Journal of Mechanical Science and Technology 27 (6) (2013) 1815~1824 www.springerlink.com/content/1738(494x DOI 10.1007/s12206(013(0432(0 Determination of the flow stress and thermal properties of ceramic powder feedstock in ceramic injection molding Jihoon Hwang 1 , Sunchul Choi 2 , Seokmoo Hong 2 and Naksoo Kim 1,* 1 Department of Mechanical Engineering, Sogang University, Seoul, 121742, Korea 2 Global Production Technology Center, Samsung Electronics, Suwon, Korea (Manuscript Received August 15, 2012; Revised December 17, 2012; Accepted January 2, 2013) ((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((((  To simulate numerically the material behavior of a ceramic powder feedstock that consist of a two(phase mixture of zirconia powder and polymer binder, a material model is needed that incorporates the change in volume fraction and temperature dependency of viscosity. Heat transfer occurs between the feedstock and the mold during ceramic injection molding (CIM). The feedstock is heavily influenced by thermal properties such as thermal conductivity and specific heat. In this study, three models are proposed to explain the material and thermal properties: a rigid(plastic flow stress model that is dependent on volume fraction and viscosity, a thermal conductivity model, and a specific heat model as a function of temperature. The material parameters in each model are obtained by using the optimization method. Error functions are defined as the differences between the experimental measurements and numerical simulation results. The parameters are determined by minimizing the error functions. The confirmation simulation for each model is conducted by applying cases that are not directly used in the optimization. The results of the confirmation simulation tend to follow the experimental results well, with correlation coefficients exceeding 0.92. The numerical simulation of the CIM process with the determined parameters is compared with the flow behavior of an actual CIM process. Simulation results, such as flow pattern and direction, are in good agreement with the measured feedstock behavior. Therefore, the method for determining the material parameters of the proposed models is feasible. Keywords: Ceramic injection molding; Ceramic powder feedstock; Rigid(plastic flow stress model; Thermal properties; Optimization    Ceramic injection molding (CIM) is a forming process used to inject a mixture of ceramic powder and polymer binder into a mold at 135°C to 200°C. CIM is a complicated process that includes several steps such as kneading, injection molding, de( binding, and sintering. CIM is an efficient process for the high(precision manufacture of large quantities of thin and complex 3D parts. Defects such as weld lines are formed in injection molding when the flow direction of the feedstock separates into two flow lines and then recombines into a single line. High(pressure is required in the CIM process for the in( jection of feedstocks into molds because feedstocks have high viscosities and are sensitive to the changes in volume fraction and temperature. The flow of feedstock is sensitive to tem( perature changes caused by heat transfer, whereas thermal properties such as thermal conductivity and specific heat de( pend on changes in temperature. Thus, defining process condi( tions, such as injection temperature, gateway location, and injection pressure, is difficult. The design of CIM process conditions generally depends on trial and error rather than on a systematic approach. Poor process conditions cause bending and twisting failures after the sintering process of CIM prod( ucts. Thus, establishing a numerical tool is important to simu( late CIM processes, predict feedstock behavior, and determine fragile locations during and after the CIM process. Previous studies have developed simulations for CIM processes. Kwak and Seo (2009) predicted the weld line location and flow pat( tern in a CIM process by using a commercial injection mold( ing simulation software (3D(TIMON) and found that changes in sintering shrinkage depend on flow direction [1]. Soykan and Karakas (2001) studied a new method in which a ceramic powder and binder can be uniformly mixed to produce a ho( mogeneous distribution of the volume fraction after injection molding [2]. Wu and Wei (2004) modeled the rheological behavior with respect to the homogeneity of the feedstock by using the Casson model to obtain the relationship between shear stress and viscosity [3]. Lee et al. (2000) compared a finite element method (FEM) simulation that applies sintering and grain growth models based on diffusion creep with ex( perimental results to analyze the sintering process systemati( cally [4]. Luo et al. (2006) studied the influence of powder characteristics such as particle size in a CIM process [5]. * Corresponding author. Tel.: +82 2 705 8635, Fax.: +82 2 712 0799 E(mail address: nskim@sogang.ac.kr Recommended by Associate Editor Dae(Cheol Ko © KSME & Springer 2013