Abstract:
With high costs of testing and rating a structural system for
fire resistance, the utilization of computer simulations that
approximate the integrity of structural subsystems could conceivably
reduce development costs. Before an analysis of a light-frame wood
system can occur, information on the components of the substructure
must be known. The purpose of this study was to systematically
study the effects of thermal load on the strength and mode of
failure in a nailed gypsum-to-wood stud connection.
Initially the distribution of temperatures within the
connection was sought by using a finite element analysis. The
analysis simulated exposure to a standard fire as dictated by the
American Society for Testing Materials standard E-119. The thermal
properties of the materials involved are documented in the literature. Close agreement was found between analytical results
and specimens exposed to standard test conditions.
With the temperatures along the connection known, the
properties of the three materials at elevated temperatures were
needed. For wood and steel, this information has been well
documented. The property values for gypsum were accumulated through
testing gypsum board in compression at various temperatures. The
results showed an increase in compressive strength and stiffness up
to 100 C followed by a decrease in strength. These results were
combined with the analytically determined temperature distribution
to obtain the material properties in the neighborhood of the
connection.
The overall model was an extension of the yield theory
available in the literature. The approach used was to evaluate the
connection at set time intervals and calculate the strength of the
connection. Each time interval had a different set of material
property values for each material in the connection as dictated by
the temperature distribution. The mode of failure did not change
from that which occurred at room temperature, compression strength
of the gypsum being the determining factor. The load at which
failure occurred increased in the first five minutes and then
sharply decreased. This can now be incorporated into models of wall
systems for room temperature to predict behavior of walls during
exposure to fire.
