Hydroforming dual-phase steel
December 7, 2004
The demand for lightweight components continues to be a primary driver in the automotive industry.
While hydroforming is considered to be one of the enabling technologies that can deliver lightweight components, the technology faces increasing competition from other lightweight solutions, such as magnesium castings. However, hydroforming can compete effectively with other lightweight alternatives if the hydroformed workpiece is an advanced material, such as a high-strength steel (HSS).
A good example of such a hydroforming application is an engine cradle (see introductory photo). Hydroforming the component from a tube results in a lighter component than a conventional cradle made from stamped parts.
This engine cradle initially was produced from a conventional roll formed and electrical resistance welded (ERW) tube. However, the tube exhibited excessive strain hardening introduced by the roll forming process and unfavorable mechanical properties caused by the heat input of the welding process.
Strain hardening introduced by the roll forming process was a result of strains in both the longitudinal and circumferential directions. The longitudinal strains are largely geometry-based and can be estimated with analytical software such as COPRA®. The circumferential strains are caused mainly by the calibration stands of the tube mill and can be estimated by modeling the tubemaking process in a finite element model. Previously made measurements and calculations on typical tubing showed that the strain was more than 6 percent.1
Annealing can be used to recover formability after strain hardening occurs in mild steel. However, dual-phase steels cannot be annealed because the process destroys the martensitic phase in the material. Therefore, the strains introduced by roll forming dual-phase steel tube are irreversible and result in a relatively poor tube for hydroforming.
The hardness profile of two tubes—one press formed and laser welded, the other roll formed and induction welded—reveals two significant hardness characteristics. First, the press formed blank's parent material is somewhat softer than the roll formed tube's parent material. Second, the hardness dip in the center of the HAZ is not as extreme in the tubular blank as it is in the roll formed tubes.
ERW introduces a large amount of heat into a local area on the circumference of the tube. The heat alters the microstructure. The melted region turns into a martensitic and bainitic structure after it cools. The melted region becomes relatively hard (430 HV, or hardness Vickers) and therefore has a high yield stress, which results in a strong but poorly formable region.
The heat-affected zone (HAZ) contains a slightly softer region (see Figure 1). This is the region that causes hydroforming problems.
It is assumed that the softened region has a slightly lower yield stress than the neighboring material (parent material and the weld seam). The hydroforming process encourages deformation in this weak region, which results in local thinning and finally failure in this area.
A typical tensile test curve for a 600-MPa dual-phase tube shows that the n value drops significantly at higher strain levels.
Dual-phase steels have a very high n value at the beginning of the stress/strain curve, which results in a rapid increase in the yield stress for small deformations, or strain hardening (see Figure 2). However, the n value at much higher strain values is much lower. A result of this lower n value is that the material is more sensitive to strain localization around pre-existing material defects or geometry constraints. In this case, the material defect is the softened HAZ.
During a free-expansion test, the tube was expanded under internal pressure over a free length at least three times the diameter. The radial expansion was measured during the test. The amount of radial expansion at failure then was used to quantify the formability of the tube.
The pressure expansion curves for the two tubes reveal that tubular blank (TuB) withstands much more pressure before failure than the conventional tube (ERW).
Free-expansion test results showed that failure takes place in the HAZ after a relatively low amount of expansion (less than 10 percent radial expansion under plane strain conditions). Figure 3 shows the pressure expansion curve recorded during the free-expansion test.
A tubular blank, which is formed by a press and welded by a laser, is more formable than a conventional roll formed ERW tube. Theory suggests that the press forming process should produce less strain hardening than the conventional process does. A finite element simulation of a press formed blank predicts strains of 2 percent, which is favorable compared to the 6 percent for a conventional tube. Furthermore, in press forming, the strain distribution is very consistent around the circumference of a tubular blank; it is uneven in conventional tube.
Production trials were performed to determine the actual characteristics of a dual-phase tubular blank. Like the microstructure of the weld seam in a conventional tube, the tubular blank's weld seam micro-structure had martensite and bainite. However, the microstructure and the hardness profile of the dual-phase blank weld seam revealed that the HAZ of the laser weld was narrower and the hardness dip was less extreme. Also, Figure 1 shows that the tubular blank parent material was not as hard as that of the conventional tube because the press forming process introduced less work hardening.
A free-expansion test shows two distinctly different failure modes. The roll formed ERW tube (top) is split along the weld seam. The tubular blank is ruptured with no discernible pattern, evidence that the weld is not significantly weaker than the parent material.
A free-expansion test showed that the failure occurred along the weld seam in conventional tube and in the parent material of the tubular blank (see Figure 4). Also, the tubular blank endured much more expansion before failure than the conventional tube did (see Figure 4). The blank reached a minimum expansion of 18.7 percent, which is nearly twice the 9.6 percent expansion of the conventional tube.
The information gathered so far suggests that a tubular blank is more formable than a conventional tube. However, this does not necessarily mean that a tubular blank's characteristics—a softer material with a smaller HAZ—will solve the hydroforming problems.
A basic finite element simulation reveals that the expected maximum strain level required under plane strain conditions is 20 percent.
Bending. Experience reveals two characteristics of tubular blanks. First, because tubular blanks are softer than conventional tubes, they have less springback. Second, the amount of springback varies significantly from batch to batch. This is related to the stress/strain characteristics (see Figure 2). It is therefore necessary to design a bending process to accommodate these properties.
Preforming. Preforming the tubular blank does not cause significant problems. However, because the springback behavior varies, the preforming step should be used to correct any geometry variations after bending.
Hydroforming. Hydroforming tubular blanks results in a dramatic drop in reject levels, which proves that hydroforming this engine cradle indeed is feasible when a tubular blank is the workpiece.
Because a strong correlation exists between the performance of a tubular blank and the material batch it comes from, continuous improvement does not stop with the bending/preforming/hydroforming process but should extend to the entire supply chain. This requires investigating every step in the steelmaking process, from slab casting to the hot and cold mills, postprocessing, and blanking. Such a rigorous and comprehensive understanding is necessary to reduce the variations from lot to lot of blanked material and thereby achieve better performance from dual-phase tubular blanks.
Maarten Kelder is a research and development project manager with Corus, Postbus 10000, 1970 CA IJmuiden, the Netherlands, 31-251-498588, fax 31-251-470432, email@example.com, www.corusgroup.com.
Kevin Edgar is a business development manager for Corus, IARC Building,University of Warwick, Coventry CV4 7AL, U.K., 44-2476-241231, fax 44-2476-241205, firstname.lastname@example.org, www.corusgroup.com.
Note: 1. R. Peeters, "3-D Modeling of the Roll Forming Process: Simulation, Setup, and Validation of the Roll Forming Process and Implementation in the Hydroforming Process Chain," Diss. University of the Twente 2001, Enschede, the Netherlands.