Enquiry on CFX FAQ
 

Dear all,

In : http://www.cfd-online.com/Wiki/Ansys_FAQ , i've an enquiry to make:

regarding the statement: You can get the residence time in CFX-Post by placing a streamline and looking at the "Time" variable on it. The maximum value of time is the residence time.

1) How in CFX-Post do i do this please? I've created a 'Streamline 3D', but i don't have the variable 'Time' to select.
2) Does 'residence time' only refers to the fluid domain, and it doesn't concern the solid domains please?

With the streamline object, colour the streamlines by time. That is how you get the time variable.

Residence time is only applicable to streamlines and they do not go through solid regions.. unless something is very wrong.

Dear Glenn,
Thanks for your reply on my enquiries.

i'm having many transient convergence issues while performing a fluid-thermal coupling analysis for my electrical machine model. Especially, oscillating residual plots in all my domains, from fluid to solids. my model consists of 10 parts. one of them is fluid, others are solids.

What advice would you have if my model size is?
Number of Nodes: 290115
Number of Elements: 792869

I know sensitivity checking is important, but am i able to narrow down which part in my model should i concentrate my sensitivity check on please? am i able to know which part i should refine its mesh or change geometry to allow better convergence?

# could you show the mesh cut setion
# residual fluctuation range /Fluid properties
# are you monitoring Temperature in all the solid domain, how about the fluctuation of that
# do you have any overlapping interface
# check the RMS and max courant number from the out file.

You can post process the results to decide which region in your geometry has the highest gradients and then refine these regions. The selection of tiemsteps is also important aspect of convergence. How have you decided it? Adaptive time-stepping may help here in quickly finding the time stepping. Without understanding the effect of physics, the number of elements in a mesh can not decide its suitability.

OJ