ICMCTF2003 Session F1/E3-1: Mechanical Properties and Adhesion

The present communication re-examines the basis of a previous model that has been published recently by the author [1], for the computation of the composite hardness of coated systems as a function of the relative indentation depth, applicable to monolayer films. The results that were initially derived from simple geometrical considerations are compared with those that are obtained taking into account the volumes of the plastic zones under in the film and substrate. Such comparisons are conducted on coated systems which include electroless Ni-P (EN) on a 6063 aluminum alloy and EN and an under stoichiometric TiNx deposit on a 7075-T6 aluminum alloy. Furthermore, the consideration of the validity of a linear law of mixtures is also re-examined and the results obtained for the coated systems mentioned above are compared with those derived by assuming that the composite hardness could also be given by a harmonic law of mixtures, rather than a simple linear law. Finally, an extension of the models examined is proposed in order to compute the absolute hardness of the different compounds that comprise multi-layer films, such as, TiAlN (multilayer) and TiAlN+WC/C deposited on different steel substrates.

[1] E. S. Puchi-Cabrera, Surf. Coat. Technol., 160 (2002) 177-186.

Recent experimental studies have shown that nanoindentation load-displacement data obtained with spherical indenters are influenced by stress in the specimen in a manner that may be useful for residual stress measurement. However, the influences are significant only in the elastic-plastic transition, a deformation regime that has largely been ignored in previous theoretical and experimental treatments. In this study, finite element simulation techniques were used to examine spherical indentation in the elastic-plastic transition and characterize the development of the constraint factor (the ratio of the hardness to the flow stress) with penetration depth for a wide variety of materials with different yield strengths and work hardening behaviors. The simulations show that the initial portion of the transition at small penetration depths is essentially independent of the work hardening behavior, apparently because deformation in the subsurface plastic zone is constrained by surrounding elastic material. This is a useful result in that it implies that some material properties may be measured from the indentation load-displacement data without a knowledge of the work hardening behavior. At a depth approximately equal to that at which the plastic zone that escapes to the free surface and plasticity is no longer elastically constrained, the constraint factor diverges and the indentation behavior depends on the work hardening characteristics of the material. Influences of friction between the indenter and specimen were also examined. Implications for mechanical property measurement by nanoindentation with spherical indenters are discussed.

* Research at the Oak Ridge National Laboratory SHaRE Collaborative Research Center was sponsored by the Division of Materials Sciences and Engineering, U.S. Department of Energy, under contract DE-AC05-00OR22725 with UT-Battelle, LLC.

A modified blister test has been developed based on helium ion implantation into the metal substrate prior to the ceramic coating deposition. After a post-deposition annealing blisters are formed by agglomeration of the implanted gas at the ceramic-metal interface. In this way implantation, followed by annealing, can be used to increase the pressure in the blister in a controlled way and induce decohesion. When a microsieve with a periodical pattern of circular pores is used during the implantation, the control over the initial blister size is assured. Two different microsieves were employed in this work with pore diameter of 1.5 and 3 microm respectively. The distance between the center of neighbor pores is twice the pore diameter. Decohesion is non-destructively monitored by means of Positron Beam Analysis (PBA) and Environmental Scanning Electron Microscopy (ESEM). A relation between blistering parameters (for example radius, height) and the adhesion parameters such as the work of adhesion can be obtained.

In this work we present the first results of this test applied to 500 nm thick Chromium Nitride (CrN) films deposited by CVD on polycrystalline copper substrates. The samples were implanted with 30 keV He⁺ ions up to doses of of 3, 4 and 5 x 10¹⁶ cm^-2 and annealed afterwards in vacuum at 973 K for 30 minutes. Experiments on uncoated copper samples will be shown in order to explain the mechanism of helium bubble growth and swelling that causes the creation of blister in the copper surface.

Thermal spraying is an attractive technique of coating manufacturing offering a wide choice of materials and processes. This technique allows many problems of wear, corrosion and thermal degradation to be resolved. Alloy powder can be processed by direct injection in a plasma medium where the powder particles are heated, fused and propelled towards the material to be coated. Particles quench and flatten rapidly forming a stacking of lamella.

Alumina-13%wt titania wear resistant coatings were obtained with this process under atmospheric conditions. In this survey, coating adhesion measured locally with interfacial indentation test was studied as a function of the process operating conditions. In order to recognize the most influencing factors on adhesion and the corresponding evolutions with few experimental sets a novel technique of optimization was used. It is based on the Artificial Neural Networks (ANN) principal considering training and test procedures permitting a prediction of the property dependence on experimental conditions.

This study points out the following conclusions: Adhesion was proved to be sensitive to coating thickness, porosity and test load. It depended largely on most parameters modifying the temperature gradient between the substrate and the coating in addition to the in-flight particles thermal status. These effects were quantitatively demonstrated and predicted with optimized neural network structures.