View Full Version : The Crab Challenge
CraigHickman
08-03-2007, 19:22
Well, here we go again...
I'm assuming you've all submitted your Inventor award submissions already, and are probably looking for something to do with your mad CAD skillage. So I'm issuing a challenge: Design a crab drive. This is a very complex thing to do in Inventor, and I'd like to see which teams can come up with a plausible design. The "submission" shall be two renders: One of an individual module, the other of the frame with modules onboard. There won't really be a single winner, just us giving props to designs that seem possible.
Go for creativity! You could go with a traditional crab, or you could go with a coaxial. You could even think up something new!
Adding onto the theme of time limits that FIRST gives us, lets set a deadline: Two weeks, or on the 22nd. All designs should be posted by then, and then we can begin to examine the details (plus this thread will serve as a good inspiration point for teams wishing to try out a crab).
So fire up Inventor!
Brandon Holley
08-03-2007, 19:28
and what is the prize?? ;)
CraigHickman
08-03-2007, 19:33
Bragging rights, and having something pre designed before the upcoming season. (yes, there's always an upcoming season, no matter when it is!)
Where do we post these pictures. Want us to upload to the CD galleries and then link to it from this thread?
CraigHickman
08-03-2007, 19:42
That sounds good, or just attach them to the thread. As long as we can get to it from here, it doesn't matter too much.
Aren_Hill
08-03-2007, 22:18
mines already posted from awhile ago and were using an iteration of it
3-speed shift on the fly swerve +coaxial
CraigHickman
08-03-2007, 22:43
The point of the challenge was to have those who are interested design a new one. I also have about 3 designed, but for the sake of the challenge, I'll be designing a very very, VERY different one.
Tom Bottiglieri
08-03-2007, 22:47
You mean a swerve drive?
CraigHickman
08-03-2007, 23:19
The difference between the two is negligable in relation to the contest, so design whichever you think is more fun/challenging!
=Martin=Taylor=
09-03-2007, 01:59
My first attempt at a crab drive was unsuccessful. The idea was to use as many off-the-shelf parts as possible to save time machining. Although I had some innovative ideas, the design was ultimately too heavy and required to many chains.
http://www.chiefdelphi.com/media/img/15f/15fbe29569069c6075fbcff98caddaaf_m.jpg
Our team basically abandoned the idea even before the season started.
However, after being inspired by 118's crab drive (they were next to us in Atlanta) I took another shot at coaxials.
http://farm1.static.flickr.com/168/415307269_c767a7db70_o.jpg
http://farm1.static.flickr.com/177/415307266_752017603a_o.jpg
Although the modules themselves look nothing like 118's, the chain-system and frame are very similar. I haven’t CAD'd the frame, so no pic...
The modules themselves are made from single pieces of U-channel, which eliminates the challenge of machining 5 bolted side-plates. By using two wheels I also eliminated the chain (which in the original design had no method of tensioning). And by eliminating the chain I made the modules much shorter.
Specs:
-5" wide
-4.5" tall (base of wheel to top of module)
-4" wheels
I'm going to machine these over summer-vacation so stay tuned for another pic!
zander_108
11-03-2007, 11:22
I am going to model a full version of a crab drive that i have designed.. Just not right this second
Daniel_H
11-03-2007, 15:03
what is a crab drive?
zander_108
11-03-2007, 15:17
A crab drive is a drivetrain That all of your wheels can rotate along the Z axis, as well as being able to drive normally.
see these links
http://www.chiefdelphi.com/media/search/results/40411
http://www.chiefdelphi.com/media/search/results/40412
negfrequency
12-03-2007, 19:46
o man its ironic you post this thread
in the preseason my team was playing around with chassis designs and one of my buddies drew up an incredible crab drive design on inventor. Id provide it but i don't have it. Maybe he'll send me a rendering.
They are difficault because you not only have to keep the systems in sync, but you also have to carry rotation to the wheels somehow, unless you have the motor mounted to the wheel in which case you have to find out how to carry wires
CraigHickman
17-03-2007, 21:06
Well folks, the deadline is coming close. And so I figured I'd post my progress on a strange new variation on a crab drive.
Enjoy!
lukevanoort
17-03-2007, 22:06
Well folks, the deadline is coming close. And so I figured I'd post my progress on a strange new variation on a crab drive.
Enjoy!
That looks like you're incorporating the top bearing into the holder for the CIM... I must say, that's pretty slick.
CraigHickman
17-03-2007, 22:36
That looks like you're incorporating the top bearing into the holder for the CIM... I must say, that's pretty slick.
Yep, my goal was to combine a coaxial with a regular crab. I haven't yet done the frame design, so I should be getting on to doing that sometime soon. (probably tonight)
Kris Verdeyen
18-03-2007, 00:19
http://farm1.static.flickr.com/177/415307266_752017603a_o.jpg
Slick. You might need a hoss steering system to drag the dual wheels around, but I like that you got rid of the the extra chain stage that we have, although we do use that stage for an additional gear reduction.
ChrisSanborn
19-03-2007, 16:13
Add a differential, then it would steer very easily.
Chris
CraigHickman
19-03-2007, 20:37
Well, here's my submission. I went with a different kind of crab, one that I haven't seen before. In this one, I attempted to limit or entirely remove the issue of wires binding when routed through a pipe on the top.
Jeremiah Johnson
19-03-2007, 23:57
Well, here's my submission. I went with a different kind of crab, one that I haven't seen before. In this one, I attempted to limit or entirely remove the issue of wires binding when routed through a pipe on the top.
Looks somewhat similar to 1625's crab this year... though they didn't include the motor in the bearing like you did, it was mounted above the assembly.
I don't have a submission, but I'd like to thank everyone that does. We made a drive train using Dewalt drill heads for gearboxes this season, but I'm planning on playing around with a swerve drive over the summer and hopefully designing and putting together my own, so now I get a head start. :) Yay!
Rich Ross
20-03-2007, 17:47
what are the specs of that model? Ground clearance? Speed? anything else of note.
CraigHickman
20-03-2007, 18:32
Speed should be about 7 fps. It weighs way too much (60 pounds). However, it's rolling with about 2 inches of ground clearance, and on 2 inch wide custom wheels. Should have massive pushing ability as well as great maneuverability.
CraigHickman
22-03-2007, 20:35
Well, folks, it's today. I'm looking forward to seeing different crab designs.
Due to lack of activity, if you have a crab designed at all, post a render so we can ooh and ahh.
whytheheckme
22-03-2007, 21:47
Would uploading AndyMark's mecanum wheel cad file be cheating? :p
That is one awesome crab drive (we have it on our bot)
Jacob
Nuttyman54
22-03-2007, 22:01
Would uploading AndyMark's mecanum wheel cad file be cheating? :p
That is one awesome crab drive (we have it on our bot)
Jacob
Mecanum and crab drive are two separate entities. Crab drive relies on wheel modules which rotate in sync to drive the robot in whichever direction is desired. Mecanums are a special type of wheel that provides a force at a specific angle away from directly forward. Vector addition allows mecanum robots to travel in any direction. Each system has distinct advantages and disadvantages, which I will not cover here.
whytheheckme
22-03-2007, 22:13
Mecanum and crab drive are two separate entities. Crab drive relies on wheel modules which rotate in sync to drive the robot in whichever direction is desired. Mecanums are a special type of wheel that provides a force at a specific angle away from directly forward. Vector addition allows mecanum robots to travel in any direction. Each system has distinct advantages and disadvantages, which I will not cover here.
OK. Thank you for the clarification.
Jacob
Woot...my very first attempt at crab...
Total weight is 49 lbs, modules are 9 lbs each. The wheels are 8" diameter, 1.5" thick IFI traction wheels (I think I'll go for 6" next time) and they have 2" clearance. There's a 12:1 reduction; 4:1 in the gears and 3:1 in the sprockets.
As for the top plates holding onto the modules, they're just pieces of 1/8". I seriously don't trust them, but I'm too tired right now to change them to something sturdier (or more aesthetic) =P
CraigHickman
22-03-2007, 23:45
Woah. 8 inch wheels? I personally would go down to 4. It would make the gearbox smaller, make the whole module smaller, and lower your cg a ton.
Looks good!
evanisthat!
22-03-2007, 23:55
My design, I actually did it over the summer then stopped working on it after I had the module Rapidly Prototyped. It is a bit flimsy near the wheel and needs some work on the bevel gears, but it is all pretty much there. I was never able to add a transmission to the design, although I had something similar to what 118 did this year, with the single transmission output, however, it was just with 4 cims.
-108
whytheheckme
22-03-2007, 23:57
My design, I actually did it over the summer then stopped working on it after I had the module Rapidly Prototyped. It is a bit flimsy near the wheel and needs some work on the bevel gears, but it is all pretty much there. I was never able to add a transmission to the design, although I had something similar to what 118 did this year, with the single transmission output, however, it was just with 4 cims.
-108
Very impressive. Especially the prototype.
Jacob
CraigHickman
23-03-2007, 00:00
My design, I actually did it over the summer then stopped working on it after I had the module Rapidly Prototyped. It is a bit flimsy near the wheel and needs some work on the bevel gears, but it is all pretty much there. I was never able to add a transmission to the design, although I had something similar to what 118 did this year, with the single transmission output, however, it was just with 4 cims.
-108
Niiiiiiiice. I like a lot. In fact, I'll probably be stealing your idea on how to roatate the module(steering sprocket around the thing) if you don't mind.
Don Wright
23-03-2007, 03:50
Ever since I saw 111 in my first year in FIRST in 2003, I've always wanted to do a swerve drive... 469 did swerve this year and Art did a lot of design work and I believe it's an absolute work of Art (pun intended). Since he probably won't do it...I'll post it.
This is the model for almost the final version and it was done in UG/SolidWorks:
http://www.freebmw.net/don/x02/FIRST_2007/images/swerve9.jpg
Some more:
http://www.freebmw.net/don/x02/FIRST_2007/images/swerve7.jpg
http://www.freebmw.net/don/x02/FIRST_2007/images/swerve8.jpg
http://www.freebmw.net/don/x02/FIRST_2007/images/swerve10.jpg
I believe it is quite unique because it uses the k.o.p wheels, a 3 speed dewalt shifting gearbox (we don't have the servos on it yet for shifting, but the capability is there), and each wheel is independently controlled for unbelievable maneuverability. Some more pictures of the parts and the modules on the robot:
http://www.freebmw.net/don/x02/FIRST_2007/images/img_7599.jpg
http://www.freebmw.net/don/x02/FIRST_2007/images/img_7706.jpg
http://www.freebmw.net/don/x02/FIRST_2007/images/img_7729.jpg
http://www.freebmw.net/don/x02/FIRST_2007/images/img_7701.jpg
The software team has also done an incredible job with the algorithms for controlling and tuning it and even wrote a self tuning program that tunes the pids independently for each module so the whole drivetrain can be tuned in about 10 minutes. So...if you see a bunch of us sitting in the pit just staring at the robot...you know what's going on. It uses independent USDigital absolute potentiometers for position feedback and those things work great.
Here is a movie of it driving for real for the first time...
http://www.freebmw.net/don/x02/HighSpeed.AVI
If you have any questions or want to see it in person, let me know or stop by our pit at West Michigan or in Atlanta...
Michael Corsetto
23-03-2007, 14:20
Ever since I saw 111 in my first year in FIRST in 2003, I've always wanted to do a swerve drive... 469 did swerve this year and Art did a lot of design work and I believe it's an absolute work of Art (pun intended). Since he probably won't do it...I'll post it.
The software team has also done an incredible job with the algorithms for controlling and tuning it and even wrote a self tuning program that tunes the pids independently for each module so the whole drivetrain can be tuned in about 10 minutes. So...if you see a bunch of us sitting in the pit just staring at the robot...you know what's going on. It uses independent USDigital absolute potentiometers for position feedback and those things work great.
Here is a movie of it driving for real for the first time...
http://www.freebmw.net/don/x02/HighSpeed.AVI
If you have any questions or want to see it in person, let me know or stop by our pit at West Michigan or in Atlanta...
Wow, that is an awesome swerve drive. It seems like you guys have a lot of different driving modes in that video: classic swerve, spin with all wheels tangent to the circle, car steering, etc. Vey cool. I will for sure stop by and check it out in Atlanta. How much does that beast wiegh?
Don Wright
25-03-2007, 20:25
The whole robot, or each swerve module? The whole robot is right at 120 now... I'm not sure of the swerve module weight...I'll weigh one when we get to WMR...
evanisthat!
27-03-2007, 23:22
Entering in this thread has inspired me to look at crab again, and the other night just for fun i threw together a frame design that uses the AM 2 speed shifters for a double crab module. The ground clearence for the motor is .75 inches I think, and the way I have set up the stearing motors (making an X pattern in the chain) allows for the robot to do tangent rotation with only 2 motors. Here are some pictures...:)
P.S. The blue lines are where the chain goes, I am too lazy to make chain in ProE, it would take 1000000 hours....
CraigHickman
27-03-2007, 23:29
Lets just say I got bored again. And so two nights ago I fully designed a Crab. Here's my latest coaxial, a single speed weighing in at 40 pounds (with the single speed gearboxes and steering setups, not pictured)
http://www.chiefdelphi.com/forums/attachment.php?attachmentid=5331&d=1174763825
The kids were nagging me to design something like this, so I started in on it today for lack of other things to do.
It's far from complete, obviously, and it's not too remarkable as compared to other crab drives.
http://img133.imageshack.us/img133/6286/2008swerve3rn8.th.jpg (http://img133.imageshack.us/my.php?image=2008swerve3rn8.jpg)
Aren_Hill
30-03-2007, 22:10
here's my entry, i'm still working on it, but its basically a revamped version of ours this year which should be lighter and more maintainable.
i've yet to add the turning setup because i need to experiment with some motors on the last one i used the fischer price and i'm wondering about the window motors.
teenmisfit
16-04-2007, 23:00
Gear, How did u import the cim on the drill transmission? did u model it? id really like to get it so impliment on my design but i dont know any of the specs for it. Thanks
Aren_Hill
16-04-2007, 23:09
i actually found the cad model linked on CD i can't seem to find it anymore but its straight out of the NBD whitepaper
teenmisfit
18-04-2007, 17:58
Thanks!! :)
Entering in this thread has inspired me to look at crab again, and the other night just for fun i threw together a frame design that uses the AM 2 speed shifters for a double crab module. The ground clearence for the motor is .75 inches I think, and the way I have set up the stearing motors (making an X pattern in the chain) allows for the robot to do tangent rotation with only 2 motors. Here are some pictures...:)
P.S. The blue lines are where the chain goes, I am too lazy to make chain in ProE, it would take 1000000 hours....
how do you carry rotation to the wheels? are there some motors im not seeing?
thanks, vivek
evanisthat!
21-04-2007, 11:52
how do you carry rotation to the wheels? are there some motors im not seeing?
thanks, vivek
The AM's drive the upper most sprockets on the Module. These then turn a bevel gear that meshs with another that then turns a sprocket that drives the wheel.
thank you.
are those window motors that are rotating the wheels?
evanisthat!
21-04-2007, 16:40
thank you.
are those window motors that are rotating the wheels?
Yeah, it takes 2 motors to get the desired turning characteristics, i.e. crab drive and tangential turning.
[527]phil
22-04-2007, 00:18
Instead of 2 motors you could also use 1 motor and 2 magnetic clutches to achieve the same effect, however i don't know if magnetic clutches are FIRST legal. I used one on a tank style prototype robot because i did not have 2 identical motors, yet had 6 magnetic clutches O_o, go figure.
Jeremiah Johnson
22-04-2007, 00:45
Here are a couple inventors of our swerve modules this year. Questions? PM me, I'm really tired and I can't really explain it right now, but it's not too complex.
Tom Line
22-04-2007, 08:41
Nice Budda. Do you have a side view of the chassis? I'd like to see the total height of your sims+trans and how close they sit to the floor.
On another note, why do you have the split ring on the outside of the miter gear shaft. Won't the side force on the miter gear keep it to one side so you don't need to hold it with the split ring? It seems like additional weight to me.
CraigHickman
22-04-2007, 16:38
Woot.
http://www.chiefdelphi.com/media/photos/28442?
http://www.chiefdelphi.com/media/photos/28443?
http://www.chiefdelphi.com/media/photos/28441?
bjimster1
22-04-2007, 19:17
im really looking into designing a swerve drive system however im concerned about the loss of ability to climb ramps and such. I dont see these things climbing much of anything.
is the ability to climb something that has to be sacrificed for the swerve drive, or am I missing something?
CraigHickman
22-04-2007, 19:50
im really looking into designing a swerve drive system however im concerned about the loss of ability to climb ramps and such. I dont see these things climbing much of anything.
is the ability to climb something that has to be sacrificed for the swerve drive, or am I missing something?
You can still have a crab climb a ramp. Just make sure the frame will clear on the outsides of the wheels (like the angles cut into 254/968's frame on the front and back), and hit the ramp with some speed, and you'll go all the way up.
Billfred
22-04-2007, 21:05
You can still have a crab climb a ramp. Just make sure the frame will clear on the outsides of the wheels (like the angles cut into 254/968's frame on the front and back), and hit the ramp with some speed, and you'll go all the way up.
Or the WildStang 2004 approach: lift the rest of the robot (http://www.chiefdelphi.com/media/photos/17849).
(Can I get an en-core?)
Here is a render of our gearbox that we used on this year's robot. It is the first ever Worm-Crab drive (AFAIK).
http://www.chiefdelphi.com/media/img/e3b/e3bb04d17154cce17f489660d87bc32e_l.jpg
http://www.chiefdelphi.com/media/img/a09/a09024f8c5599d5d043c12df83e3b79a_l.jpg
Jonathan Norris
23-04-2007, 18:01
Or the WildStang 2004 approach: lift the rest of the robot (http://www.chiefdelphi.com/media/photos/17849).
(Can I get an en-core?)
Simbotics 2004 approach:
http://www.chiefdelphi.com/media/photos/17474
similar idea.
vBulletin® v3.6.4, Copyright ©2000-2017, Jelsoft Enterprises Ltd.