View Full Version : 2 Gear Drive Train
Is any team using a multi-geared drivetrain? If so what are some pros and cons.
John Bono
06-02-2003, 23:02
You must be joking. I suppose if one of those teams that can afford to live in a machine shop could do it, but there's not that good of a reason. Either you push or you zoom. Chances are your bot only has to do one of these during the match depending on its design. And I know our team isn't going to waste the valuable servo/pump to hook up the system.
For that matter, if you're asking that question this late, chances are you're pretty screwed.
Originally posted by John Bono
You must be joking. I suppose if one of those teams that can afford to live in a machine shop could do it, but there's not that good of a reason. Either you push or you zoom. Chances are your bot only has to do one of these during the match depending on its design. And I know our team isn't going to waste the valuable servo/pump to hook up the system.
For that matter, if you're asking that question this late, chances are you're pretty screwed.
Our team already has the two gears and they work beautifully. Thanks or the comment. I am just trying to see if anyone else tried it.
Jeremy_Mc
06-02-2003, 23:35
Originally posted by John Bono
You must be joking. I suppose if one of those teams that can afford to live in a machine shop could do it, but there's not that good of a reason. Either you push or you zoom. Chances are your bot only has to do one of these during the match depending on its design. And I know our team isn't going to waste the valuable servo/pump to hook up the system.
For that matter, if you're asking that question this late, chances are you're pretty screwed.
that was rather harsh....
you don't have to live in a machine shop to do it. you just need experienced machinists. we didn't do two gears, and we probably couldn't if we tried.
and it's definitely not too late to change to a two gear system.
until you hit the field, it's never too late.
*jeremy
you dont need to machine squat. use a servo and shift the drill motor. I dont know what else youd use it for that is better than shifting. PS shifting will be essentila this year, it is ALWAYS better to be able to go slow/torque, or fast
Cory
John Bono
06-02-2003, 23:48
Oh yeah--I suppose you could do that. But you'd have to be at a perfect stop and hope your gears come together jsut right.
Ken Leung
06-02-2003, 23:55
To a team without any experience making a servo shifter, it would seems impossible to do without a machine shop. However, many have learned that it take mearly a hand drill and a couple pieces of metal to make a successful one. All it take is simply looking at other's design, and trying to duplicate it.
Given a week of time, ANY team can make one work as long as they've seen it before.
If you want to hear what people have to say about shifting gears on drive train, take a look at this thread: http://www.chiefdelphi.com/forums/showthread.php?s=&threadid=16810
Having a speed shifter on your robot simply give you more freedom in the match. It is true that you may choose one speed during one match, but over all, there will be many different situations, and you may find yourself needing more torque, or more speed depending on what teams you are playing with. All it take is a few pounds, and a few days of work, with many things to gain.
Originally posted by John Bono
Oh yeah--I suppose you could do that. But you'd have to be at a perfect stop and hope your gears come together jsut right.
It is really ez actually just put in the program to not allow the shifting button to workl until the motors are at a complete stop. The gears will mesh automatically there really is no degree of precision when shifting with a servo
Cory
having a gear switching device is really easy (if you're using drill motors). we slapped one on our robot in less than an hour last year.
Andy Baker
07-02-2003, 01:19
Originally posted by John Bono
You must be joking. I suppose if one of those teams that can afford to live in a machine shop could do it, but there's not that good of a reason. Either you push or you zoom. Chances are your bot only has to do one of these during the match depending on its design. And I know our team isn't going to waste the valuable servo/pump to hook up the system.
For that matter, if you're asking that question this late, chances are you're pretty screwed.
Just so you know, you should check the white papers on this site.
There are a number of gear switching desings that are posted on there... from numerous teams. Our team has posted 3 separate designs (with all details included) that shows the design of a gear shifting system. 2 include shifting while stopping and 1 includes shifting while moving. We have published one design during each of the last 3 summers, and these designs have been downloaded thousands of times.
We've been shifting gears since '99 and we will be again this year.
Andy B.
sevisehda
07-02-2003, 04:44
Um I'm taking it your new to FIRST Bono. ALOT of the teams use multi-geared transmission. Some actuate the drills others build customs rigs like the X-Cats. I think someone even has a C.V.T.
The benefits are amazing. If you want to push you have to go slow, <3ft/s, but it would take you forever to get places like that, but being able to run >10ft/s would mean you'd be able to knocked out of the way easier than a box. Its a cylinder and 2 extra gears to shift not that much considering the benefits, cost/benefit wise its the best thing you could do in most cases.
sevisehda
07-02-2003, 04:51
I'm not sure if you meant to but you came off pretty bitter and angry. I think that AT Man was trying to start a discussion on why or why not to use a multi geared transmission. And theres still plenty of build time left, especailly considering people have built entire bots during the practice times before regionals.
2 Speed Gear Boxes...
Pros:
Versatile in the fact they can push and run.
If you can shift on the fly you can get better acceleration.
It looks really cool.
Teaches the team alot if its the first time you've tried it.
Cons:
Heavier than a 1 speed.
Have to use pnematics,servo,motor.
Complex to fix/build.
Larger space wise.
John Bono
07-02-2003, 19:03
I'm sorry, I was working on the assumption that the person asking was rather new (although this is my second year with first and I indeed have multi-gear drive trains). It just seems pointless unless you have a lot of shoving to do--and our team doesn't want to use the valuable space for the servo or piston, and it seems to me not many teams would have (we ALWAYS max out our pistons in any concept sketch). And servos just don't have a whole lot of umph to them.
Also, the real force you should be worried about is friction with this year's surfaces, I wouldn't expect you to be able to get a friction coefficient of more than .3 or so (I guess as high as .7 is possible if you're awesome cool), and the drill motors can put out more power than that on low if I recall (then again, I could be a ranting moron--chances of which are high).
Originally posted by John Bono
I'm sorry, I was working on the assumption that the person asking was rather new (although this is my second year with first and I indeed have multi-gear drive trains). It just seems pointless unless you have a lot of shoving to do--and our team doesn't want to use the valuable space for the servo or piston, and it seems to me not many teams would have (we ALWAYS max out our pistons in any concept sketch). And servos just don't have a whole lot of umph to them.
Also, the real force you should be worried about is friction with this year's surfaces, I wouldn't expect you to be able to get a friction coefficient of more than .3 or so (I guess as high as .7 is possible if you're awesome cool), and the drill motors can put out more power than that on low if I recall (then again, I could be a ranting moron--chances of which are high).
Umm yeah... I know teams that have a COF of >1
Cory
John Bono
08-02-2003, 00:06
Wait-wait. It takes more force to move thier robot than it weights?:yikes: Thanks for breaking several laws of physics.
sevisehda
08-02-2003, 00:31
Having a COF over 1 is possible. I don't know why that breaks some physical laws. FIRST even gave you some ideas for high-friction materials in the kit of parts.
John Bono
08-02-2003, 00:47
Hmm.., let's think here. One newton down--force required to move left or right, two newtons. That doesn't make sense--it's more than just macroscopic differences in the elevation of the surface.
Now TRACTION is different all together. If you can some how "dig in" to the carpet or mesh, then yes. Or possibly if you've gone to the incredible effort of having small suction cup studs on your wheels--well no, there you still have the friction of the suction cup. Not friction, but traction, I suppose. And a ramp dominator could deffinately use being able to gear down, I suppose (but again, with the only simple fix, it would have to stop first)--but I'm sure you would've asked already if it were that crucial to the design. I suppose if you have an extra few man hours to spare, then you could try it out. We're so tight and heavy we couldn't even dream of it.
sevisehda
08-02-2003, 02:10
I'm not a physicst but its obvious you have no clue about friction. The coefficent of friction is F = µ R. where F is if friction, R is the weight of the object and and mu is the COF.
You're not thinking about this right. A robot weighs 130lbs but if the coef of friction equals 1.1 it would take you 143lbs of force laterally to move him, but you could lift him with just 130 lbs. There is a difference in how you apply the force.
Some COF:
rubber on concrete 0.8
glass on glass 0.9
steel on steel 0.74
aluminum on aluminum is 1.1
Also your analysis of traction vs friction is wrong. Friction is the resistance to movement. Now aluminum on aluminum is due to cold welding same with steel on steel. Its more due to chemical properties of the material. Where rubber on carpet there is more interlocking going on so its more due to the physical properties.
I would define traction is the weight of the bot time the coefficent of friction.
If you don't believe me do a search and on the internet and it will confirm everything I've said. IM me if your still confused.
Stephanie
08-02-2003, 04:13
the coefficient of friction can also be greater than one in cases dealing with circular motion. IE if the robot is making a turn.
$0.02
John Bono
08-02-2003, 16:03
$@#$@#$@#$@#--I'm wrong again. Just isn't my week I guess. Although I don't count sticking as friction (which cold welding basically is). I would say that you'd better have a low to the ground center of mass if you're planning on more than 1.0 being you're total friction (since I'm still not convinced that you could get more than that without actually sticking on to the carpet somehow). And of course, then the bot would be lifted clear off the ground before its total friction were reached.
And remember, the easiest thing to do to get your traction up at least a little is to be as close to the weight limit as possible.
not neccessarily, lets say that you have wings that come down, these wings have file cards on the end to anchor them to carpet. if those wings only weigh fifteen pounds, youve got even less than 15lbs of downforce on those cards. If you have file cards on the bottom of your robot however, you would have much much more downward force on the cards, giving you better traction.
The Point: unless you have weight over your traction material, youre not gonna have much traction
Cory
Stephanie
08-02-2003, 18:55
uhm... bono, i don't mean to be rude or anything, but i'm curious: have you ever taken a physics course?
if not, quit while you're ahead :(
Matt Leese
10-02-2003, 10:59
Originally posted by Cory
not neccessarily, lets say that you have wings that come down, these wings have file cards on the end to anchor them to carpet. if those wings only weigh fifteen pounds, youve got even less than 15lbs of downforce on those cards. If you have file cards on the bottom of your robot however, you would have much much more downward force on the cards, giving you better traction.
The Point: unless you have weight over your traction material, youre not gonna have much traction
Cory
File cards are a bad example. The amount of traction you get with file cards has very little to do with friction. Instead, the file cards mesh with the carpet which is a force completely unrelated to friction.
Matt
sevisehda
10-02-2003, 11:07
File cards are a bad example. The amount of traction you get with file cards has very little to do with friction. Instead, the file cards mesh with the carpet which is a force completely unrelated to friction.
It may not technically be friction but for the purposes of calculations it is. Also i doubt filecards will do much on hdpe.
Jeff McCune
10-02-2003, 13:10
Originally posted by John Bono
You must be joking. I suppose if one of those teams that can afford to live in a machine shop could do it, but there's not that good of a reason. Either you push or you zoom. Chances are your bot only has to do one of these during the match depending on its design. And I know our team isn't going to waste the valuable servo/pump to hook up the system.
For that matter, if you're asking that question this late, chances are you're pretty screwed.
We have a Continuously Variable Transmission up and running nicely. Our drivetrain consists of 6 motors (3 compose each left / right assembly.) with many custom made gears and belts. We get a very nice and well balanced blend of speed AND torque without the need to "change gears" so to speak, since it's continuously variable.
I don't believe that we "live" in a machine shop any more than most teams with shop access, but the point is that a decent transmission certainly is possible. Could we produce a CVT without shop access? Probably not. However, last year we did put a block of HDPE on two pins. The HDPE fit over the drill transmission switch and it was very easy to actuate with a small servo. Teams can and do produce transmissions such as this without shop access every year.
Why not push, zoom AND hope to do it better than everyone else? (Not that we will of course... Just *hope*)
It is clear that you guys are either uninformed on the subject of friction or confused.
You have to specify whether you are talking about static friction or kinetic friction.
In static friction, coefficients over 1 are easily possible.
Static friction in meaning means still, unmoving, not slipping. In kinetic friction, moving friction, a coefficient over 1 is practically impossible.
Both static and kinetic friction is applied in the case of robots, but you have to make sure you are using the right one.
Traction, also called adhesive friction is the same thing. The most dependant variable in traction (lets use this word because it puts more of a gripping image in ones mind) is the maximum static friction.
To determine max static friction between two surfaces, one must multiply the static coefficient of friction between the surfaces by the normal force (also the weight). The result of this is the maximum forward force that a robot can apply on a vertical surface (another robot or wall) before the wheels start slipping.
When the wheels slip this changes traction from static friction to kinetic friction and greatly lowers the amount of force that a robot can apply to actually push forward.
Its a little hard to explain in words. Wish I knew the coefficient of the wire mesh and hdpe. But you can determine that through testing (determining torque off wheels through motor and gearing/Fn). Havent done that yet.
Steve
In my eight years we've yet to try a multi-speed transmission......but I'd love to try.
We've designed our machines to be mostly offensive in the past.(we don't usually concern ourselves with torque because we try to avoid the hard pushing match-ups). This year was especially hard for us because we are using every motor and all the pneumatics........we don't use the drill transmissions so servo mounted gear switching is probably out of the question.
I love the CVT...that would be a great project, but for reliability sake, I'd probably just duplicate one of the three systems already developed by ultra-seasoned teams like 45.(their on-the-fly tranny was super sweet last year.)
Good luck to all those ramp bots this year, we'll see you under the bar!
Jeff McCune
10-02-2003, 15:01
The "object" need not be something vertically upright that the robot makes contact with either... For example, when a robot is driving up the ramp, you can essentially think of an imaginary rope attached to the back of the robot, and gravity "pulling" on this rope in the exact opposite direction that is "up the ramp."
This imaginary rope exerts a force on your robot, which may or may not (may not hopefully) be enough in and of itself to overcome the force required to break the wheels free and spin.
Remember, you're really multiplying the coefficient of friction times the NORMAL force. The normal force happens to be exactly the weight of the robot when situated on a level surface, but this normal force decreases as incline increases.
This means that if the ramp and floor were the same surface, less lateral force (relative to the robots plane) would be required to break the wheels free on the ramp than on the level ground. This makes perfect sense, since if you imagine the extreme case of a 90 degree incline, NO force is required to break the wheels free from the surface, since there is no normal force.
exactly.
i just with these people understood it.
John Bono
12-02-2003, 17:26
so... I wasn't entirely incorrect...
jzampier
12-02-2003, 18:18
There is a fascinating manual on vehicle mechanics called: "Mechanics of Vehicles" by Jaroslav J. Taborek. It contains a fairly detailed analysis of vehicle statics and dynamics which could be applied to robot motion. In there, the coefficient between the wheels and surface is referred to as the Coefficient of Road Adhesion.
It deals with the, "particular behavior of the elastic wheel, the coefficient of friction for rubber tires has been more appropiately called coefficient of road adhesion." p. 7
I highly recommend this as reading as you will feel smart afterwards even tho it makes no sense whatsoever :)
What do i know about this, i'm a CE.
hixofthehood
12-02-2003, 19:05
http://www.chiefdelphi.com/forums/pictures.php?s=&action=single&picid=3627&direction=DESC&sort=date&perrow=3&trows=4&quiet=verbose
S.P.A.M.'s gearbox is on the left.
The applications?
Well, "Lo-gear" is for slow and low, and "Hi-gear" is for high and fly.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.