Why negative temperature

  • Last Post 11 February 2019
  • Topic Is Solved
lakshmi posted this 23 January 2019


Can someone help me understand why there is a negative temperature developed in solid-fluid heat transfer simulation (steady state) ?The boundary where it is showing negative temperature is given as a solid fluid contact (interface) and the surrounding temperature is given as 23 °C.


Order By: Standard | Newest | Votes
rwoolhou posted this 23 January 2019


Check the mesh quality and boundary conditions. The former if poor can give all sorts of interesting results.   The latter can force the solution: if you have a fixed heat flux and little flow you can get very high or low temperatures. 

What does the convergence plot look like?

  • Liked by
  • lakshmi
rwoolhou posted this 23 January 2019

When you say negative, do you mean it's lower than you're expecting or trying to be below 0 Kelvin? Please can you post the image in the thread as I'm not allowed to open attachments. 

lakshmi posted this 23 January 2019


By negative, I mean its lower than expecting value.

The minimum temperature in the system must be 23°C. However, at a certain location, its showing -77°C. There is no way for the temperature to go this low. In fact the temperature of this place should be greater than 23°C.



I tried to change the below settings hoping to overcome the problem, but did not work.

 1. changed initial temperatures to 23°C

2. Tried to adjust the tolerance

3. Tried to change convergence value



kkanade posted this 24 January 2019

there must be some bad elements. please check the mesh quality and improve your mesh. 

the min. orthogonal quality should be more than 0.1.

lakshmi posted this 05 February 2019



Looks like again the problem is poping up. The minimum othogonal quality is > 0.1 and mesh resolution is high and the initial temperature is assigned 23°C. But the solution result is showing a negative value.

Note: Its a solid-fluid simulation and the negative temperature is developing in the fluid region.

Are you aware of any other possibilities?

Thanks in Advance

rwoolhou posted this 07 February 2019

What are your material properties? Ie is gas expanding quickly in the system?  Also please plot some contours to show the problem area. 

lakshmi posted this 11 February 2019

Hi rwoolhou,

A hollow cylinder made of pzt is the hot body (low thermal conductivity 1.1 W/mK, heated upto 150°C). It is attached to a stainless steel (higher thermal conductivity 25W/mK) on both ends. Air passes through pzt to stainless steel. It is the air inside the stainless steel that shows negative temperature. Air pressure is 6 bar. A thermal simulation on the same scenario shows correct result. But when fluid is introduced (air), then things arent working as expected.

Below is the convergence plot



Thanks in advance


rwoolhou posted this 11 February 2019

Please can you put some more images up: pressure, temperature & velocity? If it's a straight pipe the only causes are mesh quality & boundary conditions: if it's more complex there could be many other reasons.