AHSS simulations for automotive design

1. What to watch for in stretched cut AHSS edges

Starting off with the number 1 issue when making forming simulations for AHSS: the stretching of cut edges. You need to be aware of any situations where you see uniaxial tension in a cut edge.

The forming limit diagram cannot be used as a guide for edge stretching, simply because when we test material in the lab and create the forming limit curve we are not testing the AHSS steel at the cut edge, but in the middle — in the body of the sheet.

There is also no correlation between edge ductility and elongation values — so the AHSS “banana curve” graph cannot be used to evaluate edge ductility.

A strong influence on AHSS edge ductility limit is how the cutting tool has been designed. At SSAB’s Knowledge Service Center we test our Docol® automotive AHSS steels to find the optimal cutting clearance for each grade.

But what makes AHSS simulations even more complicated is that edge ductility changes during serial production due to wear of the cutting tool. Some simulation software has started to include ways to account for edge stretching, with default values for laser cut edges being the largest, followed by new punch edges, then worn punch edges.

So, in AHSS simulations, paid attention to:

• Where is the edge stretching?
• What is the amount of edge stretching?
• What kind of edge stretching is it?

2. Use a practical test to validate edge strain on AHSS cut edges

There are many ways to generate different AHSS behaviors and strain gradients: in the sheet plane, in the thickness direction, and also in the strength and concentration along the cut edge itself.

SSAB created a practical test — the double-bending test — that checks for the maximum bending angle of AHSS steel before cracking occurs.

We take our results from the double-bending test and compare it to the hole expansion test. And there can be huge differences between the tests in their acceptable strain levels. For example, a 1 mm 980 DP steel can have a maximum strain of 46% in the hole expansion test but only 11% in the double-bending test.

 

3. Look for major strains crossing pre-strained AHSS areas

The ISO 16630 hole expansion test for edge ductility is conducted with zero pre-strain in an AHSS sample. In reality it is common that the AHSS sheet is pre-strained prior to cutting and final cut edge straining. It is hard to design a general test for this situation as pre-straining the large specimen (100 x 100 mm) for the hole expansion ratio (HER) test is challenging. How can you predict the AHSS steel’s capability for this situation?

Instead of depending solely on the HER test, you can simulate the part — watching for major strains crossing pre-strained areas. If you find any, you have a few options. You might opt to change material to an AHSS grade with a better cut edge ductility to provide an extra margin of safety. Or you might adjust your design to keep the pre-strain at a lower level. Or try to move the pre-strain to another area where the final strain of the part is lower.

4. Have SSAB determine strain levels when bending

The forming limit diagram (FLD) is valid for elements that have the same strain throughout their thicknesses. But when you bend AHSS steel, you have stretching on the outside, compression on the inside, and a neutral undeformed layer in the middle. Using the standard view in simulations, you’re looking at the neutral layer.

Instead you should look at the amount of strain at the outer layer of the sheet. But when doing this you should not use the FLD plot for determining failure for the outer surface: doing this can lead to an overly conservative result.

So what levels are safe for AHSS bending? You should ask us for these values. For example, we made a test in our forming laboratory for 2.0 mm Docol® 1400M for a customer in Germany. On this bend, we measured 18% strain, which is much higher than the 10% strain we get from the forming limit curve for this material in the state of equal strain through thickness (FLD test).

5. Use incremental forming simulations to catch phenomenon like bending and unbending

If you bend any metal, and then bend it back the opposite direction, and keep doing that, back and forth, the metal will eventually break — you’ve accumulated damage in the material. This behavior cannot be caught by the forming limit curve and is challenging to model.

For example, we had a customer whose simulations showed no issues in AHSS forming — no strains that were over the limit. Still there were cracks during production! So we ran an incremental forming simulation that delivered a special result value called “accumulated strain” (see image).

 

6. Be careful not to be overly dependent on the high mechanical tolerances of AHSS steels

Sometimes we hear that the argument that all production instability comes from material variation. Consistent AHSS materials definitely matters, but that’s not the whole picture.

In fact, we do repeatability analyses that compare our Docol^® grades to general VDA grades. In one case, we looked at a simple flange made with 980 complex phase (CP) grade AHSS with a tolerance of ±1° in accordance to VDA 239. You can see the full analysis process in our on-demand webinar entitled: AHSS simulations for automotive design: top 10 considerations.

The analysis showed that that particular part, when made from the Docol^® 980 CP, was 628 times less likely to be out of tolerance than one made from the general VDA 980 CP — due to the Docol^® material’s higher mechanical tolerances.

High material consistency is always desirable, especially for AHSS/UHSS/Gigapascal steel applications really dependent on tight mechanical tolerances. But it's risky to design AHSS parts that are only depending on high mechanical tolerance. Many other factors come into play during production: process variations, tool wear, lubrication, etc.

What we like to say is that single most important parameter for a highly repeatable AHSS process is to have a robust part design, taking full advantage of high-stiffness geometries, small radii, the strategic use of gainers, and so on.

 

7. Optimize your AHSS forming layout

For optimizing the forming layout, you need to take a lot of parameters into account, including feasibility, repeatability, available press, and tool wear.

In our Simulations webinar, you can see how we stimulate the same AHSS auto part using three different forming approaches: Draw + Flange; Flange + Cam trim; and Flange with Cams.

For this particular side member design, the Draw + Flange simulation produces a maximum springback displacement of 10 mm and otherwise looks good. The Flange + Cam simulation has a maximum springback displacement of 13 mm but has tolerance problems in the convex surface radius. The Flange with cams simulation suffers from high strains in cut edges and huge deviations in shape accuracy due to folded radii.

8. Your wrinkling simulations may be too conservative

On AHSS parts with highly compressed edge flanges and no possibility to use a blank holder, you need simulate the part to try to detect wrinkling. Shown here is a part made from 4 mm thick AHSS. We simulated this part using three different approaches to compare to the actual prototypes:

1. A simulation performed with shell elements and no self-contact produced a result with low ability to recover from a wrinkle once it occurs. But reality shows no wrinkles after forming.
2. A simulation using fully integrated Solid elements and no self-contact. This result was closer to reality but still had remaining wrinkles after forming.
3. A simulation using Solid element and self-contact. This delivered a result with good consistency with reality.

For AHSS stamping simulations, the most common approach is using shell element with no self-contact. For determining wrinkling tendency, this is a very conservative element type. At the very least, what you can say is that if you have no wrinkling using shell elements with no self-contact, you will have no wrinkling in reality. However, as demonstrated in this example, this approach can set some limits to an AHSS part that are not present in reality.

9. Is your AHSS simulation detecting sheet reaction forces that could lead to tool opening?

When using AHSS/UHSS/Gigapascal steels, the reaction force from the sheet will increase when using blank holders. And if the AHSS reaction force is larger than the blank holder force, tool opening will occur. This leads to a highly uncontrolled process: you can get wrinkles and cracks and a very low correlation between your AHSS simulation and reality.

So, check closely that the forces on blank holders and pads are enough. Some simulation software has ways to detect the AHSS sheet’s reaction force during tool opening. Some software quietly adds more blank holder force to keep the tools shut — but it is extremely important to check if this is happening or not in your simulation software.

10. Take non-linear deformations into account

It is important to consider non-linear failure because the forming limit curve is developed for linear strain paths – meaning the forming takes place in only one way until failure occurs.

This means that when you are forming and reforming one area of an AHSS part, like in a multi-step forming tool, you get a situation that is not similar to the FLC. And in reality, the result can be either better or worse, depending on the path of deformation.

Some simulation software can take non-linear deformations into account. For example AutoForm has a non-linear forming diagram, which calculates and transforms non-linear strains and maps these on the FLD. This can be really helpful when using multi-stage forming – but also sometimes even when the forming is just in one stage, like in the following example.

The image on the left shows the traditional FLD plotted on the AHSS part (in this case, made from Docol^® 1000DP), showing red in one area, meaning strains are above the forming limit. The image on the right, however, is the non-linear (transformed) result, which shows that the part is actually OK.