Simulation software helps build better drawn and extruded parts.
Scientific Forming Technologies Corp. Columbus, Ohio www.deform.com
Edited by: Leslie Gordon
Precision metalforming processes involve complex and strong interactions between the workpiece (material and shape) and forming equipment, as well as lubrication, temperature, and other manufacturing conditions. Many parameters affect metalforming, including: the number of operations, volumetric displacement, final part geometry, starting material size, forming equipment, and material behavior of the workpiece. A major problem until now has been the wide gap between the designer’s concept of a progression and actual shop trials. Unexpected metal flows can result in excessive loads on the parts, broken dies, and other production problems. Fortunately, simulation software can help close this gap and build better designs.
For example, simulation software helped a fastener manufacturer that had noted a small defect during the shop trial of an automotive part. Company engineers analyzed the forming process in our Deform simulation software, which reproduced the defect and helped the manufacturer understand its root cause. The simulation also revealed a severe lap that technicians had originally overlooked. Slicing open real parts revealed the defect, as predicted. From then on, engineers analyzed each station of a redesigned progressive process before performing any shop trials.
Pinpointing chevron cracks
In another example, in the past, automotive manufacturers were continually thwarted by axle-shaft-breakage problems. The axles were forward extruded, a process where the billet is pushed through a sizing die. There were no external defects, but internally, chevron cracks formed. This defect is the result of axial tensile stress during deformation along the centerline. To avoid 100% inspection, engineers at Chrysler Corp. derived mathematical expressions based on the die’s cone angles as well as process reduction and friction conditions to describe bursting phenomenon during extrusion. Results provided good flow, shave, and dead-zone-formation data that helped engineers reduce chevron cracks.
More recently, engineers have used the finite-element method (FEM) in conjunction with ductile fracture criteria to determine the likelihood of central bursting. Here, the parameter damage is the cumulative measure of deformation under tensile stress. Engineers evaluated several different damage scenarios in the Deform software. Various criteria expressed ductile fracture as a function of the material’s plastic deformation, accounting for geometry, damage, and strain within the workpiece. The upshot: fracture happens when the maximum damage value (MDV) of the material exceeds a particular critical damage value (CDV).
Under ideal conditions, strain distribution across a drawn or extruded component would be uniform but, in reality, this is rarely the case. In fact, subsurface deformation increases with larger die angles. This can cause high tensile stresses that lead to microvoiding and, ultimately, cracking. In the first forming operation, there may be high stresses in the component but low damage because, by definition, damage accumulates with deformation. Thus, a part does not usually fracture until the second or third draw or extrusion operation.
To determine a material’s CDV, one way to proceed is to perform compression and notched tensile tests until parts crack and then average the values. Next comes running a simulation using the geometry and process conditions of the experimental tests to calculate damage values. The software’s prediction of MDV at the instant of fracture is a good representation of the material’s CDV.
In one example, engineers used Deform software to simulate and compare two automotive-shaft designs manufactured using a double extrusion. The company in question built 500 steel (AISI-1024) shafts with a nominal 22.5° extrusion die angle. It also produced 500 units with a 15° die angle. There were chevron cracks in 1.2% of the shafts made from the 22.5° die and no cracks at all in the other shafts. Engineers checked these results in Deform. The simulation indicated a higher damage value for products extruded with the 22.5° die. The high damage value correlated well with the location of the chevron cracks.
Where damage levels are high, such as in a shearing operation, fracture takes place consistently. In processes well below the ductility limit, parts do not usually fracture. There is a narrow range between these two regions where the chance of cracking is probabilistic. A higher damage prediction means a greater likelihood of fracture.
Cold heading die failures
There are several ways dies can get damaged or even fail. Catastrophic die failure takes place when stress levels exceed the ultimate strength at temperature of the die material. Dies plastically deform when the stress exceeds the yield strength at the operating temperature. In addition, low-cycle fatigue can take place due to mechanical or thermal tensile stresses acting on the die as each part is produced. Dies can also wear due to the combination of friction and surface pressure.
The FEM software provides the option of a decoupled analysis to simulate metal flow on the die and establish the loads and pressure distributions. One example comes from a hex-head flange screw blank, formed on a traditional four-die progressive header. The die insert in the fourth station was averaging about 40,000 parts to failure. The company in this case wanted to move the part to a faster forming machine, but poor tool life would mean stopping the machine too often to change inserts.
The first step to improving the die design was to establish the root cause of the insert failure. Company engineers used Deform to simulate the forming of the hex-head flange-screw blank. They used the forming pressure to predict the stress and deflection of the die components. The assembly consisted of a one-piece die insert and a die case. In real-world tests, the highest load happened at the end of the stroke while the flange filled out.
The engineers then performed a subsequent die analysis at the step where they had observed the highest die loads. Results indicated that the stresses were tensile. The engineers used the analysis of the stress components to help redesign the assembly. The new design used two split inserts at the lower shoulder and an intermediate sleeve. Initially, the faster machine made over 400,000 parts before failure. After a few tweaks to the design, it eventually made 1,400,000 pieces — without failure.