Discussion Forum

Hi everyone,

The case I'm working on is typical conjugate heat transfer. Two tubes are arranged in concentric configuration. The inner tube is heated on the inner surface and cooled by water flowing between the tubes. There is no coating on the surfaces of inner tube that would act as thermal resistance. My purpose is to find out the highest temperature in the cooling water region.

After running some simple tests (Reynolds number is >10000), I have found that when the k-episilon turbulence model is used, the temperature is not continuous at the solid/fluid interface. However, there is no such a problem when I change to laminar flow model.

So, here are my questions.
1. When I try to estimate the highest temperature in the cooling water region from turbulence model, it turns out the results given by the surface maximum and volume maximum functions are not the same (I made sure only the boundaries or the domain of cooling water region are selected). Since my concern is the highest temperature in fluid, which value should I trust?

2. Is the result of discontinuous temperature from k-episilon model caused by the adoption of wall function because the viscous sublayer is not considered? If this is true, can I solve this problem by using Low Reynolds number turbulence model instead, since it solve the Navier-Stokes equations all the way down to the wall?

Any comment would be highly appreciated and thanks for anyone's kind help in advance.