Hot Stamping


Introduction


The advantage of hot stamping vehicle structural components is the exceptional as-formed ultimate tensile strength (1,500MPa) and the complex geometries that can be formed. The high strength of the parts is due to a fully martensitic phase transformation that occurs during the in-die quenching process. The elevated strength of the hot stamped parts allows the weight of the components to be reduced by using thinner gauge sheet metal while maintaining structural integrity and crash performance. A schematic of the hot stamping process is shown below and typically requires a boron steel blank to be austenized for approximately 5 minutes at a temperature greater than 900°C. After the soaking time, the blank is then transferred to a cooled stamping die. At the bottom of the forming press stroke, the formed part is quenched within the die. The cooling rate must be greater than ~30°C/s in order for a solid-state phase transformation (from austenite to martensite) to occur as shown by the continuous cooling transformation (CCT) diagram below.


A schematic of the hot stamping process and a USIBOR® 1500P CCT diagram

In addition to the good intrusion resistance and preservation of structural integrity of hot formed parts, some components, such as a B-pillar, may benefit from regions where the mechanical properties have lower strength and improved ductility for improved energy absorption as shown in the schematic below. One method of producing a single continuous part without additional process steps is to introduce daughter phases within these regions by reducing the cooling rate within the die during the stamping operation, which in turn causes the austenite to transform into bainite, ferrite and possibly pearlite. Optimizing the hot stamping process to create a part with tailored properties is a major component of the hot stamping research program at the University of Waterloo. Equally as important is the thermomechanical processing and heat transfer characterization of this process, which is researched collaboratively with Professors Wells and Daun, respectively.


Schematic of a hot stamped B-pillar with tailored properties

When modeling (using finite elements) the crash behaviour of a hot formed structural component, the strain rate response of the as-formed material must be taken into account in order to accurately model deformation and energy absorption at the elevated rates of strain. The strain rate sensitivity of the various microstructures within the tailored part is another key focus of the research at the University of Waterloo.


B-Pillar with Tailored Microstructures


At the University of Waterloo, the approach taken to produce a part with tailored properties is referred to as the in-die heating process. Tooling has been developed to control the in-die cooling rate of the blank by heating and cooling specific regions of the die. Heat transfer from the blank to the tool is dependent on several factors; die surface temperature being one of the most important. As the temperature of the die is increased, the heat transfer, and thus the cooling rate, is decreased. A preliminary forming and quenching (coupled thermo-mechanical) simulation is shown below using LS-Dyna, where a hot blank is formed in a die with a temperature gradient from hot to cold. Four elements are examined, and their temperature is plotted over time (average cooling rates are indicated).


Thermo-mechanical FE simulation of hot stamping with a heated and cooled die section

The tooling used for this work was developed in-house and the geometry is representative of a scaled-down B-pillar as shown by the CAD image below. The upper and lower dies are composed of a heated and cooled section, and are separated by a small air gap to insulate the two halves. A total of 15 cartridge heaters are used, with a combined power of 8.6 kW. A feature similar to a draw-bead was designed into the die to improve the formability of the blank without the use of a blank holder.


CAD image of the hot stamping die with heated and cooled sections

The press used for this work is shown below and contains a 125 ton actuator with an accumulator bank that allows for a punch speed of up to 5 in/sec. This punch speed is representative of an industrial hot stamping process. The furnace has an inner area of 24” x 36” and a 3-zone control system is installed to ensure a uniform temperature distribution which is able to heat the blanks to 950°C in less than 3 minutes.


Hot stamping facility at the University of Waterloo

Coupled thermo-mechanical FE simulations of the hot stamping process are performed using LS-Dyna as shown below for a predicted and actual hot formed part. Hot stamping material models are used to predict the decomposition of austenite to its daughter phases (martensite, bainite, ferrite and pearlite) during the hot stamping process. The accuracy of the modeling procedure must be validated against the experiments if it is to be used as a design tool in the manufacturing stage of structural components (with tailored properties) that are to be used in vehicles.



A predicted and actual hot stamped part

High Strain Rate Characterization of Tailored Properties


Small-scale quenching experiments are conducted on blanks that are machined into miniature dogbone style specimens. These specimens are then tested in tension at low to high strain rates as discussed in the High Strain Rate Material Behaviour. The forced air quenching apparatus used to quench the blanks at various cooling rates below 30°C/s is shown below along with the temperature vs. time curves for the various cooling rates tested.


Forced air quenching apparatus and the resulting temperature vs. time at various cooling rates

Through micro-hardness measurements and tension tests, it has been shown that the hardness vs. cooling rate follows a linear trend, as does the ultimate tensile strength (UTS) vs. cooling rate (see below). The increase in hardness and UTS are due to an increasing volume fraction of martensite in the martensite/bainite mixture (for cooling rates <30°C/s)


Micro-hardness and UTS as a function of cooling rate

Preliminary high strain rate test results have shown that a Voce type hardening law (along with logarithmic strain rate dependence) can accurately predict the flow stress of the tailored properties associated with a bainitic and martensitic mixture. A single constitutive model is then created to describe the flow stress as a function of strain, strain rate and cooling rate (or hardness). A user defined constitutive model for LS-Dyna is currently under development for application to crash testing of vehicle structures that have tailored properties.


Gleeble Thermo-Mechanical Apparatus


A Gleeble 3500 thermo-mechanical testing system at the University of Waterloo (see below) is used to study the effect of cooling rate and plastic deformation on the tailored microstructures formed during hot stamping. The apparatus can accurately impose both a temperature and deformation vs. time boundary condition, which can be applied to match the experimental conditions. A quench head is used to impose a variety of linear and non-linear cooling rates during the simulation of the quenching process. A width gauge (C-gauge) measures the thermal expansion and dilation of the specimens due to phase transformations. The experimental results (along with hardness and metallographic analysis) are used to study the quenching process parameters and validate the hot forming FE material model.


Gleeble 3500 thermo-mechanical system at the University of Waterloo

Heat Transfer Coefficients During Hot Stamping


Determining the heat transfer coefficient during hot stamping is a part of the hot stamping program at the University of Waterloo. The heat transfer coefficient is critical in determining the cooling rates that are imposed during the hot stamping of parts with tailored properties. These coefficients are an input to the hot stamping FE models and determine the accuracy of the predicted phases that form in the tailored regions. Flat die quenching experiments are conducted to characterize the dependence of pressure, temperature and surface roughness on the heat transfer coefficient. Using temperature readings from thermocouples that are embedded below the surface of the flat die (see schematic below), an inverse heat conduction method is used to infer the surface heat transfer coefficient.


Schematic of the flat die tests used for the inverse analysis of heat transfer coefficients


Contacts


Alexander Bardelcik, PhD
abardelc@uwaterloo.ca

Ryan George, Research Engineer
r3george@lagavulin.uwaterloo.ca

Etienne Caron, PhD
ejfcaron@uwaterloo.ca

Michael Worswick, Professor
worswick@lagavulin.uwaterloo.ca

Mary Wells, Associate Professor
mawells@mme.uwaterloo.ca

Kyle Daun, Assistant Professor
kjdaun@mme.uwaterloo.ca
www.kjdaun.uwaterloo.ca

Back to Top