Pool shooting robot needs help with rail-gun pool-cue thruster

[BVarv]

I'm building a robotic system that plays billiards. This thread is related specifically to the subsystem of that machine that thrusts the pool cue. The general plan is to embed powerful magnets in the handle of a pool cue and to then have the cue pass through a copper tube wrapped in magnet-wire. By energizing the coils, I should be able to retract the cue and then, by reversing current direction, thrust the cue through the tube with varying stroke power based on the voltage applied.

I plan to use two 12V batteries in series to charge a large capacitor to up to 24V (depending on desired shot strength). Then, by having a controller give a signal to a relay, I hope to dump the stored energy from the capacitor through the coils to accelerate the cue stick with force hopefully comparable to that of a human.

Here are the components I'm planning to start with:
Controller: Arduino
Batteries: 12V 8ah buggy batteries
Relay: Automotive starter
Capacitor: Car Audio 8 Farad
Magnet wire: 14 GA, 156ft
Magnets: Disc Neodymium N52

Questions:
What red flags to you see?
What is dangerous about this system?
How can I calculate the expected force on the cue stick?

Thanks for reading and thanks even more for any help!

 

[kellys_eye]

I'd consider using some form of control for the discharge rather than the 'brute force' method as you will have much more control over the cue ball.

If you consider a simple loudspeaker and how it operates, you can easily see the cone movement on larger speakers so consider the possibility of a similar arrangement for the cue holder - albeit on a larger scale of course.

Any pool player will tell you 'it's all about ball control' ....

 

[BVarv]

Thanks Kellys_eye!
Yes, this is similar in application to a loud speaker... if the speaker only made one loud bump.

I wonder what you mean by "some form of control for the discharge." Do you mean pulse width modulation? Can that even happen with capacitor discharge at the current level necessary in my application?

I did some tests (see the vid) stacking 25,000Mfd capacitors to total ~.5 Farad and observed that the capacitor discharge is SO fast that the coils didn't seem to have time to fully accelerate the test stick. How can I extend the force duration?

Yes, cue ball control is paramount. We agree on that.
I have a subsystem to manage where on the cue ball the cue stick strikes and also one to control the direction that the cue stick is aimed. The subsystem discussed here is focused on controlling the energy the cue stick transfers to the cue ball.

Time to invest in the full scale components?
Any thoughts or insights are much appreciated. Thanks again.

 

[Bluejets]

Perhaps consider having the cue with a spring and a gearmotor with cam trip at the end. Would not have to travel very far , maybe 1 to 2 inches of pre-load I'd imagine. Much easier to set up than coils, magets and the like. When shot is required, cue lines up close to ball and gearmotor retracts the cue to load the spring and then discharge via the trip cam.

 

[BVarv]

Thanks Bluejets!
I should add that some goals in this subsystem are to provide the full spectrum of shot strength: Break shot all the way to super-soft, 6" shots, to be fully autonomous (no human interaction between shots to reset, etc.) and to be elegant in action (look nice).

Here are some ideas considered for the thrust subsystem:

Slingshot (elastic tension):
Good: repeatable/variable shot strength related to distance cue is retracted, light weight, similar to human in stroke appearance.
Bad: Some sort of threaded rod system is needed to mechanically retract the cue against the elastic. A trigger system is needed to release then reset for next shot. Complex, noisy and lots of parts to potentially fail.

Spring (metal compression):
Good: repeatable/variable shot strength related to distance cue is retracted, light weight
Bad: Stroke length is likely shorter than a human, making pool shots "look weird." Some sort of threaded rod system is needed to retract the cue. A trigger system is needed to release then reset for next shot. Complex, noisy and lots to go wrong.

Trebuchet-like system (falling mass transfers energy to the cue stick):
Good: Elegant and interesting to watch
Bad: The orientation of the shooting system must be consistent relative to downward gravitational force. (jacked up shot angles mess with the shot force). Heavy system. Complex system will be necessary to wind-up/release/reset the counterweight.

Rail-gun (electromagnetic force):
Good: No moving parts other than the cue. No mechanical triggers. Quiet.
Bad: Heavy! Likely a lot of heat involved. Super strong magnetic fields near humans and magnetically sensitive sensors on other subsystems.

I'm going with the rail-gun system first because weight is not a critical concern for this machine and because I am interested in keeping this subsystem mechanically simple. I'm ignoring the heat and magnetic interference issues for now. (eek!)

Thank you for helping me flush this out!
Other thoughts, ideas, concerns, reasons to not order parts?

 

[kellys_eye]

By having a mechanical (electrically controlled of course) system of spring compression - but with the position of spring also adjustable - you can get different stroke length and power.

I'm envisioning a fixed length spring (compressible) with a variable back-stop for pre-load.

You could even have the back end of the spring 'powered' for movement during the release period.

Have you graphed the various 'strokes' in terms of mechanical movement? I suspect it may look like the output of an ADSR (envelope shaper) - and another for any control of the 'back end' of the spring - by which results you can replicate digitally.

Of course pneumatic and/or hydraulic would do it all.

 

[BVarv]

No. I have not graphed any of the stroke positions through various strokes... break shots... soft shots.
That is a great idea. on it.

After thinking about it, I agree that the output would have characteristics that could be replicated digitally. This gets gears turning... interesting.

Having the stroke "Look human" is a HUGE goal for this machine. The missing link seems to be: How to control the output of a capacitor dump to shape the force curve created in the coils? Is there a device that exists already that can do this?

I'm still hooked on the rail gun idea because I noticed that the force created when misaligned magnets move towards alignment starts low and ends high. Compressed springs and stretched elastic seem to provide the opposite: strongest force initially that decreases to zero. My thinking is that, because the rail gun pushes the hardest at the end of the stroke, at the moment of contact with the cue ball, at the moment when the coil magnetic field aligns with the cue stick magnetic field, the machine would have a better strike on the cue ball... maybe? It feels like a good concept.

Additionally, if I reverse the polarity of the coils as the cue strikes the cue ball, the coils will then repel the cue stick, extending the cue through the cue ball, effectively "following through." Optional follow through feels like a tool that could help the machine compete at a human level.

A pneumatic solution was feasibility tested. I lost the video of the system operating with the cue stick attached. It needed a LOT of pressure. The on board pump to generate pressure to propel the cue stick was super loud. I'm sure that could be overcome, but one of the goals of the project is to have humans enjoy playing against the machine and I went towards the rail gun before investing energy into silencing the pump or getting the pressure otherwise.

 

[kellys_eye]

Your emphasis on the 'power stroke' is causing you problems that you don't need to (immediately) address.
When watching professional pool players take a shot the vast majority are done for 'placement' (to take the next shot) rather than 'effect' and such power shots are infrequently used. Indeed I've won more pool games using slow and careful play than those using hard/fast shots that only achieve the scattering of all the balls!

For the (greater than 90%?) of shots that DON'T need 'power' you could concentrate on the shots that win the games instead.

If you were talking SNOOKER (a much longer table) then power becomes a greater issue but on a small table it's less important.

 

[BVarv]

I appreciate your stance. In my time around pool tables, I've heard many times: "There are two types of pool shots, soft and softer." I agree and I don't feel my need to access break-shot-strength implies I'm emphasizing any type of shot over another.

Regarding the rail gun thruster idea creating "problems that I don't need." This makes me chuckle. Making a machine that plays pool is, in itself, a collection of problems that nobody needs.

I'm going to order components. Thanks for your insights.

 

[kellys_eye]

It's an interesting project. Whilst I've seen end result videos of such pool playing machines I've never followed the concept from start to finish so if you're prepared to do so, could you keep the thread updated on progress?

Good luck with it, whatever you decide <thumbs up icon that doesn't exist>

 

[BVarv]

Thanks. I'm happy to post updates. It seems my initial post/title limits the topic of discussion on this thread to the thrusting mechanism only. Do I need to make a new thread to discuss broader topics of the project, or can I just expand here and hope that people with the technical expertise I seek (on future questions/hurdles) stumble across the thread?

For a quick overview of progress in other areas of the project, here is documentation I submitted to a robotics contest:
https://www.instructables.com/contest/robotics2017/

Edit: https://www.instructables.com/id/Pool-Playing-Robot/
Whoops, the first link is to the contest results, here is the project link.

Thank you again for reading

 

[kellys_eye]

Thanks for those links - I'm a fan of the Instructables site and often in awe of the capabilities of the members!

I'll be looking forward to seeing progress.

 

[davenn]

Thanks for those links - I'm a fan of the Instructables site and often in awe of the capabilities of the members!

yeah there is some good stuff there

as we have discovered on the forums, there also lots of crap that the various electronics forums have to try and sort out for people when their projects don't work as expected