Elimination of Hot Tears in Steel Castings by Means of Solidification Pattern Optimization PETR KOTAS, CEM CELAL TUTUM, JESPER THORBORG, and JESPER HENRI HATTEL A methodology of how to exploit the Niyama criterion for the elimination of various defects such as centerline porosity, macrosegregation, and hot tearing in steel castings is presented. The tendency of forming centerline porosity is governed by the temperature distribution close to the end of the solidification interval, specifically by thermal gradients and cooling rates. The physics behind macrosegregation and hot tears indicate that these two defects also are dependent heavily on thermal gradients and pressure drop in the mushy zone. The objective of this work is to show that by optimizing the solidification pattern, i.e., establishing directional and pro- gressive solidification with the help of the Niyama criterion, macrosegregation and hot tearing issues can be both minimized or eliminated entirely. An original casting layout was simulated using a transient three-dimensional (3-D) thermal fluid model incorporated in a commercial simulation software package to determine potential flaws and inadequacies. Based on the initial casting process assessment, multiobjective optimization of the solidification pattern of the considered steel part followed. That is, the multiobjective optimization problem of choosing the proper riser and chill designs has been investigated using genetic algorithms while simulta- neously considering their impact on centerline porosity, the macrosegregation pattern, and primarily on hot tear formation. DOI: 10.1007/s11663-011-9617-z Ó The Minerals, Metals & Materials Society and ASM International 2011 I. INTRODUCTION TWENTY years after the introduction of simulation software for foundries into the industry, casting process simulation has become an accepted tool for process and design layout. Metal casting process simulation is used to provide detailed information about the mold filling, solidifica- tion, and solid-state cooling, and with that, also information about the local microstructure, nonuniform distribution of mechanical properties, and subsequently residual stress and distortion buildup. [1–9] Casting sim- ulation tries to use physically realistic models without overtaxing the computer. At the same time, the simu- lations need to give applicable results in the shortest time possible. Because of the multitude of factors affecting the quality of castings and the complex interactions of physics, metallurgy, and casting geometry, empirical knowledge is the main source on which ‘‘optimized manufacturing engineering’’ is based. Foundry simula- tion can quantify experience, but unfortunately, it can test only one ‘‘state’’ or layout. It provides insights into the root causes of problems, whereas conclusions from calculations or subsequent optimization still require an engineer’s interpretation and decision after each simu- lation run. This means that a continuous improvement involves ‘‘trial and error’’—both in reality and in simulation. In recent years, the usage of simulations software has improved and now integrates parallel processing com- puters. It is feasible to calculate numerous versions and layouts in almost unlimited configurations. The advan- tage of having such short calculation times only can be used providing that a computer can analyze calculated variants automatically with respect to predefined objec- tives (e.g., maximum feeding, low porosity, low air entrapment, etc.) and subsequently create new variants and analyze them in the same manner to achieve the optimal solution. By integrating such software for casting process simulation with an optimization algo- rithm, a computer-based optimization tool is established that can determine the optimal values of user-defined design variables, thereby optimizing a given casting process with respect to predefined objectives. [10] Autonomous optimization uses the simulation tool as a virtual test field. By modifying pouring conditions, gating designs, or process parameters, the software tries to find the optimal route to fulfill the desired objective. Several parameters can be changed at the same time and be evaluated independently from each other. Autono- mous optimization tools combine the classic approach of foundry engineers to find the ‘‘best compromise’’ with validated physics. This not only reduces the need for trial runs to find the optimal process window but also allows an in-depth evaluation of many parameters and PETR KOTAS, Postdoc, CEM CELAL TUTUM, Assistant Professor, and JESPER HENRI HATTEL, Professor, are with the Department of Mechanical Engineering, Technical University of Denmark, 2800 Kgs. Lyngby, Denmark. Contact e-mail: pkot@ mek.dtu.dk JESPER THORBORG, Software Development Engineer, is with the MAGMA GmbH, D-52072 Aachen, Germany. Manuscript submitted April 13, 2011. Article published online December 23, 2011. METALLURGICAL AND MATERIALS TRANSACTIONS B VOLUME 43B, JUNE 2012—609