[tjushang]

Hey guys,

I was modeling a fluid zone inside a cooling plate. The fluid zone has an inlet and an outlet. I set an constant inlet flow rate of about 0.5 kg/s (corresponding to 0.5 L/s considering liquid parameter) with constant temperature of 280 K. The outlet is set as pressure outlet with 0 Pa.

The models of the fluid I chose are implicit unsteady、segregated flow and segregated fluid temperature. The initial temperature of the fluid zone is 298 K. The volume of the fluid zone is about 8 L.

What confused me is the calculation result: The outlet temperature remained nearly constant (298K) for about 40 s，until the cold fluid from inlet reached outlet (which can be observed from temperature scalar). However, as the flow rate is 0.5 L/s and the total volume is 8 L, it takes at most 16 s to replace all the initial fluid with cold fluid from inlet, in theory. That means outlet temperature has to decrease after 16 s, not 40 s.

Can someone enlighten me about this issue? Are there something unsuitable models chosen that can cause this problem?

[arjun]

Originally Posted by tjushang

Hey guys,

I was modeling a fluid zone inside a cooling plate. The fluid zone has an inlet and an outlet. I set an constant inlet flow rate of about 0.5 kg/s (corresponding to 0.5 L/s considering liquid parameter) with constant temperature of 280 K. The outlet is set as pressure outlet with 0 Pa.

The models of the fluid I chose are implicit unsteady、segregated flow and segregated fluid temperature. The initial temperature of the fluid zone is 298 K. The volume of the fluid zone is about 8 L.

What confused me is the calculation result: The outlet temperature remained nearly constant (298K) for about 40 s，until the cold fluid from inlet reached outlet (which can be observed from temperature scalar). However, as the flow rate is 0.5 L/s and the total volume is 8 L, it takes at most 16 s to replace all the initial fluid with cold fluid from inlet, in theory. That means outlet temperature has to decrease after 16 s, not 40 s.

Can someone enlighten me about this issue? Are there something unsuitable models chosen that can cause this problem?

most probably the energy equation did not converge enough during the time step. You can double the inner iterations and see if this time reduces from 40s. If this reduces then what i wrote is the reason.

[tjushang]

Originally Posted by arjun

most probably the energy equation did not converge enough during the time step. You can double the inner iterations and see if this time reduces from 40s. If this reduces then what i wrote is the reason.

Thank you arjun! I doubled the iteration steps and it was observed that temperature field spread faster than before. It seems that it was the lack of iterations or the overlong time step that caused this problem.

Allow me to ask more: how can we tell if the iteration step is enough by simply seeing residuals? My normalized residuals of energy decreases by only 1 order of magnitude and it almost takes a lot of total iteration steps to finish whole simulation I desired. How to balance the accuracy and time cost is really a important matter to me.