Fretting studies on self-mated stainless steel and chromium carbide coated surfaces under controlled environment conditions

Abstract

Adhesive wear has been widely accepted as the type of wear which is most frequently encountered under fretting conditions. Present study has been carried out to study the mode of failure and mechanisms associated under conditions where strong adhesion prevails at the contact interface. Mechanical variables such as normal load, displacement amplitude, and environment conditions were controlled so as to simulate adhesion as the governing mechanism at the contact interface. Self-mated Stainless Steel (SS) and chromium carbide with 25% nickel chrome binder coatings using plasma spray and high-velocity oxy-fuel (HVOF) processes on SS were considered as the material for contacting bodies. Damage in the form of plastic deformation, fracture, and material transfer has been observed. Further, chromium carbide with 25% nickel chrome binder coatings using HVOF process on SS shows less fretting damage, and can be considered as an effective palliative against fretting damage, even under high vacuum conditions.

Introduction

Fretting occurs whenever a small amplitude oscillatory movement between two contacting surfaces is sustained for a large number of cycles. The most important mechanical variables that control the fretting process are the normal load and slip amplitude. These two variables characterize the degradation that occurs either in the form of cracking or wear. Cracking is mainly encountered in pre-sliding condition or generally referred as partial slip condition, whereas wear is favored under gross sliding condition. Waterhouse [1] first indicated a correlation between sliding and damage evaluation, as shown in Fig. 1.

Fretting maps provide great help in the identification of contact condition, whether partial slip or gross sliding prevails at the contact location. The concept of fretting map was originated by Vingsbo and Soderberg in 1988 [2], after fretting experiments were performed under ambient condition on low carbon structural steel, an austenitic steel and pure niobium. Later in 1992, Zhou and Vincent [3] proposed independently two kinds of fretting maps, that is, running condition fretting map (RCFM) and material response fretting map (MRFM). The features of fretting loop were also used to characterize the contact conditions prevailing at the contact interface [4], [5]. Occurrence of partial slip condition was identified by elliptical fretting loops, whereas gross sliding condition was identified by quadratic fretting loops. A schematic of typical fretting loop is shown in Fig. 2. The loop can be characterized on the basis of interfacial displacement amplitude (D_s), friction force (Q), slope of the loop (K_l), and energy dissipation (E_d). The interfacial displacement amplitude (D_s) is a function of independent governing parameters, namely, the fretting or displacement amplitude, the normal load, and the elastic slope of the loop, which is actually the combined tangential stiffness of the contact interface and the experimental set up [6].

The mechanism of fretting involves two major processes, viz., adhesion and metal transfer, and material separation [7]. Bowden and Rowe [8] showed that when two like or unlike metal surfaces were pressed together, normally, there was little or no adhesion when the load was removed. This was explained based on two reasons, firstly, oxide films were not readily broken and dispersed, when two surfaces approach each other normally, and secondly, the elastic stresses released on removal of the load were sufficient to break any junction that gets formed. Thus, if steps are taken to remove the oxide films and to eliminate the elastic stresses by annealing, the surfaces do adhere. Tabor [9] shows that if two normally loaded surfaces were subjected to a tangential force, then there is an increase in the adhesion at the contact interface. Bethune and Waterhouse [10] investigated the mechanism of fretting in steels, using a rotating bending fatigue machine, where the fretting was produced by bridges of 0.1% carbon steel clamped on the cold drawn 0.16% carbon steel specimen. Their studies concluded that adhesion rises rapidly, in the early stages as existing oxide films were ruptured, and leads to the formation of weld junctions. The extensive welds formed in this stage were broken by fatigue causing roughening of the surfaces, and resulting in lower adhesion. The roughened surface continues to fret against each other, and contacting asperities produce strong but smaller welds, resulting in an increase of adhesion. Further, the smaller welds suffer failure under fatigue action and the adhesion falls again. The process continues till the loose debris between the surfaces becomes appreciable and prevents further welding. Although their studies concluded that adhesion occurred is the governing mechanism under fretting conditions, but the mechanisms involved in the initiation and propagation of the damage were not apparent.

Several investigators have suggested that the frictional resistance is affected by the properties of the sliding materials. Mokhtar et al. [11] carried out experimental studies to find the correlation between frictional behavior and physical properties of the metals. From their studies, they concluded that strong adhesion occurs in the material having low melting, boiling, and recrystallization temperatures, while hard metals with strong bonds such as nickel and chromium exhibit low adhesion and, hence, low values of coefficient of friction. Metals having large values of elastic modulus, hardness, and resistance to plastic flow reduce adhesion at the contact interface and, thus, result in low values of coefficient of friction.

Material and environment conditions play an important role not only in the increase or decrease of frictional resistance, but also in the degradation mechanism due to fretting. Stainless steel (SS) is often used in nuclear industry, especially in sodium cooled nuclear power plants, because of its excellent mechanical properties under high temperature and irradiation environment, but on the other hand it is characterized as having relatively poor wear and galling resistance. In sodium cooled nuclear power plants, during differential thermal expansion or flow-induced vibration or loading and unloading events, different components move relative to each other [12], and such conditions can be categorized under fretting.

Hard wear-resistant metallurgical coatings on the component structural material can be used to improve tribological properties of the contacting surfaces. Farwick et al. [13] studied wear and corrosion performance of metallurgical coatings of various refractory metal carbides in metallic binders, nickel-base, and cobalt-base alloys and intermetallic compounds such as aluminides and borides. Apart from the evaluation of metallurgical coatings, their studies also evaluated coating processes including plasma spraying, detonation gun, sputtering, spark deposition, and solid state diffusion. Experimental studies indicate that these coatings significantly reduce the dynamic coefficients and self-welding tendencies of type 316 stainless steel, and among several coatings, chromium carbide detonation-gun-applied and nickel aluminide diffusion coatings were qualified for service in sodium cooled reactors. Further, a great reduction of self-welding tendency of SS has been observed when coated with chromium carbide with 15% nickel chrome binder, even at very high temperature. Results also indicate that coating on any one of the wear surface remains as effective as coating both surfaces. Johnson [14] studied various tribological coatings and their processes. His studies concluded that nickel aluminide coatings provide outstanding wear, friction, and corrosion performance in sodium for components that can tolerate high processing temperatures, whereas detonation gun coatings of chromium carbide provide good low friction surfaces on stainless steel components under conditions where most other mechanically bounded coatings usually fail by cracking and spalling. Although most of the coatings were qualified based on the performance criteria for friction coefficients, wear rates, galling resistance, and self-welding resistance, these coatings and the processes are still not evaluated under fretting conditions.

In the present study, the choices of the coating processes are limited such that the substrate must be maintained in a particular metallurgical condition, by limiting the substrate temperature. For example, the processes can be used for reactor components that are required to be maintained at 20% cold worked condition. Based on this, plasma spray and HVOF spray processes were considered for the studies. Plasma spray process is considered to be one of the most sophisticated and versatile thermal spray methods. The process provides a controlled atmosphere for melting and transport of the coating material, thus minimizing oxidation, and the high gas velocities produce coatings of high density. HVOF spray process, also known as detonation gun process, represents the state-of-the-art for thermal spray coatings. The HVOF process uses extremely high kinetic energy and controlled thermal energy output to produce very low porosity coatings that exhibit high bond strength, fine as-sprayed surface finish, and low residual stresses [15].

The present work is focused to evaluate the damage mechanisms involved under conditions where strong adhesion prevails at contact interface. Identification of contact conditions prevailing at the contact interface has been carried out based on mechanical responses, viz., variation of coefficient of friction with number of cycles, running condition fretting loops, and total energy dissipated. Mechanical features were then correlated with morphological features, obtained from scar profile and micrographs.