Class MiscellanyTest

Performance test that runs the clock_with_gears test.

Shows an error discovered by ERN on Sept 15, 2015.

TO DO* This error state should be fixed or prevented from happening.

Shows an error discovered by ERN on Sept 15, 2015.

TO DO* This error state should be fixed or prevented from happening.

Two balls rotate at the same initial angular velocity, the larger ball has a bigger moment of rotation and therefore spins longer under the standard DampingLaw.

Test the feature whereby a body can selectively not collide with certain edges of another body; here a ball and an pendulum are bouncing on the floor under gravity; the ball collides with the pendulum bob, but not with the pendulum rod; however, the pendulum rod does collide with the floor. *

Setup a RigidBody connected to a 'roller coaster' circle shaped NumericalPath with a PathJoint; also there is a PathEndPoint that limits travel of the RigidBody.

Tests a RigidBody connected to a 'roller coaster' circle shaped NumericalPath with a PathJoint; also there is a PathEndPoint that limits travel of the RigidBody.

Setup a RigidBody connected to a 'roller coaster' hump shaped NumericalPath with a PathJoint.

Tests a RigidBody connected to a 'roller coaster' hump shaped NumericalPath with a PathJoint.

Sets up error condition found on Sept 15, 2015. Results in simulation getting 'stuck', because ComputeForces is unable to find reasonable set of forces. The problem is the rectangle gets jammed between two fixed objects, and the distance between them is very tight. *

Sets up same error condition found on Sept 15, 2015, but at an earlier time. This results in success the first time it is run, but clicking the 'rewind' button results in 'stuck' error. The difference is due to random number generator being different. The first time thru the rectangle doesn't quite get its corner around the vertex of the fixed block, so it never gets wedged. *

Three blocks are connected by joints to mimic the arrangement of test/RopeTest.RopeTest.double_rod_spin_setup. This is intended to be a further test of Joints.

For three-body-spin setup, confirm that energy is conserved and joints stay tight.

For three-body-spin setup, confirm that with smaller time step, energy is conserved and joints stay tight, with much smaller tolerance. Here we use time step of 0.0025 and tolerance of 0.0001 for energy compared to 0.025 and 0.01.

Uses experimental AdaptiveStepSolver on three-body-spin setup, confirming that energy is conserved and joints stay tight, although a rather large time step is specified.

Same test as three_body_spin_test3, but shows that setting AdaptiveStepSolver to use second differences results in greater accuracy (at expense of taking more steps). The tolerance on energy difference is much tighter here.

Runs the three_body_spin setup with damping and a very small step size of 0.001, using the standard RungeKutta DiffEqSolver. This small step size is needed to get the same accuracy as the AdaptiveStepSolver, see test/MiscellanyTest.MiscellanyTest.three_body_spin_test4B.

Runs the three_body_spin setup with damping using the experimental AdaptiveStepSolver. This gives same accuracy as regular RungeKutta solver with a very small step size of 0.001, see test/MiscellanyTest.MiscellanyTest.three_body_spin_test4A.