On the Interplay Between Adhesion Strength and Tensile Properties of Thermal Spray Coated Laminates—Part I: High Velocity Thermal Spray Coatings

Abstract

Adhesion of thermal spray (TS) coatings is an important system level property in coating design and application. Adhesive-based pull testing (ASTM C633) has long been used to evaluate coating/substrate bonding. However, this approach is not always suitable for high velocity spray coatings, for example, where adhesion strengths are routinely greater than the strength of the adhesive bonding agent used in the testing. In this work, a new approach has been proposed to evaluate the adhesion of TS coatings. A systematic investigation of the effects of substrate roughness on both the uniaxial tensile yield strength and traditional bond pull adhesive strength of HVOF Ni and Ni-5wt.%Al, as well as cold-sprayed Ni-coated laminates revealed a strong correlation between these two test methodologies for the respective materials and processes. This approach allows measurement of the adhesion response even where the adhesive method is not applicable, overcoming many of the issues in the traditional ASTM C633. Analysis of cracking patterns of the coatings after 10.5% strain was used to assess the adhesion and cohesion properties. The mechanisms which determine the load transfer between the substrate and the coating are also briefly discussed.

Appendix

The stress distribution in a single coating segment was estimated by a one-dimensional “shear-lag” model. The coating/substrate system is schematically shown in Fig. 9, in which Cartesian coordinates are used. The coating is bonded to the substrate at y = 0, and the free surfaces of the substrate and the coating are located at y = −s and y = h, respectively. Origin (x = 0, y = 0) is located at the center of the segment. The maximum normal stress (σ_max) in a coating segment can be calculated by:

$$\sigma_{\hbox{max} } = \frac{{1 - v_{\text{c}} v_{\text{s}} }}{{1 - v_{\text{c}}^{2} }}E_{\text{c}} v_{\text{s}} + \sigma_{\text{res}}$$

where v is the Poison’s ratio, E the elastic modulus, and $\sigma_{\text{res}}$ the residual stress in the coating. The subscripts s and c denote the substrate and coating, respectively. According to Frank et al. (Ref 34), the stress distribution along the tensile direction (x) and evolution in the coating segment follows:

$$\sigma \left( {x/L} \right) = \sigma_{\hbox{max} } \left[ {1 - \frac{{\cosh \left( {\alpha x} \right)}}{{\cosh \left( {\alpha L/2} \right)}}} \right]$$

where,

Fig. 9

A schematic showing modeling of periodic coating cracking with a segment length/cracking spacing L

$$\alpha = \sqrt {\frac{3}{{2sh\left( {1 + v_{\text{s}} } \right)}}\left( {\frac{h}{s} + \frac{{\left( {1 - v_{\text{c}}^{2} } \right)E_{\text{s}} }}{{\left( {1 - v_{\text{c}} v_{\text{s}} } \right)E_{\text{c}} }}} \right)}$$

Because both sides of the substrate are coated in this study, the half thickness of the substrate is taken as s.

The interfacial shear stress, τ is governed by:

$$\tau = \frac{{{\text{d}}\sigma \left( {x,L} \right)}}{{{\text{d}}x}}h.$$