Community > Discussion Forum > Non isothermal flow

Discussion Forum

Non isothermal flow

Topics: Heat Transfer, Fluid Flow, 4.1, 4.2a

Thread index  |  Previous thread  |  Next thread  |  Start a new discussion

RSS FeedRSS feed   |   Email notificationsTurn on email notifications   |   6 Replies   Last post: July 23, 2012 9:55am UTC
Pavan Kumar Penumakala

Pavan Kumar Penumakala

May 24, 2012 1:32pm UTC

Non isothermal flow

Hello all,,

I am solving Non isothermal flow ( with solidification included) in a domain.
both turblent flow and heat transfer are involved..

I followed the continuous casting model in model library..

When i am trying to solve my model i get the error

Failed to find a solution.
Segregated group 1

Singular matrix.

There are 16 void equations (empty rows in matrix) for the variable mod2.nitf.TWall
and similarly for the degrees of freedom (empty columns in matrix)
Returned solution has not converged.


what is Twall ..

is there any other model in comsol which demonstrates the solution of turblentflow and heat transfer to gether

thanks

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Ivar Kjelberg

Ivar Kjelberg

May 24, 2012 6:29pm UTC in response to Pavan Kumar Penumakala

Re: Non isothermal flow

Hi

For me it probably means you have not given enough BCs to get an unique solution to your PDE. When using multi-physics models you still need to give enough BC valus FOR EACH physics to force your PDEs to have an unique solution. It's quickly done to overlook one or two.

Check all your dependent variables do they have 1-2 BC forcing them to a constant value or defining the corresponding fluxes ?

With NITF you must also select if the inflow is rather from the fluid (driving the equations) or from the solid side

--
Good luck
Ivar

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Pavan Kumar Penumakala

Pavan Kumar Penumakala

May 25, 2012 3:08pm UTC in response to Ivar Kjelberg

Re: Non isothermal flow

Dear Mr.Ivar

thanks for the reply...

i uploaded my model here ..

boundary 9 is the inlet for fluid..3, 4, 5, 6,7,8,10 are walls..

for the heat transfer, on wall 10 the HTC is specified ...
and i gave fluid input tempearure..
it involves solidification so i added source terms accordingly..

i do not figure out what BC i am missing

thanks

Attachments:   uploadetoivar.mph  

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Ivar Kjelberg

Ivar Kjelberg

May 26, 2012 10:56am UTC in response to Pavan Kumar Penumakala

Re: Non isothermal flow

Hi

I cannot do much ,as it seems that you have some issues with the turbulent flow set-up and CDF is not my speciality, if I play with laminar flow sometimes, I extremely seldom run turbulent flow, as this is a speciality of its own.

Newertheless a few comments:
Mesh: no need to use a minimum mesh size larger than the maximum, that should be made mode consistent

You need a wall boundary on all no-slip wall, this is not the case, try using the default COMSOL physics induced mesh as starting point

I would leave the outflow so your temperature might follow the fluid flow.

It could also be your equations that are triggering the error, but the way I read it you are missing somewhere on the wall or turbulence setting something that prevents COMSOL from finding an unique solution

--
Good luck
Ivar

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Pavan Kumar Penumakala

Pavan Kumar Penumakala

June 16, 2012 1:42pm UTC in response to Ivar Kjelberg

Re: Non isothermal flow

Hello Mr.Ivar

i found that i need to add a source term in the governing equations of Turblent KE and Turbulent KE dissipation...

while solving CC in model libarary , to dampen the velocity in ths solid region source term F is used.
now in turblence equation to dampen the turblent KE generation and KE dissipation in the solid region..
i want to use similar kind of source term in the governing equations of k and e.

rho*u*div k = ...........+.......+....+....+Fk
rho*u*div e = ...........+.......+....+....+Fe

where Fk and Fe are the terms to add in governing equations of k and e..
how can i d o it..i am trying to use weak contribution but not succesful..

Fk and Fe are similar to Fr an dFz in CC prsent in model library
Fr and Fz are added in navier stokes ..
i need Fk and Fe to add in k and e equations..

can you please suggest how to do that

i know that i am solving a highly nonliner problem and my system is crying to get the solution..

please suggest the better solver steps to get the solution faster with desired accuracy

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Ahmed Eissa

Ahmed Eissa

July 22, 2012 9:07pm UTC in response to Pavan Kumar Penumakala

Re: Non isothermal flow

Dear Pavan
I know it might be off the topic, but what is the CC that you refer to? I need to add a stress tensor to NS but unsuccessful .
Awaiting for your reply
Thanks in advance
Ahmed

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse

Pavan Kumar Penumakala

Pavan Kumar Penumakala

July 23, 2012 9:55am UTC in response to Ahmed Eissa

Re: Non isothermal flow

Dear ahmed
CC mean the continuous casting example given in the model library...
heat trasnfer>processand manufacturing>continuous casting

Reply  |  Reply with Quote  |  Send private message  |  Report Abuse


Rules and guidelines