FEM Modeling > Verify Action > Introduction to Verification
XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX''">   
Introduction to Verification
Model verification consists of a number of different tests which can be performed to check the validity of a finite element model. These tests include checks of element distortion, element duplication, model boundaries, nodal connectivity, and node⁄element ID numbering.
In the case of distortion checking, Patran provides a series of automated tests to measure the “distortion” of elements from an “ideal” shape through measurable geometric properties of the element. The results of these tests are compared to user specified criteria and a determination is made whether the element is acceptable or not. The pass⁄fail criteria is analysis code dependant and is updated automatically when the Analysis Preference is changed.
Verification tests are always performed on the current group of the active viewport except in the case of duplicate elements in which case the entire model is checked.
To get an overview when checking a specific element type, there is a test choice of All. When this is selected Patran will display a spreadsheet showing a summary of the total number of elements that exceed a threshold value for each of the distortion checks, and the actual test value and element ID number for the most extreme element.
Model Verification provides visual feedback of the selected test. Element distortion checks allow the selection of a threshold value using a slidebar. During the check, any element, which exceeds the threshold value, is highlighted and its value is listed in the Command Line. Upon completion of the check Patran will color code the elements based on the computed test value. Elements with a value higher than the threshold are colored with the highest spectrum color, all other values are assigned uniformly through the other spectrum levels. The current group will be rendered using the Element Fill style. Verification forms for Quad elements include an icon that allows a selection to split failed elements or simply highlight them.
Other checks, such as element duplication and connectivity, give options only to highlight any offending elements, or automatically correct the model.
Model boundaries may either be displayed as edge lines, showing unshared edges in the model, or as shaded faces, showing unshared surfaces.
All verification tests that involve color-coding, shading, or some other method of re-rendering the model have a Reset Graphics button on the form. Selecting this button will undo any rendering procedures performed by the most recent verification activity. The render style and spectrum display will be returned to the pre-test settings they had before the Apply button was selected. If you will be performing more than one type of verification test, it is recommended to choose Reset Graphics after each test is completed. Remember Reset Graphics resets to the settings prior to the current verification activity, not to those at the start of all verification.
All element specific verification forms have a Normalize button. By default, the normalize option will not be selected, and the slidebar will represent an actual value for the verification test threshold. If the normalize option may be selected, the slidebar will now represent a range of values from zero to one. The value of zero will represent the most reliable shape for this element type.
All element specific verification forms also have a Reset button. Selecting this button returns the slidebar and all toggles to the settings they had when the form was opened.
The information obtained from verification procedures can assist the engineer in deciding if the finite element model is satisfactory, or should be adjusted through remeshing or element modification.