rcuapi.dll: concrete models

[Romanich]

I've created little macro to compare the stress-strain relationship for different concrete models. The results for some models are not what I expected.

- Effective Height Reduction Factor doesn't work and Rec model looks like linear:

- Results for Par model are completely incorrect:

- If Exponent not equal to 2 and strain at the top is bigger then Strain Ultimate Limit, StrainToStress method return -1.#IND in Pow model (for strain bigger then the Strain Ultimate Limit). If Exponent equal to 2 (default value), results seems to be true:

[Rafal.Gaweda]

Thanks for reporting

[Romanich]

Short resume after installing SP4:

Rec model - added warning message:

Par model - results are seems to be true:

Pow model - results should be close to Par model, now strain-stress relationship looks like:

Warning message has been added:

[Rafal.Gaweda]

These exceptions are made on purpose - to display error message below description what is wrong with model \ model data.

[Romanich]

Dear Rafal, what you can say about Pow model?

[Rafal.Gaweda]

What can you say ?

[Romanich]

I think that this part is missing (according to Eurocode 2 EN 1992-1-1:2004 p. 3.1.5 eq.3.14):

[Rafal.Gaweda]

That's clear that something is not cleared in rcuapi.dll 😞
Thanks for reporting

[Romanich]

After installation of the Robot 2015 two different versions of rcuapi.dll were found.

Vesrsion 27.3.0.5135 by path:

c:\Program Files\Autodesk\Autodesk Robot Structural Analysis Professional 2015\RESC\RC\rcuapi.DLL

c:\Program Files\Autodesk\Autodesk Robot Structural Analysis Professional 2015\RESC\rcuapi.DLL

This version of rcuapi.dll can be succesfully registered and added as reference in VBA. 

Vesrsion 28.0.1.5354 by path:

c:\Program Files\Autodesk\Autodesk Robot Structural Analysis Professional 2015\System\Exe\rcuapi.DLL

This version of rcuapi.dll after registration gives an error in VBA :

P.S. Is it possible to share what was modified/added in rcuapi.dll in the current version?

[Romanich]

Ok, I found the reason of error from the previous post - Excel should be x64. But Spreadsheet Calculator works only with x32 Excel. I don't know how we can use those features simultaneously.

[Romanich]

Results for the Par model became strange during the increasing of the design strenght:

30 MPa:

40 MPa:

70 MPa:

[Rafal.Gaweda]

Oh Roman....

Need to be blocked = limit should be checked
basing on 3.14 formula the limit which can be used is 2.9 and you use 3.5

[Romanich]

After the installation of SP2 for Robot 2015 results looks like:

[Rafal.Gaweda]

See : http://forums.autodesk.com/t5/Robot-Structural-Analysis/rcuapi-dll-concrete-models/m-p/5014128#M2332...

Diagram is correct.

Your data is out of range data (regarding table 3.1) which should be used in formula 3.14

For your data : point in red circle becomes point in yellow one and then algorithm returns 0 as strain becasue strain is greater than eps_cu1

[Romanich]

Let's return to concrete models.

Model Pow (from the help - Exponential relationship for non-linear structural analysis - Eurocode 2 EN 1992-1-1:2004 p. 3.1.5 eq.3.14. Atypical) is completly incorrect.

I've created own concrete model NonLin based on the above equation.

Function NonLinStress(eps_c, f_ck, E_cm, eps_c1, eps_cu1)
'Function describes stress-strain relation for structural analysis (EC2 f.3.14)
'eps_c - compressive strain in the concrete
'eps_c1 - strain at peak stress according to Table 3.1
'eps_cu1 - nominal ultimate strain
'f_ck - Characteristic compressive cylinder strength of concrete at 28 days
'E_cm - secant modulus of elasticity of concrete (according to Table 3.1 ???)
eta = eps_c / eps_c1
f_cm = f_ck + 8 'mean value of concrete cylinder compressive strength [MPa]
k = 1.1 * E_cm * eps_c1 / f_cm

If eps_c = 0 Or eps_c < 0 Then NonLinStress = 0
If 0 < eps_c <= eps_cu1 Then NonLinStress = f_cm * ((k * eta - eta * eta) / (1 + (k - 2) * eta))
If eps_c > eps_cu1 Then NonLinStress = 0
NonLinStress = Round(NonLinStress, 2)
End Function

Please find below comparison of results:

[rsousa_]

Hi @Romanich ,

Any update on this subject? I saw your exceltent webinar👏 and it looked that everything was solved, but i tried and i have the same problems you report here...

Thank you in advance.

Kind Regards

[Romanich]

Hi @rsousa_ ,

The new version of this great dll was implemented in Robot 2021.

I hope @okapawal  and @gwizdzm  can give us more information.

[rsousa_]

Thank you @Romanich .

I don't have that version 😢. I'm digging this dll in visual studio (dll 64 bits and 32 bits office... ). It is super, but those errors are a mess... and it was phenomenal if you could set 2 types of steel for prestressing sections...

More information would be awesome. 

kind regards

[Romanich]

It seems that the BiLinear Rectangular concrete model does not work properly:

Expected results are shown in green:

[okapawal]

Hi Roman,

Your remark has been reported to the dev team.