Surface quality prediction in FDM additive manufacturing

Additive manufacturing technologies produce objects which present a characteristic surface texture. This is an inevitable and systematic error and has a predictable shape feature which depends on certain process parameters. In order to reduce manufacturing costs and obtain the best results from the point of view of quality, it is essential to predict this error in advance and choose the process parameters which minimise it. For the purpose of measuring the surface quality, the index R (a) (ISO 4287, 1997) is used in the related literature. In this paper, it is first demonstrated that the use of roughness parameters in FDM-manufactured surfaces is not adequate to quantify the surface error. The parameter P (a) (ISO 4287) is proposed as a more appropriate index to evaluate the surface quality; it is investigated and critically analysed in comparison with the R (a) index. A new original model to predict P (a) for FDM-manufactured surfaces is presented. The model prediction is compared with experimental data and with the estimation performed by the models described in the literature, within the limits of their capability to predict P (a).