We are currently working on some IFC improvements for our field software. We are trying to understand the thickness definition for steel profiles with slopes (e.g. flange slope, leg slope, web slope), such as on the drawings that are available on buildingsmart’s website.

The affected shapes are U, T, and L shapes. From the drawings and descriptions on the buildingsmart website, it is not clear what is exactly described by the “thickness” values. It seems to be defined differently for all those shapes.

U-Shape: FlangeThickness seems to define the distance from the top line to the flange along the vertical axis (x=0)

L-Shape: Thickness seems to define the distance from the leftmost line to the end of the upper round edge.

T-Shape: The FlangeSlope (which is given as an optional parameter) is not mentioned at all in the drawing. So there is no way to find out which distance is meant in that case.

Does anyone know if these calculations are correct for L- and U-shapes, and how the thickness can be calculated for a T-shape?

Many thanks!


We just had a similar question this week and in my perception there is a consistent way of measuring possible if we ignore the image of the L-profile. I would expect the most logical way would be to measure the thickness (in case of a slope) exactly at half of the ‘leg’, i.e. in case of L shape at half of the depth / width (in contradiction to the image), in case of U shape at half of the FlangeWidth / Depth and in case of T shape at 1/4 or 3/4 of the FlangeWidth and half of the Depth. Similar measurements would also make sense for Z shape etc.

Note that in case of a slope the circle parts are less than 90% and in the L shape the center of the Edge for EdgeRadius could be outside the actual shape.
Note: the consistent way of measuring would be my preference, it is however not as it is defined (at least for L shape) at this moment.

I’m also not too convinced that the L-shape is deliberately different. It could be that it’s just a bit imprecise (or that there are quantisation artefacts from the heavily compressed dwf format).

Anyway I agree with Peter:

exactly at half of the ‘leg’, i.e. in case of L shape at half of the depth / width

It’s not entirely clear either where that would start considering the fillet and the intersection point between the two sloped lines.

So I would indeed phrase it as sth like:

Thickness is measured at 0, 1/4, 1/2, 3/4, 1 of the relevant measure away from the origin defined by Position.

1 Like