Impact Testing and Analysis of Polycarbonate Shields, 18-R9562

Printer Friendly Version

Principal Investigators
P. A. Cox
Mark Griffin

Inclusive Dates:  07/15/05 – 11/15/05

Background - SwRI has conducted many impact tests that involve the use of transparent materials for debris shields. These types of debris shields are typically made of polycarbonate or acrylic materials and are designed to be lightweight and tough. The goal of this quick-look project was to demonstrate an analytical/numerical design capability for transparent shielding, taking advantage of a timely offer of free target materials from a manufacturer. Transparent shielding is desirable because it affords the user an extra measure of safety, being able to see what is occurring on both sides of the shield. It is also generally more attractive than conventional metal shields. In some cases, such as the canopies and windows of aircraft, transparency is required. Transparent shielding also is frequently employed in commercial buildings, commonly in places where money changes hands. Several companies market transparent shields, designed to resist specific threats, but these designs are all determined by trial and error experimental testing – there are virtually no validated methods for analytically predicting their performance. This statement is also valid for aircraft canopies.

Approach - Shields were provided by the manufacturer for impact testing at the SwRI Ballistics Range. To assist in the development of the shields, SwRI performed simplified analyses to assess the performance of the shields when impacted by a wooden 8-foot long 2 x 4-inch plank weighing 9.5 pounds. The analyses were performed to guide the selection of polycarbonate thickness in the shields. Three each of four unique shields were provided to allow for repeat testing. The shields were tested at the SwRI Ballistics Range Gas Gun facility. Numerical simulations were performed with LS-DYNA (see illustration) as an aid to understanding shield behavior, including failure, observed in the tests. Correlations were made between test results, analytical predictions of deflection, and numerical predictions of deflection and failure.

Accomplishments - Dade County Hurricane Protection requirements were followed for impact locations and missile size. Nineteen tests were conducted on the twelve shields, all with nominally 8 ft, 9.5 lb, wooden 2 by 4's fired at different impact velocities. Impacts were at the panel center and near one corner. The tests resulted in four failures of the shields. In three cases, the polycarbonate panel was permanently displaced from the frame of the shield and in one case the polycarbonate panel shattered. Post-test numerical simulations were performed for correlation to experiment. Results of the simulations for the shield in which the polycarbonate panel shattered were inconclusive. Neither the maximum stress nor strain occurred at the initiation site of the failure; however, failure initiated at a boundary that was rough to the touch (as received from the factory) and appeared to have been formed by partial cutting followed by breaking. Flaws formed by the operation used to cut the panel to size may have contributed to the failure. Funds were insufficient to investigate the contribution of the flaws to the observed failure.

A quarter-symmetry finite element model for center impact of a wooden 2 x 4 plank against the polycarbonate shield. The UniStrut frame is blue, the backing frame and UniStrut frame clip are red, and the polycarbonate is gray.

2006 Program Home