Compositional factors controlling solidification cracking in C-Mn steel laser welds

TWI Technology Briefing 623 - September 1997

G I Rees

FULL REPORT

The possibility of using lasers to weld steel of greater than 10mm thickness is attractive for many reasons. The speed with which laser welding can be performed, due to high power delivery and the deep penetration characteristic of the process, offers the possibility of significant increases in productivity over conventional welding, as does the ease with which laser processes can be automated.

Background

Many of the characteristics of laser welds are advantageous in themselves. The deep penetration effect produced by 'keyhole' welding and the small weld volume result in significantly lower levels of distortion than would be the case for arc welding. Low hydrogen levels are also easy to maintain.

Laser welds are, however, prone to solidification cracking and porosity. Recent work at TWI, investigating the relative susceptibility of various steels to solidification cracking, has revealed that the relationship between cracking susceptibility and the composition of the steels is different from that seen in arc or EB welds.

The task of the present research project's first year was to identify suitable test apparatus for ranking the relative solidification crack susceptibility of C-Mn and low-alloy steels. It was concluded that the Russell test gave results which were in accordance with those obtained from laser butt welds and which were a significant improvement on those produced by the test previously used at TWI. The test involves producing melt-runs, rather than actual welds, and can therefore be used only to rank material crack sensitivity, rather than determining suitable parameters for real welds. However, the morphology of the cracks formed in the Russell test matches that of cracks in actual laser butt welds. The cracks are short, typically <5mm, and are intermittent, located at the plate mid-thickness.

This report details further work carried out to rank the solidification cracking susceptibility of a range of steels of different compositions, using the Russell test.

Objectives

• To determine the typical degree of scatter for repeated tests using the Russell test apparatus.
• To rank the solidification cracking susceptibility of a range of C-Mn steels, using this test.

Experimental work

Eleven C-Mn steels ranging in composition between 0.09-0.16%C, of 12mm thickness, were tested using the Russell test apparatus, with a load of 12kN applied to each specimen. This load had been shown in previous work to produce cracking in even the most crack resistant steels tested. The welding speed and laser power used in these tests were 1 m/min and 8.5kW respectively. Preliminary trials on four steels indicates that three repeat tests were required to allow for the typical scatter in the test results.

Following laser welding, the specimens were X-radiographed to reveal the extent of cracking. Optical metallography was carried out on selected welds, to reveal the solidification structure. A relationship between cracking sensitivity and steel composition was also determined from the test results by regression analysis.

Results and discussion

Results of the Russell test show a strong decrease in the cracking tendency of the steels examined as their carbon content increases in the range 0.09-0.15%C. This trend is opposite to that found for the effect of carbon on solidification cracking by many other workers. The regression relationship fitted to the test data also indicates that increasing the sulphur content of the steel increases the tendency for cracking, but its effect is weaker than that for carbon.

A possible explanation for the observed results is that susceptibility of these steels to solidification cracking is determined by the temperature range for solidification, which decreases with carbon content over the composition range which they cover.

The larger the solidification range, the greater will be the strains developed due to thermal contraction between the onset and completion of solidification. The magnitude of such strains may be the most important factor in determining the relative susceptibility of C-Mn laser welds to solidification cracking. Since laser welds cool and solidify rapidly, the opportunity for impurity segregation may be less than in conventional welds. This may explain why the effect of sulphur on the cracking sensitivity, as determined by the regression analysis, is weaker than that of carbon.

Results of the tests on the 11 steels studied indicate that for most steels, scatter in the results of repeated tests is not great. One steel, however, did show appreciable scatter. A possible explanation for the scatter in results from this steel is that it has both a low carbon content and a low sulphur content. These two elements appear to have opposing effects on cracking sensitivity. Variations in their distribution may cause different susceptibility to cracking in different specimens.

Main conclusions

• Results obtained using the Russell test, at an applied load of 12kN, indicate that scatter is significant for some steels and that an average of the total crack length from three tests is required to give a suitable representationof crack susceptibility.
• In the steels examined, the crack susceptibility is strongly affected by the carbon and sulphur content, with increasing carbon decreasing the risk of cracking. This effect may be due to reduction in the freezing temperature rangeof the steels with increasing carbon content, over the range of compositions examined.
• A composition-based index representing crack susceptibility of the steels tested has been determined by regression analysis. While the correlation between the test results and the predictions of this index is good, the equationshould not be used for steels outside the composition range for which it was determined.

If L is the summed length of cracks in a specimen, the equation is

The equation should be regarded as specific to the material thickness and welding conditions used.

Member Report No. 623-1997
Compositional factors controlling solidification cracking in C-Mn steel laser welds