The in-service loads on wood pallets are often uniformly distributed in nature. Laboratory test methods to evaluate pallets typically use point or line loads because of the difficulty of simulating the actions of a uniform load. In support of research directed at developing design procedures for pallets, a uniform load test machine was developed. This device loads through an air bag and records load-deflection measurements automatically in a micro computer. Test results using the machine were compared to those obtained from tests using bagged goods to simulate a uniform load. The machine is superior in efficiency, speed of testing, and accuracy when compared to bagged goods loading. The assumption of a uniformly distributed load is very convenient in the design of pallets. However it may be significantly inaccurate for simulation of intrinsically stiff unit loads. The stiffness may cause the load to bridge or redistribute itself into a series of discreet loads. Pallets of varying stiffness were tested with several bridging and non-bridging type loads to determine the potential error from ignoring load bridging. The results indicated that for stiff pallets where deflection is likely the primary design criterion, the effects of bridging are relatively negligible. However, for flexible pallets where the only restriction on deflection is likely to be load stability, bridging can result in an error greater than 50 percent. The deflections of pallets stored in a racked mode, whether across the deckboards (RAD) or across the stringers (RAS), depend on the effective span between supports. A study was conducted to determine if the effective span is equal to the clear span and what advantage is gained by increasing rack support width. Tests of pallet sections supported RAD indicated that a pallet functions as a frame with semi-rigid joints whose function differs with the width of the support. A 53 percent reduction in centerline deflection was observed if support width was increased from 1 to 4 inches.

M.S.