# Problems with optimize cross section

If I want to optimize a beam on 3 supports (pinned, roller and roller), I do not get the smallest possible profile. can anyone explain to me how this is possible

Hi,

please refer to the comment I made here: Optimize cross section material beam

yes, thanks! the question to optimize the material has been resolved. The question of the beam on 3 support points is not yet. Although he does just optimize, he does not take the smallest profile possible. The unity check is also not at 100%. However, this should be the case if it takes the smallest possible profile.

Hi,
i do not understand the problem. If you set the profile to be HEB100 (the smallest in the list) then the utilisation is 1880%. Karamba3D is optimising the setup to use the HEB 320 beams for your setup to reach a 87% utilisation.

the attached file shows that the beam has a maximum bending moment of 2071.58 kNm. If you divide 2071.58 by the yield strength of steel (S235) you get the minimum moment of resistance that the beam needs to meet its strength. (2071.58e + 6) / (235) = 8815.23e+3 mm^3 (Wy).

A HEB800 has a moment of resistance of 8977e+3 mm^3, which means that it is the smallest profile possible because a HEB700 has a moment of resistance of 7340e+3 mm^3. However, an HEB900 is chosen during optimization. How is that possible?

Hello @b.hunen,
in the definition above the selection of the HEB900 is due to lateral torsional buckling. In the file below the lateral torsional buckling length is set to zero and a HEB800 results:
ligger-2 steunpunten verdiepingsligger lijnlast optimalisatie_cp.gh (53.1 KB)
– Clemens