Treadmill Dyno

[taylorp035]

We could run the starter for the entire burn.  The starter pinion has no spring on it, so only acceleration will cause it to go into the flywheel.  Could be violent though if it catches the flywheel at 4k rpm... have seen this many times on the dyno.  Usually small pieces of plastic will shatter off the pinion gear.

[Madscientist267]

Eek. I wouldn't. But I'll get to that in a second...

The two possible means of fuel delivery are:

1 - If the engine is carburetted, the engine will suck fuel in whenever the intake valve is open and the piston is going down.

2 - If the engine is fuel injected, the engine will 'use' fuel whenever the injector is open. I VERY seriously doubt that an EFI system for a small engine such as this would be 'aware' of whether or not the injector should be enabled or disabled automatically. To my knowledge, it doesn't even happen on yer fancy(er) four bangers.,,  If the 'ignition' is on, so is the fuel injection system, and the injector will fire fuel based solely on crank position.

All that being said, neither requires the engine to actually be firing to 'use' fuel. #1 does it due to natural aspiration pulling fuel into the venturi. #2 does it because the fuel is under pressure, and the valve that lets it out is open. Either way, you're dumping fuel into the engine, and if it's not providing the 'bang' part of 'suck-squeeze-bang-blow', the fuel is going out the tailpipe, lowering your MPG.

And the starter thing is kinda unrelated to the MPG thing, although would SERIOUSLY increase it if this should happen. But something tells me you'd be disqualified for cheating (I'm going to assume that 'pushing' is not a valid form of power input  ). You really want to risk losing your starter in the middle of the race because you shaved the pinion clean?

If you're fuel injected (and I think last count was that you were), you'd be better off having the coil firing whenever the engine is physically turning, 'running' or not, then modulate the fuel as the very last thing. Saving fuel IS the idea here, right? Or did I miss something...  

Disabling the injector may or may not be a trivial task, but I'd view it as an important consideration in saving fuel.

My 10-steps-to-win proposal:

1 - Crank the starter up to full speed (as you proposed)

2* - Don't forget - CLUTCH BEFORE ENGINE!!!  

3* - Enable coil

4* - Enable the injector

5 - Shut down the starter once the engine fires.

6 - Run the calculated burn.

7* - Kill injector.

8* - Kill coil.

9* - Release clutch

10 - Rinse and repeat as conditions require to win the race!

* - You can probably combine 2, 3, and 4, as well as 7, 8 and 9 into just two discrete steps, by careful selection of a switch (and maybe relays) to do the work for you, leaving only a 6 step process.

Since you're not charging your cranking/ignition batteries by any means (from the car's engine), there's no MPG penalty whatsoever to letting the waste take place there. Provided you have enough capacity to provide cranking and ignition for your needs throughout the race, you'll be all set.

If you're not comfortable with the amount of time left to try and coordinate such a design, pray the drivers can memorize everything! LOL

Good luck either way - Will be cheering for you guys from here in VA!

Is this thing televised by chance?

Steve

[taylorp035]

Is this thing televised by chance?

Umm... not really.  It should be and maybe the local stations might have something.  Shell came out last week with a nice 5 part mini series video (15 min total) from the Shell Ecomarathon last week in Europe.  Laval was in the video with 2909 mpg.... I think they have 1st place in the bag at this point.  I watched the whole video last night and I think it was worth it.

If you're fuel injected (and I think last count was that you were)

Nope, we are back to the coil and carbed engine.  I runs really really good.


As for the firing plan, I think we will stick to our original:

1) Flip kill switch -->  Turns on iginition system and coil
2) Turn servo tester to activate the starter
3) Hold on for dear life
4) In about 2 seconds, estimate your speed and when you hit 20 mph, cut the engine
5) Make sure everything is off (starter, coil)
6) Coast until speed is down to 10-12 mph, depending on race plans and my calculator


If we use our servo operated plate clutch, there will be extra steps and another servo tester.


Spinning up the flywheel isn't necessary illegal, so long as all of the energy that propels the car comes from the fuel.  At 10 mph, we are at about 1500 rpm, which is lower than the starter rpm.  In theory, we could go 24v and boost the starter to 2000 rpm, but that would be cheating and our instructor would let us walk back all the way to Pennsylvania.

[zap]

Sitting in the minivan, the car doesn't appear symmetrical.  Optical illusion or is that one of the "flaws" you mentioned earlier?

The list of Shell video on youtube:
http://www.youtube.com/user/Shell?feature=mhsn&#g/c/027E2B6D9900A88F

[taylorp035]

You are right.  It is not symmetrical.  There are two center lines about 2" apart.

[taylorp035]

Sorry for not posting sooner.  I had a giant mega post last weekend and my browser crashed

1011 mpg and 5th place.  18 cars put a number on the board out of 25 that showed up.  32 teams payed the $600 entry fee. 

[zap]

1011 mpg...NICE!

Where did that put your team... how did that mpg rank?

[Madscientist267]

Pretty good dude.

Congrats!

Steve

[taylorp035]

Our old record was 777 mpg....

1st place was 2158 mpg, 2nd = 1615 mpg, 3rd = 11xx, 4th = 1070 mpg

[Bruce S]

Hey, not bad for all the work put into it, you increased your mileage by nearly 31%.
Do you have any info on how the car and the engine performed?
Thanks for the update
Bruce S

[taylorp035]

Well, we stuck with the 100:12 gear ratio.

The first run:  All cheap bike tires from schwalbe - smooth with no tread.  They were pumped to 85-90 psi.  Both LiFe batteries for the starter and the carpet for driver comfort + a mostly empty water bottle.  The result = 769 mpg running a 15.5 mph average.  21 to 10 mph.

Second run:  Front two tires got switched to the radial tires.  Pumped to 85 psi.  New driver.  Average speed of 15.1 mph going from 21 to 10 mph.  Result = 1011 mpg.

Third run:  Canceled after 3 laps due to severe weather.  Pumped up the tires to 103 psi, lost the carpet, water bottle, original driver, 19.5 mph to 11 mph.  Driver thought things were going really well.

Fourth run: 2 hours later...  Same as 3rd run.  Just as the driver was starting the last lap, the front right tire popped like a balloon.  The car then dropped down onto its wheel skirt and ground to a stop.  Probably would of had 1200 mpg and 3rd place.  After later inspection, both front wheels had numerous holes in the rubber that extended down to the threads in the tire.

Fifth run:  Replaced both front tires for the super thin blue tires.  Also put in the experimental clutch in the rear that completely disengages the rear wheel from everything.  Once on the track start line, the clutch wouldn't engage. 

Here are the competition photos and the team photo at the end.  We are currently working on building ourselves a custom 180 tooth timing pulley for a carbon fiber belt.  This should boost drive train efficiency by about 3-5% at least and save 2-3 lbs.

http://behrend.orgsync.com/org/societyofautomotiveengineers22440/Pictures

[ghurd]

Uh, I am sure it was covered elsewhere, but these are 'bike size' tires right?
I see knuts at the Presque Isle (and other) time trials put 140PSI tubeless tires up to 175.  Or 200.  Carbon rims or not.
Can get that grade of tire for <$30 on sale.
Just thinking out loud,
G-

[zap]

I'm pretty sure they were "bike size"... I think 20"?

And speaking of tires...
Looking at the pictures, I was surprised there seemed to be a number of vehicles with exposed tires and of those, none seemed to have "areo spokes" nor were the spokes covered in any way, such as this entry from  Michigan Tech:
https://d1q9wbuypc40mm.cloudfront.net/lrdys76phdts7c_710.jpg

[ghurd]

Some spokes are more modern than expected.
Wife's spokes look like wire from the side, but are actually flattened to be very thin.  Almost like narrow stips of thin sheet metal.

I should not have narrowed it down to tubless earlier.
Oh, and I also sometimes see those over-pressured tires pop.
Not very often, but it happens.
If it happened often they wouldn't do it, because blacktop moving past flesh at 35MPH is painful.
G-

[DanG]

Sweet.

Which was 2nd & 3rd?

[taylorp035]

^DanG -   Those are old photos, at least 2008 or older.  The white car in the middle was the winner for sure (University of British Columbia).  That car got 3100 mpg.

Uh, I am sure it was covered elsewhere, but these are 'bike size' tires right?

20" tires are really only made for kids bikes and BMX bikes.  If you go to any website that sells high performance bike parts, they usually don't even list the tire diameter, because they are all 700mm ~ 26-27".  There are hunderends of choices in the 26" category, and only about 10 in the 20" section.  Now try to find aerodynamic rims in 20".....  or even light weight aluminum.

To make things harder, spokes for 20" rims that are "bladed" or flat are also very rare.

For tires, there are some recumbent bike tires, but that's it.  This year, there were ZERO supermileage tires manufactured (our blue $75 paper thin tires sold out in February).

Oh, and I also sometimes see those over-pressured tires pop.

I talked to several Laval students who went to Europe about the tires.  They have the true tubeless radials over there.  Even the top teams sometimes opt not to run them tubeless because one small rock will cause them to leak.  They say that they pump them to 205 psi.....    In Michigan, Laval ran tubes even though they have their own custom made carbon fiber tubeless rims.  I would kill for a set of those.  We could of had 2000 mpg with those with our stock engine and heavy car.

Looking at the pictures, I was surprised there seemed to be a number of vehicles with exposed tires and of those, none seemed to have "areo spokes" nor were the spokes covered in any way, such as this entry from  Michigan Tech:
https://d1q9wbuypc40mm.cloudfront.net/lrdys76phdts7c_710.jpg

Akron had exposed 26" wheels with really nice wheel covers.  Else, everyone else decided not to do anything.  Keeping the wheels inside of the car reduces frontal area and lessens the disturbance caused by outside wheels + the steering knuckles.

I see knuts at the Presque Isle (and other) time trials put 140PSI tubeless tires up to 175.  Or 200.  Carbon rims or not.
Can get that grade of tire for <$30 on sale.

But their rims aren't 1.75" wide either.  Or $30 tires from Schwalbe were their best 20" tires.  Good for 6 bar (88psi).  If I could get the european Michelin Radial tires, I would be willing to pay $300+ per tire.  Would need 4 tires most likely. 

[taylorp035]

Here are three videos.  I might get the slalom and brake test videos up if I'm ambitious.  I'm currently working on a new C.F. timing sprocket.  it will be 18" in diameter and be able to support 720lbs of tension  from the belt.  The sprocket should weigh about 1.9-2.3 lbs.

http://youtu.be/K31VZ4Kt0Nc

http://youtu.be/VOGaS4O3wXE

http://youtu.be/K31VZ4Kt0Nc

[taylorp035]

http://youtu.be/KRorTecia4w

First car:  The winner at 2158mpg.  Universite de Sherbrooke

Second car: Car from Hawaii.  It got 197 mpg.  You can probably guess that the squeaking as it drives by is not good.

Third car:  First through tech inspection.  I liked to call it the "snowplow".  Probably was 10-11 feet long.  651 mpg.  Not bad for outside wheels and a lumpy body.

You can see that it was quite windy... should of used my windmill to power the lap timer.  The good cars still were going at 10-15 mph when they passed the camera, after coasting ~0.5 miles from the hill.  Of course you had to driver into the wind on the other side



And here is the three speed transmission car.  Waaaaay too complicated for my taste and they are probably only 60-70% efficient though the entire transmission.

http://youtu.be/vRKM-40dR2w

During this video, they only used 2 gears, but while our team was driving around the track, they said you could hear them rowing through all three gears on every burn.  Probably hitting 5k rpm on every gear.

[zap]

Second car: Car from Hawaii.  It got 197 mpg.  You can probably guess that the squeaking as it drives by is not good.

Ok... so...
What would your best guess be for mpg if I were to make just some quick and cheap 3 or 4 wheeler and toss your garden variety B&S on it with say... a centrifugal clutch?

[Madscientist267]

Uhh, aint that called a 'go-kart'? LOL 

Steve

[taylorp035]

What would your best guess be for mpg if I were to make just some quick and cheap 3 or 4 wheeler and toss your garden variety B&S on it with say... a centrifugal clutch?

Umm... if you had a steel frame, a 10% efficient engine (fairly conservative), 3 or 4 wheels, and a crude body, cheap tires, simple #35 chain...

250-300 lbs + driver weight and a drag coefficient of 0.3-0.5...

Driving an average of 20-30 mph on a paved road....


100-200 mpg if you ran the engine constantly, better of you could use a 50cc OHV honda.

300-400+ mpg if your burned and coasted with a simple bicycle clutch on the rear.


There have been 500 lbs supermileage cars that look like hummers that have gotten 200 mpg.  Rolling resistance is key


If you literally mean a four wheeler (ATV), then the low pressure tires are going to hurt you badly.  The stock engine (a 250cc honda, Yamaha, or equivalent), will probably be better than the B&S.  Just put some small car tires on it or bicycle tires.  Bicycle tires are about 2x better than car tires for rolling resistance.

[taylorp035]

My next project     180 tooth super custom carbon fiber timing sprocket.




I am cutting out the spokes as I type this.  The outside will be much thinner in the aluminum version.  Later this week we will buy the belt, which won't be cheap.  This way we can make sure the tooth profile is correct and nothing will bind up.

[taylorp035]

Sprocket Loading Conditions

Max tire normal force:  80 lbs
Max coefficient of friction:  0.9 (more like 0.8 )
Tire Diameter:  19.56”
Sprocket diameter:  18”  (really close)

Max torque applied to the wheel = 80 lbs *  0.9 * ( (19.56/2) / 12 ) feet   =  58.68 lb*ft
Equivalent belt tension =   58.68 lb*ft * (12 / (19.56/2)) = 72 lbs.


From the engine side:
From previous engine tests, max average torque is 4.1 ft*lbs
Considering the lightweight flywheel and the total of 4 cycles 
4.1 ft*lbs * 4 cycles * 2x peak during the combustion stroke = 32.8 ft*lbs
Correct for the ratio :  32.8 * 180 / 22 teeth = 268.36 ft*lbs  
Belt tension would then be 329.28 lbs.

EDIT:

And here is a sketch of my new engine design.  I know it is really rough, but maybe you guys will get the idea.

Link to full size:  http://i55.tinypic.com/i7963s.jpg

[zap]

Uhh, aint that called a 'go-kart'? LOL 

Steve

Ahhh... yes... a go-kart

100-200 mpg if you ran the engine constantly, better of you could use a 50cc OHV honda.

Thanks taylor!

I asked because a few years ago I was out on the ebike and I saw an older gentleman climbing aboard a rather large delta trike at WalMart.  It had a 5 hp or so B&S or similar, hooked to a NuVinci drive.  I'm not sure how he did it without getting busted by the cops because IIRC, there were no pedals but the thing wasn't licensed either and he'd ride it on the sidewalks.  The build quality was excellent.

He told me he had ridden it all the way to Greeley which is about a 50 mile drive from where he lived.

He took off and I followed him for a ways till he went west and I went east.  The trike zipped right along.

[taylorp035]

In the last few weeks, we have been testing our new tires on our rolling resistance testing rig (the design and exact data is a secret for now).

The results:    So called "low rolling resistance" tires are complete garbage when compared to our blue supermileage tires and the radial tires.  We found that the thin construction, wide supermileage tires at 25 psi were still much better than the so called "low rolling resistance" tires at 100+ psi.  One reason for this is because nearly every tire sold has some type of puncture protection, which increases rolling resistance by a very large amount.  At the same pressure, the difference was about 2-2.5x worse/better is rolling resistance.

We found increased pressure affected the rolling resistance of the various tires differently (some linear, some logarithmic).  I found this to be very interesting.

The worst tires we tested were our "snow" tires, which were a set of $9 Michelin Diablo tires.  The thick, sticky tread pattern is the main problem.  These tires were about 6x worse then our best runs with the other tires.

Other things that helped included finding the right tire tubes.  Obvious thick ones are going to be worse, but the question was by how much.  Well, the difference between our two different thickness tubes was about 20%.  Latex tubes would be even better, if you could find them in 20" format. 

The best would be tubeless, which our tires just so happen to be rated to be     So, this week's hard work has yielded us with a semi-working tubeless set up using our paper thin tires and normal bike rims.  With I little more work, we should have them totally sealed (a few more runs to walmart to the kids bike section will be needed).  The secret construction is still under wraps until we get some data off of it (or you can PM me).

Next up will be trying the radial tires without tubes.  Supposedly a Crr of 0.0008 can be achieved... way lower than about 0.006 for a good bike tire (notice the missing zero).   Inflation pressure will be modified too for this year, hopefully going past 100 psi after a few runs are put on the board.  100 psi is a lot for a tire that is about 3 layers of paper thick .

For you bicyclists out there, big 700c racing tires for bikes have rolling resistances in the 0.005 range  (data from continental).  The only advantage to the big ones are that they are more aerodynamic... which I doubt it matters much considering the size of the rider going through the air.

Else, we turned a few hundred pounds of copper today for scrap money...    Work on our acceleration dyno should be complete by this weekend.  It's an old senior design project that never ran, so it will be nice to see it running.  It has an air disk brake and a rpm gauge. 


The new car design was mostly completed over the Thanksgiving holiday.  Hopefully construction will start in the next 3 months.... it's going to be a great car, hopefully weighing in at about 10 lbs, which would include the windows and the structural frame.  If it is 15 lbs, I won't be too unhappy, as long as it isn't 41 lbs like our current one.

Classes will be done in 2 weeks 

[SparWeb]

... Supposedly a Crr of 0.0008 can be achieved...

   

No I noticed the extra zeros!!

That's 20x less than your average auto's steel-belted 4-ply.

[taylorp035]

Here is a video of our new flywheel dyno in action      Even though the stand and flywheel was already built, it took us a good 20+ man-hours to get the thing running and hooked up to the engine.

http://www.youtube.com/watch?v=IT1sqSlVmss

The rpm's are in the description of the video.  In general, they were between, 333 and 1000 rpm.  At competition, we would shoot for about 333 rpm to 700 rpm. The loud clicking sound is from the brake rotor rubbing.

The flywheel and the correct gear ratio fairly accurately represents the load of the actual car, which should allow us to do some more testing and find out how to make the engine run the best.  Starting the engine below ~400 rpm on the flywheel is a real chore (as you can see), since the starter only goes to about 230 rpm.  Maybe I will kick up the voltage to 26v from the current 16v for next time.

After getting it going, we did 12 burns in a row to simulate half of the race.  After crunching the numbers, the fuel efficiency came out to ~1400 mpg, which seems just right.

[hydrosun]

 Very interesting following this thread. It shows all the compromises made for a reliability and comfort on my bike. I ride a 700cm road bike with Nuvinci gearing, 31cc suburu robin engine,from Statoninc, Zipper fairing, slime tubes and get over 200mpg with me pedalling too. changing a flat tire is hard with the extra chain coming from the engine. So the compromise fo a heavier tire with slime in the back tire. I have a body stocking to further cut wind resistance but riding on the road with traffic was too unerving with car side winds. A tail would help as much as the fairing but would get in the way. Since I'm pedalling too I need air flow on me to keep cool on long rides. So I have to be content with 200 mpg on a slightly modified road bike. I've put in over 5000 miles with this setup. I suppose the engine could be improved but I just put in enough maintainence to keep it running.
Chris

[taylorp035]

200 mpg sounds pretty good to me for a bike.  The tire for sure is hurting the mpg a lot, especially with heavy, protection guarded tire tubes.  I would suggest using some lightweight tubes as long as you have a big enough tire to avoid pinch flats.  I would also use wider tires.   Wind resistance is fairly high too as you pointed out for a bicyclist, so slower speeds should help.  My supermileage car would would be the happiest if I could average 10 mph most likely instead of 15 or 20 mph.  Maybe some spoke covers would help.

The hard part about projects like yours is that saving another 50 mpg doesn't really justify large investments on parts.  More air in the tires will always help, but once you get much past the recommended max pressure, sharp objects become a problem for flats.  It would be easy to drop $100, $1,000, or even $10,000 on upgrades to your bike 

[Madscientist267]

Here is a video of our new flywheel dyno in action     Even though the stand and flywheel was already built, it took us a good 20+ man-hours to get the thing running and hooked up to the engine.

http://www.youtube.com/watch?v=IT1sqSlVmss

The rpm's are in the description of the video.  In general, they were between, 333 and 1000 rpm.  At competition, we would shoot for about 333 rpm to 700 rpm. The loud clicking sound is from the brake rotor rubbing.

The flywheel and the correct gear ratio fairly accurately represents the load of the actual car, which should allow us to do some more testing and find out how to make the engine run the best.  Starting the engine below ~400 rpm on the flywheel is a real chore (as you can see), since the starter only goes to about 230 rpm.  Maybe I will kick up the voltage to 26v from the current 16v for next time.

After getting it going, we did 12 burns in a row to simulate half of the race.  After crunching the numbers, the fuel efficiency came out to ~1400 mpg, which seems just right.

Man that thing bucks like a bronco at the lower RPMs LOL - Guessing that it is this phenomenon added to the close proximity of one's appendage to the wheel with teeth that prompts one to say something like "never been so scared for the sake of my foot in my life" hahaha

Thing of it is, what do you do in the actual car? You don't just crank on the starter until you get it "up to speed" and then apply spark do you? (Assuming that's how you modulated the engine here - took me a couple passes to realize that it wasn't difficulty starting the engine, but difficulty starting the flywheel) LOL

I had a brainy-fart kinda idea while I was watching the vid... I realize something like a slip clutch is not going to cut it in the mileage arena because it would just waste the absorbed energy as heat, but what about something that returns [most of] the energy it absorbs? A heavy axial spring that gives the crankshaft a couple rotations worth of freedom but that can then be locked hard (much like a torque converter lockup in an automatic tranny) once the driver deems it's time? Something along the lines of a recoil spring for a pull start mechanism comes to mind, only both parts are of course rotating, and the version for the car would have much less angular freedom and a much stronger spring...

Just a thought... That way you could get the most out of your engine, I would think. Might require a slightly heavier flywheel to 'carry through' the lower RPM range, but might be worth a shot... ?

Steve

[taylorp035]

Those were some good thoughts Madscientist267.

When the engine is in the car, we have the centrifugal clutch hooked up, which provides the slipping.  It slips very well at 1000 rpm on the crank (starter motor max rpm), but obviously doesn't slip much at 1500 rpm (at our lower starting speed of 10 mph).  Since our clutch is of the fancy adjustable weight type, we could (and will) do some testing to get the best efficiency and make some observations.

Unfortunately, as you saw in the video, the engine currently has no clutch, so the starter has to spin up the freewheel (probably pulling 1000 watts ).   If the engine was unloaded, the first fire of the piston would kick the rpm up from the 1,000 rpm to something much higher and everything runs smoothly.  Without the clutch, it's running into a brick wall.... and the engine has to survive at 1,000 rpm.  Not a good thing for a tiny single cylinder engine.


The torsional spring idea could work while the car is moving, but when you start from 0 mph, the starter motor would effectively drive the car = a big no-no.   You could still do it if you had a second clutch, like a centrifugal clutch with a lower rpm on the crank, or a mechanical clutch on the rear axle.  All of these add weight and complexity.  But certainly not out of the question.

Sometime soon, our timing pulley for the centrifugal clutch should be finished and we could then drive the car.  A few other things need to be finished, like our new front steering (95% done), and the rear engine mount (75% done, and a bunch of tape + zip ties to hold all of the components down.


Adventures with the tubeless tires should be continuing also, and we may be able to do some rolling resistance / steering alignment tests in our long hall way.

[taylorp035]

Today I did some calculations to figure out engine efficiency from the flywheel dyno results.  As I would of guessed, efficiency during acceleration is much lower than at constant speed.  The numbers worked out to about 7-8% efficient, compared to the previous 12-16% efficient from the treadmill motor set up.  More tests need to be done for sure, especially since we were finding air bubbles in the fuel line, mostly likely from the float bowl.  We also found out that there is a high speed idle screw on the carburetor, so we can more easily control our high speed AFR

EDIT:

Our fuel usage from last year for 1011 mpg = 36 cc for 9.6 miles.  This translates to about 15 cc on our flywheel dyno test.  So far, our current engine (over 6 dyno runs) has used 11-16 cc.

[Madscientist267]

Thats a pretty decent improvement as far as I can tell... Hopefully it translates to the real thing with the same ratio.

Steve