Simulation using Jinnigine + Jinnigine with Java 3-D

65 views
Skip to first unread message

Navonil Mustafee

unread,
Jul 10, 2011, 5:04:25 AM7/10/11
to jinn...@googlegroups.com, Eberhard Bischoff, Navonil Mustafee (GMAIL)
Hi Morten

Thank you for making Jinnigine open source!!

I have been experimenting with Jinngine. The problem context is container loading, wherein
containers are distinguished by their dimension properties and have different mass. I would
like to do vibration analysis (as the containers are being transported) of the Newtonian motion
of the individual boxes of cargo. However, I would like to do this "faster than real time" (as
this is a simulation experiment).

Just to make it clear as to what I mean by "faster than real time"- In the "ragdoll example", if
I were to pull one of the constituents then the doll would collapse. And the different stage
of the collapse would mirror what would happen in real time. I just need to speed-up this 
collapse. However, the end end result of the "real time" collapse and the "speeded-up collapse" of the rag doll would be identical. Is the "faster than real time" simulation possible
in Jinnigine.

Finally, I would like to work with Jinnigine and Java 3-D. Are there any tutorials available? I 
just need an example and then work out the rest.

Many thanks with your patience in reading this.

Nav

On 10 July 2011 08:04, Morten Silcowitz <mor...@silcowitz.dk> wrote:
Hi again
Yeh, I can understand your worries :) but i'm still here :)
I took a look at your code, seems to be the 0.7 version. I think you
should check out the geometry_rework branch and use that instead. (the
newest version of jinngine) It might fix your problem, and it would
only require smaller adjustments in your code. Let me know how it
works out for you.

Morten


On Sun, Jul 10, 2011 at 12:05 AM, Abhishek Verma
<toabhis...@gmail.com> wrote:
> Hi Morten,
> Thanks for your mail.
> As there was no activity , I was just windering weather jinngine is still on
> or dead.
> Will wait for your responce.
>
> Thanks and Regards
> Abhishek Verma
>
>
> On Fri, Jul 8, 2011 at 6:04 PM, Morten Silcowitz <mor...@silcowitz.dk>
> wrote:
>>
>> hi Abhishek
>> I'll take a look as soon as I can. Its just that I just moved to
>> Dublin, and still don't have internet connection in my apartment. Hang
>> in there.
>>
>> Morten
>>
>>
>> On Wed, Jul 6, 2011 at 8:32 PM, Abhishek Verma
>> <toabhis...@gmail.com> wrote:
>> > Dear friends,
>> > I just joined community and I was trying to understand jinngine physics.
>> > I picked up RagdollExample and tidy up code littlebit and changed few
>> > boxes
>> > to Capsule.
>> >
>> > But output I didnt expect. After falling down, ragdoll capsules keep
>> > jumping
>> > and go out or scene.
>> >
>> > Please advice me what to change to make it work.
>> >
>> > Source Attached..
>> >
>> > Regards,
>> > Abhishek
>> >
>
>

Morten Silcowitz

unread,
Jul 10, 2011, 5:45:26 AM7/10/11
to jinn...@googlegroups.com
Hi :)

You can always do things as fast as you want to. Every time you call
Scene.tick() all object motions are computed and positions are
integrated forward in time. In the demos the framerate is locked to
around 60hz (vsync) to match the refresh rate of the display device.

But you can just compute all the frames you want (as fast as possible)
and the only display some of them, or record positions/velocities onto
the disk for later analysis.

You most likely want the time-step size to be lower than the size used
in the demos, to increase accuracy.

Morten

Reply all
Reply to author
Forward
0 new messages