Shock mounting motors

[fluppet]

Just to keep two generic threads on the go at once (and hoping this is of use to more than me)...

Shock mounting things is a good idea, in a modern day and age. I can see how to insulate my electronics from someone folding my frame in half, and I know about mounting the gearbox as a sub-frame so that when the outer frame gets a whack the drive keeps going. I know about Bigger Brothers two-level frame, and Ive been the instigator of a long rant about whether frames should be designed to bend or not.

I know about soldering the end cap on to motors, and using something sturdy to hold the magnets in place. I certainly know not to leave anything other than a very short bit of axle cantilevered, and obviously not to make the motor axle take the weight. I even know to put enough slip in the wheels that the motor cant stall (unless Ive got current limiting, at least), although how your gearbox is supposed to survive Growler grabbing hold of you, putting his weight on the robot, and spinning the wheels quickly is somewhat lost on me (Im surprised there werent more casualties).

I can just about suss mounting an axle with suspension and a chain drive with a spring-mounted tensioner so that the axle can move a bit relative to the motor - although once youre into long chains with active movement it seems to be mostly a matter of time before something goes pear shaped. And Im aware of the suspension wonders of pneumatic tyres (with a foam filling, as WJ demonstrated), and the benefits of leaving a bit of slack around everything should your frame get bent (demonstrators: George Francis and Andrew Marchant, several times).

So, question: how do you shock mount a drive train? Its not clear to me how one big whack on an axle doesnt throw everything out of step (although an enormous axle obviously helps). My moral objections to the Drop Zone were mostly based on it giving the drive train a whack from a direction it wasnt designed to support, and possibly doing something expensive as a result. Lots of robots fail due to something happening to their drive, some less than others, and Im sure people have given it a lot of thought. Rather than reinvent the wheel (as it were, unless thats the solution), I thought Id ask. :-)

To clarify the question: how do you allow movement in the wheel in all the directions it could get shocked without that shock being transmitted to something delicate and expensive (like a motor)? How do you handle the axle being bent without the drive train being compromised? Other than trying to make the thing so solid that nothing can happen to it, which isnt an approach I really believe in, how do you transmit torque with as many degrees of freedom as possible? How do you stop excess torque (Growler) from knackering the drive?

Am I being overly paranoid here? I can come up with something bizarre with CV joints, rubber shock absorbers (as Diesector uses for the hammers) and some form of slip-release, but Ive seen no sign of any such thing on existing robots. Perhaps I should build it and see whether anyone can break it. :-) (Of course, itll weigh so much than I can only put 1mm of polycarb over the top, but axes are getting rare these days...)

I do have one or two other ideas, but theyre based on odd robot shapes and arent all that useful in the general case. Anyone care to share their techniques? (Gary - this pretty standard shock mounting of yours? Ed - or do you just build everything out of 20mm Ti?)

Thanks everyone for their informative answers to by general questions in the past. Here we go again. :-)

--
Fluppet

[Mouldy]

I believe Storm 2 and Tornado both use fixed axles and the wheels have pressed bearings inside and the wheels turn around the solid axle. This is mainly a good idea as you can run (to a certain extent) with a bent axle and keep going.

I believe razor have 2 parts to there rear axles. They have either side supported and the centre of the axle is cut in 2 so it can flex....Or something like that. I didn€™t quite understand when James cooper was explaining it all to me, best to get in contact with Ian Lewis on that.

If I understood your questions I think the 2 options above are good ways of mounting your wheels and axles.

Regards
Dave moulds
Team Turbine/PLF

[jim]

I use a very simple method of suspension. I have 20mm silver steel axles sitting in needle roller bearings on each side of the wheel. The outer diameter of these bearings is about 26mm, which is perfect for sliding on 2 layers of mountain bike inner tube. This then sits inside a 28mm hole in the axle mount. They dont come out as easily as you might think. Theres not a lot of travel in there but it does help a bit.

[fwsorg]

My rather uniformed input:

A problem I suffer from is trying to consider and compensate/design for every possible eventuality. The problem is, I never get anything built as a result :-)

I reckon a decent pneumatic tyre will do the trick. It absorbs shocks in most directions, and as long as its well mounted (fat axle supported on both sides) it will probably do the trick. The best trick with robot build is KISS- or at least, if making ti complicated, allow it to run still when all the complicated bits fail. If you bear that in mind, you cant do wrong.

I have yet to see (tho it may have happened at a live event or similar) Storm 2 take a massive vertical blow- ie a very hard landing from a great hight, or something like beta coming at it from above. Its front 4 axles are 7075 according the their website, which ordinarily I might be a little dubious about, especially with 7075s preference to breaking instead of bending/flexing, but seeing as their are 4 of them and a further 2 steel axles, they get away with it. Would be interesting to see what beta makes of storm 2- whilst the ti would hold up (although I believe its thinner on the top) I certainly wouldnt subject a 3mm walled box section chassis to beta and it expect it to come out in any usable shape. Oh well.

Slightly more on subject, if you mounto your motor and gearbox assembly to the chassis and stick a length of non-slip rubber mat or similar, thatll help absorb shocks surprisingly well. Certainly all my armour panels have a sheet on non-slip between them and the chassis- it makes a huge difference to the knocks the robot can take.

[fluppet]

Fixed axles mean that your wheel wont wobble when its bent, although obviously it will be running at an angle; its true that this probably helps, but Wild Thing ran with a very bent axle through a Hypno-Disk encounter at one point. If the axle bends and the wheel is off true, you have the same problems with the sprocket attached to the wheel as I mentioned before: the distance from the drive sprocket to the wheel sprocket might have changed, meaning you either need slack and a tensioner or everything is going to shift; the angle of the wheel might be enough to move the chain beyond tolerance so it jams or gets thrown; rotational wheel impact still gets transmitted to the drive (although the tensioner might take some of it).

It does make it easier to make the wheel mount strong, though.

On a related note, Ive not seen many robots which bottom out when impacted from above - strikes me as one way to protect the drive train (the frame takes the impact). Obviously you have to ensure that you dont bottom out as soon as another robot is sitting on you, but the tolerances are reasonable. Perhaps Ive just not seen it because suspension is rare.

As for Razer... interesting. Ian and Simon dont tend to post as much as wed like, but if James is reading this, can you elaborate? Razer essentially has no give in the wheels, of course (plus the wheels arent round), so some form of suspension would make a big difference to them.

Whoah - and Jim and Eddy post since I started!

Thanks for the feedback, guys. Ill invest in rubber. :-) I was thinking in terms of shock-mounted armour panels (like SMIDSY) anyway, although Im not sure to what extent that weakens their attachment if something digs in and tugs (spinner blades, etc.)

Anyone doing anything more substantial than a thin layer of rubber? Many people can testify that its not always enough. :-)

--
Fluppet

[storm]

Eddy - 6mm Titanium on the top as well Also, the main drive axles at the back arent Steel - theyre 20mm Grade 5 Ti.

The Vulcalon tyres take out a lot of the shock, the axles are 20mm thick, and there are 6 of them. The unsupported distance is only 70mm. Bending 6 x 70mm long 20mm thick pieces of material is a tough job.

Seems to be holding up ok. Stripped down the drivetrain last weekend to check it over, and nothing bent anywhere

Ed
http://www.teamstorm.comhttp://www.teamstorm.com

[fluppet]

Cheers, Ed.

In Storms case Im less worried about the axles than I am about the whole thing shaking itself apart. Since you seem to rely partly on your robots weight as a hammer (as opposed to smacking the opponent hard into a wall and having a relatively cushioned landing yourself) Id have thought the drive train would take a heck of a whack from its own weight and the sudden shock load as the wheel speed changes.

Have you done anything to the LEMs to strengthen them (Ti axle?) Im not sure how thick the axle is on them anyway. Any tricks on stopping your gears from throwing teeth/clashing together on impact? Are they insulated from the frame at all, or do you just rely on it being built very solidly?

Cheers,

--
Fluppet