ainex
Geotechnical
- Jan 23, 2014
- 1
Dear all,
currently I am modelling a filtration process in ABAQUS/Standard. The aim is to change material properties of an actual element/integration point after a certain discharge Q is reached (step:soils, consolidation, C3D8PR Elements). Initially, a parameter set "material 1" is assigned to the elements, which changes to "material 2" when Q is reached. Then it stays constant. The enquiry of Q and change of material parameters is done in *UMAT and *USDFLD and works ok. My problem is, that the enquiry should only happen for elements which are directly adjacent to elements which already are "material 2".
Our first idea was to use the state variable TEMP to identify elements lying directly at the border between material 1 and 2. 'This was done by assigning initial condition TEMP=0 to Elements with "material 1" and TEMP=1 to Elements with "material2". The integration point of elements with material 1 just adjacent to elements with material 2 get the value TEMP = 0.5, due to integration from initial conditions defined at nodes to the integration point. So, Q was only assessed at elements with TEMP=0.5, and if the desired Q is reached, TEMP was changed to 1. Unfortunately this TEMP value is assigned to the integration point and the next element does not get TEMP=0.5 due to interpolation between nodes...
So my question would be:
1. Do you thing using TEMP to identify the regions in the model is the right way? If so, is it possible to change the TEMP values at nodes rather than integration points within UMAT or USDFLD, so that integration gives TEMP=0.5 at the integration point of the next adjacent element?
2. Does anyone have an other idea how in a subroutine adjacent elements can be identified and properties/state variables imported to a subroutine?
3. Do you have any other idea how to identify elements which are are located next to elements with a certain parameter/variable?
Any suggestions welcome!
Best regards,
ainex
currently I am modelling a filtration process in ABAQUS/Standard. The aim is to change material properties of an actual element/integration point after a certain discharge Q is reached (step:soils, consolidation, C3D8PR Elements). Initially, a parameter set "material 1" is assigned to the elements, which changes to "material 2" when Q is reached. Then it stays constant. The enquiry of Q and change of material parameters is done in *UMAT and *USDFLD and works ok. My problem is, that the enquiry should only happen for elements which are directly adjacent to elements which already are "material 2".
Our first idea was to use the state variable TEMP to identify elements lying directly at the border between material 1 and 2. 'This was done by assigning initial condition TEMP=0 to Elements with "material 1" and TEMP=1 to Elements with "material2". The integration point of elements with material 1 just adjacent to elements with material 2 get the value TEMP = 0.5, due to integration from initial conditions defined at nodes to the integration point. So, Q was only assessed at elements with TEMP=0.5, and if the desired Q is reached, TEMP was changed to 1. Unfortunately this TEMP value is assigned to the integration point and the next element does not get TEMP=0.5 due to interpolation between nodes...
So my question would be:
1. Do you thing using TEMP to identify the regions in the model is the right way? If so, is it possible to change the TEMP values at nodes rather than integration points within UMAT or USDFLD, so that integration gives TEMP=0.5 at the integration point of the next adjacent element?
2. Does anyone have an other idea how in a subroutine adjacent elements can be identified and properties/state variables imported to a subroutine?
3. Do you have any other idea how to identify elements which are are located next to elements with a certain parameter/variable?
Any suggestions welcome!
Best regards,
ainex