Performance Modeling of Explosively Actuated Devices

• Main Author: Braud, Adam M.
• Other Authors: Keith Gonthier
• Format: Others
• Language: en
• Published: LSU 2006
• Subjects: Mechanical Engineering
• Online Access: http://etd.lsu.edu/docs/available/etd-09052006-093802/

Explosively actuated devices (pin pullers, cable cutters, valves, etc) are used extensively to perform critical functions for aerospace, industrial, and defense related applications. The failure of these devices have led to a greater effort to quantify device design and performance. This thesis describes the actuation process of an explosively actuated valve, including: 1) the burning of the solid explosive HMX (C₄H₈N₈O₈) and production of its high pressure gas products, 2) the mass transfer of gas products through an actuator to an expansion volume including choked flow effects, 3) the resulting piston motion due to high pressure gas products, and 4) the effects of device deformation on valve performance. Although the model presented is validated with a valve, it is kept general such that it can be applied to other explosively actuated devices. A key model objective is to qualify the effect of design modification (geometry, propellant mass, etc.) on device performance. A focus of this paper is to describe the leading order effects component deformation has on device performance, including the effects of material strain hardening and internal gas pressure. Model results for the axial resistive force exerted on the piston during actuation are compared to nonreactive quasistatic compression tests and a finite element study. Results from the compression tests and FEA indicate there is significant piston bending induced by the housing corner during skirt insertion. Results reasonably predict both the compression tests and finite element results if two friction coefficients are used as a simple way to describe piston bending. To characterize reactive valve performace, data from a reduced number of experiments was used to determine model parameters which are difficult to measure (propellant linear regression rate, friction coefficient, etc.) and characterize baseline valve performance. Results from a sensitivity study suggest that the piston is being overdriven by its current propellant load (150 mg HMX). As such, valve performance is insensitive to slight modifications around the baseline case. Valve performance does show sensitivity to propellant mass and friction coefficient. Valve failure is predicted with a propellant mass between 30 and 40 mg, and with elevated friction coefficients (≈ 1.0).