Learning Forum

[manelaero2020]

Hi all,

I am performing an unsteady simulation for conjugate heat transfer. The global time step is set by default as 5s, the time step is 0.1. I have obtained a result, but I don't know the physical time step that corresponds to the outcome. 

Can you tell me please how can i estimate the physical time step in that case? 

Thank you a lot for your help 

Kind regards 

[Rob]

Please can you show what you mean by timestep of 0.1 and global time of 5s?  In Fluent we typically set a flow time step to pick up the flow features and the model then runs for some timesteps. The "time" of the solution is then "time step x number of time steps".  If you look in preferences for Titles and turn those on the flow time will be shown when you plot graphics. 

There are a couple of examples in Help and I think a lecture on transients in the Learning section on here. 

[manelaero2020]

Hi @Rob

Effectively the delta t = 0.1 s is the time step, which can be estimated in the case of heat conduct as L^2/ nu with L being the width of the plate (the solid) and nu being the heat transfer diffusivity (lambda / (rho*Cp) ) . 

I guess the physical time is equivalent to the transient state time before the solution becomes stationary.

The exponential heat transfer curve can estimate it 

The delta t = 5 s is the global time step, and the number of iterations is 5/ 0.1 

Thank you a lot 

 

 

 

 

[Rob]

The fluid dt is (approximately) how long it takes the flow to pass through a cell. Typically, we'd aim for a tenth of that value to get the model started and work from there. You may find the solid time step can be bigger depending on the conductivity. 

The total solution time (NOT duration) is then number of time steps multiplied by time step size. If the system reaches an equilibrium and you don't care how long it take to get there then steady state may be a better option. 

[manelaero2020]

@ROB 

Thanks