High-Temperature Flow Behavior of Ceramic Suspensions T.-M. Gabriel Chu* and John W. Halloran* Department of Materials Science and Engineering, University of Michigan, Ann Arbor, Michigan 48105 The objective of this study was to investigate the effects of temperature and solids loading on the viscosity of two non- aqueous ceramic suspensions. In this article, the viscosity of Al 2 O 3 suspensions with 5%–50% solids loading and hydroxy- apatite suspensions with a solids loading of 5%– 40% were measured at temperatures of 25°, 45°, 65°, and 75°C. The high-shear Newtonian viscosity at various temperatures was reduced to a single curve by the reduced viscosity and the temperature-adjusted solids volume fraction of the suspen- sions. The Krieger–Dougherty model, with the intrinsic viscos- ity corrected for particle geometry, was fitted to the data and was observed to provide a satisfactory description to the solids-loading–viscosity data for both suspensions. I. Introduction T RADITIONALLY, ceramic slurries are processed at room temper- ature. Lately, processes for making complicated ceramic parts have been developed by jetting and printing ceramic slurries at elevated temperatures. 1–3 An understanding of the effect of solids loading on the viscosity of the suspensions at elevated tempera- tures is fundamental and essential in these processes. Theories on the effect of solids loading on the viscosity at room temperature have been developed by many researchers. 4–8 Many of these groups used idealized systems that contained monodispersed hard spherical particles and a one-component liquid medium to verify their theories. However, in practice, in ceramic processing, highly loaded ceramic suspensions deviate from the ideal system in several ways. The liquid medium usually contains more than one component, the ceramic particles are not always monodispersed, and dispersants usually are used to modify the interparticle force. Mathematical models on the solids-loading–viscosity relation of these “nonideal” ceramic suspensions are needed. Relatively little literature was found regarding the temperature effect on the viscosity of highly loaded suspensions. Krieger 9 used the relative shear stress and the reduced viscosity to explain the temperature effect on Brownian force. Tsutsumi and Yoshida 10 used the product of the viscosity of the suspension medium and the shear rate to describe the effect of temperature on hydrodynamic force. An important factor in the high-temperature flow behavior of the highly loaded suspension is the effect of thermal expansion of the liquid medium. The flow behavior of a highly loaded suspension is known to be very sensitive to the variation in the solids loading. 4 One possible reason for a variation in solids loading is the temperature change in the suspensions. Because of the large difference in the coefficient of thermal expansion between solid and liquid, a change in the suspension temperature can have a strong effect on the solids loading, which, subsequently, would change the flow behavior. Although the effect of thermal expansion of the liquid phase has been mentioned, 11 no special treatment in this aspect has been noticed. The first objective of this article is to study the effect of temperature on the viscosity of nonaqueous Al 2 O 3 suspensions and hydroxyapatite (HA) suspensions. The intent is to correlate the viscosity to the temperature-adjusted solids loading of the ceramic suspensions at various temperatures. Generally, if a well-dispersed dilute ceramic suspension is prepared from a Newtonian fluid, the flow behavior of the suspension will remain Newtonian; the viscosity will be constant, regardless of the shear rate. However, in well-dispersed highly loaded suspensions, the viscosity of the suspension may or may not change with the shear rate, depending on the magnitude of the shear rate. In most cases, at extremely low shear rates, Brownian motion dominates and the structure of the particles in the suspension remains relatively undisturbed, despite the shearing. The viscosity in this region is independent of the shear rate and usually is called the “low-shear Newtonian limit” of the suspension. As the shear rate increases, the particle structure in the suspension is disrupted; the viscosity of the suspension decreases as the shear rate increases. This region usually is called the “shear-thinning region” of the flow curve. As the shear rate increases further, the particle structure in the suspension becomes grossly oriented; the viscosity of the suspension remains constant, irrespective of the shear rate. This region usually is called the “high-shear Newtonian limit” of the suspension. 4,12 For our research, we have focused on the high-shear Newtonian limit of the suspensions, because this region is most relevant to our processing condition. The second objective of this study is to provide a mathematical description to the solids-loading–viscosity relation for these highly loaded ceramic suspensions. The Krieger–Dougherty model was applied to the solids-loading–viscosity data of both ceramic suspensions, where the solids packing factor and liquid intrinsic viscosity each were corrected for the particle shape. The usefulness of the model was evaluated. II. Experimental Procedure Al 2 O 3 powder (Product A-16, Alcoa, Pittsburgh, PA) with an average size (d 50 ) of 0.4 m was used in this study. The 90th percentile particle size (d 90 ) of this powder was 1.2 m, and the 10th percentile particle size (d 10 ) was 0.14 m (according to data from the manufacturer). Scanning electron microscopy (SEM) (Model S-800, Hitachi, Tokyo, Japan) indicated that the shape of the particles was approximately spherical. The density of the powder was 3.92 g/cm 3 , and the specific surface area of this powder was 9.5 m 2 /g (according to data from the manufacturer). The HA powder used was acicular, with a diameter of 60 nm and a length of 600 nm. The density of the powder was 3.14 g/cm 3 , as measured via helium pycnometry (Model AccuPyc 1330, Micro- meritics Instrument Corp., Norcross, GA.). The Brunauer–Em- mett–Teller (BET) specific surface area of this powder was 17 m 2 /g, as reported in the literature. 13 The liquid medium was a 1:1 mixture of propoxylated neo- pentoglycol diacrylate (PNPGDA) and isobornyl acrylate (IBA). The dispersant used for the Al 2 O 3 powder was a commercial product that was based on quaternary ammonium acetate (Emcol CC-55, Witco Corp., Houston, TX). First, the optimal dose of V. A. Hackley—contributing editor Manuscript No. 188998. Received October 22, 1999; approved March 14, 2000. *Member, American Ceramic Society. J. Am. Ceram. Soc., 83 [9] 2189 –95 (2000) 2189 journal