Looking for help running a simulation

[selhamda]

Intro:
I'm a student in a french scientific prep school. I've taken upon myself as a TIPE project ( personal investment project) to study and helicopter's flapping hinges on rigid blades and how they help solve the dissymmetry of lift phenomenon.

I've come to the latest stage of my study. Did the meshing with Salome, and wanted to run the simulation but failed to do so as code-saturne sends me a "Process not running" message.

Mesh:
I do apologize for the complete mess that geometry is, but this is what i tried to do:
1st Blade Tip creation:
- Import my points to define the airfoil profile.
- Translate said points and create a face at given position.
- Draw lines between each point of intrados and extrados to make centers of mass for each line.
- Using these centers of mass create vectors normal to the face then rotate them by 90 degrees to make them normal to the line.
- use said normal vectors to create a third point by rotating points of the extrados by 90 degrees.
- Draw circles with set tri-points and make faces out of these.
- There are a few i added manually to define the spherical shape at the top.
- Extrusion across path which is a polyline going through all centers of mass
Making the fluid's domain:
- Cut the geometry using a box to leave only half.
- Extrude generated planar face to have the main blade.
- Create a box which will be the fluid's domain.
- Cut both Wing Tip and main Wing out of the box.

At this stage i first thought i was ready for the simulation. But I realized that what i wanted is a rotating blade not a still one so i decided to make the wind's speed vary instead of the blade's. Then again i did not know how to make fluid speed a function of space in code saturne. Thus i decided to add a part on which i'll have multiple inlets(faces) each with a speed higher than the one before.

Making the extra part:
-Create a simple box.
-create a smaller box with desired length of discretization.
- Use multi translation and cut function to have the crank shape which guarantees our multiple inlets.

Making the mesh:
- After creating a partition of the two solids and exploding it did the mesh for each of them and then compounded the meshes.
- Then created the groups with walls inlets and outlets.

Code-Saturne part:
- Thermophysical Models : Unsteady flow, k-epsilon turbulence model, the rest at default.
- Physical properties at default except i chose for there to be gravity.
- Volume conditions at defaults.
- Definition of boundary conditions set as in the names in the mesh.
- Boundary conditions : Inlets fluid speed norm set as [ u_norm = 55.6+value*sin(47.1*t) ] , 55.6 being the helicopter's forward speed, and value*sin(47.1*t) being the blade's rotation speed, with "value" varying given the inlet.
- Then i set results to be MED format.

That's about it. Thanks for giving this some of your time. I'd very much appreciate any kind of help in this endeavor.

Cordially,
Salim ELHAMDANI

Mesh 3D Reverse F.part1.rar

1st part of the Mesh on Salome

(10 MiB) Downloaded 487 times

Mesh 3D Reverse F.part2.rar

2nd part of the Mesh

(4.51 MiB) Downloaded 507 times

[Yvan Fournier]

Hello,

You posted in the wrong forum, so I moved your post...

Could you post the whole error message(s), as per the forum recommendations, in addition to the input data (which will be useful as a "last resort")?

Until more people help me answer on this forum, I prefer to avoid going all the way towards running cases when error logs are enough to guess at the problem, as it takes less time and allows to focus on more difficult cases... (you might have an installation issue, but only the details in the forum recommendation will help me, or any other volunteers, determine that more easily...)

Regards,

Yvan

[selhamda]

Thank you Yvan for the pointers.

Here is a screenshot of the error, (Windows 96 skin):


In attachment is the full case.

Best regards,

Salim

Case Alpha.part1.rar

(8 MiB) Downloaded 528 times

Case Alpha.part2.rar

(8 MiB) Downloaded 542 times

Case Alpha.part3.rar

(7.12 MiB) Downloaded 534 times

[Yvan Fournier]

Hello,

Could you please simply post the log files mentioned, not the whole directories ?

Regards,

Yvan

[selhamda]

Greetings,

First i'd like to thank you for your help. Then i'd like to apologize because i couldn't find the compile.log, listing and preprocessor.log files.
However, I think it is natural because my calculation does not start running when i select "start calculation" in the prepare batch calculation menu (CPU usage load does not change).


Thanks again for your help Mr. Yvan.

Cordially,

Salim

[Yvan Fournier]

Hello,

Do you have a blank screen ? In this case, try running directly the run_solver script in the case's SCRIPTS subdirectory directory (after you have "saved" the configuration with the GUI).

Regards,

Yvan

[selhamda]

Hello,

Thanks for showing me that helped me a lot to get thz simulation running. However i still get an error: "SIGSEGV signal (forbidden memory area access) intercepted!"

the error, listing and preprocessor.log files are in attachment. Could not find the compile.log.

Thanks again for the help.

Salim

(lisiting,preprocessor.log,error).rar

(9.43 KiB) Downloaded 527 times

parametres.xml

(14.8 KiB) Downloaded 567 times

[Luciano Garelli]

Hello Salim,

The error that you get it is because you created a mesh which is build from 2 independent domains (Mesh_1 and Mesh_2), wherewith you get a duplicated face at the union and you will get two surface meshes. To ensure the continuity of the fluid domain you have two options, one is merge the coincident nodes and elements when you built the compound mesh at Salome, ir order to use this option you have to be sure that both meshes are identical, so you can use Projection 1D-2D to project one mesh to the other face. The second option is use the "Faces joining" option at CS, (pick the plus button and leave the default values, I have used this option to test your case). Also, as a first run don't use turbulence model and remove the gravity.

Additionally, reading the problem that you want to solve I think that the temporal variation of the inlet velocity it is not representative of the rotating of a helicopter blade, check this link http://www.aerospaceweb.org/design/heli ... nges.shtml.

I attach a listing with 8 time steps.

Regards,

Luciano

[selhamda]

Hello,

Thank you very much Mr Luciano for your help. Thanks to you, I can finally run the simulation.
I'd like to thank you for pointing out the problem with the temporal variation of the inlet velocity. I'm currently struggling to figure that out.
Anyways, thanks again for the help.


Cordially,

Salim

[Luciano Garelli]

Hello Salim,

You are welcome. I'm glad you were able to run your simulation.

Regards

Luciano