Surface area of Box2

[pnacht]

I have created the following Box2 for a few tests:

I'm checking the surface given by Robot and am finding some odd behavior.

By my math, the surface should be: 2*(b+2*tf+hw) + 2*(b-2*tw-b1)

For the start node, this should be equal to 82.

Robot, however, is presenting a value of 86.

For the end node, this should be equal to 126.
Robot, however, is presenting a value of 138.

To protect against my own idiocy, I created this section in AutoCAD using a polyline and got its length. The result is 82.

And before anyone suggests considering the internal perimeter, that doesn't do it either. The start becomes 124 (not 86) and the end, 188 (not 138). Also, this would be the only section to return the internal perimeter (all the others return only the external perimeter, since the value is primarily for painting).

EDIT:

Also, it occured to me it was possible b1 was (despite the schematic drawing) from axis-to-axis of the webs, but thats not the case either (the schematic is correct). This would increase the values calculated by 2*tw, which is still insufficient (start = 84, end = 132). In fact, in this particular case, if b1 was between axes, the webs would be face-to-face (no distance between them) in the start node and would overlap in the end node.

[t.sautierr]

Maybe a silly question : why do you you have 2 different names for the section : one in the table and another one in the dialog box? (BOX2 V1 and BOX2 V2)

[Rafal.Gaweda]

You are right.

Something is wrong. Need to be corrected.

Thanks for reporting.

[t.sautierr]

Rafal could you investigate this quickly please because I'm currently struggling with a model for one month which is exactly in the same situation with many tapered welded I or double I sections and I have not checked this particuliar point.

We have be analysing differences in the results between two models with two different software and if there are errors in the calculation of the section properties this could come from it.

Thanks a lot.

PS : If you can fully report if there are also differences for other properties, it will be great.

Have nice day.

[Rafal.Gaweda]

Which sections?
Which properties?
What are Robot results? What are hand made calcs results?

So far I see wrong result for BOX2 section for SURF value (painting area)

[t.sautierr]

Which sections?

Same type than above
Which properties?

I was asking if some other properties may be affected, for example inertia
What are Robot results? What are hand made calcs results?
Impossible to make hand calcs (the structure is a grid of beams fixed fixed and is highly hyperstatical) - Robot results are different from the other software :

sum of reactions are the same in the two models, but values at each support are locally very different, there is two line of supports with 22 nodal supports at each line)
So far I see wrong result for BOX2 section for SURF value (painting area)

nothing to notice in the other properties?

[Rafal.Gaweda]

[pnacht]

So, bringing up this subject once more.

I've noticed Box2 is not the only section with the incorrect surface. The T section of RC beams (didn't check steel) is also incorrect.

Surface area should be 2*(bf+h) = 10,256m, but is instead calculated as 2*(b+h) = 6,336m.

[Rafal.Gaweda]

Ups, you are right.

[pnacht]

Well, I just saw the ReadMe of the new Service Pack 3 and Box2 has apparently been fixed.