Scratch adhesion testing of coated surfaces -Challenges and new directions

摘要

The scratch test is extensively used, in industry as well as in research laboratories, to assess the adhesion strength of coated surfaces. The test consists of drawing a diamond stylus across the specimen surface under increasing normal load until a failure event occurs. The load at which the failure event occurs is called the critical normal load. Only some of the observed failure events, however, are related to coating detachment at the coating/substrate interface and are therefore relevant as a measure of adhesion. Other failure events such as cracks and cohesive damage within the coating or substrate clearly cannot be used to assess adhesion properties, but may be equally important to determine the tribological behaviour of a coated component. Indeed, the scratch test is increasingly being regarded as a repeatable tribological test to assess the mechanical integrity of a coated surface. In spite of its widespread use, however, the results, i.e. critical load values, obtained from scratch testing may thus reflect a variety of failure events. Since these are not always clearly defined when explaining results, the scratch test is often held in low esteem. Much of this confusion could be avoided and the value of the scratch test would be greatly enhanced, if the scratch test procedures were standardised. In order to gain an increased understanding of the possibilities and shortcomings of the scratch test method a European EC SMT project on "Scratch Testing of Coated Surfaces" was established in 1995. The present paper reviews the main results of the project and discusses possible future directions for the scratch test method. The sensitivity of the scratch test method to various test parameters (ambient test conditions, cleaning procedures, styli properties, and transfer layers originating from preceding scratches) has been assessed. As a major step forward, an atlas of scratch test failure modes was prepared to enable users of the scratch test to assign critical load values to agreed failure events. In addition, an extended round robin exercise has been conducted to evaluate the statistical scatter in scratch test results. The main conclusion of this work, however, is that uncertainties in the Rockwell C stylus tip shape constitute the major error source for the scratch test method.