element beamColumnJoint -> problem with the recorder comm

[Macedo]

Good afternoon,

I'm modeling a steel moment resisting frame, using the joint element.

If I do a Pushover analysis and do the following recorder:

recorder Element -file $Dir/PZ_ShearPanel_233.out -time -ele 233 shearPanel stressStrain

In the recorder file I get the load factor, the Equivalent Joint Moment and Distortion of the Panel Zone. The results are the expected for the panel zone law that I had defined.

But when I do the same recorder for the Time History analysis the results I get are completely incomprehensible. I don't understand what I get..

The recorder command for the joint element works well?

Thank You very much for your help.
Best Regards
Luis Macedo

[fmk]

the recorder works the same whether the element is in a static or a transient analysis .. so yes the recorder command works well.

[Macedo]

Thank you very much fmk.

My problem is that I've done many non-linear static analysis and the results for the panel zone equivalent moment and deformation are what I expected.

But when, using the same Data file, I do the transient analysis (adding the masses and changing the input from nonlinear static to transient analysis) the results I obtain for the forces and deformation on the panel zone are not consistent. But the results I obtain for the beams and columns are what I expect.

Do you have any idea of something that could be wrong?

The only changes that I've done for the transient analysis are the following:

# TIME SERIES ANALYSIS------------------------------------------------------------
# Time series Input
#set GroundM "Series -filePath groundM.txt -dt $deltaT -factor 9.81"
set GroundM "Series -filePath Accelerogram.g3 -dt 0.02 -factor 9.81"

# Define load pattern
# --------------------
pattern UniformExcitation 2 1 -accel $GroundM

set Damp 0.025
eigen 5; # Number of modes considered
rayleigh 0 0.0188 0.0 0.0; # aplha betaKT betaKi betaES? [C] = alpha*[M]+beta*[K]

# Define Time History Analysis Option
# -------------------------------
constraints Transformation
test NormDispIncr +1.000000E-003 250 1
integrator Newmark 0.5 0.25; # If using values other than 0.5 0.25 here, will get artifical damping
algorithm Newton
numberer RCM
system BandGeneral
analysis VariableTransient
# analyze $numStepsd $deltaTA [expr $deltaTA/10] $deltaTA 10
analyze 2500 0.02 [expr 0.02/10] 0.02 10

puts "Time History Done"

puts "Analysis has finished!"

Thank you very much again,

Best Regards
Luis Macedo

[Macedo]

Dear fmk,

I still do not solve the problem.I spoke with some colleagues who have done similar analysis and they also had the same problem.

Firstly I thought that the problem was because I'm using a trilinear behavior law for the panel zone. But I changed it for a elastic law with a high stiffeness and I have the same problem.

Thank you very much for your help,

Best Regards,
Luis Macedo

[fmk]

you will have to post the rest of the i/p and show me what it is you are plotting .. however before you do .. first try tightening up your convergence tolerance 1e-3 is rather large, use something a lot smaller 1e-8 or 1e-10 .. also if you need 250 steps for convergence something is wrong, cut it down to a max of 10.

[Macedo]

Dear frank,

My code is sourced from some files so it is difficult to post here all the code..

Can I send you my code by email and then you can tell me your opinion about it??

I changed the convergence tolerance criteria and the number of steps but the output that I get for the panel zone equivelent moment and distorsion are the same..

Thank you again for your help,

Best Regards,
Luis Macedo

[Macedo]

Dear Frank,

I still have the same problem with the record in the joint element.

I have already changed the convergence tolerance criteria and the number of steps as you told me...but the results are the same..

I've tried already difference K for the joint springs.

What is strange for me is that for the pushover I get the values expected but in the time history something happens and the values are not what I expect.


Thank you again for your help,


Best Regards,
Luis Macedo