First post, and some Jawa questions

[Salmon On The Beach]

Hi everyone,

I'm from the University of Newcastle SAE team and this is my first year of being involved in SAE. Whilst I am doing little bits of work here and there, I am primarily focusing on optimising the performance of our current CBR600 engine and also engine selection and associated design for next years car. We have a dynamometer in an engine test cell that will hopefully allow us to carry out all the required testing.

We have more or less decided that the good power and torque at high rpm of the CBR just doesn't get used enough to justify its weight and fuel consumption. We are keen to try out a CVT next year as opposed to the fixed ratio integrated bike gearboxes that we have used in the past, and at the moment we have been thinking it would be cool to use a 500cc Jawa speedway engine. I'm sure many of you are at least somewhat familiar with this engine, but for those that aren't it is a single cylinder, air cooled design. You can check out the engine here if you'd like to know a bit more about them http://www.speedwaybikes.com/cody/cody03.htm

So, after looking into the Jawa I have a few questions:

- Fitting an electric start: Anyone know if this has been done, or have any ideas how it could best be done?

- Fitting an alternator: Same question really, anyone know if another team has done this or have any idea of the best way to fit a charging system to an engine that doesn't have one standard?

Any advice would be much appreciated, and I will do my best to document on this site my progress with the engines.

Have a good time all the time guys

[Scott Wordley]

I think Kev from ECU said it best...

"In FSAE you buy an appropriate gearbox and get the engine for free"

I would have to agree with him.

I would also argue that in the history of the competition we are yet to see any team extract a significant performance advantage from a custom engine and/or gearbox. Sadly most are plagued by reliability issues or additional weight or complexity, some or all of which combine to negate any theoretical advantages.

I have a huge amount of respect and admiration for teams like WWU and their V8, Melbourne WatArd, Aucklands V twin transaxle, Aachens Mahle CVT and the recent Mercedes HPE turbo inline twin but to date comp results have shown that it is not necessary to go to those lengths to build a fast and successful car, if that is your aim. OEM motorcycle engine gearbox units have proven hard to beat. So far.

Take a hard look at your goals and teams available resources. If your aim is to learn a lot about design and manufacture of engine and gearbox internals and you have the time, money, team members and facilities to bring to bear on the problem and you can tolerate non completions of the car and even non event finishes for the sake of the learning experience, then go for it. You wouldn't be the first team.

If you just want to build a reliable, high power light weight engine and gearbox with minimal effort, minimal cost, maximum simplicity and your best chance at reliability... Just get yourself an off the shelf single, Honda, Yaha or KTM.

That Jawa would also require significant mods for fuel injection (like some OEM bike engines)

[Moreboost]

The electric start and alternator are the easy bits!

If you have money or are friends with a motorcycle wrecker go and have a look at any engine, even your current cbr. Aukland/ Monash have run a alternator just of a belt i believe easy stuff!

Will you be making your own CVT? I have done a little bit of sim on this, if you can get the response time town on the CVT your definately on a winner!

The problem is - Most are big and unresponsive.

Look to piaggio scooters for inspiration.

Good luck!

[Moreboost]

Originally posted by Moreboost:
The electric start and alternator are the easy bits!

If you have money or are friends with a motorcycle wrecker go and have a look at any engine, even your current cbr. Aukland/ Monash have run a alternator just of a belt i believe easy stuff!

Will you be making your own CVT? I have done a little bit of sim on this, if you can get the response time town on the CVT your definately on a winner!

The problem is - Most are big and unresponsive.

Look to piaggio scooters for inspiration.

Good luck!

P.S - That engine is nuts, you can lie it down! a CoG that low would be nuts.

[TurboTom]

I agree with Moreboost - go for it. If you have the backing you will learn a whole bunch of stuff like Scott said as well.

I believe there is a thesis on the suzuki burgman CVT for FSAE in the forums somewhere.

However given such a minimalist approach you might consider a clutched direct drive as well?

The weight advantage of running the jawa over a traditional dirtbike 450 would most likely be offset by a CVT gearbox, and I'd expect power to be similar.

[Warpspeed]

Originally posted by Salmon On The Beach:
- Fitting an electric start: Anyone know if this has been done, or have any ideas how it could best be done?

- Fitting an alternator: Same question really, anyone know if another team has done this or have any idea of the best way to fit a charging system to an engine that doesn't have one standard?

This is very often done with small gasoline engines such as lawn tractors and golf carts by linking a "starter generator" with a vee belt to the original rope start pulley.

Just Google "starter generator" for more info.

[Salmon On The Beach]

Unfortunately I'm not the guy responsible for investigating the CVT, but I really don't think we have the capability to make one ourselves. When you said you have done some simulations on this Moreboost, what sort of sims do you mean? I'm sort of new to all this and aren't really down with all the popular software etc.

I was reading a previous thesis paper from a team that used a CVT mated to a Yamaha Genesis engine, and I think they said theirs weighed about 8 kg from memory. And yeah, I have looked at some scooters that use them, although it looks like most of them feature an engine that is attached to a swing arm.

We certainly aren't set in our ways about using either a CVT or the Jawa engine, we just thought that it would be good to do something interesting and a bit different in terms of design. You may well be right that a more common dirt bike single would be better. I will have a look into the starter generator.

Thanks for all the advice guys, very impressed with the quick responses. I'm sure you'll be hearing more form me and I'll let you know how we go with engine/gearbox selection

[Z]

Salmon,

Both Rob Woods and myself have had quite a lot to say about the Jawa engine and CVTs. The "Any way to objectively choose engine?" (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905/p/1) thread has some of those ramblings.

Generally speaking, I think the Jawa would make a great FSAE engine. The only thing holding it back is that someone has to be the first to do the grunt work of figuring out all the sorts of things you mentioned. Once that is done I reckon they would sell like hot-cakes. One of their main advantages is that the engine and its parts are readily available all around the world, and at reasonable cost (~$3,500US for the base engine, same as the Drexler diff!).

Note that there are lots of different variations of Jawas. Lay-down and near vertical, 2-valve and 4-valve, and I believe a 600cc is possible. Primary-drive (belt or chain), multi-plate clutches, and a separately mounted 2-speed gearbox are also available from Jawa distributors.

The standard engines are carbed and run on methanol. The carbs could possibly be rejetted for E85 or petrol, but a bespoke EFI would be better. You say you have a dyno cell, so this shouldn't be any harder than other FSAE engines (easier than the fours, because only one pipe/injector required). Exhaust should be no different than any other single.

For alternator you can use the smallest one you can find, from car or motorbike. This can be direct-drive from the crank (ie. in-line), or you can belt-drive it from any part of the car that rotates. It is common on racecars to drive the alternator from the final drive. This way it only charges when the car is actually moving, but not when the car is stationary, even with engine running. The point is that racecars don't spend a lot of time stationary and with the headlights, sound system, +++, all draining current.

For electric start, and also for a fan for better cooling, you might look to smallish industrial engines like these (http://www.genquip.com.au/petrol-stationary-motor-13hp-electric-start/). The parts should be able to be bought individually (very cheap from China) and they shouldn't be too hard to bolt on via some homemade brackets.
~~~o0o~~~

For overall layout I would have a lateral crank, rotating the opposite direction to the wheels, and the cylinder leaning backward. So maybe something like this, with the front of the car to the left...

http://www.atvriders.com/images/yamaha/2009-atv/yamaha-2009-preview-grizzly-550-fi-eps-4x4-atv-engine.jpg

In fact, this Yamaha Grizzly 550 (http://www.atvriders.com/atvmodels/yamaha-2009-grizzly-550-fi-4x4-eps-atv-preview.html) would be a good donor for an FSAE car. Aside from engine (with EFI, electric-start, balance-shaft...), CVT, diffs, axles, brakes, etc., etc., it even has Electric-Power-Steering (see right of above pic)! You could start with this Yamaha 550 at about 36 hp, and later upgrade to the Jawa in a similar layout with almost twice that horsepower...

BTW, U of Newcastle in which country?

Z

[Francis Gagné]

CVT, been there, done that. We used CVT on CBR600F4i and on Turboed Yamaha 80FI (phazer) engine. Granted, both engine were pretty powerful (80+ hp) the CVT was quite huge. It was relatively light, using snowmobile race parts. (around 9kg if I remember well). And it was not overkill in size, considering we sheared a drive rubber plate in the few first test days last year.

There were three major problems :

1 - Packaging : It was difficult to position the CVT, the differential, and the 2 Stage custom chain reducer that was needed after the CVT. The 2 stage reducer was needed for us to shorten the rear of the car and keep an acceptable mass repartition and wheelbase. The chassis was a mess in the rear and the suspension needed special care to go around the CVT. The chassis needed to be very precise to ensure that there would be no interference between parts.

2 - Hidden weight : The chain reducer weighed as much as the CVT itself, the chassis in the rear had horrible load paths.

3 - Tuning : Tuning of a CVT is somehow a black art form known by very few, tuning for accel seemd to be the easiest done, but having good and predictable response in corner exit was never achieved in the 11years the team ran a CVT.

You can look at our 2011 car pictures on our facebook of website, this was probably the best drivetrain packaging we could have made using these componments.

On the upside, the no shifting aspect made the learning curves of driver much better.


If you are going to use an engine with no gearbox, why not use a two "gear" setup? Depending on your car setup and philosophy two gear might be sufficient, Lapsim it to evaluate the difference in performance.

I have seem some home made karts and buggy using two chain or belt reductor that are couple with one clutch on each chain and switching the clutch from one to the other. It's simple to design, cheap, and lightweight. Probably what you are looking when packaging a completely new engine. (an exemple for the visual ones : http://www.thepixelpump.com/wi...eed-transmission.php (http://www.thepixelpump.com/widdershins/how-to-2speed-transmission.php))

[Mbirt]

OP, be sure to check out Aaen's CVT tuning book to see what you're getting into. A potential issue with the Jawa is the high peak power speed for CVT use. Rubber belt CVT efficiency really falls off as engine speed increases, thus Yamaha snowmobiles feature an internal primary gear reduction before the CVT primary. Other manufacturers seem to stick to peak power speeds of 8000 rpm and below.

Also, as Francis has mentioned above, the final drive will need to be a higher numerical ratio than is typical for a motorcycle engine. Snowmobiles have tiny "wheels" that drive the track, so something like a 7:1 final drive might be necessary on an FSAE car. Check out the Univ. of New Mexico's car's belt final drive they use with a Yamaha Phazer drivetrain.

If the team decides a CVT is the way to go, check out the Rotax 600 ACE: http://www.ski-doo.com/technol...chnologies/4-strokes (http://www.ski-doo.com/technologies/engine-technologies/4-strokes). No one has used it in FSAE, but we've got lots of experience with it in the Clean Snowmobile Challenge. I'm convinced that it would run well on a stock ECU with a restrictor placed between the stock throttle and big stock 4.6 liter plenum. The stock exhaust headpipes would even package well on an FSAE car, just don't use the stock 17 lb muffler. The e-Drive CVT was designed for four-strokes and is very smooth and drivable. It will make more than 60 hp with less conservative fuel and spark mapping on a standalone ECU.

[Z]

Regarding CVTs, I would suggest only using them on a lower power engine, as in the above pic. Using all those Yamaha 550 parts means they are all matched to each other, so less reliability issues.

Also a lower power engine NEEDS the different overall ratios, whereas a high power engine can just about smoke the tyres anywhere on a FSAE track in a single gear. A high power engine really only needs a means of keeping revs up at very low car speeds, such as by slipping the clutch, or by using a hydro-torque-convertor, while negotiating hairpins, etc.

For controllability of CVTs, has anyone ever tried "manual over-ride"? I am thinking of a lever in the cockpit that the driver can push or pull to "help" the pulleys expand/contract for ratio changes. Or since some of the CVTs nowadays have electric motor control, maybe just some buttons on the steering wheel to help "up/down" the ratio?
~~~o0o~~~

Regarding the Jawa, my current thinking for easiest/best layout is roughly as in the Yamaha 550 above, with front of car to left. So, heavy Jawa crank tucked up hard against seat-back, and laydown cylinder pointing backward, or the more upright cylinder leaning backward against the diff (with diff roughly just under the exhaust manifold).

Depending on type of Jawa the overall ratio required is about 10:1 (~12,000rpm short-stroke, 4-valve-engine:1,200rpm wheels), or maybe ~7:1 for the lower reving, longer stroke, 2-valve Jawas. This could be done by a primary chain-drive back to a clutch and layshaft under the diff, then gears up to the diff. This conveniently gets the wheels turning the right direction (because with cylinder pointing backward, the Jawa is also turning backward).

The primary chain-drive could be done as a two-speed chain-box, as in Francis's link. Or else the secondary gear-drive (up to the diff) could use two pairs of gears "borrowed" from any small car's manual gearbox (1st and 2nd would give best ratios). These would include synchromesh for free (or could be converted to dog-clutches), and reasonably high quality, quiet, helical gears, and all you have to build is the case. Err, and mount the output gears to your choice of diff, or just run as a spool...

Anyway, the Jawa's power (~60+hp) in a lightweight car should manage just fine with a low gear for AutoX and Accel starts, and maybe also for Skid-Pad and coming out of hairpins, and then high gear everywhere else.

Z

[Warpspeed]

If only two speeds are sufficient, how about just a hi stall torque converter plus a means of locking the converter solid with a parallel friction clutch?

You could even control the clutch with a microcontroller and then map it for road speed and throttle opening.

[Moreboost]

Originally posted by Salmon On The Beach:
Unfortunately I'm not the guy responsible for investigating the CVT, but I really don't think we have the capability to make one ourselves. When you said you have done some simulations on this Moreboost, what sort of sims do you mean? I'm sort of new to all this and aren't really down with all the popular software etc.

I was reading a previous thesis paper from a team that used a CVT mated to a Yamaha Genesis engine, and I think they said theirs weighed about 8 kg from memory. And yeah, I have looked at some scooters that use them, although it looks like most of them feature an engine that is attached to a swing arm.

We certainly aren't set in our ways about using either a CVT or the Jawa engine, we just thought that it would be good to do something interesting and a bit different in terms of design. You may well be right that a more common dirt bike single would be better. I will have a look into the starter generator.

Thanks for all the advice guys, very impressed with the quick responses. I'm sure you'll be hearing more form me and I'll let you know how we go with engine/gearbox selection

No programs. Just excel. If you have some old data and a track map easy to replicate.

From memory the phazer one was 10kg - wasnt so much the issue of weight with it for me. More the size, so bloody hard to package not to mention all of that weight is in the pulleys. Then there was the tuning of the CVT which is abit unknown.

Some of the scooters use 2 tapered rollers with a belt. They all look small. Would be cool to see a small one tucked in somewhere.

You could always just make your own gearbox using donor internals as well.

[jlangholzj]

Originally posted by Z:

For controllability of CVTs, has anyone ever tried "manual over-ride"? I am thinking of a lever in the cockpit that the driver can push or pull to "help" the pulleys expand/contract for ratio changes. Or since some of the CVTs nowadays have electric motor control, maybe just some buttons on the steering wheel to help "up/down" the ratio?
~~~o0o~~~

Z

Our 1/4 scale tractor pulling team on campus tried such a system two years ago and I (think) they're planning on developing it further this year. Was a cool concept but I don't know if they quite had the resources to pull it off to the degree that they wanted. IIRC, it did function just not as well as they wished, issues with compliance, etc.

The thought of a manual CVT is kind of cool in itself. Would serve basically like a mechanical hydro-stat drive....without all the hydraulic crap and weight associated with it.

[Moreboost]

Originally posted by jlangholzj:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:

For controllability of CVTs, has anyone ever tried "manual over-ride"? I am thinking of a lever in the cockpit that the driver can push or pull to "help" the pulleys expand/contract for ratio changes. Or since some of the CVTs nowadays have electric motor control, maybe just some buttons on the steering wheel to help "up/down" the ratio?
~~~o0o~~~

Z

Our 1/4 scale tractor pulling team on campus tried such a system two years ago and I (think) they're planning on developing it further this year. Was a cool concept but I don't know if they quite had the resources to pull it off to the degree that they wanted. IIRC, it did function just not as well as they wished, issues with compliance, etc.

The thought of a manual CVT is kind of cool in itself. Would serve basically like a mechanical hydro-stat drive....without all the hydraulic crap and weight associated with it. </div></BLOCKQUOTE>

holy shit, you have a tractor pulling team.

[jlangholzj]

Originally posted by Moreboost:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by jlangholzj:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:

For controllability of CVTs, has anyone ever tried "manual over-ride"? I am thinking of a lever in the cockpit that the driver can push or pull to "help" the pulleys expand/contract for ratio changes. Or since some of the CVTs nowadays have electric motor control, maybe just some buttons on the steering wheel to help "up/down" the ratio?
~~~o0o~~~

Z

Our 1/4 scale tractor pulling team on campus tried such a system two years ago and I (think) they're planning on developing it further this year. Was a cool concept but I don't know if they quite had the resources to pull it off to the degree that they wanted. IIRC, it did function just not as well as they wished, issues with compliance, etc.

The thought of a manual CVT is kind of cool in itself. Would serve basically like a mechanical hydro-stat drive....without all the hydraulic crap and weight associated with it. </div></BLOCKQUOTE>

holy shit, you have a tractor pulling team. </div></BLOCKQUOTE>

not to completely derail this thread but...

dude...we're from fricking south dakota....are you really THAT surprised?