Rotating Joints Should have a "Motor" Option

[elihuisland]

Many designs involve motors powering rotating components.  Why not have an option when assembling rotating joints to simply cause the joint rotate at a selectable speed.-- in essence a motorized joint.  I already model devices with motion links, then couple them (using more motion links) to a manually-rotated crank modeled near the edge of the screen.  This usefully animates the whole device in the model space, but is hand-cranked!   Almost everything to do this is already in the program, since joint animation is part of the preview in joint assembly. 

[otomasa]

I completely agree, it would be wonderful with real motor specs though in future iterations..

[JRKuyper]

I would take it a huge step further (I have been thinking about this already):  Add a motion module.  Currently, we can create models in the Modeling module, and analyze stresses in the Simulation model.  But we are missing the basic "calculate forces" step.  It would be great to have motors (rotational, linear), springs, gears (both linear - rack & pinion, and rotational), inertia, momentum, cylinders (either hydraulic or pneumatic), etc. to calculate either static or dynamic resultant forces (not the same as internal stresses calculated in Simulation).  It could even include gravity so you could simulate how an object falls - yielding impact forces.

This module would handle things like:  How quickly can I accelerate a box on a conveyor with a 1 hp motor?  How much pressure would I need in my hydraulic cylinder to lift 500 lbs of dirt in my back hoe?  If I sit on the edge of my table, how much force is transmitted through the cross beam?  Basically, it would handle all the stuff engineers learn in Statics and Dynamics classes.

I have seen video games with this sort of kinematic engine, so it shouldn't be a MAJOR investment.

Thanks!

[jkelindberg]

Did you try "Motion study"? Using this tool you do animations of joints.

[elihuisland]

reply below...

[elihuisland]

I did try the Motion Study feature. The feature is good for setting up a quick series of precise joint movements. The feature does not apparently support continuous motion of a joint, however, and even on the slowest setting completes its action in a couple of seconds. If set to repeat, it starts over so that linked joints reset repeatedly rather than continue from where they left off. In my designs, I need to set rotary joints in continuous rotation. By linking these to other joints in the device design that run at slightly different speeds (ie, as in geared shafts), I can watch the action of the device as it runs for several minutes and the relative aspects of its parts change. (imagine a clock design, for example.) Right now I do this by slowly turning a manual crank modeled off to one side of the screen that simulates a motor driving the whole train of events. It would be great to just have a joint with a rotation speed selection (down to fractions of an RPM), direction of rotation, and an on-off switch. The Model space already supports motion that respects joint construction. It just feels odd to make it run with a hand crank.

[JRKuyper]

I also tried playing around with Motion Study.  Like elihuisland, I found it severely lacking.  In his case, he could try rotating it a few thousand times, but then the devise would fly around making it impossible to interpret data.  Only allowing 100 steps is very limiting where each step represents a very small slice of time.  I also found that the motion study does not seem to save.  So I can evaluate a motion, find a problem, fix it, then I have to set up the motion study again.

Lastly, it does not provide any meaningful data.  For example, how big does elihusisland's motor need to be?  Fusion should calculate inertia, and acceleration should be an input.  Then the motor size could be determined (power, torque, speed).  Or, the motor could be defined with a certain power / torque / speed (any of the two), and the actual acceleration would be calculated. Suppose elihusisland's device needed to lift a load.  We would need a way to add loads and gravity to the system (much the same way loads are added to simulation).  He may have a shaft going through a lever.  How much force is acting on the shaft?  It would be very cool to include friction so the system would evaluate more than just applied / inertial loads.  It would also be very cool if you could import the results from the Motion Study into Simulation for stress analysis (will that shaft fail?).

If this is too much to include in Model > Motion Study, it could be a separate Workspace.

Just search the internet for [pick a major CAD system] & "motion simulation" for inspiration.

Thanks

[promm]

Thank you for your idea, the team is reviewing the requirements and will change the status once a decision is made.

Regards,

Mike Prom

[hans]

CNC Gcode interpretation.

I asked SolidWorks way back when (10 years ago) to offer a feature where their mate limited references (Advanced mates) could be script-able, so for example a design assembly of a 3-axis gantry type CNC machine could run a script and execute Gcode.

This feature would vary each "axis" assembly's length mate distance by a value derived from the script (like gcode execution) and all programmed mates move at the same time (interpolated motion). They never did anything with it.

I wonder if this would fit well here.

The idea is to be able to access joint limits or values (I don't yet know F360 well enough) from a text file (like a script) where all joint adjustments would be made in an interpolated way (at a programmed speed) to all finish at the same time.

This would make it easily possible for users  to write a text interpreter that could convert gcode to this script format so when it was run, it could move a CNC machine designed in Fusion360 around on the screen, lazy man's animation.

It would work also for angles so a design of a robot could move through a script.

Just an idea.

[keqingsong]

A large portion of our team is currently focused on sheet metal and improving some of our core modeling tools. This is a good idea, but we'll not confident that we'll be able to tackle it soon. We have this captured and will report back when we are ready to do this. In the meantime, it'd be interesting to see if our API can help get you closer to a motorized joint solution.  https://autodeskfusion360.github.io/

[elihuisland]

Thanks for the update. I can certainly understand how you folks are busy— Fusion 360 is great and I’m sure it didn’t get that way at an easy pace! Writing an API program for a motorized joint would be a fun challenge, but I’m fully engaged in making stuff already. Maybe some of your talented users want to give it a shot?

I’d like to add how very useful Fusion 360 has become to me for visualizing the math-based kinetic sculptures that I build. Many thanks to your team!

[salesukprecision]

Just thinking this today setting a certain R.P.M. would be brilliant....

[jimad3]

I would like to add my voice in support of a motor function with a calibrated speed control in RPM with a lower limit near .001 rpm and an on/off switch. I have a mechanism with many gears and the motion study is too fast and too short-lived to be helpful. I would be neat to have a speed slider.

[nibbleshrew]

As jimad3 said.

I'm just looking for slower animations when setting up a Motion Link and adding a motor option to Drive Link to simply set it to an RPM and have it continuously rotate instead of doing a single rotation and then resetting.

For instance I'm making a little gear driven turntable model and it would be nice to be able to set the driving motor/gear to a specific RPM and observe the realistic rotation of the table and see if it feels right.