SimCenter Forum

Research in Natural Hazards Engineering => Uncertainty Quantification (quoFEM) => Topic started by: Rashad on October 15, 2021, 05:10:09 PM

Title: Optimal parameters values
Post by: Rashad on October 15, 2021, 05:10:09 PM
I have chosen values of some parameters in opensees model, and I want to know if my choice is satisfied (values are not low, not high ). I tried using the parameter Estimation method, but quoFEM gives me the same values that I input in the initial point values in ‘RV’. I do not have an experimental and the parameter Estimation method is obtaining the parameters that best fit the experimental results. so could please, guide me on which method can I use for conforming parameters values?

Title: Re: Optimal parameters values
Post by: AakashBS on October 15, 2021, 09:47:28 PM
Hello Rashad,

1) The parameter estimation methods require data to be provided (either from experiments or otherwise), since the data are necessary to define how 'good' different parameter values are relative to each other.

In the absence of data, if there are no recommendations in literature of good parameter values for your problem, you might want to first explore how the parameter values map to the outputs to get an idea if the parameter values that you have chosen are satisfactory. You can do this by performing either forward propagation or global sensitivity analysis in quoFEM by defining uniform distributions for the parameters in the RV panel and setting wide bounds corresponding to the range of values that you want to explore. Inspecting the scatter plots of the parameter values vs. the outputs in quoFEM's RES panel after the analysis (as in the example shown in the screenshot attached) will show you the range of outputs that are produced corresponding to different parameter values. Sensitivity analysis results will also tell you how much influence each of the parameters in the range that you have defined will have on the outputs from your model. You can check whether the parameter values you have used can produce the outputs that you want. If not, try expanding the range of the parameter values and run the forward propagation/sensitivity analysis again. You can then pick a particular target output value from this plot, provide it as data and perform parameter estimation in quoFEM.

Once you have chosen some parameter values, you can specify a narrow Normal distribution centered at those parameter values and run a forward propagation analysis. You can check the outputs produced and judge if the parameter values are satisfactory.

2) If the finite element analysis is running successfully, then the parameter estimation methods should find parameter values that produce outputs from your model that best match the data. If your analysis is running successfully but you still are having any issues performing parameter estimation in quoFEM, please get back to us with more details (the 'ops.out' and 'dakota.out' files) and we can help resolve the issue.

Title: Re: Optimal parameters values
Post by: Rashad on October 18, 2021, 02:01:41 PM
Thank you,
I tried running the model but there are no results all displacements are zero. Could you please, have a look at screenshots 1,2. in ''LocalWorkDir'' file there is ''tmp.SimCenter'' file has ''templatedir'' and  ''workdir.1-300'' all of the folders are empty.
Title: Re: Optimal parameters values
Post by: Sang-ri on October 18, 2021, 03:17:49 PM
Hi Rashad,

Thanks for sharing the screenshots. The quoFEM inputs in your first screenshot look good. For the workdir.1-300 folders, probably you will see the files inside from windows explorer, but not from the quoFEM browser shown in the second screenshot (because it is showing you only folders and not files).

I think the opensees model is either not converged or giving only zeros for all input parameters. We can have a look at your input files if you could share the tcl files.
Title: Re: Optimal parameters values
Post by: Rashad on October 18, 2021, 03:45:25 PM
Thank you, please check the Attachments.
Title: Re: Optimal parameters values
Post by: AakashBS on October 18, 2021, 05:35:29 PM
Thanks for sharing the files, Rashad.

The model script was running the OpenSees analysis successfully, but the issue was caused by the post-processing script. This post-processing script uses the QoI name that you provide in quoFEM to extract the corresponding output. In this case, the QoI name provided was 'Node_16_Disp_1', so the script was trying to find the 1st component of the displacement of node 16. But, since there were only 6 nodes in the model, the post-processing script was unable to find a node with the tag 16 and was instead simply writing an output of 0.

There are two ways you can fix this issue -

a) Simply rename the recorder output file name to 'results.out' instead of 'Model1.out' in Line 74 of the 'Model1.tcl' script.
Then, the post-processing script is not necessary, and should not be provided in the FEM panel of quoFEM. If a post-processing script is not being used, then you are also free to set any valid variable name in the QoI panel.

b) Alternatively, if you would like to use this post-processing script, then you should

Title: Re: Optimal parameters values
Post by: Rashad on October 18, 2021, 06:06:43 PM
actually, I always use the right variable name 'Node_16_Disp_1' in the QoI panel of quoFEM but I change node 16 to node 2 when I sent the opensees file. anyway, the issue was due to using ''wipe'' in Tcl file.
Now I have got the results. THANK YOU SO MUCH