Stopping instability

Share with us how are you using the powerrrr of the force

Moderator: Alain

Stopping instability

Postby martinw » Tue Aug 24, 2010 11:10 am

Hi,

Can joint instability feedback from a joint be interrogated via the API?

Martin
User avatar
martinw
 
Posts: 30
Joined: Wed Jun 01, 2005 1:19 pm
Location: Aberdeenshire

Re: Stopping instability

Postby Julio Jerez » Tue Aug 24, 2010 1:52 pm

give me an example.
Julio Jerez
Moderator
Moderator
 
Posts: 12249
Joined: Sun Sep 14, 2003 2:18 pm
Location: Los Angeles

Re: Stopping instability

Postby martinw » Tue Aug 24, 2010 2:51 pm

For example when the framerate of my application ............

http://www.youtube.com/watch?v=DbsJlP7_ChQ

........... drops due to joint interactions I suppose (there are quite a few ball joints in the yellow umbilical which in real life supplies power and communication to the underwater robot), if I then pull on the umbilical with the ROV (Remotely Operated underwater Vehicle) the umbilical breaks and jumps around in a mad fashion. I can increase the mass of the umbilical elements but then it acts less like a cable and more like a jointed rod. If I had some feedback from Newton that instability was occuring at a particular joint I could back off the ROV power like it was being pulled back. The umbilical has a kevlar sheath in real life so is very strong.

Thanks Julio.
User avatar
martinw
 
Posts: 30
Joined: Wed Jun 01, 2005 1:19 pm
Location: Aberdeenshire

Re: Stopping instability

Postby Julio Jerez » Tue Aug 24, 2010 3:14 pm

Ok I see. Also hey that is really cool :mrgreen:
are you simulating every peia as a rigid body?
are you using custom joint?

one way to solve is by special casing the cable joint an add a kenematic pass to correct the error and or measure the force tenstion on the joint.
I said in teh pass tha I was goin to add a funtionalitu to newton to solev tehis probelem automaytically.
you are not the only one with that problem, KellaSprint, pJani, and many oeth prople also have it with the Mutibody car.

He game me a demo to work with, but was so long ago that I do not ememer where is it.
anyhway can I have a demo of this?

I am about to start woking of the SDK demos again this weekend, maiby I can take a look at how to adress this issue by either custmizing the joint,
or with a combination form the engine and the joint.

I am sure we can adress that problem to work with it the parameters of the real thing.
Julio Jerez
Moderator
Moderator
 
Posts: 12249
Joined: Sun Sep 14, 2003 2:18 pm
Location: Los Angeles

Re: Stopping instability

Postby Julio Jerez » Tue Aug 24, 2010 5:44 pm

Hey I hope those are not the one BP is using here in the Golf of Mexico. 8)
Julio Jerez
Moderator
Moderator
 
Posts: 12249
Joined: Sun Sep 14, 2003 2:18 pm
Location: Los Angeles

Re: Stopping instability

Postby martinw » Wed Aug 25, 2010 10:23 am

Hi Julio,

Thank you, it seems to be coming together quite well 8)

I am simulating ever piece of the yellow umbilical cable as a rigid body with ball joints in between (I presume they're custom as they're through TV3D) 25 in total so far.

I have heard of soft body physics but don't really know much about it, do you think the umbilical would be better simulated as a soft body?

Being able to know the force/tension in the joints would be a great help I think, as I could use this information to limit the power of the underwater robot as things became unstable.

The types of ROVs we sell are very similar to the ones in the GOM but they have electric thrusters rather than hydraulic ones and are a bit smaller.

I'll send you a zip file of the program in an e-mail.

Martin
User avatar
martinw
 
Posts: 30
Joined: Wed Jun 01, 2005 1:19 pm
Location: Aberdeenshire

Re: Stopping instability

Postby Julio Jerez » Wed Aug 25, 2010 2:47 pm

Soft Body can be a solution of for the cable cource.
Newton does not have that now, it will come later afte I have a editor.

Mean time I beleive that teh special casing for the cable joint will suffices.
Soft bodies solve one part but it will also open a while new can of worms in other areas.
Julio Jerez
Moderator
Moderator
 
Posts: 12249
Joined: Sun Sep 14, 2003 2:18 pm
Location: Los Angeles

Re: Stopping instability

Postby PJani » Thu Aug 26, 2010 5:38 pm

Julio Jerez wrote:Soft bodies solve one part but it will also open a while new can of worms in other areas.


Nice phrase. :D
| i7-5930k@4.2Ghz, EVGA 980Ti FTW, 32GB RAM@3000 |
| Dell XPS 13 9370, i7-8550U, 16GB RAM |
| Ogre 1.7.4 | VC++ 9 | custom OgreNewt, Newton 300 |
| C/C++, C# |
User avatar
PJani
 
Posts: 448
Joined: Mon Feb 02, 2009 7:18 pm
Location: Slovenia

Re: Stopping instability

Postby martinw » Sun Sep 26, 2010 2:46 pm

I think I understand Newton a little better now and have limited the mass ratios of interacting bodies and things are a lot more stable and faster. The values may not be exactly what the real world masses are but Newton (and more importantly my PC) seem to cope a lot better with them :D

http://www.youtube.com/watch?v=6EQ3fvZyT7o
User avatar
martinw
 
Posts: 30
Joined: Wed Jun 01, 2005 1:19 pm
Location: Aberdeenshire

Re: Stopping instability

Postby Julio Jerez » Sun Sep 26, 2010 3:02 pm

Ha much better, :D

I have not forgotten about your problem, and I definitely wan to address it, It is just that I am a little short on time now, and the enthusiasm is not as strong as it was 5 year ago.
Every week end the time I spend in the engine I try to push the File format, because this is what will help me and the users to move more quickly in setting up scene, and finding Bug.

The file format will allow me to work in my own framework rather than setting up debug environment for each person with a bug. It got to the point when I have hundreds of demos that I do not know what is what and some of them stop working as I make a change to the DLL.
Plus some people have their application that work in exclude more, and when the modify the bug do not show up, other people use Newton in Linux and Mac, all in all it is a really bad way to debug.
You case is one example of what I say, I cannot really debug any of the TV problems because I can step in secure code, and I do not have the TV SDK,
but if you were able to export the Scene, (The physics part only) then it will be easier to work on it.

I am hopping the File format solve those problems and only leave the debug with an demo for the rare cases.
Can you please repost the bug in this forum, so the when I get to the point to debug and come up with a solution for that big problem I can search in one place.
viewtopic.php?f=24&t=6127

PS:
wow I wacth teh comple video, Wholly molly teh robot arm manilunation it looks awesome when capping the well. :mrgreen:
what are you dowing and underwater simulation?
It looks better than the symulations shown in televsion during the British Petruleum Oil slick in Luisiana cost
Julio Jerez
Moderator
Moderator
 
Posts: 12249
Joined: Sun Sep 14, 2003 2:18 pm
Location: Los Angeles

Re: Stopping instability

Postby martinw » Wed Oct 06, 2010 12:27 pm

Hi Julio,

Thanks for that, yeah I'm making a 3d underwater training environment for our ROV operators.

http://www.sub-atlantic.com

I've just about got the 3D environment sussed, I'm now writing the code to interface to our operator interface hardware through the RS232 ports on the PC.

I'm parsing the commands that would normally go to the sub-sea hardware (ROV, TMS, Valve Packs etc) and translating those commands into actual movement of the ROV, TMS, Manipulators etc and also ambience changes (such as lighting levels etc) within the virtual 3D environment.

Too much fun :D

Newton is great, thank you very much for writing it 8)

Martin
User avatar
martinw
 
Posts: 30
Joined: Wed Jun 01, 2005 1:19 pm
Location: Aberdeenshire


Return to User Gallery

Who is online

Users browsing this forum: No registered users and 12 guests

cron