[RJE]

i'll try that thanks

[RJE]

i tried both , the first when preload is defined with coordinate (5mm) w the other case using preload option (-877N)
height of the fluid is 30mm
CG of the first case is at 25mm
CG of the second is 30mm
i defined a constraint also that can't move more than 35mm
results are differents

[vinerm]

If the fluid height, whatever that means, is 30 mm, how could it reach 35 mm. Or do you mean that the total length of cylinder is 30 mm and then it can expand to 35 mm? Do note that it is not the length but actual values of the coordinates that matter. If the axis of cylinder is parallel to y-axis, then you have to report y-coordinate for the piston. That is its current position. If you provide same value for C.G., then spring is assumed to be in normal state.

[RJE]

yes i mean that the total length of cylinder is 30 mm and then it can expand to 35 mm


what do u mean by " If you provide same value for C.G., then spring is assumed to be in normal state."
i'll find the same result ?

[vinerm]

Check the position of the piston at time t=0. If you provide those coordinates as coordinates in C.G. then Fluent assumes that the spring is relaxed, i.e., neither under tension nor under compression.

[RJE]

yes i checked that and it's relaxed
i have another question, i have made 2 models (see photos)
i have made the same tips in both but the simple model works and the ather get me out with the error "negative volum cell"
ps: i tried to change size of the mesh , time step but i get same problem

[vinerm]

Ensure that six-dof setup is as in your simpler case, i.e., only one rigid-body object with layering. Usually, layering does not cause any negative volume until and unless there is an interface in the mesh or the ideal height is set wrong or 0. So, check for these.

[RJE]

the height is what i draw in red right ?
here there's no interfaces and i have only one rigid body which is the piston

[vinerm]

Then it should not cause negative volume. Which method did you choose for meshing? Plot the mesh on the surface passing through the center of the chamber and see if the mesh is proper sweep mesh. If mesh is alright, then try with less stiff spring or very small time-step.

[RJE]

layering method , the mesh is hexaedral using multizone , size =2mm
time step is already 5e-4 s

[vinerm]

Multizone method sometimes creates tetrahedral core. That's why you should check the inside of the mesh as well to ensure that whole of the mesh is hexahedral.

[RJE]

i thought that i enable tetra mesh by this (photo)?
how could i check if there's any ?

[vinerm]

Yes, as per that option, you have not given the Meshing tool a freedom to put tet mesh. However, the best option is to create a midsurface in Fluent or in Meshing and then check the mesh. Because, if the mesh is pure hex, then there should not be a negative volume with only one dynamic mesh object and properly set height, provided time-step is not very large.

[RJE]

i have made midsurface like what u said to my model (fig .cc) which was at the beginnig with multizone method , when i turn it to a midsurface , the method change to automatic method and then i get quad and tri meshing in that surface (fig .en)

[vinerm]

In Meshing, it is called section; it is used only for display and cannot have any effect on the method chosen. If method is multizone, it should stay as multizone. Look for a symbol with a square and a vertical line passing through its center. Use that to look at the mesh at the center section.

In fact, to look at the mesh, I did not expect you to use Meshing. You could create a mid-surface in Fluent using Iso-surface and then look at the mesh right there. No need to come back to Meshing.

[RJE]

ah sorry i didn't understand well at the beginning, as u can see it's all hexa meshing, so that's not the problem

[vinerm]

Yes, the mesh look good. It is very strange then that the one-DOF model gives issue at all. Check by displaying only the boundary that you are assigning as piston. Ensure that it is only one face and not multiple faces are assigned to the same name. So, if the name of the boundary that you are assigning as one-DOF object is piston, then display only piston. It should just be a flat circle. If it shows more boundaries, then separate the boundary using Mesh > Separate.

If the boundary is already one flat circle, then it should work. Just try without any spring and give it some motion but slowly increasing the pressure inside the chamber.

[RJE]

it was only one face
and , even without spring i've got the same problem
can i talk to u in PM?

[vinerm]

Yes, you can but it's weird that you are getting a problem without spring as well.

[RJE]

this is th error and i attached photos for

Reversed flow on 142 faces (100.0% area) of pressure-inlet 7.
Stabilizing pressure coupled to enhance linear solver robustness.
23 1.5398e+02 4.4420e-01 3.3843e-02 3.3821e-02 0:00:13 29
24 1.7483e+02 2.8342e-02 1.0309e-02 1.0063e-02 0:00:10 28
25 2.8234e+02 6.5862e-02 3.1072e-02 3.0710e-02 0:00:08 27
26 3.6826e+02 3.4298e-02 8.1878e-03 8.2251e-03 0:00:11 26
27 3.4823e+02 2.3900e-02 7.2094e-03 7.5912e-03 0:00:09 25
28 2.9650e+02 3.0177e-02 8.6201e-03 9.1219e-03 0:00:07 24
29 3.5638e+02 3.6948e-02 1.3693e-02 1.3471e-02 0:00:05 23
30 3.5370e+02 2.6336e-02 7.1233e-03 7.0900e-03 0:00:08 22
31 3.4965e+02 3.6287e-02 1.1690e-02 1.2089e-02 0:00:06 21
32 3.5708e+02 2.5979e-02 7.0180e-03 7.1939e-03 0:00:05 20

iter continuity x-velocity y-velocity z-velocity time/iter
33 4.4590e+02 3.9006e-02 1.2661e-02 1.4545e-02 0:00:04 19
34 4.4161e+02 2.6285e-02 8.8537e-03 8.2315e-03 0:00:06 18
35 4.1913e+02 3.0431e-02 1.1330e-02 9.9433e-03 0:00:05 17
36 4.2571e+02 2.5494e-02 8.2326e-03 8.1606e-03 0:00:04 16
37 5.1447e+02 3.5696e-02 1.3666e-02 1.5358e-02 0:00:06 15
38 5.4650e+02 2.3416e-02 8.1957e-03 7.3638e-03 0:00:04 14
39 5.4999e+02 2.8488e-02 1.2434e-02 1.0686e-02 0:00:03 13
40 5.0307e+02 1.9782e-02 6.9956e-03 7.3232e-03 0:00:02 12
41 5.5054e+02 3.3832e-02 1.1992e-02 1.3449e-02 0:00:04 11
42 5.3466e+02 1.9598e-02 7.7441e-03 7.0434e-03 0:00:03 10
43 5.4959e+02 2.8097e-02 1.0991e-02 1.0397e-02 0:00:02 9

iter continuity x-velocity y-velocity z-velocity time/iter
44 5.2286e+02 1.8236e-02 7.1502e-03 7.7405e-03 0:00:01 8
45 5.3241e+02 2.8210e-02 1.1326e-02 1.0963e-02 0:00:02 7
46 5.6324e+02 1.8853e-02 7.7905e-03 7.8986e-03 0:00:02 6
47 5.7266e+02 2.8713e-02 1.0718e-02 1.1446e-02 0:00:01 5
48 5.5391e+02 1.7443e-02 7.8248e-03 8.3175e-03 0:00:01 4
49 6.0637e+02 2.8408e-02 1.3360e-02 1.1162e-02 0:00:01 3
50 6.1500e+02 1.9230e-02 7.8138e-03 7.4366e-03 0:00:01 2
51 7.1730e+02 3.3720e-02 1.4138e-02 1.5768e-02 0:00:00 1
52 7.2535e+02 1.8543e-02 8.7249e-03 8.7646e-03 0:00:00 0
(update-animation-object "animation-1")()
Flow time = 0.001s, time step = 2
48 more time steps

Updating solution at time level N...
done.

Updating mesh at time level N... done.

WARNING: 1703 cells with non-positive volume detected.
Error at host: Update-Dynamic-Mesh failed. Negative cell volume detected.

Error at Node 0: Update-Dynamic-Mesh failed. Negative cell volume detected.

===============Message from the Cortex Process================================

Compute processes interrupted. Processing can be resumed.

================================================== ============================

Error: Update-Dynamic-Mesh failed. Negative cell volume detected.
Error Object: #f

fig e5 is the simulation at t=0s