Hydrogen-Induced Cracking Susceptibility in High-Strength Weld Metal

A formula is proposed to estimate minimum preheat and interpass temperatures which help prevent hydrogen-induced cracking

N. OKUDA, Y. OGATA, Y. NISHIKAWA, T. AOKI, A. GOTO AND T. ABE

ABSTRACT. The hydrogen-induced cracking (HIC) susceptibility of weld metals of high-tensile strength ranging from 70 to 100 kgf/mm 2 (100 to 143 ksi) was investigated to estimate the lowest possible preheat and interpass temperatures to prevent cracking. The investigation was conducted using a multipass submerged arc welding (SAW) process with several types of electrodes and a highly basic flux of the bonded type. The flux was prepared at several baking temperatures to change the diffusible hydrogen contents ([H]D) in the weld metals without changing any other characteristics peculiar to the flux. It was found that the critical preheat and interpass temperatures required to prevent HIC were in proportion to the tensile strength of the weld metal and the logarithm of diffusible hydrogen content of the weld metal as measured by gas chromatographic method ([H] D _GC)- and that they could be safely estimated within those variables.

Introduction As welded structures such as penstocks, bridges, pressure vessels and offshore drilling rigs become larger, higher tensile strength steels, such as HT80 and HT100, as well as thicker steels, are increasingly used in their construction. In accordance with the development of higher tensile strength steels for such projects, research and development is being carried out on welding materials. Some welding consumables have become available that meet the require-

ments for mechanical properties, such as strength and notch toughness, even for low-temperature service. However, particular attention should be paid to preventing HIC, because in general, susceptibility to cracking tends to increase as the plate thickness and/or tensile strength increases. Recently, the weldability of steels has improved significantly. Therefore, welding precautions to prevent HIC in both the heat-affected zone and the weld metal should be considered. For the base metals of HT60 and HT80, many studies of HIC in multipass weld metals were carried out, and some quantitative analysis for the prevention of cracking was reported (Refs. 1-3). However, as far as the tensile strength of weld metals up to 100 kgf/mm 2 (143 ksi) is concerned, there have been few studies of HIC. In general, the susceptibility to HIC is affected by the tensile strength, diffusible hydrogen content, microstructure and residual stress in the weld metal, and for more accurate evaluation, those factors should be investigated at the crack initiation site of the weld metal. For instance, local hydrogen accumulation during multipass welding should be taken into account, as has been proposed (Ref. 4).

In this investigation, the HIC susceptibility of multipass high-tensile strength weld metals ranging from 70 to 100 kgf/mm 2 (100 to 143 ksi) was estimated with reference to the variables of tensile strength and diffusible hydrogen content in the weld metal. The welding variables and plate thickness were fixed so as to simplify the evaluation and to establish an evaluation that can conveniently be performed in industry.

Experiment Procedure
Electrodes and Fluxes Used Several types of electrodes were used to change the tensile strength of the weld metal from 70 to 100 kgf/mm 2 (100 to 143 ksi). The chemical compositions of the electrodes are shown in Table 1. The diameter was fixed at 4 mm (%> in.). The electrodes ranging from A through D are commercially available for HT70 and HT80 steels, and the electrode E was produced experimentally to obtain a high-tensile strength weld metal of 100 kgf/mm 2 . A commercially available flux of the bonded type was modified to produce experimental fluxes which would change the diffusible hydrogen content in the weld metals. The weld metal deposited with the basic flux had a very low hydrogen content, up to 1.5 m/100 g, as measured by the gas chromatographic method. Therefore, the potential hydrogen content of the fluxes for this test was increased by reducing the flux baking temperatures or by adding raw materials containing crystal water. The ranges of baking temperature and times were 150-600C (310-1120F) and 30-120 min, respectively. Talc [rv.g3Si.4Cho (OH) 2 ] and/or Kaolinite [Al2Si20s (OH) 4 ] were added in order to increase the potential hydrogen content of the fluxes in this experiment. The relationship between the square

KEY W O R D S Weld Metal Cracking Hydrogen Cracking Induced Cracking Crack Susceptibility SAW Cracking Tests Basic Bonded Flux Diffusible Hydrogen Flux Bake Temperature Critical Temperature 100-140 ksi Steel

N. OKUDA, Y. NISHIKAWA, T. AOKI, A. GOTO and T. ABE are with Kobe Steel, Ltd., Fujisawa, lapan. Y. OGATA is with Kobe Steel America, Inc., Houston, Tex. Paper presented at the 67th Annual AWS Meeting, held April 13-18, 1986, in Atlanta, Ga.

WELDING RESEARCH SUPPLEMENT 1141-s

Table 1-Chemical Composition of Electrode (wt-%) Base Material' 3 ' HT70 HT80 HT80 HT80 HT100

Wire A B C D E
<a)

C 0.10 0.09 0.10 0.12 0.12

Si 0.12 0.14 0.20 0.11 0.07

Mn 1.84 2.35 2.40 1.99 1.90

70 80 100

P 0.012 0.008 0.010 0.006 0.007

(kgf/mm2)

s
0.005 0.003 0.005 0.005 0.005
100 (ksi) 114 143

Ni

Cr 0.59 0.44 0.79 0.61

Mo 0.53 0.78 0.87 0.75 0.76

2.67 2.78

Minimum tensile strength of steel HT7Q HT80 HT100

root of the moisture content of the fluxes measured at 1000 C C (1832F) and the diffusible hydrogen content of the weld metals is shown in Fig. 1. It was found that the diffusible hydrogen content is proportional to the square root of the moisture content of the flux, precisely in accordance with Sieverts' Law. In such a manner, the diffusible hydrogen content of the weld metal changed from 1.0 to 5.1 m2/100 g, depending on the moisture content of the fluxes and no changes in any other characteristics. The measurement of the hydrogen content was conducted by the gas chromatographic method in accordance with WES 1003-1984. This method has the same collect rate as the mercury method standardized by the IIW (Ref. 5). The difference in diffusible hydrogen content as measured by the glycerin method (JIS) and the gas chromatographic method is shown in Fig. 2. The gas chromatographic method seems to be a more suitable method for the accurate evaluation of diffusible hydrogen content, especially when the hydrogen is at low levels.

Base Metal Used Only one kind of base metal was used, even though numerous combinations of electrodes and fluxes were used. Its chemical composition is shown in Table 2. In this investigation, the base metal had little influence on the cracking susceptibility, since the cracking susceptibility was estimated directly from the tensile strength of the weld metal in the multipass weldments. Experimental Process The test assembly used to evaluate the multipass weld metal cracking is shown in Fig. 3. Fillet welds restrained the plate to prevent distortion. However, after the welding was completed, considerable residual stress remained in the weld metal because the distortion caused by the stress from this process was restricted after the first several passes. It is reported that the maximum residual stress in the weld metal tends to level off to a constant value with plates of a thickness of more than 50 mm (2 in.) (Ref. 6). There-

fore, it seems that the cracking susceptibility can be estimated accurately by using U-groove prepared test plates. Welding was conducted under the conditions shown in Table 3. All welding was performed with AC power. The temperature of the test assembly was maintained within a tolerance of 5 C (9F) from the designated temperature during welding. These temperatures appear in Fig. 4. After the groove joints were filled, the test plates were left for more than 96 h in the atmosphere. The cracking in the weld metal of the plates, from which the reinforcement and restraint plates had been removed, was then investigated. The investigation was conducted mostly by using an ultrasonic detector. In some cases, dye penetrant tests and visual observation tests with a magnifying glass were carried out on the longitudinal and transverse sections of the weld metal. It is well known that the tensile strength of the weld metal depends on the cooling rate, and the cooling rate is influenced by the preheat and interpass temperatures when the materials and welding variables are the same. Therefore, all the weld metal tensile specimens were taken from the sound portion of the test coupon, and they were tested after evaluation for HIC. Additionally, in order to ascertain the influence of the hydrogen, the diffusible hydrogen content of the weld metal deposited from each combination of electrode and flux was measured. The flux was also removed once it had been used so as to keep the potential hydrogen in the flux constant. Furthermore, the influence of the welding variables on the diffusible hydrogen content of the weld metal was

CJl CD O

u o ; Agglomerated Flux

" ^
(JD

/ /.

O o S

yC% ' o
of
/O o

; Fused Flux-

/ /

4
a> o c:

/
6 5 H

/ y

o o o o
o

jr

: % y/
o s

%Ar,

^ >
a>
JO

cn o

o
0/

:y 4 /
/ "

-D

>
t
$/

/t

o
i i . i ,

,
0.5 1.0
1.5
1/2

(Moisture Content in Flux Measured at 1000'C) : :. ft) Fig- 1 Relationship between square root of moisture content in flux measured at 1000"C (1832F) and diffusible hydrogen content in the weld metal

1 7 3 4 5 Diffusible Hydrogen Content by JIS Z3116 m^/lOOg (glycerin method) Fig. 2 Relationship between diffusible hydrogen content as measured by the glycerin (/IS) and gas chromatograph methods

142-s | M A Y 1987

Table 2 - -Chemical Composition of Base Material Base material HT80 Thickness (mm) 50 Chemical Composition (wt-%) C 0.14 Si 0.28 Mn 0.78 P 0.005 S 0.005 Cu 0.27 Ni 0.99 Cr 0.46 Mo 0.39 V 0.042 B 0.0010 Ceq<a< (WES) 0.50
p c m (b)

0.276

(bl

Ceq (WES) = C + 1/24S + 1/6Mn + 1/40Ni + 1/5Cr + 1/4Mo + 1/4V Pcm = C + 1/30Si + 1/20Mn + 1/20Cu + 1/60Ni + 1/20Cr + 1/15MO + 1/10V + 5B

investigated. T h e w e l d i n g variables such as current, v o l t a g e and travel speed h a d negligible effects o n the diffusible h y d r o g e n c o n t e n t . T h e e f f e c t o f v o l t a g e is s h o w n in Fig. 5.

Test Results and Discussion

T h e influence o f the tensile strength (<rB) and t h e diffusible h y d r o g e n of t h e w e l d metal w e r e investigated t o establish the critical preheat and interpass t e m p e r atures f o r the p r e v e n t i o n o f HIC. T h e w e l d i n g w a s multipass using t h e S A W process a n d several electrodes a n d p r e heat and interpass t e m p e r a t u r e s . A n e x a m p l e o f a typical transverse w e l d crack w h i c h w a s f o u n d in the longitudinal cross-section of t h e w e l d m e t a l , w i t h a tensile strength o f 101 k g f / m m 2 , is s h o w n in Fig. 6. T h e SEM f r a c t o g r a p h o f the transverse crack near t h e initiation site is s h o w n in Fig. 7. A n intergranular fracture surface a n d a quasi-cleavage fracture surface w i t h a s e c o n d a r y crack are o b s e r v e d o n the crack surface. This is the typical f r a c t u r e m o r p h o l o g y caused b y HIC. This t y p e o f cracking t e n d e d t o p r o p a g a t e w i t h increases in the tensile strength o r the diffusible h y d r o g e n c o n tent o f t h e w e l d m e t a l , a n d in e x t r e m e cases, the cracking r e a c h e d t h e surface of t h e plates. O n the o t h e r h a n d , the cracks had t h e t e n d e n c y t o contract as those factors diminished. T h e m a j o r i t y o f the cracks in t h e w e l d metal w e r e o b s e r v e d w i t h i n a range o f 5 t o 15 m m (0.2 t o 0.6 in.) b e l o w t h e plate surface, a location in w h i c h the residual stress and local h y d r o g e n accumulation are said t o reach a m a x i m u m . Finally, n o cracking w a s e v i d e n t e v e n w h e n the preheat a n d interpass t e m p e r a t u r e s w e r e maintained at 5 0 C (122F), p r o v i d e d that those t w o factors w e r e k e p t t o a m i n i m u m . T h e length a n d f r e q u e n c y o f HIC in t h e w e l d metal d e p e n d e d o n t h e tensile strength, diffusible h y d r o g e n c o n tent and preheat a n d interpass t e m p e r a tures. H o w e v e r , t h e critical p r e h e a t and interpass t e m p e r a t u r e s f o r the p r e v e n t i o n o f HIC w e r e r e a d j u s t e d in accord a n c e w i t h t h e diffusible h y d r o g e n c o n tent o f the w e l d m e t a l . T h e relationship b e t w e e n diffusible h y d r o g e n c o n t e n t a n d the critical preheat a n d interpass t e m p e r a t u r e s f o r 89.2 t o 93.8 k g f / m m 2 (127 t o 133 ksi) w e i d metals is s h o w n in Fig. 4A. It w a s f o u n d that t h e critical t e m p e r a t u r e is p r o p o r -

tional t o the logarithm of t h e diffusible h y d r o g e n c o n t e n t w h e n t h e tensile strength of the w e l d metal is almost the same. In t h e same m a n n e r , t h e relationship b e t w e e n diffusible h y d r o g e n and the critical t e m p e r a t u r e f o r p r e v e n t i n g HIC f o r each level o f tensile s t r e n g t h in t h e w e l d metal is s h o w n in Fig. 4B. It w a s found that the critical temperature increased in p r o p o r t i o n t o the logarithm o f diffusible h y d r o g e n , a n d the gradient

f o r each strength level w a s the same, irrespective o f the tensile strength levels in the w e l d metals. Tensile strength level E (77.0 t o 79.3 k g f / m m 2 ) resulted in a v e r y l o w cracking susceptibility c o m p a r e d w i t h o t h e r tensile strength levels. It w a s e s t i m a t e d that t h e d i f f e r e n c e in t h e cracking susceptibility resulted f r o m the varying microstructures of t h e w e l d m e t als (Ref. 7). Bainitic structure was o b s e r v e d mainly in strength level E, w h i l e

Table 3Welding Conditions Welding Position Electrode Dia. (mm) Welding Variables Heat Input (kj/cm) Preheat and Interpass Temp. (C)

Current
(A) 600

Arc Voltage (V) 32

Welding Speed (cm/min) 30

Flat

4.0

38.4

50-175

400(16!

Restraint Weld

Section A-A'
Fig. 3 Plate preparation for multipass weld metal cracking test

Unit ; m m (inch)

WELDING RESEARCH SUPPLEMENT 1143-s

200

o: No Crack : Cracked C ,
o /

CF) I 392

CO 200

150
CO

302

150

o a.

o/
O e(

100
TJ
(1)
i_

212

100

cz CJ
4-J

~5 4 >

1 50
o o /

n i_ cu a) JZ n a> L_
O
Q_

122

50

CD

4 5

10

32

1 m^/lOOg

2 3 4 5

10

32

Diffusible Hydrogen Content Ihln-GC

Tensile Strength c f Weld Metal kgf/mm" A B C D E 107,8 (106,3 -v 109.5) 101,0 ( 98,3 v 103.6) 91,5 ( 89,2 % 93,8) 84,2 ( 82,1 * 86,1) 77,8 ( 77,0 % 79,3) ksi 153 (151 * 156) 144 (140 * 147) 130 (127 1- 133) 120 (117 * 123) 111 (109 * 113) n 7 13 11 8 6

Fig. 4 - Relationship between diffusible hydrogen content (Ho) and critical preheat and interpass temperature for a 70 to 100 kgf/mm2 (100 to 143 ksi) class of high-tensile weld metal

a mixture of bainitic and martensitic structures, or a mostly martensitic structure, corresponding to the tensile strength of the weld metals, was observed for levels A to D. The typical microstructures for that group of weld metals are shown in Fig. 8. Throughout the range of conditions encountered during the experiment, the

LSkA

N**JrW
Fig. 6Example of transverse cracks in a longitudinal section of 101 kgf/mm metal
2

. "&**;

(144 ksi) weld

Fig. 7 Fractograph of transverse weld crack in weld metal of 101 kgf/mm2 (144 ksi)

144-slMAY 1987

2.0 P o l a r i t y : AC 500 Amp-40 C P M

CJl
"%

2,0

O)

a +-> a; E
4->

CD CD

cn

^yyA^-?m*rt

, OO.
Qi Q *-> I J Cl) 1=3 !^ V .
C O
.c

TJ CO

4-> Q

f=
.. J

CJ3 I

CO

_ >
co C cy

c o CJ
X _J
,1

0.5 "

u
X

: Flux consumption
O: [HlD-GC

0.5

LL

_) * '

27

30 Voltage

33 (V)

36

.<.rt^ir\v.a<K/?-,-Vs^;-r- -.-,-->V. :)-; : ; H'a.C-

Fig5- - The influenceof welding variables on flux consumption and the diffusible hydrogen content Fig. 8Microstructure of the weld metal (400 in the weld metal X) critical temperature for preventing HIC was determined from diffusible hydrogen content, while the tensile strength was kept constant. Therefore, the critical temperature was estimated in accordance with the parameters of tensile strength and diffusible hydrogen, by the method of multiple regression analysis, under the assumption that the tensile strength and diffusible hydrogen content as measured by the gas chromatographic method ([H]D-CC) are independent of each other in terms of their influence on cracking susceptibility. The critical temperature Tc was determined with the following formula: Tc = 5.24 r/B + 277 log [H] D . C c - 482 where crB : 77 to 110 kgf/mm 2 (109 to 156 ksi)
[H]D-GC : 1.0 to 5.1 mC/100 g

250

200CO T J CO CU 1J LQ Z3 L_ CO CJ
a ^

No Cracking zone
A/

ft
y/
/ / / / // / a> / X o / /
/

(F) 482

.-Cracking zone 392

o c

150

302

cu
s_

5=

cn

TJ C
a J

'OJ'
i_ _J 4-J
i >

100

-o

4-1

i_ o a ; ai n sz UJ u_ CJ_

//

oy/

/ /> o o
ooo*

212
45

CD

50

//
' i

: N o Crack

122

The maximum residual stress in the weld metal tends to level off to a constant value with plates not less than 50 mm (2 in.) in thickness. However, the cracking susceptibility increases as local hydrogen accumulates in accordance with the increase in weld thickness. Therefore, estimating cracking suscepti-

: Cracked
0

50 100 150 200 250 Tc(Calculated)=5,24CTB+277 log [H.D-GC-482

32

Fig. 9 Comparison of cracking susceptibility variable, Tc (calculated), and preheat and interpass temperatures (measured)

WELDING RESEARCH SUPPLEMENT 1145-s

bility by the above formula is practical for plates of 40 mm (1.6 in.) in build-up thickness, and in the case of doublegroove joints, this formula can safely be used for base plates of 80 mm (3.1 in.) or less in thickness. Figure 9 shows a comparison of the cracking susceptibility between the critical temperature (Tc), as calculated from all experimental plots in Fig. 5, and the actual preheat and interpass temperatures. The critical temperature is equivalent to Tc, and a very close estimate of the cracking susceptibility is possible from the Tc. In Fig. 9, one plot was out of the estimated range. The reason was presumed to be that the mean values of each level of tensile strength were used for the multiple regression analysis. Thus, it is safe to estimate the practical minimum temperature as 1.15 times the Tc (dashed line in Fig. 9). As a result, the practical minimum temperature T (C) estimated within the above range is as follows:

In addition, the influence on cracking susceptibility of weld thicknesses ranging from 20 to 80 mm (0.78 to 3.1 in.) was investigated. In the light of the very low diffusible hydrogen obtained for this experiment, as far as the no cracking zone in Fig. 9 is concerned, it was concluded that the influence of weld thickness on cracking susceptibility was sufficiently minimal, in that no cracking was recognized in weld metals ranging from 20 to 80 mm in thickness.

by the gas chromatograph method (1.0 to 5.1 mC/100 g). The relationships are: T = 1.15 Tc and Tc = 5.24 r/B + 277 log [H]D-CC 482. In order to apply the above formula widely in industry, it will be necessary to conduct a detailed investigation of the influence of plate thickness or weld metal thickness on cracking susceptibility.

References

Conclusion
Multipass SAW cracking tests were conducted to investigate the hydrogeninduced cracking susceptibility of hightensile strength steel weld metals. In the case of a weld thickness of 40 mm (1.6 in.) on a 50-mm (2-in.) plate, the practical minimum preheating and interpass temperature (T) was related to the tensile strength (<xB) of the weld metal (77 to 110 kgf/mm 2 ), and to the diffusible hydrogen content in the weld metal, as measured

T(C)= 1.5 Tc
Tc = 5.24 crB + 277 log [H] D . CC - 482

1. Yatake, et al. 1981. journal of the lapan Welding Society 50(3). 2. Kishimoto, ef al. 1971. journal of the bopan Welding Society 46(8). 3. Evans, G. M., etal. 1971. Metal Construction 3(5). 4. Suzuki. 1982. The Fourth International Symposium of the lapan Welding Society, 4JWS-1I-10. 5. Godai, er al. 1980. R & D Kobe Steel Engineering Report, Vol. 30, No. 3. 6. Takahashi, et al. 1979. journal of the japan Welding Society 48(10). 7. Asakura, etal. 1980. journal of the Japan We/ding Society 49(8).

WRC Bulletin 318 September 1986

The primary objective of this Bulletin, which contains t w o papers, is to present a comprehensive picture of the research work conducted to establish the current techniques and procedures for specifying the ferrite content of austenitic stainless steel weld metal and measuring its level.

Factors Influencing the Measurement of Ferrite Content in Austenitic Stainless Steel Weld Metal Using Magnetic Instruments
By E. W. Pickering, E. S. Robitz and D. M. Vandergriff This report describes a program conducted under the auspices of the Welding Research Council (WRC) Subcommittee on Welding Stainless Steel to identify the o p t i m u m procedure for the preparation of austenitic stainless steel weld samples for Ferrite Number (FN) determination.

Measurement of Ferrite Content in Austenitic Stainless Steel Weld Metal Giving Internationally Reproducible Results
By E. Stalmasek This report is a summary of the results of 14 years' work by the IIW Commission 2 in the field of ferrite content measurement, done prior to 1978. The publication of these reports was sponsored by the Subcommittee on Welding Stainless Steel of the High Alloys C o m m i t t e e of the Welding Research Council. The price of WRC Bulletin 318 is $24.00 per copy, plus $5.00 for postage and handling. Orders should be sent with payment to the Welding Research Council, Suite 1 3 0 1 , 345 E. 4 7 t h St., New York, NY 10017.

146-s | M A Y 1987