Simulating the Physics World

Simulating the physics scene

Now, we’ve only had some theory so far, but haven’t seen any simulation yet. To simulate, we will need to keep calling the quickStep(stepSize) function on the OdeWorld instance. stepSize is how much time should be simulated in one step. To get the most stable simulation, it is recommended that the stepSize be kept constant.

The problem with using the delta time of a task to step the simulation is that the time between tasks might not be consistent. To get around this, a deltaTime accumulator is used to figure out how many steps must be taken. When a step is performed, the world is iterated a few times, you can specify how much times the world is being iterated by calling the setQuickStepNumIterations(num) function on the OdeWorld instance.

Here’s a small example showing a simple simulation showing an iron ball falling from a ridge:

from direct.directbase import DirectStart
from panda3d.ode import OdeWorld, OdeBody, OdeMass
from panda3d.core import Quat

# Load the cube where the ball will fall from
cube = loader.loadModel("box.egg")
cube.reparentTo(render)
cube.setColor(0.2, 0, 0.7)
cube.setScale(20)

# Load the smiley model which will act as our iron ball
sphere = loader.loadModel("smiley.egg")
sphere.reparentTo(render)
sphere.setPos(10, 1, 21)
sphere.setColor(0.7, 0.4, 0.4)

# Setup our physics world and the body
world = OdeWorld()
world.setGravity(0, 0, -9.81)
body = OdeBody(world)
M = OdeMass()
M.setSphere(7874, 1.0)
body.setMass(M)
body.setPosition(sphere.getPos(render))
body.setQuaternion(sphere.getQuat(render))

# Set the camera position
base.disableMouse()
base.camera.setPos(80, -20, 40)
base.camera.lookAt(0, 0, 10)

# Create an accumulator to track the time since the sim
# has been running
deltaTimeAccumulator = 0.0
# This stepSize makes the simulation run at 90 frames per second
stepSize = 1.0 / 90.0

# The task for our simulation
def simulationTask(task):
    global deltaTimeAccumulator
    # Set the force on the body to push it off the ridge
    body.setForce(0, min(task.time**4 * 500000 - 500000, 0), 0)
    # Add the deltaTime for the task to the accumulator
    deltaTimeAccumulator += base.clock.dt
    while deltaTimeAccumulator > stepSize:
        # Remove a stepSize from the accumulator until
        # the accumulated time is less than the stepsize
        deltaTimeAccumulator -= stepSize
        # Step the simulation
        world.quickStep(stepSize)
    # set the new positions
    sphere.setPosQuat(render, body.getPosition(), Quat(body.getQuaternion()))
    return task.cont

taskMgr.doMethodLater(1.0, simulationTask, "Physics Simulation")

base.run()