Estimation of shrinkage for near net-shape using a neural network approach

A neural network approach is presented for the estimation of shrinkage during a hot isostatic pressing (HIP) process of nickel-based superalloys for near net-shape manufacture. For the HIP process, the change in shape must be estimated accurately; otherwise, the ®nished piece will need excessive machining and expensive nickel-based alloy powder will be wasted (if shrinkage is overestimated) or the part will be scrapped (if shrinkage is underestimated). Estimating shape change has been a very dif®cult task in the powder metallurgy industry and approaches range from rules of thumb to sophisticated ®nite element models. However, the industry still lacks a reliable and general way to accurately estimate ®nal shape. This paper demonstrates that the neural network approach is promising to estimate post-HIP dimensions from a combination of pre-HIP dimensions, powder characteristics and processing information. Compared to nonlinear regression models to estimate shrinkage, the neural network models perform very well. Furthermore, the models described in this paper can be used to ®nd good HIP process settings, such as temperature and pressure, which can reduce operating costs.