Chief Delphi

Chief Delphi (http://www.chiefdelphi.com/forums/index.php)
-   Technical Discussion (http://www.chiefdelphi.com/forums/forumdisplay.php?f=22)
-   -   Drives in general (http://www.chiefdelphi.com/forums/showthread.php?t=59522)

lukevanoort 09-11-2007 23:21

Re: Drives in general
 
Quote:

Originally Posted by Gamerboyb1 (Post 650830)
no when i say 4 wheel i mean with omni wheels.
that would be a very big act of lessened intelligence to just use regular wheels.

Not really, you just need to be more careful about how you do it to make sure you get what you want. There is nothing inherently bad or good about any given drive system, just things is does well and poorly. For example, in my opinion, the most universally applicable* drivetrain used by FRC robots is tank treads; however, I can guarantee with almost 100% certainty that I will never advocate building such a system to my team. Here's why: properly designed, a tank tread system is reliable, has good speed, good traction, unsurpassed performance climbing obstacles, and good maneuverability; unfortunately these come at a cost: tank tread systems are expensive, heavy, and hard to design well. For me, that is a trade-off I am not willing to make; thus, I don't choose treads, but that doesn't mean there is anything wrong with them.

To the specific matter of 4WD with omni wheels, forward traction would be similar to a 4WD/6WD robot without omnis (this, of course, depends on the omni roller's material; PVC roller will be much lower traction than most normal wheels while SBR rubber rollers provide similar traction to their normal counterparts); however, the sideways traction (parallel to the driveshaft) is not as good, which is both the advantage and disadvantage of omni wheels. There is always a trade-off, in this case you lose resistance to sideways pushing in favour of maneuverability. Maneuverability should be similar to a 6WD while being lighter, possibly cheaper (depends on wheels), and simpler.


*by universally applicable I mean it works the best in the most games

PS: If you are curious, I have a PowerPoint I put together with some vague plan to present to my team at some point that goes over some of the basic pros/cons of various common drive systems.

joeweber 10-11-2007 09:05

Re: Drives in general
 
We have been working on a list of drive types on our web site http://team1322.org/omni_drive.htm , the list is not complete but getting close. Everybody asks for what use, but you want to know for all uses. You want Traction, Power, Speed and Maneuverability but is all of that possible? You would have to have semi soft rubber wheels across the bottom of the whole robot that turn 360 degrees each, groups of them would have to turn independent of each other have a two speed transmission and independent suspension so that it can climb. As soon as some one figures all this out with the limits we have let all of us know. The closet we have come to this is the four wheel drive crab drive. Our system turns all of the wheels the same direction including the arm. The arm would always be pointing the direction you are moving.

Gamerboyb1 10-11-2007 11:49

Re: Drives in general
 
yeah but why not an 8 wheel system
wouldn't that have the most pushing power and mobility

GMAdan 10-11-2007 11:57

Re: Drives in general
 
An 8 wheel setup doesn't necessarily mean more mobility. An 8 wheel setup could create a lot of scrub making it hard to turn quickly and you really can't do a lowered middle wheel to reduce the scrub. Also pushing power is not only a function of the number of wheels. There is more to pushing power, all you have to make sure is that your wheels can provide enough traction to with stand a pushing match. 6 or even some 4 wheel setups can do that.

Madison 10-11-2007 12:45

Re: Drives in general
 
Quote:

Originally Posted by Gamerboyb1 (Post 650872)
yeah but why not an 8 wheel system
wouldn't that have the most pushing power and mobility

Can I ask that you try to do a better job of explaining your questions and thoughts when you post? You don't need to be familiar with all of the proper or common terminology -- we can help you to learn that -- but I think you'd be able to learn a lot more if you help us out a bit and share some more of your thoughts.

Why do you think that 8WD will have the most pushing power or mobility?

Gamerboyb1 10-11-2007 12:50

Re: Drives in general
 
Surface area on the wheels and the fact that there would be 2 more transmissions that a 6 wheel system with a center drop. That and you would not need a center drop system.

AndyB 10-11-2007 13:32

Re: Drives in general
 
I believe it was in a powerpoint done by Madison Krass (M. Krass) that discussed various drive types.

In it she said, after 6 wheels, you are pretty much just adding weight.

This is because traction is only dependent on the friction of the wheel and the weight of the robot. It is not dependent on the contact area of the wheel or the number of wheels.

Of course you could add another set of gearboxes which would of course add more power, but assuming KOP transmissions, you can really only use four maximum anyways. There is no reason why you couldn't use all four gearboxes on a four wheel drive, six wheel drive, eight wheel drive, or an 18 wheel drive.

The reason why teams then use six wheel drives over four wheel drives is for the fact that you retain mobility without sacrificing the ability to not be pushed around.

Madison 10-11-2007 13:37

Re: Drives in general
 
Quote:

Originally Posted by Gamerboyb1 (Post 650879)
Surface area on the wheels and the fact that there would be 2 more transmissions that a 6 wheel system with a center drop. That and you would not need a center drop system.

Okay, let's take this bit by bit:

1 -- Increased surface area increases traction.

Simply speaking, this is not true. Our robots move because of the force of friction between their wheels and the playing field. Friction is defined by an experimentally derived constant, mu, and by an object's "normal force". To say that something is experimentally derived means that the only way to know the right value is to test materials together and record the results. A "normal force," again simply speaking, is the weight of your robot.

The largest "mu" -- usually called a coefficient of friction -- we see in FIRST is about 1.3. To determine the absolute maximum force ("traction") that our robot can exert, we have to multiply "mu" times "normal force". So, generally speaking, robots can exert about 1.3 times their weight in "traction," or 160 lbs.

All of the above assumes fundamentally that all of your robot's wheels are attached to a motor somehow. I can explain why that's important later. What's important to realize here is that the number of wheels or the surface area of those wheels does not affect either "mu" or the weight of your robot.

2 -- 8WD would have two more transmissions than 6WD

Why does 8WD necessarily have more transmissions -- or, perhaps, can you better describe what you mean by "transmissions"? Most FIRST robots will have just one transmission or gearbox for each side of their drive train.

Like surface area, the number of transmissions or even motors on a drive train does not affect your robot's maximum possible "traction." Transmissions only help to make the motors given to us by FIRST more suitable for our needs by slowing them down. Drive design can be pretty complex, so we can talk more about that later.

3 -- 8WD does not require a center drop

Whether or not a drive requires the "center drop" is not dependent upon the number of wheels, but upon the "coefficient of friction" I mentioned above. A high coefficient of friction means that our wheels can transmit higher force, but the reverse is also true. It requires a similarly increased force to move those wheels from the outside. Since turning involves moving the wheels sideways -- since most FIRST robots don't have "steering" in the way that we think of it from a car -- that increased coefficient of friction means it's harder to turn. Again, I'll repeat what I've written a few times before. The number of wheels isn't relevant to determining the force required to turn if the length of the wheelbase -- the distance between wheels -- is the same in all designs.

"Center drop" works because it shortens the wheelbase.

Maximizing traction in a FIRST robot has very little to do with number of wheels or their surface area. What is important are the following things:

A -- Weight Distribution


It's of paramount importance to distribute the weight of the robot over wheels connected to a motor. By further shifting the weight distribution around the robot, it's possible to maximize your traction while minimizing turning difficulty.

B -- Coefficient of Friction


Selecting the right wheel material is also very important. Most teams select their wheel material before designing the rest of the robot, and while that's more practical, ideally you want to select wheel material only after you've finalized the weight distribution on your robot. Only at that point can you optimize your design by choosing materials with a coefficient of friction that matches your desired operating characteristics. It's always a compromise between mobility and traction. You cannot have the best of both worlds.

C -- Gearing


This is the hardest part of drive train design. Gearing is the design or selection of a gearbox or transmission that makes the best use of the motors we're provided to move your robot around. It's here that you make decisions that determine how fast your robot will go, how close to the maximum available traction you will come, and whether or not you will exceed the allowed current draw of ~40A per motor and pop breakers. This is an advanced subject.

EricH 10-11-2007 14:07

Re: Drives in general
 
Here are a couple of whitepapers to help you. One is by M.Krass:
http://www.chiefdelphi.com/media/papers/2037

And one form team 670, into the bargain:
http://www.chiefdelphi.com/media/papers/1705

A quick CD-Media search turned up a lot more--I think these will be most useful at this point.

Molten 12-11-2007 18:18

Re: Drives in general
 
I like the idea behind the direct drive system. It looks great on paper. However, when we used this idea last year. The only problem I learned from this is that I despise the motors we used. They are so long, that we had only two inches to spare. The wheels were almost scraping the edge of the robot. So, no matter what kind of drive you use, just keep in mind the overall width. Unless we have shorter motors this year, do not use the direct drive system.

joeweber 12-11-2007 19:19

Re: Drives in general
 
We have tried to develop dive systems with the motors inside the wheels. It worked pretty well (pictures http://team1322.org/omni_drive.htm ) but it was used for omni drive. I suppose you could put a direct drive system inside a drive wheel, you would save space but the clearances to make it work would be really tight.

For the discussion about six wheels or eight wheels or more, if you can make a perfect robot that has all the traction you can get and have sensors to check for slippage and you can some how out maneuver every body you still will be against three other robots that are working over time to try to shut you down. Teams will be contemplating on strategies and think of way around you. So you solve one problem than another turns up and no matter how good your drive system is it is only as good as your drivers. I would take a standard drive system with great drivers any time over a perfect drive. What we should be working on is ways to make our drivers better. Find ways to teach them on the fly strategies, ways they can learn to drive the robot that has already been shipped.

JesseK 12-11-2007 19:28

Re: Drives in general
 
Quote:

as well as the fact that defending a robot with omni wheels in the front is relatively easy. With six wheel drive... not so easy.
By concept, the 6WD robot that drops the middle wheel only has 4 traction wheels on the ground at any given time anyways. Turning a bot like that is just as easy as a 4Traction/2Omni bot. The difference is where the weight is placed -- inherently most 4Traction/2Omni bots have their center of mass somewhere over the back 4 traction wheels so that more of the robot's weight helps with traction. 6 Traction wheels that rock back and forth usually just put the center of mass somewhere near the middle.

It's a small distinction of concepts, however it can lead to big misinterpretations down the road that lead to disappointment when you design for one thing and get another.

To the OP:
What ^^^ they ^^^ said. There's tons of information available; you'll have to make a tough engineering decision come build time.


All times are GMT -5. The time now is 06:48.

Powered by vBulletin® Version 3.6.4
Copyright ©2000 - 2017, Jelsoft Enterprises Ltd.
Copyright © Chief Delphi