Hello there,

I'm trying to solve a natural convection problem inside a metal enclosure, where air is contained as well as heat sources corresponding to the primary and secondary winding and the magnetic core in a dry type power transformer.

On the outside of the metal casing I declare natural convection with a relevant heat transfer coefficent h and a free stream temperature.

I consider the fluid as ideal gaz for the density computation. As for the rest I declare physical properties at average working temperature 60 deg. C. The measured temperature distribution oscillates between 29 and 99 deg C. So I initialize the velocity at zero, the temperature at 65 deg. C and the gauge pressure at +8000 inside the metal enclosure.( This is a guessed value). I declare a working pressure of some kPa above ambient pressure and a working temperature of 65 deg. C.

This kind of problems are normally solved by considering a temperature gradient, between two regions of interest. In my case I want to deduce tempeature distribution in solids by the action of fluid circualtion by body forces. For pressure I use Presto, SimpleC and second degree upwind for the rest. My underelaxation coefficients are all set at 0.8, is the only way I can get the residual for continuity to move.

I do not get proper solution, there is too much oscillation in the continuity residual and the pressure computation does not move at all, any suggestions?

Thanks in advance

Carlos Ortiz Universite du Quebec a Trois Rivieres Canada

I think for Pressure you need to use Body forced weighed option instead of Presto. May be this helps.

Sridhar.

Hi Sridhar,

Body Force weighted helped in deed, I also used power law for energy and momentum. There is a big change in the computation but I still do not get convergence.

Thanks a lot for your message.

Carlos