why it doesn't abide the connuity law

February 28, 2001, 07:24

When I sumulate air flow around rectangle objects, I set the INLET attribute "velocity in x direction=5m/s, and in y or z direction v=0m/s". Using Ke-epsilon model, and get a converged result. But when I see the result with VR Viewer, I found the velocity near the INLET all greater than 5m/s. Then it doesn't abide the connuity law.I also find that the speed is large near the domain boundary and small in the middle in the place when x=0.[ in addition: that I set the INLET in the position of x=0]. What wrong with the result? Can anyone help me?

February 28, 2001, 15:19

Hi Yj

It is good that you are very observant and study your results carefully. It is very difficult to see exactly what is wrong with your simulation from the information you give. If you are sure that the simulation has converged, then it is likely that something has gone astray with your boundary and input conditions. It is always useful to check your input conditions against a case that works. From your description a case that could be similar to yours is library case V407 which is the same as VR tutorial no 7. My first suggestion is to check your input data and boundary cnditions against those of V407.

Good luck

Arne

March 1, 2001, 05:47

If the run is truly converged and your boundary conditions are correct then I suspect a density mismatch. Is is possible that the density used within the domain (RHO1) differs from that employed for the boundary condition specification? Is the flow is incompressible? If so, check the ratio of predicted velocity/specified inlet velocity to see if it can be explained by an inconsistency between the inlet and field densities. Check in the RESULT file the NETT SOURCE printout of R1 for your inlet patch, i.e this records the mass flow rate in kg/s for the INLET. Divide this quantity by the inlet velocity and area, and this will confirm the inlet density used for the inlet boundary condition. In the RESULT file, RHO1= will tell you the density used for internal cells. However, if the density is variable throughout the domain you will have to store DEN1 and look at those field values to see what is used internally.

March 3, 2001, 11:16

thank you for your help

February 28, 2001, 07:24	why it doesn't abide the connuity law	#1
YJ Guest Posts: n/a	When I sumulate air flow around rectangle objects, I set the INLET attribute "velocity in x direction=5m/s, and in y or z direction v=0m/s". Using Ke-epsilon model, and get a converged result. But when I see the result with VR Viewer, I found the velocity near the INLET all greater than 5m/s. Then it doesn't abide the connuity law.I also find that the speed is large near the domain boundary and small in the middle in the place when x=0.[ in addition: that I set the INLET in the position of x=0]. What wrong with the result? Can anyone help me?

February 28, 2001, 15:19	Re: why it doesn't abide the connuity law	#2
Arne Holdo Guest Posts: n/a	Hi Yj It is good that you are very observant and study your results carefully. It is very difficult to see exactly what is wrong with your simulation from the information you give. If you are sure that the simulation has converged, then it is likely that something has gone astray with your boundary and input conditions. It is always useful to check your input conditions against a case that works. From your description a case that could be similar to yours is library case V407 which is the same as VR tutorial no 7. My first suggestion is to check your input data and boundary cnditions against those of V407. Good luck Arne

March 1, 2001, 05:47	Re: why it doesn't abide the connuity law	#3
Mike Malin Guest Posts: n/a	If the run is truly converged and your boundary conditions are correct then I suspect a density mismatch. Is is possible that the density used within the domain (RHO1) differs from that employed for the boundary condition specification? Is the flow is incompressible? If so, check the ratio of predicted velocity/specified inlet velocity to see if it can be explained by an inconsistency between the inlet and field densities. Check in the RESULT file the NETT SOURCE printout of R1 for your inlet patch, i.e this records the mass flow rate in kg/s for the INLET. Divide this quantity by the inlet velocity and area, and this will confirm the inlet density used for the inlet boundary condition. In the RESULT file, RHO1= will tell you the density used for internal cells. However, if the density is variable throughout the domain you will have to store DEN1 and look at those field values to see what is used internally.

March 3, 2001, 11:16	Re: why it doesn't abide the connuity law	#4
xyj Guest Posts: n/a	thank you for your help