Simulation Ship Hull, CFX solver interrupts

dkmred · March 11, 2017, 07:00

Hey Guys,

I am trying to simulate a kayak ship hull. At the start I have simulated a simple cube in two phases. That worked well and I could clearly post process and visualize the wave pattern. With those settings I tried to get the ship hull running. But since months I did not get the simulation to run.

Most error messages are:

- negative volume elements (in spite of good aspect ratio / orthogonal quality and skewness)
- fatal overflow

I have tried modifing the mesh, I have tried a lot of different boundary conditions for the walls, changed the turbulence model and so on....I am really desperate and have no glue anymore.

I hope some of you did handle with anything similar before and can have a quick view over my ccl file if I used any trouble causing settings. Additionally I've added a picture of my meshing and the domain in cfx pre.

Mesh.jpg

settings.zip

domain.jpg

Thanks for your help in advance,

dkmred

ghorrocks · March 12, 2017, 05:27

You are using a rigid body solver for this, presumably for the motion of the kayak. This makes the simulation a zillion times harder and is not recommended. The recommended approach is to assume a water level/pitch and model that with a simple fixed mesh. If you are not sure exactly what level/pitch your kayak will run at, then do a series of different levels/pitches and interpolate to find the correct one. Even though this is multiple simulations you will find it much easier than the rigid body solver.

In fact you will find this very challenging to go good verified results even with just a fixed mesh. You model is quite a challenging one (I have tried doing it on a kayak model myself

)

dkmred · March 13, 2017, 04:29

Hey ghorrocks,

thanks for this answer.

Excatly, I want to simulate the sinkage and pitch angle as well as the x-Direction due to stream forces.
I will try your suggest of the non rigid body solution today.

But I am still asking how to handle the rigid body solver. My idea was to start simulating just the sinkage/water level (the sinkage of the boat I have roughly calculated manually) with a water/air speed of 3 ms. As a second simulation, with initial values of previous simulation, I thought about adding rotation and x-direction (driving direction) to stabilize the simulation. But as you see, still stuck at the first step

.

How are your experiences about simulating your kajak?
I have also tried a couple of ted and hex variants, with inflation layer and without? How about your experiences there?

Best Regards,

dkmred

ghorrocks · March 13, 2017, 05:39

My experience is that getting a mesh independent solution on a stationary mesh model is challenging as you have a few issues:
1) A turbulent boundary layer to resolve to get accurate drag
2) A laminar to turbulent transition which means a significant section of the hull is actually laminar
3) A wave pattern which significantly affects the drag results

You need to have all these basic issues under control before you even consider a rigid body simulation. So you will my suggestion of a series of stationary mesh simulations much easier to do.

March 11, 2017, 07:00	Simulation Ship Hull, CFX solver interrupts	#1
dkmred New Member Andy Join Date: Jul 2016 Posts: 7 Rep Power: 9	Hey Guys, I am trying to simulate a kayak ship hull. At the start I have simulated a simple cube in two phases. That worked well and I could clearly post process and visualize the wave pattern. With those settings I tried to get the ship hull running. But since months I did not get the simulation to run. Most error messages are: - negative volume elements (in spite of good aspect ratio / orthogonal quality and skewness) - fatal overflow I have tried modifing the mesh, I have tried a lot of different boundary conditions for the walls, changed the turbulence model and so on....I am really desperate and have no glue anymore. I hope some of you did handle with anything similar before and can have a quick view over my ccl file if I used any trouble causing settings. Additionally I've added a picture of my meshing and the domain in cfx pre. Mesh.jpg settings.zip domain.jpg Thanks for your help in advance, dkmred

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
CFX Solver stopped with error when requested for backup during solver running	Mfaizan	CFX	40	May 13, 2016 06:50
fluent divergence for no reason	sufjanst	FLUENT	2	March 23, 2016 16:08
ANSYS CFX simulation: Post CFX error	MInXD	CFX	3	May 19, 2013 19:24
CFX Solver Stops the Running	ashtonJ	CFX	1	December 4, 2011 04:33
CFX 5.5	Roued	CFX	1	October 2, 2001 16:49

March 12, 2017, 05:27		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	You are using a rigid body solver for this, presumably for the motion of the kayak. This makes the simulation a zillion times harder and is not recommended. The recommended approach is to assume a water level/pitch and model that with a simple fixed mesh. If you are not sure exactly what level/pitch your kayak will run at, then do a series of different levels/pitches and interpolate to find the correct one. Even though this is multiple simulations you will find it much easier than the rigid body solver. In fact you will find this very challenging to go good verified results even with just a fixed mesh. You model is quite a challenging one (I have tried doing it on a kayak model myself )

March 13, 2017, 04:29		#3
dkmred New Member Andy Join Date: Jul 2016 Posts: 7 Rep Power: 9	Hey ghorrocks, thanks for this answer. Excatly, I want to simulate the sinkage and pitch angle as well as the x-Direction due to stream forces. I will try your suggest of the non rigid body solution today. But I am still asking how to handle the rigid body solver. My idea was to start simulating just the sinkage/water level (the sinkage of the boat I have roughly calculated manually) with a water/air speed of 3 ms. As a second simulation, with initial values of previous simulation, I thought about adding rotation and x-direction (driving direction) to stabilize the simulation. But as you see, still stuck at the first step . How are your experiences about simulating your kajak? I have also tried a couple of ted and hex variants, with inflation layer and without? How about your experiences there? Best Regards, dkmred

March 13, 2017, 05:39		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,703 Rep Power: 143	My experience is that getting a mesh independent solution on a stationary mesh model is challenging as you have a few issues: 1) A turbulent boundary layer to resolve to get accurate drag 2) A laminar to turbulent transition which means a significant section of the hull is actually laminar 3) A wave pattern which significantly affects the drag results You need to have all these basic issues under control before you even consider a rigid body simulation. So you will my suggestion of a series of stationary mesh simulations much easier to do.