Modelling complex geometries in heat load calcs

[jistre]

I want to pick some brains here and see what kinds of strategies you guys use when you are in the position I'm in now. I use a computer program to model heat loads on my projects, and when external wall exposures are input, I have the option to select 16 compass points to describe what direction the wall is facing. However, I have curved walls in my current project, and the curved section spans 90 degrees of the compass. I'm wondering what you guys do in this situation. My first instinct is to compute the total wall area, then divide that by four and apply that value to each of the 4 compass points that are in the quadrant the curved wall spans.

Any thoughts?

 

[trashcanman]

Breaking it up into 4 areas should work. 8 areas would be closer. It is a bit like a calculus problem. The more areas you break it into, the more accurate the load, but also more work involved. Try it with 4 areas, then with 8. Compare results and see if 8 areas are significantly more accurate.