This report of the research project "Improved prediction of seismic resistance of Part 9 Houses" under the CMHC External Research Program consists of a review and assessment of analysis methods; numerical evaluation of current seismic design requirements in Canada; and new formulations for seismic design of conventional wood-frame construction in Canada.
The relative performance of three mechanics-based methods is ascertained by comparing the test data of lateral capacities of partially restrained wall specimens having window openings with the predicted results from three calculation methods: Method 1 by Ni and Karacabeyli (2000, 2002) is the simplest to use and gave the most conservative results; Method 2 by Källsner et al., (2001, 2002) is less conservative but more complicated to apply to practical problems, and Method 3 by Källsner and Gurhammar (2005, 2006) gives non-conservative results. The suitability of other methods of analysis, (e.g. SAWS, Drain2D-X) was also examined. Method 1 was chosen as the principal analysis tool for this investigation.
The adequacy of the seismic provisions of the CWC 2004 Design Guide and of the proposals for Part 9 of the 2010 NBCC are assessed by the seismic design methods specified in Part 4 of the 2005 NBCC and utilizing the analysis Method 1. Two building types were used: a square building of 15.0 x 15.0 m plan and a rectangular one of 4.8 x 15.0 m, each of 1, 2 and 3 storeys height. The analysis indicates that neither the current CWC Guide nor the proposals for the 2010 NBCC Part 9 meet the seismic requirements of Part 4 of the 2005 NBCC for the higher seismic zones. The discrepancies are particularly pronounced for the shorter side of the rectangular buildings.
It must be noted that the buildings studied in this investigation represented worst case scenarios. In reality, wood-frame houses would generally contain more walls than the minimum wall lengths required by the CWC Guide and the proposed NBCC 2010, and thus would possess larger lateral resistance.
Following the numerical results of a parametric study of different wall constructions, two new approaches for the seismic provisions of conventional wood-frame construction in Canada are presented, an area-based method, and a method based on percentages of braced wall lengths. Both methods conform substantially to the seismic requirements of Part 4 of the 2005 NBCC.
For heavy construction the provisions for 1 and 2 storey buildings give reasonable agreement with those for 2 and 3 storeys of light construction.
Additional parameter studies should be carried out for irregular buildings, for heavy wall cladding such as stucco and masonry, and for minimum size of braced wall panels.