Robot Suspension System Design

[theonelinerkid]

those wheels would be brilliant for my friends diy scooter he made from boxes and wood cos you dont need srpings cos theyre in the wheel

[TheChickenLoop]

Tires achieve a nearly identical result. Whether you can argue the benefits of this design is moot - neither tires nor this design provide damping to the degree that is necessary to achieve an effective suspension system. With no way to extract the energy from an impact, a bump can send this design airborne, unlike a traditional suspension. Keep going, though - I'm sure you can incorporate an equally simple solution to achieve damping.

[kubarebo]

@TheChickenLoop This design does not provide damping because it was designed with no damping at all. It can be easily modified to add damping, and that's what I was saying all along :) So we agree.

[palmisano]

@TheChickenLoop It's a suspension system to help keep all four wheels on the ground when on uneven terrain . . . it's not for high speed damping . . .

[RustyNex]

its a nice idea but the reason this isn't in cars and bicycles etc (because i had a similar idea for a bicycle wheel) is because of torque, for larger equipment with more weight and a powerful drive system, this design becomes a torque absorber and ends up eating the rotary force that you are tying to apply to the wheel.

[kubarebo]

@RustyNex It takes a reasonably simple modification to this design to decouple the torque transmission from the thin suspension-providing sections. I'll leave it as an exercise to the reader. Hint: it can be done using very simple kinematics -- no need for gears or anything of that sort. Besides, it won't "eat" infinite amount of rotary force. You'll twist it until you stiffen it up enough to resist further twisting. At that point it's not good at being soft suspension anymore, of course.

[kubarebo]

@kubarebo Moreover, it takes a reasonably simple modification as well to provide for damping: fill some unused space with shaped memory foam.

[RustyNex]

@kubarebo yes, and that works great in a small scale like this, but i mean in lager scale it just wouldn't work, there would not be a simple way to absorb the rebound of the spring effect and dampen it, and the longer you make the "spokes" to make them more springy, the bigger problems you have with horizontal drift outside the hub. imagine this on a car and all the stresses it goes through. i am not saying its a bad idea, just pointing out why is not used on a larger scale.

[RustyNex]

just to add, i come from a car based background so i understand a lot of forces involved with a car and the most important thing to think about (when making a high performance car) is rotational mass. you want the wheels as light as possible. any part of the car that has to be rotated by the engine must be as light as possible, (alloy flywheel, magnesium wheels, carbon drive shafts etc) to put all this weight into the wheels will not only slow the car down, but would be a nightmare to balance.

[kubarebo]

@RustyNex Rotational mass is not an issue. Overall mass is an issue in cars. The real issue is really unsprung mass -- that's why the wheels have to be light, not because they rotate, but because the heavier they are, the poorer the car's response to shocks. Since this is a composite wheel, not all of it is unsprung mass. The parts that are rigidly coupled to the shaft are stuck to the body of the car and can be as heavy as you wish, pretty much.

[RustyNex]

@kubarebo of course rotational mass is an issue, the more weight the engine has to rotate the more torque it needs. Rotational mass, over all mass and unsprung mass, all dictate how the car performs, but few components are in all 3 areas, wheels are one of them. But you are right that wheel weight is more of an issue to how the suspension performs, but its also an issue to acceleration, anyway in my experience with mechanics, this design wouldn't be effective on a larger scale with larger forces

[kubarebo]

@RustyNex The moment of inertia of wheels on my car adds about 200kg of equivalent weight to the car per m_eq=I*(1/r)^2, so it's not as horrible as it'd sound. I don't feel much difference in acceleration with 3 extra adults in the car... I don't exactly do WOT every 20 minutes. Well, maybe every 40 minutes ;)

[RustyNex]

@kubarebo lol well it sounds like you have a beast of a car, because my car is very much affected by extra weight, i bought some alloy wheels of the same diameter as the original steel wheels but a lot heavier and it slowed the car right down, add extra passenger and its very noticeable. so maybe engine torque is what it comes down to, i know bikes with low engine torque but high RPM, weight makes a huge difference.

[Gig540]

I think its perfect for your application. Very simple not complex with parts and no need to worry about centrifugal force keeping out of round. Well I assume it wont be going to fast. lol

[henrik2k]

all u have done is replace soft tires with soft rims..... an idea and technology developed long ago by the car industry

[kubarebo]

@henrik2k Nope. This approach, when designed to do so, can replace soft tires, suspension springs, and shock absorbers. All in a composite wheel.

[chinico68]

Very clever...

[spam1712]

i allwys wanted to build robots...

plz give audio name

[blundellp]

bored. Could of atleast drove round

[gorrammit]

Both the flexing arm and flexible spoke wheel suspensions have long been around in RC airplanes. You have thin flexible metal wire holding the wheels and some wheels with hub designs similar to yours to absorb landing impact.

It's great, though, to see you've done some optimizing for your designs

[cormatches]

awesome!!!

[pzpepsi]

im sorry but i expected something more impressive. the wheel-suspension is pretty good. does its job and saves weight but this is like the common sense solution. especially with the the front setup. thats the first idea i ever had making a suspension, but i made it front thin steel. it worked good on lego cars for awhile back when i was 6

[pomeroy600287]

I like the front wheels that can roll sideways as well. I suppose castors would be too innacurate?

[palmisano]

@pomeroy600287 They are called 'omni-wheels' =P

[madybear72]

I am gonna make it easy for you " don't make big sentences"

[SvenOkonomi]

Those designs might be a little concerning when it comes to lateral bend though. I have a feeling the center hub would pop out of alignment too easily when reasonable load is applied or when some sort of sideways motion is forced upon the unit like an uneven surface. The hook prong omniwheel suspenders are quite good though, you might even find its latteral bend could work WITH your design rather then against it. Interesting work! Thanks for sharing

[palmisano]

@SvenOkonomi Yeap I also analyzed lateral bend, even though you don't see it in the video. You can limit the bend by selecting appropriate material thickness.

[FutureInventions]

An interesting idea! I would've never thought of putting suspension in a wheel.

[urstupidipshit]

very nice!!!! BUT did you know that the mars rovers by nasa used these kind of suspension system????

[palmisano]

@urstupidipshit Not quite. The Mars rover suspension is specifically meant to protect the robot from high frequency high force shock during the initial airbag bouncing phase. Its not used for terrain suspension, hence the rocker-bogie design.

[urstupidipshit]

@palmisano actually, right before the rover lands, it triggers large airbags that completely cover the rover. and wheel suspensions are no help when its landing. If nasa designed for the wheel suspension for the landing, there is a very high chance for the rover to break due to such a force

[palmisano]

@urstupidipshit I went to a NASA talk that specifically claimed what I said earlier. I just googled it, and found a contradiction, "The futuristic-looking "hubcaps" were chosen over dozens of other flecture and spoke options and are designed to absorb shock and to protect the rest of the vehicle during driving."

[wilddice77]

could this in some way be fitted on a go kart or anything bigger ???

[palmisano]

@wilddice77 Yeap. A car tire manufacture is already experimenting with this concept, but can't remember the name off-hand.

[urstupidipshit]

@palmisano but its rlly gonna look weird on cars... it looks goon on robots

[LIKUS007]

Great work, thanks for sharing!

[kght222]

pretty cool application for an old idea (look around real close and you will find devices that use a similar concept for vibration control)

[gonepishing]

looks a tad stiff on the back wheels

[wwwKx]

Nice idea!! 5* :D

PS: what is the software used in 0:30?

[palmisano]

ANSYS

[palmisano]

ANSYS

[gizmo2477]

reminds me of tweels =D

[Jesus45U]

Interesting idea.

But it has limited usage.

Any usage that puts high sideways stress on the wheel, would not be appliable, since it would flex too much.

Still, very nice idea. :)

[palmisano]

If high sideways stress was expected, it can easily be designed around by making the wheel thicker. =P

[Jesus45U]

True enough, up to a point.

A train for example, would never work with this. Nor motorcycles or Cars.

Luckily, most robots are slow, and therefore make little sideways flex.

Another issue, would be shockabsorbtion on this wheel ... it has none. Which again, is fine as long as you don't need it.

Which is the case in a slow mowing environment.

Just look at tractors, they have no suspension either, apart from the tires, and they do allright, as long as they move slowly. :)

[palmisano]

Actually, there is a new car tire that works similarly.

As for shock absorption, it *does* absorb shock. The question is how much shock, and this particular design won't do very well for large translations/forces at high speed.

[Jesus45U]

Ah, you must be thinking about the Tweel.

It has it's own problems.

Particularly at high speed or particularly rough surfaces, which kinda takes us back to what I think we both agree on .. it has limitations, but should be good for most robots.

As for shockabsorbtion, I would call the absorbtion by material to be so low, as to be almost nonexistant.

The real question then really becomes, does it fit this application ?

And it probably does.

As long as there isn't too much force.

[suprTOONCES]

What about torques? Would it be a suitable torsional damper maybe if it was more resilient?

[palmisano]

More resilient?

It can be designed to dampen any mechanical force desired.

[kubarebo]

@palmisano I don't think you'll be able to get optimal shock absorption with just a single monolithic material. It'd need to be a composite -- decouple torque transmission from suspension from shock absorption. Perhaps you'd be able to use nonlinearities in the material, but then you'd need a mucho multiscale design -- with small scale structure doing damping, larger scale structure doing suspension spring, and a yet another class/size of strucure constraining torque transmission.

[kubarebo]

@Jesus45U This design has a rather wide margin for optimization. The biggest problem is manufacturability and quality of materials available I'd think. There's nothing special about "slow" speed of robots. You rescale a couple of things and slow becomes "fast". It's only for "slow" uses when you are using this particular material and shape.

[Jesus45U]

@kubarebo

Whoa .. a blast from the past, someone reply to a 2 year old post. :)

Anyways, the problem with speed, is that it doesn't scale liniearly.

Old Swordfighters found this out by experimentation.

If you swung a sword, double the weight, and you double the impact.

Double the speed however, and you quadruple the impact - hence why most swords weigh only 3-4 pounds.

This means that you need quite a lot of dampening at higher speeds, and that becomes a problem in designs like this.

[kubarebo]

@Jesus45U The active volume of a hydraulic damper in a typical passenger car suspension is on the order of 10 cubic inches. This "quite a lot of dampening" is very compact. You make it seem like a much bigger deal than it really is :)

[Jesus45U]

@kubarebo

The volume of fluid in a shock absorber is less important than the viscosity of the fluid and the size of the gap/holes in the piston that moves trough the fluid. Hence how they can use magnetically enhanched newtonian fluids to enhanche shock absorbing without adding more volume.

And yes, it's a big deal - Try and take off the shockabsorbers (keep the springs offcourse) of a car and drive down a road with potholes. At 10 mph, no problem, at 60mph, you will be all over the place.

[Omnigeek6]

That flexible wheel is freaking genius!

[OmegaDude22]

Oh wow I love simple and functional design like that! Great job!

[jsz1998]

how do you make this and please tell me in a way i can understood (i'm a kid and don'tknow muchabout this kinda stuff)

[palmisano]

see info box.

[TopGunMan]

who the !@#$ has a cnc mini mill god damn so jealous. I swear ide be rocking this boat if i got enough resources for one.

[palmisano]

Don't be too jealous, I currently don't have access to one either. Or any form of modern equipment actually. Living in Bangkok kinda sucks like that . . .

[80amnesia]

nice homepage man

[palmisano]

Although the Mars Rover wheels inspired me, the design and purpose is different. The Mars design is for high frequency shock protection, while my design is for suspension and conforming to rough terrain. Anyone can take my idea, just give me credit with a link back =P

[ptraider2002]

Very cool website, congratulations u r the best! can't wait to build my first robot! the Mars rovers have similar design for the suspension, is this where u got the idea from? or was this your idea? anyway can anyone use this design? or is it like protected?

[RobotGrrl91]

I remember seeing this video... and I thought of it when I saw the wheels on the MSL. They're quite similar- is there more of a story to this suspension system than meets the eye? ;)

[ccc1123]

cool man..... hey, what is society of robots???

[98989898787878]

its his amazing website that teachs ppl like myself how to make robots.

it has lots of his robots, a robot forum, user sumitted robots, and much more

[ExodusVI]

Woah those suspention systems are quite cool i'd love to build a robot but haven't quite got he time or the materials at the moment

[lemuelbecc]

You're a clever man.