Hello ppl,

I'm trying to formulate the flow past a block in an open domain. I've a fractional step scheme on hand and I'm using the finite volume method. I've used it to test Ghia's lid driven cavity and results are very similar.

I believe I just need to modify the boundary conditions (BCs) for my new problem. I was reading Blazark's CFD book and these are the BCs given for subsonic inflow and outflow (mine is an incompressible problem)

For inflow a belongs to outside of domain, b is on the boundary, d is inside domain. a .... b .... d

p(b)=0.5*{p(a)+p(d)-rho(0)*c(0)*[nx*(u(a)-u(d))+ny*(v(a)-v(d))]}

rho(b)=rho(a)+(rho(b)-rho(a))/c(0)^2

u(b)=u(a)nx*(p(a)-p(b))/(rho(0)*c(0))

v(b)=v(a)nx*(p(a)-p(b))/(rho(0)*c(0))

rho(0) & c(0) represents reference state

For outflow a belongs to outside of domain, b is on the boundary, d is inside domain. d .... b .... a

p(b)=p(a)

rho(b)=rho(d)+(rho(b)-rho(d))/c(0)^2

u(b)=u(d)nx*(p(d)-p(b))/(rho(0)*c(0))

v(b)=v(d)nx*(p(d)-p(b))/(rho(0)*c(0))

p(a)=prescribed static pressure

So, is this BCs applicable since this is an incompressible problem?

Also in most problems, ppl usually specify inlet velocity (e.g. u=1,v=0). Is there a need to specify pressure at

inlet as well?

When you specify velocity you do not usually specify pressure.Have a look at peric and ferziger for more details.

-H