Heat treatment is an indispensable step in the manufacture of steel products, as mechanical properties such as hardness, static, and dynamic strength and toughness are selectively controlled by deliberate manipulation of the chemical and metallurgical structure of a component. However, apart from the desired effects, the heat treatment process can be accompanied by unwanted effects such as component distortion, high material hardness, low material strength, a lack of toughness—which can lead to crack formation—and inadequate hardness depth, which can lead to fatigue failure. Therefore, success or failure of heat treatment not only affects manufacturing costs but also determines product quality and reliability. Heat treatment must therefore be taken into account during development and design, and it has to be controlled in the manufacturing process. Part designers and heat treatment practitioners are looking for process feasibility, a specific microstructure fitting to the in-service requirements, minimum part distortion, and proper distribution of residual stresses.
In simulation-based design and manufacturing it is desirable to calculate the effects of heat treatment in advance and to optimize them by varying materials and workpiece shape. Once the part shape is designed it is of utmost importance to make sure that the heat treatment process is correct and that the process window is safe against process parameter variation. With the aid of the finite element analysis software SYSWELD, such calculations can be carried out for all generally applied heat treatment processes, taking all significant physical effects into account. Thus, the part designer/heat treatment practitioner can have a deliberate influence on minimizing manufacturing costs and optimizing product reliability and quality.
SYSWELD is a powerful tool that can be used to judge the heat treatment process on an actual part and efficiently provide answers to these basic questions:
• Is the selected heat treatment process feasible?
• Is the selected steel feasible?
• Is the selected quenching media suitable?
• Is the process window safe against process tolerances?
• Is the part hard where it should be hard?
• Is there any crack risk occurring during the process?
• Are the obtained distortions acceptable?
• Are the residual compressive stresses high enough and well positioned?
Heat treatment is a complex process involving heat transfer, and mechanics, including phase transformation. With the help of the heat treatment advisor, the set-up of a numerical computation is extremely fast. This does not mean that the simulation engineering is simple. The use of a user-friendly, intuitively driven graphical interface does not change the fact that the physics behind a heat treatment simulation is complex. Heeding the adage from Albert Einstein that “One should never do too much, but never less than necessary” to avoid obtaining incorrect results from an incorrect model, SYSWELD provides detailed training for heat treatment simulation that covers real-life situations.
Software and Applications
Dedicated solutions for both heat treaters and part designers can be obtained. Heat treaters focus on the feasibility of the heat treatment process and need answers to their questions instantly. Dedicated solutions are available that fit the needs of a heat treatment job shop. Part designers focus on trying to determine the optimum balance of cost, part shape, material, and heat treatment process. In this case, dedicated solutions are available that provide meshing and computation capabilities.
CAD Data Import and Export
Visual mesh for heat treatment provides graphical modeling capabilities for manipulating finite element meshes. Native CAD data is imported, automatically cleaned, and meshed by a batch-meshing algorithm dedicated to heat treatment applications (Figure 1). Simulations are performed on the real geometry. It is not necessary to work with simplified objects.
Specific technical capabilities are provided for the finite element modeling of the heat-treated structure. The demand for high quality in computed results requires a refined layered mesh from the surface through a few millimeters of thickness of the part. A guided layered mesh generator is available for two-dimensional (2D) and three-dimensional (3D) structures (Figure 2-3), which drastically reduces the time to mesh parts while offering high quality finite element models. The group concept allows simple, complete interfacing to any existing meshing tool and, therefore, the definition phase of the numerical problem is extremely short and simple.
In cases where information for a specific quenching medium is not yet accessible in the SYSWELD database, it is necessary to adjust the convective heat transfer coefficient of the quenching medium as a function of the temperature (Figure 4). Measurements of cooling rate and temperature over time performed using ISO or JIS probes are usually available through the vendor of the quenching media. However, the quality-assured European and Japanese measurement specifications have been included in the database. By performing a few simple recursive computations, the convective heat transfer coefficient of the quenching media can be easily evaluated for the defined specimen (Figure 5). The values obtained provide a good starting point to compute good behavior tendencies on the real parts. The higher the BIOT numbers typical for the real part, the more precise the computations using the evaluated heat transfer coefficient will be. The lower the BIOT numbers, the higher the sensitivity of the computed results against variations in the convective heat transfer coefficient depending not only on temperature, but also on space coordinates. The adjustment has to be done only once for each quenching media, and the results stored in a database for future use.
Fitting the Martensite Transformation
In cases where information is not yet accessible in the database it is necessary to adjust—especially for case hardening simulations—the martensite transformation depending on the carbon content, the martensite start temperature depending on the carbon content, and the retained austenite proportion at room temperature depending on the carbon content (Figure 6). Martensite start temperatures and the retained proportions of austenite usually are known. Based on this data, fine-tuning of the parameters relationship to the Koistinen-Marburger Law—which describes the martensite transformation by mathematical means—can be performed using the SYSWELD PHASE module. The adjustment has to be done only once for each steel. The results can be stored in a database and are then accessible for further computations.
Fitting the Continuous Cooling Diagram
It is necessary to adjust the continuous cooling transformation diagram of the steel if specific information is not available in the database, extracting basic parameters from an ITT (isothermal temperature transformation) diagram and parameters for the fine-tuning from the CCT (continuous cooling transformation) diagram. For numerical reasons, it is preferable to describe the cooling behavior of steel by differential equations rather than by pairs of temperature-proportion values. Those differential equations have been defined, for example, by Johnson-Mehl-Avrami and Leblond and contain phenomenological parameters that have to be adjusted individually for each CCT diagram. Using the PHASE module and the ITT/CCT display tool, the adjustment of a CCT diagram is straightforward and simple (Figure 7). The adjustment has to be done only once for each steel and the results stored in a database for future use. The major steels used in heat treatment are already available in the database.
Material Thermal and Mechanical Properties
The thermal, metallurgical, and mechanical material properties of a heat-treated steel depend on temperature, phases, and carbon content. A comprehensive material database includes the major steels that are used for case hardening, surface hardening, and through hardening. The values in the database are average values extracted from experiments and literature; missing values have been completed by best simulation engineering practice. In addition, steel properties depend on the manufacturer, the year, and the country, etc. Therefore, the material properties represent an average material that will give good tendencies; the data, in any case, do not fit precisely to an individual steel.
Simulation Setup and Solution
A graphical user interface (heat treatment advisor) allows an intuitive and process-driven methodology to set up simulations (Figure 8). Once a dedicated project is defined and stored then parts, process, and material parameters can be exchanged within the project and a computation of a variant can be started in less than one minute.
With the help of the advisor, case hardening and through hardening processes can be fully defined. In case of surface hardening, a few additional simple operations with the standard capabilities of the software are needed to adjust the energy input through the surface.
Heat treatment problems are solved automatically, covering all related complex mathematics and material physics. Depending on temperature, phase proportions, and proportion of chemical elements, thermal and mechanical properties are computed, including large strains. Isotropic and kinematic hardening (including phase transformations), transformation plasticity, nonlinear mixture rules for the yield stress of phases, phase-dependent strain hardening, restoring of strain hardening during diffusion-controlled phase transformations, melting and solidification of material, material properties depending on temperature, phases and proportion of chemical elements, and all features dedicated to the methodology of finite elements are taken into account. The user need not be familiar with the mathematics involved to perform heat treatment computations, but only needs to load the project and to start the solver (Figure 9).
Multi-physics processing capabilities provide instantaneous process information for the evolution of:
• Temperature field
• Heating and cooling rates
• Metallurgical structure of the material
• Yield stress of the modified material
• Plastic strains
A variety of techniques for reviewing process results include:
• Contour plots
• Iso-lines and iso-surfaces
• x-y diagrams
• Symbol plots
• Numerical presentation
• Cutting planes
Examples of some of these capabilities are shown in Figs. Figure 10-17. The system also provides the capability to review movies of the step by step evolution of results on the surface or through the structure. The simultaneous display of the evolution of results gives a deep understanding of process and computed results.
The Jominy test is implemented as a predefined, ready-to-run simulation project in SYSWELD. The user defines the chemical composition of the steel, and the computation of the Jominy test is done automatically. The most important results—e.g., hardness profile—are displayed at the end of the computation (Figure 18).
The Jominy test is key to a precise heat treatment simulation. When the computed hardness coincides well with the measured hardness it ensures that the CCT-diagram of the steel being examined is numerically well implemented for a full range of possible cooling rates. In case of discrepancies the CCT diagram can be modified to precisely meet the measured hardness profile.
Because the formulas used for the hardness computation are empirically approved, existing CCT diagrams can be tuned following hardness measurements. Based on the optimized CCT diagram the core hardness of complex parts can be precisely predicted, which is of utmost importance for the lifetime of parts and components under dynamic loads.
1) Denis S., et. al., Stress/phase transformation interactions: basic principles, modelization and their role in the calculation of internal stresses, Proc. Int. Symp. on the Calculation of Internal Stresses in Heat Treatment of Metallic Materials, Vol. 1, p 157-190, Linkoping (Sweden), 1984
2) Bergheau J.M. and Leblond J.B., Coupling between heat flow, metallurgy and stress-strain computations in steels – The approach developed in the computer code SYSWELD for welding or quenching, Proc. 5th Eng. Found. Conf. on Modeling of Casting, Welding and Advanced Solidification Processes Davos (Switzerland), 1990
3) Leblond J.B. and Devaux J.C., A new kinetic model for anisothermal metallurgical transformations in steel including effect of austenite grain size, Acta Metallurgica, Vol. 32, No.1, p 137-146, 1984
4) Fernandes F., Modélisation et calcul de l’évolution de la température et de la microstructure au cours du refroidissement des aciers, Thèse de Doctorat de l’INPL, Nancy, 1985
5) Leblond J.B., Mottet G., and Devaux J.C., A theoretical and numerical approach to the plastic behavior of steels during phase transformation, I: Derivation of general relations, II: Study of classical plasticity for ideal-plastic phases,” Jour. of the Mech. and Phys. of Solids, Vol. 34, No.4, p 395-432, 1986
6) Leblond J.B., Devaux J., and Devaux J.C., Mathematical modelling of transformation plasticity in steels, I: Case of ideal-plastic phases, II: Coupling with strain-hardening phenomena,” Int. Jour. of Plasticity, Vol. 5, p 551-591, 1989
7) Fortunier R., et. al., Récents développements dans la simulation numérique des traitements thermochimiques des aciers, 9th Int. Conf. on Heat Treatment and Surface Engineering, Nice (France), 1994
8) Großer Atlas Schweiß-ZTU-Schaubilder: Fachbuchreihe Schweißtechnik, DVS-Verlag GmbH Düsseldorf, 1992
9) Dieter Liedke, Rolf Jönsson: Wärmebehandlung, Expert Verlag, 1996
10) Karl Heeß und 14 Mitautoren: Maß- und Formänderung infolge Wärmebehandlung, Expert Verlag, 1997
11) Ruth Chatterjee-Fischer und 8 Mitautoren: Wärmebehandlung von Eisenwerkstoffen, Nitrieren und Nitrocarburieren, Expert Verlag, 1995
12) Atlas zur Wärmebehandlung der Stähle, herausgegeben vom Max-Planck Institut für Eisenforschung, in Zusammenarbeit mit dem Werkstoffaussschuss des Vereins Deutscher Eisenhüttenleute, Band 2 von Adolf Rose und Hans Hougardy, Verlag Stahleisen m.b.H., Düsseldorf
About the Authors
Dr. Frederic Boitout, Dr. Damian Dry, Yogendra Gooroochurn, Philippe Mourgue, and Harald Porzner are with the ESI Group, Paris, France. For more information contact Olivier Morisot at (248) 203-0642 or firstname.lastname@example.org. Visit online at [www.esi-group.com].