Full-scale fire resistance tests on cross-laminated timber

Volume 17, Number 4, December 2012

Cross-laminated timber panels

Full-scale tests carried out by NRC Construction researchers have demonstrated that cross-laminated timber (CLT) assemblies can achieve good levels of fire resistance, even when unprotected under full loading conditions. The tests are part of a study recently launched by NRC in collaboration with FPInnovations to develop a methodology that will foster the design of fire-safe CLT or hybrid buildings in North America. The study will also facilitate the acceptance of future code provisions for the design of CLT panels with regard to fire resistance.

CLT is a relatively new building system in North American construction that is helping to define a new class of products known as massive timber. The wood industry in Canada is interested in using CLT panels in certain types of buildings, particularly those currently using non-combustible construction. The panels are typically manufactured with three, five or seven plies that are glued together with the grain perpendicular to the previous layer.

Since CLT is a new product, there is a need to carry out research to fully understand its behaviour as a structural system in general, and in fire in particular.

As wood burns it forms a thick char layer, which acts as a low-density insulator that protects the wood underneath it from elevated temperatures. Understanding charring rates (a measure of the char depth over time) is fundamental to estimating the remaining thickness of full-strength wood, which in turn designers can use to calculate the residual strength of members for a given fire exposure. It is also important to understand the effect of other key factors on the performance of CLT in fire, such as the type of adhesive used between plies, the number of plies (thickness of the panel), the joint configuration, the protection methods used, and the type of fire exposure.

NRC wall furnace

Full-scale experiments

NRC researchers conducted eight full-scale experiments to obtain fire resistance ratings for a number of CLT panels and to obtain data on some of the other key factors. The panels were subjected to the standard ULC S101 fire exposure. The assemblies tested consisted of three wall tests and five floor tests. The tests were carried out in the NRC floor and wall furnaces.

Assemblies consisted of three or five CLT panels, which were constructed of No. 1, No. 2, No. 3 or MSR lumber boards, and came from different manufacturers from across Canada. Some of the CLT panels were fully exposed to fire (unprotected) while some were protected by Type X gypsum board. The configuration details of each test as well as the results are summarized in Table 1.

Table 1: Results from CLT fire resistance tests
Wall or Floor # of Plies Thickness (mm) Gypsum Board Protection Load Charring from Data (mm/min) Failure Mode Fire Resistance (min)
Wall 3 114 2 x 12.7 mm 333 kN/m 0.41 Structural 106
Wall 5 175 Unprotected 333 kN/m 0.65 Structural 113
Wall 5 105 Unprotected 72 kN/m 0.80 Structural 57
Floor 3 114 2 x 12.7 mm 2.7 kPa - No failure 77 Note 11
Floor 5 175 Unprotected 11.8 kPa 0.64 Integrity96
Floor 3 105 1 x 15.9 mm 2.4 kPa 0.60 Integrity 86
Floor 5 175 1 x 15.9 mm 8.1 kPa 0.75 Integrity 124
Floor 7 245 Unprotected 14.6 kPa 0.65 Structural 178


Note 1

Test was stopped due to equipment safety concerns. Failure was not reached.

Return to note 1 referrer

The tests demonstrated that CLT assemblies can achieve significant fire resistance that is close to three hours in some cases with even unprotected CLT under full loading conditions. The failure modes were a mix of integrity and structural failures.

NRC and FPInnovations researchers will use these test results to validate a generic fire resistance procedure developed by FPInnovations and currently published in the 2011 Edition of the Canadian CLT Handbook. The validation will consist of comparing the NRC test results to fire resistance values obtained by the generic method and then modifying the method to enhance its accuracy, if required.

The project was funded by the National Research Council of Canada and FPInnovations through the Natural Resources Canada Transformative Technology program.

For more information

Contact Noureddine Bénichou at noureddine.benichou@nrc-cnrc.gc.ca or 613-993-7229.