Advantages/Disadvantages of West Coast Drive

[jfox100]

Hello Mentors/Teams/Students. I'm hoping to learn more about the advantages and pitfalls for a West Coast Drive type of chassis/drive train compared to the traditional drive train. Which of you have recently transitioned from traditional drive train to WCD? Please share your experience.. the good, the bad and the ugly!

What are unique design challenges? Maintenance issues? Challenges with buying WCD parts during the rush of build season? Cost? Reliability? Since we dont yet know what the 2017 game: is WCD drive better for certain playing field surfaces but perhaps not-so-good for other surfaces? Discuss some of the varoius versions of WCD.

Anxious to hear your thoughts!

Thanks

Jim

[Isonine]

What would be your definition of a traditional drivetrain? I thought that WCD was the traditional drivetrain...

[SenorZ]

What would be your definition of a traditional drivetrain? I thought that WCD was the traditional drivetrain...

Maybe 6-wheel drop-center KoP drive is traditional?

[D.Allred]

Hello Mentors/Teams/Students. I'm hoping to learn more about the advantages and pitfalls for a West Coast Drive type of chassis/drive train compared to the traditional drive train. Which of you have recently transitioned from traditional drive train to WCD? Please share your experience.. the good, the bad and the ugly!

What are unique design challenges? Maintenance issues? Challenges with buying WCD parts during the rush of build season? Cost? Reliability? Since we dont yet know what the 2017 game: is WCD drive better for certain playing field surfaces but perhaps not-so-good for other surfaces? Discuss some of the varoius versions of WCD.

Anxious to hear your thoughts!

Thanks

Jim
First and foremost for me, WCD is a tube and gusset construction technique that is handy for making the full frame. VersaFrame as a system makes it simple without needing complex tooling.

Making bearing blocks used to be the toughest part for us. Again VersaFrame COTS parts have fixed that.

Mounting the gearboxes is the next challenge. There are COTS solutions for some of the drive train gearboxes. We have a milling machine to make custom mounting plates for any gearbox we use.

Design challenges: You can quickly design your base frame shape to fit the scoring section configuration. Cut and rivet the tubes... (to borrow a phrase) Boom. Done.

Maintenance: Wheels are easy to change if your tread wears out.

Buying parts: We commit to WCD style before seeing the game and buy tubing, gussets, and bearing mounts before the season starts. We've pondered swerve drive but haven't yet committed the resources to develop one.

Surface: I usually consider WCD a 6 or 8 wheel skid steer system. There are several variations on this theme. However, the wheels don't care which construction technique you use. Skid steer driving performance comes down to designing for scrub / traction forces, good construction, good control, and good weight distribution.

David

[connor.worley]

Pros:

Reliability. If you have a 6-wheel WCD, you drive the center wheels directly off the output shafts of your transmissions. Even if every chain to the outer wheels fail, you should still be able to drive your center wheels.
Maintainability. Because the wheels in a WCD are cantilevered, removing a wheel is as easy as removing a snap ring. No need to take apart your frame to get at a wheel.
Simplicity. No outlandish parts or software required to use.
Drivability. Driving a good 6-wheel WCD is incredibly smooth, thanks to the dropped center wheel.

[mman1506]

I've found a good WCD to be far more rigid than other drivetrains. This gives it better, more consistent driving characteristics and allows you to run less drop.

[GeeTwo]

While my team has never built what I would consider a true WC drive train, we have incorporated at least one element of WCD in all but our rookie year. First of all, my understanding of the definition of WC vs traditional skid steer:

Common elements:

Both traditional and WC normally have 6 simple (non-roller) wheels with a drop center, though variants of each with additional wheels and/or omni on the corners exist.
There is no notable difference in software required; you have two banks of motors, one on either side. Either may have one, two, three, (or occasionally more) motors per side.
Either can be used with a single speed or shifting gearbox.
[Edit]Any articulated drive train does not properly fit either category, though an articulated gearbox may be extended from either design.


Distinctions:

West coast drives feature cantilevered drive shafts; traditional has a bearing on either side of the wheel.
West coast drives provide direct drive to at least one wheel on either side, usually (if not always) the one closest to the center of gravity of the robot; traditional usually had chain or belt to all wheels, though direct drive is now common in non-WC drive trains, including the 2014-2016 KoP chassis.
WCD usually employs tube (though occasionally channel with the opening on top or bottom) in order to support the cantilever; traditional usually employs channel with the opening inboard or outboard, or vertical plate, most commonly bent inboard or outboard at the top and/or bottom.


Note that there are also drives which incorporate these features mix-and-match. The common example which comes to mind first is the (recently discontinued?) Andy Mark nano tube chassis. It has direct drive on four cantilevered axles, but these are the two corners, NOT the one usually nearest the CoG. Our second and third year drive trains also featured direct-drive wheels on cantilevered axes, but for similar reasons were clearly not WC drive trains [Edit: Our 2013 robot only had two driven wheels and two idle omnis, whereas our 2014 robot had mecanum drive; we learned a lot from these mistakes].

Advantages of WC:

Easier to change wheels by removing a single lug screw, snap ring, shaft collar, or clip.
Reliability despite failure of chain, as the axle nearest the CoG is direct driven (though this feature is now on some newer "traditional" drive trains)
Track width can be a greater proportion of robot width, as there is no need for an outer plate/channel (other than to meet bumper support rules)


Advantages of traditional skid steer:

You are much less likely to bend an axle. (I never saw an FRC axle with a bearing on both sides of the wheel bent until STRONGHOLD.)
Requires less precision in machining (though this can now be mitigated with dollars by buying a COTS WC chassis kit).
[in recent years] Available through the KoP chassis program at little or no dollar cost, as a kit which can be assembled in a remarkably small number of worker hours.

[ToddF]

Well, let me tell you all about our experiences at 2363!

Quick Build Drive Train Concept (https://www.youtube.com/watch?v=UD3xK9CC4L4)

Triple Helix West Coast Drive Trains (https://www.youtube.com/watch?v=vhy5nV88BBM)

[RoboChair]

Here is my only question for you, has 254 ever not done a WCD?

I'd bet $100 that 254 does a WCD next year, and another $100 that that is what we settle on as well.

It is quick, easy, and robust. Most of the design drawbacks for a particular game can be designed around with WCD without much effort, this year it was wheel spacing.

[Michael Hill]

One of the great things about it is that it's almost formulaic in design, which allows for super quick design and construction. Many of the problems have already been solved so you can focus on the harder elements of robot design. Obviously there are some exceptions (this year was a year of drivebase exceptions), but in a flat floor game, it becomes much easier.

[Nuttyman54]

The primary reason to build or not build a WCD versus a "traditional" skid-steer drivetrain is what your resources are. Some teams resources lend themselves to a more traditional drivetrain, some towards a West Coast Drive. It's worth noting that time is most definitely a resources. Probably the most important one. Just because you have the fabrication capability and money to make something doesn't mean it's worth time investment during the season if it's not an efficient use of your time.

The primary disadvantage of a WCD over a more "traditional" drivetrain is that it usually requires more precision machining (yes there are COTS options like the VersaBlocks that can let you build one with hand tools). WCD requires live axles and does not offer as many simple solutions for chain tensioning. Dead axles tend to be a little more forgiving, but nothing that can't be achieved with time and attention to detail. Bumper mounts also takes more thought.

The primary advantage is typically maintenance and access, especially with replacing wheels and worn tread. There are some marginal performance benefits that can be gained by putting your wheels a little farther outboard than usual.

Sheetmetal construction methods are also generally not well suited to making WCD either, which is why you see many teams with sheetmetal sponsors opting for other drivetrain styles.

It's worth mentioning that many of the challenges associated with building an effective WCD are obsolete due to new COTS options and the proliferation of exact C-C belts and chains. Machining the bearing blocks, developing a good chain tensioning method, wheel retention, ability to make custom wheels with hex bores and building a transmission with the correct output speeds on a live axle shaft are all aspects of WCD that are now available as COTS or able to be fabricated without a mill/lathe. These weren't available even a few years ago, which made WCD out of reach or not worth the resource investment for many teams. At the same time, it also means that many of the advantages of a WCD are now available to "traditional" drivetrains as well, so the lines are blurred a bit.

From a COTS standpoint, if you're not doing any precision fabrication yourself it's still likely cheaper to take your KOP chassis in the kit or purchase the VexPro 2014 Drive-in-a-day (both of which are "traditional" drivetrains) rather than spend money on the various bearing blocks, bearings, etc for a WCD.

Don't make drivetrain decisions in a vacuum, especially if your team already has a particular drive style that you like using. I'm not saying don't switch, but be sure you're considering what the cost of switching would be (time, money, learning curve, etc) and what problems you're solving by making a change. Is it cheaper? Maintenance? Better performance? Fewer or easier to make parts? Then figure out how switching to a different drivetrain style addresses that.

[Oblarg]

While WCD maintenance, on the whole, is quite nice, one very annoying aspect is access to the belts or chains, which generally are driven by pulleys/sprockets inside the gearbox. In the case of belts, the gearbox has to be taken apart for any maintenance that requires removing the belt or putting on a new one.

[notmattlythgoe]

While WCD maintenance, on the whole, is quite nice, one very annoying aspect is access to the belts or chains, which generally are driven by pulleys/sprockets inside the gearbox. In the case of belts, the gearbox has to be taken apart for any maintenance that requires removing the belt or putting on a new one.

Do you not run into this issue in other cases but with having to take wheels off instead?

[GeeTwo]

Do you not run into this issue in other cases but with having to take wheels off instead?

I don't recall having to take a gearbox apart to change belts or chains on a traditional drive train. Robot chassis, but not the gearbox. Then again, I have seen drive trains which were described as WC which had easily replaceable chains, and others where it was apparently impossible in an FRC pit period - or six. I would list this more as a "gotcha" of too-well protected chain and belt, not a WC feature per se.

[Oblarg]

Do you not run into this issue in other cases but with having to take wheels off instead?

Yes, but it's usually somewhat easier to slip an axle out than to take a gearbox apart.

[EricH]

Here is my only question for you, has 254 ever not done a WCD?
As a matter of fact, yes, I believe they have.

BEFORE the WCD was developed (2002-2004 timeframe), of course, but seeing as 254 is one of the developers of the WCD...

The WCD was officially unleashed in late 2003/2004 season, the result of a collaboration between 254 and 60. There are photos in CD-Media from 2001 showing both teams running elements of what is now known as a WCD, though 254 for some reason ran their chains outside their frame from a central gearbox that year (either one of which would, under the traditional definition, disqualify the drive as being WCD), and neither had more than 4 wheels on the ground (also disqualifying as a WCD under the traditional definition, which requires 6 (or more, depending just how traditional you want to get)).

So, more detailed answer: yes, but it's been quite a while and they were developing WCD when they did.


And now back to your regularly-scheduled discussion of the pros and cons of the WCD.

[Andrew Schreiber]

Here is my only question for you, has 254 ever not done a WCD?

I'd bet $100 that 254 does a WCD next year, and another $100 that that is what we settle on as well.

It is quick, easy, and robust. Most of the design drawbacks for a particular game can be designed around with WCD without much effort, this year it was wheel spacing.

Here is my only question for you, why does "254 does it" make something a good design decision for a team without 254's resources?

[dirtbikerxz]

Here is my only question for you, why does "254 does it" make something a good design decision for a team without 254's resources?

Because it does not take many resources to do a WCD, and by 254 doing it, they are proving that it is a good solution. Which means that is a good solution for not much resources. (Man my grammar was bad in that sentence.)

We are a low resource team, like really low resource, yet we do WCD with success.

[Billfred]

Here is my only question for you, why does "254 does it" make something a good design decision for a team without 254's resources?

Because it does not take many resources to do a WCD, and by 254 doing it, they are proving that it is a good solution. Which means that is a good solution for not much resources. (Man my grammar was bad in that sentence.)

I think our definitions of "not many resources" differ. I've been on teams where the build budget was maybe $1,000, the machining resources stopped at that of a disused wood shop, and the mentor expertise didn't really include machining. I would wager my dollars to your donuts that that situation describes More Than A Couple Teams. I would also say those teams are less likely to have an inspiring season if they reached for WCD without more of a grasp of the situation than "254 does it and does well".

[Andrew Schreiber]

Because it does not take many resources to do a WCD, and by 254 doing it, they are proving that it is a good solution. Which means that is a good solution for not much resources. (Man my grammar was bad in that sentence.)

Then that would have been a better argument than "254 does it".

I'd disagree with that argument since it's not a function of raw number of resources but more a function of what type of resources and the allocation of them. For teams with tight budgets and resources using the KoP or VP DIAD may be better uses of tight resources. I don't see how 254-copy describes advantages or disadvantages for the purposes of resource allocation.

[JesseK]

For me, it's either WCD or KOP. Either can be purchased and put together with nominal amounts of machining and programming work.

Presuming you use all-COTS for either drive train, these are IMO, of course:
WCD is easier to customize around an overall robot design since the frame consists of only 5-6 basic parts.
WCD maximizes robot stability by putting the wheels at the outer-most maximum dimensions
WCD makes it far easier to experiment with and find the right wheels - and then maintain them throughout the season
WCD makes dealing with chains super-easy. If a chain snaps (literally) because of a wonky new offseason defense (@Nate @Matt .... ;)) then a fix requires just a few parts of adjustment rather than removal of anything.



WCD reduces flexibility when it comes to bumpers. There are ways to design bumper mounts to make them secure, and most of them over-complicate the frame design when compared to the simplicity of WCD itself.
WCD can be done with rivets, but IMO it is far simpler to design and execute if a team has a welding sponsor. The middle cross rail usually drives me up the wall, unless we have a welding sponsor.
If welded, WCD frames are less forgiving if not properly cooled. The reliability of chains presumes the side rails are perfectly straight.
WCD setups have very few COTS "plug-and-play" options that are designed specifically for WCD. Most gearboxes used in FRC will require some sort of custom mount in order to direct-drive one of the wheels in WCD. This isn't an insurmountable issue to fix for an individual team. Yet don't expect to go to the 3 major FRC gearbox suppliers to find anything that adapts non-WCD gearboxes to a WCD setup.

[Bertman]

I find this discussion interesting from the standpoint of the definition of WCD, KOP, & traditional when related to past baseline definitions. They seem to vary with how long you have been around. These definitions are very fluid and I believe will continue to be so. Just a thought.

[MichaelBick]

WCD setups have very few COTS "plug-and-play" options that are designed specifically for WCD. Most gearboxes used in FRC will require some sort of custom mount in order to direct-drive one of the wheels in WCD. This isn't an insurmountable issue to fix for an individual team. Yet don't expect to go to the 3 major FRC gearbox suppliers to find anything that adapts non-WCD gearboxes to a WCD setup.


Both Vex and WCP make pretty much COTS "plug-and-play" gearboxes for WCD.

[frcguy]

Both Vex and WCP make pretty much COTS "plug-and-play" gearboxes for WCD.

http://www.wcproducts.net/media/catalog/product/cache/1/image/650x/040ec09b1e35df139433887a97daa66f/o/i/oie_2865057t1hlpbkx.jpg (http://www.wcproducts.net/wcp-ss-gearbox)

Example of said WCP gearbox with integrated sprockets and bearing block. If we go WCD next year, we'll probably be using these.

[Billfred]

Both Vex and WCP make pretty much COTS "plug-and-play" gearboxes for WCD.

And before I even thought about working at AndyMark, we made a Toughbox Mini work for a chain-in-tube-with-coulda-been-cantilevered-wheels (but weren't cantilevered because this game is brutal) by getting the flat front plate, using the long output shaft cut to length, and pressing a bearing into the tube. Done.

[RoboChair]

I think our definitions of "not many resources" differ. I've been on teams where the build budget was maybe $1,000, the machining resources stopped at that of a disused wood shop, and the mentor expertise didn't really include machining. I would wager my dollars to your donuts that that situation describes More Than A Couple Teams. I would also say those teams are less likely to have an inspiring season if they reached for WCD without more of a grasp of the situation than "254 does it and does well".

You can make a WCD these days with a hack saw, a hand drill, and some aluminum tube. The gear box and bearing block options from Vex make producing a WCD a trivial task these days as long as you can drill some straight holes. This is how 5458 made their drivetrain this last season. You can make a 6 wheel drive for around $450 depending on your choice of gearboxes(~$150 and up) and motors.

[JesseK]

Both Vex and WCP make pretty much COTS "plug-and-play" gearboxes for WCD.

To an extent. I wouldn't call what we have a plethora of options that are specifically designed with WCD in mind. For a given drive train design we have 1 great product and a few other adaptable choices. If a team can't make an adapter plate of some sort, there's really only 1 SS and 1 DS product.

AM - practically non-existent because the mount points interfere with the chain runs

WCP - yes! We've used these for 2 years now and love both the single and double-speed options. WCP-WCD-DS lasted through 85 matches in 2016 during the preseason, competition season and post-season with zero issues. Zero. None. We snapped quite a bit of chain (broken links), but the gearbox kept on chugging.

VEXPro -
Single Speed: The single-speed double-reduction looks like it will work for WCD as-is, yet I have some reservations about longevity. There are effectively about 5.5 lbs cantilevered off of 2 mount points through plastic housing. I'd trust such a setup for a single competition. Yet I have reservations for so many matches between March 1 and November 31 in a game like 2016 (with considerations for transporting the robot so many times). It would require additional support somewhere, either from adapter plates or a super structure. So IMO it's not quite plug-and-play in WCD. My team could easily get it done, but I'm not sure for a typical team.

There also seemed to be a very large sensitivity to assembly tolerances on the VP Ball Shifter. I heard of multiple teams who had gear teeth sheer on the 3rd stage due to slight misalignments, and one such team was our alliance partner in the Finals at their 2nd event. It was something manageable, but I know my team would have to consider how rough the game is on drive trains before purchasing ball shifters.

These are, of course, my perspective based upon what I know about my teams strengths and weaknesses.

[dirtbikerxz]

VEXPro -
Single Speed: The single-speed double-reduction looks like it will work for WCD as-is, yet I have some reservations about longevity. There are effectively about 5.5 lbs cantilevered off of 2 mount points through plastic housing. I'd trust such a setup for a single competition. Yet I have reservations for so many matches between March 1 and November 31 in a game like 2016 (with considerations for transporting the robot so many times). It would require additional support somewhere, either from adapter plates or a super structure. So IMO it's not quite plug-and-play in WCD. My team could easily get it done, but I'm not sure for a typical team.

There also seemed to be a very large sensitivity to assembly tolerances on the VP Ball Shifter. I heard of multiple teams who had gear teeth sheer on the 3rd stage due to slight misalignments, and one such team was our alliance partner in the Finals at their 2nd event. It was something manageable, but I know my team would have to consider how rough the game is on drive trains before purchasing ball shifters.

These are, of course, my perspective based upon what I know about my teams strengths and weaknesses.

We've used VEXpro for 3 years now, and it relatively pretty easy to do WCD. We use bearing blocks (http://www.vexrobotics.com/vexpro/all/bearingblocks-g.html), and all you have to do is turn the middle one upside down to get the required drop. This year we used this system, and the triple cim ball shifter (http://www.vexrobotics.com/vexpro/motion/3cimballshifter.html) and it worked like a charm. We supported the gearboxes by having a crossbar over the gearboxes and screwing into the mounting points on top of the plastic housing. Never snapped a chain, and just generally never had problems with the drivetrain at all this year. And we were definitely one of the more "violent" bots while crossing defenses.

[GeeTwo]

And before I even thought about working at AndyMark, we made a Toughbox Mini work for a chain-in-tube-with-coulda-been-cantilevered-wheels (but weren't cantilevered because this game is brutal) by getting the flat front plate, using the long output shaft cut to length, and pressing a bearing into the tube. Done.

While not canonical WCD, I bet that a couple of TB Micros with long hex shafts at each end of the drive train would interface with rectangular tubing even easier. I find myself liking the idea of using six TB micros and six CIMS on six wheels, with chains inside a 3"x1" tube to transfer torque among the wheels on each side as needed. Not really cheap, but much less than swerve, and eminently functional in case of chain failure. If I get to call the name, make it Gulf Coast Drive.

Sounds really heavy.

Yes. I wouldn't want to use it every year, but it might have been worth it to play Stronghold with a drive train that works almost as well even if you lose a chain. We could have spared the extra few pounds this year.

[roboruler]

While not canonical WCD, I bet that a couple of TB Micros with long hex shafts at each end of the drive train would interface with rectangular tubing even easier. I find myself liking the idea of using six TB micros and six CIMS on six wheels, with chains inside a 3"x1" tube to transfer torque among the wheels on each side as needed. Not really cheap, but much less than swerve, and eminently functional in case of chain failure. If I get to call the name, make it Gulf Coast Drive.

Sounds really heavy.

[JesseK]

Sounds really heavy.

We didn't something like this with toughbox nanos for our 2010 and 2014 robots. It wasn't that heavy, TBH. It wasn't as light as we possibly could have made it, but the weight difference didn't make or break us those years.

The toughbox nanos were mountable without alot of tooling, but I also wouldn't say the robot drove straight in autonomous.