Seismosoft | Earthquake Engineering Solutions

Dear seismostruct team.
I want to catch the Moment-curvature section of the cantilever beam with constant axial load at top of beam.
So i modeled the element(FB) then do Pushover analysis and get the moment and curvature during the analysis.
(geometrical nonlinearity is off - 5 section per element - section mesh:500 -weight of all section is gotten zero)
I compare this result with exact and powerful software like "XTRACT". but seismostruct result is very higher than "XTRACT".
Please see the image.
http://www.4shared.com/photo/I8weQIKIba ... ature.html
I searched in forum but i couldn't find well discussion.
Please tell me how can i get well result from seismostruct.

Hi salar. Just some things to check:

1. Does XTRACT use the same stress-strain model for confined concrete that you specified in SeismoStruct?

2. Does XTRACT use the same stress-strain model for reinforcing steel that you selected in SeismoStruct?

3. Is it possible that strain hardening is being considered in SeismoStruct and not in Extract?

4. Do you have failure criteria set in SeismoStruct? XTRACT likely has some default criteria for maximum strains in the longitudinal and transverse reinforcement, while much of that is up to the user in SeismoStruct.

Just some thoughts salar. Best of luck.

Hi huffe
How can i get sismostruct strain-stress of material.
In material part we can see just time-strain!

Yes. That is correct salar. The best you can do is to specify confinement factors, compressive strengths, etc. . . which will most closely match the input used in XTRACT.

However, you do have options you can control.

Does XTRACT include strain - hardening of the reinforcement? If not, then use bilinear steel (not mp or mn) material with 0.000 (instead of the default 0.005) strain hardening parameter for your SeismoStruct reinforcement.

Does XTRACT use a code-based fracture strain of 0.06 or something similar for reinforcement? If so then with the same bilinear steel model for your reinforcement in SeismoStruct, specify such (instead of the default value of 0.10).

Does XTRACT use the Mander model for core concrete? If so, then use the Mander material for your core concrete (but not for the shell concrete) in SeismoStruct with appropriate confinement factor as mentioned above.

thanks huffe for your help.
My confinement factor in seismostruct is 1.1958 and for xtarct is 1.2092. As you see they are very close.
XTRACT used mander's formula confine and unconfine concrete and steel, my model has a strain hardening.
Please see the image:
http://www.4shared.com/photo/8RB415Evba/12_online.html
I correct the cover to 5 cm and My and Mu is gotten better but curvature not change well.
what should i do?

Hi salar. As I look over the screen shots you sent, I notice that you are, in fact, using the "steel-mn" model for reinforcing. For comparison to XTRACT results, I believe that this is part of the discrepancy between the two.

I am not saying that the steel-bl model is more accurate than steel-mn, but for trying to compare with XTRACT, I think it is more similar to what is being used in XTRACT.

Did you try my suggestion of using steel-bl model with strain hardening parameter set to 0 for reinforcing instead?

I say this because I notice in your XTRACT images that your strain hardening begins at a strain of 0.08 and fracture at a strain of 0.09. What is the fracture strain in your SeismoStruct reinforcing model?

In addition to these comments, I'll add that I don't believe that the two results are as far apart as you may initially suspect. While SeismoStruct has the capability to continue loading the structure once a bar fracture occurs, XTRACT does not. And from your two curves (the SeismoStruct curve extending much further that the XTRACT curve) it does appear that the negative stiffness occurs at approximately the same curvature value in both cases. The higher moment in SeismoStruct at this curvature is likely due to the different reinforcing models used in the two cases. SeismoStruct is able to continue the predicted behavior beyond this point, which is where you would likely limit the designed response.

Best of luck salar.

Dear huffe,
Thanks for your help. I checked steel-bl and it doesn't changed.
XTRACT stop the analysis while confined strain reached to ecu. if assume
seismostruct continue loading the structure once a bar fracture occurs so the curvature is okey, but as you see the Moment strength of seismostruct is bigger than XTRACT. I checked my both modeling for wrong input and inputting data is right.
http://www.4shared.com/photo/9j2kZRd0ce/14_online.html
So please tell how can i match the Moment strength of seismostruct and XTRACT?

I am afraid that since I do not have XTRACT, that we must rely on you for this potential issue, salar. If you could e-mail the SeismoStruct file to me, I will compare the results of said model to my own methods of developing moment-curvature relationships to see if I get similar differences.

Regards,

huffte.

Thanks, I sent to your email.

Hi salar. I fear your e-mail may have been lost somewhere in cyber-space, for I have not received anything yet. You may want to try a re-send.