Abstract
Our interest in multivariate capability indices (MPCI) is based on a case from a thermal spraying process at Volvo Aero Corporation, Sweden, where they wanted to determine the process capability of a three-dimensional variable. The process can be characterized as a high temperature “spray painting” to protect the sprayed surface against, e.g. heat, corrosion and erosion. Different porosity is desirable depending on what the thermally sprayed coating should protect against. There is a relation between the porosity and the in-flame variables and it would be preferable to secure the porosity during spraying instead as of today, afterwards. Since the three-dimensional, in-flame, variables are correlated one way to do this could be to calculate a MPCI of the in-flame variables. There are a number of different MPCIs described in the literature, but for only a handful of these confidence intervals have been derived. In practise the conclusion about process capability must be drawn from a random sample. Hence, confidence or tests for MPCIs are important. Four different available methods for calculating confidence intervals of MPCIs are being reviewed and compared. Current investigation shows issues that need to be solved before the studied methods can be applied more generally in practice.