The press hardening process is a relatively new technology, which allows ultra-high strength steels to be formed into complex shapes. In this context, a new material model has been developed that allows measuring and modelling material behaviour along measurable phase paths. The reward of precise material and process modelling is the realistic determination of the deviation of the formed shape from the nominal part shape, accurate quenching time determination, correct material properties and residual stresses at the end of the process, and significant time savings in the try-out period.
Steps for hot forming process
As in most sheet metal forming operations, hot forming also starts with a coil from which the blanks are cut from and end with the final useful component. The first operation is decoiling and blanking. Then, the blanks are stacked. Another material handling unit takes one blank at a time and places them into a furnace. The furnace heats the blank over their austenitizing temperature.
The blanks are then transferred quickly to a press where forming and quenching are done consecutively. The parts are moved out of the press and typically laser trimming/piercing is the last step.
Process design and validation
Process design and validation are carried out in a close interaction between design of tool faces and simulation. Once the process is validated, the next step is to optimise quenching and tool surfaces to achieve the best possible process performance.
Quenching and virtual die spotting
Quenching is one of the most time consuming steps of hot forming process. Most of the research and development in the field is done to reduce the time required for quenching. Although this can be achieved also by novel quenching methods, the most common and most applicable are: (1) to optimise the cooling channels and fluid, and (2) to improve the contact between the blank and the tool.
Heat transfer from the blank is a function of gap and pressure in the real world. In hot stamping, after the dies are machined, they are installed to a press. The blank is covered with a water-soluble compound called ‘spotting blue’ and a few stampings are done. The dies are then manually ground (die spotted) to have at least 90% of the area covered by the blue color. This ensures that only a small portion of the blank will have heat transfer through air gap. This process can take several days of labour.
‘Virtual Die-Spotting’ allows shortening this time. First, the simulation is done using ‘nominal (CAD) surfaces’. As the thinning of the blank is calculated, a new ‘spotted’ die face can be automatically generated. In the next iteration, the spotted geometry can be used. With the spotted die set, better quenching results can be achieved. In addition, distortion after air cooling is also reduced.
Mechanics of material modelling
In order to get good results in terms of distortion, it is necessary to model the behaviour of the material precisely. The mathematical description of the continuous and isothermal cooling behaviour yields in the correct description of the phase transformations as a function of the measured cooling rates. At each increment of the simulation, the actual cooling rate for the computed increment is determined and then used to compute the phase transformations for this increment. Like that, any cooling path can be accommodated and it is possible to simulate correctly all possible quenching and cooling processes.
Next to correct flow stress curves as a function of phases/phase paths/temperature and strain rates, the material model must reflect the evolution of the thermal strains as a function of phases and temperature to be able to compute accurate change in shape due to forming and quenching. In the hard section of the part, there is only a martensite transformation and in the soft section of the part, it is mainly a bainite transformation.
One key effect to compute correctly distortion after quenching is transformation plasticity – plastic deformation that occurs only due to phase transformation evolution and happens at applied stresses much lower as the nominal yield strength would be.
Stress relief due to transformation plasticity is measured with the Satoh test. It allows measuring the stress that accumulates during cooling. The specimen is heated until a desired peak temperature is reached. Thermal contraction causes the accumulation of stresses, which are partly relieved during phase transformations. A new material model has been implemented in PAM-STAMP, which allows measuring and modeling material behaviour along measurable phase paths. Depending on the evolution of phase transformations, those properties are mixed. To validate the new material model, SYSWELD is used, which is a reference in the domain of modelling heat effects of welding and heat treatment including phase transformations and restoring of strain hardening during phase transformations. It uses a per phase model to compute stresses.
Accumulated stresses, with and without transformation plasticity, for a 100% transformation into martensite, along a cooling rate of -50K/s, for 22MnB5 steel, are illustrated in Figure 1. In Figure 2, the stress relief is illustrated for a cooling rate of -7.5K/s that yields in a combined ferrite, bainite and martensite phase transformation. Both material models, the per-phase and phase path model yield in the same results.
To simulate a part with ‘tailored-heating’ process, the flow stress data of steel before austenitisation is required. This data is currently being generated at Atılım University Metal Forming Centre of Excellence for several materials. Tests are done at Bähr DIL 805 Forming Dilatometer. Material data of non-austenitised 22MnB5 and other grades (6MnB6, 37MnB4, etc) will be added to a stamping material library after validation studies are completed. Note that, before austenitisation, the total elongation is limited to <0.1, whereas after austenitisation, total elongation is >0.3 true strain.
Trends in hot forming industry
Tailor rolled blanks: Tailor rolled blanks have been increasingly used in the automotive industry. The new VW Passat, Peugeot 308, Renault Twingo/smart ForFour and Volvo XC90 are just a few examples for cars introduced in 2014 that had hot stamped TRB components.
Patchwork blanks: In 2014, Subaru Impreza and Fiat 500X were introduced and both cars had patchwork reinforced hot stamped B-pillars. A ‘spot-welded’ patch reinforced hot stamped blank. In this case, thickening regions had wrinkles but as the press went down, they were flattened.
New grades – tailor welded blanks: Up until now, we have only seen 22MnB5, uncoated or typically AlSi coated steels in hot forming industry. There have been some examples of Zn-based coatings as well, but the application level was low. However, recently Volvo has started using 6MnB6 (commercially known as DUCTIBOR 500, MBW 500 or phs-ultraform 490). Renault has also shown an example of front rail with this material. 6MnB6 is typically welded to 22MnB5 and hot stamped.
There are also new developments among steel makers: (1) to achieve >1800 MPa UTS levels and (2) having high elongation grades at 1000-1300 MPa UTS levels. These materials will also be studied at Atılım University.
Realistic virtual prototyping
Simulation of hot forming requires a complex material model that has to cover mechanical, thermal and metallurgical effects. A new material model has been developed that allows measuring and modelling material properties along measurable phase paths. Different materials, and welding techniques (patchwork spot welded, laser welded), tailored (rolled/ welded/quenched) as well as all kind of process variations can be included in the simulation model, which allows realistic virtual prototyping.