Analysis of Computational Methods for the Treatment of Material Interfaces

Abstract

The U.S. Air Force has a test facility at Holloman Air Force Base that specializes in the field of very high speed impact testing. This Holloman AFB High Speed Test Track (HHSTT) is currently working to increase the speed of their test vehicle to above Mach 10. The vehicle runs on rails and carries a test article that impacts a stationary target downrange. In the past, the HHSTT has experienced serious problems with the material interaction between the test vehicle and the rail upon which it rides. The test sled's are made from VascoMax 300 (high strength steel) and they run along a set of rails constructed of 1080 Steel. As the test sled's speed increases into the Mach 8.5 range, the shoes have tended to gouge the rails and start a process that can result in catastrophic failure. In the tests that do not structurally fail, the rails and shoes are damaged by this high- speed impact of the shoe and the rail. In order to model the physics present within this high- energy impact event, the authors have chosen to use the CTH code. Because high velocity impact creates a situation in which there is coupling between strain, strain-rate, pressure and temperature, the appropriate stress-strain relationship must include various characteristics of this coupling. The CTH code is a hydrocode that has the capability to model a high strain-rate, high pressure, large plastic strain and high temperature impact. By using CTH, the authors have created a fairly accurate model of the shoe/rail interaction and have duplicated this phenomenon of gouging. Since gouging development is dependent on the impact surface conditions, the method used to model material interfaces directly affects the accuracy of the solution. Three methods are available in CTH to handle material interfaces: 1) materials are joined at the interface, 2) a frictionless slide line is inserted, and 3) a boundary layer interface is established. An axisymmetric sliding scenario is used to compare how each method affects the computed material shear response. A more general comparison study of the effects of these methods on high energy impact solutions is also conducted with an axisymmetric impact scenario. When the various methods are compared, the frictionless slide line method is shown to produce numerical instability and the boundary layer interface method is too limited for two dimensional applications. The method of joining materials at the contact surface, however, generates good results and appears well suited to the simulation of high energy impact events and the hypervelocity gouging process. Therefore, a recommendation is made, based on the results, for future hypervelocity gouging simulation.