Shear Fracture

Formability

Automotive product designers target small radii for springback control, sectional stiffness, packaging constraints, and design features. These small radii lead to new challenges as applications for AHSS grades continue to increase. One of these challenges is an increased sensitivity to crack formation in those designs with small die radius to material thickness (R/T) ratios. Cracks forming at small R/T in AHSS grades are known as shear fractures.

General forming limit curves or other press shop criteria cannot predict shear fractures, nor are they flagged when traditional approaches are used in forming simulation packages. However, these shear fractures do occur in die tryout. Shear fractures are another form of local formability failure  associated with multiphase AHSS such as DP and TRIP.

Shear fractures on AHSS may exhibit similarities to edge fracture, specifically the absence of necking prior to failure. This is in contrast with global formability failures, which are characterized by significant thinning near the fracture. Shear fractures occur almost immediately (within 1 mm of displacement) after reaching maximum load, meaning there is essentially no post-uniform elongation. This is contrary to the tensile behavior where significant post-uniform ductility remains prior to fracture.^K-9

Figures 1 and 2 highlight the appearance of a crack on the bend radius caused by shear fracture. No thinning is observed, which is consistent with failures limited by local formability concerns.

Figure 1: Shear fracture in DP780.^F-5

Figure 2: Shear fracture in DP980.^D-7

Figure 3 shows a typical shear fracture on a DP780 part viewed from different angles. The shear fracture occurred on the sharp radius on the left whereas the larger radius on the right experienced no failure. Depth of draw and draw bead configuration were the same on both sides of the draw panel. Restraining force was also similar on both sides of the blankholder. The significant variable was the die radius.

Figure 3: Different Views of a DP780 Part with Small R/T Leading to Shear Fracture at Bead Radius. An R/T = 1.67 led to shear fracture on the left side of the image, while the symmetric area on the right side had an R/T = 4.4, with no shear fracture even though it had the same depth of draw, draw bead configuration, and restraining force.^U-6

It is helpful to describe the characteristic differences between conventional tensile fracture and shear fracture, as shown in Figure 4. A conventional tensile fracture is called a Type I fracture, and has been the typical fracture type historically encountered. Type I fractures occur off the radius, and is preceded by necking or metal thinning. Successful prediction of this type of fracture occurs with conventional application of strain analysis and Forming Limit Diagrams. This is in contrast with shear fracture – categorized as a Type III fracture – which occurs within the die radius, with no thinning from necking (typical for local formability failures). Type II fractures occur at or near the tangent of the radius in metal drawn over the radius.

Figure 4: Schematic descriptions of different fracture types ranging from shear fracture to conventional tensile fracture.^S-19

Numerous studies show that radius to thickness ratios (R/T) are significant indicators of performance with respect to shear fracture on AHSS. This research led to the establishment of R/T ratio guidelines. While bend testing also categorizes products based on achievable R/T, the significant difference is that the ends are not restrained in a standard bend test. Shear fracture testing typically involves some type of restraining force, such as that seen in a Bending Under Tension test.

As with edge fracture, AHSS grades may be available at a similar strength but with improved minimum R/T ratio. Guidelines have been established for minimum R/T ratios based on bend test results as well as real world case studies. For DP340/590 and above, the R/T ratio should be at least 3T for product features such as embossments where there is relatively limited metal motion. Pulling these grades across a radius or through a draw bead under tension increases the minimum R/T ratio to at least 5T.

As strength levels increase, it is necessary to increase R/T ratios to avoid shear fractures. One study recommended a minimum R/T of 8 for DP800^W-4 while another has a critical R/T of at least 7T for DP980^S-19. Differences such as these are likely due to different test conditions such as the tension applied during bending, test speed and lubrication. Higher deformation rates (forming speed) and better lubrication tends to promote shear fracture and cause fracture on material close to the die radius.^S-19  A combination of high forming speed and back tension leads to DP590 and DP780 having a critical R/T value of 12, with DP980 having a critical R/T of over 16.^S-20

Shear fracture is sensitive to rolling direction. If the radius is running in the rolling direction, the bend will be transverse to the rolling direction which is the worst-case scenario when trying to avoid shear fracture. Understanding the directionality of minimum R/T ratios when designing the part to avoid shear fractures is therefore important. One study evaluated DP780, which discovered a critical minimum R/T of 5 to avoid shear fracture, although this varied with back tension and pulling speed. At all R/T ratios tested, the samples oriented so the bend radius was parallel to the rolling direction failed at lower stress than if the bend radius was perpendicular (transverse) to the rolling direction. Tighter R/T magnifies this effect, as does a higher strength grade. At 1.5 R/T, shear fracture with bends perpendicular to the rolling direction occurred when the stress reached 81% of the tensile strength, where in samples with the bend parallel to the rolling direction, failure occurred at 66% of the tensile strength.^S-21

Microstructure also plays a role. Phase distribution uniformity, fine microstructural phases, and a decrease in hardness ratio between martensite and ferrite all increase formability as measured by the smallest achievable R/T resulting in split-free panels.^{W-4, H-6}  These are the same factors that result in improved hole expansion values.

Forming Limit Curves are the limiting strain states based on the onset localized deformation, or necking. Generating conventional FLCs typically involves using a 100mm diameter hemispherical punch to deform 1mm to 2mm thick sheet steels, resulting in R/T ratios of 25 or higher. Many decades of use have shown that conventional FLC approaches can be used to assess part robustness and predict failure in areas having large R/T ratios and failure modes affected by global formability. These show up as Type I fractures.

When the R/T ratio is within the shear fracture limit but above the simple bending limit (no stretching imposed to bending), failure may occur in the radius or outside the radius, depending upon the tension level applied during forming. If the failure occurs inside the radius, the failure limit derived from the shear fracture tests should be used as the criterion to predict failure. Finally, when the R/T ratio is smaller than the limit from the simple bending test, the failure occurs in the radius.

In addition to large R/T, most Forming Limit Curves are generated from tests are conducted at a very low strain rate, maintaining an isothermal condition (no heat generated from deforming the sheets). While these conditions are consistent with those adopted for the simulations, they differ significantly from those encountered in industrial practice.

Ignoring the role of deformation-induced heating is one of the most significant reason for the limited success of shear fracture prediction in conventional forming simulation. Strain rates for stamping are typically 1,000x to 10,000x greater than the strain rates for tensile testing. Contact between the sheet metal and the tooling at tight radii also drives up the local temperature. Maximum temperature at the die-sheet interface of 70°C has been found on DP590, but 105°C for CP800.^F-6 Stamping DP780 leads to die temperatures of 180°C and blank temperatures of 108°C.^P-11 Stamping DP980 produces contact temperatures of over 200°C in certain conditions.^F-28

Using a simulation model that accounts for the thermal effects of forming and the associated response in the sheet metal, along with the traditional mechanical deformation response, has been shown to dramatically improve simulation accuracy to predict when shear fracture will occur.^{K-9, S-20, S-22}

Therefore, it is important for the part designer to design an appropriate die radius for a given AHSS product and forming conditions / processes used to manufacture the part. Forming simulation is an excellent tool to derive an appropriate die radius for the specified part and forming process, recognizing that the failure criterion in the simulation must incorporate all conditions and failure modes encountered, including shear fracture.

Key Points

1. • AHSS grades are at risk of crack formation in designs with small die radius to material thickness (R/T) ratios. These cracks are known as shear fracture.
   • The strains associated with shear fracture are below that which are associated with the Forming Limit Curve.
   • As with edge fractures, shear fractures are also a function of microstructure, strength level, and rolling direction. Consider these variables when designing parts and conducting die try-outs with prototype steel.