[hamidciv]

hi all
i modeled ogee spillway in ansys cfx, my result in low discharges in compared with experimental is almost good, but with increase of discharge my result is excellent, i read in article , that in cfd software bottom of model dont simulated in form completely flat induced of meshing , and contain ripples, that this ripples in high discharge reduced effect on flow, whether this subject is correct?
thanks a lot

[ghorrocks]

Can you post an image of what you mean?

[hamidciv]

hi dear glenn
my mean hydraulic parameter like velocity, depth in Q=1350 cubic metric/s compared to Q=600 qubic metric/s is better and close to experimental data, is really mesh can do change geometry model and do regged bottom surface and finally in high discharge , effect of this regged decrease on flow,my mean is different result obtained of cfx in two case, that with increase discharge, my result have been close to experimental data,and also i read one article in simulation using flow-3d taht it written bottom surface is not flat simulated in model and have regged, that in high discharge effect this regged decreased.
dear glenn i runned in two discharge(q=600,1350), in Q=1350 my result closed to experimental data, but i dont know why i have sudden jump in velocity in Q=1350 while in Q=600 my result varied in form Gradual as shown follow, i have experimental data only in 3 points that in two case is good agreement, in your order What is the reason this suuden increase in velocity? is it can dus of high discharge( Q=1350) in my model? Or it could be another reason?
thanks a lot in advance

[ghorrocks]

How are you extracting the data for the velocity plots from your simulation?

I can see the mesh you are using is very coarse, I am surprised you got results so close to experimental results with such a coarse mesh. You should consider refining your mesh, especially in the high velocity section.

[hamidciv]

dear glenn
i meshed model with 988117 node and 5188525 elements, i attached one image of mesh, I do not think my mesh be coarse, i refined mesh in areas i want extract results, as aeport experimentl data i extracted velocity in 0.6 of free surface with define polyline, i runned for 1 weeks with properties system contain 8gb ram, core i5 such as mesh, and all result i is good, only in Q=1350 I have sudden chande in velocity value that i dont know cause it?
thanks for spending time for me.

[ghorrocks]

I do not think my mesh be coarse,

I have heard that so many times on the forum and the problem ends up being the mesh.... What do you base your opinion on? Have you done a sensitivity check and actually checked your mesh is OK? Or does it just look kind of fine so it must be good?

And have you checked what your mesh resolution is in a cross section of the middle of your mesh? I suspect you will find it is much coarser in the middle than the wall.

Everything is still suggesting the jumps in the graph are caused by coarse mesh.

[hamidciv]

dear glenn
i done meshed for two case , that in second case my result have a very good agreement with experimental data, therefore i considered this mesh for runned,what your mean resolution is in a cross section of the middle of your mesh? how do this?
thanks for help me.

[ghorrocks]

In CFD-Post put a plane in the middle of the flow, and draw mesh lines on it.

[hamidciv]

hi dear glenn
i done your order,your order is correct unfortunately my mesh is coarse, but i dont know because my result have a good agreement with experimental, whether exist way in cfx-mesh like method your expressed in top , how i should understand my mesh is coarse in cfx mesh and before running i refined my mesh? which metric parameter can using for this?
very thanks for help me

[ghorrocks]

Do the tutorials on meshing which come with workbench. Also more are available on the ANSYS website customer area.

You should always know what the mesh size in the important areas of your simulation are. A wise CFD person will have done a mesh sensitivity check on the mesh size required for accurate simulations.