Application of statistical models to predict roof edge suctions based on wind speed

  1. Get@NRC: Application of statistical models to predict roof edge suctions based on wind speed (Opens in a new window)
DOIResolve DOI:
AuthorSearch for: ; Search for: ; Search for:
Journal titleJournal of Wind Engineering and Industrial Aerodynamics
Pages4253; # of pages: 12
SubjectAerodynamic loads; Building codes; Dynamic loads; Gaussian distribution; Gaussian noise (electronic); Loads (forces); Office buildings; Random processes; Roofs; Structural dynamics; Wind; Wind stress; Full-scale experiment; Gumbel distribution; Low-rise buildings; Non-Gaussian process; Peak pressure; Translation process; Wind pressures; Wind effects
AbstractThis work compares and predicts the response of roof edge components to wind load. The edge components consist of three different parapet coping configurations on the edge of a commercial building's roof system. Full-scale highly non-Gaussian data acquired on a low-rise building is used for analysis. The comparison shows that strong suction is observed on the front flashing of all configurations, contrarily to what is specified in building codes. The prediction of the edge component response to wind load is accomplished with both a Gumbel distribution model and a translation method recently proposed in the literature, which estimate the extreme value distribution of the pressure coefficient. A Gumbel model is commonly used to represent the distribution of the peak pressure coefficient. The model parameters are determined from observed peaks, defining the Gumbel method. Recent work has proposed an alternative, the translation method, using the pressure coefficient entire time history instead, modeled as a translation from a Gaussian random process. Major gains include accurate and stable performance for strongly non-Gaussian data. The present results show that the translation method produces a more realistic estimate of the peak pressure coefficient distribution than the Gumbel method.
Publication date
AffiliationNational Research Council Canada
Peer reviewedYes
NPARC number21277455
Export citationExport as RIS
Report a correctionReport a correction
Record identifierbea35ff7-f312-47e4-b9a9-6e19ae042483
Record created2016-03-09
Record modified2016-05-09
Bookmark and share
  • Share this page with Facebook (Opens in a new window)
  • Share this page with Twitter (Opens in a new window)
  • Share this page with Google+ (Opens in a new window)
  • Share this page with Delicious (Opens in a new window)