bracing with two material

[skhosravi]

Here is my model. I want not to have any compression in the bracings.
https://www.dropbox.com/s/5kkjmybcu93fb ... x.spf?dl=0

Also I want to see the hysteretic behaviour of the SMA in the bracings. So, I have to use truss element to define that.
Please help me with this case.

[z.gronti]

Dear Skhosravi,

Example 7 of Chapter 2 of the Verification Report of SeismoStruct might hepl you in the compression limits, and example 2(Seven storey, full-scale, RC shear wall building) of Chapter 4 of the Verification Report in combination with the Help System and the User Manual will help you to model the truss elements.

[skhosravi]

Thanks for your guide. But still it has problem! If I use truss member for part 1 and 2, then there will be error! I must use infrm member.
Then again the results are not ok. for the connection between the two truss member I tried both GAP and lin_asym with 0 stiffness in compression. but again it fails!

[skhosravi]

I am sure there is something wrong in this case. It is working just for one bracing without any compression. but for two material bracing it is not able to analyze that!

[seismosoft]

Hi skhosravi,

The main problem is that you model is completely 'unbalanced' in the out-of-plane direction. You can check this if you run an eigenvalue analysis (change the eigen-solver from Lanczos to Ritz+Jacobi, which is better suited to small models, see Project Settings>Eigenvaue). You will notice that the 10 first modes of vibration are in the Y direction, and most of them have really strange shapes.
If you restrain the out-of-plane DOFs in the free nodes (Y+RX+RZ) you will make the analysis run. Some other comments:
1) Having a maximum tolerance value of 1e100 (1 with 100 zeros at the end!) is the best way to numerically crash the program (stack overflow, range check error etc). The default value is perfectly good. If you get norm values larger than 1e20, you will never get your model to convenrge!
2) What do you need the rigid diaphragm constraints for? Simply remove them
3) The way you define damping is wrong. Read the help or the manual. You kept the default values for T1 and T2 (0.7 and 0.3sec) whereas your first period in the X direction is 0.01sec. You will get deformations totally wrong.
4) I do not know if this is intentional, but the pga of the recond that you have input is more than 1.0g. A very big ground motion!

Seimosoft Support

[skhosravi]

this is a very simple model for a concrete frame with steel braces. there is an error when I use the brace diameter 0.085 or 0.09!!!
but it works when the diameter is 0.1 or even 0.07!!! I think the model is OK but the error is strange.

https://www.dropbox.com/s/5oj5p62cnzr94 ... c.spf?dl=0

[z.gronti]

Dear skhosravi,

From what i can see, your project settings are too strict, that's why the model cannot converge. You need to change the settings both in convergence criteria and iterative strategy tab of Project settings module. If you use the Program Defaults values the analysis will be completed.