|
|
|
 darn design trade-offs! the game design committee does this to us every year! - KenWittlief [more] |
 |
|
|||||||
|
||||||||
By: 548swimmer
New: 08-05-2010 13:23
Updated: 08-05-2010 21:33
Views: 1253 times
Weight ~ 30lbs
The omnis ARE NOT driven, if anyone has any idea's on whether or not the suspension is worthwhile or how to drive with suspension, please let me know.
Discussion
view entire thread
08-05-2010 23:54
Dave McLaughlin
Re: pic: Prototype Drivetrain v2
I would recommend powering all of the wheels that will be in contact with the ground, including the Omni wheels. In addition to that, are you planning on using the custom traction wheels shown in this model? If so, do the spokes require 3D contouring?
08-05-2010 23:58
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Dave McLaughlin
I would recommend powering all of the wheels that will be in contact with the ground, including the Omni wheels. In addition to that, are you planning on using the custom traction wheels shown in this model? If so, do the spokes require 3D contouring?
|
09-05-2010 01:03
Eugene Fang
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
I'm not sure how to power the omni's when they're in a suspension. We may use the custom wheels, the spokes are flat, no fillet's into the rim.
|
09-05-2010 01:21
kgzak
Re: pic: Prototype Drivetrain v2
could you possibly design a suspension that could keep a chain tensioned? Possible have it move around a point that would have the chain stay tensioned as it goes up and down?
Or could you have a chain tensioner that keeps a chain tensioned by use of springs and have the springs that hold the chain tensioned be easier to compress/expand than the spring or whatever you are using to make the wheels have suspension.
I am trying to design a suspension for our robot next year but the wheels would be powered in one way or another I don't see a point in having a suspension if you are not going to power the wheels. What is the reason behind having a suspension on unpowered wheels?
09-05-2010 01:30
gyaniv
Re: pic: Prototype Drivetrain v2
Why exactly do you need suspension?
Wouldn't it be much easier to lower the center wheels by about 1/8" and power the other wheels with chain? or just going flat 6WD?
09-05-2010 01:46
548swimmer
Re: pic: Prototype Drivetrain v2
The suspension is there to improve handling, much like a rocker, only smoother.
I was thinking about widening the carriage for the suspension and putting a dual sprocket with a bearing on the dead axle that is the pivot point. Then run chain to the omni.
09-05-2010 01:51
Eugene Fang
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
The suspension is there to improve handling, much like a rocker, only smoother.
I was thinking about widening the carriage for the suspension and putting a dual sprocket with a bearing on the dead axle that is the pivot point. Then run chain to the omni. |
And if you want it to rock, why don't you just drop the center and forget about the suspension?
If you want to still have the suspension, you could either do what you suggested or make the suspension go up and down along an arc with radius equal to the distance between the front and middle wheels, so the chain has a constant distance.
09-05-2010 01:54
sanddrag
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
The suspension is there to improve handling
|
09-05-2010 02:00
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EugeneF
The point of using omnis is to avoid needing to "rock." I don't get how it would be smoother to have it rock AND have omnis.
And if you want it to rock, why don't you just drop the center and forget about the suspension? If you want to still have the suspension, you could either do what you suggested or make the suspension go up and down along an arc with radius equal to the distance between the front and middle wheels, so the chain has a constant distance. |
There is currently no rock at rest. Once you begin moving the suspension absorbs any jarring movements like sudden stops or turns, making it handle more smoothly.
09-05-2010 02:08
Eugene Fang
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
There is currently no rock at rest. Once you begin moving the suspension absorbs any jarring movements like sudden stops or turns, making it handle more smoothly.
|
Don't get me wrong, I think your idea is really cool. It's just the matter of seeing if all the effort to get it to work is worth it, and if the benefits are true benefits.
09-05-2010 02:11
Eugene Fang
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
How can I make the carriages pivot around the center axel?
|
The other option would to be have a long arm pivoting on the center axle, but that might be too heavy/bulky for practicality.
09-05-2010 02:16
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EugeneF
As sanddrag said, have you had any problems with a rigid drive? How much are you planning to have it rock? Too little rock and it would be practically the same as a rigid drive, and too much rock might affect accuracy of manipulating the game object due to uncontrollable front/back rocking.
Don't get me wrong, I think your idea is really cool. It's just the matter of seeing if all the effort to get it to work is worth it, and if the benefits are true benefits. |
With springs tensioning the suspension, it will rock only as far as it needs to. Using a diff. eq. we can figure out the specs on the spring. Ideally, at rest, the spring keeps the wheels exactaly level. Once the robot starts moving and the CG "shifts" the springs will expand cushioning any sudden blow (like a rapid acceleration). Additionally, if this robot starts to get flipped, all wheels can stay on the ground providing some traction. The front ones would lift off the most, increasing the Fn on the rear and center
That still leaves the problem of chain runs though. I'm not sure a slot would have low enough friction, plus the exterior has to be smooth for the bumpers.
09-05-2010 02:56
EricH
Re: pic: Prototype Drivetrain v2
How about this: Run the chain so that it is long enough to handle the longest dimension, and stick a couple of McMaster part 5896K1 in there as tensioners. That should help keep the chain tensioned correctly, if you don't have too big of difference between longest and shortest chain runs.
09-05-2010 03:04
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EricH
How about this: Run the chain so that it is long enough to handle the longest dimension, and stick a couple of McMaster part 5896K1 in there as tensioners. That should help keep the chain tensioned correctly, if you don't have too big of difference between longest and shortest chain runs.
|
Do you have any experience with whether or not these fall out or how they work? I don't see how they could tension in a situation like this.
09-05-2010 03:08
EricH
Re: pic: Prototype Drivetrain v2
Not these--they go in the chain, and don't contact the robot frame at all. Should be negligible effect on the suspension.
09-05-2010 03:19
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EricH
Not these--they go in the chain, and don't contact the robot frame at all. Should be negligible effect on the suspension.
|
09-05-2010 11:10
Edoc'sil
Re: pic: Prototype Drivetrain v2
Forget the suspension if you are not doing two things:
Driving over rough terrain like the bumps
Powering all wheels
You speak of a "smother ride" and I really do not understand what it will benifit you. If you are concerned about your electronics mount them on a foam surface, or some other kind of protection. If you don't want to jostle the rest of your bot, well, almost all of the jostling comes from collisions, not deceleration from turning and stopping.
Also if your bot cam into any kind of pushing match it would distinctly be in the featherweight class. Not only will the omnis make it so that you well be easily spun, the 4 useless wheels will give you significantly pushing power. On your v1 thread you mentioned that you mechanical guy said he could put 75% of the weight in the center. I do not see how that is possible when you don't know what next years challenge is.
Here is one way to keep the chains in tensioned
http://www.flickr.com/photos/47595136@N03/4592232432/
http://www.flickr.com/photos/47595136@N03/4591612471/
http://www.flickr.com/photos/47595136@N03/4591612485/
Power the middle wheel run chain to the other axles, and then add a suspension to the angled bars with closed pneumatics.
09-05-2010 11:21
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Edoc'sil
Forget the suspension if you are not doing two things:
Driving over rough terrain like the bumps Powering all wheels You speak of a "smother ride" and I really do not understand what it will benifit you. If you are concerned about your electronics mount them on a foam surface, or some other kind of protection. If you don't want to jostle the rest of your bot, well, almost all of the jostling comes from collisions, not deceleration from turning and stopping. Also if your bot cam into any kind of pushing match it would distinctly be in the featherweight class. Not only will the omnis make it so that you well be easily spun, the 4 useless wheels will give you significantly pushing power. On your v1 thread you mentioned that you mechanical guy said he could put 75% of the weight in the center. I do not see how that is possible when you don't know what next years challenge is. Here is one way to keep the chains in tensioned http://www.flickr.com/photos/47595136@N03/4592232432/ http://www.flickr.com/photos/47595136@N03/4591612471/ http://www.flickr.com/photos/47595136@N03/4591612485/ Power the middle wheel run chain to the other axles, and then add a suspension to the angled bars with closed pneumatics. |
As for the weight distribution, that 75% was assuming the CG was in fact in the geometric center of the robot, and it usually is pretty close to that.
This whole concept may never get built. This is just to inspire the younger designers that I'm training over the summer and to provide us with a jumping off point for next year's challenge. If there is a surface feature like a large washboard or uneven, low steps going up to some important game piece, the suspension will be useful.
09-05-2010 12:11
Edoc'sil
Re: pic: Prototype Drivetrain v2
Your team doesn't like pneumatics? That makes life harder. The same idea works with springs, they just may wear out and cannot be adjusted.
09-05-2010 12:13
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Edoc'sil
Your team doesn't like pneumatics? That makes life harder. The same idea works with springs, they just may wear out and cannot be adjusted.
|
Any ideas on how I could tension the chains in the current configuration?
09-05-2010 12:33
Edoc'sil
Re: pic: Prototype Drivetrain v2
It seems foolish to throw away a useful tool.
And simply no, I do not see a good way to tension it like you have it set up now, mainly do to the fact that springs in a tensioner will allow chains to slip.
Something like this could be converted into that tensioner from mcmaster mentioned earlier. I have never used this style to tension, I have no idea how they work. http://www.mcmaster.com/#sprockets/=70knoq
The link isnt working here is the pn 6793K19
09-05-2010 12:47
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Edoc'sil
It seems foolish to throw away a useful tool.
And simply no, I do not see a good way to tension it like you have it set up now, mainly do to the fact that springs in a tensioner will allow chains to slip. Something like this could be converted into that tensioner from mcmaster mentioned earlier. I have never used this style to tension, I have no idea how they work. http://www.mcmaster.com/#sprockets/=70knoq The link isnt working here is the pn 6793K19 |
09-05-2010 14:54
LLogan
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
We haven't had any issues with handling, but why not investagate this in the off season.
With springs tensioning the suspension, it will rock only as far as it needs to. Using a diff. eq. we can figure out the specs on the spring. Ideally, at rest, the spring keeps the wheels exactaly level. Once the robot starts moving and the CG "shifts" the springs will expand cushioning any sudden blow (like a rapid acceleration). Additionally, if this robot starts to get flipped, all wheels can stay on the ground providing some traction.The front ones would lift off the most, increasing the Fn on the rear and center That still leaves the problem of chain runs though. I'm not sure a slot would have low enough friction, plus the exterior has to be smooth for the bumpers. |
I am unsure of why you are so set on a suspension.
|
Originally Posted by Edoc'sil
Forget the suspension if you are not doing two things:
Driving over rough terrain like the bumps Powering all wheels |
My advice: a suspension is not worth the hassle (and relatively useless) unless you are powering all wheels, or going over a surface that is not flat. To compensate for the sudden stops and starts, perhaps you could try to scale the acceleration and turning to your discretion through programming.
09-05-2010 15:15
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by LLogan
The bolded part does not entirely make sense for what you want to do. There is absolutely no point in keeping all the wheels exactly level if only the center ones are driven. Your rear wheels will not be driven, so it's quite useless to increase the normal force on them. In fact, it's detrimental because you're decreasing the normal force that is on the center wheels.
I am unsure of why you are so set on a suspension. The whole point of a suspension is to make sure that the normal force is distributed as evenly as possible. Doing this is only advantageous when each wheel is powered, seeing as your tractive force is based on the sum of the forces of friction and thus the normal forces. My advice: a suspension is not worth the hassle (and relatively useless) unless you are powering all wheels, or going over a surface that is not flat. To compensate for the sudden stops and starts, perhaps you could try to scale the acceleration and turning to your discretion through programming. |
. I'm currently rendering it, and will post it once it's done. This is just in case we don't have a flat surface next year. If we do, it'll most likely get nixed.
09-05-2010 15:24
kgzak
Re: pic: Prototype Drivetrain v2
I personally like the suspension even if we have a flat surface. Last year the field was flat but at nationals we kept having issues with one or two wheels not touching the ground because of our pick up mechanism which had to touch the ground. There were bumps in the actual field which caused our pickup to touch the ground but our front two wheels did not. I am designing a suspension so this does not happen to us agian. We were using 4 Wheel Drive which is different but the idea is still the same. My design will be for mecanum.
09-05-2010 15:34
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by kgzak
I personally like the suspension even if we have a flat surface. Last year the field was flat but at nationals we kept having issues with one or two wheels not touching the ground because of our pick up mechanism which had to touch the ground. There were bumps in the actual field which caused our pickup to touch the ground but our front two wheels did not. I am designing a suspension so this does not happen to us agian. We were using 4 Wheel Drive which is different but the idea is still the same. My design will be for mecanum.
|
09-05-2010 15:38
kgzak
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
I'm not sure if you're interested, but what I did to solve the solution may be applicable to you guys too. Feel free to be inspired by it.
|
09-05-2010 15:44
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by kgzak
Well I was going to have the motor move on the suspension as well but your idea might be better.
|
09-05-2010 15:50
kgzak
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
I'm not sure with mecanums, cause you need one motor/wheel, and mine is designed for all of the wheels on one side to spin the same way.
|
09-05-2010 22:17
EricH
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
Yeah, I realized that after I posted. How would these work in this situation? If they tension, how do you get them in? Do they allow for the variable chain lengths?
|
09-05-2010 22:19
Hawiian Cadder
Re: pic: Prototype Drivetrain v2
woah, if that only weighs 30 lbs then something is wrong, a cim in a 1-12 bane bots gear box weighs 12 lbs, there is no way 2 cims and the whole rest of that is under 6 lbs
09-05-2010 22:25
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Hawiian Cadder
woah, if that only weighs 30 lbs then something is wrong, a cim in a 1-12 bane bots gear box weighs 12 lbs, there is no way 2 cims and the whole rest of that is under 6 lbs
|
There is also a new post with the tensioning solution for you guys to look at.
09-05-2010 23:30
Jack Jones
Re: pic: Prototype Drivetrain v2
| I'm not sure how to power the omni's when they're in a suspension. |
Confucius say: Man who think physics is religion must sit in own pew.
09-05-2010 23:40
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Jack Jones
Lengthen the omni wheel swing arms and pivot them on the center drive axels. That way the distance between centers remains constant as the arms rotate. Use slot mounted dead axels on the omnis to tension the chains.
Confucius say: Man who think physics is religion must sit in own pew. |
http://www.chiefdelphi.com/media/photos/35715?
Already got the tensioning down. I prayed to the physics gods, and they answered!!
10-05-2010 00:26
sdcantrell56
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Jack Jones
Lengthen the omni wheel swing arms and pivot them on the center drive axels. That way the distance between centers remains constant as the arms rotate. Use slot mounted dead axels on the omnis to tension the chains.
Confucius say: Man who think physics is religion must sit in own pew. |
10-05-2010 00:33
548swimmer
Re: pic: Prototype Drivetrain v2
http://www.chiefdelphi.com/media/photos/35715?
10-05-2010 00:34
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by sdcantrell56
Better yet, remove the suspension as it is a solution to a nonexistent problem and greatly complicates things as well as adds more points of failure. There is a lot to be said for simple well thought out designs. Look at any drivetrain from 254/968 for inspiration
|
10-05-2010 00:35
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Jack Jones
Lengthen the omni wheel swing arms and pivot them on the center drive axels. That way the distance between centers remains constant as the arms rotate. Use slot mounted dead axels on the omnis to tension the chains.
Confucius say: Man who think physics is religion must sit in own pew. |
10-05-2010 06:26
Jack Jones
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
There is no good way for 2 arms to mount on one axle evenly though.
|
10-05-2010 11:01
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by Jack Jones
If you say so.
|
10-05-2010 21:16
EricH
Re: pic: Prototype Drivetrain v2
You could do a dovetail mount. It would take some machining, but it would probably work. Something like the image below (= is the axle).
__ __
=| ||=|| |=
10-05-2010 22:24
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EricH
You could do a dovetail mount. It would take some machining, but it would probably work. Something like the image below (= is the axle).
__ __ =| ||=|| |= |
10-05-2010 23:29
EricH
Re: pic: Prototype Drivetrain v2
*curses at VB's not putting spaces in right*
If you've ever seen a dovetail joint, it should be fairly simple to figure out. Look at a standard door hinge; that's a fairly common dovetail application. Now replace the pin with an axle, and the screw plates with swing arms. You'd need to change the geometry and check the stresses, but that's pretty simple to do.
10-05-2010 23:45
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EricH
*curses at VB's not putting spaces in right*
If you've ever seen a dovetail joint, it should be fairly simple to figure out. Look at a standard door hinge; that's a fairly common dovetail application. Now replace the pin with an axle, and the screw plates with swing arms. You'd need to change the geometry and check the stresses, but that's pretty simple to do. |
/_ or _/-_/-_/ with the dashed part cutout and bottom line at the top
10-05-2010 23:48
EricH
Re: pic: Prototype Drivetrain v2
|
Originally Posted by 548swimmer
Door hinges aren't dovetails, are they? I always thought dovetails looked like this:
/_ or _/-_/-_/ with the dashed part cutout and bottom line at the top |
Either way, I was referring to the door hinge style.
10-05-2010 23:58
548swimmer
Re: pic: Prototype Drivetrain v2
|
Originally Posted by EricH
It can go either way, straight or angled. I think the technical term for a straight dovetail is a mortise and tenon.
Either way, I was referring to the door hinge style. |
So I would make a tenion joint between the carriage and the shaft? That doesn't make sense... Or between the carriage and the side plate?
I think a better graphic would help the most honestly, I don't want to place any extra work on you though, but if you could I'd love to see your way of fixing this problem.
11-05-2010 01:13
EricH
Re: pic: Prototype Drivetrain v2
Door hinge is still the best description I can come up with.
For modeling, take a door hinge and replace the plates with arms of the length you'd need. Replace the pin with the axle.
Note that some geometry would probably need to be changed to accommodate robot dimensions and stress.
view entire thread
Tags
loading ...