Year

2004

Degree Name

Doctor of Philosophy

Department

Department of Materials Engineering - Faculty of Engineering

Abstract

Hydrogen assisted cold cracking (HACC) represents a significant threat to the integrity of welded steel constructions. HACC occurs at near ambient or lower temperatures and may be delayed for several hours, or even days, after welding is completed.

Significant advances in developing thermomechanically controlled processed steels, with reduced carbon and alloy content, resulted in steels that are more resistant to HACC in the heat affected zone (HAZ). This has enabled the use of lower preheat temperatures during fabrication of welded structures. However, the improvement in the resistance of the HAZ to HACC has shifted the problem of cold cracking to the weld metal.

The objective of this current work was to establish the effect of flux cored arc welding(FCAW) welding parameters, such as welding current, contact-tip toworkpiece distance (CTWD) and shielding gas type, on diffusible hydrogen (HD) content for single run, bead-onplate welds using low strength seamed and seamless gas shielded rutile wires of E71 T-1 classification. The work has shown that under most conditions investigated, the weld metal HD levels for the seamed rutile wire were above the 10 ml/100g specified by the consumable’s classification (H10). The measured range of diffusible hydrogen for the H10 wire was 8.3 to 17.0 ml/100g, with the highest hydrogen content being obtained at the lowest welding current of 280 A, shortest CTWD of 15 mm and deposited using 75Ar-25CO2 shielding gas. In contrast, the seamless wire met requirements of the H5 classification (HD ≤5 ml/100g) for all welding conditions investigated, with a range of HD levels of 0.9 to 3.5 ml/100g.

In general, lower HD levels were achieved when using CO2 shielding gas, although the effect is less significant with the H5 seamless rutile wire.

The work included an investigation of arc characteristics under typical welding conditions, using high speed digital imaging and laser backlighting, in order to provide information on metal transfer and arc length. Several tests were carried out using the H10 consumable. It was established that the amount of heat generated by resistive heating of the wire prior to melting can exert a strong influence on the weld metal HD content and is more pronounced in welds deposited using 75Ar-25CO2. The measured arc length was reduced significantly when welding under CO2 shielding gas. Despite suggestions in the literature there was no evidence of a change in metal transfer mode to spray transfer on increasing the welding current from 280 to 320 A, transfer mode was globular for all conditions used.

Following the weld metal diffusible hydrogen testing and welding arc imaging work, the weld metal susceptibility to cold cracking was assessed using the gapped bead-on-plate (G-BOP) test at different preheat temperatures. For this part of the work, identical welding parameters those used for the diffusible hydrogen testing were selected. It was found that the H5 wire weld deposits did not reveal any cracking at ambient temperature, whereas all the welds deposited using the H10 wire exhibited cold cracking with no preheat. Weld metal deposited using 75Ar-25CO2 shielding gas resulted in higher susceptibility to cold cracking than with CO2, which correlated to lower HD levels the CO2 tests. Besides the higher hydrogen content, it was also found that higher weld metal hardness corresponded to the greater crack susceptibility in the welds deposited using 75Ar-25CO2 shielding gas. These factors are considered to contribute to higher susceptibility to transverse cold cracking compared with CO2 shielding gas.

Although the overall results indicate that the weld metal susceptibility to cold cracking corresponds to the relevant levels of HD, this relationship was found to be ambiguous in welds deposited at the shortest CTWD of 15 mm, using CO2 shielding gas at all welding currents investigated. While the amount of diffusible hydrogen was marginally increased from 11.7 to 12.8 ml/100g, resulting from the welding current increase from 280 to 320 A, the amount of cold cracking at room temperature was significantly decreased from 89 to 25 %RTC. This is explained by a significant difference in the cross section of the weld beads, suggesting a need to more closely evaluate the G-BOP testing, particularly examining the effects of weld bead profiles on the weld susceptibility to HACC.

Preheat was found to decrease the amount of cold cracking in the H10 welds and it was concluded that preheat significantly reduces the main contributor to decrease the HD in the weld metal. Although the cracking susceptibility of welds using 75Ar-25CO2 shielding gas decreased more slowly with an increasing preheat temperature, compare to those deposited using CO2, no cracking was observed at 120 °C in welds under both shielding gases. This indicates that the same welding consumable (H10) deposited using different shielding gases can result in a different response to preheat temperature.

Based on the results of this work, a number of changes are proposed to hydrogen testing standards AS 3752-1996 and ISO 3690-2000, particularly with respect to the effects of CTWD and shielding gases on levels of diffusible hydrogen in weld metal deposited using gas shielded rutile flux-cored wire.

Share

COinS