Gravity Vehicle C

[illusionist]

I'd love to see a video of that launching and braking system at the end of this year!

[iwonder]

Thanks for the help, I will definitely keep those tips in mind while making plans for next year, and I agree, I would love to see that launch and braking system at the end of the year... it sounds like it took a lot of work to make it happen.

[twototwenty]

I concur; a system like that sounds amazing. How much did it end up costing?

[Balsa Man]

I will definitely try to get video, and share it- it's pretty cool to watch; State is tomorrow.

Cost?- let’s see. If you had to go out and buy everything, right at $200- carbon fiber- $35, Bearings- $30, wheels & tires- $25, titanium front axle- $10, Epoxy- $15, Rubberized CA glue- $7, Lexan & Plexi- $10 (bought as scrap pieces fm local plastics company), Balsa- $10, nylon 3/8 threaded rod- $10, gears- $10, T6061-T6 aluminum- $10 (w/ a lot left over), rod & tube for gravity supercharger (stainless steel rods, brass tube that slip-fits over rod)-$10, miscellaneous hardware (screws, nuts, bushings)- $15, a (2+kg) chunk of lead- $5 (fm a metal recycler). Actual cost to us was more like $150, because we had things around from other events/previous years.

Bringing this together was one of the neater/better “design and implement exercises” I’ve been involved in over the years- the guys learned a lot. Good up-front analysis of what scored more points, the factors going in maximizing speed & precision, analysis of materials & properties, constructability analysis (how to get needed level of precision w/ tools available), and adjustability analysis- what things needed to be designed/built to be adjustable, and how to do that. Some really good proof-of-concept/prototyping work early-on, with good design changes and evolution from that, so that when the actual build was done, it…..worked- it was linear, and they didn’t have to circle back and make significant changes.

As one example, the gravity supercharger went through a very interesting evolution. First thing last fall, we put together a half scale, simple test chassis; ¼” plywood chassis plate, disc wheels on bushings. Used some chunks of steel bar for getting the weight up. Basic physics analysis said we wanted as much of the mass starting as high above the floor as possible. Found that with the mass stacked over the back axle, it didn’t want to roll straight at all- major wandering; with it in the middle, though, it did roll straight.

So, I asked the question, “can you think of any way to have the best of both worlds?”- have the mass starts high/to the back, and end up in the middle??. The first answer was a “trapeze.” Two vertical bars up from midpoint with a rod between them at the top; two arms down from that rod, with the steel bar (with a length a bit less than chassis width) hanging on them- pull the steel bar back/up, let it go, would swing down to the middle- put a stop block in to stop it at the middle. Realized two things; that going from around 200gr of moving mass to around 2kg (in final/actual vehicle) would present all sorts of problems in how to get the trapeze strong/stiff enough (without being really heavy), and having the weight end up above the chassis plate did not maximize the downward end of the distance the center of mass could fall.

So, cut a “bay” out of the back half of the chassis plate- so front half was solid, back half was “legs” extending back, with open space between them. Extended the trapeze arms so the weight ended up swinging in at the bottom edge/side of the chassis plate. Did some brain-storming on "how else can we get the weight to move from high/over the back to low in the middle"? First iteration was a plate- a piece of 3/32nds plexi- low end at the front of the bay, upper end above/over the real axle; steel bar weight would slide down it, but would rotate unpredictably on the way down.

From work on our robot arm, we’d found out about how well tubes sliding on rods worked (with tube i.d. just slightly bigger than rod o.d). That led to the configuration we ended up with. Rods (1/8” stainless steel, highly polished) are anchored in lexan blocks mounted on the underside of the chassis plate at the edge of the “bay”; they angle back (at about 25 degrees from horizontal) to a bit behind the rear axle; lexan bracket pieces glued to the inside of the chassis plate “legs” hold/mount the upper/back ends of the rods. With the vehicle on the ramp, the rods are at about 85 degrees from horizontal, so the lead weight falls almost vertically. 2kg falling 10-ish cm is a pretty good shot of energy. The lead weight is held in a framework of ¼” thick lexan pieces, and that framework holds 5/32nds” brass tube pieces (i.d. just a hair bigger than 1/8th”). Getting the rods parallel, and in a plane, and the tubes parallel and aligned with the rods was a challenging exercise in precision, and for this system to work, it has to be precise. But when you do that, even with the vehicle sitting horizontal on the floor, the weight slides freely down the rods. Front face of the weight assembly is shaped/angled so that it gets full/close contact with the edge of the bay. 1/16th thick lexan plate on the edge of the bay; ¼” thick lexan plate on the front edge of the weight provides a good elastic collision when weight meets bay edge (i.e., maximum energy transfer - of the 2kg falling ~10cm -from the collision). The carbon fiber/balsa core sandwich of the chassis plate acts as a very effective sounding board; the “whack” of the weight hitting home really resonates. With the energy shot, the vehicle launches much faster than it does from just a rolling start.

So, especially when you add in the complexities/challenges of the braking system, it is in one sense certainly…..a bit over the top; a lot of time & effort, and a significant cost. The Team 2 vehicle went down a significantly simpler pathway, and ended up with performance pretty close to T1. A very good demonstration of an important concept; once you’re at a “good” level, the incremental “cost” of improvement (time/energy, and $s) tends to really ramp up. Each year, in coaching the building events, I try to end up with a mix of “intensity” put into the various events; for some, the ‘game” is how to get a decent device at minimal time and cost; for some, its “let’s go for it- do something really well.” Depends on who wants to really invest time, and who is fighting the time demands of things other than Science-O. This year, my two Team 1 guys really wanted to go for it on GV from the beginning, and so we did. They ended up with a way cool machine, and they learned a heck of a lot along the way. As a coach, what more can you ask for….. it’s been a lot of fun.

Couple last thoughts, then, on the incremental improvement in speed that the Team 1 guys were able to produce – and the nature of this event.

The improvement – the speed difference between the two vehicles – is absolutely real; enough carefully timed runs on both vehicles now to know for sure there is a 0.2 (at 5m) to 0.4 (at 10m) second difference. Scoring-wise, that’s the same value as 2 to 4cm off-target. But, given the fact of hand timing, the difficulty the folk timing being able to clearly see both start of movement and end of movement, reaction time variations in different people – the unavoidable human factor, a difference of this magnitude is at a high risk of getting lost in actual competition timing and scoring. I’m quite certain that at many competitions across the country, the time scoring component scores/places were different than “reality.”

But, as I’ve said before, there is no practical way to get precise, actual times. So, there is an inherent, unavoidable “roll of the dice factor” in the event, and where it comes into play the most is at the “top end”- those little differences that could/should be ‘the winning edge.” Even given this, though, I have to say,I think it’s a neat event.

Last thought is on what might be done next year. Precise timing being impractical, how else might you more precisely measure/score the “speed” factor? It comes from and is dictated by how well you do two things; a) maximize the velocity off the ramp, and b) minimize the friction loss rate through the run. Those same two factors happen to determine a precisely measurable value – how far the vehicle can roll. Out of curiosity, we did this test on our vehicles last weekend- T1 goes a bit over 30m; T2 is a bit over 20m. If (for space practicality), the ramp height were reduced a bit, and maybe the weight reduced some (less momentum = less distance capability), a ‘total distance capability’ factor could be used as a practical and precise scoring factor. Just a thought….

[physicsphan]

Last thought is on what might be done next year. Precise timing being impractical, how else might you more precisely measure/score the “speed” factor? It comes from and is dictated by how well you do two things; a) maximize the velocity off the ramp, and b) minimize the friction loss rate through the run. Those same two factors happen to determine a precisely measurable value – how far the vehicle can roll. Out of curiosity, we did this test on our vehicles last weekend- T1 goes a bit over 30m; T2 is a bit over 20m. If (for space practicality), the ramp height were reduced a bit, and maybe the weight reduced some (less momentum = less distance capability), a ‘total distance capability’ factor could be used as a practical and precise scoring factor. Just a thought….

I like the idea of the distance you are trying to optimize for scoring purposes. Maybe having a variable mass or initial height that the vehicle had to be release at to add the challenge that the variable distance did this year.

[iwonder]

Last thought is on what might be done next year. Precise timing being impractical, how else might you more precisely measure/score the “speed” factor? It comes from and is dictated by how well you do two things; a) maximize the velocity off the ramp, and b) minimize the friction loss rate through the run. Those same two factors happen to determine a precisely measurable value – how far the vehicle can roll. Out of curiosity, we did this test on our vehicles last weekend- T1 goes a bit over 30m; T2 is a bit over 20m. If (for space practicality), the ramp height were reduced a bit, and maybe the weight reduced some (less momentum = less distance capability), a ‘total distance capability’ factor could be used as a practical and precise scoring factor. Just a thought….

I like the idea of the distance you are trying to optimize for scoring purposes. Maybe having a variable mass or initial height that the vehicle had to be release at to add the challenge that the variable distance did this year.

Maybe not the variable height, as it might make some release mechanisms more complicated, but a weight that would change and had to be used to propel the vehicle would be cool. But since this is a new event and we're just getting started, it would be nice to not see to many major changes next year

Also, I'm kinda confused with it, but whats the benifit of small diameter axels? I've heard people mention rotational inertia, but I just can't seem to understand how that effects speed that much. And from a building standpoint... How did other teams use bearings with the threaded rod? We ended up turning our own axel on a lathe, it seems out of reach for most teams(in fact, it took us a couple months to find someone who would help us turn it)

[illusionist]

What's a material that can be used for covering CD's to give them grip. I've tried balloons and latex gloves, but the issue is that they rip and break down due to the friction of braking. Any other suggestions?

[twototwenty]

Balsaman-So does your release system release the car and the weights, or just the weights, which jolt the car to a start? It seems that the first would be the best option, but it also seems like it would be much harder to get a system that release the weights and the car at the same time.

[Balsa Man]

Twototwenty, first trigger (activated by a pencil) releases the weight. The weight trips a second trigger on the underside of the vehicle, releasing it from the ramp.

Iwonder- it is small wheel size that reduces rotational inertia. That can help.... a fair amount. Energy goes into "spinning up the wheels"- more mass, more radius, more energy goes into it Small axle diameter just means smaller bearings- less friction than larger ones. We''re running 1/8th i.d, 1/4 o.d. For the threaded rod axle on our Team 2, we just had to sand down a few thousandths to get a nice fit.

Illusionist, that's why we went to wheels that would take O-rings. I really can't think of anything to slip on a CD that will hold up....

[Saturiea]

What's a material that can be used for covering CD's to give them grip. I've tried balloons and latex gloves, but the issue is that they rip and break down due to the friction of braking. Any other suggestions?

I have had great luck using rubber bands, they tend to last a good period of time and add enough grip to help.

[twototwenty]

Twototwenty, first trigger (activated by a pencil) releases the weight. The weight trips a second trigger on the underside of the vehicle, releasing it from the ramp.

Iwonder- it is small wheel size that reduces rotational inertia. That can help.... a fair amount. Energy goes into "spinning up the wheels"- more mass, more radius, more energy goes into it Small axle diameter just means smaller bearings- less friction than larger ones. We''re running 1/8th i.d, 1/4 o.d. For the threaded rod axle on our Team 2, we just had to sand down a few thousandths to get a nice fit.

Illusionist, that's why we went to wheels that would take O-rings. I really can't think of anything to slip on a CD that will hold up....

Thanks...that is quite impressive, compared to my school especially. Is your ramp anything special, or is it just a well made curve that's as tall as possible?

[Balsa Man]

Thank you. It's fun when you have a team that wants to ... really push it in an event.

Our ramps are pretty straightforward. Surface is a piece of formica sheet. Glued to a piece of pressboard shelf, The bottom, oh 18", of the sheet hangs off the end of the shelf board, so the upper part is flat/straight; the bottom, where the formica extends past the board is curved. That transition curve is at a radius of about 30cm. Base board has a wooden frame to hold the shelfboard. When you put it down (with a concrete block on the baseboard), the floor pushes the last part of the formica into the transition curve. One very useful/functional thing it has is a guide rail. Its a 1/4 x 1/2 cross section strip of HDPE plastic, screwed onto the ramp (can't glue HDPE). Vehicles have finger units front and back- they carry a pair if 1/8 nylon rod pieces that engage the sides of the rail. Circular cross section of the rods means very small contact area. Two/three thousandths clearance; nylon to HDPE friction is very low. With the rail, vehicles come off the ramp on a VERY consistent line. If the ramp doesn't move, they'll run the same line within a centimeter at 10m.

[Balsa Man]

One of our coaches got video at State yesterday- YouTube at:
http://www.youtube.com/watch?v=nHhQAAdU ... e=youtu.be
I'll try to get a few pictures together for better close detail.

It was a good day; everything worked. 7.8m, 3.5cm, 3.1 sec. Our timers had it at 2.8 sec; 2.7 on the first run. Got a bit of brake rebound on the second run that added time.

As a matter of principle, simpler is generally better. But in this case, a fairly complicated set of things to optimize a number of performance parameters.....worked like it was designed to. Hard work and careful testing and practice paid off

[illusionist]

One of our coaches got video at State yesterday- YouTube at:
http://www.youtube.com/watch?v=nHhQAAdU ... e=youtu.be
I'll try to get a few pictures together for better close detail.

It was a good day; everything worked. 7.8m, 3.5cm, 3.1 sec. Our timers had it at 2.8 sec; 2.7 on the first run. Got a bit of brake rebound on the second run that added time.

As a matter of principle, simpler is generally better. But in this case, a fairly complicated set of things to optimize a number of performance parameters.....worked like it was designed to. Hard work and careful testing and practice paid off

Congrats on the gold! Your vehicle looked nothing like I imagined it, but I'd love to see some close up pictures of the individual assemblies.

[Balsa Man]

Thanks.

One thing that surprised me at our State tournament, that has relevency for those going to Nationals; almost everyone was running the paper clip in a fixed position. Rule 5(g) explicitely allows adjustment between runs.
If you run it in a fixed position, to improve on your distance score on the second run, you have to adjust both braking distance and L/R alignment. Braking distance is/should be pretty straightforward, easy, and pretty precise. Alignment is a much more challenging thing to do consistently/reliably.
With an adjustable clip positioner, the problem becomes much easier.

For this to work/help, you do have to get to where it rolls close to the same line each time; this can for sure be done with a guide rail (we were seeing within a cm L/R at 10m); I suspect with a good solid/stiff chassis, good tightly positioned bearings/wheels you could get pretty good linear repeatability with wheel positioning/alignment blocks on the ramp. You have to be able to do your first run alignment to get within the L/R range of your adjustment "frame." In the video of ours, you can see that frame sticking off the left side. Its about 15cm wide, and 30cm long. Two carbon fiber rods sticking out, with a bar that can be slid L/R along them; the clip is mounted on that bar, and cam be adjusted fore and aft. With good linear consistency, how you adjust for the second run depends on how close the first is. Let's say first run, the clip is 6cm to the left of the mark, and 7cm past it (a bit over 9cm fm the mark). You adjust the clip right 6cm, and back 7cm, set the brakes the same. If it runs the same line, and the brakes go off at the same distance, the clip ends up right over the point. If the braking distance is off more than your frame allows, you have to adjust number of revs- let's say 30cm off - long. Wheel circumference on ours is about 18cm. Take off 1 rev (-18cm), adjust clip back another 12cm, and you've adjusted to -30cm