Return to search

Weldability of a Dual-Phase Sheet Steel by the Gas Metal Arc Welding Process

Dual-phase (DP) sheet steels have recently been used for automotive manufacturing to reduce vehicle weight and improve fuel economy. Dual-phase steels offer higher strength without reduced formability when compared to conventional high strength low alloy (HSLA) steels and so thinner gauge DP sheet steel can be used to meet the same design requirements. The DP steel microstructure is comprised of dual-phase mixture hard martensite particles, which provide strength, in a soft ferrite matrix, which provides ductility. Fusion welding processes, such as gas metal arc welding (GMAW), are used to join DP sheet steels; however, the heat input from fusion welding can cause the martensite islands to decompose into softer islands of tempered martensite. This can reduce the joint efficiency and cause premature localized necking in the region where tempered martensite forms.
The weldability of coated 1.65 mm Cr Mo DP600 (dual-phase 600 MPa) sheet steel welded using the pulsed gas metal arc welding (GMAW-P) process was assessed. Processes with a range of GMAW P weld heat inputs were developed to make full penetration bead-on-plate welds that had similar bead geometry. The range of weld heat input was between 193 J/mm and 347 J/mm. Uniaxial transverse weld tensile tests of welds that were made at high heat input fractured in the heat affected zone (HAZ), welds that were made at low heat input fractured in the base metal (BM), which is most desirable, and at intermediate welding heat inputs, fracture locations were mixed. Heat input was compared to corresponding weld HAZ half-width measurements and it was shown that as heat input increased, HAZ half-width increased as well; this followed an expected linear trend. The ultimate tensile strength (UTS) was not diminished in specimens that exhibited BM fracture and 100% joint efficiency was achieved. Welded DP600 specimens that failed in the HAZ had minimal (< 5%) reduction of UTS.
During the welding process development phase, the same range of heat input was used to make bead-on-plate full penetration welds onto coated 1.80 mm HSLA (high strength low alloy) sheet steel to assess its weldability. It was found that all of the welds fractured in the BM during uniaxial transverse weld tensile testing and, therefore, had achieved 100% joint efficiency.
It was shown that by increasing the strength grade of DP sheet steel to DP780 and DP980, 100% joint efficiency was not retained. To better understand why high heat input welding caused HAZ fracture, low heat and high heat input welds that had consistently fractured in the BM and HAZ, respectively, were used to assess the differences between BM and HAZ fracture mechanisms.
Fractographic analysis of BM and HAZ fracture surfaces of the dual-phase steels showed that fracture had occurred due to micro-void coalescence for both types of failure; however, the HAZ fracture had greater reduction of cross-sectional area and the surface had more numerous and smaller shear tearing ledges. Examination of the microstructure showed that there were decomposed martensite islands in the region the HAZ fracture; these likely increased ductility and led to a more significant tri-axial stress state. However, decomposed martensite was also found in the HAZ of welds that had BM fracture. The low and high heat input welds had similar reduction of martensite percentage (~3 – 4%) in the subcritical (SC) region of the HAZ; immediately below the Ac1 temperature where transformation from a BCC ferrite to FCC austenite occurs. Each weld HAZ was assessed with an average through-thickness microhardness (ATTH) profile. Four distinct regions of hardness were identified: hard intercritical (IC), which was formed by heating between Ac1 and Ac3 temperatures, soft subcritical (S SC), hard subcritical (H SC), and base-metal (BM). The width of the S-SC was slightly larger (~10%) for the HAZ fracture weld; however, the degree of softening (~8 – 11 VHNATTH/200g) compared to BM hardness was similar for both. It appeared that HAZ fracture could be shifted to the BM by reducing the width of the S SC so that the surrounding hard IC (+40 – 50 VHNATTH/200g) and H-SC (+5 – 10 VHNATTH/200g) could support the S SC and prevent a tri-axial stress state from developing; this is similar to increased strength of brazed joints caused by optimal gap width.
Using this knowledge base, new welds were made onto different sheet thickness (1.20 mm and 1.80 mm) Cr-Mo DP600 sheet steels and onto higher strength grades of 1.20 mm Cr-Mo DP780 and 1.20 mm Mn –Si DP980 sheet steels. These were compared with the heavily studied 1.65 mm Cr Mo DP600 sheet steel described above. The 1.80 mm DP600 sheet steel (welded with the same range of heat input) fractured in the BM during all uniaxial transverse weld tensile tests; this was caused by a 4% increase in sheet thickness. The majority of thinner 1.20 mm welds fractured in the HAZ; there was one BM fracture for the DP600 sheet steel. Only the DP980 had a significant drop in UTS (~28%), and the DP600 and DP780 approached 100% joint efficiency (based on the UTS). The same distinct regions of hardness were observed for Cr Mo DP600 and Cr-Mo DP780. The Mn Si DP980 did not exhibit an H SC and had a significantly wider S SC (~80% wider) when compared to welds of similar heat input and sheet thickness. This suggested that the presence of an H SC region could improve joint efficiency. It also suggested that material chemistry played an important role in reducing the extent of softening during welding; however, the martensite percentage for the DP600, DP780, and DP980 were different (approximately 7.5%, 20%, and 46%, respectively) and this could also have affected the observed S SC widths.
It was concluded that GMAW-P welded DP600 sheet steel shifted from a HAZ fracture to a more desirable BM fracture location during uniaxial transverse weld tensile testing as the S-SC region of hardness was narrowed. A narrow S-SC was supported by the adjacent hard IC and H-SC regions, which limited diffuse necking in the vicinity of the S-SC region. Diffuse necking continued to thin out material in the BM region, where there was a greater reduction in cross-sectional area prior to the onset of localized necking, and, therefore, the BM entered a state of higher stress than the S-SC and failed once it reached UTS. This was not observed for a higher strength grade of DP780 sheet steel, which had higher degree of softening, because, diffuse necking was not sufficient to reduce the BM cross-sectional area and hence the level of stress in the S-SC reached the UTS before the UTS was reached in the BM.

Identiferoai:union.ndltd.org:LACETR/oai:collectionscanada.gc.ca:OWTU.10012/4928
Date January 2009
CreatorsBurns, Trevor
Source SetsLibrary and Archives Canada ETDs Repository / Centre d'archives des thèses électroniques de Bibliothèque et Archives Canada
LanguageEnglish
Detected LanguageEnglish
TypeThesis or Dissertation

Page generated in 0.0024 seconds