[nitinnitin]

If I have two concentric rings, A and B. Initially at time T=0 sec, both are at 373 degree Celsius. A time T= 0 Sec, heat flux of 45000 W/m2 is applied at A. Now transient analysis takes place. The gap between A and B is filled with Steam. If we have to simulate Natural convection for 30 minutes, So what options we should take?

Initial condition is that , gap between A and B is at atmospheric pressure of 1 bar and saturation temperature of 100 Degree Celsius.

Figure is attached here with.

Can we use Boussinesq parameter for it , for transient analysis or not?

[virendra_p]

Sf. A bc is solid wall with heat source...sf. B bc is solid wall...select steam as fluid....turn gravity on with specified operating density of steam...patch the entire domain with temp. value of 100 deg. for initial condition....and the flow will be non-turbulent so viscous model should be laminar....hopefully this helps.... please someone double check me

[nitinnitin]

In my knowledge , Boussinesq approximation is used for small temperature difference.

But as the inner ring is getting constant heat flux , its temperature will be very high with time. So in transient case can we use boussinesq approximation?

[virendra_p]

Based on area of inner s/f and time of simulation...we can determine whether the heat source is weak or strong....and then use BA if it turns out to be a weak source...but you can also run simulations with both methods and compare results...

[nitinnitin]

Originally Posted by virendra_p

Based on area of inner s/f and time of simulation...we can determine whether the heat source is weak or strong....and then use BA if it turns out to be a weak source...but you can also run simulations with both methods and compare results...

Ok , i undestood what you said.
Suppose with transient of 30 minutes , inner circle temperature reaches to 800 degree Celsius and outer circle reaches to 300 degree Celsius. In that case Boussinesq Approximation is valid?

Or we have only options to check results be comparing with other methods?

As in starting at T=0 sec, both circles are at 100 degree celsius.

Actually what happens , when i am using boussinesq approximation , i am getting convergence. While i am using other options of density (Peicewise linear profile), I am not getting convergence. Does it happens with others also?

Regards
Nitesh

[virendra_p]

See if this works....use BA initially...choose small time step say 5 millisec...and longer convergence like 50 iterations per time step....after some 10 time iterations..switch to linear profile..if divergence occurs adjust the timestep/rf....i guess comparison is the only way to check applicability of BA for this case...good luck!

[nitinnitin]

Thanks for giving your ideas