Load duration behavior of steel-doweled wood connections

The load-carrying capability of wood decreases as the duration of the applied load increases. Current design practice for wood structures accounts for this reduced capacity by applying load-duration factors (National 1997) and time-effect factors (Standard 1996) to the wood strength. These factors were developed through long-term testing of wood members in flexure, yet they are applied to wood connections as well. This study develops time-effect factors for steel-doweled wood connections and compares them to those presently used in load and resistance factor design (LRFD) (Standard 1996).; Long-term tests are performed on small-scale wood connection samples, and the resulting time-to-failure data is used to calibrate two cumulative damage models (CD models), the Exponential Damage Rate Model (EDRM) and the Barrett and Foschi Model II (BF-II). The calibrated CD models are then used in a relatively simple analysis to determine approximate time-effect factors for steel-doweled wood connections. To test the adequacy of using the CD models for full-scale connections, step-load tests are performed on full-scale bolted connection samples and the results are compared to the results predicted by the CD models.; A numerical method is developed for calibrating existing damage models to test data. The adequacy of the EDRM and BF-II models are investigated using several techniques to quantify the quality of fit. The best fit is provided by the BF-II CD model, using a fitting technique that utilizes maximum likelihood estimation.; The LRFD code stipulates a time-effect factor of 0.80 for both snow load and occupancy live load. The best fitting models for the long-term test data produce 7 and 14 day time-effect factors between 0.68 and 0.70 for EDRM and 0.74 to 0.78 for BF-II. The best fitting model for the step-load test data gives time-effect factors between 0.86 and 0.88.