Multiple support problem

giacomodemarie · Post by **giacomodemarie** » Wed Jan 27, 2010 10:07 am

Dear all,

I made a model of a bridge and performed successfully a seismic analysis with a uniform excitation pattern (in terms of acceleration). Now I have moved to analysing the same structure with different support excitation, imposed in term of displacement, and I cannot manage to get convergence.

Then, to make a comparison, I have imposed the same displacement time histories at all the support nodes, calculating this time history from the support acceleration used in the uniform excitation case (in theory, the two analysis should give the "same" results). Even in this case, it was impossible to get convergence, both using Transient dynamic analysis and Variable Transient dynamic analysis, and using all the integrator schemes available in Opensees.

Can anyone help me please?!

Giacomo

Post by **fmk** » Wed Jan 27, 2010 11:27 am

the problem probably lies in your input displacements .. if you are using some well known earthquake record i suggest you get the filtered ground displacement or it as simple numerical integration of the accelerations does not work.

giacomodemarie · Post by **giacomodemarie** » Thu Jan 28, 2010 2:55 am

Thanks for your reply, Frannk.

I'm using in fact some earthquake acceleration recored, from which I have determined the displacement by band pass filtering and double integration with respect to time. The time histories of displacement seem to be good.

Moreover, I have in output stresse and strain in four point of each section, I have checked the stress strain curve obtained till the program stops and it seems that the structural response is substantially linear.

Can you give me any suggestion please?!

Post by **fmk** » Thu Jan 28, 2010 11:53 am

do you have damping in your model.

giacomodemarie · Post by **giacomodemarie** » Thu Jan 28, 2010 11:43 pm

Thanks for yor your kind answer!

Yes. I have used Rayleighdamping (proportional to the mass and initial stiffness matrix), and have calculated the constant considering the first mode and the 75th mode (this correspond, in the linear case, to the mode with non-negiglible modal mass in one of the three principal direction).

The model is made of 156 nodes, 75 fiber elements for the piers (Concrete04 + Steel02), 78 elasticBeamColumn elements for the deck, 3 linear elastic spring element for the deck-piers connections.

... ?

Giacomo

Post by **fmk** » Fri Jan 29, 2010 8:20 am

another reason to get differences is the rayleigh damping . the damping is done on the velocity terms at the nodes . in uniform case there are relativ velocities, in multi-support case they are absolute acelerations.

try turning off damping to see how results differ.

giacomodemarie · Post by **giacomodemarie** » Fri Jan 29, 2010 10:46 am

Dear Frank,

I have tried an analysis without Rayleigh damping, but no convergence is got even in this case.

What else can I do?!

Thanks very much for yout kind answer.

Giacomo

giacomodemarie · Post by **giacomodemarie** » Tue Feb 02, 2010 6:18 am

Dear Frank,

I have tried an analysis without Rayleigh damping, but no convergence is got even in this case.

What else can I do?!

Thanks very much for yout kind answer.

Giacomo

Post by **fmk** » Tue Feb 02, 2010 10:24 am

i will look at it .. but can you delay sending it until after i get the next release out as i am totally swamped at the moment .. you will get an email when the release comes out .. hopefully sometime tomorrow.

giacomodemarie · Post by **giacomodemarie** » Wed Feb 03, 2010 6:16 am

Dear Frank,

Thank you very much for your kind answer! I will wait for your new suggestion!

Thank!!

Giacomo

giacomodemarie · Post by **giacomodemarie** » Thu Feb 04, 2010 11:40 pm

Dear Frank,

Thank you very much for your kind answer! I will wait for your new suggestion!

Thank!!

Giacomo

giacomodemarie · Post by **giacomodemarie** » Fri Feb 05, 2010 6:07 am

Dear Frank,

I continued to look for some solution to my problem. In the imposed motion, I have also defined time series for velocity and acceleration, in addition to displacement.

Now it tells me that it cannot finde compatibility between forces and deformations of some elements, even if the behavior is still linear.
Moreover, the program quits abnormally, and windows gives me the following error repèort files:

<?xml version="1.0" encoding="UTF-16"?>
<DATABASE>
<EXE NAME="OpenSees.exe" FILTER="GRABMI_FILTER_PRIVACY">
<MATCHING_FILE NAME="OpenSees.exe" SIZE="7465472" CHECKSUM="0xF5E5A9E" MODULE_TYPE="WIN32" PE_CHECKSUM="0x72CD25" LINKER_VERSION="0x0" LINK_DATE="01/16/2009 21:34:15" UPTO_LINK_DATE="01/16/2009 21:34:15" />
</EXE>
<EXE NAME="kernel32.dll" FILTER="GRABMI_FILTER_THISFILEONLY">
<MATCHING_FILE NAME="kernel32.dll" SIZE="1009152" CHECKSUM="0x6444867F" BIN_FILE_VERSION="5.2.3790.3959" BIN_PRODUCT_VERSION="5.2.3790.3959" PRODUCT_VERSION="5.2.3790.3959" FILE_DESCRIPTION="Windows NT BASE API Client DLL" COMPANY_NAME="Microsoft Corporation" PRODUCT_NAME="Microsoft® Windows® Operating System" FILE_VERSION="5.2.3790.3959 (srv03_sp2_rtm.070216-1710)" ORIGINAL_FILENAME="kernel32" INTERNAL_NAME="kernel32" LEGAL_COPYRIGHT="© Microsoft Corporation. All rights reserved." VERFILEDATEHI="0x0" VERFILEDATELO="0x0" VERFILEOS="0x40004" VERFILETYPE="0x2" MODULE_TYPE="WIN32" PE_CHECKSUM="0x100243" LINKER_VERSION="0x50002" UPTO_BIN_FILE_VERSION="5.2.3790.3959" UPTO_BIN_PRODUCT_VERSION="5.2.3790.3959" LINK_DATE="02/17/2007 13:58:24" UPTO_LINK_DATE="02/17/2007 13:58:24" VER_LANGUAGE="English (United States) [0x409]" />
</EXE>
</DATABASE>

...

Please, gve me some sugestion!

Thanks!!!!

Giacomo

Post by **fmk** » Fri Feb 05, 2010 7:20 am

send me the model and the ground motions.

giacomodemarie · Post by **giacomodemarie** » Mon Feb 08, 2010 4:56 am

I sent it!
Thank you very much!!!

giacomodemarie · Post by **giacomodemarie** » Thu Feb 11, 2010 1:14 am

I sent it!
Thank you very much!!!