From the last CD threads on this topic, it seems as if the responses were a bit dated so I figured I’d ask the question again, with the idea that in 2016 we might have similar field topography to that of 2014.
With equal weight (normal force), why do teams choose:
6wd pure tank
6wd drop center (WCD)
8wd
…when their resistivity to movement and therefore traction are all the same? If the point of WCD is to emulate a four wheel movement system, why don’t teams simply resolve to using four wheels in contact games like 2014? I understand that motor performance changes (or it used to in 2005, according to the old CD threads), but considering the current wheels we have like colsons, what really is the difference? Some of the older responses said that 4wd drives straighter but has a greater impact on current draw than a distributed 6wd.
And to clarify, my idea of a 4wd is a gearbox chained or belted to the two driving wheels, not individually ran wheels like mechanum.
Please read slides 63-74 of this PowerPoint: http://www.simbotics.org/files/pdf/drivetraindesign.pdf. That should do a good job of explaining why turning with a 4 wheeled drive doesn’t often work. Doing six wheels with no drop would run into the same issues (yes, I know 25 is the exception…).
Also, WCD /=/ 6 wheel drop center. A WCD is a specific, popular way of building a 6 wheel drop center drive but there are other ways of doing it. Look at the AM14U2 for example.
An advantage of 8w over 6w that the PowerPoint doesn’t mention is rocking; with a 6w drop center your robot is constantly rocking back and forth (by a very, very small amount) between the front four and back four wheels contacting the ground. With an 8w drive, the 4 wheels in contact with the ground are almost always the middle four wheels, which are all dropped. Whether this is significant or not depends on the rest of the robot’s functions.
Wheelbase. Shorter wheelbase means easier turning. 6WD drop and 8WD (2-drop) are common because they provide a shorter wheelbase, while still maintaining the same trackwidth–turning is easier, and you don’t have to worry as much about losing traction or tipping as if you just built a smaller 4WD.
That being said, many of the 6WD and 8WD teams started doing it when the drivetrains were 36"x30" or 38"x28", when the high wheelbase/trackwidth ratio really showed up–as “bouncing” in many 4WDs. Once you have something relatively standard, why change it?
This post was more-or-less inspired by the old 4wd physics analysis posted a while back and I was just verifying if any of the old information is pretty much the same considering how much drivetrains have evolved since then.
Based on wheel base sizes, why do so many teams opt for the long chassis? In 2012 I feel that chassis shape would have taken a change for wide, but alas I was not around in FRC back then to see what the trends were.
This year (as rookies) we ran with the KOP chassis and had a significant rock. As our weight distribution ended staying with the part of the robot with the lift, we mainly stayed on the four front wheels the entire time. Has anybody experienced how bad the scrubbing is on long 4wd systems? And would have a 4wd been have feasible in 2014? Thanks for the input.
One other reason to go 4 or 8wd (4 center wheel drop) over 6wd (2 center wheel drop) is to avoid rocking on the center wheels. The center of gravity will be closer to the middle of 2 sets of wheels on an 8wd, wheras it will be directly on top of the center wheels of a 6wd (assuming you have a COG in the middle of the bot and non-offset center wheel).
It’s up to the team to decide whether this rocking is significant enough to merit attention. There have been very successful teams on both sides of that particular coin.
75-25 in favor of long, or larger. There were a significant number of wide robots, though, valued for their ability to fit on the bridges as part of a 3-robot balance. And they could have a wider intake.
The primary reason is going to be front-back stability. Again this goes back to the 28x38 base size, but if you’re running a long wheelbase, you’re less apt to drive right out from under yourself, as wide wheelbases can do fairly easily.
Your secondary reason? It’s easier to fit through a gap if needed.
This year (as rookies) we ran with the KOP chassis and had a significant rock. As our weight distribution ended staying with the part of the robot with the lift, we mainly stayed on the four front wheels the entire time. Has anybody experienced how bad the scrubbing is on long 4wd systems? And would have a 4wd been have feasible in 2014? Thanks for the input.
It’s bad. That’s all I’m going to say. I’ve been around a while; my first two years were 4WD long. (Then the team switched to 6WD drop, long.) But, it can be worked around by several methods, including tread patterns and careful driving.
As far as feasible in 2014… I’d be seeing that type as a goalie or inbounder, using limited motion other than for defense. Not so much on a finisher.
It might be good to summon the cow-signal and get someone who was on 1538 in 2013 to talk about their experience running what I believe they considered a 4-wheel west coast drive on their robot Daisy Thunder. (I think they drew from this well again in 2014? My memory at 2am exacpes me) I assume they were operating with something like a 15 inch wheelbase on that machine. On a fairly standard 6 wheel drop center drive from the old 38"x28" days you would probably see up to and around a 14.5" wheelbase.
The reason that you would not see a four wheel drivetrain that often is because it has a large scrub force compared to many other options in drive trains.
Scrub forces is determined by how far apart the front and back wheels are. I don’t know the equations but the smaller length your drive base has the less scrub force and the more manuverable it is.
It is because of scrub force that teams will use 6 wheel drop center. Wheen the middle wheel is lower than the rest it effectively gives you a drive base that behaves like it is half the lenght when it comes to performance but has all the stability of the full length.
If you have any more questions feel free to PM me and I will help as much as I can. I had a longer post I was going to submit but my phone deleted it.
Actually the main reason that a drop center 6wd turns easier is not the shorter effective wheelbase, it is the fact that the normal force and thus the scrub friction is lower on the outer set of wheels that are in contact with the ground assuming that the CoG is near the center wheels. A 4wd’s turning ability can be greatly improved by putting the CoG very close to one end because it also reduces the normal force and thus scrub friction on the set of wheels farther away from the CoG. Of course offsetting the CoG dramatically may not be feasible and it increases the likelihood of tipping.
Obviously physics hasn’t changed in a while, so lets see if we can use the conclusions of the paper to try to understand the justifications of using a six-wheel drive.
The “rules of thumb” listed at the end of the paper were:
Make the track width greater than the wheel base (LTW > LWB)
If possible, reduce the lateral friction coefficient while keeping the longitudinal
friction high (i.e., use holonomic wheels or choose a good wheel tread pattern).
Try to move the center of mass slightly away from the center of the robot. Use
caution to not move the COM far enough so that the robot becomes unstable.
Lets ignore #3 for now, for many teams the COG is set by all the other mechanisms on the robot, and modifying those to get better drive performance seems like a lot of extra work if we can improve the drive by just changing the drive. If you’re not at 119.9lbs, then you do have room to add some extra ballast and might be able to use this to your advantage.
#1 is exactly what a 6w accomplishes, it effectively cuts your wheel base in half while keeping the track width the same. So, why use #1 and not #2? First off, maximizing your longitudinal friction has several benefits with minmal drawbacks, mainly it’ll take more effort to make your wheels slip (which you don’t want to happen in most scenarios, such as pushing matches or rapid accelerations). However, part of friction is material selection. Naturally, if a material has higher longitudinal (forward) CoF compared to another on carpet, it probably has a higher lateral (sideways) CoF. This can be mitigated with various tread patterns, but trying to figure out therotically how the tread pattern affects these coefficients is diffifcult; you’re better off just testing for it.
I agree, wide was an extremely good choice in 2012. The first step to many of our pick lists in 2012 was simply crossing off all the long-orientation robots at the tournament.
For a long time, the rule of thumb with a 6w, drop center drive was to have about .125" of drop, sometimes a bit more for some treads that dug deeper into the carpet. However, this was with the bigger robot dimensions, when most people were 38" long. With the shorter robot frames we have now, I also think that a full .125" of drop is a bit excessive. I understand why the kitbot has a similar drop to that though; it is built to work for all teams, regardless of how they use it; erring on the side of a bit of rocking and being able to turn well is much better than not being able to turn well.
Not exactly - all other things equal, assuming a rigid frame, 6 wheel no drop drives of the same wheelbase length will turn about 1/3 better than a 4 wheel drive of the same wheelbase length. 25’s drivetrains aren’t magic - they just combined a nearly square wheelbase with the benefits of weight being supported by middle wheels. In the era of 28x28 frames, you might not need any drop at all. Of course, it’s easy to put a little drop in and just be safe about it, so unless you really need the resistance to turning and can spare the current of more turning resistance, there’s no reason not to drop your 6WD center. But I think at this point honestly teams are dropping the middle wheel out of habit / “we know it works” when the practice is to a small extent a relic of the 38x28 era.
I’ll pull up the link to the old CD whitepaper in a bit, but the section on tracked drivetrains / statically indeterminate drivetrains applies to this six wheel no drop case.
The big change that happened in (2013?) was the new frame perimeter rules. Prior that we had max dimensions in each direction and it meant you were either building a wide bot that was unstable front to back or a long bot that had a longer wheel base[1]. With the change to a 112" perimeter we were able to optimize differently. Example - 125 ran a 6 wheel flat drive in 2014[3] and had no issues turning because we were effectively 28x28, and when looking at actual points of contact, much wider than we were long.
2015 allowed teams to do whatever they wanted.
Personally, I hope to see a return to frame perimeter rules as it is both easier to inspect than transporting a box[2] and allows teams more flexibility to optimize their chassis shape.
[1] 25 Addendum - No secret sauce, their large wheels usually ended up with contact points roughly square (or close to it) as folks have said in this thread.
[2] That isn’t square or always spec’d right
[3] Correction - We started w/ .125 drop, shaved to .09 and then by end of the season we were effectively flat but our frame was also bent to heck because New England.
Alternatively, if you want 4wd, two omni’s and two traction wheels works pretty good chained tank. It’s a setup that worked well for 955 in 2014 under defense.
I had always thought that yoy had 6wheels with four traction and 2 omni guess i was wrong all this time. It was a very well built drive train either way and you drove it really well.
In this situation, your forward or direct traction would be similar to that of a standard four wheel drive, but laterally against the faces of the wheels, your friction force would be about half. It also changes your center of rotation, but making a 6wd with two front omnis would be a good compromise: wide wheel base for 4 traction wheels and the stability of a 6wd.
We used this setup in 2013 and loved it. In 2014 we switched to 8wd tank, with no drop but the far front and far back wheels being omnis. We hated it. Partially because our gear ratios were wrong and we had misaligned chains, but it didn’t work for us – we tripped breakers, popped chains, and still couldn’t push very well. This year we went back to the 2013 setup (back traction, front omni) and again loved it. The problem is finding (non-custom) gearboxes to direct drive wheels that bolt on easily to 1x2" aluminum frames. Back in 2012 I was on a team that build a wide 4wd traction setup, which worked very well for bridge balancing, but “hopped” when it tried to turn and drive at the same time. Being wide made the drive better than 4wd long, but it still wasn’t great.
We’re getting ready to build a “Texas Tube” drivetrain – 6wd drop center, chain in tube – setup in the coming weeks, and may well run that in 2016, pending the game. The 6wd benefit over the 4wd traction/omni setup is traction, especially in turning situations with defense. We avoided most defense in 2013 (and our driver was insane – he did spin-moves around other robots, it was crazy), so it wasn’t a problem, but it can be.
I wouldn’t recommend building a drivetrain for the first time ever during build season if you can help it. You never know that kind of problem might pop up, or how it will perform for you until you’ve tried, and it’s far harder to rebuild a chassis and drivetrain after the robot it complete than during the offseason.
Thank you! We ran the 4wd for all PNW events, then swapped to 6wd at worlds. I can say with confidence that for 2014 (and 2013) we preferred the 4wd, 2omni, 2-tank. Friction pins were much less of an issue, and having the pivot point at the front (or rear) of the robot was nice. It does require a bit more driver skill when turning, but its easily learn-able in an hour or so of drive time.
Major downside was the limited traction when pushing, but you can adjust play-style to compensate.
Depends how you do it. You could do an “opposite corner omnis” 4WD, which has been done before–say you have an omni on your left front corner, you put another on your right rear corner. It has a similar effect in terms of friction reduction, but it keeps your center of rotation in the center of your robot. Now whether that’s a good idea depends on your robot and the game–you might actually want to pivot elsewhere!
I think the 112" frame perimeter rule was an evolution to FRC design that I’m pretty sure was universally lauded. I really hope to see that return in 2016.
My biggest regret in 2014 was not running a flat 6-wheel because with a robot would have had a wheelbase of around 22" that year, I think. The reason I cited Daisy Thunder is because a Hall of Fame team took the opportunity with the 112" rule in 2013 to develop a 4 wheel drive and looking at results, they must have liked it well enough.
There is a lot of inertia when it comes to FRC strategic design. That’s mostly a good thing. Now is a good of time as any to go out (or in… to your shop) and build and test something like this and see how it goes. In fact, I’m probably going to look into crashing something like this together!