Year

2003

Degree Name

Doctor of Philosophy

Department

Faculty of Engineering

Abstract

High strength, cold rolled, corrosion resistant ZINCALUME™ (ZA) and ZINC-HIT EN™ (ZHT) G550 sheet steels are widely used in the construction and automotive industries. The sheet steel is commonly joined by gas metal arc welding (GMAW) because of its low cost and high productivity. However, two significant problems can occur: porosity in the weld bead and loss of strength in the heat affected zone (HAZ). As a result of the strength loss, the design strength is currently down-graded to the fully annealed yield strength (250 or 300 MPa). The full strength potential of the steels cannot be utilised unless the strength loss in the HAZ can be quantified.

In this investigation the microstructures and mechanical properties of the weld and HAZ of gas metal arc welding and flux cored arc welding (FCAW) welded ZA and ZHT G550 1.0 mm sheet steels have been studied and the strength loss in the HAZ has been quantified. A major aim was to correlate the mechanical properties with HAZ microstructure and to correlate microstructure with the weld thermal cycle. The effect of welding process variables on the HAZ microstructure, the structure and composition of the weld metal and on the Zn alloy coating have been investigated. HAZ thermal simulations were used to elucidate the structure evolution in the actual weld HAZ.

As expected, the reduction in HAZ mechanical properties is dependent on the heat input. After GMAW at the highest heat input of 130 J/mm the tensile strength of Zn-Al coated steel decreased to 460 MPa (with a yield strength of 400 Mpa) from a pre-weld minimum specified longitudinal tensile strength of 550 MPa. However, a lower decrease in tensile strength to about 520 MPa (with a yield strength of 505 MPa) occurred at the lowest heat input of 50 J/mm. All tensile pieces were fractured in the grain refined region closed to the grain coarsened region of the HAZ .

Although the microstructure of the HAZ was successfully simulated, the simulated HAZ had a much lower tensile strength than the real welds at similar heat inputs due to the absence of the 'brazing effect' or 'two phase aggregate effect' present in the real welds.

It was found that at the same nominal heat input the strength of the welded joint is significantly affected by individual welding parameters, particularly, torch angle, shielding gas and welding consumable. At the same nominal heat input FCAW generated welds with a lower tensile strength than GMAW . GMAW using perpendicular and forehand welding techniques tended to produce welds with a higher tensile strength than the backhand technique. A higher strength loss was also evident for GMAW with pure C02 or a shielding gas containing helium.

No obvious macro-defects or Zn penetration cracks were found in the weld bead or HAZ . However, gas pores and aluminum rich particles were observed in the weld metal. These features did not adversely affect the overall weldment tensile properties.

A more globular weld bead with a higher surface tension, corresponding to a lower oxygen content in the weld metal, was found for GMA welds shielded by Argoshield 60 (98.5%Ar+1.5%02), compared with shielding gases with a higher oxygen potential. Flatter and smoother weld beads were observed in the FCA welds despite a significantly lower oxygen content than the GMA welds. The presence of a slag layers on the weld pool and a significantly higher nitrogen content m a y account for this weld profile.

The oxygen content in the GMA weld metal increased with an increase of the oxygen potential of the shielding gas. This resulted in an increase in volume fraction and mean particle size of the non-metallic inclusions, consistent with the prediction of the Franklin equation. A high volume fraction and large mean particle size were also found in the FCA welds even though the FCA welds contained significantly Lower oxygen and higher nitrogen and aluminum contents. The inclusions in this case were predominantly AlN, not oxides, and therefore the 'Franklin' equation is inappropriate for the prediction of inclusion volume fractions of self shielded FCA welds.

Damage to the coating by the welding thermal process has also been characterised and is concluded that some form of surface coating remains after welding even in close proximity to the weld fusion boundary. A white layer with a composition close to Fe3Al was formed adjacent to the weld bead for the Zn-Al coated steel and the Fe-Zn alloy phase Ƭ was detected in the high peak temperature region for the Zn coated steel. However, the changes to the coating had little effect on the soundness and mechanical properties on the joints.

The investigation confirms the feasibility of arc welding of high strength Zn alloy coated sheet steels using GMAW and FCAW . The strength loss has been quantified, the processing factors exerting the most significant effect on strength loss have been identified, and the mechanisms of strength loss have been established.

The research provides guidelines for the optimum welding process and welding conditions to minimise the strength loss during arc welding of cold rolled sheet steel. The quantification of the strength loss for GMA and FCA welding high strength sheet steel should be of a considerable value in structural design and the results have the potential for incorporation into appropriate Australian Standards.

Share

COinS