A tribological study of a detonation gun coating of tungsten carbide for use in a subsea gate valve
Detonation gun coatings of tungsten carbide have been widely recognised as one of the most effective anti-wear coatings for oilfield applications. However, very little fundamental tribological information exists for the material, which hinders coating development and the evolution of correct specifications. This study redresses this problem by conducting adhesive, abrasive and erosive tests upon the coating and relating the findings to the coated microstructure. The intention has been to simulate the in-service behaviour of parallel gate valves, which are used primarily to control flow in remote locations where reliability and freedom from maintenance are essential. Although problems with such valves are rare, costs associated with replacement are exceptionally high and therefore a high research priority has been placed on valves of this type. Currently, new designs of valve are tested using a pipe loop rig at BP Research Centre. However, such tests are both expensive and time consuming and with the increasing desire to bring products to market more quickly an alternative is sought. Probably, the area offering most scope for improvement is in material specification of the sealing surfaces and this work sets out to produce a first stage selection procedure for candidate materials. Uniquely, the study has taken one component, systematically categorised its failure mechanisms using non-destructive replication techniques and then reproduced them in the laboratory. The failure analysis has pointed to three-body abrasion, erosion and adhesion being the dominant failure modes and therefore, a suite of tribo-test methods have been developed to replicate them These are namely reciprocating diamond-on-flat, slurry erosion and reciprocating pin-on-plate tests. The material studied was a proprietary detonation gun coating of tungsten carbide, LW45, which is currently the most popular seal facing material specified for gate valves. A conformal contact geometry was chosen for the reciprocating pin-on-plate tests and problems with alignment were overcome by using a pre-test running-in procedure with 1 μm metallographic paste. Wear of LW45 occurring during the pin-on-plate test was not affected by test speed over the range selected, but was highly dependent upon load. Four different categories ranging from minimal wear to catastrophic wear have been identified. Extensive post test analysis using optical and scanning electron microscopy has further classified the failure that occurs into two groups, termed mild and severe. In the mild regime wear occurs by preferential removal of the binder phase, which is minimised on further sliding by protruding carbide particles. Eventually sufficient binder is removed for carbide fall-out to occur, upon which the cycle is repeated. A greater wear volume is produced by the severe wear mechanism which is caused by the interlinking of cracks present within the microstructure of the coating. To ensure operation in the mild regime, continuous sliding under operating pressures of above 7.84 MPa should be avoided. Abrasive wear simulated by the diamond-on-flat test increased with load. However, the failure mechanisms produced were independent of load and consist of a combination of plastic deformation and brittle fracture with plastic deformation representing the rate controlling step in the wear process. Slurry erosion tests have shown that LW45 wears by a brittle erosive mechanism and is therefore best able to resist erosion at low impingement angles. The volume loss per particle impact for LW45 is proportional to the kinetic energy of the impinging particles. The failure mechanism involved the growth of cracks in the microstructure by a fatigue action eventually leading to crack interlinking and material fall out. For all wear conditions, it is suggested that the removal of microcracking from the coating microstructure will lead to significant improvements in wear performance. A simplified design guide has been produced that gives a weighted importance to the various failure modes attributable to the respective tests. A significant improvement in performance was recorded by LW45 in comparison to typical substrate materials such as AISI 410 and Ferralium F255 stainless steels.