat affected zone

and filler materials and of the cooling speed of the joint. An unsatisfactory structure composition promotes very much the formation of

transverse cracks in weld metal br-er09-24.cdr

© ISF 2002

Cold Cracks in the Heat Affected Zone and Weld Metal

cold cracks (hardening by martensite). Figure 9.24

9. Welding Defects

119 Another cause for increased cold crack sus-

1,2 18 °C

% 1,0 Water content of coating

90 % RH

1,12 1,0

0,8

1,17

ceptibility is a higher hydrogen content. The

basic stick electrode Mn - Ni - Mo - Typ

hydrogen content is very much influenced by

0,83

the condition of the welding filler material

0,6

0,4

0,46

0,2

0,35 0,27 0,17 0,1

70 % 0,43

(humidity of electrodes or flux, lubricating

50 % 0,22 35 % 0,18

grease on welding wire etc.) and by humidity

0,4 0,21

0,16

on the groove edges.

0 0

1

2

3

4 5 Storage time

6

7

days

9

The cooling speed is also important because

4,0

Water content of coating

it determines the remaining time for hydrogen

20 °C / 70 % RH

%

effusion out of the bead, respectively how

3,0

much hydrogen remains in the weld. A meas-

2,0

ure is t8/1 because only below 100°C a hydro1,0

0

gen effusion stops.

0

1

10

Tage

Storage time

br-er09-25.cdr

100 © ISF 2002

A crack initiation is effected by stresses. De-

Water Pick-up of Electrode Coatings

pending on material condition and the two already mentioned influencing factors, even

Figure 9.25

residual stresses in the workpiece may actu-

ate a crack. Or a crack occurs only when superimpose of residual stresses on outer stress. Figure 9.24 shows typical cold cracks in a workpiece. An increased hydrogen content in the weld metal leads to an increased cold crack tendency. Mechanisms of hydrogen cracking were not completely understood until today. However, a spontaneous occurrence is typical of hydrogen

cracking.

Such

1.0 % 0.9

cracks do not appear diafter

welding

but

0.8

hours or even days after

0.7

cooling. The weld metal hydrogen content depends on humidity of the electrode

Water content of coating

rectly

basic electrode 1 year storage time at 18 - 20 °C 0.74

0.6 0.5 0.4 0.39 0.3 0.28 0.2

coating (manual metal arc

0.1

welding) and of flux (submerged arc welding).

AWS A5.5 stored and rebaked

0 30

40

50 60 Relative humidity

70

Water Content of Coating After Storage and Rebaking

Figure 9.26

%

80 © ISF 2002

br-er-09-26.cdr

9. Welding Defects

120

Figure 9.25 shows that the moisture pick-up of an electrode coating greatly depends on ambient conditions and on the type of electrode. The upper picture shows that during storage of an electrode type the water content of the coating depends on air humidity. The water content of the coating of this electrode type advances to a maximum value with time. The lower picture shows that this behaviour does not apply to all electrode types. The characteristics of 25 welding electrodes stored under identical conditions are plotted here. It can clearly be seen that a behaviour as shown in the upper picture applies only to some electrode types, but basically a very different behaviour in connection with storage can be noticed. 60 ml 100g

In practice, such constant

preheat temperature in °C 80

20

100

Diffusible hydrogen content in weld metal

50

storage conditions are not to be found, this is the rea-

40 cellulose coated stick electrode

son why electrodes are

30

backed before welding to 20

limit the water content of basic coated stick electrode

the coating. Figure 9.26

10

shows the effects of this 0

100

200 300 Cooling time between 800 and 1000°C

400

s

500

measure. The upper curve © ISF 2002

br-er-09-27.cdr

Influence of Preheat Temperature on Cooling Speed and Hydrogen Content

Figure 9.27

shows the water content of the coating of electrodes which were stored at constant air humidity before

rebaking. Humidity values after rebaking are plotted in the lower curve. It can be seen that even electrodes stored under very damp conditions can be rebaked to reach acceptable values of water content in the coating. Figure 9.27 shows the influence of cooling speed and also the preheat temperature on hydrogen content of the weld metal. The values of a high hygroscopic cellulose-coated electrode are considerably worse than of a basic-coated one, however both show the same tendency: increased cooling speed leads to a raise of diffusible hydrogen content in weld metal. Reason is that hydrogen can still effuse all the way down to room temperature, but diffusion speed increases sharply with temperature. The longer the steel takes to cool, the more time is available for hydrogen to effuse out of the weld metal even in higher quantities.

9. Welding Defects

121 The table in Figure 9.28 shows an assessment of the quantity of diffusible

Designation

Hydrogen content ml/100 g deposited weld metal

hydrogen in weld metal

high

>15

according to DIN 8529.

medium

£ 15 and > 10

low

£ 10 and > 5

very low

£5

in ISO 2560 classified as Hcontrolled electrodes

Based on this assessment, a classification of weld metal to DIN 32522 into groups depending on hy-

© ISF 2002

br-er-09-28.cdr

drogen is carried out, Fig-

Assessment of Diffusible Hydrogen During Manual Metal Arc Welding

ure 9.29.

Figure 9.28

A cold crack development can

be

followed-up

by

means of sound emission

Abbreviation

measurement. Figure 9.30 represents the result of such a measurement of a welded component. A solid-borne

Hydrogen content ml/100 g deposited weld metal (max.)

HP 5 HP 7 HP 10 HP 15

5 7 10 15

sound microphone is fixed to

a

component

which

measures the sound pulses

© ISF 2002

br-er-09-29.cdr

H2-Bestimmung nach DIN 8572,2

Diffusible Hydrogen of Weld Metal to DIN 32522

generated by crack development. The intensity of the pulses provides a qualitative

Figure 9.29

assessment of the crack size. The observation is carried out without applying an external tension, i.e. cracks develop only caused by the internal residual stress condition. Figure 9.32 shows that most cracks occur relatively short after welding. At first this is due to the cooling process. However, after completed cooling a multitude of developing sounds can be registered. It is remarkable that the intensity of late occurring pulses is especially high. This behaviour is typical for hydrogen induced crack formation. Figure 9.31 shows a characteristic occurrence of lamellar cracks (also called lamellar tearing). This crack type occurs typically during stressing a plate across its thickness (perpen-

9. Welding Defects

122 dicular to rolling direction). The upper picture shows joint types which are very much at risk to formation of such cracks. The two lower pictures show the cause of that crack formation. During steel production, a formation

of

segregation

cannot be avoided due to the casting process. With following production steps, such

Figure 9.30

segregations

are

stretched in the rolling direction. Zones enriched and depleted of alloy elements are now close together. These concentration differences influence the transformation behaviour of the individual zones. During cooling, zones with enriched alloy elements develop a different microstructure than depleted zones. This effect which can be well recognised in Figure 9.31, is called structure banding. In practice, this formation can be hardly avoided. Banding in plates is the reason for worst mechanical properties perpendicular to rolling direction. This is caused by a different mechanical behaviour of different microstructures. When stressing lengthwise and transverse to rolling direction, the individual structure bands may support each other and a mean strength is provided.

Figure 9.31

Such support cannot be obtained perpendicular to rolling direction, thus the strength of the workpiece is that of the weaker microstructure

9. Welding Defects

123

areas. Consequently, a lamellar crack propagates through weaker microstructure areas, and partly a jump into the next band takes place.

100

vulnerable. Depending on joint shape, these welds

12

show to some extent a conA

welded construction which

8

shrinking.

12

siderable

shrinkage value 1,7 mm

such t-joints are particularly

shrinkage value 0,6 mm

shrinkage value 0,4 mm

Figure 9.32 illustrates why

6r

100

50°

3

greatly impedes shrinking of joint,

may

2

this

50°

generate

© ISF 2002

br-er-09-32.cdr

stresses the

perpendicular

plane

of

to

Shrinkage Values of T-Joints With Various Joint Shapes

magnitude

above the tensile strength.

Figure 9.32

This can cause lamellar tearing. Precipitation

cracks

occur mainly during

stress relief heat treatment of welded components. They occur in the coarse grain zone close to fusion line. As this type of cracks occurs often during post weld heat treatment of cladded materials, is it also called undercladding crack, Figure 9.33. Especially susceptible are steels which concrack formation in these areas of the coarse grain zone

tain alloy elements with a precipitation hardweld bead

coarse grain zone 1

2

3

ening effect (carbide developer like Ti, Nb, V).

4

During welding such steels, carbides are dissolved in an area close to the fusion line. Durbase metal: ASTM 508 Cl (22NiMoCr3-7)

ing br-er09-33.cdr

© ISF 2002

Undercladding Cracks

Figure 9.33

the

following

cooling,

the

carbide

developers are not completely re-precipitated.

9. Welding Defects

124

If a component in such a condition is stress relief heat treated, a re-precipitation of carbides takes place (see hot ageing, chapter 8). With this re-precipitation, precipitation-free zones may develop along grain boundaries, which have a considerably lower deformation stress limit compared with strengthened areas. Plastic deformations during stress relieving are carried out almost only in these areas, causing the cracks shown in Figure 9.33.