GearHeads Forum

Bill:

Yup, just finished watching it do 12 hours. Not sure why it might lose its pivots information,
but then I suspect a great many bugs exist in there..

(nice looking simulation by the way...)

Art

Art

Thanks for the quick reply.

I was surprised this morning when I reloaded my local copy of MotionWorks-Moving.vcp file.  It too failed because did not contain the pivots that connect the gears/hands/pinions to the movable frame.  I also downloaded my posted version of the file and it too did not work.  When activating the simulation toggle I got eight error messages "Joint removed! No chain found."

I reconnected the pivots, saved the project and also re-created a mech file of the entire project.  This time I used the mech creation function while the simulation toggle was "on"  The resulting mech file did contain the pivots...I'll call this problem as my operator error.  The reconnected .vcp file also could be reloaded and worked properly.

I am not sure why the local and posted MotionWorks-Moving.vcp file was in failure mode this morning.

Did the file work for you? 

Bill Michael

Bill:

  lol, youve probably already stretched its limits..

  Im not sure just how deep it will nest, shouldnt be any limit in reality
but honestly its been so long Im not sure exactly what limits it anymore.


Art

 

Art

I'm continuing my experiments with the Auggie Vector Processor.  I decided to take an example of a clock's motion works model that I developed a couple of years ago and recreate it using the Vector Processor.

For non-clock folks motion works are a set of gears that provide a 12:1 reduction between co-axial input/output shafts.  Tower clocks use such a mechanism to take the shaft that drives the minute hand and reduce the rotation to power an hour hand.

Attached are two VCP files:

The first file MotionWorks-Stationary.vcp shows gear set and hands with the pivots attached to the background.  I created a mech file to serve as the starting point for the moving version.

The second file MotionWorks-Moving.vcp imports the above mech file and attaches it to a movable platform that crawls along a long rack. All the gears and clock hands from the above mech file had to be detached from the background and then reattached to a movable platform using the pivot feature of the vector processor (this is a total of six pivots).  A pinion/rack set attaches via two more pivots to the movable platform the left pinion has a motor attached to drive the entire assembly to the right along the rack.

The third file is a zip file of the mech files that supported the development of the above two simulation models.

My goal is to develop sets of mech files of components that have been tested and working.  Multiple mech can be then imported to serve as the basis of more complex simulated mechanisms.

Possible bug?::

I'm not sure if this is a bug or feature of mech files.

The mech file MotionWorks-Moving-error.mech  did not save any of the pivots that make the model work. 

I did try importing your "foot-walker" mech file and placed it above a static floor.  The pivots in your foot-walker model were preserved and the foot-walker model worked as expected.  Perhaps the complexity of my moving model exceeded the capabilities of the program that creates the mech files.

I am quite impressed with the physics behind the vector processor and do plan to continue exploring its features.  I will continue to stretch the limits of the program.

Bill Michael