You are on page 1of 140


Weldability of Metals

1. Weldability of Metals 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 1 Forming

Main group 2 Deforming

Main group 3 Separating

Main group 4 Joining DIN 8593

Main group 5 Plating

Main group 6 Changing material characteristics

Group 4.1 Assembling

Group 4.2 Filling

Group 4.3 Pressing

Group 4.4 Joining by forming

Group 4.5 Joining by deforming

Group 4.6 Joining by welding

Group 4.7 Joining by soldering

Group 4.8 Bonding

Sub-group 4.6.1 Pressure welding

Sub-group 4.6.2 Fusion welding


ISF 2002

Classification of Production Techniques to DIN 8580

Figure 1.1

Weldability of a component is determined by three outer features according to DIN 8528, Part 1. This also indicates whether a given joining job can be done by welding, Figure 1.2.
ility sib os g p ture ldin fac We anu M

Material Welding suitability

Weldability of a component
in De g saf sig ety n
ISF 2002

Influencing Factors on Weldability to DIN 8528 Part 1

Figure 1.2

We ld

1. Weldability of Metals Material influence on weldability, i.e. welding suitability, can be detailed for a better 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 physical characteristics of the material.
Figure 1.3



three subdefinitions, Figure

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

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 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 charac-

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

Converter: Removal of C and P through oxygen and CaO

Injection of solid material or feeding cored wires

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

Continuous casting: casting of billets, blooms, slabs

br-er01-04-E.cdr ISF 2002

teristics (see chapter 2).

Important Process Steps During Steel Production

Figure 1.4

1. Weldability of Metals

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.

+ +

+ +

+ +


+ +

+ +

+ +


Tensile strength

+ + -





Hot cracking Creep resistance

Critical cooling rate Formation of seggregations
Formation of inclusions
+(-400C) +
(+) ++

(+) ++ ++


+ -

+ -

+ + + with Al +

+ + with Mn with S

+ Increase of property ++ Strong increase of property



Decrease of property Strong decrease of property

ISF 2002

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 higher C contents, there is a
E355 (St 60), C35 (C 35) about 0,40

S185 (St 33) [EN 10 025]

C-content (%) (Melt analysis)

unlimited (up to 0,30) up to 0,21 up to 0,17 up to 0,22 up to 0,18 (up to 0,24)

Fusion weldability
Not guaranteed, however mostly no problem with low C-content up to 0,22% C: good weldable (exception: plate thickness <0,5 mm, special order condtions), as long as content of impurities (P,S etc.) not too high

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] S355J0 (St 52) E295 (St 50) Steel for mechanical engineering

up to 0,22 and higher contents of Mn and Si unlimited, about 0,30

Weldable Restricted weldability with electric arc methods, no gas welding of thin plates.

danger of hardening, and welding becomes only possible by observing special precautions (e.g. pre- and post-weld heat treatment).
Figure 1.6
E360 (St 70), C45 (C 45)

Weldable with special electrodes and mostly pre- and post-welding heat treatment. Very restricted weldability in spite of special measures
ISF 2002

about 0,50

Fusion Weldability of Unalloyed Quality Steels

1. Weldability of Metals

In addition to material behaviour, weldability is also essentially determined through the design of a component. The influence of the design is designated as welding safety, Figure 1.7.

Welding Safety

(Welding safety due to design)

e.g. Power flow in workpiece Arrangement of joints Material thickness Notch effect Stiffness differences

Stress condition
e.g. Type and Level of strain in workpiece Dimensional degree of strain Stress speed Temperature Corrosion


ISF 2002

Welding Safety

Figure 1.7

Welding Possibility

The influence of the manufacturing process to weldability is called welding possibility, Figure 1.8. For example, a preand post-weld heat treatment is not always possible, or grinding the weld surface before welding the subsequent pass cannot be carried out (narrow gap welding).
Figure 1.8
Welding preparation
e.g. Welding method Consumble type and auxiliaries Joint type Groove shape Preheating Actions in the case of unfavourable weather conditions

(welding possibility due to manufacture)

Execution of welding
e.g. Heat control Heat input Welding sequence

e.g. Post-weld heat treatment Grinding Pickling

ISF 2002

Welding Possibility

10. Testing of Welded Joints

10. Testing of Welded Joints

126 The basic test for determination of material



in test area


behaviour is the tensile test. Generally, it is carried out using a round specimen. When determining the strength of a welded joint, also standardised flat specimens 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 specimen is measured. With these measurement values, tension and strain are calculated. If is plotted over , the drawn diagram is typical for this test, Figure 10.2. Normally, if a steel with a bcc lattice structure is tested, a curve with a clear yield point is obtained (upper picture). Steels with a fcc lattice structure show a curve without yield point. The most important characteristic values which are determined by this test are: yield stress ReL, tensile strength Rm, and elongation A. To determine the deformability of a weld, a

in test area

L0 Lc Lt

total length head width width of parallel length plates tubes parallel length ) ) radius of throat
1 1 2

Lt b1 b b Lc r

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

depends on test unit b + 12 12 with a 2 25 with a > 2 6 with D 50 12 with 50 < D 168,3 L S + 60 25

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

L0 = measurement length (L0 = k S0 with k = 5,65) Lt = total length S0 = initial cross-section within test length
ISF 2002

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

Figure 10.1

Rm ReH Rel sf

ALud Ag A



bending test to DIN EN 910 is used, Figure 10.3. In this test, the specimen is put onto two supporting rollers and a former is pressed through between the rollers. The distance of the supporting rollers is Lf = d + 3a (former diameter + three times specimen thickness). The backside of the specimen (tension side) is observed. If a surface crack develops, the test will be stopped and the angle to which the specimen could be bent is measured. The

s Rm RP0,2 RP0,01 sf

0,2 % 0,01 % Ag A

ISF 2002

Stress-Strain Diagram With and Without Distinct Yield Point

Figure 10.2

10. Testing of Welded Joints


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 specimens thickness. on notch the test of plate effects.

thickness is Side the

transverse and longitudinal bending

specimens are normally only used with very thick plates, here the specimen thickness is fixed at 10 mm. A determination of the
Figure 10.3

toughness of a material or 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. notch Figure 10.4 (Iso-Vrepresents sample shape, shape 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-


Weld centre

Designation VWS a/b (fusion weld)

Fusion line/bonding zone

perature, and the average values as well as


VWS a/b



the range of scatter are entered on the impact energy-temperature diagram (AV-T curve).

VWT 0/b


VHT 0/b

This graph is divided into an area of high impact energy values, a transition range, and an

VWT a/b

VHT a/b
a a

area of low values. A transition temperature is


VWT 0/b

assigned to the transition range, i.e. the rapid drop of toughness values. When the temperature falls below this transition temperature, a transition of tough to brittle fracture behaviour takes place. As this steep drop mostly extends across a certain area, a precise assignment of transition temperature cannot be carried out. Following DIN 50 115, three definitions of the transition temperature are useful, i.e. to fix T to:

VHT a/b
a RL

VWT a/b
a RL

VHT a/b
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 ISF 2002

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

Figure 10.5

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-

J 300

specimen position: core longitudinal

specimen shape: ISO V 2% Mn 1% Mn 0,5% Mn

tent. Starting with a certain Nickel content (depends also from other alloy elements), a steep drop does not happen, even at lowest temperature the steel shows a tough fracture behaviour.
0% Mn



Charpy impact energy AV

27 200 J 100 13% Ni 8,5% 5% 3,5%

In Figure 10.7, the AV-T curves of some

2% Ni 0% Ni

commonly used steels are collected. These curves are marked with points for impact en0,1% C

27 200 J 100

ergy values of AV = 27 J as well as with points where the level of impact energy has fallen to half of the high level. It can clearly be seen 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 improvement of impact energy values and in

0,4% C

27 -150

0,8% C


-50 0 Temperature


C 100
ISF 2002

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

Figure 10.6

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. Another showed treatment imthe (conprovement chanical

application of a thermometrolled rolling during heat treatment). The application of this treatment resulted in an increase of strength and
Figure 10.7

10. Testing of Welded Joints


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
0,55h 0,25 1,2h 0,25

test speed in the impact energy test, thus there are no reliable findings about crack growth and fracture mechanisms.


b L

CT - specimen

Figure 10.8 shows two commonly used specimen shapes for a fracture mechanics test to determine crack initiation and crack growth. The lower figure to the right shows a possibility how to observe a crack propagation in a compact tensile specimen. During the test, a current I flows through the specimen, and the tension drop above the notch is measured.

h 1,25h 0,13 specimen height h = 2b 0,25 specimen width b total crack length a = (0,50 0,05)h test load P

2,1h S specimen width b


SENB -specimen 3PB

bearing distance S = 4h total crack length a = (0,50 0,05)h U F UE,aE U UO V

sample height h = 2b 0,05 F,U crack initiation


ISF 2002

As soon as a crack propagates through the material, the current conveying cross section decreases, resulting in an increased voltage
Figure 10.8

Fracture Mechanics Test Sample Shape and Evaluation

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 stressF

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


the stressed

cross-section. If the voltage drop is plotted over the force, then the start of crack initiation can be determined with suitable accuracy, and the crack propagation can be observed.
d1 d


Hardness Testing to Brinell and Vickers

Figure 10.9

10. Testing of Welded Joints


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 on soft materials up to 450 BHN (Brinell
0,200 mm

3 6 7

100 0

0,200 mm

Hardness Number). Hardness testing to Vickers is analogous.

4 5 3 8

3 10
130 30 0

1 6

3 7

4 5 3 8

3 10

specimen surface 100

specimen surface 130

hardness scale

0,200 mm

0,200 mm

hardness scale

This method is standardised to DIN 50133. Instead of a ball, a diamond pyramid is pressed into the workpiece. The lengths of the two diagonals of the impression are measured and the hardness value is calculated from their average and the test load. The impressions of the test body are always geometrically similar, so that the hardness value is normally independent from the size of the test load. In practice, there is a hardness increase under a lower test load because of an increase of the elastic part of the deformation.

8,9 10

reference level for measurement

6 8,9 7 10

reference level for measurement

30 0

Abbreviation 1 2 3 4 5 6 7 8 9 F0 F1 F t0 t1 tb e HRC HRA cone angle = 120 test preload test load total test load = F0 + F1

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

radius of curvature of cone tip = 0,200 mm

penetration depth in mm under test preload F0. This defines the reference level for measurement of tb. total penetrationn depth in mm under test load F1 resulting penetration depth in mm, measured after release of F1 to F0 resulting penetration depth, expressed in units of 0,002 mm: tb / 0,002 Rockwell hardness = 100 - e HRB HRF e =


Rockwell hardness = 130 - e


ISF 2002

Hardness Test to Rockwell

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 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

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 workpiece and whose hardness is then determined in a test machine. If a workpiece on-site is to be tested, a dynamical hardness test method will be applied. The advantage of these methods is that measurements can be taken on completed constructions with handheld
br-er10-11.cdr ISF 2002

reference bar


Poldi - Hammer

Figure 10.11

10. Testing of Welded Joints


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 kp 4.0

the steel. The measurement tip is excited to emit ultrasonic oscillations by a piezoelectric crystal. The test tip (diamond pyramid) penetrates the workpiece under the test pressure


2.0 Federweg

caused by the spring force. With increasing 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

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-


ISF 2002

Figure 10.12

10. Testing of Welded Joints


urement is carried out quickly and easily. Measurements can also be carried out in confined spaces. This measurement method is not yet standardised.

+ tension

m > a

m < a

m = a

m = 0
Dye penetrant method crack is free, surface is clean

all materials with surface cracks

compression -

m < a

m = a

m > a


crack and surface with penetrant liquid cleaned surface, dye penetrant liquid in crack surface with developer shows the crack by coloring Magnetic particle testing

pulsation range (compression)

alternating range

pulsation range (tension)

Whler line

II failure line Stress

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

III 0 1 10 102 103 104 105 106 Fatigue strength (endurance) number lg N
I area of overload with material damage II area of overload without material damage III area of load below fatigue strength limit


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


ISF 2002

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 Whler 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 Whler diagram is shown in Figure 10.13. The upper line, the Whler line, indicates after how many cycles the specimen ruptures under tension amplitude a. The

10. Testing of Welded Joints


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

Application Mainly for defects with orientation in radiation direction.

W ire diameter
Tolerated deviation

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

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

mm 0,03




Abbreviation FE 1/7 FE 6/12 FE 10/16 CU 1/7 CU 6/12 W ire number to Table 1 1 to 7 6 to 12 10 to 16 1 to 7 6 to 12 10 to 16 1 to 7 6 to 12 10 to 16 W ire length mm 50 50 or 25 50 or 25 50 50 50 or 25 50 50 50 or 25 aluminium copper W ire material Material groups to be tested

mild steel

iron materials

radiation source workpiece

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

copper, zink, tin and its alloys aluminium and its alloys

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 ISF 2002


ISF 2002

Non-Destructive Test Methods Radiographic Testing

Determination of Picture Quality Number to DIN 54105

Figure 10.15

Figure 10.16

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.






specimens taken out of the workpiece and a partly very accurate sample preparation. A testing of completed welded constructions is impossible, because this would require a destruction of the workpiece. This is the reason why various non-destructive test methods were developed, which are not used to determine technological properties but test the
ISF 2002

sound head oil coupling workpiece defect ultrasonic test device radiation pulse defect echo backwall echo


workpiece for defects. Figure 10.14 shows

Non-Destructive Test Methods Ultrasonic Testing II

Figure 10.17

10. Testing of Welded Joints 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
Figure 10.18

size of the smallest visible defect. Radiation 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


pulse, backwall and defect echo is carried out with an oscilloscope.

Figure 10.19

Ultrasonic Testing of Fillet Welds

10. Testing of Welded Joints


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

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

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

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

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


ISF 2002


ISF 2002

Defect Identification with Ultrasound

Defect Identification With Ultrasound

Figure .10.20

Figure 10.21

Figures 10.20 and 10.21 show

macro section



display of various defects on an oscilloscope. A cor50

base material

ferrite + perlite

coarse grain zone


rect interpretation of all the signals requires great experience, because the

2,5 mm

fine grain zone

ferrite + perlite

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


shape of the displayed signals is often not so clear. Figure 10.22 illustrates the potential of metallographic examination. Grinding and


Metallographic Examination of a Weld

Figure 10.22

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



% Cr

% Fe 80

reflecting the light in a different way. The macrosection, i.e. without magnification, gives a complete survey about the weld and fusion line, size of the HAZ, and sequence of solidification. Under adequate magnification, these areas can still not be distinguished precisely, however, an assessment of the developed

20 60 40 15

20 0

10 % Ni 8 10

6 4 2 0 0 200 mm 100 0 100 5

microstructure is possible. An assessment of the distribution of alloy elements across the welded joint can be carried out by the electron beam micro-analysis. An example of such an analysis is shown in Figure 10.23. If a solid body is exposed to a

Distance from fusion line

br-er10-23.cdr ISF 2002

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

Figure 10.23

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 and spectrum quantita50 20 20 50


tively. A detection limit is about 0.01 mass % with this method. Microstructure areas of a minimum diameter of about 5 m can be analysed. If the electron beam is moved across the specimen (or the specimen under the

1. weld 2. weld
0 10


20 20


axis of bending former


axis of bending former

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 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.

beam), the element distribution along a line across the

Figure 10.24

Strau - Test

10. Testing of Welded Joints


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 Strau test is

standardised in DIN 50 914. it serves to determine

tack welds measurement points



the resistance of a weld against intergranular corro40 20

base plate weld2 weld1 20 40

sion. Figure 10.24 shows the specimen shape which is normally used for that test. In addition, some details of the test method are explained.


a a a
80 120


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 a tensioning plate specimen base body guidance plates


Figure 10.26

a a

a a

Test of Crack Susceptibility of Welding Filler Materials to DIN 50129

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 bending the weld from the root. Test results record any

Tensioning Specimen for Crack Susceptibility Test

10. Testing of Welded Joints


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 groove shape 60 cross-section 60 weld metal support plate Wd./2 Wd. Wd./2 H Hc implant electrode welding direction

2 load temperature in C Tmax load in N

specimen shape


end crater

crack coefficient



x 100 (in %)

800 500

2 3

4 5 150 100 60 anchor weld t8/5 time in s rupture time

sections 60 anchor weld 80 test weld


ISF 2002


ISF 2002

Tekken Test

Implant Test

Figure 10.27

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


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

curves remain free from cracks, below these curves cracks are present. Evident for both graphs is that with increased preheat

C Implant-Test 150 Rcr = Rp0,2 = 358 N/mm

C Tekken-Test 150

Preheat temperature


Preheat temperature


50 starting cracks crack-free ml/ 40 100 g



fractured starting cracks crack-free 0 10 20 30 ml/ 40 100 g

higher hydrogen contents are tolerated without any crack development because of the much better hydrogen effusion. If both graphs are com-


20 0 10 20 30

Diffusible hydrogen content


Test Result Comparison of Implant and Tekken Test

Figure 10.29

pared 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


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 conditions. under This
capillary side meniskus1 to pump hydrogen under reduced pressure evacuated VT

air pressure B


weight are used to calculate,

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

commonly used method to determine the hydrogen

Figure 10.30

Burettes for Determination of Diffusible Hydrogen Content

content in welded joints.

volume and the coupons

2. TTA / TTT - Diagrams

2. TTA / TTT Diagrams

9 An essential feature of low alloyed ferrous materials is

a -Iron body-centered

g -Iron face-centered

the crystallographic transformation of the bodycentred cubic lattice which is stable at room temperature (-iron, ferritic structure) to the face-centred cubic lattice (-iron, aus ISF 2002

Lattice constant 0.286 nm at room temperature


Lattice constant 0.364 nm at 900 C

tenitic 2.1.

structure), The


Body- and Face-Centered Lattice Structures


where this transformation occurs, is not constant but depends on factors like

Figure 2.1

alloy content, crystalline structure, tensional status, heating and cooling rate, dwell times, etc.. In order to be able to understand the basic
L1 S TsA T1 T2 S+ a 3 2 5 So 4 Li TsB L1

processes it is necessary to have a look at the basic processes occuring in an Figure 2.2 shows the state of a binary system with
a 1

Temperature T

a - ss

complete solubility in the liquid and solid state. If the melting of the L1 alloy is cooling down, the first crystals of the composition c1 are formed with reaching the temperature T1. These crystals are depicted as mixed crystal , 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 enFigure 2.2
A (Ni) c1 c2 c0


Concentration c

B (Cu)

Temperature T




Time t


ISF 2002

Binary System With Complete Solubility in Liquid and Solid Phase

2. TTA / TTT Diagrams


riched with component B, following the course of line Li (liquidus line, up to point 4). In parallel, always new and B richer -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 TsA S 2 1 TsB Li Li L2 L1 L2

formation on transformation behaviour in relevant specialist literature. The definite use of the principles occurs in the iron-iron carbide diagram.

Temperature T


So Te

S+B 3 4




c1 Concentration c


Temperature T

Time t
ISF 2002

Transformation behaviour of carbon containing iron in the equilibrium condition is described by the

Binary System With Complete Solubility in Liquid Phase and Complete Unsolubility in Solid Phase

Figure 2.3

2. TTA / TTT Diagrams


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 system is shown by a
dmelt + d - solid solution melt melt + graphite

dashed, the metastable by a solid line, Figure 2.4.

solid sol.

d -+gsolid sol. melt + austenite

Fe3C (cementite)

Temperature C


melt + cementite




austenite + graphite austenite + cementite

diagram is limited by the formation of cementite with a carbon content of 6,67 mass%. The of strict the

austenite + ferrite

ferrite ferrite + graphite ferrite + cementite


stable equilibrium metastable equilibrium

Mass % of Carbon
ISF 2002


formed carbide phase can be read off at the top Xcoordinate of the molar
Figure 2.4

Stable and Metastable Iron-Carbon-Diagram

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 - arrt (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) -iron, 1392 C: A4- point - iron,


2. TTA / TTT Diagrams 911 C: A3- point non-magnetic - iron, 723 C: A1- point (perlite point).


with carbon containing iron: 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 -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 -solid solution is formed below the solidus line A-B (-ferrite). In accordance with the peritectic transformation at 1493 C, melt (0,51% C) and -ferrite (0,10% C) decompose to a -solid solution (austenite).

The transformation of the -solid solution takes place at lower temperatures. From -iron with C-contents below 0.8% (hypoeutectoidic alloys), a low-carbon -iron (pre-eutectoidic ferrite) and a fine-lamellar solid solution (perlite) precipitate with falling temperature, which consists of -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 -iron because of falling carbon solubility.

2. TTA / TTT Diagrams


The most important distinguished feature of the three described phases is their lattice structure. - and -phases are cubic body-centered (CBC lattice) and -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 -iron) is the result. However, diffusion speed in -iron is always at least 100 times slower than in -iron because of the tighter packing of the -lattice.

Although - and -iron show the same lattice structure and properties, there is also a difference between these phases. While -iron develops of a direct decomposition of the melt (S ), -iron forms in the solid phase through an eutectoidic transformation of austenite ( + Fe3C). For the transformation of non- and low-alloyed steels, is the transformation of ferrite of lower importance, although this -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 sysAc3

tems. Figure 2.5 shows a schematic cut through the


three phase system Fe-M-C. During precipitation, mixed carbides of the general composition M3C develop.
br-eI-02-05.cdr ISF 2002

In contrast to the binary

Description of the Terms Ac1b, Ac1e, Ac3

system Fe-C, is the three

Figure 2.5

2. TTA / TTT Diagrams


phase system Fe-M-C characterised by a temperature interval in the three-phase field + + M3C. The beginning of the transformation of + M3C to is marked by Aclb, the end by Acle. The indices b and e mean the beginning and the end of transformation. The described equilibrium diagrams apply only to low heating and cooling rates. However, higher heating and cooling rates are present during welding, consequently other structure types develop in the heat affected zone (HAZ) and in the weld metal. The structure transformations during
Figure 2.6

ISF 2002

TTA Diagram for Isothermal Austenitization

heating and cooling are described by transformation diagrams, where a temperature change is not carried out close to the equilibrium, but
ASTM4; L=80m ASTM11; L=7m

at different heating and/or cooling rates. A representation during of the transformation austenitizing processes isothermal

shows Figure 2.6. This figure must be read

20m 20m

exclusively along the time axis! It can be recognised 800 C. that several transformations austenite means during isothermal austenitizing occur with e.g. Inhomogeneous 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


br-er02-07.cdr ISF 2002

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

TTA-Diagram for Continuous Warming

Figure 2.7

2. TTA / TTT Diagrams


mogeneous austenite is passed. A growing of the austenite grain size (to ASTM and/or in m) 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 obvious that the Fe-C diagram is only valid for infinite long dwell times and that the TTA diagram applies only for one individual alloy. Figure 2.9 shows the dif-

Ac1e Ac1b


ISF 2002

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

ferent and down.

time-temperature subsequent cooling

passes during austenitizing

Figure 2.8

The heating period is comAc3 continuous Ac1e

posed of a continuous and an isothermal section.



During cooling down, two different ways of heat control can be distinguished: 1. : During continuous temperature control a


ISF 2002

Heating and Cooling Behaviour With Several Heat Treatments

cooling is carried out with a constant cooling rate out of

Figure 2.9

2. TTA / TTT Diagrams 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
br-eI-02-10.cdr ISF 2002






Isothermal TTT-Diagram of Steel C45E (Ck 45)

marked with F, P, B, und M represent areas where ferrite, perlite, Bainite and martensite are formed. The

Figure 2.10

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


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
27 19 40


the same content of carbon as the steel Ck 15. Here you can see that all transformation processes are strongly postponed in relaTime


235 220


tion to the mild steel. A

ISF 2002


Continuous TTT-Diagram of Steel C15E (Ck 15)

completely transformation

martensitic is carried

Figure 2.11

out up to a cooling time of about 1.5 seconds, comC 0,13 Si 0,31 Mn 0,51 Al P S 0,023 0,009 0,010 Cr 1,5 Mo 0,06 Ni 1,55 V < 0,01

Chemical composition %
1000 C 900 800 700

pared with 0.4 seconds of Ck 15. In addition, the completely diffusion controlled transformation processes of the perlite area are postponed to clearly

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

Ac3 Ac1
75 75 75

600 500




55 67

10 22

75 75 25 25

25 23

60 72 55 37



300 200 100

417 400 396

22 9

longer times.
167 152 151

314 304 287 268 251 224 192

0 10




10 Time



ISF 2002


The hypereutectoid steel C 100 behaves completely different, Figure 2.13. With this carbon content, a pre-


Continuous TTT-Diagram of Steel 15 CrNi 6

Figure 2.12

2. TTA / TTT Diagrams

18 eutectoid ferrite formation cannot still be car-

Chemical composition %
1000 C 900 800

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
austenitizing temperature 790C 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 "austenitizing temperature means the temperature, where the workpiece transforms to an austenitic microstructure in the course of a heat treatment. Dont mix up this temperature with the AC3 temperature, where above it there is only pure austenite. In addition you can see that only martensite is formed from the austenite, provided that the cooling rate is sufficiently high, a formation of any other



A+C 100 P








100 AC1b

600 500 400 300 200 100 0

1000 C 900 800

2 15 180

914 901 817 366







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



C A P 100 100 100








500 400 300




MS RA40 M 876 887 867 496 457 442 347 289 246 227 200

microstructure is completely depressed. With 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 called critical cooling rate.
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
100 101 102 103 104 s 105



ISF 2002

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

Figure 2.13

1000 C 800 900C 1300C

Ar3 Ar1 Perlite 100%


Cr, V, Mo

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

A F 600 MS 400 M B P

Cr, V, Mo

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

Low hardening temperature (special carbides), austenite above bainite

200 100


Structure distribution

% 75 M M

Ms Martensite C, Mn, Cr, Ni, Mo, V, high hardening temperature, preprecipitation in bainite Co, Al, deformation of austenite, low hardening temperature

50 25 0 10-1 1 10 102 s 103

Transition time
br-er02-14.cdr ISF 2002 br-er02-15.cdr

Cooling time (A3 to 500C)

ISF 2002

Influence of Alloy Elements on Transformation Behaviour of Steels

Temperature Influence on Transformation Behaviour of Steels

Figure 2.14

Figure 2.15

2. TTA / TTT Diagrams


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
S355J2G3 (St 52-3) Chemical composition % C 0,16 Max. temperature 1350 C Si 0,47 Welding heat cycle Cu 0,17 Ni 0,06 Mn P S Al N Cr 1,24 0,029 0,029 0,024 0,0085 0,10

an extension of the diffusion lengths which must be passed during the transformation. As a result, the "noses" in the TTT diagram are shifted to longer times. The lower part of the figure shows the proportion of

900 C 800








449 420 400 363 334 324 270 253 251 249 243


8 10


20 Time


60 80 100



ISF 2002




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

Figure 2.16

Bainite depending on cooling time. You can see that with higher austenitizing temperature the start of

15 Mo 3 Chemical composition %
900 C 800

Max. temperature 1350 C C 0,16 Si 0,30 Mn P S Mo 0,68 0,012 0,038 0,29

Welding heat cycle

Bainite formation together with the drop of the martensite proportion is clearly shifted to longer times. As Bainite formation is not so much impeded by the coarse austenite grain as with the completely diffu-

Ac3=861C Ac1=727C F
32 8 4 45 17 53 32





MS 14

74 87 95






















8 10

20 Time


60 80 100



sion controlled processes of ferrite and perlite formation, the maximum Bainite proportion is increased from about 45 to 75%.


ISF 2002

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

Figure 2.17

2. TTA / TTT Diagrams


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
%C 1 1000 C 800 0 1000 C 800 F 600 400 MS B M 200 200 0 10-1 10

0,45 0,5

iron-carbon diagram was developed as an equilibrium diagram for infinite long cooling time and that



a TTT diagram applies always oy for one alloy.



10 Time



ISF 2002


Relation Between TTT-Diagram and Iron-Carbon-Diagram

Figure 2.18

3. Residual Stresses

3. Residual Stresses

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

examples in Figure 3.1.

grinding disk pressure weld tension





causes of origin. In a produced workpiece, material, production-, and wearcaused residual stresses are overlaying in such a



ISF 2002

Various Reasons of Residual Stress Development

way that a certain condition of residual stresses is created. Such a workpiece shows in service more or

Figure 3.1

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 distribution of residual stresses. Residual stresses, which build-up around dislocations and other lattice imperfections (III), superimpose within a
forming deforming residual stresses due to inhomogenuous deformationanisotropy separating residual stresses due to machining joining residual stresses due to welding plating layer residual stresses

relevant material



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

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

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

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

grain causing stresses of the 2


changing material characteristics induction hardening, case hardening, nitriding

type and if spreading


e.g. thermal residual stresses

around several grains, bring

br-eI-03-02e.cdr ISF 2002

out residual stresses of the 1 type. The formation of

Development of Residual Stresses


Figure 3.2

stresses in a transition-free

3. Residual Stresses


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

tension s

General Definition of the Term Residual Stresses

s III s II
+ 0 -

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.


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

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.

grain boundaries

sE = s I + sII sI sII sIII



< < <

= residual stresses between several grains = residual stresses in a single grain = residual stresses in a point
ISF 2002


ISF 2002

Definition of Residual Stresses

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

Figure 3.3

Figure 3.4

homogeneous. The left part of Figure 3.5 shows the T-tcurve of three different measurement points in the cylinder.

1000 C 900 1000 C 750 1 35 mm diameter water cooling 500 1 edge 2 50 % radius 3 core MS 2 3

0s 10 s 15 s 20 s 25 s 5s 1s

800 700 600 500

Figure 3.6 shows the results of quenching on the stress condition in the cylinder. At the beginning of cooling, the cylinder edge starts shrinking faster than the core (upper

35 s 400 300 200 45 s 53 s


0 -2 10

68 s 10-1 10-0 101 102 Cooling time 103 s 104 100 280 s 0 17,5

14 10,5

7 3,5

0 3,5 Radius

7 10,5
ISF 2002

mm 17,5

figure). Through the stabilising effect of the cylinder core,

Figure 3.5

Temperature in a Cylinder During Water Cooling

3. Residual Stresses


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.

tension pressure

300 N/mm 200 E D


Stresses in the central rod

Volume differences between edge and core at start of cooling


tension pressure


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


pressure tension

-200 B



-300 0
ISF 2002


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




Temperature of the central rod

br-er03-07e.cdr ISF 2002

Volume Changes During Cooling

Residual Stress Development by Warming the Central Rod

Figure 3.6

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


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 lattice deformation). In the case of a homogenous transition, the weld will consequently unfold pressure stresses. If the transition of
+x +y -x

the edge areas happens earlier than the transition of the slower cooling core, plastic deformations of the core area may be present similar to quenching (see above: quenching stresses). In this case, the weld surface will
3. Transformation stresses
-x +x

-y 1. Shrinking stresses
+s +y

2. Quenching stresses
-x +x

-s -y

show tensile stresses after cooling. Generally these mechanisms cannot be separated accurately from each other, thus 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 mechanisms makes a forecast of the remaining residual stress condition difficult.
Figure 3.8
homogenuous transformation

+s +y
inhomogenuous transformation

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

+s +y





-s -y

-s -y
ISF 2002

Stress Distributions and Superpositions Perpendicular to Welded Joint

3. Residual Stresses

26 Figure 3.9 shows the building-up of residual


Temperature distribution
1. cut A-A x
DT ~ 0

Stress distribution sX

stresses crosswise to a welded seam in analogy to the 3-rod model of Figure 3.7. This figure considers only shrinking residual stresses. Before application of welding heat, the seam area is stress-free (cut A-A). At the weldpool


2. cutt B-B

weldpool B
area of plastic deformations

the highest temperature of the welding cycle


can be found (cut B-B), metal is liquid. At this point, there are no residual stresses, because molten metal cannot transmit forces at the weldpool. Areas close to the joint expand

3. cut C-C

M' 4. cut D-D

DT = 0 residual stresses

through welding heat but are supported by areas which are not so close to the seam. Thus, areas close to the joint show compres 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.
1 15 mm 3
15 mm material S235JR (St 37)

103 a a

Figure 3.10 shows how much residual stresses are influenced by constraining effects of adjacent material. The resulting stress in the presented case is calculated according to Hooke: = E Elongation is calculated as l/a ( l is the length change due to shrinking). With conFigure 3.10
1. 2. 3. 4. 5. a = 100 mm a = 150 mm a = 200 mm a = 250 mm a = 300 mm s = 800 N/mm s = 530 N/mm s = 400 N/mm s = 300 N/mm s = 270 N/mm


ISF 2002

Shrinking Stresses in a Firmly Clamped Plate

3. Residual Stresses


stant joint volume will shrinking and l always have the same value. Thus the elongation 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.

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

Longitudinal expansion Dl

firm clamping

t rri fe





force sensor thermo couples links


nit ic

to calculator 800 N 1000 C

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










heat affected zone


400 force

200 temperature

Temperature [C]


0 -1 10



102 Time





ISF 2002


ISF 2002

Longitudinal Expansion of Various Steels

Force Measurement During Cooling of a Weld

Figure 3.11

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


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 consumable, the transition temperatures and/or temperature areas for the individual zones of the welded joint can be determined. With these temperature it can be clearly determined in which part of

steel consumable electrode sample shape (V-groove, 60) type of welding position of the HAZ

austenitic austenitic

S690QL (StE 70) austenitic

S690QL (StE 70) high-strength

surface weld

surface weld

surface weld

residual stress distribution sL



data and with the course of

ISF 2002

the curve the force drop is caused by the transition of the welding consumable and in
Figure 3.13

Influence of Material Combination on Residual Stress Distribution in a Weld

which part by transition in the heat affected


zone (HAZ). These results can be used to determine the longitudinal residual stresses transversal to the joint, as shown in Figure 3.13. During welding of austenitic transition-free materials only tensile residual stresses are caused in the welded area according to Figure 3.8. If an austenitic electrode is welded to a StE 70, transitions occur in the area of the heat affected zone which lead to a decrease of tensile stresses. If a high-strength electrode which has a martensitic transition, is welded



140 %
Angle change

100 80 60 40 20

f = 1

f = 3

f = 7

f = 13




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.


ISF 2002

Influence of Welding Sequence on Angle Distortion

Figure 3.14

3. Residual Stresses


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 as conditionally destructive, Figures 3.15 and 3.16. In both cases, present residual stresses are released through partial material removal and the resulting deformations are then strain
br-eI-03-15e.cdr ISF 2002





measured by wire

gauges. An essential advantage of the borehole method is the very small material removal, the diameter of the
Figure 3.15

Residual Stress Determination Using Bore Hole Procedure

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 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

Figure 3.18 shows a surcutting in layers

assumption of stress distribution

f f 0 x y z

measured variable
bending deflection f curves reduced curves

residual stresses
sy sz tzy

vey of the completely destructive procedures of residual stress recognition.

biaxial any uniaxial locally different linear, tensile residual stresses on top, down pressure stresses tripleaxial independent of smple length sL, sT, sR


tear f

partial residual stress relief by Dsz

eT e45 eL

length change eL circumference change eT

sL sT sR


uniaxial linear symmetrically with reference to rod axis

tear f

partial residual stress relief by Dsz


ISF 2002

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


In the European Standard DIN EN

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)].

10020 (July 2000), the designations (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 grades in accordance with their

Classification in accordance with the chemical composition:


unalloyed steels stainless steels other, alloyed steels

chemical composition and the main quality classes.

Classification in accordance with the main quality class: unalloyed steels stainless steels other, alloyed steels - unalloyed quality steels - unalloyed special steels

- alloyed quality steels - alloyed special steels

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


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


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 l P = Steels for pressure vessel construction e.g. P265GH, P355M l L = Steels for pipeline construction e.g. L360A, L360QB l E = Engineering steels e.g. E295, E360 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

According to the utilization of the steel and also to the mechanical or physical properties, the steel grades of the first group are designated with different main symbols (Fig. 4.3).

Classification of steels in accordance with their designated use

Figure 4.3

4. Classification of Steels, Welding of Mild Steels


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


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) +A + AR + AS + AZ + CE + Cu + IC + OC +S + SE +T + TE +Z + ZA + ZE + ZF + ZN
1 2

Coating 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)

) 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

br-er-05-05.cdr ISF 2004

Symbols for the coating type

Figure 4.5

Symbol1)2) +A + AC +C + Cnnn + CR + HC + LC +Q +S + ST +U

treatment condition 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/mm 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

1 2

) 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,

for example +TA

br-er-05-06.cdr ISF 2004

Symbols for the treatment condition

Figure 4.6

4. Classification of Steels, Welding of Mild Steels


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
Steel for structural steel engineering further property (RR = normalised) test temperature 20C impact energy 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
br-er-05-07.cdr ISF 2002

ReH 235 MPa/mm2

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 0,20 0,21 0,26 0,15 0,18

Si 0,55

Mn 1,60

P 0,040 0,035 0,05

S 0,040 0,030 0,05 0,035 0,05

Cr / 0,30 / 0,40 0,80 /

Al / 0,020 / / /

Cu / 0,20 / 0,25 0,5 /

N 0,009 0,020 / / /

Mo / 0,1 / 0,30 /

Ni / 1 / 0,65 /

Nb / 0,05 / / /

V / 0,22 / 0,02 0,12 /

0,1 - 0,6 1 - 1,7 0,35 0,6

0,50 0,5 - 1,3 0,035 0,1 0,7 - 1,5 0,05 0,35

Tensile strength Zugfestigkeit RmRm [N/mm]

yield point Re Streckgrenze eH H [N/mm]

elongation after fracture Bruchdehnung A A [%]

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

impact energy AV Kerbschlagarbeit V [J] -20C

510-680 610-780 460-550 510-610 400-490

355 500 285 355 355

20-22 16 >18 22 >22

27 31-47

27 21-39

49 (bei +20C)

76 (bei -10C)
ISF 2004

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


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


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


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

4. Classification of Steels, Welding of Mild Steels


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 austen13
Energy-per-unit length in kJ/cm

ite grains in the HAZ and the width

18 36

Austenite grain size index according to DIN 50601

11 9 9 12







coarsened austenite grain decreases the critical cooling

3 0


0,4 0,6 Distance of the fusion line



ISF 2004

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

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

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 With

mm 1 8 6 -4

43 higher temperatures, these

precipitations dissolve and cannot suppress a grain growth any more. Steel 3 contains mainly titanium carbonitrides of a much lower grainrefining effect than that of AIN. Steel 4 is a combination of the most effective properties of steels nos. 2 and 3.
Steel 1 Steel 2 Steel 3 Steel 4 1000 1100 1200 Austenitization temperature 1300 C 1400


Grain size index according to DIN 50601

Medium fibre length

10 8 6 4


10-2 8 6 10


The importance of grain refinement for the mechanical properties of a steel is shown in Figure 4.13. Provided the temperature keeps constant, the yield strength of a steel increases with decreasing grain size. This influence on the yield point Rel is specified in 1 d the Hall-Petch-law:
ISF 2004

12 900

Steel 1 2 3 4

%C 0,21 0,17 0,18 0,19

% Mn 1,16 1,35 1,43 1,34

% Al 0,004 0,047 0,004 0,060

%N 0,010 0,017 0,024 0,018

% Ti / / 0,067 0,140

Austenite grain size as a function of the austenitization temperature

Figure 4.12

Rel = i + K to the
900 N/mm 800
Yield point or 0,2 boundary


above-mentioned law, the increase of the yield point is inversely proportional to the root of the medium grain diameter d. i

Temperature in C:

700 600 500 400 300 200 0 1

-193 -185 -170 -155


-100 -40 +20

2 3 4 5 6 -1/2 Grain size d 7 8 mm-1/2 10

stands for the internal friction stress of the grain measure material. The is a the


ISF 2004

Figure 4.13

Connection between yield point and grain size

resistance K for

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


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 Yield point Streckgrenze Plate thickness Blechdicke Weld cross-section Nahtquerschnitt Welding wire Schweidraht 1.2 1.2 Welding wire costs Schweidrahtkosten Steel costs Stahlkosten Weld metal costs Schweigutkosten Special weld costs Spez. Schweinahtkosten Costs ratio inclusive base Kostenverhltnis inklusive materials Grundwerkstoffe

S235JR (St37-2) N/mm2 mm mm2 mm

Ratio Verhltnis Ratio Verhltnis Ratio Verhltnis Ratio Verhltnis

S355J2G3 (St52-3) 345 31 370 SG3 1 1,2 2,3 5,1

S690Q (StE690) 690 14,4 100 NiMoCr 2,4 1,9 1,5 1,8

S890Q (StE890) 890 11 60 X 90 3,2 2,3 1,16 1,18

S960Q (StE960) 960 10 50 X 96 3,3 2,4 1 1

Verhltnis Ratio S235JR - S960Q 1:5 5:1 17 : 1 1 : 3,3 1 : 2,4 5,3 : 1 12 : 1 5:1

tages of microalloyed fine-grained with are structural steels in comparison tural shown steels in conventional strucFigure

215 50 870 SG2 1 1 5,3 12

Randbedingungen: Boundary condition:

Schweiverfahren = MAG welding process = MAG

Deposition rate = 3 kg wire/h, weld /shape X -60 X - 60 Abschmelzleistung =welding 3 kg Schweidraht h, Nahtform Costs ofund labour and equipment = = 60 30/h LohnMaschinenkosten DM / h Special costs = weld filler materials + welding Spez. weld Schweinahtkosten = Schweizusatzwerkstoffe + Schweien

4.14. Due to the considerably better

ISF 2004

Berechnungsgrundlage =szul = Re / 1.5 Calculation base = szul = Re/1.5


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

mechanical properties of the finegrained structural steel in comparison

Figure 4.14

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 Figure on 4.2, the
unalloyed steels

alloyed low-alloyed high-alloyed


of Figmild steel higher-carbon steel Hardening Underbead cracking

ure 4.15 divides the steels with regard to their problematic processes during welding. When it comes to unalloyed

rimmed steel cutting of segregation zones

killed steel duplex killed steel

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 grain increase in the weld interfaces pearlitic-martensitic hardening embrittlement formation of chromium carbide austenitic grain desintegration stress corrosion cracking hot cracks (sigma phase embrittlement)

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

Post-weld treatment for highest corrosion resistance

ISF 2004

steels, only ingot

Classification of steels with respect to problems during welding

Figure 4.15

4. Classification of Steels, Welding of Mild Steels


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 clearly zone segregated

where unfavourable elements like sulphur and phospho0,025 0,012 0,003

rus are enriched. During rolling, such zones are stretched

ISF 2004

fully killed steel

Figures: mass content of oxygen in %

semi-killed steel

rimmed steel

along the complete length of the rolling profile.

Ingot cross-sections after different casting methods

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 In to as, hotin

and are inclined to hardening. sensitive cracking, addition, they are

these zones, the elements phosphorus are and sulphur enriched.

Figure 4.17

ISF 2004

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

4. Classification of Steels, Welding of Mild Steels


Therefore, touching such segregation zones during welding must be avoided by all means. In the case of lowalloy
Microstructures Ferrite Austenite Perlite (granular) Perlite (lamellar) Sorbite Troostite Cementite Martensite Average Brinell Hardness (Approximately) 80 250 200 300 350 400 600 - 650 400 - 900

steels, of must

the HAZ

problem hardening welding observed. ness

during be Figure of

4.18 shows hardvalues various microstruc ISF 2004


tures. The highest hardness martensite values and can be found with

Hardness of Several Microstructures

Figure 4.18

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 influence of the martensite content in the HAZ on the crack formation of welded martensite joints. formaHardening through tion is not to be
Br-er-05-19.cdr ISF 2004

maximum hardness HV root cracking presumable root cracking possible no root cracking sufficient operational safety without heat treatment 400 HRC 41

strength, calculated at max. hardness N/mm2 1290

with maximum martensite content % 70

400 - 350 350 280

41 - 36 36 28

1290 - 1125 1125 900

70 - 60 60 30

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.

expected with pure carbon steels up to about 0,22%,

Figure 4.19

Influence of Martensite Content

4. Classification of Steels, Welding of Mild Steels


because the critical cooling rate with these low C-contents is so high that it normally wont 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 - qu. = C + Mn Cr + Mo + V Cu + Ni + + 6 5 15
Mo Ni Cu Mn Cr + Mn + + + 6 10 20 40

bon, all other alloy elements are important site when it in comes to martenformation the welding cycle, as they have substantial influence on the transformation behaviour of steels (see


C - qu. = C +

Ito and Bessyo

PCM = C +

Si Mn + Cu + Cr Ni Mo V + + + + + 5B 30 20 60 15 10
Si Mn + Cu Cr Ni Mo V + + + + + 25 16 20 60 40 15


C - qu.PLS = C +


C - qu. = C +

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



Mn + Mo Cr + Cu Ni = C+ + + 10 20 40
PLS = pipeline steels PCM = cracking parameters (%)
ISF 2002

C-qu.= carbon equivalent (%)


Definition of C - Equivalent

Figure 4.20

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 and as for the
Tp [ C]
250 100

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

200 80

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


delta Tp [C]



hardening tendency the general conditions like plate heat thickness,





0 0,2 0,3 0,4

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

0 0 5 10 15 20

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


Kohlenstoffquivalent CET [%] Carbon aquivalent

Wasserstoffgehalt HD Hydrogen content of des theSchweigutes weld metal [%]


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



delta Tp = (53 CET - 32)-Q - 53Q CET + 32 delta Tp = (53 CET 32) - 53 CET + 32

CET = 0,4 %

CET = 0,4 %

CET = 0,2 %

CET = 0,2 %

CET = 0,2 %

CET = 0,2 %


delta Tp [C]

delta Tp [C]

input, etc., are also of importance, the carbon equivalent cannot be a common limit value for the weldability.






0 0 20 40 60 80

CET =0,4 0,4 CET = %% HD = 2 2 HD Q= =1 Q 1kJ/mm kJ/mm



-100 0 0,5 1 1,5 2 2,5 3 3,5 4

d =d50 mm = 50 mm =8 HDHD =8
4,5 5

Blechdicke d [mm] Plate thickness

Wrmeeinbringen Heat input Q [kJ/mm]

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



+ (53 CET - 32) Q - 328

ISF 2005

Source: Quelle: DIN EN 1011-2

Calculation of the preheating temperatures

For the determina- Figure 4.21

4. Classification of Steels, Welding of Mild Steels


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 formaTmax

tion in the welding cycle is the cooling time. As a measure of cooling time, the

Temperature T


500 t8/5

time of cooling from 800 to 500 C (t8/5) is defined (Fig. 4.22).

t800 Time t t500 s
ISF 2004




Definition of t8/5

range was selected 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 measured temperature shows


distriTemperature T

1500 A 1000 B 500 C 0 0 50 100 150 Time t


butions in the vicinity of a weld. Peak values and dwell times depend obviously on the location of the and measurement




ISF 2004

are clearly strongly determined by the heat conduction Figure 4.23 conditions.

Temperature-time curves in the adjacence of a weld

4. Classification of Steels, Welding of Mild Steels


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
t8 / 5 = h U I 1 1 2 p l v 500 - T0 800 - T0 1 ) Uv I 500 - T 1 h N 3 800 - T0

3 - dimensional:
universal formula:

are covered by the formulas given in Figure 4.24, which provide a method of calculating the cooling time t8/5 of low-alloyed steels. In the case of a
ISF 2004

extended formula For low-alloyed steel:

t8 / 5 = 0,67 - 5 10 - 4 T0

2 - dimensional:
universal formula:

t8 / 5 =
2 2 2 h2 1 1 U I 1 - 800 - T 4 p l r c v d 2 500 T 0 0

extended formula For low-alloyed steel:

2 2 2 2 1 1 U I 1 t8 / 5 = 0,043 - 4,3 10 -5 T0 2 - h N 2 v d 500 - T0 800 - T0

formula for the transition thickness of low-alloyed steel:

d =

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

1 1 500 - T + 800 - T 0 0


three-dimensional heat dissipation, t8/5 it independent of plate thickness.

Calculation equation for two- and three-dimensional heat dissipation

Figure 4.24

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 welding method on the heat dissipation. With the same heat energy base depends input, which the is

TIG-(He)-welding TIG-(Ar)-welding MIG-(Ar)-welding MAG-(CO2)- welding Manual arc welding SA welding 0 0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9 1

transferred to the material on the

Figure 4.25

Relative thermal efficiency degree h

ISF 2004

Relative thermal efficiency degree of different welding methods

4. Classification of Steels, Welding of Mild Steels


welding method. This dependence is described by the relative thermal efficiency . The influence of the groove geType of weld 2-dimensional heat dissipation 1

weld factor 3-dimensional heat dissipation 1

ometry is covered by seam factors according to

0,45 - 0,67


Fig. 4.26. Empirically determined, these factors were introduced for an easier calculation. For other groove geometries, tests Figure 4.26 to measure the cooling time are recommended.





ISF 2004

Weld factors for different weld geometries

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 TA=20C 25 cooling time t8/5 [s] 10 20 30 TA=200C 30 40 60 100 2-dimensional 0 0 10 20 30 40 50 0 10 20 30 40 50 2-dimensional 60 80 100 150 40 50


Plate thickness

3 3-dimensional 2 3-dimensional

Heat input E.h.Nn [kJ/cm]

Br-er-05-27.cdr ISF 2004

Transition From Two to Three Dimensional Heat Flow

Figure 4.27

4. Classification of Steels, Welding of Mild Steels

51 Fig. 4.28 shows the possible range of


heat input depend-spray arc


ing on the electrode diameter. It is clear that a relatively large working


Heat input

4 -short arc

range is available
2,5 3,0 4,0 5,0 SA-welding

3,25 4 5 6 Manual metal arc welding

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

ISF 2004


welding A the be of can

procedures. variation length


Heat Inputs of Various Welding Methods


Figure 4.28

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) Current intensity (A)

2,5 90 75

3,25 135 120

4,0 180 140

5,0 235 190

6,0 275 250

input during manual metal arc welding. The shorter the fused electrode distance, i.e., the shorter the extracted length, the higher the energyper-unit length.

Current intensity (A)

35 kJ/cm 25
Energy-per-unit length

20 15

6,0mm x 450mm

5,0mm x 450mm

10 5 0

4,0mm x 450mm 3,25mm x 350mm 2,5mm x 350mm

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


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


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 20 10 7 50 40 30 20 10 7 50 40 30 20 10 7 50 40 30 20 10 7 5

T0 200C 150C 100C 20C d = 7,5 mm

Cooling time t8/5 in s

T0 200C 150C 100C 20C d = 10 mm

T0 200C 150C 100C 20C d = 15 mm

T0 200C 150C 100C transition to 3-dimensional heat flow 20C d = 20 mm

7 8 9 10

15 20


kJ/cm 50
ISF 2004

Heat input E

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


50 mm 40

With the transition to thicker plates,

aera of 3-dimensional heat flow

Transition thickness d

the diagrams in Fig. 4.31 apply. The

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 present. For the three-dimensional heat dissipation, the lower diagram applies
kJ/cm 50

20 15

10 9 8 7 50 s 40 30

area of 2-dimensional heat flow

7 8 9 10

15 20


Heat input E

where the same information can be determined, independent of plate thickness, as with Fig. 4.30.

Cooling time t8/5

20 15

25 T

C 0 C 0 C 10 0 C 20 C 15


10 9 8 7

7 8 9 10

15 20


kJ/cm 50
ISF 2004

Heat input E

Dependence of E, T0, t8/5 And d

Figure 4.31



be35 V 30
gas composition: C1 100% CO2 M21 82% Ar + 18% CO2 M23 92% Ar + 8% O2

tween current and voltage for MAG

Welding voltage

C1 M21 M23

welding is shown in Fig. 4.32 and the used shielding gas is one of the parameters. Welding voltage and welding current, or




short arc contact tube distance ~15mm

mixed arc

spray arc

contact tube distance ~19mm

150 3,5

200 4,5 5,5

250 Welding current 7,0 Wire feed 8,0

A 9,0

300 10,5 m/min

ISF 2004

wire feed speed, determine the type of arc.

Figure 4.32

Dependence of Current And Voltage During MAG-Welding, Solid Wire, 1.2 mm

4. Classification of Steels, Welding of Mild Steels


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

strates the dependence of plate thickness, heat input E and cooling time t8/5 for fillet welds at a preheating

60 kJ/cm 50
toughness affection

fillet welds T0= 150 C

kJ/cm 59 53 47

temperature of T0 = 150 C. If d and t8/5 are given, the acceptable range of heat input can be determined with the help of this diagram. The kinks of the curves mark the transition between two-dimensional and threedimensional heat dissipation.


Heat input E SA - welding


35 30 25 20 15
6s 10s 15s

41 35 29 23 18 12
cracking tendency

10 5 0 0 5 10 15

6 mm 0 40
ISF 2004

20 25 30 Plate thickness


Permissible E-Range During SA - And MAG - Welding

Figure 4.33

Fig. 4.34 shows the same dependence for butt welds with V groove preparation.
h'UP = 1 h'MAG = 0,85 dU max = 34 mm dU min = 15 mm F3 = 0,9 F2 = 0,9 t8/5 max = 30 s t8/5 min = 6 s Emax = 49 kJ/cm Emin = 10 kJ/cm

Heat input E MAG - weldind

60 kJ/cm 50
toughness affection butt welds T0= 150 C

70 kJ/cm 59


53 47

Heat input E SA - welding

40 35 30 25 20 15 10 5 0 0 5 10 15
cracking tendency
15s 25s

41 35 29 23 18 12 6




20 25 30 Plate thickness


0 40
ISF 2004


Permissible E-Range During SA - And MAG - Welding

Figure 4.34

Heat input E MAG - welding

4. Classification of Steels, Welding of Mild Steels


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 V 25 kJ/cm 20
2 1 3 4 5 6 7 8 1 2 3 4 5 6 7 8

have been taken from Figures 4.32 and 4.34 and apply only for related wire welding conditions like wire diameter, feed,









300 275 250 225 200 175 150 125 5.5 4.5 3.5 3.0

vZ(m/min) 10.5 9.0 8.0 7.0

wor king

Heat input E

15 10 5 0 10 15




30 35 40 45 Welding speed vS

50 cm/min 60

voltage, etc.
ISF 2004

MAG/ M21 (82% Ar, 18% CO)


E as a Function of Welding Speed, Solid Wire, 1.2mm

Figure 4.35

Figure 4.36

60 kJ/cm 50
SA - welding
toughness affection butt welds T0= 150 C

a reading example for such diagrams (according to DVSReference Nr. 0916). In this example, a plate thickness of 15 mm and a cooling time t8/5 between 10 and 20 s are given. In this case, the maximum and 300A is used. Sheet

70 kJ/cm 59
MAG - welding

curve V A









300 275 250 225 200 175 150 125 5.5 4.5 3.5 3.0


53 47

vZ(m/min) 10.5 9.0 8.0 7.0

40 35

41 35 29

Heat input E

Heat input E

30 25 20 15 10 5 0 0 5 10 15
cracking tendency


25 kJ/cm 20
2 1 3 4 5 6 7 8

23 18

heat input E



16 15 13
10 5 0 10 15





16 12 13
6 0 40

20 25

50 cm/min 60

20 25 30 Plate thickness


30 35 40 45 Welding speed vS


ISF 2004

Determination of Welding Speed for MAG - Welding

Figure 4.36

cooling time for MAG welding is 15 s. A solid wire with a diameter of 1.2 mm at 29V 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 the determination of the microstructural composition and cooling time subject to peak temperatures which occur in the welding cycle. In the lower diagram, the point of the plate
Peak temperature Temperature

800 C 700 F 600 B 500 400 300 200 HV30=400 1400 C 1200 M B+M F+B F+P 300 200 M Peak temperature 1000C 1400C P

thickness at the top line is linked with the point of heat input at the lower line. The point of intersection of the linking line with the middle scale represents the cooling time t8/5 . If the peak temperature of the welding cycle is known, one can read from the middle diagram in which transition field the final microstructures are formed. The advantage of the determination of microstructures compared

1000 Arc3 800 Arc1

600 plate thickness 40

1 30 25 20

10 15

100 10 9 8 7


1000 5 mm 4


three-dimensional two-dimensional

300 200 100

2 3



50 100 200 400 s 1000

0 100 C 200


preheating temperature

energy-per-unit length 6





50 kJ/cm 70
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


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

steels. They can be singled out as heat- and scale-resistant and stainless steels, depending on service temperature. Stainless steels are used at room temperature conditions and for water-

scale- and heat-resistant steels

perlitic martensitic









non-stabilized (austenite with delta-ferrite) X12CrNi18-8


stabilized (austenite without delta-ferrite) X8CrNiNb16-13

ISF 2002

based media, whilst heatand scale-resistant steels are applied in elevated 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 d A4 g A3 a A4 g A3 a(d) A3 a g T T A4 d

Classification of Corrosion-Resistant Steels

Figure 5.1

phase diagram (left figure), Figure 5.2 shows the effects of two different groups of alloying elements on the equilibrium diagram.
Alloy elements in % Nickel Manganese Cobalt

Alloy elements in %

Alloy elements in % Chromium Vanadium Molybdenum Aluminium Silicon




chromium as the most important element cause a

ISF 2002


Modifications to the Fe-C Diagram by Alloy Elements

strong reduction of the aus-

Figure 5.2

5. Welding of High-Alloy Steels, Corrosion


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 Carbon l l l Chromium l Nickel l l Oxygen l Niobium l Manganese l l Steel type, no. All types l l l All types l All types Special types l 1.4511,1.4550, 1.4580 u.a. All types l l Effect Increases the strength, supports development of precipitants which reduce corrosion resistance, increasing C content reduces critical cooling rate Works as ferrite developer, increases oxidation- and corrosion-resistance Works as austenite developer, increases toughness at low temperature, grain-refining Works as strong austenite developer (20 to 30 times stronger than Nickel) Binds carbon and decreases tendency to intergranular corrosion

Austenite developers cause an extension of the austenite area to Figure 5.2 (right figure) and form a purely austenitic and transformation-free steel. The table in Figure 5.3 summarises the effects of some selected elements on high alloy steels.

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

Silicon l Titanium l l Aluminium l Copper l l 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 Type 17-7 PH, 1.4505, 1.4506 l l Improves corrosion resistance against certain media, decreases tendency to stress corrosion cracking, improves ageing
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 -phase, which decomposes in the lower temperature range into a low-chromium solid solution and a chromium-rich -solid solution. Both, the development of the -phase and of the unary --decomposition cause a
Figure 5.4

5. Welding of High-Alloy Steels, Corrosion


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 the binary system Fe-Ni
1600 C 1400 1200 g S+d S+g

Fe Ni3

decomposition range take place.



in the lower temperature Along two cuts through the ternary system Fe-Cr-Ni, Figure 5.6 shows the most
Fe Ni3


1000 800 600 a 400 200 0 Fe a+g

90 % Ni
ISF 2002



develop in high alloy steels.

10 20 30 40 50 Nickel 60 70 80


Binary System Fe - Ni

A solidifying alloy with 20% Cr and 10% Ni (left figure) forms at first -ferrite. ferrite is, analogous to the

Figure 5.5
70 % Fe 1600 C 1500 1400 1300
S+d S S+d+g S+g

60 % Fe 1600 C 1500 1400 1300 1200

S+d S S+d+g S+g

Fe-C diagram, the primary from the melt solidifying body-centred cubic solid solution. However -ferrite


d d+g g


1100 1000 900 800

d+ g+ s

1100 1000 900 800

d+s g+s

is developed by transformation of the austenite, but

d+g+s d+s g+s

is of the same structure from the crystallographic

15 25 20 20 25 15 % Ni % Cr


700 5 25 10 20 15 15 20 % Ni 10 % Cr 0 40 5 35 10 30

0 30

point of view, see Figure 5.4.

ISF 2002

Sections of the Ternary System Fe-Cr-Ni

Figure 5.6

5. Welding of High-Alloy Steels, Corrosion


During an ongoing cooling, the binary area ferrite + austenite passes through and a transformation into austenite takes place. If the cooling is close to the equilibrium, a partial transls ls st ee ls ee iti cfe st ee st rri tic

formation of austenite into the brittle -phase takes place in the temperature range below 800 C. Primary ferritic solidifying alloys show a reduced tendency to hot cracking, because -ferrite can absorb hot-crack promoting elements like S and P. However primary austenitic solidifying alloys show, starting at a certain alloy content, no transformations during cooling (14% Ni, 16% Cr, left figure). Primary austenitic solidifying alloys are much more susceptible to hot cracking than primary ferritic solidifying alloys, a transformation into the -phase normally does not take place with these alloys. Figure 5.7 shows some typical compositions of certain groups of high alloy steels.

si tic

M ar te n


iti c

rri tic



C Si Mn Cr Mo Ni Cu Nb Ti Al V N S

0.1 max. 1.0 max. 1.0 15 18 up to 2.0 1.0



0.1 1.2 max. 1.0 max. 1.5 12 18 up to 1.2 2.5

0.1 max. 1.0 max. 2.0 17 26 up to 5.0 7 26 up to 2.2

0.1 max. 1.0 max. 2.0 24 28 up to 2.0 4 7.5

+ + + +

+ +




+ + +

st en



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

st ee

ISF 2002

Typical Alloy Content of High-Alloy Steels

Figure 5.7

The diagram of Strau 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
28 % 24 20 16 12 8 4 ferrite / perlite 0 0 2

5.1 is based on this diagram. If a steel only contains C, Cr and Ni, the

lowest austenite corner will be at 18% Cr and 6% Ni. And also other elements
austen ite / ma rt ensite
austenite / ferrite


than Ni and Cr work as an austenite or ferrite developer. these The influence is of denickel elements and

martensite / troostite / sorbite 4 6 8 10 12 14 Chromium 16


/ martens

ite / ferrite
24 % 26

martensite / ferrite 18 22
ISF 2002

Maurer - Diagram

scribed by the so-called chromium

Figure 5.8

5. Welding of High-Alloy Steels, Corrosion


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 grain growth (area 1).
Nickel-equivalent = %Ni + 30x%C + 0,5x%Mn
30 28 26 24 22 20 18 16 14 12 10 8 6 4 2 00

A cause for this strong grain growth during welding is the greatly increased diffusion speed in the ferrite compared with austenite. After reaching a diffusion-start Figure 5.10 temperature, have a


t rri Fe 5%


20 %

A+F A +M

80 %


martensite F + M
2 4 6 8 10

A+M+F M+F ferrite

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 considerably


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

sigma embrittlement between 500-900C grain growth above 1150C

ISF 2002

stronger grain growth than austenites. Therefore high alloyed ferritic steels are to be considered as of limited weldability. The area 3 marks a possible embrittlement of the material
4000 6000

Schaeffler Diagram With Border Lines of Weld Metal Properties to Bystram

Figure 5.9


grain size

due to the development of -phase. As explained in 5.6, this risk occurs with increased increased ferrite contents, conchromium



austenitic steel

ferritic steel

tents, and sufficiently slow cooling speed.




600 temperature




ISF 2002

Grain Size as a Function of Temperature

Figure 5.10

5. Welding of High-Alloy Steels, Corrosion


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 -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 A 20 10 20 24 A A+M 16 12 8 4 0 0 4 8 12 16 20 24 28 32 36 0 0 4 8 12 16 20 24 Chromium-equivalent 28 32 36 Chromium-equivalent

S235JR (St 37) Welding consumable

10 20 40




40 80

A+M 16 12 8 4 F M+F M 20% : =1:1 +


3 1 20%

80 A+F 100 % F

: =1:1 + F M+F


100 % F


X8Cr17 (W.-Nr. 1.4510)

21% Cr, 14% Ni, 3% Mo

S235JR (St 37) Welding consumable

X10CrNiTi18-9 (W.-No. 1.4541)

21% Cr, 14% Ni, 3% Mo

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


ISF 2002

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


ISF 2002

Application Example of Schaeffler - Diagram

Application Example of Schaeffler - Diagram

Figure 5.11

Figure 5.12

5. Welding of High-Alloy Steels, Corrosion


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

20 19 18 17 16 15 14 13 12 11 10 16 austenite

te rri fe 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


r be m 0 2 4 6 8 10 12 14 16 18

The most important feature of high alloy steels is their corrosion resistance starting with a Cr content of 12%. In addition to the problems during welding
austenite + ferrite

described by the Schaeffler diagram, these steels can

26 27



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

ISF 2002

be negatively affected with view to their corrosion resistance caused by the

De Long Diagram

Figure 5.13

welding process.


air O 2Fe+++O+H2O 2Fe++++2OHOHFe+++ water

5.14 shows schematically the processes of electrolytic corrosion under a

O2 OH H2O cathode 4eFe(OH)3

drop of water on a piece of iron. In such a system a potential difference is a precondition for the development of a local element consisting of an anode and a cathode. To develop


anode 2Fe 2Fe+++4eiron

O2+2H2O+4e 4OH


ISF 2002

Corrosion Under a Drop of Water

Figure 5.14

5. Welding of High-Alloy Steels, Corrosion


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

passive layer

active dissolution

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




ISF 2002


ISF 2002

Figure 5.15

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.


Pitting Corrosion of a Steel Storage Container

With pitting corrosion, the chemical composition of the attacking medium causes a

Figure 5.17

local break-up of the passive layer. Especially salts, preferably Clions, 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 tion.


repassivation corrosion mainly

causes a crack propagaStress takes cracking





passive layer;

metal surface;


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


Model of Crack Propagation Through Stress Corrosion Cracking






Figure 5.18

5. Welding of High-Alloy Steels, Corrosion


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 may release stress corrosion cracking.

Sensitivity to stress corrosion cracking

complete cover layer

tough fracture



e 2

e 1

Elongation speed e


ISF 2002


ISF 2002

Influence of Elongation Speed on Sensitivity to Stress Corrosion Cracking

Transgranular Stress Corrosion Cracking

Figure 5.19

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 chro-

to Bain and Aborn 1200 C 1100 1000 900 800 700 600 0 0.05 0.1 0.15 0.2 Carbon content 0.25 % 0,3 A

mium to carbon, which causes the precipitation of chromium carbides Cr23C6 on grain boundaries, Figure 5.22. Due to these precipitations, the austenite grid is depleted of chromium content along the grain boundaries and the Cr content drops below the parting limit. The diffusion speed of chromium in austenite is considerably lower than that of carbon, therefore the chromium reduction cannot 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.

Heat treatment temperature


ISF 2002

Carbon Solubility of Austenitic Cr - Ni Steels

Only after the steel has been subjected to sufficiently long heat treatment, chromium will diffuse to the grain boundary and increase the

Figure 5.21

C concentration along the grain boundary (line 3 in Figure 5.22). In this way, the complete corrosion resisChromium content of austenite
resistance limit 3 1 - homogenuous starting condition 2 - start of carbide formation 3 - start of concentration balance 4 - regeneration of resistance limit 1 2 4

tance can be restored (line 4 in Figure 5.22). Figure 5.23 explains why the IC is also described as intergranular disintegration. Due to dissolution of depleted areas along the grain boundary, complete grains break-out of the steel.


Distance from grain boundary

ISF 2002

Sensibility of a Cr - Steel

Figure 5.22

5. Welding of High-Alloy Steels, Corrosion 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 (see also Figure 5.21). Below this temperature, a carbon precipitation takes place. As it is a diffusion controlled certain process, the time
Reciprocal of heat treatment temperature 1/T
br-er-06-23e.cdr ISF 2002


Grain Disintegration

Figure 5.23

precipitation occurs after a incubation

unsaturated austenite austenite chromium carbide (M23C6) no intergranular disintegration

which depends on temperature (line 1, precipitation characteristic curve). During stoppage at a constant temperature, the parting limit of the steel is

austenite + chromium caride (M23C6) sensitive to intergranular disintegration

oversaturated austenite

Heat treatment time (lgt)

regained by diffusion of chromium.


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


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 so-called than 0,03% During welding, the considerable influence of
cooling curve

lower temperatures ELC-steels



quench temperature precipitation curves for 17% Cr steel 18-8-Cr-Ni steel

times. This fact initiated the development of where the C content is decreased to less
Tempering temperature


carbon is also important for the selection of the shielding gas, Figure 5.27. The higher the CO2-content of the shielding gas, the
br-er06-25e.cdr ISF 2002

Tempering time

stronger is its carburising effect. The Ccontent of the weld metal increases and the steel becomes more susceptible to intergranular corrosion.
Figure 5.25

Precipitation Curves of Various Alloyed Cr Steels

1000 C 900 800

An often used method to avoid intergranular corrosion is a stabilisation of the



steel by alloy elements like

0.05%C 0.03%C


niobium and titanium, Fig0.025%C


ure 5.28. The affinity of these elements to carbon is significantly higher than that of chromium, therefore


400 1 10


103 Time



ISF 2002

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

Influence of C-Content on Intergranular Disintegration

Figure 5.26

5. Welding of High-Alloy Steels, Corrosion


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.

Heat treatment temperature

600 550

M2 M1

0.018 % C 0.57 % Nb Nb/C = 32

Heat treatment temperature

0.058 % C 0.53 % Nb Nb/C = 9

0.030 % C 0.51 % Nb Nb/C = 17

800 C 700 650 600 550 500 450 0,3 1 3 10 30 100 Time W.-No.:4301 (0,06%)

1050C /W

500 450 400 0,2 0,5 1 2,5 5 10 25 50 100 250 h






Heat treatment time

Heat treatment temperature
800 C 700 650 600 550 500 450 0,3


C o m p o sitio n S hie ld ing g a s S 1 M 1 M 2 A r [% ] 99 90 82 C O2 / 5 18 O2 1 5 /

1300C /W 1050C /W

1 3 W.-No.:4541



100 Time







ISF 2002


ISF 2002

Influence of Shielding Gas on Intergranular Disintegration

Influence of Stabilization on Intergranular Disintegration

Figure 5.27

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


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 hereof. Figure 5.30 shows one of the possible grouping of nickel-base-alloys.
Figure 5.29

Knife-Line Corrosion

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 Group A Nickel 200 Nickel 212 Nickel 222 Group B Monel 400 Monel 450 Ferry Group C Inconel 600 Nimonic 75 Nimonic 86 Incoloy 800 Incoloy 825 Inco 330 Ni 66.5, Cu 31.5 Ni 30.0, Cu 68.0, Fe 0.7, Mn 0.7 Ni 45.0, Cu 55.0 Ni 76.0, Cr 15.5, Fe 8.0 Ni 80.0, Cr 19.5 Ni 64.0, Cr 25.0, Mo 10.0, Ce 0.03 Ni 32.5, Fe 46.0, Cr 21.0, C 0.05 Ni 42.0, Fe 30.0, Cr 21.5, Mo 3.0, Cu 2.2, Ti 1.0 Ni 35.5, Fe 44.0, Cr 18.5, Si 1.1 Ni 99.6, C 0.08 Ni 97.0, C 0.05, Mn 2.0 Ni 99.5, Mg 0.075 Chem. composition Alloy Group D1 Duranickel 301 Ni 94.0, Al 4.4, W 0.6 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 Group D2 Monel K-500 Inconel 718 Ni 65.5, Cu 29.5, Al 2.7, Fe 1.0, W 0.6 Ni 52.0, Cr 22.0, Mo 9.0, Co 12.5, Fe 1.5, Al 1.2 Chem. Composition

characterized by moderate 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 condition, and cold-drawn material is recommended

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 Nimonic 105 Incoloy 903 Incoloy 909 Inco G-3 Inco C-276 Group E Monel R-405 Ni 66.5, Cu 31.5, Fe 1.2, Mn 1.1, S 0.04 Ni 76.0, Cr 19.5, Fe 112.4, Al 1.4 Ni 39.0, Fe 34.0, Cr 18.0, Mo 5.2, W 2.3, Al 0.8 Ni 58.0, Cr 19.5, Co 13.5, Mo 4.25, W 3.0, Al 1.4 Ni 38.4, Fe 42.0, Cu 13.0, Nb 4.7, W 1.5, Al 0.03, Si 0.15 Ni 38.4, Fe 42.0, Cu 13.0, Nb 4.7, W 1.5, Al 0.03, Si 0.4


ISF 2002

Typical Classification of Ni-Base Alloys

for best machinability and smoothest finish.

Figure 5.30

5. Welding of High-Alloy Steels, Corrosion


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 (') 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


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 ' 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 ' phase develops inside the grains and M23C6 precipitates onto the grain boundaries. Adjacent to the grain boundary, there is a ' depleted zone. During precipitation hardening (700 C/20 h) ' phase develops in these depleted zones. ' 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 ' 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


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 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 als, steel, as special cast materidue to their poor weldability, are of no importance in welding.
Figure 6.1


Designation according to the material code (DIN EN 1560)

Figure 6.2 shows the designation of the cast material in accordance with DIN EN 1560. A distinction is made between the designation according to the material code and the designation according to the material number. In Figure 6.2, examples of two materials are specified.

e.g.: EN-GJ L F 150 1

Position 1: Position 2: Position 3: Position 4: Position 5: Position 6: EN GJ L F 150 -

2 34

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

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 -


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


ISF 2004

Designation of Materials

Figure 6.2

6. Welding of Cast Materials


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 Mechanical Properties 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

Chemical Analysis A % 15 12 25 C Si Mn P S

this way. Cast iron with lamellar and spheroidal graphite has carbon of 2,8 beand contents tween

250 200 190

300 400 380 380

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

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

- lamellar graphite cast iron - nodular graphite cast iron - decarburizing annealed malleable cast iron (former : white -heart malleable cast iron) - cast steel
ISF 2004

4,5%. Through the addition of alloying elements, above all Si, these mate-

Characteristics and Analyses of Cast Materials

Figure 6.3

rials solidify following the stable system, i.e., the carbon is precipitated in cast the from metal, the form of graphite. Malleable iron shows similar C-contents, solidification the molten however,

follows the metastable system. The

Figure 6.4

C-contents of cast steel, on the other

6. Welding of Cast Materials 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 obtained, the residual stresses inside the workpiece are, on the other hand, reduced.
Figure 6.5


ISF 2002

Microstructure of Normally Glowed Cast Steel

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

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

Microstructure of Lamellar Graphite Cast Iron

Figure 6.6

6. Welding of Cast Materials


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

br-er-07-06e.cdr ISF 2002

Nodular Graphite Cast Iron




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 timeconsuming heat treatment.


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 ISF 2002

tite and precipitated. During a subsequent annealing process, the iron carbide disintegrates into graphite and iron.
Figure 6.9

Decarburizing Annealed Malleable Cast Iron EN-GJMW-350

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 atmosphere 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.

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

white-heart malleable cast iron


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

the region of the HAZ occurs. For carrying out constructional welds made of
Hardness after Brinell


malleable cast iron parts, a special material quality has been developed. Figure 6.11 shows that this material, EN-GJMW-400-12, is characterised by considerably less hardening. This material is weldable without any problems up to a wall thickness of 8 mm.




material thickness: 7 mm Testspeciem 50

20 mm 10 Distance of center welding seam



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 Atomic weight Specific weight Lattice E-module R pO,2 PO,2 R mm spec. Heat capacity Melting point Heat conductivity Spec. el. Resistance Expansion coeff. [N/mm] [N/mm] [N/mm] [J/(g*C)] [C] [W/(cm*K)] [nW m] [1/C] [g/Mol] [g/cm]

Al 26.9 2.7 fcc 71*10 ca. 10 ca. 50 0.88 660 2.3 28-29 24*10

Fe 55.84 7.87 bcc 210*10 ca. 100 ca. 200 0.53 1539 0.75 97 12*10 FeO

of steel and aluminium. Side by side with different mechanical behaviour, the following differences are important for aluminium welding: - considerably lower melting point compared with steel - three times higher heat conductivity - considerably lower electrical resistance - double expansion coefficient - melting point of Al203 considerably higher than that of Al; metal and iron oxide melt approximately at the same temperature. Figure 7.2 compares some mechanical properties of steel with properties of some light metals. The important advantages of light metals compared with steel are especially


Al2O 3

Fe 3O 4 Fe 2O 3 1400

Melting point of oxydes



1600 (1455)


ISF 2002

Basic Properties of Al and Fe

Figure 7.1

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 not
Figure 7.2

Deflexions and Weights of Cantilever Beams Under Load

subject to a lattice trans-

7. Welding of Aluminium Alloys


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 600



1000 1200

800 -2



-4 8 cm aluminium 6 4

100C 200



2 300 400 500 600 -2 -4 -6

br-er08-03.cdr ISF 2002

-8 -18











ISF 2002

Comparison of Stress-Elongation Diagrams of Al and Steel

Isothermal Curves of Steel and Al

Figure 7.3

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 Figure 7.6 shows typical applications of some Al alloys together with preferably used welding consumables.
Al Mg Si



Al Cu Mg

Aluminium alloys are often welded with consumable of the same type, however, quite often over-alloyed consumables are used to compensate burn-off losses (especially with Mg and Zn because of their low boiling point) and to improve the mechanical properties of
Al Si Al Zn Mg

Al Zn Mg Cu



Al Si Cu

the seam. The classification of Al alloys into two groups is based on the characteristic that the group of the non-hardenable alloys cannot increase the strength through heat treatment, in contrast to hardenable alloys which have such a potential. The important hardening mechanism for this
Figure 7.5

Al Mg


Al Mg Mn

Al Mn

ISF 2002

Classification of Aluminium Alloys

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 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 SG-AlMg4,5Mn SG-AlMg5; SG-AlMg4,5Mn; SG-AlSi5 SG-AlSi5 SG-AlMg3; SG-AlMg4,5Mn SG-AlMg5; SG-AlMg3; SG-AlMg4,5Mn SG-AlMn1;SG-Al99,5T

temperature in this condition, no precipitation will take place. The alloy elements are forced to remain dissolved, the crystal is out of equilibrium. If such a structure is subjected to an age hardening at room or elevated temperature, a

AlMg2Mn0,8 AlMn1

base material - aluminium percentage of alloy elements without factor

br-er-08-06.cdr ISF 2002

Use and Welding Consumables of Aluminium Alloys

precipitation of a second phase takes place in ac-

Figure 7.6

non-hardenable alloys

hardenable alloys


7. Welding of Aluminium Alloys


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

stable condition

solution heat treatment

solidification of alloy elements in solid solution

repeated hardening

regeneration cold ageing (RT ageing)

oversaturated solid solution, metastable condition

precipitation takes place in three possible forms: copartly coherent precipitations, warm aged condition

ageing at slightly increased temperature

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

herent particles (i.e. particles deviating from the matrix in their chemical composition but having the same partly lattice coherent structure), particles

longer warm ageing

partly coherent and incoherent precipitations, softening

longer warm ageing

stable incoherent equilibrium phase stable condition
br-er-08-07.cdr ISF 2002

Ageing Mechanism

(i.e. the lattice structure of the matrix is partly retained), and incoherent

Figure 7.7

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 dislocation movement in the matrix lattice, thus

700 liquid liquid and solid Q copper containing aluminium solid solution 500 P

leading to an increase in strength. The finer the precipitations, the stronger the effect. At an increased temperature (heat ageing, Figure 7.7) a maximum of second phase has precipitated after elapse of a certain time. Consequently a prolonged stop at this temperature does not lead to an increased strength, but to coarsening of particles due to diffusion processes and to a decrease in strength (less bigger particles in an extended space).
Figure 7.8



aluminium solid solution and copper aluminide (Al2Cu)


100 copper content of AlCuMg 0 1 2 3 Copper 4 5 mass-% 7

ISF 2002

Phase Diagram Al-Cu

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 500 P C 400

solution heat treatment

tical with the starting condition. A deterioration of mechanical properties only happens during hot ageing, if the ageing time is excessively long.





heat ageing

The complete process of hardening at room temperature is metallographic also called age hardening, at elevated temperature heat age-

age hardening
0 2 4 6 8 10 12 h 14


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

ISF 2002


Temperature - Time Distribution During Ageing

Figure 7.9 shows a schematic representation of time-temperature curves during hardening with age hardening and heat ageing.

Figure 7.9





380 water quenching (~900C/min) air cooling (~30C/min) 320

0.2% yield stress s0.2 in N/mm

strength increase of AlZnMg 1 in dependence of time. The difference between age hardening and heat ageing is here very clear. Due to improved diffusion conditions is the strength increase in the case of heat ageing much faster than in the case of age hardening. The
Figure 7.10

260 120C 200 RT 140

80 10-1 quenched 100 Ageing time in h

ISF 2002




Increase of Yield Stress During Ageing of AlZnMg1

strength maximum is also reached considerably earlier. 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


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.
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 incubation time. The lower the ageing temperastress and tensile strength. If a low ageing temperature is selected, the ageing time as well as the incubation time become extremely long. 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 increase of the ageing temperature shortens the duration of the complete precipitation process 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.
400 N/mm 300
Tensile strength Rm

N/mm 400

110 135 150 180

300 230

190 205 260C

200 N/mm 400

0.2% yield stress s0.2


ture, the higher the resulting values of yield


135 150

200 230 260

180 190 205C

Fracture elongation d2

30 % 20 10 260 205




135 110C

30 min 1 day
0 1

1 week

1 1 month year





10 10 Ageing time


103 h 104


ISF 2002

Influence of Ageing Temperature and -Time on Ageing

Figure 7.11

As the lower part of the figure shows, the fracture elongation

AlMg5 AlMg3 200



proportional to the strength values, i.e. the strength increase caused by ageing is accompanied by an embrittlement of the material.

100 Al99,5

0 0



ISF 2002

Age Hardening of Al Alloys

Figure 7.12

7. Welding of Aluminium Alloys


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 N/mm 250

one hand, tensile strength increases with increasing content of alloy elements (solid solution strengthening), on the other hand, this increase is caused by a stronger deformation of the lattice. Figure 7.13 shows the effect of the welding
0,7 0,6


Rm or Rp0,2


process on mechanical properties of a coldworked alloy. Due to the heat input during welding, the blocked dislocations are released (recovery), in addition, a grain coarsening will start in the HAZ. This is followed by a strong drop in yield point and tensile strength. This strength loss cannot be overcome in the case of a welding process.


0,5 50

0,4 0,3 0,2 80 60 40 20 0 20 40 Distance from Seam Centre 60 mm 100


ISF 2002

Non-Hardenable Al Alloy

Figure 7.13





N/mm 350 21 days RT 300

90 days RT Rm

mechanisms in the case of a hardenable aluminium alloy.


1 day RT

As a consequence of the welding heat, the precipitations are solution heat treated and the strength values decrease in the weld area. Due to the age hardening, a restrengthening of the alloys takes place with increasing time.

250 90 days RT 21 days RT 200



1 day RT


4 mm plates of: AlZnMg1F32 start values: Rp0,2=263N/mm Rm=363 N/mm welding method: WIG, both sides, simultaneously welding consumable: S-AlMg5 specimens with machined weld bead 20 20 60 0 40 Distance from seam centre 80 100 mm 140
ISF 2002

50 80



Hardenable Al Alloy

Figure 7.14

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

Hot Cracks in a Al Weld




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 that the manganese content
Weld cracking tendency

100 % 80
Cracking susceptibility

influences significantly the hot crack susceptibility. The maximum of this hot crack susceptibility is likely with about 1% Mg content (corresponds with alloy 1). With increasing MG content, hot crack susceptibility decreases strongly (see also alloy 2 and 3, left part).

2 60 1 40









500 0 1 2 Alloy content 3 % 4

Preheat temperature

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


ISF 2002

Influence of Preheat Temperature and Magnesium Content

Figure 7.16

To avoid hot cracking, partly very different preheat temperatures are recommended for the alloys. Zschtge 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

92 7.17, together with the re-

TVorw. = TS TS Tvorw. lAl-Leg.


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

in C in C in J/cm*s*K

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

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



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

melting point pure aluminium

Recommended preheat temperature

400 300 200 100

0 Increasing better weldability


ISF 2002

mild steel (0.2%C) without preheating


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

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


Recommendations for Preheating

Figure 7.17

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
br-er-08-18.cdr ISF 2002

Excessive Porosity in a Al Weld

Figure 7.18

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 humid air (nitrogen, oxygen, water) nozzle deposits and too steep inclination of the torch cause turbulences

Solubility of hydrogen in aluminium changes abruptly on the phase transition melt-crystal, i.e. the melt dissolves many times more of the hydrogen than the just forming crystal at the same temperature.

poor current transition

humid air


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

pores feuchte Luft

Poren solid weld metal festes Schweigut base material


H2 H2


ISF 2002

Ingress of Hydrogen Into the Weld

Figure 7.19

7. Welding of Aluminium Alloys


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 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 2 0 -2 -4 -6 -14

300C 400C 500C

600C 700C 800C 900C

6 cm 4 2 0 -2 -4 -6

500C 400C 300C

600C 700C

itself, but also the surrounding base material (HAZ) is influenced by the supplied heat quantity. The temperature-field, which appears around the weld when different welding procedures are used, is shown in Figure 8.1. Figure 8.2 shows the influence of the material












0 cm 2


C 1750

1250 1000 723C 750 500 250 -60 mm -40 -20 0 20 40 mm 60

manual metal arc welding oxy-acethylene welding

properties on the welding process. The determining factors on the process presented in this Figure, like melting temperature and interval, heat capacity, heat extension etc, depend greatly on the chemical composition of the material. Metallurgical properties are here characterized by e.g. homogeneity, structure and texture, physical properties like heat extension, shear strength, ductility. Structural changes, caused by the heat input

distance from weld central line

heat affected zone during oxy-acethylene welding


heat affected zone during manual metal arc welding

ISF 2002

Temperature Distribution of Various Welding Methods

Figure 8.1

(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. Based on the binary system, the formation of the different structure zones is shown in
4 Heating and melting the welding consumable Melting parts of base material Specific heat, melting temperature and interval, melt heat, boiling temperature (metal, coating) Specific heat, melt temperature and interval, heat conductivity, heat expansion coefficient, homogeneity, time Compositionof atmosphere, affinity, pressure, temperature, dissotiation, ionisation, reaction speed Solubility relations, temperature and pressure under influence of heat source, specific weight, weld pool flux Diffusion and position change processes, time, boundary formation, ordered - unordered structure Affinity, temperature, pressure, time Melt heat, cooling conditions, density and porosity of slag, solidification interval Phase diagrams (time dependent), heat conductivity, heat coefficient, shear strength, ductility Phase diagrams (time dependent), texture by warm deformation, ductility, module of elasticity Phase diagrams, operating temperature, mechanical and chemical strain, time

Reaction of passing welding consumable with arc atmosphere

Figure 8.3. So the coarse


Reaction of passed welding consumable with molten base material Interaction between weld pool and solid base material (possibly weld passes)
Reaction of metal and flux with atmosphere Solidification of weld pool and slag Cooling of welded joint in solid condition Post-weld heat treatment if necessary Sustainable alteration of material properties

grain zone occurs in areas

6 7




austenitising temperature for example. At the same time, hardness peaks appear in these greatly areas because of

7 8

2 5 9 10



ISF 2002


Figure 8.2

Classification of Welding Process Into Individual Mechanisms

cooling rate and the coarse

8. Technical Heat Treatment


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 formation of the individual structure zones and the zones of unfavourable mechanical properties can be influenced. Applying an electroslag one pass weld of a 200
Temperature Hardness

hardness peak

hardness sink weld bead

1 1500

incomplete melt coarse grain

C 1300 1200 1000 G 800 P 600 400 300 100


2 1147 3

mm thick plate, a HAZ of approximately 30 mm width is achieved. Using a three pass technique, the HAZ is reduced to only 8 mm. With the use of different procedures, the differ-

standard transformation incomplete crystallisation recrystallisation

ageing blue brittleness

4 5 S


ISF 2002

ences in the formation of heat affected zones become even clearer as shown in Figure 8.5. These effects can actively be used to the advantage of the material, for example to adjust calculated mechanical properties to one's choice or to remove negative effects of a weld-

1 2 % 3 carbon content

heat affected zone (visible in macro section)


Microstructure Zones of a Weld Relation to Binary System

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 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 1536 C

A melt + d - solid solution melt

1600 C 1400 heat colors

melt + austenite


electron beam welding

d - solid solution A4 1392 cbc atomic lattice

H B d - solid solution + austenite N


diffusion heat treatment 2,06

1300 yellow white 1200 1147 1100 1000

yellow red light yellow yellow

1200 1100

submerged arc welding pass / capped pass

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 769C O S K A1 P 723C soft annealing ferrite700 (a-solid solution) recrystallisation heat treatment


coarse grain heat treatment

light red



700 600 500 400

dark red brown red dark brown

cbc 500 atomic lattice recrystallisation heat treatment Q

stress relieving






MS hypereutectoidic steel

gas metal arc welding

200 100 20 0 Fe 0

200 100 2 30
ISF 2002

hypoeutectoidic steel
eutektoidic steel

0,5 5

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


10 15 20 25 Cementite content in weight %


ISF 2002


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

Metallurgical Survey of Heat Treatment Methods

Figure 8.5

Figure 8.6

As this diagram does not cover the time influence, only constant stop-temperatures can be read, predictions about heating-up and

C 900


intense heating

long time several hours

austenite + ferrite 700 ferrite + perlite 500



cooling-down rates are not possible. Thus the individual heat treatment methods will be explained by their temperature-time-behaviour the following. in

300 0,4 0,8 C-Content


ISF 2002

Coarse Grain Heat Treatment

Figure 8.7

8. Technical Heat Treatment


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 schematically time-temperature behaviour in a TTT-diagram.

900 C 700 600 500 400 300 MS

martensite bainite




(Note: the curves explain running structure mechanisms, they must not be used as reading off examples. To determine distribution, t8/5, are hardness values, or microstructure TTT-diagrams always read continuously or isothermally. Mixed types like curves 3 to


e lin

200 100 2 0 0,1


3 1 10

4 10

6 s

1 10

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


ISF 2002

TTT-Diagram With Heat Treatment Processes

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 900


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


30 C above AC3. Also in this case one must watch that the austenite grains remain as small as possible. To ensure a complete

austenite + ferrite


700 ferrite + perlite 500

quick heating

air cooling

300 0,4

transformation to marten0,8 C-Content % Time

ISF 2002

site, a subsequent quenching 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 stopping a water quenching
C 900 austenite + ferrite

austenite about 30C above A3 A3


process at a certain temperature and continuing the cooling with a milder cooling medium (oil). With longer holding on at elevated temperature level, transformations can also be carried through in the bainite area (curves 5 and 6).


700 ferrite + perlite 500 start of martensite formation 300 0,4


quenching in water

start of martensite formation Time

ISF 2002

0,8 C-Content


Figure 8.10

8. Technical Heat Treatment


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
C 900 austenite + ferrite


hardening and tempering about 30C above A3 A3


e.g. strength and toughness.


700 ferrite + perlite 500

quenching slow cooling

Figure 8.12 shows the pro300 0,4


cedure of soft-annealing.
0,8 C-Content % Time
ISF 2002

Here we aim to adjust a soft and suitable microstructure for machining. Such a structure is characterised by mostly globular

Hardening and Tempering

Figure 8.11

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 transformation is carried out because

C 900

austenite time dependent on workpiece 10 to 20C below A1


the surface area is strongly reduced tion). thermodynamisteels cally more favourable condiHypereutectic have in addition to the lamellar structure of the perlite a cementite network on the

austenite + ferrite 700 ferrite + perlite 500

A3 A1

oscillation annealing + / - 20 degrees around A1


300 0,4 0,8 C-Content % Time cementite


ISF 2002

grain boundaries.

Soft Annealing

Figure 8.12

8. Technical Heat Treatment


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 900 austenite + ferrite


place. The repetition of this


process leads to a steptime dependent on workpiece


700 ferrite + perlite 500

wise spheroidizing of cebetween 450 and 650 C

mentite and the frequent transformation annealed avoids a grain coarsening. A soft-

300 0,4

0,8 C-Content

ISF 2002


represents frequently the delivery condition of a material.

Stress Relieving

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 make such deformations
Stress releaving Heat treatment at a temperature below the lower transition point A1 , mostly between 600 and 650C, with subsequent slow cooling for relief of internal stresses; there is no substantial change of present properties. 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. 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. Heat treatment to achieve a high ductility with defined tensile stress by hardening and subsequent tempering (mostly at a higher temperature. Fast cooling of a workpiece. Also fast cooling of austenitic steels from high temperature (mostly above 1000C) to develop an almost homogenuous micro-structure with high ductility is called 'quenching heat treatment'. 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.

possible because the yield strength limit decreases with increasing temperature. A stress-relieve heat treatment should not cause any other change to properties, so that tempering steels are heat treated below tempering


Hardening (quench hardening) Quenching and tempering Solution or quenching heat treatment



ISF 2002

Figure 8.14

Type and Purpose of Heat Treatment

8. Technical Heat Treatment


Figure 8.14 shows a survey of heat treatments which are important to welding as well as their purposes.
Types of heat treatments related to welding

Figure 8.15 shows principally the heat treatments in

combination heat treatment after welding (post-weld heat treatment)

heat treatment before welding


accompanying heat treatment

connection with welding. Heat treatment processes are divided into: before,
solution tempering heat treatment

simple step-hardening welding

pure step hardening welding

modified step hardening welding


stress releaving

stress releaving

annealing hardening quenching and tempering postheating (post weld heat treatment)

during, and after welding. Normally a stress-relieving or normalizing heat treatment is applied before welding to adjust a proper



simple preheating

increase of working temperature

constant working temperature isothermal welding

local preheating

preheating of the complete workpiece

heat treatment of the complete workpiece

local heat treatment


ISF 2002

material condition which for

Heat Treatment in Connection With Welding

welding. After welding, alFigure 8.15

most any possible heat 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

treatment which accom-panies the welding. The most important processes are explained in the following.
Temperature T

C 700



Figure 8.16 represents the influence of different accompanying heat treatments during welding, given within a TTT-diagram. The fastest cooling is achieved with welding without preheating, with addition of a small share of bainite, mainly martensite is formed (curve 1, analogous to Figure 8.8, hardening). A simple heating before welding without additional stopping time lowers the cooling rate according to curve 2. The proportion of martensite is reduced in the forming structure, as well as the

400 MS 300 (1)






0 0 1 10 tH (1): Welding without preheating, (2): Welding with preheating up to 380C, without stoppage time (3): Welding with preheating up to 380C and about 10 min. stoppage time TA: Stoppage temperature, tH: Dwell time 102 Time t 103 104 s 105


ISF 2002

TTT-Diagram for Different Welding Conditions

Figure 8.16

8. Technical Heat Treatment


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



used in the following, Figure 8.17 shows a superposition of all individual influences on the materials as well as the resulting T-T-course in
Temperature T


A3 A1 transformation range

the HAZ. As an example, welding with simple preheating is selected. The plate is preheated in a period tV. After removal of the heat source, the cooling of the workpiece starts. When tS is reached, welding 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 temperature-time-behaviour of the HAZ.


Time t tV
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


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.
ISF 2002


The temperature time course during welding with simple preheating is shown in Figure 8.18.
Figure 8.17

Temperature-Time-Distribution During Welding With Preheating

During a welding time tS a drop of the working temA3

Temperature T

perature TA occurs. A further air cooling is usually carried out, however, the cooling rate can also be



Time t tV tS tA

reduced by covering with

Temperature of workpiece, Temperature of weld point

heat insulating materials. Another variant of welding

ISF 2002

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

Welding With Simple Preheating

with preheating is welding at constant working This is temperature.

Figure 8.18

8. Technical Heat Treatment

104 achieved through further


warming during welding to avoid a drop of the working temperature. In Figure 8.19 is this case (dashed line, TA needs not to be above MS) as well as the special case of isothermal welding illustrated. During isothermal welding, the workpiece
ISF 2002

Temperature T



Time t tS tV tH tA

tH = 0

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

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


Welding With Preheating and Stoppage at Working Temperature

is heated up to a working temperature above MS (start of martensite formation) and is also held there

Figure 8.19

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

Temperature T

welding with post-warming (subsequent heat treatment, see Figure 8.15). Such a treatment can be carried out very easy, a gas welding torch is normally used for a local preheating. In this way, the toughness properties of some steels can be greatly improved. The lower sketch shows a combination of pre- and postheat treatment. Such a treatment is applied to steels which have such a strong tendency to hardening that a cracking in spite of a simple preheating before welding cannot be avoided, if they cool down directly from working temperature. Such materials are heat treated immediately after welding at a temperature between 600 and 700 C, so that a formation



Time t tS tN tA

2. Pre- and post-heating

Temperature T




Time t tV tS tR tN tA

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

tS: tA: tN: tR:

Welding time, Cooling time (room temperature), Postheat time Stoppage time
ISF 2002

Welding With Pre- and Post-Heating

Figure 8.20

8. Technical Heat Treatment of martensite is avoided and welding residual stresses are eliminated simultaneously.


Aims of the modified stepTHa A3

Temperature T




not be discussed here, Figure 8.21. Such treatments are used for transformationinert materials. The aim of the figure is to show how complicated a heat treatment

TSt TAnl


MS TAnl Time t tS tH
TA: Working temperature, TAnl: Tempering temperature, TH: Hardening temperature,


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

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


Temperature of workpiece, Temperature of weld point

can become for a material in combination with welding. Figure 8.22 shows tempera-

ISF 2002

Modified Step Weld Hardening

Figure 8.21

ture distribution during multipass 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 welded without cooling down to a certain temperature. As a result, working temperature increases with the number of passes. The
Temperature T

weld pass heat affected zone 4 3 weld pass 2 1 observed point


second pass is welded under a preheat temperature which is already above martensite start temperature. The heat which remains in the workpiece preheats the upper layers of the weld, the root pass is post-heat treated through 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 multi-pass welding is the warming of the HAZ of each previous pass above recrystallisation temperature with the corresponding crystallisa-



Time t tS tV tA

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 ISF 2004

Temperature-Time Distribution During Multi-Pass Welding

Figure 8.22

8. Technical Heat Treatment


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


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.
in the unaffected base metal

undercut, continuous

unfused longitudinal seam edge

longitudinal crack
in the HAZ in the unaffected base metal

in weld metal in fusion zone


open end crater

end crater with reduction of weld cross section

transverse crack

in weld metal in the HAZ in the unaffected base metal in weld metal in the HAZ globular gas inclusion


star shaped crack

weld reinforcement pore too small throat thickness


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
ISF 2002 br-er09-02.cdr ISF 2002


Defect Class: Shape Defects

Defect Class: Cracks and Cavities

Figure 9.1

Figure 9.2

lack of fusion between passes

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
br-er-09-03.cdr ISF 2002

Defect Class: Lack of Fusion, Insufficient Through-Weld

Figure 9.3

9. Welding Defects


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

slag line
different shapes and directions

as unfused area between weld metal and base material or previously welded layer. This happens when
irregular slag inclusions

single slag inclusions

the base metal or the previous layer are not completely or insufficiently

pore nest
locally enriched

molten. Figure 9.6 explains

ISF 2002


the influence of welding parameters on the development of lack of fusion. In the upper part, arc characteristic lines of MAG welding are shown using CO2

Defect Class: Solid Inclusions

Figure 9.4

welding joint defects

welding defects due to manufacture

welding defects due to material

and mixed gas. The welding voltage depends on welding current and is selected according to the joint type. With present tension, the welding current is fixed by the wire
ISF 2002

external weld defects

internal weld defects

hot cracks

cold cracks

cavities with weld metal

spatters and start points undercuts seam shape defects

lacks of fusion slag inclusions mechanical pore formation

solidification cracks remelt cracks

hydrogen cracks hardening cracks lamellar cracks precipitation cracks

metallurgical pore formation crater formation






Welding Defects

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

Figure 9.5

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


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 direction
Welding voltage


mixed gas

Welding current

positive torch angle torch axes


negative torch angle torch axes

Welding current

correct correct Wire feed Melting rate lack of fusion due to too low performance

false false

approx. 45 1...2
ing ra ng e

Welding speed





lacks of fusion due to preflow



Melting rate

ISF 2002


ISF 2002

Influence of Welding Parameters on Formation of Lack of Fusion

Influence of Torch Position on Formation of Lack of Fusion

Figure 9.6

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


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 1 2 9 Uhr 3
9 Uhr 12 1 2 3


an example, the upper fi1 2

gure shows the desired torch orientation for usual welding speeds. This orientation depends on parameters like and shape, workpiece thickness, melting

9 Uhr


ISF 2002

Influence of Torch Position on Formation of Lacks of Fusion

diameter groove

Figure 9.8

rate, and welding speed. 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


gas/gas developing material air -nitrogen -hydrogen



too low shielding gas flow through: too low setting leaking lines too small capillary bore hole too low supply pressure for 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) turbulences through: to high shielding gas flow spatters on gas nozzle or contact tube irregular arc 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 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 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

thermal current - possibly increased by chimmney effects with one-sided welding too high weld pool temperature too high work piece temperature injection effects water carbonmonoxide leaking torch (with water-cooled types) remelting of seggregation zones remelting of rust or scale


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 development 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 a) low crystallisation speed

cannot raise to the surface

br-er-09-12.cdr ISF 2002

of the weld pool, the pore will be caught in the weld metal. Figure 9.10 shows a X-ray
Figure 9.12

Growth and Brake Away of Gas Cavities at the Phase Border

photograph of a pore which developed in this way, as well as a surface and a transverse sec-

9. Welding Defects


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 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
Figure 9.13

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


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 Figure

0010 TS 0011 0012 0021 800 0020 400 MS 0022 0023 0024 0025 103 Time 0027

115 9.16 allocates during the

1600 C 1200

cracks according to their appearance welding heat cycle. Principally there is a distinction
0026 0028 104 105 106 s 107

0 1 10

between the group 0010 (hot cracks) and 0020 (cold cracks).


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

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
ISF 2002

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 simple case of a binary system, under the provision that a complete concentration balance takes place in the melt ahead of the solidification front, but no diffusion takes place in the crystalline solid. When a melt of a composition
tension tension


C5 TmB A C0 C5 B CB

C0 cools down, a crystalline solid is formed when the liquidus line is reached. Its concentration can be taken from the solidus line. In the course of the ongoing solidification, the rest

a tension tension


of molten metal is enriched with alloy elements


in accordance with the liquidus line. As defined in the beginning, no diffusion of alloy elements in the already solidified crystal takes place, thus the crystals are enriched with alloy elements much slower than in a case of the binary system (lower line).
Figure 9.17

segregation in base metal
aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa aaaaaaaaaa

ISF 2002

Development of Remelting and Solidification Cracks

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


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 segreb b




above, the melt between the crystalline solids at the


end of solidification has a considerably solidus

a: non-preferred bead shape b <1 t

decreased As


b: preferred bead shape b >1 t

c: non-preferred bead shape

indicated by the black areas, rests of liquid may be trapped by dendrites. If tensile stresses exist (shrinking stress of the

ISF 2002

Crystallisation of Various Bead Geometries

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 welding, 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
Figure 9.19

9. Welding Defects


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 occurrence stresses, of hot shrinking

cracks may develop. In the

Figure 9.20






shown in the middle part of


ISF 2002

Macrosection of a SA-Weld

Figure 9.21

Figure 9.22

9. Welding Defects


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 that an existing crack develops through the following layers during over-welding. Figure 9.22 classifies cold cracks depending on their position in the weld metal area. Such a classification does not provide an explanation for the origin of the cracks.

chemical composition (C-equivalent)

welding consumables humidity on welding edges

residual stresses (yield stress of steels and joints)

cooling rate (t8/5)

cooling rate (t8/1)

additional stresses (production conditions)

5 mm section plane
br-er09-23.cdr ISF 2002

Causes of Cold Crack Formation

Figure 9.23

0,2 mm

5 mm

Figure 9.23 shows a summary of the three main causes of cold crack formation and their main influences. As explained in previous chapters, the resulting welding microstructure depends on both, the composition of base
crack in heat affected zone etching: HNO3 5 mm

and filler materials and of the cooling speed of the joint. An unsatisfactory structure composition promotes very much the formation of cold cracks (hardening by martensite).
Figure 9.24

transverse cracks in weld metal

ISF 2002

Cold Cracks in the Heat Affected Zone and Weld Metal

9. Welding Defects
1,2 % 1,0
Water content of coating

119 Another cause for increased cold crack sus-

18 C 1,12 1,0 0,83

90 % RH basic stick electrode Mn - Ni - Mo - Typ


ceptibility is a higher hydrogen content. The hydrogen content is very much influenced by the condition of the welding filler material (humidity of electrodes or flux, lubricating grease on welding wire etc.) and by humidity


0,6 0,46 0,35 0,27 0,17 0,1 0 0 1 2 3 4 5 Storage time 6 7 days 0,16 0,4 0,21 50 % 0,22 35 % 0,18

0,4 0,2

70 % 0,43

on the groove edges. The cooling speed is also important because it determines the remaining time for hydrogen effusion out of the bead, respectively how much hydrogen remains in the weld. A measC a hydroure is t8/1 because only below 100 gen effusion stops.

4,0 %
Water content of coating

20 C / 70 % RH





1 Storage time



ISF 2002

A crack initiation is effected by stresses. Depending on material condition and the two already mentioned influencing factors, even residual stresses in the workpiece may actu-

Water Pick-up of Electrode Coatings

Figure 9.25

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 rectly cracking. welding Such but
Water content of coating

cracks do not appear diafter hours or even days after cooling. The weld metal hydrogen content depends on humidity of the electrode coating (manual metal arc welding) and of flux (submerged arc welding).

1.0 % 0.9 0.8 0.7 0.6 0.5 0.4 0.39 0.3 0.28 0.2 0.1 0 30 40 50 60 Relative humidity 70 % 80
ISF 2002

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

AWS A5.5 stored and rebaked

Water Content of Coating After Storage and Rebaking

Figure 9.26

9. Welding Defects


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 20 50
Diffusible hydrogen content in weld metal

preheat temperature in C 80 100

In practice, such constant storage conditions are not to be found, this is the rea-

40 cellulose coated stick electrode 30

son why electrodes are backed before welding to limit the water content of


basic coated stick electrode 10

the coating. Figure 9.26 shows the effects of this

400 s 500
ISF 2002



200 300 Cooling time between 800 and 1000C

measure. The upper curve shows the water content of the coating of electrodes which were stored at constant air humidity before

Influence of Preheat Temperature on Cooling Speed and Hydrogen Content

Figure 9.27

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 high medium low very low

Hydrogen content ml/100 g deposited weld metal >15 15 and > 10 10 and > 5 5 in ISO 2560 classified as Hcontrolled electrodes

hydrogen in weld metal according to DIN 8529. Based on this assessment, a classification of weld metal to DIN 32522 into groups depending on hy-


ISF 2002

Assessment of Diffusible Hydrogen During Manual Metal Arc Welding

drogen is carried out, Figure 9.29.

Figure 9.28

A cold crack development can be followed-up by

Abbreviation HP 5 HP 7 HP 10 HP 15 Hydrogen content ml/100 g deposited weld metal (max.) 5 7 10 15

means of sound emission measurement. Figure 9.30 represents the result of such a measurement of a welded component. A solid-borne sound microphone is fixed to a component which measures the sound pulses generated by crack development. The intensity of the pulses provides a qualitative


ISF 2002

H2-Bestimmung nach DIN 8572,2

Diffusible Hydrogen of Weld Metal to DIN 32522

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 segregations are

Figure 9.30

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. Such support cannot be obtained perpendicular to rolling direction, thus the strength of the workpiece is that of the weaker microstructure
Figure 9.31

9. Welding Defects


areas. Consequently, a lamellar crack propagates through weaker microstructure areas, and partly a jump into the next band takes place. Figure 9.32 illustrates why such t-joints are particularly joint shape, these welds show to some extent a con12 100



welded construction which greatly impedes shrinking of this the joint, plane may of generate

100 3




ISF 2002




Figure 9.32

Shrinkage Values of T-Joints With Various Joint Shapes

above the tensile strength.

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 weld bead 2 3 4

tain alloy elements with a precipitation hardening effect (carbide developer like Ti, Nb, V). During welding such steels, carbides are dissolved in an area close to the fusion line. Dur-

coarse grain zone 1

base metal: ASTM 508 Cl (22NiMoCr3-7)

br-er09-33.cdr ISF 2002





shrinkage value 1,7 mm


shrinkage value 0,6 mm

vulnerable. Depending on

shrinkage value 0,4 mm


developers are not completely re-precipitated.

Undercladding Cracks

Figure 9.33

9. Welding Defects


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.