Examining the effects of push assist on the formability of aluminum tubes

A cold forming process, bending influences the tube's properties and consequently its formability in subsequent hydroforming. Because plastic deformation in the cross section of a tube is not uniform during bending, the wall thickness decreases on the outside of the bend (in the tension zone) and increases on the inside of the bend (in the compression zone). This restricts the hydroformability of the tube and affects the wall thickness uniformity of the end product.

Several manufacturing factors—using the most appropriate lubricant, bend tools, and settings for bend parameters—help to minimize the bent tube's wall thickness variance. Another factor, applying axial force, or push assist, during the bending process also can minimize the wall thickness variance (see Figure 1). Push assist improves material flow in the forming zone and leads to less thinning of the wall on the outside of the bend, which is the most critical area when hydroforming prebent tubes.

Figure 1 A rotary draw bender with accessories (left) and a prebent tube hydroforming die insert for outside corner fill (right) were used to determine the effect of push assist on tubular components.

Experimental Parameters

Researchers bent and hydroformed aluminum tubes to determine the effects of push assist on the tube's formability. Welded tubes (3.5-millimeter wall thickness and 76.2-mm OD) from a 5000-series aluminum alloy were bent with low and high levels of push assist using a rotary draw bender to investigate the influence of different levels of push assist on the tubes' hydroformability.

The bent tubes then were internally pressurized inside a hydroforming die in a process that radially expands the outside of the tube (see Figure 1). The thickness and strain distributions of tested tubes were measured both after bending and after hydroforming.

Rotary Draw Bending With Push Assist. Rotary draw bending using push assist is achieved by applying both moment and transverse loads as well as axial force to a straight tube. An electrohydraulically driven rotary draw bender with digital controls for various bending parameters was used for the bending in the experiment.

Tubes were bent using a standard mandrel and dies (a flexible ball mandrel, wiper die, clamp die, and pressure die). Before each bending operation, the tubes were cut to the desired length and were marked with circle grids for subsequent circle grid analysis. The initial wall thickness of each tube was measured around its circumference to enable a rigorous comparison of the wall thicknesses after each forming process. Second, the tubes were bent to 90 degrees by applying two different levels of push assist, low and high. The ratio of bend radius to OD was 2.5 for every bend. A special lubricant was used for all of the bending operations. When the tubes were inserted into the bender, the weld seam was located in the neutral plane.

During the bending process, all relevant parameters, such as bend angle, bend torque, bend reaction forces, and push assist, were measured and controlled online. After bending, thickness distributions and strains were measured (using the circle grids) on both the outside and the inside of the bend, as well as around the circumferences of the tubes.

Figure 2 The initial wall thickness for tubes used in this experiment is 3.5 mm. A higher level of push assist during bending provides a greater thickness distribution on the inside and outside of the bend than a lower level of push assist does.

Prebent Tube Corner Fill Hydroforming. The prebent tube corner fill test involves a bent tube inside a die that has essentially the same geometry as that of the bent tube, except for an additional cavity either on the outside of the tube (for outside corner fill tests) or on the inside of the prebent tube (for inside corner fill tests). The outside prebent tube corner fill process was used for the hydroforming tests in this experiment.

Analyzing the Results

The tubes were analyzed in two stages. One group of tubes was analyzed after bending; the other group was examined after hydroforming.

Thickness Distribution After Bending. Bent tubes from each level of push assist were cut along the longitudinal axis and at 45 degrees to the bend. The tube thickness was measured lengthwise along the outside and inside of the bend and around the circumference (see Figure 2 and Figure 3). The thickness distribution was measured at 30 points along the inside and outside of the bend. The left side of the curves represents the clamped area of the bent tubes.

Figure 3 Thickness distribution around the circumference is greater for tubes bent with high push assist than those bent with low push assist.

• Figure 2 shows that the thickness of the straight portions of the tube at 0 and 90 degrees was different from the initial wall thickness. It therefore was concluded that the plastic deformation zone extends beyond the beginning and the end of the bend.

• The curves in Figure 2 and Figure 3 show that applying higher push assist resulted in more thickness on both the inside and outside of the bend as well as around the circumference of the tube.

Thickness and Strain Distributions After Hydroforming

Figure 4 The hydroforming process in this experiment caused an increase in wall thickness on both the inside and outside of the bend.

The effect of different levels of push assist on thickness distribution after hydroforming was determined by expanding the prebent tubes until the onset of necking. The hydroformed tubes then were cut along the longitudinal axis and at 45 degrees to the bend. The thickness was measured at 30 points along the inside and outside of the bend (see Figure 4) and at 25 points around the circumference (see Figure 5). Only necked tubes were measured; burst tubes were not included in the data.

• The distribution curves in Figure 4 have almost the same characteristics as those in Figure 2. Hydroformed tubes had less wall thickness along the outside of the bend compared to tubes that were bent but not hydroformed. The outside corner fill test resulted in smaller thickness changes on the inside of the bend as compared to the outside of the bend.

Figure 5 The thickness distribution around the circumference of tubes bent and hydroformed with the corner fill process shows no appreciable increase in wall thickness in the neutral zone, which is represented by the dotted lines that run between the tension and compression zones.

• The thickness measurements for Figure 3 and Figure 5 start at the inside of the bend (0 degrees) where the wall was thickest. The curves overlap each other in the neutral plane (90 degrees and 270 degrees). This indicates that applying a push assist in bending did not significantly influence the wall thickness in the neutral plane.

• Figure 6 shows the approximation curves for measured axial, thickness, and hoop strain around the circumference of the hydroformed tubes. The strain distribution curves show that a hydroformed tube prebent with higher push assist was less strained than one prebent with lower push assist. This implies that in hydroforming, a bent tube with higher push assist can endure more expansion (has greater formability) than a bent tube with lower push assist.

Conclusions

Figure 6 Strain distributions around the circumference of bent and hydroformed tubes show that tubes bent with higher push assist undergo less strain than tubes bent with low push assist.

The most significant effect of different push assist levels is on thickness distribution, particularly on the inside and outside of the bend. The tube wall thickness increases with an increase in the applied push assist in the bending operation. Thus, hydroformed hollow parts made from bent tubes with higher push assist always have thicker and more uniform wall thickness compared to parts made from tubes bent with low push assist.

In some cases, significantly raising the level of push assist could allow the use of a thinner-walled tube than otherwise would be possible. This would reduce the material usage and, therefore, the weight and cost of the finished part.

Ghafoor Khodayari, Ph.D., is senior project leader, metal forming, with Industrial Research & Development Institute (IRDI), 649 Prospect Blvd., P.O. Box 518, Midland, ON L4R 4L3, 705-526-2163, fax 705-526-2701, khodayari@irdi.com, www.irdi.com.

Michael Worswick, Ph.D., is a professor at the University of Waterloo, 200 University Ave. W., Waterloo, ON, N2L 3G1, 519-888-4567 ext. 5830, fax 519-888-6197, worswick@lagavulin.uwaterloo.ca, www.uwaterloo.ca.

Acknowledgments

The experimental work was performed at IRDI and the University of Waterloo, Ontario, Canada, in cooperation with the United States Automotive Materials Partnership (USAMP) Aluminum Hydroforming committee, which is part of the United States Council for Automotive Research (USCAR, www.uscar.org). The authors thank the research team at the University of Waterloo for providing the bender and conducting the bend tests. The authors extend a special thanks also for the invaluable assistance of Xin Wang, Ph.D., for measurements and H. Herbert for performing all hydroforming tests with the assistance of H. de Boer, G. Gravel, and B. Perkins.

This research was supported, in whole or in part, by the Department of Energy Cooperative Agreement No. DE-FC05-95OR22363 (and for work conducted after Dec. 31, 2002, DE-FC05-02OR22910). Such support does not constitute an endorsement by the Department of Energy of the views expressed herein.