cvt options

[NRBaer]

we have been throwing ideas around about trying to switch to a cvt for the 2009 and/or 2010 year.

In short, a belt-driven cvt (snowmobile) will require some pretty large pulleys and quite the packaging problem so I am looking more into other options. I'm looking at toroidal right now but please talk about all options.

The specific questions I am trying to get answered are:
1. feasibility/practicality - done in 1 year(approx. hrs to design)
2. cost - high-tech materials and lubricants
3. gear ratio range - sufficient to operate these cars
4. reliability - how significant is wear/typical hrs until rebuild
5. size - small enough to replace a typical 5-speed

Thanks,

Nick

[Mike Macie]

Gaged Engineering (http://gagedengineering.com/)

Very small and easy to package. Tell them U at Buffalo sent you and they will help you out.

[J.R.]

Definitely a feasible project, are you running a bike engine with a built in transmission, or something else?

[NRBaer]

We are keeping our options open at this point but if we did run a bike engine we would chop out the transmission. The Genesis 80FI snowmobile engine is more what we are looking at.

When you guys were designing your system did you look into other types of cvt's? Any idea as to the max torque capacity of the Gaged pulleys?

[J.R.]

One of the guys who was on our team a while ago (Rob Woods) mentioned JR Race Cars, but he said that they were not great with customer service. Also, I couldn't find it in a quick search on their website.

Also, I think Dearborn used to run a Baja clutch.

[rjwoods77]

NRBaer,

If you ran a 4 cylinder engine with transmission you would have to do way too much work to get it work be space and weight efficient to worth its while. If you are going to run a CVT then get an engine with a PTO output so that the primary will bolt on. That being said there are a couple choices out there. The Yamaha Phazer sled engine(parallel twin) would be a great choice if you want a water cooled engine. You would be a fool to run anything for a CVT then what comes on it stock. Drivetrain matching a CVT to an engine is very difficult if the clutch wasnt designed for it. Dearborn ran this engine and tranny combo on their car this year after switching from a Briggs V-twin and Comet 790 clutch. They took a step backward if you ask me and their result showed it this year. We had a 440lb car with about 35 hp and a cvt and took 18th at Detroit so you need to think very carefully about how much power you need for your goals. The team is still working to get 40-50hp out of the briggs with some decent cvt tuning the and some weight loss the car should be able to advance into the top ten. Dearborn was at this point before they switched so that is food to think about also. Not to say their potential was less but it is hard to beat a simplistic package with minimal problems. Also nothing packages like a briggs twin. Do some searching on this site and you'll find a link that I wrote on that had some CAD side views of how small it is. So...... if you were to run a Briggs v-twin then the gaged cvt is the best you can get. If you run a Yamaha twin then you should run the cvt with it. If you run a Polaris single then run the p-85 clutch that comes with it. Other than what I mentioned there are hardly any other PTO shaft engines out there that you can use and the couple that do exist you should/must use the cvt that comes with it.

[NRBaer]

Sorry if I wasn't clear with what I was saying with the 4 cyl. I meant that we would probably be converting it to a PTO output whether that means redesigning the bottom end or cutting it out. Yes it is very complicated and time consuming but it would allow us to make a tight package.

The Phazer engine is the Genesis 80FI, just to be clear, and the problem we see with it is the overall width of the system (24"-30", the dealer wouldn't let us take the sled apart too much)in comparison to the space we have in our cut-n-fold chassis.

That said, all of the cvt's that you mention are belt-driven. I'm hoping to explore other options that might allow a more flexible/smaller package for the 40-80hp and the 30-50 ft-lb range. When designing your system did you look into toroidal or stick to what is available in the market?

You had mentioned that "drivetrain matching...is very difficult if the clutch wasn't designed for it." Could you elaborate on this further?

[rjwoods77]

Here are a couple tasty nuggets from the past. I am an unabashed supporter of lawn mover engines owning wannabe Formula 1 kids so take that into consideration when you read stuff I have written on.

http://fsae.com/eve/forums/a/tpc/f/125607348/m/42510133...10473031#70510473031 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/42510133031?r=70510473031#70510473031)

http://fsae.com/eve/forums/a/tpc/f/125607348/m/20210647...10747341#27510747341 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/20210647341?r=27510747341#27510747341)

http://fsae.com/eve/forums/a/tpc/f/125607348/m/55310441...10082241#38410082241 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/55310441241?r=38410082241#38410082241)

Also I cannot even begin to calculate or explain how easy it is to drive a cvt car. It is JUST the same as driving a go-kart but with more power. Taking shifting out of the drivers concentration reserve just does unbelieveable things that cant really be quantified. With a properly setup up clutch it is a dream. When they are not then its like slitting your wrists. If I had to quantify what UB had theres at it would be about 60 percent right.

[rjwoods77]

All of the other types of cvts out there are too big and heavy for what you are trying to do. Everyone says dry belt friction cvts are inefficient but you make that back up on shift time and driver ease. Toroidals or any other high pressure traction fluid designs are big, heavy and not scaled to the right size. Wet metal belts cvt's (Jatco) have the same problem. You could always go with a double cone cvt like Aachen had since it is a perfect engineering marvel that is way better than a belt cvt. Just kidding. That was a big joke that goes way back. Dont even look at that rattle trap. I dont know how you think you will package a inline 4 with cvt if you cant package a parallel twin. You could cvt a 4 cylinder but you could do much better with a lower power engine that is much smaller. Even if you did some goofy gear drive to offset the cvt to the front or rear of the engine is just adding more shit for not much gain. My comment is to use an exsisting engine/cvt combo or use a PTO engine and use a gaged. I would be more than happy to help you with questions if you do that. If you go the "custom because I wanna be cool" route then I wont be much use since I will just keep telling you that you are dumb for doing it. Go with the phazer or the polaris engine and you will have a dead reliable setup. By matching I mean the internals of the cvt that make it work good are tuned for that particular engine which means to make changes to it isnt hard since you wont be changing it much. Do match an engine and tranny that were designed together is a nightmare. It took mini-baja teams about 8 years of trail and error just to get the Polaris P-90's working right on their engines. If you have to ask what drivetrain matching is then you should definitely use a setup meant for each other. Let me know what I can do to help but please think twice about 4 cylinder with a cvt. It's just a bad look homie. Just a bad look.

This is a neat setup but that particular cvt wont hold dick for power in a car FSAE sized. I absolutely love the intergral planetary.

http://www.karting1.co.uk/continuously-variable-transmi...n-for-karts-test.htm (http://www.karting1.co.uk/continuously-variable-transmission-for-karts-test.htm)

[NRBaer]

Thanks for the post, I had completely forgotten about the utility vehicles.

[vreihen]

Originally posted by NRBaer:
The Phazer engine is the Genesis 80FI, just to be clear, and the problem we see with it is the overall width of the system (24"-30", the dealer wouldn't let us take the sled apart too much)in comparison to the space we have in our cut-n-fold chassis.

Would it be possible to package the engine turned 90 degrees? If you put the CVT's driven pulley onto the input shaft of an automotive-style rear differential and can get the crankshaft pulley into that belt plane, it may just fit inside the narrow chassis:

(broken image link removed)

This is from a snowmobile engine powered dragster, and the engine has apparently been raised to get the crank pulley above the diff pulley. Maybe you can use a jackshaft or go with the pulleys side by side?

Not exactly the most elegant solution (especially raising the engine! http://fsae.com/groupee_common/emoticons/icon_eek.gif ), but one packaging option to consider if for no other reason than to rule it out.....

[rjwoods77]

You could do anything if you put your mind to it. Sherbrook and Bradly universities both use 4 cylinder engines with cvts. Sherbrook I believe designed a cover that bolts on to the side of the engine that makes the primary coaxial to the crankshaft. Bradley takes off of the trans output I think. If you look on youtube bradley posted a video of the cvt operation. Sherbrook style allows you to point the exhaust out of the back because the engine is rotating in the right direction To be honest the packaging sucks with a cvt and a 4 cylinder any way you cut it. You dont need that much power in an FSAE car and you'll end up with a car heavier than it needs to be. Stick to the parallel twin, v twin or single. You'll never beat the really good teams at their own game so you have to beat them by going in a different direction which is simple and lightweight. Longitudal engines with a 90 degree gear reduction are not efficient weight,space or friction wise I have seen a couple and honestly dont see the point for all the extra length and work just to make the rear of the main roll hoop a little more narrow. Look at the link where I have the drawing of our car with a single and the briggs twin. The dimension to our solid axle of 19 inches is also what you can fit given about a 4.5" diameter diff housing. Food for thought.

[rjwoods77]

I just re-watched that kart cvt video. I love the the guys commenting. "You know I'm a bit of a braker."

I so wish I could use this motor for the formula comp. It weighs 35lbs loaded and it is 300cc. Shit at that weight I would run two of them. One for each rear wheel with a kart cvt like in the other video. Listen to the thing on the straights.

Torque and HP Curves:

http://www.woelfle-engineering.com/GB/Produkte/Leistungsdiagramm.xls

http://www.karting1.co.uk/wankel-kart-test.htm

[rjwoods77]

bump

[D J Yates]

Okay, i give up, this has been dangling here for too long not for me to comment.


Originally posted by rjwoods77:
...Toroidals or any other high pressure traction fluid designs are big, heavy and not scaled to the right size...

Rob, are you sure?

[rjwoods77]

DJ,

Yeah I should probably correct that statement. My comments were based on the existing toroaid designs that I have seen which are mainly the nissan setup. I am sure with enough money and time a better lighter system could be designed just like anything else out there. But in terms of right now, off the shelf and correctly sized setups I havent seen one other than a dry belt setup. Prove me wrong please because I would love to see more options on the table. I see on your website that you guys are working on some stuff that is smaller in nature.

I will also say this though about dry belts setups. For some of theri disadvantages they sure as hell make up for it in other things. A purely mechanical setup that is very easy to chnge once you understand what is going on is hard to beat for this type of racing. Another huge advantage is the packaging the system we use. I dont think there is a more efficient longitudal size package than what you can achieve with current off the shelf jr drag and snowmobile stuff. Not that we do it this way but you can do all sorts of interesting stuff with cvt packaging.

http://www.ritzsite.demon.nl/DAF/DAF_cars_p17.htm

Just for comparison purposes lets talk turkey. What compareable system that you guys make or can make meets these numbers that our Gaged Engineering cvt has.

4:1 low
1:1 High
7.5" OD Secondary, Weighs 2.7 lbs
5" OD Priamry, Weighs 4 lbs.
9" Center to center distance.
Hold up to 60ish hp.

Also you guys are doing the KERS system but have you toyed with the idea of a toroidal F1 tranny? I know current rules ban CVT's because they just wouldnt be fair but with the trend of turning F1 cars into superfast/super efficient Prius' it would make sense for them to allow cvt's back in the future. I am sure if you guys were working on something that your wouldnt talk about it anyway but just a thought none the same.

Also I know it is a different animal all together but here is someone who is messign around with stuff here in Rochester. They even helped RIT with the diff they used that if above and beyond what else is out there. Type 1 torsen without the wheel lifting power loss problem with the huge TBR that type 1 had and beyond.

http://www.torvec.com/products_ivt.html

[mech5496]

Well, back from the dead thread (and I am sorry about that) but this seemed the most appropriate place to post on from the few the search function returned. Anyway, I "accidentally" found this (http://cvt.com.sapo.pt/MCVT/MCVT.htm) and this (http://people.bath.ac.uk/enscjb/milner.pdf), and seemed interesting to me, but I would like an opinion of someone with more expertise on the subject. ;)

(P.S. The ATV version (http://www.orbitaltraction.com/atvs.html) of the Orbital traction sounds promising for FSAE...)

[mech5496]

Well, doing a bit of a research lately on transmissions, and I found out that the new 2012 Honda Four Trax Foreman Rubicon uses Honda’s Hondamatic hydrostatic transmission, which seems nice to source parts for a hydro-planetary IVT.

2012 Honda Four Trax Foreman Rubicon Specifications:Engine Horsepower: 19.9KW/6500RPM
Torque: 37.8Nm/5000RPM
Hondamatic Working Pressure: 35.6MPa (max 60MPa)

The same model uses also electric PS and has a diff which seems nice for FSAE too... :D Really nice donor, aint it Z?

[Z]

Well, doing a bit of a research lately on transmissions, and I found out that the new 2012 Honda Four Trax Foreman Rubicon uses Honda’s Hondamatic hydrostatic transmission, which seems nice to source parts for a hydro-planetary IVT.
...
Really nice donor, aint it Z?

Harry,

Thanks! This is the first time I have heard of these Hondamatic hydrostatic transmissions. I know there are many similar (and different!) versions on all sorts of tractors, but never knew Honda fitted them to their ATVs. Well done Honda!

The whole package seems very well suited to FSAE conversion. The 500cc OHV single with dry-sump (with built-in tank in engine nose) gives a low overall height and CG, plus the in-line drive to the hydrostatic trans (ie. longitudinal crank), means that it might (?) be possible to very compactly package it laterally in an FSAE car, with either gear or chain drive to the diff. (I think keeping it longitudinal would make it too long.)

It would probably be necessary to take a hacksaw or angle-grinder to parts of it to make it really suitable for FSAE, but I reckon these types of ATV powertrains are a much better starting point than the "standard FSAE sportsbike engine".

I managed to find quite a bit of info on the web, but no good pictures of the engine/drivetrain layout without all the bodywork. If anyone can find any good pics (or links) of it "naked", then please post! :)

Z

[mech5496]

Not quite a photo, but I uploaded the actual patent (with some rough drawings of the general layout), you can find it here (https://docs.google.com/file/d/0B9AIXdZgarikdEFwdGtqR05Hazg/edit?usp=sharing). Moreover I found some pictures of a disassembled transmission. You can see the fixed swash plate (http://www.hondaatvforums.net/forums/attachments/honda-rancher/10198d1349099599-some-great-hondamatic-unit-info-pics-img_0448.jpg), adjustable swash plat (http://www.hondaatvforums.net/forums/attachments/honda-rancher/10199d1349099624-some-great-hondamatic-unit-info-pics-img_0449.jpg)e, adjustable swash plate (http://www.hondaatvforums.net/forums/attachments/honda-rancher/10200d1349099715-some-great-hondamatic-unit-info-pics-img_0451.jpg) showing the rectangle slot where angle sensor attaches via joint shaft, swash plate worm gear drive (http://www.hondaatvforums.net/forums/attachments/honda-rancher/10201d1349099749-some-great-hondamatic-unit-info-pics-img_0454.jpg) and ball nut, and finally a cutaway view (http://www.hondaatvforums.net/forums/attachments/honda-rancher/10202d1349099847-some-great-hondamatic-unit-info-pics-tran22.jpg) showing inside parts with another good view of what is controlling the angle sensor (small rectangle slot in middle of pic). All and all it seems like a nice small unit, no idea about the weight though.

EDIT: Found another cutaway picture (http://world.honda.com/motorcycle-technology/hondamatic/image4.html) of the engine/transmission assembly, and it seems rather compact... Apparently Honda uses them since 2001, here's (http://www.hondaatvforums.net/forums/attachments/repair-maintenance/1322d1285591259-hondamatic-continuously-variable-transmission-operating-principles-dsc03521.jpg) an earlier version.

EDIT 2: Well, seems that Honda Ranchers used the same (or almost the same) transmission too! Plus side, more spare parts! Minus side, I found out that the 2004 Racher 400 powertrain weighs some 50kg...

[Z]

Harry,

Thanks. I also found many other photos, plus some tortured explanations of how it all works. I reckon I know how it works now, and I like it!

It works as a CVT, so it needs a separate centrifugal clutch to disengage the engine at low vehicle speeds. This is unfortunate since many of the hydrostatic drives work as IVTs, so do not need a clutch. Nevertheless, the Hondamatic has "hydraulic lock" at direct-drive (ie. no oil flow at 1:1 ratio) so has good efficiency. They quote ~75-80% efficiency at the low ratios, and 90+% at direct-drive, which is quite good. I think the main losses at direct-drive are from the spool valves which still move in and out.

In principle it could be run in-line with the engine (ie. coaxial crank, input, and output shafts), but on the Rubicon the CVT is mounted parallel and next to the longitudinal crankshaft, and geared to it (as seen in your cutaway picture (http://world.honda.com/motorcycle-technology/hondamatic/image4.html) above, and page 6 of the patent, and not what I originally said in earlier post). This makes it very compact, and even more suited to FSAE. Mount it with lateral crank, remove some of the reverse gear bits (at right of photo), and fit a gear or chain drive up to a diff mounted just above the CVT.

Weight-wise, I don't know... Honda developed this transmission back in 1990s, fitted to a 250cc motocrosser that won some Japanese championships, so it cannot be TOO heavy (see on same site as engine pic, where slightly different hydrostatic drive originally used on 1960s Honda scooter!).

Overall, the Rubicon drivetrain looks very solid, and the various ATV forums suggest it is very reliable (used for ploughing!). I guess that things like the starter motor are also very reliable (to keep Farmer Brown happy), so probably a bit oversize and heavy. It also has a balancer shaft (adds a bit more weight) and is rubber mounted in the frame (again to keep FB happy, and which I might keep for FSAE). I am sure the CVT could take a lot more power, and the engine at ~20kW should also be able to give a lot more.

I note that along with the Rubicon hydrostatic drive, Honda also have ATVs with all-manual gearboxes, torque-converter style automatics (like on cars), and (I think) also belt-and-cone style CVTs. I guess they are covering all options...

Z

[mech5496]

Well, I do not like the engine it comes with at all, but I guess the actual unit can be used (or its parts) to build an IVT for FSAE, using any gearbox-less engine out there, like Jawa, Genesis 80Fi or those 4stroke 250cc kart engines... ;)

[bob.paasch]

I note that along with the Rubicon hydrostatic drive, Honda also have ATVs with all-manual gearboxes, torque-converter style automatics (like on cars), and (I think) also belt-and-cone style CVTs. I guess they are covering all options...

Z

Also of interest, the Fourtrax Rancher with its 5 speed dual clutch transmission. It can operate in either automatic or push button mode. The efficiency of all-gear drive combined with very quick shifting. I'm guessing we might see this style transmission on other Honda ATVs in the future.

[mech5496]

Another piece of info to add here; the Hondamatic unit weighs about 17lbs or 7.7kg, which IMO is a bit on the heavy side...