The design of wood-frame structural systems to withstand exposure to fire depends on knowledge of the fire endurance (time-to-failure) of the wood members used in the system. In fires, wood looses part of its load-carrying capacity due to charring and part due to strength degradation. This thesis examines the reduction in compression strength experienced by dimension lumber when exposed to elevated temperatures.
A program of experimental testing of nominal 2×4 Machine Stress Rated (MSR) lodgepole pine lumber concentrically loaded in compression and exposed to elevated temperature was undertaken by Forintek Canada Corp., Canada's wood products research institute. A computer program entitled HTExposure was written to simulate the experimental time-to-failure data gathered in Forintek's testing program. This computer program combines a modification of an existing heat-transfer model with various published compression-strength reduction models. This was done in order to determine which of those strength-reduction models could predict times-to-failure comparable to the observed values. As well, a new compression-strength reduction model was proposed. When predicted results were compared to the observed data, it was determined that the computer program predicted results closest to those observed when using the new compression-strength reduction model proposed in this study.
This supplemental report has been written to present aspects of the experimental testing not fully described in the final report for this project, and discusses in greater detail the recommendations for future work made in that report. Much of the experimental data gathered under this project and its predecessor has not been fully analyzed. It contains within it much more information regarding the behaviour of structural dimension lumber. Knowledge of this behaviour is important because the wood industry must be prepared to provide adequate and relevant performance information to gain code acceptance under the new proposed objective-based national building code.
WALL2DN is a simple, user-friendly, integrated two-dimensional computer model to predict the fire resistance of either insulated or uninsulated, non-loadbearing wood-frame wall assemblies finished on both faces with gypsum wallboard.
WALL2DN is written in Microsoft Visual C++ 6.0, and can be run on most personal computer systems with WINDOWS 95® or better operating systems. WALL2DN has a simple, user-friendly interface that permits architects, building-code officials and engineers to easily create custom-designed walls and then estimate the fire-resistance ratings of those assemblies. As well, WALL2DN allows users to compare the model's theoretical predictions with the empirical results from any of six standard fire-endurance tests.
WALL2DN can be used to investigate the fire resistance of wood-stud walls with the following construction options:
· three grade-species of lumber: SPF, Douglas fire, southern yellow pine,
· two sizes of studs: 28 mm x 89 mm (nominal 2 x 4) and 38 mm x 140 mm (nominal 2 x 6),
· two spacing distances between studs: 400-mm o.c. (16 in.) and 600-mm o.c. (24 in.),
· two types of gypsum board: generic Type X and proprietary Type C,
· two thicknesses of gypsum board: 12.7 mm (1/2 in.) and 15.9 mm (5/8 in.),
· single and multiple layers of gypsum wallboard attached directly to the two sides of the walls,
· varying spacing-distances between the fasteners used to attach the gypsum wallboard to the studs,
· if insulated, either glass-fibre or rock-fibre insulation filling the cavities between the studs, and
· if insulated, varying densities of insulation.
There is very good agreement between WALL2DN’s predictions and the observations and measurements taken during standard fire-endurance tests for the following:
· two dimensional heat transfer (temperatures) in wood-frame walls,
· deformation or gypsum wallboard during fire exposure,
· loss of joint compound,
· fall-off of the gypsum wallboard,
· melting/shrinking of glass-fibre insulation,
· charring of the wood studs, and
· “insulation failure’ of wood-frame walls.
While the following features are not currently available in WALL2DN, many will be added to future versions:
· heat transfer through floor-ceiling and roof-ceiling assemblies,
· dimensional shrinkage of wood members resulting from loss of moisture during fire exposure,
· application of structural loads on wall assemblies and the ability of the assemblies to support those loads during fire exposure,
· validation data for other wall designs,
· pre-loaded non-standard fire exposures,
· attachment of the gypsum wallboard to the studs using resilient channels,
· wall assemblies constructed with two rows of studs (e.g. either staggered or double-row), and
· wall finishes other than gypsum wallboard (e.g. OSB or plywood).