legacycentral bbs

For an engineering project, we need to move a cart down a track at a constant velocity (no slower than 1.2ft/sec).

First, I thought of going simple, using gears and having a smaller gear spinning opposite the larger gear. Might not work, don't feel like trying.

Then I thought slightly more high-tech and using a vicious coupling. I could make one myself, and I figure it'd work well. It's gotta travel down a ramp 3ft high and 6ft long. Heat wouldn't be that great, and I could get it to work.

But in the instructions, he said electricity (motor or the like) could not be used to move the cart down the track. I don't think he'd have a problem with electrics controlling brakes on free-rotating wheels.

So I'd need to program it, and that's where I am unsure. We can use the Lego Mindstorm to control things (coded in basic).

Should I just do it so that I use a light sensor where pulses/sec = X, and 1.2ft/sec = Y. Should I just set a switch to activate (apply brakes) when X>Y, and set the switch to deactivate (release brakes) when X<Y?

Is it that simple, or would there be a better way to program it?

I talked to my professor, and he said that I COULD do it this way if I wanted. My group was impressed, and drawing it out, I could see how it would work well. His concern was running in the dark. Would it still perform as predicted should the lights be turned out. I was sure it would, since the sensor for the lego kit emits a light, and that would be simple enough.

I could probably make a magnetic sensor though, and simply tie it into the systems interface. It's a pretty easy setup to use.

But he said that if there was a more mechanical way of doing it, that would be better.

So what I intend to do, is build a viscous coupling using CAD and a 3-D printer. But in order for a viscous coupling to work, I'd have to design the layout. Since there is no drivetrain, I need to impliment it inline with an axle.

I think I could just build it from the inside out, and have one solid axle with finned teeth on it, and a 2 peice case that I could seal up and rigidly mount to the body of the cart. That outta work out alright. Plus, since I would need to design it in CAD and print it on our 3-D printer, that would knock out a requirement (we have to render in 3-d our whole project, so doing this would mean it would be a good portion of our prototype would be drawn up in CAD already.

I could have the case tapped for 2 grease fittings so that I could change fluids to experiment with viscosity. Only problem is, would it really be constant velocity? Would the momentum of the cart build and build and overcome the braking force of the fluid, thus slowly accelerating the vehicle?

Points are awarded for ingeniuity and simplicity. The means to the ends don't have to be simple, but it must appear to be simple. A box mounted with an axle running through it, would appear quite simple, but in reality it'd be more than that and thus extremely ingenius.

So many options.

Gravity is accelerating it a constant rate. So you have to have something decelerate it at an equal rate. So the resistance has to increase from what i can figure.

The good thing is you can accurately calculate the gravity part of the equation.

Yes, but since mass is low, and the slope of the ramp is so shallow, the acceleration due to gravity isn't going to be that significant.

That's kind of on my side. A 2 pound vehicle going down a ramp with a .5 slope won't take much to overcome the acceleration due to gravity.

However, can a constant resistance work "well enough" to slow the vehicle?

A spring couldn't be used, nor could a rubber band, because all of those would increase resistance beyond the rate of acceleration.

But with fluid and the mass I am given, I think I have a good chance of being able to limit the velocity of the car. I think.

I'm no fluids engineer, so I'd like the input of people who know more than I do.

I was also toying with electro-magnetic, but that quickly flew out the door when I found that we couldn't design the track any way we wanted. I was going to use a rail-gun type system to launch the golf-ball. Might go further than 15" though.

How lame is this project? A friend of mine is taking the same course at a different campus and gets to make a tennis ball launcher. 3 tests. Max height, max distance, and land the ball in a trashcan 50 feet away. Why couldn't I do THAT!? Much more fun than a 1.2ft/sec cart!

This is a more interesting task than a tennis ball launcher.

The acceleration due to gravity is very significant, Phil. Realize that the only two forces acting on the system, aside from anything you introduce, are gravity and friction.

Your idea of an electronic feedback control is an interesting one. I would suggest you consider looking at not only the velocity but the derivative of velocity (that is, acceleration) so that you can start braking before you actually reach the desired velocity, to help avoid overshooting.

But may I propose another approach? You know centrifugal clutches? And centrifugal advance distributors? What about making a centrifugal brake?

You could have a spring with one end at the axis of a rotating assembly. The other end of the spring would have a mass attached to it. You'd somehow set it up so that when the spring stretched the right amount, the mass would somehow inhibit further rotation.

You'd probably use at least two springs and weights to keep it from being rotationally imbalanced.

That idea would be easier to adjust during testing to get the desired result. Changing masses on the springs or the springs themselves would give more adjustability than whatever fluids you have available.

I was thinking of a very simple design. imagine an axle with a tiny pinion gear on it. you could then have a large sun (spur) gear on a shaft in a way that you could raise or reduce the amount of friction on that gear. you could then set the friction to your desired speed (with test runs) and then it should stay constant. imagine driving downhill in 1st gear and not touching the gas. just my .02

Some things to consider: friction clutches, flywheels, (Vikash already mentioned centrifugal clutch). Can you use an exterior force? Something like an arrestor hook grabbing a line with a controlled resistance? I think it's an interesting project. My favorite at school was the elastic powered car. I had a three wheeled Morgan aero trike-like thing with the elastics turning a 'power' axle connected to a 1:3 gearing running the fwd axle. Points were for distance, speed, and aesthetics. I came in 3rd due to too much torque necessitating a last-minute design change that made the vehicle too unstable to clinch 2nd. If only I'd had the time to instal a wider axle...

The entire assembly must be no greater than 2lbs going down a ~6.5 foot long ramp @ ~45 degrees. I guess I could figure out the acceleration and top un-assisted speed of the vehicle (I think aerodynamics are almost insignificant).

The centrifugal clutch is a good idea. The car has to be capable of making 1000 passes without breaking or malfunctioning, so it should meet that requirement. For the centrifugal clutch, I could probably even CAD that and use our 3d printer for it. It prints in a sort of ceramic type of material, which is kind of cool. At the very least, I could use it to make the "brake pads" for the ends of the springs.

The problem I see with the centrifugal brake idea, is that when it reaches that point, the pad is going to be contacting the friction surface. Will it be durable enough to withstand 1000 passes? The way he is measuring that, is whether or not there is any waste material after 10 or 20 passes. If there is friction material deposited on the cart, it won't get the durability win. I'd use metal, but that would cause heat, and it's supposed to be a toy for a toy company, so children need to be considered. Kids touching hot things is a liability. Significant heat? I dunno. Maybe not.

How would you program, in extreme basic, motor operation based on the derivation of acceleration?

The lego system uses commands like "motor speed 1" and stuff. It is totally incapable of doing calculations on-the-fly. It requires hard input data, and can't really store that much information.

So far, I'm thinking the fluid damper would work pretty well. I had a fairly good sketch of what to do, and making it would be easy with the printer.

I like the centrifugal clutch idea, espically because it wouldn't require resetting when moving the car back to the top of the track. Same with the fluid and the electro-brake system.

I'd like to know more about fluid dynamics so I would have an idea of where to start with the fluid for the VC idea... if that were the method I chose to persue.

evolutionmovement wrote:Some things to consider: friction clutches, flywheels, (Vikash already mentioned centrifugal clutch). Can you use an exterior force? Something like an arrestor hook grabbing a line with a controlled resistance? I think it's an interesting project. My favorite at school was the elastic powered car. I had a three wheeled Morgan aero trike-like thing with the elastics turning a 'power' axle connected to a 1:3 gearing running the fwd axle. Points were for distance, speed, and aesthetics. I came in 3rd due to too much torque necessitating a last-minute design change that made the vehicle too unstable to clinch 2nd. If only I'd had the time to instal a wider axle...

We built composite bridges in high school for an engineering class. I still hold the record. 3lbs and supporting over 800lbs of force. If I get beaten, I'll do the challenge over to reclaim my title.

Hmm, yeah, a thousand passes is a lot. I don't know how you'd build that frictional interface.

Now how does this work? Do you somehow launch it at the desired velocity, and then your mechanism is required to maintain that velocity? Or do you actually accelerate from a standstill to the desired velocity and then hold it?

The viscous fluid idea -- you mean to use a fluid that thickens as it heats up, with some rotating fins going through it, so it effectively increases resistance as velocity increases? Seems like it might be hard to calibrate... Although I suppose you could just build it once and then use gearing to select the velocity.

Hmm... maybe you could intentionally make wind resistance matter. Make a sail that slows it down. At low speeds, drag is roughly proportional to velocity, so you could get the kind of feedback you want. No moving parts (aside from the axles and wheels of course) would be nice.

I don't know anything about Mindstorms or extreme BASIC or whatever... but if you have the means to measure time, you can estimate acceleration if you can record the times of the last three pulses.

Say the three pulses occured at T_a, T_b, and T_c. Then the time span (T_c - T_b) would be inversely proportional to velocity, right?

Well, the difference between the last two time spans would be proportional to acceleration. This is (T_c - T_b) - (T_b - T_a) = T_c - 2 * T_b + T_a

You've gotta be able to do simple arithmetic.

It's been a long time since I played with Legos or the Mindstorms setup, so I'd have to read the instructions on the limitations of just what the software is capable of.

I know there is a light sensor that can count pulses per second (though I can't recall the limit), and I know that it can tick off seconds (i.e. wait 3 seconds).

I seem to recall programming the Mindstorms to be similar to programming basic games into a Ti-83. But with the Ti-83, you could program it to set a pace, then have it count time as you so like (start at zero, count to Z integer, and on input store the T value and display. T1 = 30. T2 = 60. From there you could extend and manipulate as you so desire). I don't recall if the Mindstorms has the computing ability of a Ti-83 though.

If you are referring to ME being able to do simple arithmetic, I'd like to think I'm somewhat proficient.

As far as the launching. Place car at top, it must accelerate and then cease to accelerate at/around 1.2ft/sec.

It's not so much a viscous coupler in the traditional sense, as that's where I originally got my idea. My idea is more like a soup can. Imagine a full soup can, like Tomato soup. Stick a spatula in there and try to spin it. More a fluid-resistance device than a viscous coupler. Espically since the way I designed it, there is no coupling anything, and it doesn't rely on the heating of a fluid to increase resistance. It solely relies on the viscosity of the fluid and a fan/shroud idea to slow the axle.

*sigh* it's been a long day.

Hehe, no, I meant that you have got to be able to do arithmetic with the Mindstorms.

It sounds to me like your soup can idea is the same as my sail idea except that we're talking about different fluids.

Ahh, the sail idea... I like that one! Paint a skull and cross bones on it and put a sponge bob sticker on it. Kids love that kind of stuff. I'll run that one by the group.

They sound exactly similar. With my soup-can, I'd be changing the fluid. With the sail, I'd be changing the spatula.

Hehe, yeah.

With the sail you're using a fluid that's free.

You could calibrate it by using a huge sail and them trimming it down little by little until it seems to give you what you want.

You might also need a tailfin to act as a rudder, though, to keep it going straight.

Or I could cut slits in the sail .

The sides of the track have a 1/2" tall rail to help keep it straight, so I don't know if a rudder would be necessary.

We may build several different arresting systems and have our project be modular (using pins). I think we are going to produce our prototype completely through 3-D printing, so that we can get perfect prototype's of what we are trying to do.

So far, I've come up with all the ideas, so I'm leaving it up to my groupmates to get the ball into the bucket. I refuse to do all the work on this project. But I'm "Team Captian" so I'll probably end up doing that.


In hindsight, I should have said that I'm changing the soup, not the fluid. Soup sounds better, don't you think.

You guys are so lucky to have a rapid prototyping system at your disposal.

But air isn't soup.

But Soup is a fluid. Like I said, it was late. I was revising an english paper at the same time, so my attention was drawn elsewhere.

It's not so much a rapid prototyping system, it's just that I'm pretty good with the rendering software so I can quickly produce a working model in 3D. We are definately lucky to have a 3 dimensional printer, that produces prints that have some stability to them. It isn't designed to print production material, but that's how I intend to use it. I wish we had a CNC mill. That'd make things convenient!

yes, a CNC mill would make things MIGHTY convenient.. I tell myself this every day

The 3D printer is what I meant by "rapid prototyping system." Is that not an accurate term? Does a rapid prototyping system include other things?

I have no idea, lol. I suppose "rapid prototyping system" is an appropriate word. But I won't only be using it to rapidly build a prototype, but also to build the final product.

I am drawing my fluid-resistor in 3d right now. I'll post pictures when I'm done.

I wanted to get a rough idea and sketch of what I am thinking of.

vrg3 wrote:The 3D printer is what I meant by "rapid prototyping system." Is that not an accurate term? Does a rapid prototyping system include other things?

I'm pretty sure that's an accurate term. I've heard them used interchangeably.
Maybe the rapid prototyping systems that remove material rather than adding it are not referred to as 3D printers?

BAC5.2 wrote:I'd use metal, but that would cause heat, and it's supposed to be a toy for a toy company, so children need to be considered. Kids touching hot things is a liability. Significant heat? I dunno. Maybe not.

Remember Easy Bake Oven? Heat is fine as long as it's inside the toy. Creepy Crawlers, on the other hand, had exposed hot metal molds kids could move around freely.
I'd be more concerned about the small parts like wheels passing the 'can a kid swallow this?' plastic tube test. I don't know if that aspect of your project is that important.

So here is what I came up with. Took about 90 minutes or so of actual work.

It's made up of 4 (really 5) main parts. The wheel/axle:



The finned sleeve:



The housing that the sleeve sits inside:



And a bearing/seal on either end of the system. I didn't think to take a picture of that, it's just a cylinder.

Put it all together:



I made the one bearing/seal glass so you could see inside.



And different, but similar view:



There would be a hole drilled and tapped in the top of the housing. I didn't include that in my drawing, but it would be there. There would be a grease fitting screwed into this, so that fluid could be changed easily. There would be a blank plug on the opposite side, just unscrew and force the fluid out.

What do you think? Feasable? Pointless? INPUT!

Looks neat.

So there's gonna be another axle with a bearing attached rigidly to the housing, to make sure the housing doesn't rotate with respect to the ramp?

How are you going to tune the resistance? Adjusting the blend between 140-weight gear oil and 5-weight motor oil or something?