Example of creating custom result

Pernelle Marone-Hitz

Do we have an example of creating a custom result based on derived values?

Pernelle Marone-Hitz

The following example illustrates how a custom result can be created to evaluate the Normal SX stress value. The following are shown:

• creating a result based on "elemnode" location
• having the possibility to define averaging option from the .xml file and result.ResultAveraging enum
• appropriate handling of the unit conversion

The .xml should read:

    <extension version="1" name="Stress Postprocessing">

        <guid shortid="Stresses">a1844c3c-b65c-444c-a5ad-13215a9f0413</guid>

        <script src="main.py" />

        <interface context="Mechanical">
            <images>images</images>

            <toolbar name="Normal X Stress" caption="Normal X Stress">
                <entry name="SX" icon="result">
                    <callbacks>
                        <onclick>CreateSXResult</onclick>
                    </callbacks>
                </entry>
            </toolbar>

        </interface>

        <simdata context="Mechanical">
            <result name="SX Stress" version="1" caption="SX Stress" unit="Stress" icon="result" location="elemnode" type="scalar">
                <callbacks>
                    <evaluate>EvaluateSX</evaluate>
                </callbacks>
                <property name="Geometry" caption="Geometry" control="scoping"></property>
                <property name="DisplayOption" caption="Display option" control="select" default="Unaveraged">
                    <attributes options="Unaveraged,Averaged,Elemental Mean"></attributes>
                    <callbacks>
                        <onvalidate>ChangeDisplayOption</onvalidate>
                    </callbacks>
                </property>
            </result>
        </simdata>
    </extension>

While the main.py Python file should read:

    import units
    import math

    def CreateSXResult(analysis):
        analysis.CreateResultObject("SX Stress", ExtAPI.ExtensionManager.CurrentExtension)


    def EvaluateSX(result,stepInfo,collector):
        '''Evaluates SX as a custom result, with different averaging options.'''
        # Reader initialization
        reader = result.Analysis.GetResultsData()    
        reader.CurrentResultSet = stepInfo.Set
        # Get the stress result from the reader
        stress = reader.GetResult("S")
        stress.SelectComponents(["X"])
        resultUnit = stress.GetComponentInfo("X").Unit
        convFactor = units.ConvertUnit(1.,resultUnit,"Pa","Stress")   
        
        # Calculate and plot result
        for elemId in collector.Ids:
            elemVal = stress.GetElementValues(elemId)
            values=[]
            # Use unit conversion factor
            for value in elemVal:
                values.append(value*convFactor)
            # set values to collector
            collector.SetValues(elemId,values)
                         
    def ChangeDisplayOption(result,prop):
        displayOpt = result.Properties["DisplayOption"].Value
        if displayOpt == "Unaveraged":
            result.ResultAveraging = ResultAveragingEnum.Unaverage
        elif displayOpt == "Averaged":
            result.ResultAveraging = ResultAveragingEnum.Average
        elif displayOpt == "Elemental Mean":
            result.ResultAveraging = ResultAveragingEnum.ElementalMean

Pernelle Marone-Hitz

Please note that more recent technologies are available to create customized plots in Mechanical : DPF and Python Results. These posts can be used as a starting point:
https://developer.ansys.com/blog/script-tip-friday-examples-python-results-mechanical-part-1
https://developer.ansys.com/blog/script-tip-friday-examples-python-results-mechanical-part-2
https://discuss.ansys.com/discussion/3074/getting-started-with-dpf-in-mechanical?utm_source=community-search&utm_medium=organic-search&utm_term=getting+started+DPF

Philb78

Hello,

I am developping an ACT extension to perform some custom results on shell elements.
And i get a warning that i don't understand. So i have tried this above example on shell elements and i get the same warning when i execute collector.SetValues :

"The result collector has been resized because the number of values provided for location 1: 12 in SetValues method does not match with the number of values expected: 6. Please check carefully your results."

As a shell, we have normally 3 results by nodes so 12 results for a quad shell. So i don't understand why the collector is resized to 6 values.
And Moreover, the results seems correct.

Regards
Phil

Pernelle Marone-Hitz

Hello @Philb78 , I'd highly encourage you to use a Python Result + DPF to create your custom result. The technology is much more powerful, it will execute faster and you won't need to handle the .xml file.
For the size of the result collector, I guess that you have degenerated elements in the model and/or are using a old version of Ansys (I'd recommend using the latest one if possible, 24R2).

Philb78

Hi,

thanks for quick answer, i am using ANSYS 2023R2 and i don't have degenerated elements.
Moreover when i use NumberValuesByElement([elemid]) on a stress result, i get 12 values :

Example of element n°381 :

The result collector has been resized because the number of values provided for location 381: 12 in SetValues method does not match with the number of values expected: 6. Please check carefully your results.

res = ExtAPI.DataModel.Project.Model.Analyses[1].GetResultsData()
stress = res.GetResult("S")
stress.SelectComponents(['X'])
stress.GetElementValues(381).Count
12

This extension will be used by several people, so i prefer to encapsulate it in an ACT extension in order to deploy it.

Pernelle Marone-Hitz

This element indeed clearly expects 12 values. The fact that the collector expects only 6 is probably an issue related to the collector itself, or simply an issue with the messaging. Please try in 24R2 to check if problem is resolved. If not, please contact your local support provider and share a simplified extension + model, we will investigate and if need be open a bug.