Experimental investigation of damage accumulation in wood

Zhiyong Cai, Purdue University

Abstract

The purpose of this study was to examine the damage accumulation (DA) process in wood. The research objectives were: (1) compare the predicted damage by the most commonly used DA models with actual damage in members subject to load processes and to determine which DA model(s) best characterized the experimental data; (2) investigate the relationship between the damage and physical parameters, nondestructive parameters, or change in these parameters. This project consisted of two phases: phase I and phase II load tests. In phase I load test, a five-day load sequence with different stress levels was used to simulate structural loading process. In phase II load test, a five-day load sequence with constant stress level was used to simplify the data analysis. Three different methods used to predict the static strength of test specimens were presented and quantitatively compared. The multiple linear regression method was the best predictor of strength. A new technique that utilizes all sample points and not requiring knowledge of the initial conditions to determine the damping properties of free vibration was developed. The new method was found to be more consistent in measuring damping than the classic method. The wide distribution of specimen failure during the five-day load sequence in both load tests was observed to be different than the predicted failure distributions by the four DA models. The significant differences indicate that no DA model successfully characterized the experimental results. Significant (9%) decrease of static modulus of elasticity (MOE) was observed after phase II load test. The difference between static MOE and the dynamic MOE could be used to indicate damage. New creep and recovery models for improving curve-fitting were developed. No useful trends were discovered in plotting creep and recovery parameters for the load sequence in both load tests. The 13.7% reduction of the residual strength in a surviving specimen indicated that its maximum failure load was smaller than its highest applied load. This observation suggested that damage occurred and accumulated not only during load pulses, but also when the applied load was removed.

Degree

Ph.D.

Advisors

Hunt, Purdue University.

Subject Area

Wood|Technology

COinS