| Author | Message | ||
     Shidrati Ali (shidrati) New member Username: shidrati Post Number: 1 Registered: 04-2004 |
Hi, I'm working on a fluid-structure interaction and I have a problem with declaring the boundary conditions at the interface. For my problem I need to declare the b.c. for not only displacements but forces like the stress tensors. However my variables are only the pressure and displacement and FLEXPDE only allow b.c. for variables. Do someone know how I can overcome this problem? Thanks, Shidrati | ||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 141 Registered: 06-2003 |
Value boundary conditions will apply displacement values, and natural boundary conditions (usually) apply forces. So you can pick the one you want to apply. You cannot apply boundary conditions to both the force and displacement at a point, because the PDE implies a relationship between the two which cannot be abrogated. A fixed displacement implies the force required to maintain it, etc. See the documentation about Natural boundary conditions to determine the meaning of the Natural for your equations. | ||
| Shidrati Ali (shidrati) New member Username: shidrati Post Number: 2 Registered: 04-2004 |
Hi Mr. Nelson, Thank you for your pointer on the boundary conditions. But my case is slightly different. I have the following b.c. applied at the fluid-structure interface: u.n = C.dn(p) n.tau.n = -p n (cross) tau.n = 0 There are two different mediums where in the structure side, I need to satisfy the balance of the forces through the stress tensor and on the fluid side, I need to satisfy the displacement of the fluid. Should I declare the stress tensor as variables as well in this case? Thanks, Shidrati | ||
| agatha de la torre (ph6undrgrad) New member Username: ph6undrgrad Post Number: 1 Registered: 07-2005 |
I have a problem declaring the conitnuity condition for heat conduction in a composite slab. How do i represent the temperature condition at the interface of two different materials? I need help | ||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 414 Registered: 06-2003 |
The Natural Boundary Condition defines both the meaning of boundary conditions on exterior boundaries and also the continuity conditions at interior interfaces. (See Natural Boundary Conditions in the Help Index.) The quantity defined by the Natural BC is assumed continuous across interior interfaces. If you have written your PDE as the divergence of thermal flux, ie div(k*grad(Temp)) or div(-k*grad(Temp)), then the Natural Boundary Condition is defined as the value of the surface-normal component of the flux, k*grad(Temp). In this formulation, the thermal flux k*grad(Temp) is assumed continuous across interior interfaces. This is true even if k is different in the two materials. So, if you want to conserve energy, use the flux-divergence form of the PDE and specify nothing internally. If you want internal surface sources or sinks, then you can define a non-zero Natural BC internally. This value will then be interpreted as a surface source (or sink). If you want to represent a contact resistance, then use the CONTACT form of boundary condition. (See CONTACT in the Help Index). | ||
| sandeep (snallagu) New member Username: snallagu Post Number: 1 Registered: 08-2005 |
Hi All, I am trying to write the energy balance at an interface to find out the temperature at a node(interface). As it is an interface of two mediums A and B, both media have their own thermal resistances. There is also a contact resistance at the interface. The temperatures of previous and next nodes are known. Every thing else is known except the temperature at the node(interface). Can somebody help me in writing the energy balance at the node(interface). I would appreciate it. | ||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 431 Registered: 06-2003 |
As pointed out in the documentation, the Natural Boundary condition is the basis of energy conservation. The quantity defined as the Natural Boundary condition is considered to be conserved across all interior interfaces. In a diffusion system, for example, with the equation div(k*grad(u))+s=0, the natural BC is the normal component of k*grad(u). This quantity is conserved across all internal interfaces, whether k is the same or different across the interface. In this case, the Natural BC is the flux of U, so flux is conserved. The standard model places nodes on the boundaries that are shared by the adjoining materials. There is therefore a single temperature at the interface, which is dictated by the flux conservation laws described above. At a CONTACT boundary (where you have specifically used the CONTACT specification), there are duplicate values at each node, one on each side of the interface. The flux is defined as the JUMP of the variable divided by the contact resistance (See the documentation of CONTACT boundaries). Since the definition is in terms of flux, the flux is again conserved. The normal component of k*grad(u) on one side of the interface will be equal to the normal component of k*grad(u) on the other, and both will be equal to the flux defined by the contact boundary condition. So, you don't have to write any energy balance at the interfaces. It is all implied by the way you write the equation and what the attendant Natural Boundary Condition means. |