Effect of moulding pressure on sink-mark defects of injection-moulded ceramics

Abstract

Injection moulding is an attractive method for fabricating ceramic articles. The advantages include high precision forming in dimensional control and the freedom in achieving complex shapes. A prerequisite in obtaining moulded ceramics with an accurate dimensional tolerance is that the moulded products should be free from any shrinkage defect. It has been pointed out that non-uniform shrinkage in injected moulded ceramics often results in defects such as voids, sink marks, curve deformations, weld lines and even cracks [1, 2]. The shrinkage defects are reported to be closely related to moulding parameters and sample size, as well as to the blend compositions of compounded mixtures [3]. However, the details about the relative significance of each potential factor and the interactions between factors remain unclear. This letter is therefore aimed at providing such information by designing a statistical L162 experiment to investigate the dominant factors for the formation of sink marks on injection moulded ceramics. Other shrinkage-related defects will be discussed elsewhere. The injection moulding process is often divided into filling, packing and cooling cycles [4]. Thermal contraction is most pronounced at the cooling stage, and the moulding ceramics are subject to nonuniform volume shrinkage. It is plausible that at least two conditions must be met for the formation of sink marks. Firstly, the channel (such as the sprue) which feeds fresh molten suspension into the shrinking mould core to compensate for the volume reduction solidifies earlier than the core. Secondly, the wall surface of the moulded sample is still soft as the core contracts, i.e. voids will form instead of the sink mark if the wall surface hardens. Hunt et al. [5, 6] have suggested that buckling of the skin of moulded bars arising from non-uniform shrinkage and local difference in the coefficient of heat transfer are two possible causes for the formation of the sink mark. Separation of the moulding ceramics from the mould wall leads to local temperature inhomogeneity and has been suspected to be the main cause [3]. Undoubtedly, solidification-related parameters should be the main focal point for tackling the underlying mechanism. However, process parameters are often interdependent in injection moulding; in fact, the interaction effect may end up influencing the overall behaviour. We therefore conducted a two-level, 15factor designed experiment to account for the most probable individual factors as well as interaction parameters. The controlling factors, as well as the selected levels, are shown in Table I. The ceramic used in this study was HSY3.0 zirconia powder with an average particle size of about 0.25 im. Low molecular weight organic vehicles composed of paraffin wax and vinyl acetal polymer in weight ratios of 50:50 and 80:20 (as shown in Table I) were used as major and minor binders, respectively. Ceramic blends with solid loading of 45 and 65 vol% were prepared, compounded and kneaded through an extruder (model 70-20 vex-6, KCK Industrial Co.), followed by injection moulding. A Battenfeld BA 250=50 CDC with a replaceable wear resistant barrel liner was used with barrel temperature series of 70–120–140– 130 8C (or 180 8C) from feed to nozzle. The screw diameter of the injection moulding machine was 18 mm. The mould temperature was controlled at either 40 or 60 8C by a water-circulating HB-Therm CW90. The nominal injection pressure was set at 80 MPa. The holding pressure varied at either 80 or 190 MPa and pressure decay time was controlled at 10 or 30 s. The moulded bars of dimensions 4 mm 3 5 mm 3 60 mm were prepared, and stored in a controlled humidity of 55–60% relative humidity (RH) at room temperature for three days, and the sink marks of the green bars were inspected visually or under polarized microscopy.