Design of experiments and surface treatments - Quantitative study of effects and interactions

Surface treatment (ST) consists of functionalizing the surface of a material by superficial addition of material (metallic and/or organic coating) or by transformation of the substrate (conversion, dissolution, etc.) to give it a technical function (anti-corrosion, anti-wear, magnetic properties, etc.) and/or aesthetic function (appearance, color, etc.). The industrial sectors concerned are as numerous as they are varied. Examples include jewelry, connectors, construction, the transport industry (aeronautics, automotive, railways, etc.), the medical sector, etc.

Surface treatments are typically carried out in three different ways:

• aqueous (TSVH) ;

• igneous (TSVI) ;

• dry (TSVS).

They all involve three main players:

• substrate (metal, ceramic or polymer) ;

• treatment medium (aqueous, molten or gaseous) ;

• process such as :

  • attachment, barrel, parade, stamp for TSVH,

  • CVD, PVD, thermal spraying, etc. for TSVS.

Our aim is to use examples to encourage practitioners in the TS sector to adopt a day-to-day approach geared towards the implementation of the Design of Experiments (DOE) methodology, all the more so as the requirements of clients are very diverse and varied. To this end, we have deliberately chosen to deal with plans with different, and often complementary, purposes.

This article deals with the screening of factors and the quantitative determination of their effects and interactions. The [M 1 429] website will be devoted to the search for the optimum and the formulation of mixtures using response surface methodology.

In order of appearance, full factorial, fractional factorial, Taguchi, Plackett & Burman and supersaturated designs will be analyzed, based on examples taken from the scientific literature. Readers will appreciate that it would have been non-confidential to consider industrial examples. Incidentally, we have reprocessed data kindly supplied by the authors using NEMRODW software.