Design Code Derived Wind Loads vs. CFD – Introduction

Rough and ready ‘CFD analysis’ using Autodesk Flow Design. Good for quick and easy visuals and money shots but not much else I don’t think.

I intend this to be a series of posts looking at using computational fluid dynamics to model wind pressures on a building facade for the structural design of the facade elements (glass, supporting framing etc). Although I work as a facade engineer I am somewhat of a noobie when it comes to CFD analysis so this is intended to be a bit of a learning process into the software and the theories behind CFD modelling techniques..

  1. I start off looking at the design pressures one gets on a simple building using a set of design codes (AS/NZS1170.2)
  2. I then run CFD models on the same geometry with similar boundary conditions
  3. The investigation finishes with a comparative look at the results and look into the reasons similarities and differences.

I have done some reading around the subject and as one would expect there is a wealth of knowledge on the subject of using CFD for building modelling. It is typically used for wind comfort studies and building air movement and thermal modelling studies. These subjects are typically the main focus of the various papers and research on the matter and these papers are generally explicit in referring to mean turbulent velocities. Comment on the application of the technique to peak ‘design’ pressures is somewhat absent except through passing comment and is probably best summarised as follows from Holmes:

This is the general consensus on the current state of play I feel. Bearing this in mind I’m not expecting to be able to reconcile design pressures with CFD-derived pressures, particularly given my limited understanding of the subject at hand. I just hope to grow (and share) a basic understanding of the Reynolds -Averaged Navier Stokes (RANS) methods and various turbulent models to, I suppose, understand the state-of-the-art.


Somewhat contradicting Holmes conclusions in the excerpt above, the design codes do not preclude the use of CFD analysis (although quoting Holmes as a source). From AS/NZS 1170.2:Supplement 1 Wind Actions – Commentary :

(a – c excluded for brevity)

I find section (e) of this interesting since my research to date has not unearthed recommended practice and methods for reconciling/ calibrating CFD with measured data, in terms of peak wind velocities and the associated pressures acting on building cladding systems. My investigations have not been exhaustive, however, and this clause from the design codes suggests perhaps methods do exist. Perhaps other design codes provide more information on this methodology.