@ashscott said:

So you're not getting the issues I described two posts ago.....?

I seem to get it no matter how I run it (apparently the raytest issue was clashing with SU8 but I can't even get it working right in SU7)

try reinstalling sketchup or SP or sumthing ...

@ashscott said:

But Raytest has a tendency to cause SU to flip out - know of a workaround or why I am having the symptoms described?

uh... nooo...
u'd better consult pros like Wacov..

@ashscott said:

I've added a little complexity that is causing this script to crash SP when I try to run it.

As far as I can tell it has something to do with the raytest function - potentailly related to the issue described on this thread: http://forums.sketchucation.com/viewtopic.php?f=180&t=30558

I have tested it in SU7 as well but am getting the same issue - basically the model explodes, the POV zooms way out and all I can find left in the model is a bunch of endpoints (like on the end of a tapemeasure guide......can anyone shed some light?

Ash

that's right sir...
i used raytest to find the point where the wheels touch the floor, then i teleport magnets to them to attract the wheels...

Hi there..
Now probably all of u guyz have used Sketchyphysics at some point, and created physics joints in them.

The problem though, is that sumtimes the joints become glitchy : they either become floppy or fall off completely... not good...

The solution to the problem is to create a joint while the simulation is running.

Ever explored the scripts which make up Sketchyphysics? If so, u've probably seen the module "NewtonServer" sumwhere... . this module is just about the kernel of the Sketchyphysics system. It creates nearly every physical feature of Sketchyphysics, from joints to forces to gravity and so on.

We need one of its methods, which is "createJoint"...

# this example creates a hinge joint
parent={grp} # a grouped object that encases the joint
child={grp} # a grouped object that is attached to the joint

type="hinge"
min=0
max=0
accel=0
damp=0
breakingForce=1000000
limits=[min,max,accel,damp,0,0,breakingForce] # parameters for the joint

xform=parent.transformation
pinDir=xform.zaxis.to_a # the axis of the joint

parentBody=DL;;PtrData.new(parent.get_attribute("SPOBJ","body",nil)) # converted parent body

childBody=DL;;PtrData.new(child.get_attribute("SPOBJ","body",nil)) # converted child body

  jnt=NewtonServer.createJoint(type,xform.origin.to_a.pack("f*"),
            pinDir.pack("f*"),
            childBody,parentBody, 
            limits.pack("f*") ) # the joint itself	

Important: in general, all array objects need ".pack("f*")" after them, and all grouped bodies need to be converted into DL::PtrData (by using "DL::PtrData.new({grp}.get_attribute("SPOBJ","body",nil))" )

The created joint doesnt have the bugs normal joints have. (hopefully... )

Have fun "Sketchyphysic"ing (!)

@hobbnob said:

I haven't had that problem in the 4 years I've been using sketchyphysics, you sure you've got your wheels geom and ontick interval set up correctly?

sure as pants man..

@hobbnob said:

Hehe not a bad idea, but you could just have put push([0,0,-1]) in the ontick field of the wheel, does exactly the same thing

(little extra note, if you want to have the vehicle turn on it's side/upside down and still have it work, then encase your steering block around the drive joint (either hinge or motor dependign on your preference) and they won't collapse)

uh..well.. push([0,0,-1]) creates a violent vibration in the wheels.. so its not that simple..

p.s. i looked in ControllerCommands.rb and found the push()method is the same as $sketchyPhysicsToolInstance.pushBody(), which creates the vibration i talked about...

i created something similar in SketchyPhysics 3 (although 2d) by switching the texture every frame...

[flash=480,360:2mrvf6eu]http://www.youtube.com/v/fa7XopOzq7M[/flash:2mrvf6eu]

sorrie its not downloadable...

@don east said:

Where did you put this script?
Don

u need to put it in the "scripted" section of the Sketchyphysics UI (neither the "ontick" or "ontouch" section would work...)

Ever built a sketchyphysics car? And theres no grip on the ground? Well never fear, 'cos MrDailyBlah is here....(!) (!)

I recently fiddled around with the raytest function, and i thought i could incorporate it in to the sim...

A floor in sketchyphysics is perfectly smooth...it doesnt create friction like a normal surface would. To create more grip, the most obvious solution is to apply negative thrust to the car's body or suspension, but this collapsed the wheel hinges.

I tried to use magnetism as a solution, and i found that if i could use the raytest function, i could find a point under each wheel on the floor, to each of which i can teleport a magnetic object, attracting the wheels to the floor and creating grip without collapsing the car...

That was it in a nutshell... here's the code i used -

# paste the following into the floor
onstart{
# a way to simplify Sketchup.active_model.raytest
def newRay(p,d)
ray=Sketchup.active_model.raytest(p,d)
return ray
end
}
# then i copied the following script 2 each wheel
ontick{
t=[0,0,-1]
ray1=newRay(position,t)
ray2=newRay(ray1[0],t)
s=@simulation.findBody("sphere...") # the magnetic object; theres a unique one 4 each wheel
s.teleport ray2[0] if (ray2!=nil) and (s!=nil)
}

heres a demo of the script

P.S. make sure u know how it works befor u copy it... ( )

NewtonServer.magnetMove? what is NewtonServer? is it a module in SP or sumthing?