The OpenSees Community

Thanks Frank.

I was able to stop building the tk version, and take away the tk dependency by:

1) Removing all references to libtk in the makefile.def (all -ltk)
2) making via "make OpenSees" which only builds the tcl version.

Is it possible to build a non-tk dependent OpenSees? If I understand correctly, tk is a GUI building tool that works hand in hand with tcl. When I do an OpenSees build, I get a normal OpenSees executable that runs without any GUI, and a tk version. I ask, because I am trying to build on a cluster ma...

JCEE-

I never ended up submitting the slip-lock material to FMK. I'll drop him an email, and see if he wants to add it. Otherwise, feel free to drop me a PM here with your email address, and I can email you the source code.

Thanks!

Question: I have a custom material I am using on a zero-length element. I know that zero-length elements are not included in Rayleigh damping, however, it looks like you can use a damping tangent value to include damping. Lets say I am using a rayleigh stiffness damping of 0.0023. I have a gap eleme...

fmk wrote:suggest get it straight from the element.

element1->getTangentStiff()

Thanks Frank.

Can you tell me the proper code to access an element from a given id? (Ex- the element1 portion of the code above)

Given an element number of a zeroLength element, and a direction how would I access that element's stiffness in the given direction? I dont mean from the interperater, but in a programmed object? Example: I create a zero-length element, id 1, and it has material 1 in direction 1 and material 2 in di...

A material has two state stiffnesses: the trial stiffness, and the committed stiffness. The trial stiffness is cacluated at each iteration (as the non-lineararities are stepping to converge), whereas the committed stiffness is the final trial stiffness that occurs at convergence. So each analysis ti...

Thanks Nero, that is exactly what I mean. In this way you can send a script to your collegues and they will immediately receive the proper error (incorrect version of openSEES) rather than a cryptic error about a material model not being present. One problem I see is this- none of the past versions...

doesn't it say it on top??? I may be wrong, but I think what he means is the ability to grab the version while executing a script. That way you could compare the current version to any requirements the script has, and then output a warning/error if the version is too old. Ex, if there was a command...

NP

2 things- 1) In the zero-length element description, it claims that the recorder command to record element stiffness is "stiffness"- this command is actually "stiff" 2) For the Truss element, there is no documented deformation recorder. To record deformations for truss elements, ...

Ever since moving up to 1.7, I cant seem to perform an Eigen analysis. I have tried an extremely simple model, and when I try to get the eigen values, I get the following errors: BandArpackSolver::Error in dgbtrf_ Warning FrequencyAlgo::solveCurrentStep() - the EigenSOE failed in solve(). EigenAnaly...

I have noticed that one of my models is about 2x faster using the UmfPack solver than any of the other solvers (and about 4x faster than the BandGeneral), but a different model is faster using SparseGeneral -piv. Is there any rule of thumb as to what types of models, etc.. perform better with a spec...

I am attempting to use OpenSees with the ACML, AMD's tuned math routines (which include BLAS and LAPACK). Has anyone here compiled against these libs? Everything compiles fine when I set my BLAS and LAPACK libs to where the ACML libs are installed, but any of the systems that use BLAS or LAPACK are ...