Bug in pendulum

If you have problems installing or running Algodoo, have found a bug or need in-game help - this is it!

Bug in pendulum

Postby vk4 » Mon Dec 30, 2013 1:38 am

Here there are no choc (it's an axis) so where the energy goes ?
Attachments
test pendule 5.phz
(14.57 KiB) Downloaded 18 times
vk4
 
Posts: 31
Joined: Sun Dec 22, 2013 5:14 pm

Re: Bug in pendulum

Postby electronicboy » Mon Dec 30, 2013 2:25 am

However, you also somewhat have a simulated gravity induced by the magnet, take a pendulum swinging, it will slowly lose energy as it's fighting against gravity, as well as forces/friction exerted onto the axle (they may be miniscule, but they are still there). (in fact, looking with the view forces tool and the show value's check box, it does look there is quite a bit of friction on the axle).

you also have to remember, that algodoo can only ever be so accurate, Algodoo uses a lot of high precision floating point calculations which produce the results in the application, however desktop CPUs aren't amazing at high precision FP calcs, so there is always going to be some minor calculation errors along the way, algodoo does take some steps to reduce the errors, however they can only take that so far before the error prevention limits the physics engine too much.
When asking for help, READ THE STICKIES!
electronicboy
 
Posts: 1694
Joined: Mon Aug 31, 2009 6:18 pm

Re: Bug in pendulum

Postby vk4 » Mon Dec 30, 2013 2:46 am

ok, but why this scene, the rotation increase alone ?
Attachments
biz2.phz
(13.7 KiB) Downloaded 16 times
vk4
 
Posts: 31
Joined: Sun Dec 22, 2013 5:14 pm

Re: Bug in pendulum

Postby electronicboy » Mon Dec 30, 2013 3:59 am

Setting legacymode to 0 on the objects axel appears to severely limit the issue, seems like the axle isn't in perfect centre, which is really a calculation issue when trying to find the dead centre of the object (with perfect precision for the calculations anyways).

However, it seams to be magnified with the new axle physics which attempt to reduce an issue in older versions of axle physics where axles would easily stretch and deform causing a magnitude of issues, which the new system fixed however did bring a few drawbacks.
When asking for help, READ THE STICKIES!
electronicboy
 
Posts: 1694
Joined: Mon Aug 31, 2009 6:18 pm

Re: Bug in pendulum

Postby vk4 » Mon Dec 30, 2013 3:06 pm

in this scene, no axis, and the system is unstable !

idem with ellipse on a circle, maybe the force from wall works finally ! compare with a circle on a circle: no energy in addition
Attachments
el7.phz
(22.92 KiB) Downloaded 14 times
el2.phz
(29.44 KiB) Downloaded 12 times
vt3.phz
(25.23 KiB) Downloaded 13 times
vk4
 
Posts: 31
Joined: Sun Dec 22, 2013 5:14 pm

Re: Bug in pendulum

Postby vk4 » Tue Dec 31, 2013 3:56 pm

I can built a scene with no shock only rotation, the energy of the system is not fixed.

In fact, even an ellipse on Earth give more and more energy (no friction and restitution = 1, loot at xct scene, it seems it's when the ellipse change its angular velocity (change positive direction to negative or - to +)
Attachments
xct.phz
(7.34 KiB) Downloaded 13 times
oeuf.phz
(5.76 KiB) Downloaded 15 times
rf5.phz
(21.3 KiB) Downloaded 12 times
vk4
 
Posts: 31
Joined: Sun Dec 22, 2013 5:14 pm

Re: Bug in pendulum

Postby jon_joy_1999 » Wed Jan 01, 2014 7:46 am

This is not a bug, this is garbage in, garbage out. If you want Algodoo to produce realistic results you have to provide it with realistic parameters. These include frictional forces, rebound forces, and any other kind of forces.
Algodoo will allow you to play God and divide by zero, but don't be surprised when it gives you nasal demons.
Image
BSrac = BoincStats recent average credit
I'd rather be network computing.
jon_joy_1999
 
Posts: 233
Joined: Fri Dec 09, 2011 12:51 am

Re: Bug in pendulum

Postby vk4 » Wed Jan 01, 2014 10:59 am

Happy new year jon_joy1999 ! I put small friction and restitution not at 1 and this give more energy too ! friction is at 0.00001 and restitution at 0.9999. And if you try with a circle all is fine with friction at 0 and restitution at 1. And even you put the frequency at 24000 Hz, it's the same result.

I try to test with friction at 0.01 and restition at 0.99 (for 2 objects) and the energy move down but sometimes (where there is rotation before schock) the energy move up. test the scene "friction1 energy inc" at time 1.4s the energy move up even friction and restitution.
Attachments
friction1 energy inc.phz
(5.7 KiB) Downloaded 19 times
friction1.phz
(5.63 KiB) Downloaded 18 times
vk4
 
Posts: 31
Joined: Sun Dec 22, 2013 5:14 pm


Return to Help / Bug reports

Who is online

Users browsing this forum: No registered users and 4 guests

cron