Degree Name

Doctor of Philosophy


Department of Materials Engineering


In the current economic climate, there is a growing pressure on the Australian fabrication industry to increase its welding productivity. In response to this situation, the high productivity welding processes of 4-wire submerged arc welding (SAW) and narrow gap welding (NGW) were recently developed by the CSIRO Division of Manufacturing Technology. However, the efficacy of these processes in practice depends ultimately on the structure and properties of the welds, particularly those of the heat affected zone.

The objective of this investigation was to gain a basic understanding of the way in which the high productivity welding conditions affect the microstructure and properties of the heat affected zone of Australian structural plate steels. The steels investigated were a C-Mn steel with a minimum yield strength of 250 MPa and two quenched and tempered steels with minimum yield strengths of 500 and 690 MPa.

Literature relevant to the high productivity welding processes of 4-wire submerged arc welding and narrow gap welding has been reviewed. In particular, published material on the welding processes and the effect of the weld thermal cycle on the structure and properties of the HAZ has been examined. Additionally, a review of weld simulation techniques has been carried out.

The investigation involved three interconnecting studies. Firstly, investigations were conducted on the effects of the welding process parameters of welding speed and heat input in 4-wire submerged arc welding (bead-on-plate and multi-ran) on the structure and properties of the welds. The heat input and welding speed ranges investigated were 2-10 kJ/mm and 400-2000 mm/min. Increasing heat input was found to decrease the toughness, despite the lower hardness; due to the larger prior austenite grain size and formation of coarse and undesirable microstructural constituents. Changing the welding speed at a constant heat input did not, in general, have a significant effect on HAZ hardness, CVN toughness and tensile properties, consistent with a relatively minor variation of the HAZ microstructure.

Secondly, the effects of current type (pulsed and non-pulsed) and weld process type (gas metal arc and submerged arc) on the structure and properties of the HAZ were examined for narrow gap welding. It was found that the current type had only a minor effect on the HAZ microstructure and therefore, a minor effect on the hardness, toughness and strength properties of the HAZ. The process type, on the other hand, was found to exert a significant effect on the HAZ toughness due to different weld bead profiles which affected the extent of reheating of the HAZ. Considerable variability in HAZ Charpy impact values was ascribed to microstructural variations at the location of the notch tip.

Finally, in the third study, the microstructures and properties of the regions of the HAZ and effects of reheating the grain coarsened region to various peak temperatures were examined through weld HAZ simulation studies. Partial CCT diagrams for grain coarsened and grain refined HAZ regions under weld thermal cycle conditions were obtained. It was found that the grain coarsened region generally showed the lowest impact toughness because it exhibited the coarsest ferritic transformation product, the largest prior austenite grain size, and the highest hardness. The latter effect is due to high hardenability associated with the large grain size. The grain refined region exhibited the highest toughness and intermediate hardness, due to the fine microstructure. The partially transformed region had low hardness and an intermediate toughness, because of the structure of recovered ferrite and retransformed austenite regions consisting of pearlite and/or martensite-austenite constituent, depending on the hardenability and cooling rate. The BIS 80 steel was an exception to this trend because a combination of maximum hardness and the highest Charpy toughness was exhibited by the grain refined region. It was concluded that this effect is related to the high hardenability of the steel and the property improvement effected by the refined structure of the grain refined region (GRHAZ).

Reheating of the grain coarsened region generally reduced the hardness and improved the toughness of the original grain coarsened region. However, HAZ simulation also demonstrated the potential for reduced toughness in the intercritically reheated grain coarsened region of high hardenability steels, as the prior austenite grain boundaries were decorated with martensite-austenite islands.

The research investigation provided detailed data on the structure and properties of the HAZ of a C-Mn steel and two quenched and tempered steels after welding by four-wire submerged arc and narrow gap welding processes. Although the data relate specifically to the steels and the processes investigated, the results contribute new knowledge and understanding of the response of the base steel to the arc welding process and have general applicability to the science and technology of arc welding of steels.

It was found that small but significant changes occurred in the width, structure and hardness of the HAZ at various positions around the fusion line of a single weld bead. The inference that the cooling rate varies locally is inconsistent with the commonly accepted proposition arising from the Rosenthal analysis that heat input alone determines the HAZ cooling rate, which is thus assumed to be constant for a constant heat input. A qualified assumption can be made that heat input alone determines the cooling rate for a selected position of the HAZ and that this position can be used to compare the effects of different heat inputs. However, small variations in HAZ structure were also found, particularly for the 250 grade C-Mn steel, for welds produced at the same total high heat input (10 kJ/mm), but for different welding speeds. These findings indicate that the Rosenthal analysis can be a useful approximation when comparing similar positions of the weld bead for different heat inputs, but it should also be recognised that the variables contributing to the heat input can exert a second order effect on the HAZ structure and properties.

Detailed surveys of the structural and property gradients in high hardenability steels established that the maximum hardness can occur in the grain refined region of the heat affected zone and not in the grain coarsened region, as is conventionally assumed. This effect is due to the high hardenability which establishes the same martensitic/bainitic constituents in both regions, but the finer austenite grain size in the grain refined region results in a marked refinement of the transformed structure which contributes to the higher hardness and strength. The refinement of the structure also resulted in a higher toughness.

On the basis of the present investigation of the structural response of the HAZ, it is concluded that, for the C-Mn and the two quenched and tempered steels and the welding conditions studied, increases in productivity can be achieved without compromising the properties of the welded joint by using higher welding speeds (1000-2000 mm/min) at a heat input ≤ 5 kJ/mm in the 4-wire submerged arc welding process.

Although the width of HAZ was narrower in weldments produced by narrow gap welding, the structure and properties of the HAZ were similar to those of the 4-wire submerged arc welding. The effectiveness of the high productivity narrow gap welding process, using either pulsed or non-pulsed current, was also confirmed for the three steels investigated.