You Have to Make the Analysis Fail

If you coded a UniaxialMaterial or any other constitutive model or element in OpenSees, you had to implement pure virtual methods commitState(), revertToLastCommit(), and revertToStart(). The behavior of each method is pretty clear: commitState() - After convergence of the global solution at the current time step, get the material ready for the next time step. This typically involves swapping history variables … Continue reading You Have to Make the Analysis Fail →

How to Use pytest with OpenSees

Despite plentiful constitutive models and analysis options, testing and verifying OpenSees has been quite limited. At this point, going back and testing all the contributions from the last 25 years is a nearly insurmountable task. But, as the saying goes: The best time to start testing OpenSees was 25 years ago.The second best time is … Continue reading How to Use pytest with OpenSees →

Counting Flops

When I was an undergraduate at Pine State University, all engineering freshmen had to take a programming course. We could pick between Fortran, Pascal, and C++. From what I recall, most civil and mechanical engineering students took Fortran because that's how you crunch numbers. I chose C++ and liked it so much I willingly took … Continue reading Counting Flops →

How to Profile an OpenSeesPy Analysis

Python has a couple of profiling libraries--pyinstrument and cProfile--for finding out where all the time goes when you run a script. But, as far as I can tell, these libraries only tell you that the ops.analyze() command is called, not what happens therein. What you really want is to drill down into the state determinations … Continue reading How to Profile an OpenSeesPy Analysis →

Getting the Number of Nodal DOFs

Do you ever get tired of typing a lot of zeros, e.g., to apply a 20 kip load in the X-direction at node 18 on a 3D frame model? ops.timeSeries('Linear',1) ops.pattern('Plain',1,1) ops.load(18,20,0,0,0,0,0) I know what you're thinking. Did I put five zeros after the 20 kip load or only four? Well, to tell you the … Continue reading Getting the Number of Nodal DOFs →

Dangling Recorders

Because I don't perform very large or time-consuming OpenSees analyses, I use recorders rather infrequently, instead preferring commands like nodeDisp and eleResponse. An issue with node and element recorders is they can segmentation fault after the node or element to which they point is removed from a model--a common scenario for progressive collapse simulations. A … Continue reading Dangling Recorders →

Is It Close Enough?

The locations and weights for Gauss-Lobatto beam integration, the de facto beam integration for force-based elements, are hard-coded in the OpenSees source code. For most cases in the LobattoBeamIntegration class, the locations and weights are written to only ten significant figures instead of 16 or more. Although I am certain that leaving six sig-figs on … Continue reading Is It Close Enough? →

Set and Get Concrete23 Response

When I wrote Concrete23, I copied Concrete01 then tweaked the unloading and reloading rules. I also added a couple of bells and whistles and it was music to my ears. But I wanted to record those sounds during an analysis. Recorders in OpenSees use the setResponse and getResponse methods to identify and obtain element, section, … Continue reading Set and Get Concrete23 Response →

Valid Range of OpenSees Tags

A few draft blog posts are swirling around the abyss of approximately 80% complete. In need of a diversion, I found a question about tags, posted on the OpenSees Facebook group, to be perfectly timed. The question was about strange behavior when using large integers for node and element tags. Not many other details were … Continue reading Valid Range of OpenSees Tags →

Monte Carlo Simulation with OpenSeesMP

The parallel computing capabilities of OpenSeesSP and OpenSeesMP are easily confused. OpenSeesSP runs your script on a single processor with the other processors awaiting instructions on what to do. OpenSeesSP is ideal for assigning subdomains of a large model to each processor. The main processor, processor 0, directs traffic and solves the governing equations of … Continue reading Monte Carlo Simulation with OpenSeesMP →