Abstract Scope |
Introduction
Since the 1990s offshore production with semi-submersible platforms and ships is expanding. In service, the structural elements reach a maximum depth of about 20 meters. Consequently, wet welding with the shielded metal arc process (SMAW) becomes a possible technique to be used in structural repairs. One of the biggest downsides of the SMAW process is that wet weldments can present hydrogen cracking. Some authors state that carbon content (%C) and carbon equivalent (CE) should be kept lower than approximately 0.15 and 0.3, respectively to prevent hydrogen cracking. Nevertheless, quantitative models for predicting the occurrence of hydrogen cracks do not exist. The purpose of this work is to present technical information on hydrogen cracking in the wet welding of ferritic structural steels. relating it with diffusible hydrogen content (Hdif), %C, CE, metal critical parameter (Pcm), hardness, and welding depth. The main contribution is the proposition of algorithms, equations, and graphs to evaluate the risk of WM and HAZ cracking.
Experimental Procedures
The welding tests were performed in the hyperbaric laboratory of UFMG university using a mechanized gravity system at simulated water depths of 20m maximum. The welds were produced by the SMAW process in the flat position with ferritic electrodes. An electronic constant current source operating in DCEN. The welding current ranged from 160A to 190A and the voltage ranged from 23V to 28V.
The experimental 3.25mm oxyrutile electrodes were manufactured in an ESAB industrial production line and insulated with vinyl varnish. Commercial rutile-type electrodes were also tested.
The diffusible hydrogen was measured by the gas chromatography method in samples welded at 0.5m water depth.
Hydrogen cracking in the weld metal (WM) was studied by multipass V groove welds (GW) tests consisted of about 20 to 30 beads deposited on a V groove with an opening angle of 45º, a root opening of 5 mm maximum, and the thickness of the plates ranged from 16 to 19 mm. The quantification of the incidence of cracks in the weld metal was expressed by the number of cracks per unit area measured in longitudinal sections along the weld axis. The metallographic samples were polished with diamond paste and etched with nital 2%. Weld metal hardness (HV10) measurements were carried out in the same metallographic samples.
Hydrogen cracking in the heat-affected zone (HAZ) was studied by bead-on-plate (BOP) tests and controlled thermal severity (CTS) tests. The crack incidence parameter (CIP), measured in transversal sections, was defined as the ratio, expressed as a percentage between the total length of the cracks and the length of the fusion line.
The weld metal and base metal chemical compositions were determined by optical emission spectrometry.
Literature results from Nóbrega, Stalker, Ozaki, et al., Nevasmaa, Gooch, and Klett were used to complement the present experimental work.
Results and Discussion
Hdif measured at 0.5m depth varied from 19.8ml/100g to 23.2ml/100g for experimental oxyrutile electrodes with standard deviations (SD) from 1.79 ml/100g to 8.13ml/100g. The respective values for the commercial rutile type electrodes were 85.4 ml/100g to 97.2 ml with SD from 2.5ml/100g to 17.6 ml/100g.
Results of the incidence of cracks in the WM are based on the variables Hdif, hardness, and Pcm. For HAZ cracking the variables were Hdif and base metal´s carbon content and carbon equivalent.
Guidelines to estimate the risk and prevent hydrogen cracking both in the WM and in HAZ are presented in the form of flowcharts.
The relationship of WM hardness versus Nº of WM cracks/cm2 was presented in the form of a graph. The trend line indicates 165HV as the maximum hardness value acceptable to avoid the occurrence of cracks, regardless of the Hdif value. When the variables involved are Hdif and Nº of WM cracks/cm2, the maximum Hid value acceptable is 23ml/100g. When the variables involved are Pcm and Nº of WM cracks/cm2, the maximum Pcm value acceptable is 0.09.
Results of WM hardness versus cracks/cm2 for rutile-type electrodes down to 20m show a decrease of cracking incidence with increasing welding depth.
The graphs with results of the HAZ crack incidence parameter (CIP), from BOP and CTS tests down to 20m, indicate no cracks occurrence when Hdif < 23ml/100g. With this Hdif value, ship steels which maxima %C and CE are respectively 0.20 and 0.40, approximately, can be welded without hydrogen cracks both in WM and HAZ. Similar data treatment of literature data for steels with maxima %C and CE of 0.28 and 0.48, respectively, indicated no cracks occurrence when Hdif < 12ml/100g.
The general graph of Hdif versus base metal %C and CE from experimental and literature data allows the prediction of the risk of HAZ cracks.
Conclusions
Predictive models for WM and HAZ cracking are suggested. The risk of cracks can be estimated graphically, analytically, or by flowcharts.
Up to 20m, the influence of pressure on Hdif and, therefore, on the incidence of cracks, is not significant in the welding with lower Hdif electrodes. As for the higher Hdif electrodes, there was a significant reduction in their content and, consequently, in the incidence of cracks both in WM and HAZ.
Combining the proposed criteria for obtaining WM and HAZ crack-free welds, 23ml/100g is suggested as the Hdif limit in the welding of any ship steel that meets the ASTM A131/A 131M-08 Standard.
Cracking in the weld metal
Cracks shall not occur for Hdif lower than approximately 23ml/100g regardless of the WM hardness.
Cracks shall not occur for hardness lower than approximately 165 HV, regardless of the Hdif.
Cracks shall not occur if the Pcm is lower than approximately 0.09 regardless of the Hdif.
Cracking in the heat-affected zone
Hdif = 32ml/100g is the maximum value suggested for avoiding HAZ cracks if the base metal %C and CE are lower the 0.20 and 0.40, respectively. If %C and CE are higher than 0.28 and 0.48, respectively, the limit Hdif value is 12ml/100g.
Keywords.
Hydrogen cracks
Underwater Welding
Wet Welding
SMAW
Diffusible Hydrogen
Oxyrutile electrodes |