Today's heat treatment software can predict distortion, residual stress, hardness, and metallurgical phase percentages during hardening of both carburized parts and through-hardened steel parts as well as assess the effect of the quenching method.



For critical engineering applications, heat treatment of steel components is often used to produce the combination of hardness, strength, toughness, and ductility that produces high performance capability. However, accompanying these heat treatment benefits is the undesirable side effect of distortion. Control of residual stress is a growing issue as the operating stress in critical parts continues to increase, making improvement in properties such as fatigue strength important. The achievement of desirable surface compressive stresses in the hardened and dimensionally correct part is a goal of all heat treaters. Past practices used by heat treaters and designers in determining the proper process routing have included experience-based rules, statistical models, and costly shop trials. By combining mechanics, thermal analysis, and phase transformation behavior with the powerful numerical methods of finite element analysis, new software-based tools have emerged for heat treat simulation.

DANTE® (Deformation Technology Corporation, Cleveland, OH, www.deformationcontrol.com) is one of several software packages that are being used by industry, R&D laboratories and academia to investigate and improve the heat treatment of precision steel parts, with benefits including increased product reliability and performance, reduced finishing costs, and decreased testing and development time. This heat treatment simulation software predicts distortion, residual stress, hardness, and metallurgical phase percentages during hardening of both carburized parts and through-hardened steel parts. It is also used to assess the effect of the quenching method, allowing comparisons between different quench processes, such as immersion quenching in oil, salt, polymer or brine, high pressure gas quenching, intensive quenching and press quenching. Tempering effects, including the size and residual stress changes due to tempering, are also predicted. The resident database contains properties for many commonly used steels, allowing the software to be used to quantitatively assess the effect of alloy substitution on part hardness.

Developed under a Cooperative Research and Development Agreement (CRADA) between a group of automotive-related companies and the Department of Energy, the project addressed the common problem of distortion due to phase transformations that occur during heat treatment of precision, critically stressed components. An equally important problem was control of residual stress so that part performance could be extended by heat treatment, i.e. achieve residual surface compression and not surface tension in the heat treated part. Using the commercial software package ABAQUS/STANDARD® (Abacus Inc, Pawtucket, RI, www.abaqus.com) as the finite element solver, software subroutines were developed that captured the metal physics of heat treatment. As these subroutines were being developed to address the diffusive and martensitic phase transformations that occur during heating and cooling of steel components, supporting material and process data were developed from many experiments, both for use in simulations and to validate model accuracy. The end result was an accurate simulation tool that offered heat treaters and designers an alternative and cost effective method for modifying existing heat treat processes or implementing new product designs and heat treat processes instead of the traditional method of shop floor try-outs. A CRADA project team member, Deformation Control Technology, Inc. (DCT) was charged with commercialization of the software, expansion of the material and process databases, and enhancement and technical support of the software package.

Heat treatment simulation software enables designers and heat treaters to understand the impact of the metallurgical events that occur during heat treatment. The ability to graphically view the internal distribution of phases as they evolve and the changes in dimensions and internal stress of the part throughout the entire heat treat process schedule is a basic key to developing this improved understanding. Detailed graphics of carbon profile development, temperature, distribution of austenite, ferrite, pearlite, bainite and martensite, and contour plots of the internal stress state and dimensional change can be readily generated and even animated. These graphics enable the user to visualize and subsequently understand why a part moves from the green shape to a particular final shape due to heat treatment. In addition, the software provides an accurate numerical method that can be used to anticipate how different steel alloys or process variations affect the final part shape, residual stress state, and hardness.

Simulation has a side benefit in that it forces the process engineer or heat treater to document the actual process in detail. Every step of the process is important. Time and temperature, in concert with the steel alloy, determine what events are taking place locally within a part. Well known process data include heat-up rate, austenitizing and carburizing temperatures and time at these temperatures, the quench media and it's temperature and level of agitation, residence time in the quench tank, and so forth. Less often, documented steps include time of transfer from furnace to quench tank, speed of immersion into the tank (in the case of immersion quenching), actual flow rate of quench media throughout the racked parts, and the local cooling rate within any part. If the simulation is to be of the highest accuracy, the entire process must be documented. This documentation procedure forces the user to think in detail about the heat treat process, which in itself is often enlightening. In many cases, the actual conditions have not been quantified, and here simulation can be used to assist in understanding the role of each process step and its impact on the final part properties.

A major limitation of all modeling methods is the scarcity of accurate data, and this was true initially for heat treat simulation. DCT has concentrated on this area and has expanded the number of steel alloys resident in the DANTE® material database. The methods used to gather the necessary data have included searching the public domain, formation of alliances with universities, companies, national laboratories, and consortia, and both funded and non-funded experiments. For example, through collaborative efforts an ASTM specification for quantitatively measuring steel phase transformations has been developed.[1] To aid in continued enhancement of the material and process databases, a set of software tools has been developed based on optimization methods that process data from standard tests. Using these utilities, published continuous cooling transformation (CCT) diagrams, isothermal transformation (TTT) diagrams, or raw data generated using ASTM Standard A-1033, the phase transformation parameters stored in the software's kinetics database can be determined. Similarly, standard temperature and strain rate controlled tension and compression test data can be used to determine the parameters for the mechanical model. The test data required is similar for any commercially available software package but how the data is used is unique and usually proprietary to each.

Simulation can take a heat treated part and determine the effects of heat up, carburization/soak, quenching, and tempering on dimensions, microstructure (phase distribution) and stress state. Other quench methods that can be simulated include quenches that use fixtures, presses, sprays, or high pressure gases. Sensitivities to variables can also be determined so that heat treaters can focus on changes that reduce processing costs and/or improve the final product. The capability of an accurate simulation tool to perform virtual experiments where a number of process variables, including input material, the use of fixtures, and quench process variations, can be assessed is a powerful method for process optimization and cost minimization.

Examples of how DANTE® has been applied to reduce costs and improve the heat treated product include:

  • Reduced carburization time, which translates directly into reduced energy costs. This was achieved while performance life was actually increased by also modifying the quench practice
  • Substitution of a lower cost 15XX series steel for a higher cost 43XX series steel, again with no loss in hardness or strength. The combination of accurate process data and phase transformation kinetics data provided the simulation accuracy needed to compare the heat treat response of these steels for this particular application.
  • Avoidance of cracking in induction hardened components around particular geometric features by modifying the cooling practice. While direct simulation of the induction heating process was not completed, the local temperature histories induced during heating were inputted to drive the simulation so that austenite formation and then decomposition during spray quenching was simulated. By monitoring the predicted maximum stress level at critical locations, the process was modified to avoid the conditions that caused cracking while still achieving the desired hardness.

Other examples are presented below.

Example 1: Improving Part Performance

A project funded by the Aviation Applied Technology Directorate has a goal of increasing the bending fatigue life of helicopter transmission gears through heat treatment process modification [2]. Software simulations predicted that replacing the conventional oil quenching process with Intensive Quenching [3] would promote a significant increase in favorable residual surface compressive stresses in the gear root (Fig. 1). This improved compression was shown to be achieved in carburized test bars, and the bending fatigue life of these bars was improved beyond the goal of 25%. Simulation was used to investigate the sensitivity of the quenching process to quenchant agitation levels and flow directions such that production quenching could be tuned to optimize the magnitude of residual compression. Furthermore, the predicted residual stress state was carried forward into a fatigue model such that the combined state of stress due to heat treatment and cyclic loading could be applied to predict bending fatigue performance of the gear tooth. The minimum principal stress predicted for oil quenched gears and the stress state at peak loading in the single tooth bending fatigue test (Fig. 2) were determined.

Example 2: Assessing the Effect of Quench Method on Distortion

Simulation was applied to investigate the possibility of using a modified free quench instead of a roller quench as a cost savings option for hardening a marine crankshaft. The result shows the predicted non-uniform length displacements and resultant bending of the crankshaft (Fig. 3) from which it was concluded that the modified process would not maintain acceptable straightness. While a knowledgeable heat treater may arrive at this conclusion based on experience, the simulation gave quantitative results to predict by how much the process being assessed missed the desired straightness specification. In addition, simulation provided an understanding of why the modified free quenching did not work since differences in evolution of phases along the length of the crankshaft were responsible for distortion. With this knowledge, alternative fixturing methods could then be explored efficiently by simulation.

Example 3: Distortion of Helical Spur Gears

Quench hardening of carburized gears is a costly problem if size and shape change cannot be anticipated. Correcting for distortion requires additional processes such as straightening or grinding, and this adds considerable expense in terms of stock allowance, energy usage, labor, tooling and equipment. For example, tooth unwind of helical gears is a significant problem. Software simulation has been applied to accurately predict tooth unwind (Fig. 4) during immersion quenching. With simulation, the proper green shape that will unwind to the desired final shape and size after hardening can be determined, and the undesirable re-work costs can be avoided.

Example 4: Avoidance of Cracking during Quenching

A truck output shaft was cracking during quench hardening. Simulations showed that during the quench high tensile stresses were developed temporarily at edge locations around oil holes (Fig. 5) and these locations were associated with the cracking problem. Using simulation results for guidance, the quench practice was altered to limit the magnitude of the tensile stress developed in these locations, and cracking was avoided while hardness and dimensions were achieved. Additional simulations investigated the possibility of substituting lower cost, leaner alloy, steel grades in place of the specified grade. Further cost savings were achieved through alloy substitution with no sacrifice in hardness.

Summary

Heat treatment simulation technology has matured to where users can effectively apply it to improve heat treat processes and part performance, while reducing cost. Whether faced with a current problem, investigating a new process, or preparing a heat treat schedule for a new part with difficult to meet requirements, process engineers, heat treaters and designers can confidently and creatively use these new software tools. As the future demands increase for faster and more cost effective changes in heat treat practices, reduced re-work and finishing costs, and improved part performance, the use of heat treatment simulation will continue to grow, just as it has for casting and forging technologies. IH