What Are Press Hardening Steel Grades and How Are They Made?
Press hardening is a special hot forming process, where the part is quenched in a forming die to receive its high hardness. It has been used in the automotive industry for over 40 years now.
The most common press hardening steel (PHS) is 22MnB5, a low carbon steel with Manganese-Boron alloying. Since it achieves a typical tensile strength of 1500 MPa after heat treatment, this material is mostly named as PHS1500 or CR1500T-MB (Cold Rolled, 1500 MPa typical Tensile strength, Manganese-Boron alloyed).B-14, V-19
The Direct Press Hardening Steel Process
The direct press hardening process involves heating the blanks over 900°C (1650°F) in an industrial furnace. The blanks are then removed from the furnace and quickly transferred to a forming die. The formed parts are not removed immediately. Instead, they are kept under force in a water-cooled tool set for quenching. With a 22MnB5 steel, the quenched part typically reaches 1500 MPa tensile strength. B-14
Coatings and Early Process Enhancements
Over the years, the first improvement on the 22MnB5 material was the application of an aluminum-silicon coating around early 2000’s. The addition of the coating did not affect its strength or elongation but improved the process as it eliminated scale formation during forming and quenching. AlSi coating however limited the process to the aforementioned direct process.B-14
Zinc Coatings and the Indirect Process
Some OEMs, especially in Europe, wanted to use Zn-based coatings for corrosion protection. The typical 22MnB5 is available with hot-dip galvanized (GI) and galvannealed (GA) coatings. Liquid metal embrittlement (LME) with Zn-based coatings is avoided with an indirect press hardening process. Forming is done at ambient temperatures, with the part subsequently heated and quenched in a press tool. These materials and techniques have been available since 2008.P-5
Figure 1 summarizes the most common 22MnB5 grades and coatings before and after the press hardening process.
Figure 1: 22MnB5 before and after the hot stamping and quenching cycle. The incoming material is similar to HSLA 380 or DP600 and can be cold formed if needed. After hot stamping, typical tensile strength is around 1500 MPa (re-created after: B-18, O-8, U-9).
Measuring Performance: VDA Standards
In 2010, German Association of the Automotive Industry (VDA) developed a new bending test (238-100) to evaluate energy absorbing capacity of PHS and PQS grades.V-4 This test gave a “bending angle” measurement, which replaced – to some extent – the use of “total elongation” value for energy absorbing calculations.
PHS1800 and Beyond
In 2011, a Japanese steel maker developed the first 1800 MPa (typical) tensile strength material. The material was AlSi coated with a modified, higher carbon, 30MnB5 chemistry and Nb alloying.H-33 One Japanese OEM applied the material in their bumper beams. The higher strength allowed using 1.4 mm thick PHS1800 material, instead of 1.6 mm PHS1500.M-28
Tailored Solutions for Specific Applications
A German OEM designed a B-pillar with a PHS1500 upper section, laser welded to a lower section formed from HSLA 340LA (340 MPa yield strength) steel for improved energy absorption.S-13 It was later found that typical HSLA steels not designed for hot stamping process may show significant variation in mechanical properties depending on the cooling rate.D-22
The Rise of Press Quenched Steels (PQS)
Steel companies subsequently developed “Press Quenched Steels” (PQS) which are also HSLA but have been specifically modified to achieve consistent material properties at varying cooling rates.H-69 PQS grades are not hardenable, even after hot stamping and quenching cycle.
In 2015, a steel company in Europe developed 20MnB8 with GI coating. Chemistry with slightly lower carbon and higher manganese allowed forming to be done at lower temperatures. The company developed a new process route where the heated blank is first pre-cooled to around 500°C (930°F) and then formed and quenched – solving any LME concerns. The grade’s mechanical properties are nearly identical to 22MnB5 after quenching.K-21 Thus, it may be called PHS1500, but to differentiate the material, they are typically named CR1500T-MB-PS (PS stands for Pre-cooled Stamping).V-9
Expanding Options: Composite Steels
In 2016, two different composite steels were developed for hot stamping. These are 3-layers, hot rolled cladded grades with PQS on the outer skin and PHS1500 in the core. These were 1200 and 1400 MPa tensile strength level grades, with significantly improved bendability. There is only a commercial name for this material. To avoid using those names, the grades may be referred to as PHS1200 Sandwich and PHS1400 Sandwich.L-68
Improving Formability and Weldability
Around 2016, steel makers started developing another PHS grade which has 1000-1200 MPa tensile strength after quenching. The grade had almost similar elongation with PHS1500 (almost 5%), but higher bendability (75° vs. 50°). These grades also have lower metallurgical notch effects when spot welded. The material may be named CR1100T-MB.V-9
Multi-Step and Air-Hardening Innovations
In 2019, a Japanese steel company developed “air-hardening” 22MnSiB9-5 alloy with GA coating. After hot forming and quenching, the material had mechanical properties almost equivalent to 22MnB5. Thus, this material can also be named as PHS1500. Since the material is air-hardenable, meaning that it hardens even at very low cooling rates, it can be hot formed in a multi-station servo-mechanical-transfer press [16]. The technique is then named as “multi-step hot forming”, with the grade referred to as CR1500T-MB-MS (the last MS stands for Multi-Step).V-9
Ultra High Strength and the VDA Naming System
Since 2020, steel companies rolled out 1900 or 2000 MPa (typical) tensile strength materials. These grades are now commonly referred to as CR1900T-MB. These grades are already available uncoated, AlSi coated or GA coated.V-9
In 2021, VDA published a new standard (239-500), which standardizes the naming, chemistry and mechanical properties of PHS and PQS grades. All the grades shown in Figure 2 (excluding the sandwich) are named based on this VDA standard.V-9
Figure 2: Stress-strain curves of commercially available PHS and PQS grades after quenching. (re-created after: B-18,Y-12, R-14).
The UniSteel Concept: One Alloy, Many Properties
In late 2021, researchers from China came up with a concept of using one chemistry (a modified 22MnB5) combined with different thermal processes to tailor the production of differing mechanical properties. Thus, it became possible to make a whole car from the same alloy, named as “UniSteel”.The different properties and their use areas are shown in Figure 3. The research was published in Science magazine.L-68
Figure 3: UniSteel concept: (a) material usage in a car body, (2) mechanical properties after heat treatments (re-created after R-14)
The Future: BQP and SIBORA Development
In 2025, a German consortium developed a new grade 37SiB6 and a new process route called Bainitizing, Quenching and Partitioning (BQP). Similar to the Chinese UniSteel concept, the new SIBORA (Silicon Boron with Retained Austenite) material can have various strength and elongation levels. Both the process and resulting mechanical properties are given in Figure 4. Different strength levels can be achieved by changing the bainitizing temperature between 360 and 460°C (680 and 860°F).O-15
Figure 4: (a) the BQP process (shown here is 360°C bainitizing temperature), (b) the mechanical properties after PHS or BQP processes (re-created after O-15).
We encourage you to visit this steel grades page to learn more about these grades available for Press Hardening, and head to this PHS and PQS Overview page for our PHS Primer. Thank you to Eren Billur for providing this information.
Thanks go to Eren Billur, Ph.D. for his contribution of this article to the AHSS Insights blog. Eren Billur is the Technical Manager of Billur Makine and Billur Metal Form, based in Ankara, Turkey, specializing in advanced sheet metal forming technologies. He holds a Ph.D. in Mechanical Engineering from The Ohio State University and has extensive experience in material characterization, sheet metal forming processes, and finite element simulations. Eren has contributed significantly to the understanding and application of hot stamping and advanced high-strength steels (AHSS) in the automotive industry. He is a regular columnist for MetalForming Magazine’s “Cutting Edge” column and has authored numerous scientific papers and book chapters, including contributions to the WorldAutoSteel AHSS Applications Guidelines. Passionate about advancing manufacturing knowledge, Eren provides engineering consulting, training, and simulation services worldwide, helping manufacturers optimize forming processes and successfully implement new-generation AHSS materials.
You’ll find this content as part of our page on Laser Blanking, but this month, we want to highlight it in our AHSS Insights blog. We thank Schuler North America for contributing this insightful case study.
Production of Class A quality and structural parts without a blanking die is possible, even for high-volume serial production. Laser blanking enables flexible, cost-effective, and sustainable manufacturing and is capable of reaching 45 parts per minute. DynamicFlow Technology (DFT) from Schuler provides highly productive, die-free blanking with lasers—directly from a continuously running steel coil. DFT combines the advantages of flexible laser cutting with the speed of conventional blanking.
Laser blanking technology addresses market challenges such as frequent die changes, the need to increase capacity, and improving plant floor utilization, material utilization, and downstream processes.
LASER BLANKING ELIMINATES FREQUENT DIE CHANGES
It is important to remember that there are no dies with laser blanking technology, and no dies mean no die changes. Overall Equipment Effectiveness (OEE) of up to 80% can be achieved with laser blanking technology. In fact, 4 to 6 million parts per year of various materials are produced with the help of DFT—including mild steel, high-strength steel, and advanced high-strength steel. Even processing press-hardening steels with an aluminum-silicon coating is possible with laser blanking. Surface and cutting quality can be maintained over this spectrum of steel grades. Laser blanking technology can even achieve effective small batch production of Class A outer body panels and structural parts typically up to 3mm thick.
LASER BLANKING INCREASES PLANT OUTPUT
Competitive high-speed and high-output results can be achieved in multiple ways with laser blanking technology. The above-ground coil-fed line, optimized for short setup time, can handle coils with material widths up to 2,150 mm, weighing up to 30 tons. The material transport is smooth and controlled, simplifying setup and leading to uninterrupted processing within the laser cell.
There are three highly dynamic and simultaneously moving laser cutting heads within the laser cell of these lines. These laser cutting heads cut the programmed blank contour from a continuously moving material coil. Cutting speeds can exceed 100 meters per minute. The material is protected against any process contamination throughout the cutting process by custom-designed cutting clearance and material transport.
Figure 2 reveals the high-speed and high-output results for outer body parts. Each part is measured by improved output per minute and hour to achieve an OEE of 80%. Laser blanking lines can achieve up to 45 parts per minute and reduce costs per blank.
Figure 2: High productivity achieved with laser blanking
LASER BLANKING IMPROVES MATERIAL UTILIZATION
Up to 90% of blank costs are determined by the material price. The most significant leverage would be to reduce scrap and save on materials. Schuler conducted research based on the production of 300,000 cars per year, at 350 kg per car and $1,000 USD per ton of steel to provide a realistic inside look at how much cost savings can be achieved with laser blanking. The result was $1 Million USD saved with just 1% of material savings. This is extremely significant as material costs keep increasing.
Laser blanking is the digital way to cut blanks. All that’s needed to create a blanking program is a drawing to be loaded and a material to be selected. The part-specific program can be created offline and modified at any time. It is designed to create optimal combinations of material utilization and output—resulting in a high level of flexibility that significantly reduces development time for optimal blanks while also allowing for need-based production. This makes production planning easier, and it also opens the door to continuous contour optimizations for the forming process. Additionally, laser cutting does not require any gaps between individual parts due to smart nesting capabilities that cannot be achieved in comparison to die nesting or flatbed laser nesting. The combined smart, flexible nesting functions unlock new potential for material savings. Manufacturers can optimize individual blanks and eliminate the separating strip or connection bridges. Scrap savings in the forming process can also be achieved as there are no geometric restrictions due to cutting dies, and manufacturers can continuously optimize or adapt parts.
Figure 3 showcases the comparison of die nesting (the two graphics on the left) versus a laser-optimized blank contour and material savings via smart, laser blanking line nesting (the two images on the right).
Figure 3: Die nesting (left) compared with laser-optimized blank contours highlighting potential material savings (right)
Overall, laser blanking lines can have an equivalent throughput to conventional blanking lines, but laser blanking lines can achieve up to 10% greater material utilization.
You can read the full Case Study, including how laser blanking reduces infrastructure costs and improves downstream processes here: Laser Blanking Case Study
Schuler will present laser blanking technology, along with a variety of digital tools that create the “Press Shop of the Future” at FABTECH Chicago 2023 (booth # D41306). Tiago Vasconcellos, Sales Director at Schuler North America, will present “How Smart is Your Press Shop?” during FABTECH’s Educational Conference. The presentation will use The Smart Press Shop, a newly formed joint venture between Porsche and Schuler, as an exemplary case study for smart manufacturing standards. Attendees will discover innovative and practical ways to incorporate digitalization into production and become a state-of-the-art stamping facility directly from Schuler.
Schuler offers customized cutting-edge technology in all areas of forming—from the networked press to press shop planning. In addition to presses, Schuler’s products include automation, dies, process know-how, and service for the entire metalworking industry. Schuler’s Digital Suite brings together solutions for networking forming technology and is continuously being developed to further improve line productivity and availability. Schuler customers include automotive manufacturers and suppliers, as well as companies in the forging, household appliance, and electrical industries. Schuler presses are minting coins for more than 180 countries. Founded in 1839 at the Göppingen, Germany headquarters, Schuler has approximately 5,000 employees at production sites in Europe, China and the Americas, as well as service companies in more than 40 countries. The company is part of the international technology group ANDRITZ.
Schuler’s global portfolio of world-renowned brands include BCN (Bliss Clearing Niagara) Technical Services, Müller Weingarten, Beutler, Umformtechnik Erfurt, SMG Pressen, Hydrap Pressen, Wilkins & Mitchell, Bêché, Spiertz Presses, Farina Presse, Liebergeld, Peltzer & Ehlers, Schleicher, and Sovema Group.
Schuler North America (Schuler), headquartered in Canton, Michigan, is the North American subsidiary of Schuler Group. Schuler provides new equipment, spare parts, and a portfolio of lifecycle services for all press systems—including preventative maintenance, press shop design and optimization, turnkey installations, retrofits for existing systems, and localized production and service. Schuler’s best-in-class position in the metalworking and materials industry serves automotive manufacturers and tier suppliers, as well as home appliance, electronics, forging, and other industries.
Roll Forming takes a flat sheet or strip and feeds it longitudinally through a mill containing several successive paired roller dies, each of which incrementally bends the strip into the desired final shape. The incremental approach can minimize strain localization and compensate for springback. Therefore, roll forming is well suited for generating many complex shapes from Advanced High-Strength Steels, especially from those grades with low total elongation, such as martensitic steel. The following video, kindly provided by Shape Corp.S-104, highlights the process that can produce either open or closed (tubular) sections.
The number of pairs of rolls depends on the sheet metal grade, finished part complexity, and the design of the roll-forming mill. A roll-forming mill used for bumpers may have as many as 30 pairs of roller dies mounted on individually driven horizontal shafts.A-32
Roll forming is one of the few sheet metal forming processes requiring only one primary mode of deformation. Unlike most forming operations, which have various combinations of forming modes, the roll-forming process is nothing more than a carefully engineered series of bends. In roll forming, metal thickness does not change appreciably except for a slight thinning at the bend radii.
Roll forming is appropriate for applications requiring high-volume production of long lengths of complex sections held to tight dimensional tolerances. The continuous process involves coil feeding, roll forming, and cutting to length. Notching, slotting, punching, embossing, and curving combine with contour roll forming to produce finished parts off the exit end of the roll-forming mill. In fact, companies directly roll-form automotive door beam impact bars to the appropriate sweep and only need to weld on mounting brackets prior to shipment to the vehicle assembly line.A-32 Figure 1 shows an example of automotive applications that are ideal for the roll-forming process.
Figure 1: Body components that are ideally suited for roll-forming.
Roll forming can produce AHSS parts with:
Steels of all levels of mechanical properties and different microstructures.
Small radii depending on the thickness and mechanical properties of the steel.
Reduced number of forming stations compared with lower strength steel.
However, the high sheet-steel strength means that forces on the rollers and frames in the roll-forming mill are higher. A rule of thumb says that the force is proportional to the strength and thickness squared. Therefore, structural strength ratings of the roll forming equipment must be checked to avoid bending of the shafts. The value of minimum internal radius of a roll formed component depends primarily on the thickness and the tensile strength of the steel (Figure 2).
Figure 2: Achievable minimum r/t values for bending and roll forming for different strength and types of steel.S-5
As seen in Figure 2, roll forming allows smaller radii than a bending process. Figure 3 compares CR1150/1400-MS formed with air-bending and roll forming. Bending requires a minimum 3T radius, but roll forming can produce 1T bends.S-30
Figure 3: CR1150/1400-MS (2 mm thick) has a minimum bend radius of 3T, but can be roll formed to a 1T radius.S-30
The main parameters having an influence on the springback are the radius of the component, the sheet thickness, and the strength of the steel. As expected, angular change increases for increased tensile strength and bend radius (Figure 4).
Figure 4: Angular change increases with increasing tensile strength and bend radii.A-4
Figure 5 shows a profile made with the same tool setup for three steels at the same thickness having tensile strength ranging from 1000 MPa to 1400 MPa. Even with the large difference in strength, the springback is almost the same.
Figure 5: Roll formed profile made with the same tool setup for three different steels. Bottom to Top: CR700/1000-DP, CR950/1200-MS, CR1150/1400-MS.S-5
Citation A-33 provides guidelines for roll forming High-Strength Steels:
Select the appropriate number of roll stands for the material being formed. Remember the higher the steel strength, the greater the number of stands required on the roll former.
Use the minimum allowable bend radius for the material in order to minimize springback.
Position holes away from the bend radius to help achieve desired tolerances.
Establish mechanical and dimensional tolerances for successful part production.
Use appropriate lubrication.
Use a suitable maintenance schedule for the roll forming line.
Anticipate end flare (a form of springback). End flare is caused by stresses that build up during the roll forming process.
Recognize that as a part is being swept (or reformed after roll forming), the compression of metal can cause sidewall buckling, which leads to fit-up problems.
Do not roll form with worn tooling, as the use of worn tools increases the severity of buckling.
Do not expect steels of similar yield strength from different steel sources to behave similarly.
Do not over-specify tolerances.
Guidelines specifically for the highest strength steelsA-33:
Depending on the grade, the minimum bend radius should be three to four times the thickness of the steel to avoid fracture.
Springback magnitude can range from ten degrees for 120X steel (120 ksi or 830 MPa minimum yield strength, 860 MPa minimum tensile strength) to 30 degrees for M220HT (CR1200/1500-MS) steel, as compared to one to three degrees for mild steel. Springback should be accounted for when designing the roll forming process.
Due to the higher springback, it is difficult to achieve reasonable tolerances on sections with large radii (radii greater than 20 times the thickness of the steel).
Rolls should be designed with a constant radius and an evenly distributed overbend from pass to pass.
About 50 percent more passes (compared to mild steel) are required when roll forming ultra high-strength steel. The number of passes required is affected by the number of profile bends, mechanical properties of the steel, section depth-to-steel thickness ratio, tolerance requirements, pre-punched holes and notches.
Due to the higher number of passes and higher material strength, the horsepower requirement for forming is increased.
Due to the higher material strength, the forming pressure is also higher. Larger shaft diameters should be considered. Thin, slender rolls should be avoided.
During roll forming, avoid undue permanent elongation of portions of the cross section that will be compressed during the sweeping process.
Roll forming is applicable to shapes other than long, narrow parts. For example, an automaker roll forms their pickup truck beds allowing them to minimize thinning and improve durability (Figure 6). Reduced press forces are another factor that can influence whether a company roll forms rather than stamps truck beds.
In addition, increasing the number of passes has been shown to be an effective technique to lower residual stresses and therefore improve dimensional accuracy. Multiple bending sequences, especially in the transverse direction also improve dimensional accuracy, providing the steel has sufficient inherent formability to accommodate the additional bends.X-4
Figure 6: Roll Forming can replace stamping in certain applications.G-9
Traditional two-dimensional roll forming uses sequential roll stands to incrementally change flat sheets into the targeted shape having a consistent profile down the length. Advanced dynamic roll forming incorporates computer-controlled roll stands with multiple degrees of freedom that allow the finished profile to vary along its length, creating a three-dimensional profile. The same set of tools create different profiles by changing the position and movements of individual roll stands. In-line 3D profiling expands the number of applications where roll forming is a viable parts production option.
One such example are the 3D roll formed tubes made from 1700 MPa martensitic steel for A-pillar / roof rail applications in the 2020 Ford Explorer and 2020 Ford Escape (Figure 7). Using this approach instead of hydroforming created smaller profiles resulting in improved driver visibility, more interior space, and better packaging of airbags. The strength-to-weight ratio improved by more than 50 percent, which led to an overall mass reduction of 2.8 to 4.5 kg per vehicle.S-104
Figure 7: 3D Roll Formed Profiles in 2020 Ford Vehicles using 1700 MPa martensitic steel.S-104
Roll forming is no longer limited to producing simple circular, oval, or rectangular profiles. Advanced cross sections such as those shown in Figure 8 provided by Shape Corporation highlight some profile designs aiding in body structure stiffness and packaging space reductions.
Figure 8: Roll forming profile design possibilities. Courtesy of Shape Corporation.
In summary, roll forming can produce AHSS parts with steels of all levels of mechanical properties and different microstructures with a reduced R/T ratio versus conventional bending. All deformation occurs at a radius, so there is no sidewall curl risk and overbending works to control angular springback.
Case Study: How Steel Properties Influence the Roll Forming Process
Optimizing the use of roll forming requires understanding how the sheet metal behaves through the process. Making a bend in a roll formed part occurs only when forming forces exceed the metal’s yield strength, causing plastic deformation to occur. Higher strength sheet metals increase forming force requirements, leading to the need to have larger shaft diameters in the roll forming mill. Each pass must have greater overbend to compensate for the increasing springback associated with the higher strength.
Figure 9 provides a comparison of the loads on each pass of a 10-station roll forming line when forming either AISI 1020 steel (yield strength of 350 MPa, tensile strength of 450 MPa, elongation to fracture of 15%) or CR1220Y1500T-MS, a martensitic steel with 1220 MPa minimum yield strength and 1500 MPa minimum tensile strength.
Figure 9: Loads on each pass of a roll forming line when forming either AISI 1020 steel (450 MPa tensile strength) or a martensitic steel with 1500 MPa minimum tensile strength. Courtesy of Roll-Kraft.
Although a high-strength material requires greater forming loads, grades with higher yield strength can resist stretching of the strip edge and prevent longitudinal deformations such as twisting or bow. Flange edge flatness after forming either AISI 1020 or CR1220Y1500T-MS is presented in Figure 10.
Figure 10: Simulation results showing flange edge flatness of a) AISI 1020 and b) CR1220Y1500T-MS. Assumptions for the simulation: AISI 1020 yield strength = 350 MPa; CR1220Y1500T yield strength = 1220 MPa. Higher yield strength leads to better flatness.
Force requirements for piercing operations are a function of the sheet tensile strength. High strains in the part design exceeding uniform elongation resulting from loads in excess of the tensile strength produces local necking, representing a structural weak point. However, assuming the design does not produce these high strains, the tensile strength has only an indirect influence on the roll forming characteristics.
Yield strength and flow stress are the most critical steel characteristics for roll forming dimensional control. Receiving metal with limited yield strength variability results in consistent part dimensions and stable locations for pre-pierced features.
Flow stress represents the strength after some amount of deformation, and is therefore directly related to the degree of work hardening: starting at the same yield strength, a higher work hardening steel will have a higher flow stress at the same deformation.
Two grades are shown in Figure 11: ZE 550 and CR420Y780T-DP. ZE 550, represented by the red curve, is a recovery annealed grade made by Bilstein having a yield strength range of 550 to 625 MPa and a minimum tensile strength of 600 MPa, while CR420Y780T-DP, represented by the blue curve, is a conventional dual phase steel with a minimum yield strength of 420 MPa and a minimum tensile strength of 780 MPa. For the samples tested, ZE 550 has a yield strength of approximately 565 MPa, where that for CR420Y780T-DP is much lower at about 485 MPa. Due to the higher work hardening (n-value) of the DP steel, its flow stress at 5% strain is 775 MPa, while the flow stress for the HSLA grade at 5% strain is 620 MPa.
In conventional stamping operations, this work hardening is beneficial to delay the onset of necking. However, use of dual-phase steels and other grades with high n-value can lead to dimensional issues in roll-formed parts. Flow stress in a given area is a function of the local strain. Each roll station induces additional strain on the overall part, and strains vary within the part and along the edge. This strength variation is responsible for differing springback and edge wave across a roll-formed part.
Unlike conventional stamping, grades with a high yield/tensile ratio where the yield strength is close to the tensile strength are better suited to produce straight parts via roll forming.
Figure 11: Stress-strain curves for CR420Y780T-DP (blue) and ZE 550 (red). See text for description of the grades.
Total elongation to fracture is the strain at which the steel breaks during tensile testing, and is a value commonly reported on certified metal property documents (cert sheets). As observed on the colloquially called “banana diagram”, elongation generally decreases as the strength of the steel increases.
For lower strength steels, total elongation is a good indicator for a metal’s bendability. Bend severity is described by the r/t ratio, or the ratio of the inner bend radius to the sheet thickness. The metal’s ability to withstand a given bend can be approximated by the tensile test elongation, since during a bend, the outermost fibers elongate like a tensile test.
In higher strength steels where the phase balance between martensite, bainite, austenite, and ferrite play a much larger role in developing the strength and ductility than in other steels, bendability is usually limited by microstructural uniformity. Dual phase steels, for example, have excellent uniform elongation and resistance to necking coming from the hardness difference between ferrite and martensite. However, this large hardness difference is also responsible for relatively poor edge stretchability and bendability. In roll forming applications, those grades with a uniform microstructure will typically have superior performance. As an example, refer to Figure 11. The dual phase steel shown in blue can be bent to a 2T radius before cracking, but the recovery annealed ZE 550 grade with noticeably higher yield strength and lower elongation can be bent to a ½T radius.
Remember that each roll forming station only incrementally deforms the sheet, with subsequent stations working on a different region. Roll formed parts do not need to use grades associated with high total elongation, especially since these typically have a bigger gap between yield and tensile strength.
Coil Shape Imperfections Influencing Roll Forming
Along with the mechanical properties of steel, physical shape attributes of the sheet or coil can influence the roll forming process. These include center buckle, coil set, cross bow, and camber. Receiving coils with these imperfections may result in substandard roll formed parts.
Flatness is paramount when it comes to getting good shape on roll formed parts. Individual OEMs or processors may have company-specific procedures and requirements, while organizations like ASTM offer similar information in the public domain. ASTM A1030/A1030M is one standard covering the practices for measuring flatness, and specification ASTM A568/A568M shows methods for characterizing longitudinal waves, buckles, and camber.
Center buckle (Figure 12), also known as full center, is the term to describe pockets or waves in the center or quarter line of the strip. The height of pocket varies from 1/6” to 3/4”. Center buckle occurs when the central width portion of the master coil is longer than the edges. This over-rolling of the center portion might occur when there is excessive crown in the work roll, build-up from the hot strip mill, a mismatched set of work rolls, improper use of the benders, or improper rolling procedures. A related issue is edge buckle presenting as wavy edges, originating when the coil edges are longer than the central width position.
Figure 12: Coil shape imperfection: Center Buckle
Coil set (Figure 13a), also known as longitudinal bow, occurs when the top surface of the strip is stretched more than the bottom surface, causing a bow condition parallel with the rolling direction. Here, the strip exhibits a tendency to curl rather than laying flat. To some extent, coil set is normal, and easy to address with a leveler. Severe coil set may be induced by an imbalance in the stresses induced during rolling by the thickness reduction work rolls. Potential causes include different diameters or surface speeds of the two work rolls, or different frictional conditions along the two arcs of contact.
Crossbow (Figure 13b) is a bow condition perpendicular to the rolling direction, and arcs downward from the high point in the center position across the width of the sheet. Crossbow may occur if improper coil set correction practices are employed.
Figure 13: Coil shape imperfections: A) Coil set and B) Crossbow A-30
Camber (Figure 14) is the deviation of a side edge from a straight edge, and results when one edge of the steel is elongated more than the other during the rolling process due to a difference in roll diameter or speed. The maximum allowable camber under certain conditions is contained within specification ASTM A568/A568M, among others.
Figure 14: Coil shape imperfection: Camber
Coil shape imperfections produce residual stresses in the starting material. These residual stresses combined with the stresses from forming lead to longitudinal deviations from targeted dimensions after roll forming. Some of the resultant shapes of roll formed components made from coils having these issues are shown in Figure 15. Leveling the coil prior to roll forming may address some of these shape concerns, and has the benefit of increasing the yield strength, making a more uniform product.
Figure 15: Shape deviations in roll formed components initiating from incoming coil shape issues: a) camber b) longitudinal bow c) twist d) flare e) center wave (center buckle) f) edge wave. H-66
Roll Stamping
Traditional roll forming creates products with essentially uniform cross sections. A newer technique called Roll Stamping enhances the ability to create shapes and features which are not in the rolling axis.
Using a patented processA-48, R-9, forming rolls with the part shape along the circumferential direction creates the desired form, as shown in Figure 16.
Figure 16: Roll Stamping creates additional shapes and features beyond capabilities of traditional roll forming. A-48
This approach can be applied to a conventional roll forming line. In the example of an automotive door impact beam, the W-shaped profile in the central section and the flat section which attaches to the door inner panel are formed at the same time, without the need for brackets or internal spot welds (Figure 17). Sharp corner curvatures are possible due to the incremental bending deformation inherent in the process.
Figure 17: A roll stamped door part formed on a conventional roll forming line eliminates the need for welding brackets at the edges.R-9
A global automaker used this method to replace a three-piece door impact beam made with a 2.0 mm PHS-CR1500T-MB press hardened steel tube requiring 2 end brackets formed from 1.4 mm CR-500Y780T-DP to attach it to the door frame, shown in Figure 18. The new approach, with a one-piece roll stamped 1.0 mm CR900Y1180T-CP complex phase steel impact beam, resulted in a 10% weight savings and 20% cost savings.K-58This technique started in mass production on a Korean sedan in 2017, a Korean SUV in 2020, and a European SUV in 2021.K-58
Figure 18: Some Roll Stamping Automotive Applications.K-58
Thanks are given to Brian Oxley, Product Manager, Shape Corporation, for his contributions to the Roll Forming Case Study and Coil Shape Imperfections section. Brian Oxley is a Product Manager in the Core Engineering team at Shape Corp. Shape Corp. is a global, full-service supplier of lightweight steel, aluminum, plastic, composite and hybrid engineered solutions for the automotive industry. Brian leads a team responsible for developing next generation products and materials in the upper body and closures space that complement Shape’s core competency in roll forming. Brian has a Bachelor of Science degree in Material Science and Engineering from Michigan State University.
During a tensile test, the elongating sample leads to a reduction in the cross-sectional width and thickness. The shape of the engineering stress-strain curve showing a peak at the load maximum (Figure 1) results from the balance of the work hardening which occurs as metals deform and the reduction in cross-sectional width and thickness which occurs as the sample dogbone is pulled in tension. In the upward sloping region at the beginning of the curve, the effects of work hardening dominate over the cross-sectional reduction. Starting at the load maximum (ultimate tensile strength), the reduction in cross-sectional area of the test sample overpowers the work hardening and the slope of the engineering stress-strain curve decreases. Also beginning at the load maximum, a diffuse neck forms usually in the middle of the sample.
Figure 1: Engineering stress-strain curve from which mechanical properties are derived.
The elongation at which the load maximum occurs is known as Uniform Elongation. In a tensile test, uniform elongation is the percentage the gauge length elongated at peak load relative to the initial gauge length. For example, if the gauge length at peak load measures 61 mm and the initial gauge length was 50mm, uniform elongation is (61-50)/50 = 22%.
Schematics of tensile bar shapes are shown within Figure 1. Note the gauge region highlighted in blue. Up though uniform elongation, the cross-section has a rectangular shape. Necking begins at uniform elongation, and the cross section is no longer rectangular.
Theory and experiments have shown that uniform elongation expressed in true strain units is numerically equivalent to the instantaneous n-value.
Deformation Prior to Uniform Elongation is Not Uniformly Distributed
Conventional wisdom for decades held that there is a uniform distribution of strains within the gauge region of a tensile bar prior to strains reaching uniform elongation. Traditional extensometers calibrated for 50-mm or 80-mm gauge lengths determine elongation from deformation measured relative to this initial length. This approach averages results over these spans.
The advent of Digital Image Correlation (DIC) and advanced processing techniques allowed for a closer look. A studyS-113 released in 2021 clearly showed that each of the 201 data points monitored within a 50 mm gauge length (virtual gauge length of 0.5-mm) experiences a unique strain evolution, with differences starting before uniform elongation.
Figure 2: Strain evolution of the 201 points on the DP980 tensile-test specimen exhibits divergence beginning before uniform elongation—counter to conventional thinking.S-113
As more companies aim to reduce their product’s time to market, research and design engineers have begun integrating predictive modeling into their process. These models, whether finite element based or artificial intelligence based, all rely on quality mechanical testing results. Companies within the automotive industry have seen that accurately predicting large scale tests such as crashworthiness trials can greatly expedite the time it takes to get their products to market. One of the more important details in predicting these expensive and time-consuming tests is to understand how the materials within the design are affected by the higher rates of deformation or their strain rates.
Traditional standardized tensile tests have been used for over a century – ASTM E8 was first approved in 1924. Testing laboratories using standards like ASTM E8, ISO 6892-1, and JIS Z-2241 produce repeatable and reproducible mechanical properties for metals undergoing tensile deformation, but each of these standards requires the test to be run at a speed orders of magnitude lower than those occurring during events like sheet metal forming and automotive crash. A tensile test run according to ASTM E8 to obtain the yield properties of a metal is run at a strain rate of 0.00035 strain per second (0.00035/s). For comparison, a stamping process has strain rates on the order of 1 to 10 strain per second, and an automotive crash can have strain rates up to 1,000 strain per second (Figure 1).
Figure 1. Strain rates of different events.
Historically, no guidelines have been available as to the testing method to obtain high strain rate mechanical properties. Decisions on specimen dimensions, measurement devices, and other important issues which are critical to the quality of testing results were made within each individual laboratory. As a result, data from different laboratories were often not directly comparable. A WorldAutoSteel committee evaluated various procedures, conducted several round-robins, and developed a recommended procedure, which evolved into what are now the first two parts of ISO 26203, linked below.
Published standards addressing tensile testing at high strain rates include:
Steel alloys typically possess positive strain rate sensitivity, or m-value when tested at ambient temperature, meaning that strength increases with strain rate. This has benefits related to improved crash energy absorption.
The specific response as a function of strain rate is grade dependent. Some grades get stronger and more ductile as the strain rate increases (left image in Figure 2), while other grades see primarily a strength increase (right image in Figure 2). Increases are not linear or consistent with strain rate, so simply scaling the response from conventional quasi-static testing does not work well. Strain hardening (n-value) also changes with speed in some grades, as suggested by the different slopes in the right image of Figure 2. Accurate crash models must also consider how strain rate sensitivity impacts bake hardenability and the magnitude of the TRIP effect, both of which are further complicated by the strain levels in the part from stamping.
Figure 2: Two steels with different strength/ductility response to increasing strain rate.A-7
Importance of Proper Testing Equipment
Knowing that the strain rate directly affects mechanical properties, many research test laboratories have run tensile tests using the same specimen geometry and machine as standardized but have increased the speed the machine runs during the test. This typically allows for tests to be performed at strain rates of up to 0.1 strain per second. From this data, an extrapolated curve can be fit to approximate the properties of the materials at higher strain rates. One model used to predict the increase in strength of a material deformed at a higher strain rate is the Cowper-Symonds model:
Equation 1
where σd is the strength of the material at a strain rate έ, σs is the strength of the material at a theoretical strain rate of zero strain per second, and C and p are model parameters.
Figure 3 shows two best fit curves for this Cowper-Symonds model to the tensile strengths of a cold rolled steel across a range of strain rates. The first considers only data that can be obtained using traditional tensile testing machines (strain rates less than 0.1 strain per second) while the second uses data up to 2,000 strain per second. Both models have minimal errors at strain rates below 0.1 strain per second, but as the limited model begins to extrapolate data beyond this strain rate regime, the associated error begins to grow exponentially causing large errors at the strain rates typical in crashes.
Figure 3. Example of extrapolation of tensile strength vs strain rate data using a Cowper-Symonds model. The dashed curve is an extrapolation based only on data acquired using traditional tensile testing machines, where strain rates are less than 0.1 strain per second. The solid curve is the extrapolation when considering data from equipment capable of achieving 2,000 strain per second.
Testing Methods and Equipment
The biggest obstacle to measuring how a material responds to different strain rates is that it requires several different types of equipment. This is due to the need to run tests at up to six or seven different orders of magnitude to fully characterize the material. Figure 4 shows which mechanical testing equipment is most used to perform tensile tests based on the strain rate of the test. A broader range of testing methods for more strain rates can be found in the ASM Handbook, Volume 8: Mechanical Testing and Evaluation.A-88 The limits for the strain rate range that each type of test equipment can achieve varies based on the design and attributes of each specific machine as well as the specimen geometry used in testing. In tensile and compression testing, going from a longer specimen to a shorter specimen allows for a specific machine to increase its upper limit on strain rates.
Figure 4. Testing equipment most used to test materials at strain rates between 0.0001 strain per second and 10,000 strain per second in uniaxial tension.
Modified Servo-Hydraulic Machines
Modified servo-hydraulic testing machines are specifically engineered to characterize the dynamic mechanical properties of materials at high strain rates, often reaching strain rates up to 500 strain per second. Unlike conventional servo-hydraulic machines, which used closed-loop control for precise lower-speed testing, these modified systems incorporate design features that overcome the limitations of standard hydraulic controls at higher speeds. Many of these machines utilize extreme high flow valves along with slack adapters. The higher flow valves allow for higher accelerations while the slack adapters decouple the actuator from the specimen while the actuator accelerates to a desired test speed.
Split Hopkinson Pressure Bars
A split Hopkinson pressure bar (sometimes referred to as Kolsky bar or simply SHPB) is an impact-based device that is designed to characterize the dynamic mechanical properties of materials at strain rates above 100/s. The SHPB system uses a striker rod to generate a stress wave which induces plastic deformation in a specimen placed between two elastic bars. The stress wave generated by the striker bar in the first elastic bar (called the incident bar) is measured by a strain gauge fixed at the midpoint of the incident bar. When the stress wave reaches the specimen, part of the stress wave is transmitted through the specimen into the second elastic bar (called the transmitted bar) where it is measured by a second strain gauge. The rest of the stress wave is reflected off the specimen and returns down the incident bar to be re-measured by the strain gauge. Figure 5 shows how the stress waves propagate through a SHPB system during a compression test. There are various options to modify the compression SHPB setup to run a tensile test, but the most common is to replace the striker bar with a tube that slides on the incident bar where it impacts a flange on the end of the incident bar causing a tensile stress wave to propagate towards the specimen instead of a compression wave.
Figure 5. Animation showing how a stress wave propagates through a
split Hopkinson pressure bar system during a high strain rate compression test.
High Strain Rate Testing Challenges
Many other challenges complicate testing materials at higher strain rates. Three of the major challenges are
Challenges of measuring strain at high speeds
Challenges of accounting for inertial effects
Challenges of accounting for adiabatic heating
Strain Measurements in High Strain Rate Testing
During standardized mechanical testing, clip-on extensometers and deflectometers provide excellent extension measurements. As the speed of the test increases, the mass of the extensometer inhibits its use due to slippage of the contact points or interference with the specimen, both of which lead to erroneous test results and potential damage to the device. During high strain rate tests, the simplest means of measuring specimen displacements is to derive them based on the stress waves from the test. A detailed derivation of this method can be found here. This method works well for compression testing, but during tensile testing, events such as slippage of the specimen within the grips or deformation of the radius section of the specimen add to the displacements of the test. These additional displacements overshoot the tensile strain of a specimen during the test. This has led to the nearly universal adoption of optical strain measurements for high strain rate tension tests.
The most common optical method is digital image correlation (DIC). DIC correlates a series of images taken during the deformation of a specimen and calculates the corresponding strains of the specimen. It does this by tracking a black-and-white speckle pattern painted on the specimen’s surface which creates a series of high-contrast features as shown in Figure 6.
Figure 6. Two examples of 2D digital image correlation (DIC) showing the true equivalent strain fields during tensile tests performed at 0.1 strain per second and 1000 strain per second.
When testing round or more complex specimens, two cameras are required to track the surface of the specimen in three-dimensional space. This is referred to as 3D DIC, and it often requires more rigorous calibration for use in testing due to its multi-camera complexity. Alternatively, there are a series of one-dimensional options for strain measurements. The simplest utilizes a high-speed line scan camera to measure the displacement of a specimen along the test direction. While the two- and three-dimensional approaches bring in more data, the one-dimensional approach has been shown to provide excellent resolution at a lower adoption cost.Z-16
Accounting for Inertia in High Strain Rate Testing
At strain rates around one to ten strain per second, inertial effects can begin to complicate the interpretation of test results. These effects can be broken into two varieties: inertia of the specimen being tested and inertia of the equipment being used. Both directly affect the load values measured during a given test. In a modified servo-hydraulic load frame, large grips can create a large difference in stiffness (known more specifically as mechanical impedance) between each grip and the specimen. This difference causes more of any generated stress wave to be reflected at the interface as opposed to transmitting through; thus, as the wave travels back-and-forth through the specimen, the load it experiences “rings up”. This is sometimes referred to as “load ringing”. During this transient period of ring-up, the wave is amplifying or changing shape each time it reaches an end of the specimen. From this, the specimen experiences a load gradient across its gauge section depending on where the wave is at that point. After a certain number of reflections have occurred, the stress wave within the specimen becomes uniform and the specimen is determined to be in stress equilibrium. The number of oscillations it takes for this to occur is greatly dependent on the maximum frequency of the stress wave that enters the specimen and the ratio of mechanical impedances between the specimen and its grips. The lower the difference in impedances, the lower the energy that is reflected back into the specimen. The overall time of this period is not affected by the strain rate of the test being performed. This is because the phenomenon is based more so on the number of times the wave traverses the specimen which is solely dependent on the wave speed and length of the specimen. The wave speed of a material (assuming one—dimensional wave propagation) is calculated by:
Equation 2
Where c is the wave speed of the material, E is the Young’s modulus of the material, and ρ is its density. When viewing test results, the load ringing can be seen as a decaying sinusoidal that begins when the specimen is first loaded. Figure 7 shows an example of how this could look across various strain rates. Each data set shares the same decay time constant, frequency, and magnitude of oscillations. Holding these three as constant simulates running tensile tests using the same specimen geometry, machine, and grips but varying the strain rate of the test. In the data set at lower strain rates, the effect of load ringing is minor with no notable difference between the actual response and the measured response at 10 strain per second. At 30 and 100 strain per second, the yield portions of the stress-strain curves are noisy, but the overall hardening profile is still clean. At 300 strain per second, the oscillations affect most of the hardening potion of the stress-strain curve. At 1000 strain per second, the overall profile of the stress-strain curve can be made out including strain to failure, but no details regarding hardening, uniform elongation, or tensile strength can be stated without large error bands. At 3000 strain per second, very little can be discerned from the data.
Figure 7. Illustration of theoretical frequency responses of dynamic tensile tests performed at various strain rates. All tests share the same decaying time constant, frequency, and magnitude of oscillations.
Some laboratories have adopted an inverse method that takes data with excessive load ringing and derives a stress-strain model. This is done by simulating the equipment used to perform the test along with the specimen tested in a finite element model. Then, repeated iterations of the stress-strain profile are sequentially optimized until the simulation best fits the data read from the test which includes the load oscillations. While this method has shown great potential, it is often too time intensive and expensive to justify in most industrial applications.
Adiabatic Heating in High Strain Rate Testing
The final challenge when testing materials at high strain rates comes from a by-product of plastic deformation of materials: adiabatic heating. Quasi-static tests allow for iso-thermal testing where the rate of heat being generated from plastically deforming the specimen is exceeded by the rate that heat is lost to the surrounding environment. As the strain rate of the test is increased, so too does the rate of plastic work and heat generation within the specimen. This internal heating is also compounded with the need for high intensity lighting to illuminate any high-speed optical methods for measuring strain of specimens. Because of these heat sources without equivalent cooling, the material response as measured by a high strain rate test is jointly affected by the higher strain rate as well as the elevated temperature of the specimen. These two effects have been shown experimentally to not be independent of one another, further complicating the analysis and interpretation of these tests. More complex multi-physics material models are often employed to account for these coupled effects in finite element model simulations.
Thanks are given to Trey Leonard, for his contributions to this page. Trey is the founder and CEO of Standard Mechanics, LLC, where he delivers advanced mechanical testing services and solutions across a wide range of applications, from automotive design to consumer electronics. His expertise spans formability, fatigue, and strain rate sensitivity testing. Beyond conducting tests, he partners with customers to ensure the development of high-quality, calibrated material cards and models for accurate finite element simulations. Dr. Leonard earned his Ph.D. in Mechanical Engineering from Mississippi State University, where he pioneered and licensed technologies in dynamic material testing and characterization. Building on this foundation, he continues to develop innovative testing methods and technologies that advance the field of dynamic mechanical testing. He also contributes to the broader engineering community through his work with ASTM International, where he serves on the E28 Committee on Mechanical Testing, regularly reviewing and improving industry standards.
Predicting metal flow and failure is the essence of sheet metal forming simulation. Characterizing the stress-strain response to metal flow requires a detailed understanding of when the sheet metal first starts to permanently deform (known as the yield criteria), how the metal strengthens with deformation (the hardening law), and the failure criteria (for example, the forming limit curve). Complicating matters is that each of these responses changes as three-dimensional metal flow occurs, and are functions of temperature and forming speed.
The ability to simulate these features reliably and accurately requires mathematical constitutive laws that are appropriate for the material and forming environments encountered. Advanced models typically improve prediction accuracy, at the cost of additional numerical computational time and the cost of experimental testing to determine the material constants. Minimizing these costs requires compromises, with some of these indicated in Table I created based on Citation R-28.
Table I: Deviations from reality made to reduce simulation costs. Based on Citation R-28.
Yield Criteria
The yield criteria (also known as the yield surface or yield loci) defines the conditions representing the transition from elastic to plastic deformation. Assuming uniform metal properties in all directions allows for the use of isotropic yield functions like von Mises or Tresca. A more realistic approach considers anisotropic metal flow behavior, requiring the use of more complex yield functions like those associated with Hill, Barlat, Banabic, or Vegter.
No one yield function is best suited to characterize all metals. Some yield functions have many required inputs. For example, “Barlat 2004-18p” has 18 separate parameters leading to improved modeling accuracy – but only when inserting the correct values. Using generic textbook values is easier, but negates the value of the chosen model. However, determining these variables typically is costly and time-consuming, and requires the use of specialized test equipment.
Hardening Curve
Metals get stronger as they deform, which leads to the term work hardening. The flow stress at any given amount of plastic strain combines the yield strength and the strengthening from work hardening. In its simplest form, the stress-strain curve from a uniaxial tensile test shows the work hardening of the chosen sheet metal. This approach ignores many of the realities occurring during forming of engineered parts, including bi-directional deformation.
Among the simpler descriptions of flow stress are those from Hollomon, Swift, and Ludvik. More complex hardening laws are associated with Voce and Hockett-Sherby.
The strain path followed by the sheet metal influences the hardening. Approaches taken in the Yoshida-Uemori (YU) and the Homogeneous Anisotropic Hardening (HAH) models extend these hardening laws to account for Bauschinger Effect deformations (the bending-unbending associated with travel over beads, radii, and draw walls).
As with the yield criteria, accuracy improves when accounting for three-dimensional metal flow, temperature, and forming speed, and using experimentally determined input parameters for the metal in question rather than generic textbook values.
Failure Conditions
Defining the failure conditions is the other significant challenge in metal forming simulation. Conventional Forming Limit Curves describe necking failure under certain forming modes, and are easier to understand and apply than alternatives. Complexity and accuracy increase when accounting for non-linear strain paths using stress-based Forming Limit Curves. Necking failure is not the only type of failure mode encountered. Conventional FLCs cannot predict fracture on tight radii and cut edges, nor can they account for dimensional issues like springback. For these, failure criteria definitions which are more mathematically complex are appropriate.
Constitutive Laws
Simple material models reduce the effort of testing, but may not be sufficiently accurate or do not apply to the spectrum of grades available today.
The yield surface constructed with Hill’48H-74 requires only data from three tensile tests. Models like Barlat-1989B-89, Barlet-2003B-90, Barlat-2005B-91, BBC2005B-92, VegterV-26, and Vegter liteV-27 all require data from more detailed advanced tests, but simulation results incorporating these yield surfaces more closely match experimental results.
Use caution with the assumptions that go into the Material Card. For example, the card may show the r-value in the rolling, diagonal, and transverse (0°, 45°, and 90°) orientations all the same (not realistic), or worse yet, all equal to 1. For high strength and advanced high strength steels, it is likely that at least one of the orientations will have an r-value below 1. In these cases assuming an r-value of 1 will lead to an underestimation of the thinning. Furthermore, use of the Hill 1948 yield criterion is not recommended since the model assumptions do not apply to r-values of less than 1.
The Keeler equation for FLC0K-71 requires only the n-value and thickness, but is based on a correlation established from grades and testing available no later than the early 1990s.
Predictive models for the yield surfaceA-96 and FLCsA-97 in cold stamping conditions were created to simplify the testing requirements while maintaining the accuracy and usefulness of these advanced models. These predictive equations have been validated against physical testing for mild steels, conventional high strength steels, advanced high strength steels, aluminum alloys, and stainless steel grades.
Similarly, for hot stamping, predictive models based on conventional tensile testing have been developed and verified.A-94, D-48, A-98 Challenges here include that elongation and r-value both vary with temperature and testing speed.
On the Forming Limit Curve shown in Figure 1a, the uniaxial strain path, plane strain, biaxial, and balanced biaxial points are predicted from total elongation (A80) and r-value determined from tensile testing in the 0, 45 and 90° orientations with respect to rolling direction.A-94 Compared with the Keeler model, the Abspoel & Scholting model is better at predicting the FLC of DP800, noting the upper limiting strains found in the experiment match the FLC, as well as a more accurate representation of the slope on the LH side. Both models appear sufficient for the conventional grade DC04 (similar to CR3).
Figure 1. a) FLC prediction locations. b) FLC comparison on drawing steel DC04 (CR3); c) FLC comparison on DP800.A-94
Yield surface correlations use Tensile strength (Rm), uniform elongation (Ag) and r-value test data as inputs to predict the equi-biaxial, plane strain and shear points in three directions. Figure 2 compares biaxial yield strength predictions between Hill’48 and Vegter 2017, showing the improved correlation in the model developed almost 70 years later.
Figure 2. Comparison of measured biaxial yield strength with prediction from Hill’48 (red) and Vegter 2017 (yellow).A-94
Figure 3 presents a comparison of the measured yield surfaces of DX54D+Z (galvanized CR3) and DP1000 with those predicted by Hill’48 and Vegter 2017, highlighting the improved accuracy found in Vegter 2017.
Figure 3. Comparison of measured yield surface with predictions from Hill’48 and Vegter 2017. a) DX54D+Z (galvanized CR3); b) DP1000.A-94
Material properties like elongation, r-value, the hardening curve, and forming limits are all likely both strain-rate and temperature dependent, meaning that a rate-dependent and temperature-dependent yield surface and forming limit curve are needed for more accurate representations of cold- and hot-stamping.
The initial implementation of the Vegter yield surface in forming simulation software packages showed satisfactory correlation with conventional stamping applications, but was sub-optimal in operations where stresses are found in the thickness direction such as coining, wall ironing and score forming processes in packaging and battery applications. In these cases, a non-convexity close to the equi-biaxial point of the yield locus was observed, likely due to extreme anisotropy or very low r-values.
Citation A-95 discusses methods for improved accuracy. DIC measurements offer improved r-value characterization over mechanical measurement approaches. Yield locus correlations at low and high plastic strain ratios were also improved. Shell elements used in the simulation of the yield surface in plane stress ignore the strains in the through-thickness direction. For the applications where thickness stresses play a role (like wall ironing, coining, score forming in packaging, and sharp radii in closures, solid or thick shell elements are required. The yield surface was extended to the thickness direction to allow for improved characterization in these applications having significant stresses through the thickness. This extension may be deployed in forming simulation software as “Vegter 2017.1.”
Constitutive Laws and Their Influence on Forming Simulation Accuracy
Many simulation packages allow for an easy selection of constitutive laws, typically through a drop-down menu listing all the built-in choices. This ease potentially translates into applying inappropriate selections unless the simulation analyst has a fundamental understanding of the options, the inputs, and the data generation procedures.
Some examples:
The “Keeler Equation” for the estimation of FLC0has many decades of evidence in being sufficiently accurate when applied to mild steels and conventional high strength steels. The simple inputs of n-value and thickness make this approach particularly attractive. However, there is ample evidence that using this approach with most advanced high strength steels cannot yield a satisfactory representation of the Forming Limit Curve.
Even in cases where it is appropriate to use the Keeler Equation, a key input is the n-value or the strain hardening exponent. This value is calculated as the slope of the (natural logarithm of the true stress):(natural logarithm of the true strain curve). The strain range over which this calculation is made influences the generated n-value, which in turn impacts the calculated value for FLC0.
The strain history as measured by the strain path at each locationgreatly influences the Forming Limit. However, this concept has not gained widespread understanding and use by simulation analysts.
A common method to experimentally determine flow curves combines tensile testing results through uniform elongation with higher strain data obtained from biaxial bulge testing. Figure 4 shows a flow curve obtained in this manner for a bake hardenable steel with 220 MPa minimum yield strength. Shown in Figure 5 is a comparison of the stress-strain response from multiple hardening laws associated with this data, all generated from the same fitting strain range between yield and tensile strength. Data diverges after uniform elongation, leading to vastly different predictions. Note that the differences between models change depending on the metal grade and the input data, so it is not possible to say that one hardening law will always be more accurate than others.
Figure 4: Flow curves for a bake hardenable steel generated by combining tensile testing with bulge testing.L-20
Figure 5: The chosen hardening law leads to vastly different predictions of stress-strain responses.L-20
Analysts often treat Poisson’s Ratio and the Elastic Modulus as constants. It is well known that the Bauschinger Effect leads to changes in the Elastic Modulus, and therefore impacts springback. However, there are also significant effects in both Poisson’s Ratio (Figure 6) and the Elastic Modulus (Figure 7) as a function of orientation relative to the rolling direction. Complicating matters is that this effect changes based on the selected metal grade.
Figure 6: Poisson’s Ratio as a Function of Orientation for Several Grades (Drawing Steel, DP 590, DP 980, DP 1180, and MS 1700) D-11
Figure 7: Modulus of Elasticity as a Function of Orientation for Several Grades (Drawing Steel, DP 590, DP 980, DP 1180, and MS 1700) D-11
Testing to Determine Inputs for Simulation
Complete material card development requires results from many tests, each attempting to replicate one or more aspects of metal flow and failure. Certain models require data from only some of these tests, and no one model typically is best for all metals and forming conditions. Tests described below include:
Tensile testing [room temperature at slow strain rates to elevated temperature with accelerated strain rates]
Biaxial bulge testing
Biaxial tensile testing
Shear testing
V-bending testing
Tension-compression testing with cyclic loading
Friction
Tensile testing is the easiest and most widely available mechanical property evaluation required to generate useful data for metal forming simulation. However, a tensile test provides a complete characterization of material flow only when the engineered part looks like a dogbone and all deformation resulted from pulling the sample in tension from the ends. That is obviously not realistic. Getting tensile test results in more than just the rolling direction helps, but generating those still involves pulling the sample in tension. Three-dimensional metal flow occurs, and the stress-strain response of the sheet metal changes accordingly.
The uniaxial tensile test generates a draw deformation strain state since the edges are free to contract. A plane strain tensile test requires using a modified sample geometry with an increased width and decreased gauge length,
Forming all steels involves a thermal component, either resulting from friction and deformation during “room temperature” forming or the intentional addition of heat such as used in press hardening. In either case, modeling the response to temperature requires data from tests occurring at the temperature of interest, at appropriate forming speeds. Thermo-mechanical simulators like Gleeble™ generate such data.
Conventional tensile testing occurs at deformation rates of 0.001/sec. Most production stamping occurs at 10,000x that amount, or 10/sec. Crash events can be 2 orders of magnitude faster, at about 1000/sec. The stress-strain response varies by both testing speed and grade. Therefore, accurate simulation models require data from higher-speed tensile testing. Typically, generating high speed tensile data involves drop towers or Split Hopkinson Pressure Bars.
A pure uniaxial stress state exists in a tensile test only until reaching uniform elongation and the beginning of necking. Extrapolating uniaxial tensile data beyond uniform elongation risks introducing inaccuracies in metal flow simulations. Biaxial bulge testing generates the data for yield curve extrapolation beyond uniform elongation. This stretch-forming process deforms the sheet sample into a dome shape using hydraulic pressure, typically exerted by water-based fluids. Citation I-12describes a standard test procedure for biaxial bulge testing.
A Marciniak test used to create Forming Limit Curves generates in-plane biaxial strains. Whereas FLC generation uses 100 mm diameter samples, larger samples allow for extraction of full-size tensile bars. Although this approach generates samples containing biaxial strains, the extracted samples are tested uniaxially in the conventional manner.
Biaxial tensile testing allows for the determination of the yield locus and the biaxial anisotropy coefficient, which describes the slope of the yield surface at the equi-biaxial stress state. This test uses cruciform-shaped test pieces with parallel slits cut into each arm. Citation I-13 describes a standard test procedure for biaxial tensile testing. The biaxial anisotropy coefficient can also be determined using the disk compression testing as described in Citation T-21.
Shear testing characterizes the sheet metal in a shear loading condition. There is no consensus on the specimen type or testing method. However, the chosen testing set-up should avoid necking, buckling, and any influence of friction.
V-bending testsdetermine the strain to fracture under specific loading conditions. Achieving plane strain or plane stress loading requires use of a test sample with features promoting the targeted strain state.
Tension-compression testing characterizes the Bauschinger Effect. Multiple cycles of tension-compression loading captures cyclic hardening behavior and elastic modulus decay, both of which improve the accuracy of springback predictions.
The Bauschinger effect leads to early re-yielding after loading reversal, and has been observed in loading-reverse loading testing. The Yoshida-Uemori (YU) kinematic hardening model accurately captures the Bauschinger effect as well as other hardening behaviors of sheet metals during loading-reverse loading.Y-7, Y-8
Characteristics of the Bauschinger Effect (Figure 8) include a) the transient Bauschinger deformation characterized by early re-yielding and smooth elastic–plastic transition with a rapid change of the work hardening rate; b) the permanent softening characterized by a stress offset observed in a region after the transient period; and c) work-hardening stagnation appearing at a certain range of reverse deformation.Y-7
Figure 8: Characteristics of the Bauschinger Effect during cyclic loading.Y-7
Citation L-78 describes an approach to calibrate the YU model on a QP1500 steel that uses a combination of physical testing and machine learning to achieve loading-reverse loading stress-strain curves over broader strain ranges. This citation also reported that the results from tension-compression testing were not the same as those from compression-tension testing – meaning that the order of deformation influences the results.
However, no standard procedure exists for determining the kinematic hardening and Bauschinger parameters and subsequently incorporating it into metal forming simulation codes. Independent of the procedure, one of the biggest challenges with this test is preventing buckling from occurring during in-plane compressive loading. Related to this is the need to compensate for the friction caused by the anti-buckling mechanism in the stress-strain curves.
Frictionis obviously a key factor in how metal flows. However, there is no one simple value of friction that applies to all surfaces, lubricants, and tooling profiles. The coefficient of friction not only varies from point to point on each stamping but changes during the forming process. Determining the coefficient of friction experimentally is a function of the testing approach used. The method by which analysts incorporate friction into simulations influences the accuracy and applicability of the results of the generated model.
Studies are underway to reduce the costs and challenges of obtaining much of this data. It may be possible, for example, to use Digital Image Correlation (DIC) during a simple uniaxial tensile testing to quantify r-value at high strains, determine the material hardening behavior along with strain rate sensitivity, assess the degradation of Young’s Modulus during unloading, and use the detection of the onset of local neck to help account for non-linear strain path effects.S-110
Application of Advanced Testing to Failure Predictions
Global formability failures occur when the forming strains exceed the necking forming limit throughout the entire thickness of the sheet. Advanced steels are at risk of local formability failures where the forming strains exceed the fracture forming limit at any portion of the thickness of the sheet.
Fracture forming limit curves plot higher than the conventional necking forming limit curves on a graph showing major strain on the vertical axis and minor strain on the horizontal axis. In conventional steels the gap between the fracture FLC and necking FLC is relatively large, so the part failure is almost always necking. The forming strains are not high enough to reach the fracture FLC.
In contrast, AHSS grades are characterized by a smaller gap between the necking FLC and the fracture FLC. Depending on the forming history, part geometry (tight radii), and blank processing (cut edge quality), forming strains may exceed the fracture FLC at an edge or bend before exceeding the necking FLC through-thickness. In this scenario, the part will fracture without signs of localized necking.
This collection of studies, as well as work coming out of these studies, show that relatively few tests sufficiently characterize forming and fracture of AHSS grades. These studies considered two 3rd Gen Steels, one with 980MPa tensile strength and one with 1180MPa tensile.
The yield surface as generated with the Barlat YLD2000-2d yield surface (Figure 9) comes from:
Conventional tensile testing at 0, 22.5, 45, 67.5, and 90 degrees to the rolling direction, determining the yield strength and the r-value;
Disc compression tests according to the procedure in Citation T-21 to determine the biaxial R-value, rb.
Figure 9: Tensile testing and disc compression testing generate the Barlat YLD2000-2d yield surface in two 3rd Generation AHSS Grades B-13
Creating the hardening curve uses a procedure detailed in Citations R-5 and N-13, and involves only conventional tensile and shear testing using the procedure included in Citation P-15.
Figure 10: Test geometries for hardening curve generation. Left image: Tensile; Right image: Shear.N-13
Characterizing formability involved generating a Forming Limit Curve using Marciniak data or process-corrected Nakazima data. (See our article on non-linear strain paths) and Citation N-13 for explanation of process corrections]. Either approach resulted in acceptable characterizations.
Fracture characterization uses four plane stress tests: shear, conical hole expansion, V-bending, and a biaxial dome test. The result from these tests calibrate the fracture locus describing the stress states at fracture.
A different approach requires only the results from conventional tensile testing and a crack growth test under simple loading to simulate post-necking strain hardening behavior and ductile fracture. The details of this approach are beyond the scope of this webpage, but are presented in detail in Citations S-124 and T-56, in addition to verification procedures.
Simulation Set-Up Parameters
One of the most basic choices when starting a simulation run is the setting related to the mesh size. Reduced processing time is associated with large mesh sizes, but that risks not having sufficiently fine mesh resolution to capture the forming strain gradient. A large mesh size averages the strains over a larger region, which is analogous to a tensile bar with an 80 mm gauge length having lower elongation than a 50 mm tensile bar cut from the same sheet steel.
Figure 11 compares the Forming Limit Curve (FLC) for an 1180 MPa steel determined from gauge lengths of 2, 6, and 10 mm, along with the associated theoretical predictions. As expected, the smaller gauge length is able to more effectively capture peak strains, and is therefore associated with a higher forming limit.A-90
Figure 11: Comparison of predicted values and experimental values of the Forming Limit Curve of an 1180 MPa steel.A-90
The stress-strain curve of a 1.6 mm 980 MPa steel tested with a 50 mm gauge length (ISO III, JIS) was captured, resulting in a strain at fracture of 0.147. A model based on a 2 mm element size was created, calibrated to the same strain at fracture of 0.147. The model was re-run with element sizes of 3 mm and 5 mm, which resulted in different stress strain curves and simulations that could not predict the fracture known to occur, Figure 12. This study also showed a technique that can be used to achieve similar performance nearly independent of mesh size, such that accuracy is not compromised when optimizing computer processing speed.A-90
Figure 12: Comparison of the tensile test result and fracture model predictions based on different element sizes.A-90
Constitutive Models
Constitutive models for steel strengthening fall into two general categories: power law behavior like HollomonH-71 and SwiftS-119 or saturation models like Voce and Hockett-Sherby.H-72 As shown in Figure 2 above, the chosen constitutive model significantly influences the extrapolation of experimental stress-strain curves to larger strain values. Model combinations such as Swift-Voce or Swift/Hockett-Sherby, typically using one for lower strains and the other for higher strains, typically provide better fit with experimental dataK-65, but more parameters are usually beneficial, especially for advanced high strength steels where the n-value is not constant with strain.
To improve the modeling accuracy of high strength steels with variable instantaneous n–value, hardening curves obtained with uniaxial tensile and hydraulic bulge tests were fit to a new proposed modelL-73 to verify its predictive capability and accuracy. This new model, based on the Swift power law (Equation 1), addresses the decrease in n-value at larger plastic strains by varying what has been termed as the strain hardening attenuation coefficient a, within a new parameter λh as defined in Equation 2.
Equation 1
Equation 2
When a=0, Equation 1 reverts to the standard Swift equation. When a>0, it allows for Equation 1 to correct for the decrease of instantaneous n value occurring at larger plastic strains. The results in Figure 133 show that the predictive accuracy of the new model is better than the individual Swift or Hockett-Sherby models.
Figure 13: Hardening Curve for two grades showing Uniaxial Tensile (ut) stress-strain curves, biaxial tensile extension from biaxial bulge testing (bt), Swift and Hockett-Sherby model fit, and new model fit with different a parameter.L-73
Experimental hardening data for a QP grade, referred to as QP1180-EL, was obtained from uniaxial tensile testing combined with bulge testing and in-plane torsion testing for strains beyond uniform elongation. These are shown as squares and black or red circles in Figure 14, along with projections from the Swift and Hockett-Sherby models. The Modified Power Law (MPL) achieves the best fit to the tested results.Z-18
Figure 14: Experimental results compared with the Swift power law hardening model, Hockett-Sherby saturation hardening model, and a newly developed Modified Power Law.Z-18
Improved vehicle crashworthiness predictions occur when the forming history of the critical structural parts, including the effects of bake hardening, work hardening, and thickness reduction, are incorporated into vehicle virtual development models. Historically, simulations did not contemplate the initial damage caused by plastic deformation. Accumulated damage can be captured within a GISSMO (Generalized Incremental Stress State Dependent Model) damage model, albeit with certain assumptions.N-30, N-31
Five different loading cases capturing the stress state of shear, uniaxial tension, stretching, plane strain and equi-biaxial stretching can be used to calibrate parameters of the Modified Mohr-Coulomb (MMC) B-83 fracture model. Schematics of these five individual tests are shown in Figure 15.H-73 The calibrated MMC model and loading path results from these tests are shown in Figure 16. The MMC model was subsequently used to calibrate a GISSMO damage model.
Figure 15: Tests coupons for fracture model calibration.H-73
Figure 16. Calibrated MMC fracture model and loading path results from the tests shown in Figure 10.H-73
Quasi-static three-point quasi-static bending tests were used to validate the MPL hardening model, the MMC fracture model, and the GISSMO damage model. An FEA model with a 2 mm mesh size was compared with one having a 5 mm mesh size for the simulation of the bending process. Figure 17 shows the predicted fracture location and test result.
Local necking was observed during the experimental bending test and 2 elements failed in when using a 2 mm mesh size, yet no failure was observed when using a 5 mm mesh. This indicates that accurate simulation results may require refined mesh sizes.
Figure 17: Experiment and simulation results of three-point bending testing of a Quenched & Partitioned 1180 MPa Steel.H-73
A subsequent studyZ-18 confirmed that ignoring the stamping forming history in the damage model results in a lower prediction of the failure risk, especially for cold stamping high-strength steel parts under large deformation conditions.
Case Studies: Benefits of using Advanced Models for Springback Prediction
The output of simulations using material models that thoroughly capture the changes in metal properties occurring during forming are more likely to match reality than those simulations based on basic models.
Kinematic hardening models where the Bauschinger effect and modulus degradation are captured have been shown to be substantially more accurate in springback prediction than isotropic hardening models based on more conventional tensile testing.
Citation S-130 investigated this difference. Different types of steels having 980 MPa or 1180 MPa minimum specified tensile strength from multiple suppliers were used to form a targeted part shape using either a draw forming process or a crash forming process. The formed panels were scanned and compared with simulation results from multiple software packages. In all cases, the simulation was capable of accurately predicting strains and the risk of necking failure.
For springback, the type of hardening model used in the simulation appeared to correlate with prediction ability. Table 2 compares the dimensional difference between the model and a scan of the physical panel, with smaller numbers representing an improved ability to predict springback in the evaluated condition. As indicated in Table 2, models incorporating Kinematic Hardening more closely matched the actual springback seen on the scanned panels.
Table 2: Dimensional Deviation Between Simulation and Scan as a Function of Hardening Model.S-130
Maximum Sectional Deviation
During Draw Forming (mm)
Maximum Sectional Deviation
During Crash Forming (mm)
Hardening Model
5.38
5.54
Isotropic Hardening
5.25
2.27
Yoshida
5.01
10.2
Isotropic Hardening
4.2
2.422
Hill ’48 Isotropic Hardening
4.17
3.127
Hill ’48 Isotropic Hardening
3.14
2.8
Yoshida
3.12
5.3
Yoshida
2.65
N/A
Yoshida
2.3
4.2
Yoshida-Uemori
2.17
7.39
Yoshida
1.98
1.64
Yoshida-Uemori
1.797
1.952
Yoshida
1.5
2.3
Yoshida
Results compiled across multiple types of simulation software,
compared with formed parts made from different types of 980 and 1180 grades from multiple suppliers.
A modified S-shape generic panel was used to evaluate springback using a 3rd Generation Steel having a minimum specified tensile strength of 980MPa.B-98 Nearly 200 points were evaluated on the panel shown in Figure 18. For the simulation that did not incorporate kinematic hardening, 48% of this panel were within 1 mm of the physical scanned part and 32% were within 0.5 mm. When the simulation incorporated kinematic hardening, 94% of the points were within 1 mm and 60% were within 0.5 mm.
Figure 18: Modified S-Shape used to evaluate springback on 3rd Gen 980 MPa steel in Citation B-98.
A generic B-pillar panel was used to evaluate springback using a 3rd Generation Steel having a minimum specified tensile strength of 1180MPa.K-72
For the simulation that did not incorporate kinematic hardening, 59.9% of this panel were within 1 mm of the physical scanned part. When the simulation incorporated kinematic hardening, 68.9% of the points were within 1 mm. Simulation results are presented in Figure 19.
Figure 19: Springback results on a B-pillar formed from 3rd Gen 1180 MPa steel.K-72
Thanks go to Eren Billur, Ph.D. for his contribution of this article to the AHSS Insights blog. Eren Billur is the Technical Manager of Billur Makine and Billur Metal Form, based in Ankara, Turkey, specializing in advanced sheet metal forming technologies. He holds a Ph.D. in Mechanical Engineering from The Ohio State University and has extensive experience in material characterization, sheet metal forming processes, and finite element simulations. Eren has contributed significantly to the understanding and application of hot stamping and advanced high-strength steels (AHSS) in the automotive industry. He is a regular columnist for MetalForming Magazine’s “Cutting Edge” column and has authored numerous scientific papers and book chapters, including contributions to the WorldAutoSteel AHSS Applications Guidelines. Passionate about advancing manufacturing knowledge, Eren provides engineering consulting, training, and simulation services worldwide, helping manufacturers optimize forming processes and successfully implement new-generation AHSS materials.
Thanks are given to
Thanks are given to 
