CFD Online Logo CFD Online URL
www.cfd-online.com
[Sponsors]
Home > Forums > Software User Forums > OpenFOAM > OpenFOAM Running, Solving & CFD

Planing boat simulation with InterDyMFoam and force ramp for make it stable

Register Blogs Members List Search Today's Posts Mark Forums Read

Like Tree5Likes

Reply
 
LinkBack Thread Tools Search this Thread Display Modes
Old   March 10, 2018, 08:50
Default Planing boat simulation with InterDyMFoam and force ramp for make it stable
  #1
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Hi guys!
I need your help because I知 trying to simulate a planing motorboat (using DTC Hull tutorial as usual) but I知 addressing the same problem some of you already faced: interDyMFoam becomes instable and terminates.
I have done many trials attempting to follow the hints I found on the forum but I think I have done a passage not in the correct way. I知 referring particularly to the thread:

interDyMFoam - planing hulls at high speed

Specifically in the dynamicMeshDict (sixDoF motion solver) file I have simply copied the rows relative to the 吐orce ramp from the dynamicMeshDict (RigidBody motion solver) file suggested by Jan @JNSN.
So I would like to ask you which is the correct way to insert the force ramp in the dynamicMeshDict (motion solver sixDofRigidBodyMotion)
Thanks in advance for the help!!
Best regards,
Emanuele

PS: Some data of my simulation
-motorboat: 7 meters runabout, beam 2.5 meters, draft 0.43 meters, displacement 2700 kg, speed 35 knots
-mass and inertia moments of half hull
-acceleration relax 0.4
-translational damper 85960
-rotational damper 115860 (I increased them by an order of magnitude wrt DTCHull values aiming at increasing solver stability but it doesn稚 work)
-Values I set for turbulence parameters (hope they are reasonable!):
k 0.01215
omega 18
nut 6.75 x 10^(-7)
-Mesh: 1.8 milion cells, built similar to the DTChull mesh through topoSet and refineMesh
-OpenFOAM v17.12
Attached Files
File Type: zip dynamicMeshDict.zip (761 Bytes, 56 views)
File Type: zip log_interdymfoam.zip (41.7 KB, 30 views)
EmaManu is offline   Reply With Quote

Old   March 12, 2018, 05:59
Default
  #2
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Dear EmaManu,

I thought the "ramp coefficients" were only available with openfoam-dev and you are using OF1712+. Are you sure they are working ? EDIT : Yes it seems they are available with of1712+ (getting mad with all these OpenFOAM versions and different functionalities ...)

Can you show your checkMesh, shemes and BC ?

Thanks
EmaManu likes this.
Fauster is offline   Reply With Quote

Old   March 12, 2018, 16:26
Default
  #3
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster,

thank you for answering!! Of course: I have just uploaded on dropbox my entire case folder... the link is the following:

https://www.dropbox.com/s/g9ukleb75z...Barca.zip?dl=0
EmaManu is offline   Reply With Quote

Old   March 13, 2018, 07:35
Default
  #4
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Quote:
Originally Posted by EmaManu View Post
Dear Fauster,

thank you for answering!! Of course: I have just uploaded on dropbox my entire case folder... the link is the following:

https://www.dropbox.com/s/g9ukleb75z...Barca.zip?dl=0
You will find here your case modified. The problem is due to excessive mesh deformation. I change your mesh (coarsened and add layer), schemes and dynamicMeshDict. I run just 1 sec of your case on 10 cores.

I am not sure the quadraticRamp is working. If you have still some problem you will need to damp the strong acceleration at the beginning. Play also with inner and outer distance in your dynamicMeshDict.

https://www.dropbox.com/s/oj35smm6xm...arca2.zip?dl=0

cheers
EmaManu likes this.
Fauster is offline   Reply With Quote

Old   March 13, 2018, 18:53
Default
  #5
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster
thank you so much for the help!!! I have downloaded the file and I'm going to launch the simulation! I will tell you how it will go
However thanks again!
EmaManu is offline   Reply With Quote

Old   March 17, 2018, 06:05
Default
  #6
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster,
the simulation with your settings works well!
I uploaded the case folder with 8 seconds of simulation so that maybe can be helpful for other people, the following is the link:

https://www.dropbox.com/sh/s0cwrcpbt...ThLOFdE2a?dl=0

Have a great day and weekend!
Best regards
Emanuele
Fauster likes this.
EmaManu is offline   Reply With Quote

Old   March 29, 2018, 04:08
Default
  #7
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Dear EmaManu,

Have you been able to stabilized your simulation ? I mean do you get stabilized trim and sinkage time evolution ?

I am doing similar simulation and I am not able to get a stabilized solution. My pressure coefficient oscillates randomly wich affect the hull position and. Or maybe it's the opposite !

F
Fauster is offline   Reply With Quote

Old   March 29, 2018, 15:20
Default
  #8
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster,
for the time being I can't say because until now I have done several short simulations but now I am doing a new calculation which I hope wuold simulate 20-25 seconds by next monday/tuesday and I will tell you how it will go! Hoping the boat reaches a stable condition
Best regards
Emanuele
EmaManu is offline   Reply With Quote

Old   April 5, 2018, 20:03
Default
  #9
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster,
sorry for the delay!
I'm still searching the useful parameter for my simulation because with 5x10^6 for both the translational and rotational dampers the bow rises too fastly and thus trim increases too mach while the aft remains immersed... so at 12-13 sec the simulations fail
I'm looking for the useful values to fix this problem: now I'm trying with 3x10^6 for translational damper and 7x10^6 for rotational one.
And you? How are going your simulations?
Best regards,
Emanuele
EmaManu is offline   Reply With Quote

Old   April 9, 2018, 12:16
Default
  #10
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Quote:
Originally Posted by EmaManu View Post
Dear Fauster,
sorry for the delay!
I'm still searching the useful parameter for my simulation because with 5x10^6 for both the translational and rotational dampers the bow rises too fastly and thus trim increases too mach while the aft remains immersed... so at 12-13 sec the simulations fail
I'm looking for the useful values to fix this problem: now I'm trying with 3x10^6 for translational damper and 7x10^6 for rotational one.
And you? How are going your simulations?
Best regards,
Emanuele
My simulations are not fine. I am not able to stabilize trim and sinkage and by consequence pressure coefficient is not stabilized.

Some keys points to help you:
  • I realized since my last message about the quadraticRamp that the key word is quadraticRamp :
Quote:
ramp
{
type quadraticRamp;
start 0;
duration 10;
}
  • After investigating it appears that the ramp is not working !! The ramp works only with rigidBodyMotion !! Change your dynamicMeshDict and rerun your case.
  • If it's still not ok try to use interFoam with LTS option (in fvschemes file) to obtain a stabilized flow.
  • Then switch to interDyMFoam with a quadraticRamp (maybe 5 sec of duration) without damping and with accelerationRelaxation of 0.4 (or maybe 0.3).
Can I ask you where the hull comes from ? The stl triangulation is nice. This improves the meshing step.

I hope this will help you.
F
EmaManu likes this.

Last edited by Fauster; April 11, 2018 at 04:54.
Fauster is offline   Reply With Quote

Old   April 16, 2018, 09:02
Default
  #11
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster,
first of all thanks for your suggestions!! ...I haven't put them in practice yet because I was attempting to find the suitable values for damping force and moment, and these trials took some time: now I have managed to find two values (7e6 for the rotational damper and 9e5 for the translational one) that work (quite) well beacause now the boat rises and rotates more normally, but I think sinkage and trim will be characterised by an oscillatory trend which doesn't end like it has happened to your simulations. I'm at 39 seconds of motion and the bow of the boat is coming down.
I have the curiosity to see how it will go, then I will applly your hints. About this: once I would have completed the simulation with InterFoam how should I practically use its results for the next simulation with interdymFoam? That is which data should I take from the output of the InterFoam simulation to put as input of the interDymFoam one?
Thanks in advance, again!!
EmaManu
PS: I uploaded the graphs of trim, sinkage and damping forces and moments If you want to see and compare to your results

And about the STL of course I can tell you! I realized it with Vacanti at first and then modified and exported with Rhino5
Attached Images
File Type: png Trim.PNG (17.8 KB, 77 views)
File Type: png Sinkage.PNG (18.7 KB, 63 views)
File Type: png DampingForce.PNG (17.9 KB, 58 views)
File Type: png DampingMoment.PNG (22.1 KB, 45 views)
EmaManu is offline   Reply With Quote

Old   April 16, 2018, 13:07
Default
  #12
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Quote:
Originally Posted by EmaManu View Post
About this: once I would have completed the simulation with InterFoam how should I practically use its results for the next simulation with interdymFoam? That is which data should I take from the output of the InterFoam simulation to put as input of the interDymFoam one?
Before answering to your whole message I can give you how I proceed for doing that (I guess there are plenty of possibilities).
  • Once interFoam (LTS) is finished you have to reconstruct the last results (reconstructPar -latestTime).
  • Then rename the created folder to 0lts.orig (or any other name).
  • Create a new clean case suitable for interDyMFoam (good fvsolution, controlDict, dynamicMesh etc...)
  • Use the 0lts.orig as your 0 folder for the interDyMFoam run.
  • Decompose the case and run interDymFoam in parallel



I hope it will help you !
Fauster is offline   Reply With Quote

Old   April 17, 2018, 04:08
Default
  #13
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Fauster of course it helps!! Thanks a lot!
EmaManu is offline   Reply With Quote

Old   April 17, 2018, 05:41
Default
  #14
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Quote:
Originally Posted by EmaManu View Post
Dear Fauster,
first of all thanks for your suggestions!! ...I haven't put them in practice yet because I was attempting to find the suitable values for damping force and moment, and these trials took some time: now I have managed to find two values (7e6 for the rotational damper and 9e5 for the translational one) that work (quite) well beacause now the boat rises and rotates more normally, but I think sinkage and trim will be characterised by an oscillatory trend which doesn't end like it has happened to your simulations. I'm at 39 seconds of motion and the bow of the boat is coming down.
I have the curiosity to see how it will go, then I will applly your hints. About this: once I would have completed the simulation with InterFoam how should I practically use its results for the next simulation with interdymFoam? That is which data should I take from the output of the InterFoam simulation to put as input of the interDymFoam one?
Thanks in advance, again!!
EmaManu
PS: I uploaded the graphs of trim, sinkage and damping forces and moments If you want to see and compare to your results

And about the STL of course I can tell you! I realized it with Vacanti at first and then modified and exported with Rhino5
The figures of trim and sinkage evolution show that your simulation is not stabilized. I guess your damping coefficients are too high and the hull can't reach a stabilized trim and sinkage position. The inertia of the movement are way too big from my point of view. Your hull is small with reasonable inertia coefficient. You should reach a stabilized position in 10 sec of 20 sec maybe.

If you still want to use interDyMFoam at the beginning I would lower the value of damping coefficients and increase the ramp time (15 sec or 20 sec for example) to see if it helps.

Try also with a lower speed and then increase to 20 m/s. I manage to stabilized the hull with a speed of 8 m/s.

Can I ask you how you monitor trim and sinkage evolution ? Can you share your program ?

Paul
Fauster is offline   Reply With Quote

Old   April 17, 2018, 19:08
Smile
  #15
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Paul
I agree with you about the eccessive slowness and inertia of my actual simulation, next days I will apply both your stategies: at first interdymfoam with dinamicmeshdict.rigidbodymotion so that the 15-20 sec ramp can work, and dampers lowered: I am thinking to mantain the same ratio but with lower vaues, for istance 6e5 the linear and 4.67e6 the rotational one
...then if it doesn't work i will apply your second way: interfoam LTS and then interdymfoam without dampers
About the graphs of course i can share my scripts! Ihave uploaded the ones for the sinkage and trim (cosine(trim))
But I don't think that is very smart my path: they are java scripts that extract data from the file log.interdymfoam wrtiting in a new text file two columns: one for the time and one for the quantity of interest (you can download eclipse java and run them)
After that I import these files in excel and do the graphs
..it's really not the smartest way I think...
EmaManu is offline   Reply With Quote

Old   April 20, 2018, 11:26
Default
  #16
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
Quote:
Originally Posted by EmaManu View Post
Dear Paul
I agree with you about the eccessive slowness and inertia of my actual simulation, next days I will apply both your stategies: at first interdymfoam with dinamicmeshdict.rigidbodymotion so that the 15-20 sec ramp can work, and dampers lowered: I am thinking to mantain the same ratio but with lower vaues, for istance 6e5 the linear and 4.67e6 the rotational one
...then if it doesn't work i will apply your second way: interfoam LTS and then interdymfoam without dampers
About the graphs of course i can share my scripts! Ihave uploaded the ones for the sinkage and trim (cosine(trim))
But I don't think that is very smart my path: they are java scripts that extract data from the file log.interdymfoam wrtiting in a new text file two columns: one for the time and one for the quantity of interest (you can download eclipse java and run them)
After that I import these files in excel and do the graphs
..it's really not the smartest way I think...
Awesome ! Thanks a lot . Let me know if your simulation run fine.
Fauster is offline   Reply With Quote

Old   April 24, 2018, 09:05
Default
  #17
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear Paul I manage to get a stabilized simulation with the 0.LTS initial conditions, 5 sec quadratic ramp and zero dampers and 15 m/s speed of the boat
Thanks again, have a nice day!
Emanuele
EmaManu is offline   Reply With Quote

Old   May 4, 2018, 22:36
Default
  #18
New Member
 
Scott Hogan
Join Date: Mar 2010
Location: Washington State, USA
Posts: 7
Rep Power: 12
shogan50 is on a distinguished road
EmaManu,
Would you mind sharing your case files where you initialize with LTS interfoam? I have a similar problem with a different hull and only 6.5 m/s.
shogan50 is offline   Reply With Quote

Old   May 5, 2018, 09:31
Default
  #19
Member
 
Paul Palladium
Join Date: Jan 2016
Posts: 83
Rep Power: 6
Fauster is on a distinguished road
See the tutorial dtcHull in interfoam. You Will find the setup. It s realy basic. Set localEuler as scheme for d/dt. Copy also fvsheme and fvsolution
EmaManu likes this.
Fauster is offline   Reply With Quote

Old   May 10, 2018, 13:34
Default
  #20
New Member
 
Emanuele Ponzano
Join Date: Feb 2018
Posts: 15
Rep Power: 4
EmaManu is on a distinguished road
Dear shogan50,

here the files, related to the first simulation with interFoam, with the boat at 24 knots

Best regards,

Emanuele
Attached Files
File Type: zip Boat_InterFoam.zip (18.8 KB, 84 views)
EmaManu is offline   Reply With Quote

Reply

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are On
Refbacks are On



All times are GMT -4. The time now is 06:11.