Hi all
I am trying to implement birth-death technique in order to implement oxidation of a material. It means I want the material assigned to the element changed after some amount of time. This material is in perfect contact with the already existing oxide layer which then grows by more oxidation. Anyone can help me to use technique for "replacement"?
I have modeled initially as one part and partitioned it in two parts assigning different materials to each part. I think I should add another part and create the elements that I want to activate after some time.
I was planning to do it using USDFLD but it is very costly in terms of time, because my model consists of many layers. The oxidation start from one layer and with time it extends to the next layers. Therefore, I should define many materials for all the layers. Accordingly, the USDFLD code will also be too inefficient and large.
Thanks
Dear Vahid Ebrahimzade Isfahani
Birth-death technique means you should model all the elements and materials at the first time. And activate what you need each time.
With USDFLD, you can have just one material and change its properties according to the value of the USDFLD by using DEPVAR and field-dependent material properties.
E.g.
*Density, dependencies=1
1000, ,0
2000, ,1
Then you program your USDFLD/UFIELD so that it changes from 0 to 1 when you do the material change and the density changes accordingly from 1000 to 2000.
It sounds like you want to use the dummy technique rather than the element birth and death technique. You can simply use field variable for the material properties for the dummy or use model change for the birth and death technique.
Thanks to all of you, it was helpful.
@Mr. Bäcker
Thanks a lot.
The problem is that I want to model oxidation of the bond coat layer in a TBC system. The interface has a sinusoidal shape and the TBC system has an edge, and because of this edge the use of Swelling option leads to instabilities of the interface especially at the edge.
I am planning to partition the BC layer several times, with a sinusoidal profile, beneath TGO. So I obtain several thin layers (in range of less than 1 micrometer: ca. 0.5 micrometer). And with increasing hold time, with consideration of TGO oxidation kinetics, these layers start to change, from the top one to the next, until a critical thickness reaches and failure happens.
Now I think using USDFLD I must define different sections for each layer with the same material(BC), but with different material names, because for each layer the FV changes after a different time, so that the sequential order of the oxidation of these layers from top to low will be hold. I checked USDFLD and it does not recognize different element sets to avoid such extra works.
Nevertheless, using only these 2 ways, laterla growth of TGO will not be considered.
But anyway maybe USDFLD is a better way. I would appreciate if you have any tip to make this process better and easier.
- Badges
- Science topic