| Author | Message | ||
     Mark (everest20) New member Username: everest20 Post Number: 1 Registered: 01-2009 |
Dear Mr. Nelson I need to apply the following boundary condition on the interface between two materials: #################################### k1*(dT/dr)1-k2*(dT/dr)2 = f(T) #################################### There is a temperature jump through the interface ,and this is what makes me confused, therefore I have few questions: 1 - Can the problem be modeled with CONTACT and JUMP commands? If yes, how should I apply it to the code? 2 - Can these commands be used with 1D spherical coordinates? Best regards, Everest | ||
| Robert G. Nelson (rgnelson) Moderator Username: rgnelson Post Number: 1212 Registered: 06-2003 |
You have requested two different things here. 1. The condition k1*(dT/dr)1-k2*(dT/dr)2 = f(T) implies an energy source (or sink) imbedded in the boundary surface. This is exactly the meaning of a Natural BC on an interior boundary. Simply define Natural(T)=F(T), and the source will be inserted on the boundary. 2. The CONTACT boundary condition is an implementation of the idea of a contact resistance. It is based on the conservation of flux across the resistance, so is incompatible with (1). There is no way to program both a variable jump and a flux gain (or loss) at the same time. | ||
| abbasali New member Username: abbasali Post Number: 1 Registered: 04-2010 |
Dear Dear Mr. Nelson I need to apply the following boundary condition on the interface between two materials: k1*(du/dn)1=k2*(du/dn)2 how can I write the related code to this boundary condition best regards | ||
| rgnelson Moderator Username: rgnelson Post Number: 1345 Registered: 06-2003 |
The meaning of the NATURAL() boundary condition depends on the equation to which it is applied, and since you have not posted your equations, I can only guess. For an equation with the second-order term term Div(k*grad(u)), the Natural boundary condition means the outward normal component of k*grad(u) at the boundary. This is also the quantity that is by implication continuous across interior interfaces. So if your equation contains the diffusion term above, the condition you require is automatically satisfied without any additional specification. See Natural Boundary condition in the Help Index. | ||
| abbasali New member Username: abbasali Post Number: 2 Registered: 04-2010 |
Dear Mr. Nelson I need to apply the following boundary condition (Ben Danial-Duke boundary conditions) on the interface between two materials: k1*(du/dn)1=k2*(du/dn)2 the related equation is -*div(k*grad(u))+v*u=lambda*u regards. | ||
| rgnelson Moderator Username: rgnelson Post Number: 1346 Registered: 06-2003 |
See my previous post. Your desired condition is the default. Define k=K1 in one region and k=k2 in the other. |