Investigation of HAC- susceptibility of multi- layer welds with the “BEAD BEND TEST” Procedure and examples

M. Fiedler, H. Königshofer, J. Fischer, G. Posch, W. Berger

Böhler Welding Austria GmbH, Kapfenberg, Austria

Abstract

This report introduces a practical research method for evaluation of the susceptibility to hydrogen-assisted cracking- (HAC) of multi-layer welds. The test method called “BEAD BEND TEST” was developed by Böhler Welding Austria GmbH. It allows to evaluate the HAC susceptibility of weld metals in practical on-site conditions. The stress condition which essentially determines the loading of the welded joint and thus causes the cracking can be changed by various clamping conditions. Therefore it is possible to perform theoretical investigations as well as evaluations of on-site conditions. In this report two application examples are shown to describe the test approach. Special emphasis is laid on practical test parameters.

Introduction

It was recognized very early that in weldments with increased hydrogen content in combination with tensile load and critical microstructures HAC occurs. This happens especially at temperatures below 200 °C [1, 2]. HAC appears primarily in the heat affected zone (HAZ) of the parent metal as longitudinal cracks and is well-known as “under-bead cracks”. Especially in high-strength joints as well as in the weld deposit hydrogen induced faults can be found. These faults are recognizable as brittle fracture spots in fracture surfaces

(e.g.

tensile specimen) as well as cracks in the cross-section of longitudinal cut through the weld seam [3]. In Figure 1 the hydrogen induced faults in transversal and longitudinal direction of the weld seam are shown. The picture indicates schematically the preferential occurrence of these faults in the upper third of V-joints. That is well founded by the locally increased tension condition and the higher quantity of stored hydrogen.

M.

Fiedler and H. Königshofer are development engineers forunalloyed welding consumables

J.

Fischer is welding technologist and specialist for welding of pipeline electrodes

G.

Posch is head of research and development department

W.

Berger is general manager for welding technology, quality affairs and research & development

Transversal direction Longitudinal direction

Figure 1: Appearance of hydrogen induced faults in the weld metal

Considering the increased danger potential due to so called hydrogen-induced cracks (HIC) many institutes laid special effort in the investigation of this type of cracks. During the last 30 years many different test methods became developed in this research field and are world-wide in use. These test methods can be divided roughly into two major groups. The first group deals with externally loaded specimen (e.g. IMPLANT- test) [4-7]. The second group contains “self-restraint specimens”. This group comprises a lot of different tests (according literature [4, 8-13]):

• TEKKEN
• U- GROOVE WELD CRACKING-/LEGHIGH-
• CTS- (Controlled Thermal Severity)
• RRC-(Rigid Restraint Cracking)/IRC-(Instrumented Restraint Cracking)
• GBOP- (Gapped Bead-On-Plate)
• RGW
• CRUCIFORM
• WIC- (Welding Institute of Canada)

• TRC- (Tensile Restraint Cracking) Test. The advantage of the second group is that the tests reflect on-site conditions more clearly. However a correlation of the obtained results to the actual welds is very difficult – and sometimes impossible. In order to achieve a good compliance between the HAC susceptibility test method and the practical weldments, joints identical to the on-site situation should be welded [9]. For reach the highest possible compliance Böhler Welding Austria was keen to develop a practical test method for evaluation of HAC susceptibility of multi-layer welds. This test method covers the requirement of identical stress conditions and delivers a high correlation between theoretical and practical aspects. The test method is described in the following, whereby two examples are exemplarily shown.

Examination procedure – “Bead bend test”

The „BEAD BEND TEST” is an examination method for the evaluation of the hydrogen-assisted cracking (HAC) susceptibility of multi-layer weldments. Special emphasis is laid on a very easy but nevertheless comprehensive investigation of welded joints. Originally the test was developed for comparative investigations of welded circumferential joints in the pipeline construction. From the obtained experience a test procedure was derived to evaluate the HAC susceptibility. Figure 2 shows the setup for welding the test specimen including the typically used specimen dimensions. For welding the specimen plates are attached to an inflexible bottom plate by fillet welds. To provide the high stiffness of the system the wall thickness of the bottom plate should be selected four times higher than the thickness of the specimen plate. Basically the BEAD BEND TEST shows a high resemblance with the RRC-test. Both tests are using self-restraint specimen where the shrinkage stress is determined by the testing facility and the specimen dimensions. In consideration of the testing facility the shrinkage stress is influenced highly by the free length between the clamps. Normally the shrinkage force increases by reducing the so called fixing length. Figure 3 illustrates two different possibilities for fixing the specimen plates on the bottom plate. Both options are interesting for practical investigations. In the first alternative (type A) the specimen plates became welded onto the bottom plate. This results in a nearly zero fixing length, which induces a high shrinkage force in the weld seam. With type A a fully suppressed shrinkage can be reached. This type of fixing differs from alternative B, where a transverse distortion of the specimen plates is possible. An angle-deformation which occurs typically during welding of V-joints is prevented. Type B is used particularly for investigations where a transverse shrinkage of the weldment can be allowed. This load situation is typically apparent by welding of circumferential welds in the pipeline industry. Two up to four welders are working simultaneously at the pipe and its movement is not substantially prevented. Therefore a very small movement in the longitudinal direction is possible but angle deviation normally doesn’t occur. Therefore the type B of the given fixing methods is the most suitable one.

II-1566-05 II-A-154-04

Figure 2: facility for welding the test specimen: 1…parent metal, 2…weld seam, 3…test specimen, 4… firm base plate, 5…anchored welds; d…wall thickness of parent metal

Figure 3: different fixing conditions by welding the test specimen; A…fully hindered shrinkage B…free transverse shrinkage – hindered angle-deformation

After fixing the specimen plates onto the bottom plate welding with parameters in line with practical recommendations is performed. By using critical welding parameters like low heat input and/or low interpass temperature and so on, cracks can be triggered in the weld metal and the heat affected zone. These cracks are formed during welding but also a delayed cracking is well known. The delayed cracking is very well described in literature [1] and [14]. To stay abreast of all developing cracks the test specimen is extricated approximately 24 hours after welding. This final specimen becomes polished at the examination surface (refer figure 4). Figure 4 shows the test specimen with the variable dimension X which equals the wall thickness of the welding joint and the dimension Y which is typically given with 10 mm.

The examination or inspection surface is the longitudinal section of the weld seam and lies generally in the middle of the seam. In the mid of weld seam the local stress shows a relative maximum and the highest susceptibility of cracking is observed. Figure 5 illustrates the distribution of residual stresses by using a strict fixing of the specimen plates (fully suppressed distortions). At the upper third zone of the wall thickness the highest tensile stress builds up and because of this the cracks initiate in this zone.

Figure 4: extricated test specimen with polished examination surface 1…parent metal, 2…weld metal, 3… specimen

Figure 5: distribution of stress in a V- butt weld after J.

B. Roelens; hindered transversal and angle-shrinkage (values in MPa)

After extricating and preparation of the test specimen a dehydrogenisation annealing at 250°C for 16 hours is performed. The reason for that is to prevent a superposing of the test results by the fish-eye-phenomena. The hydrogen effuses out of the specimen and will be not activated by bending. After the dehydrogenisation the deformation of each test specimen takes place via bend in order to make the hydrogen-assisted cracks visible. Due to the bending,

II-1566-05 II-A-154-04

the micro cracks are flared and grow to a visible size for inspection. Figure 6 exhibits the bending facility with a moderate bend radius of 90 mm. In figure 7 two specimen of bended specimen are shown. By looking at the characteristic appearance of the examination surface it is easy to distinguish between those which have typical imperfections due to HAC and those which haven’t. The relatively large inspection surface provides good information for the HAC susceptibility by testing of only one test specimen per parameter setting.

Figure 6: bending facility

Figure 7: bended test specimen with HAC

Example 1: Determination of the minimum interpass temperature when welding with cellulosic electrodes

Cellulosic electrodes are primarily used for the economic welding of circumferential joints in the pipeline construction. The name of this electrode type is descended from the special ingredients in the coating of the electrodes. The main components of the coating are cellulosic powders and other organic materials. Due to these ingredients a high hydrogen input of approx. 40– 50ml/100g deposited weld metal (acc. AWS A 4.3) is achieved. The high amount of hydrogen brought in by welding with cellulosic electrodes requires a preheating of the base material and the keeping of the chosen interpass temperature. From practical experience the dependency of temperature against strength level and wall thickness is more or less known. To estimate proper interpass temperatures for appropriate customer recommendations and to conduct welds without HAC test series became performed where the BEAD BEND TEST was used. Special emphasis is laid on welding with practical welding parameters. The welding parameters were derived from welding procedures usually used in the pipeline industry. The following electrode types (whole tensile strength variation of cellulosic types) were examined in this test series:

• E 6010
• E 7010
• E 8010 and

• E 9010 The electrodes became welded on pipeline steel plates with a wall thickness of 5 up to 25 mm. Dimension of air gap and root face are shown in figure 8.

Figure 8: generally used joint bevel preparation by welding with cellulosic electrodes

For applied fixing alternative of the specimen plate was type A (refer figure 3). The root and the hot pass welding was performed with diameter 4 mm electrodes. For welding of fill and cap layers diameter 5 mm electrodes were used. The whole examination program were included interpass temperatures of +20, 50, 80, 100, 120 und 140 °C. The welds were made with 5, 10, 15, 20 and 25 mm thick plates. That means for each electrode a total of 30 samples was welded to investigate the relationship of HAC with interpass temperature and wall thickness. Figure 9 exhibits schematically the analysis of the BEAD BEND TEST examination surfaces for a 15 mm thick joint welded with a high strength electrode E9010 at 3 different interpass temperatures. It can be seen that an interpass temperature of 120°C is enough to obtain a welding joint without HAC. By decreasing the interpass temperature down to room temperature the crack susceptibility increases dramatically.

Figure 9: examination surfaces of BEAD BEND TEST specimen at 3 different interpass temperatures for a 15 mm thick parent plate – used electrode: AWS E 9010 Figure 10 shows the HAC susceptibility when welding with an electrode type E9010 for different interpass

II-1566-05 II-A-154-04

temperatures and different wall thicknesses. The absolutely minimum interpass temperature is defined marginal above the temperature where HAC begins to show on the examination surface of the BEAD BEND TEST specimens. The minimum recommended interpass temperature for welding on site is chosen 20°C higher than this minimum interpass temperature.

Figure 11 illustrates the recommended interpass temperature as a function of wall thickness and weld metal strength. From this relationship an equation can be derived for the interpass temperature where no HAC will occur.

31413

IPT rec = 58 + 42 · ln( wt )

.

weldmetal eh

;

rec.IPT…recommended interpass temperature [°C] wt…wall thickness [mm] Reh,weldmetal…yield strength of joint weld metal [MPa]

This simple formula can be used to estimate the recommended interpass temperature by welding with Böhler cellulosic electrodes. The hydrogen content of the weld metal is not taken into account in this equation due to generally high hydrogen contents when welding with cellulosic electrodes. It shall also be noted that the hydrogen input cannot become significantly changed by exposing the electrodes to humid atmospheres. Further the rebaking of electrodes is not allowed by manufacturers.

Example 2: HAC susceptibility depending on diffusible hydrogen content and weld metal strength when using a combination of cellulosic and basic vertical-down (BVD) electrodes

Example 2 covers a welding technology with high practical relevance in the pipeline industry. It deals with the combined use of cellulosic electrodes for the welding of root and the hot pass layer and basic vertical-down electrodes for filling the remaining joint. By using cellulosic electrodes a high pipeline erection velocity can be reached due to the high welding speed in the root pass. This bears in mind that the velocity of root pass welding is the most important influencing factor for high-efficiency pipeline welding. The fill and cap layers become welded with basic vertical-down electrodes. This electrode type provides low hydrogen content and a weld metal with generally high toughness. To examine the influence of the weld metal strength together with different diffusible hydrogen contents on HAC a test series was started using these mentioned variables. The test series includes seven different strength levels of the weld metal achieved with simple manganese alloying. The yield strength levels covered a band width of 670 up to 814 MPa. The next step was to generate 3 different hydrogen groups of all the 7 strength levels. The generation of three different hydrogen content groups was obtained by an exposure of the electrodes to an atmosphere with 80% rel. humidity/27°C in a climate chamber for different time spans. Table 1 exhibits the three hydrogen groups and the reached diffusible hydrogen content according AWS A 4.3.

Hydrogen group	Diffusible hydrogen content per 100g weld metal [ml]
group I	6
group II	9
group III	12

Table 1: diffusible hydrogen content for the basic vertical down electrodes acc. AWS A 4.3.

In sum this investigation covered 21 different conditions for the basic electrodes includes three hydrogen content groups together with seven strength levels. These electrodes were used to weld fill and cap layers. The investigation used a pipeline steel X80 according API 5L with a specified minimum yield strength (SMYS) of 550 MPa. The wall thickness of the test plates were 18,3 mm. The joint preparation can be seen in figure 8.

II-1566-05 II-A-154-04

The welding was performed in PA Position holding a constant interpass temperature of 100°C. The fixing of the specimen plates was made with type B (refer to figure 3) enabling a transversal distortion and hindering the angle-deformation. Figure 12 shows the layer sequence for all 21 test plates.

Summary

1. For an accurate prediction of HAC in weldments it is necessary to develop examination methods which allow the welding of specimen within on-site conditions. Further more it is important to be able to test the welds with simple facilities and high significance. With the BEAD BEND TEST these criteria can be fulfilled. Now it is possible to weld identical test-welds for many applications for evaluation of the HAC susceptibility.
2. For self-restraint specimen the tensile stress depends highly on the dimension of specimen, the welding procedure and the fixing method. The BEAD BEND TEST uses different fixing types. A very rigid alternative is suited for theoretical evaluation of HAC; a second alternative with more degrees of freedom has a high relevance for practical usage in the pipeline industry.
3. The BEAD BEND TEST provides a good prediction of HAC due to the relatively large examination/inspection surface.
4. For weldments in the pipeline industry produced with cellulosic electrodes the HAC susceptibility was measured as a function of wall thickness and strength level of the all-weld metal. In that way it was possible to prepare recommendations for the appropriate interpass temperature.
5. The modern pipeline construction often uses a combined welding procedure with cellulosic and basic vertical-down (BVD) electrodes. The welding of the root and hot pass are made with cellulosic electrodes to achieve a high pipeline erection velocity. The fill and cap layers are welded with BVD electrodes to obtain high toughness and low hydrogen values. The BEAD BEND TEST allows also evaluating weldments with different strength levels and moisture contents of the electrode coating. By varying critical parameters indicators for the HAC susceptibility can be found.
6. By increasing the hydrogen-level and the yield strength of the weld metal the HAC susceptibility increases.
7. By increasing the wall thickness of the weld joints the HAC susceptibility increases causing an unfavourable stress distribution in the joint.

Acknowledgement

The authors wish to thank all co-workers at the Böhler Welding Austria for there friendly and helpful assistance to compile this work.

II-1566-05 II-A-154-04

References

1. G. Gnirß. Hydrogen during welding. Part 1. TÜ 17 (1976), No. 11, page 367
2. G. Gnirs. Wasserstoff und seine Wirkung beim Schweißen. Teil 2. TÜ 17 (1976), No. 11, page 414

- 432

3. E. Perteneder, H. Königshofer, R. Bischof.

Capabilities and limitations of modern welding consumables suitable for girth welding of pipelines

– a producers perspective. 3^rd international pipeline technology conference, Brugge, Belgien, May 21-24.2000.

1. IIW- Doc. No II-1513-03. Cold cracking tests.
2. H. Granjon. Information on cracking tests. IIS/IIW-93- 62
3. IIW-Doc. No IX-1240-82. Cold cracking test methods using Implants
4. IIW- Doc. IX-1729-94. The investigation of implant cold cracking test
5. Pedder et al. CTS testing procedures: the present position. The welding institute research Bulletin, Sept. 1975
6. Satoh et al. Japanese studies on structural restraint severity in relation to weld cracking. Welding in the world 15 (1977), No. 7/8, pp. 155
7. H. Hoffmeister. Concept and procedure description of the IRC-test for assessing HACC. Steel Research (AEW), Vol. 57 (1986), No. 7, pp 344
8. Britisch Standard (BS) 7363, 1990. Controlled thermal severity (CTS) test and bead-on-plate (BOP) test for welds
9. A.P. Chakravarti, S.R. Bala. Evaluation of weld metal cold cracking using the GBOP test. Welding Journal Vol. 68 (1989), p. 1s-8s.
10. JIS Z 3157, 1993. Method of U-groove weld cracking test.
11. A. R. Troiano. The role of hydrogen and other interstitials in the mechanical behaviour of metals.

Edward de Mille Campbell Memorial Lecture (1977)

p. 54-79