MSPhysics 1.0.3 (16 October 2017)
-
Cristyan,
There is not much I can do about the chain instability. Artificial chains, connected by joints, was a workaround. There is another way to get the chains to work properly and it would be by reducing Update Timestep, i.e 1/240. The smaller the update timestep the more accurate and stable the simulation is, however everything will be in slow motion when you run simulation.
Speaking of the Domino_2016, to improve performance, open MSPhysics UI, activate the Simulation tab and set Solver Model to Iterative 4 Passes and Update Rate to 3. This change should increase performance.
Speaking of the dominoX thingy, currently there is a bug in the physics engine when colliding with the compound bodies. I even assume that that compound collision bug could be the cause of the unstable chains in MSPhysics. To fix that, I advise you to select all the dominoes and set their shape to convexhull (Context Menu->MSPhysics->Shape->Convel Hull). Then set floor shape to Static Mesh. Then set World Scale to 20. You can as well, optionally, adjust Update Timestep, Solver Model, and Update Rate for the desired speed, performance, and stability. For that particular model, I figured an update timestep of 1/120, a solver model of iterative 4 passes, material thickness of 0, and an update rate of 2 would guarantee stable results. The idea is that you adjust simulation options for the most desired effect.
-
Yes, msp_lib_no_sdl.so is there.
-
Okay, I'll see what I can do about it.
-
Thank you !! if I adjust Update Timestep the chain simulation is working
And another thing in the last version the PLAY button is not working.. -
Cristyan, I fixed the replay play button. Try version 0.7.2. It's in the first post.
-
Play and Reverse buttons work fine. Play, Stop and then Play again or Reset is running into problems...
Emitted objects are not reset. Please test the model.
-
faust07, Thanks for posting, the accumulation does not grow, is that a shortcoming of MSPhysics. or something you control?
-
This you can control with MSPhysics. The lifetime of the objects is very short, so that they will be deleted after 100 Frames.
-
@faust07 said:
Play and Reverse buttons work fine. Play, Stop and then Play again or Reset is running into problems...
Emitted objects are not reset. Please test the model.Thanks for the report faust. I will fix it and also get the feature to save replay into model in the next release.
-
Where can I find documentation on how the servo works? There was a comment way back that had...
@unknownuser said:
CODE: SELECT ALL
rightx Left - right arrows
righty Up - down arrowsleftx A - D
lefty W - Skey('1') 0-9
key('a') A-Z keys
key('enter') All the Windows function keys. (lctrl, rctrl, space, backspace, etc...)
key('numpad1') 0-9 Right numpad keys
key('+') Numpad operations (+, -, *, /)key('lbutton') Mouse keys (mbutton, rbutton, xbutton1, xbutton2)
get_var('variable') ... for script.
What do I do with this?
-
Xbox controller please !!!! -
Announcing version 0.7.3. Download in first post.
Faust07, I fixed the replay bug. I also added the feature to save replay into model. The replay is saved into model automatically just like in SP.
SynSuka3d, I added joystick controls similar to SketchyPhysics.
Here is a sample model: See controls in modelNote that in order for the joystick to work it must be attached to USB before SketchUp is opened; that is, restart SketchUp if joystick doesn't work in the first place.
-
Good news, as always.
-
The boat-car is a good example to learn the use of thrusters and joints (motor joint as a component copy).
Version 0.7.4 (May 26, 2016) has the same problem in Replay that Play, Stop and then Play again is running into problems...
Maybe it is my computer... -
The emitter test you posted in one of the earlier post works as expected. Can you elaborate more on what problems you encounter?
-
Sorry. Maybe I have the function of the stop button just now understood. I think that with the stop function the model can be held and saved in an intermediate state of the simulation. After that you should delete the replay storage, otherwise unexpected intermediate phases of the simulation arise if you start or play again.
-
Announcing version 0.8.0!
This version comes with many bug fixes and improvements. As well as with two new joints: CurvySlider and CurvyPiston. A demonstration of these two joints is available at 3DWarehouse: https://3dwarehouse.sketchup.com/model.html?id=d21984a9-477b-44d7-9eca-c191ae85afe4 -
The new Curvy Joints improve animation along paths fundamentally. The possibilities are not yet in sight. Great is that you can take curved sections and even edges with the joint in a group without to weld them, and therefore the curve in turn can be moved. Your examples show this excellent.
Two questions:
What are the rules for the orientation of the joints to an individual curve - exactly parallel?
It seems that old Servo Joints with Oscillator as a controller are not functioning as in version 0.7.0? Tested for example, with birdy and orca. -
Well I hopefully thought I'd try again.
http://gettheworkdone.net/images/MSPhysics/6-13-2016_0.8.0.jpg
the error at least looks a little different.
Windows 10
-
Hello Faust,
Curvy joint orientation has no effect on a curve or whatsoever. The curve may be parallel or not parallel to the actual joint - it doesn't affect the way joint behaves. The joint object just represents the origin of the joints. The edge closest to the origin of the joint is considered as the starting edge and all its connected edges are considered as the consequent edges. As long as the curve (or a set of connected edges/arcs) is created within the joint, with the angle between each edge set to a maximum of 45 degrees, the curvy joint will work fine. To my experience CurvyJoints operate better on iterative solver than on exact solver. So if you use curvy joints, setting a solver model to iterative 4+ passes will guarantee the optimal behavior.
The servo joint was reworked in this version. Now instead of accel and damp, it uses angular rate, and the SP Mode was removed. If your joints relied on SP Mode, their controller values ranged from 0.0 to 1.0. Now that is changed. The controller values range from -min_angle to max_angle. So for your birdy the servo controllers should be changed to something like,
oscillator(100)*30. The limits should be disabled or adjusted to -30,30.
Ralph,
Sometimes the zip file fails to extract all content. Try checking if Plugins/MSPhysics/win64/2.0/msp_lib_no_sdl.so exists.
Advertisement