Fantasy engine

[DannytheRadomski]

If you didn't need to worry about cost or availability, what would be the best (<610cc) engine. Not worrying about fuel ecnomy, My thoughts are either V6, V8, or V10. If I was worrying about fuel economy, I would go with a V-twin or V4. I know everyone that has ever expressed any interest in FSAE has thought about using something other than a Honda CBR600RR engine. Also, include some reasons (other than badass factor) why you would choose that specific setup.

This isn't proposed for people to bring up logic in this; this is all just because I want to hear other peoples opinions on this. WWU is the only place to ever make any of this come true.

If you want to include transmissions, the V8 and V10 would be longitudinally mounted and have 8-speed tranaxles. The V6 and V4 would be longitudinally mounted and have 6-speed transaxles. I would also slap some turbochargers and superchargers on there. And maybe an anti-lag system just for the sound.

THis is all pipedreams, nothing real. At least keep it feasible though(I would prefer to leave out the ridiculous V200 derived from 3cc model plane engines that is inevitably mentioned on engine threads).

[Racer-X]

All those cool engines you listed would be heavy and impossible to package IMO. With the restrictor we can't really take advantage of the revs either. I feel the same about the 8 speed, no one can use that.

I would do something like a 600cc snowmobile twin tuned for peak power with a huge plenum to help mitigate the pulsing at the restrictor. I would want to use some sort of custom electronically controlled CVT with a clutch to keep it in the power and avoid slipping. My reasoning is you can't escape reality and keeping it simple pays in our world.

Edited the first paragraph so it made more sense.

[Jon Burford]

Racer-X,
I like your common sense.
More people should use snowmobile engines.
An electronically controlled CVT, Enter Rob...

[murpia]

FSAE is all about packaging. I think a shallow angle V-twin with the gearbox cluster in the V would be about the shortest possible package. (Longitudinal crank).

Next up could be a transverse inline twin, intake / exhaust routing is a pain compared to the above, though.

Regards, Ian

[Owen Thomas]

Closest "ideal" engine to reality for me would have to be a turbocharged 500cc parallel twin. 4-speed sequential transmission, ~10k rpm max. Not as balanced as a wacky V8, but it would be easier to package for sure. Large enough for the power, small enough to make a restricted turbo worth it, with torque for days. Racer-x's idea is more practical, since 600cc twin snowmobile engines already exist.

If you're going to throw out the idea of V2 through V10, I'm going to go out on a limb and suggest horizontally opposed 4 cylinders (boxer). You could drop the thing super low, and would probably end up with a transaxle setup (not necessarily good or bad). 600cc, direct injection, naturally aspirated.

[RenM]

v10 seriously?

a dream engine would have to have a lot of power and even though you don't care about your bsfc, you should care about your brake specific air consumption. With the air restricted you should be getting as much power as possible with the air that is available. Having the mechanical and thermal losses of 10 cylinders wont help much http://fsae.com/groupee_common/emoticons/icon_wink.gif

i´d build a 2 cylinder inline and supercharge it. Neat package, lots of power, low weight.

[Jon Burford]

Owen,
the Genesis Yamaha is a 500cc inline twin and it is worth turbocharging that if your using a CVT.

[Owen Thomas]

Owen,
the Genesis Yamaha is a 500cc inline twin and it is worth turbocharging that if your using a CVT.
I know, that's why I threw in the "4-speed sequential" part. What with it being a fantasy and all.

[Jon Burford]

Well there was a fairly famous engine to that effect produced in Australia some years ago.

[Charles Kaneb]

I'd like a V-twin that doesn't have the "character" of an Aprilia engine. If Suzuki had only made the SV610 instead, or FSAE had chosen 646 rather than 615cc...

To go "absurd" rather than practical, how about a tandem-twin like the old Rotax 256 two-stroke?

One cylinder ahead turning a crank that turns one way, one cylinder behind turning the other that turns the other way - reduced vibration, and no "replace everything if something breaks" like a vee- or inline-engine.

[Bus_Lengths]

Well,if we are going the $$$$ and time route of custom building a 600cc V10 or equivalent, I'd just ditch the restrictor and silly 610cc limit altogether. Why spend years of manpower and $$$$ for something you get less than 1 weekend out of?

Instead, I'd get a higher mileage LS7 out of a totalled vette. Throw in the cam and valvetrain required for 8500 RPM's along with the necessary portwork. I'd just run one gear because a full transmission wouldn't really be necessary. It would weigh about 320 lbs for the engine (as opposed to 100 for your typical 4 cyl) and your driveline would be heavier, BUT it'd also make about 740 bhp. With no transmission at all, gear changes will be infinitely quicker than an F1 car too.

I'd run a 3.42:1 total gear reduction, giving its single gear a 150mph top speed (about as fast as I'd feel safe in a 700 hp student built open wheel car). This would also give it about 2700 lb of thrust at the tire at peak torque. Taking a 400 lb FSAE car, adding 220 lbs for the engine, and 150 lbs for the driveline, and 50 lbs more to the chassis, this car could weigh just under 1000 lbs with driver. This would give it a much better power to weight ratio than a measly F1 car.

Such a car should be able to pull 2.0g of launch accel off idle, upwards of 2.7g at peak torque, and back to down to 2.2g at redline. Might need a little assistance from some VHT for these numbers, but the point stands. If you can get the car to hook, top speed (150 mph) should be reached in approximately 3.1 seconds according to my tractive effort spreadsheet.

You could roll people up at the drag strip on Friday nights and destroy the autocross courses on Sunday. That's what I'd do.

[JT A.]

8 or 10 cylinders just seem like too much of a hassle to make an intake & exhaust for.

I would go with a parallel twin around 500cc, if they could get it to weigh somewhere around 60-70 pounds like the aprilia instead of 90-100 pounds like the current snowmobile motors. A dry sump so it could be laid down horizontal, and direct injection cuz why not. Ideally I'd want a well set up CVT for a transmission.

[Jan_Dressler]

300cc W-3 (30°+90°, gearbox in the 90°-angle), main bearings on the webs, Bishop-type rotary valves, supercharged, turbocompound for fuel efficiency, 3-speed epicyclic gearbox (a la Wilson preselector), one clutch for each wheel both for starting and torque vectoring (no diff).
http://fsae.com/groupee_common/emoticons/icon_smile.gif

[mech5496]

Jan, wow mate! Rotary valves might be a good idea, but I do not know if they worth the hassle. I prefer to be a bit more realistic (possibly wrong for a "fantasy" topic, but, hey, I would love someone to get a bit nuts and actually build something inspired from that very thread). So, onto my "perfect IC FSAE engine"!

As RenM mentioned, more cylinders means more losses. Thus I would go for a single (or inline twin for smoother operation), laid down with the cylinder heads facing rearwards and about 10-20deg from horizontal to make space for the intake port (and a nice big diffuser except for around the intake ports), thus bringing the relatively heavy cranckshaft/flywheel near the CoG (minimize yaw inertia) and as low as possible. Twin (electronically controlled) CVTs, one on each side of the engine for torque vectoring. Turbocharger in the middle of the CVTs, directly above the exhaust port (which would be placed facing upwards), E85 and as many CCs as possible, redline at about 10000RPM, tuned for peak power. Machined/cast motor and CVT casings, backing up as suspension/damper/jacking point mounts and designed to withstand all frame forces. Possibly hollow, so it could also be used as oil tank. Neat small package to bolt directly after the main hoop on a monocoque, with no need for any additional subframe. To aid efficiecy add some Honda tricks, like roller-bearing camshaft(s?), two (instead of three) piston rings and low friction coating (DLC?). I would love to see a team doing it, as mine won't do for sure!

EDIT: To get a bit crazier, maybe direct injection, electric water/oil pump and no alternator. Two really small motors on the front hubs would charge the battery during braking, improving overall efficiency. Thoughts?

[Markus]

3-cylinder with a simple two-speed or three-speed planetary (epicyclic) transmission. Both longitudinal or transverse mounting possible options, can't decide.

Almost constant inertial torque, simple packaging, super fast shifting especially with the two-speed. No forced induction because of restrictor, packaging and weight.

If anything with a "normal" gearbox then zero-shift technology.

I also liked Owens boxer engine idea. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[DougMilliken]

One of the key features of the FSAE rules is no minimum weight. Whatever engine configuration (& unitized frame and suspension mountings as already noted) came out of the packaging study, I'd go for a really light engine. One way to do this is minimize parts count and joints. For example unitize the cylinders and head like an Offy (inline-4 DOHC). A bit of a pain to fit the valves up through the cylinder...

More extreme is Tony Rudd's Lotus Microlight small aircraft engine. This was a pushrod boxer with a split at the crank and one casting for each half-crankcase + cylinder + head. Light rocker covers were the only other gasketed joints. This also had the propeller reduction drive unitized with the 2:1 cam shaft drive (cam above crank to raise the prop center above the engine).

[Ash47]

500cc V-Twin with a 1 or 2 speed transaxle gearbox http://fsae.com/groupee_common/emoticons/icon_wink.gif

[mech5496]

Doug,
being a weight weenie myself, I couldn't agree more; however, unifying the head,cylinders (and crankcase) would only save a few grams in the cost of much increased complexity. So, does it worth the hassle? IMO, no... Weight (and efficiency) is my worry on a twin CVT setup like the one I proposed, and it is true that I really like those epicyclic gearboxes, so maybe using them in a similar twin-clutch setup (like already proposed) looks promising and maybe lighter. BTW does anyone know why those gearboxes are not seen very often? Combined with electronic clutch control and shifting they should kick ass!
On a side note, tilting the cylinder by a bit higher (say 30deg) and swapping intake/exhaust sides, gets the heavy turbo as close to the ground as possible, so this might have a slighter lower CG. (A look on Swanseas S12, note the neat turbo/plenum/intercooler package)

Ash12, I think you have something like the one you describe lying around there...http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Markus]

Harry, epicyclic transmissions are one of the oldest automotive gearboxes ever used and even still are employed inside multi-speed automatic transmissions. What you suggested (epicyclic transmission + electric clutches) is a basic part of a modern auto trans. http://fsae.com/groupee_common/emoticons/icon_wink.gif

FSAE is special in the way that it would not require such a high gear-count as "real-life" conditions so it could really make use of this "ancient technology".

We're already repeating history by re-employing electric motors to produce traction and epicyclic transmissions followed shortly after... So who knows how soon we'll see them. http://fsae.com/groupee_common/emoticons/icon_smile.gif

[theTTshark]

500cc 120 degree V4, twin scroll turbo sitting in the V with a small intercooler for each bank. Mounted to the engine would be a CVT/differential transaxle. It would be dry sumped, oil to water cooler, and a charging system capable of 45 amps. There would be mounting points for all of the rear suspension and rear aerodynamics, and of course they would be stressed chassis members. Whole thing would weight 120 lbs. The engine would utilize an electronic throttle only to be able to idle. Once running on track the gas pedal would be a torque pedal (ala Formula 1 blown exhaust) so that the engine could utilize hot blowing (effectively anti-lag) on the rear wing. Hot blowing instead of cold blowing because flames are cool, and everybody knows that a racecar is instantly cooler if it spits some fire! Plus hot blowing produces higher exhaust gas velocities.

[mech5496]

Markus, I'm not a transmission guy, so forgive my ignorance... Regarding Owens' idea of a boxer engine, there was a thread a while ago about a 2cyl boxer w. a transaxle based on a Suzuki SV650, which seems pretty feasible. Much was also discussed in the "optimum powertrain" (or something like this) a few months back. As we probably seek for the most badass engine, WWU's V8 and Auckland's V2 are by far the most impressive,followed by UAS Graz/Karlsruhe/AMG 2cyl

[Markus]

Isn't Auckland using a longitudinally installed single cylinder with a transaxle? Anyway that's quite a neat package.

I'd recognize WWU's V8 as the most impressive custom powertrain, followed by Mahle's 3 cylinder transaxle, Auckland single and the Wattard engine. For all these the performance is unknown to me but I'd guess it was around what custom engines usually tend to have (especially for the first two), which is not that good. Wattard engine might have been best power-wise and Auckland's engine most likely takes home the overall performance award.

Honestly AMG 2cyl was a bit of a disappointment for me, it was (really) heavy and underperforming. But I must admit that I expected both and given it was their first take I'm interessed to see the progress for next year. Medium-weight and less underperforming, hopefully.

I'm also eager to see if there's a team to utilize the SZEngine and how it will perform - they're not trying to do anything fancy but a purpose built FSAE single cylinder engine.

[mech5496]

Markus, the Aucland team uses a single, but they have also built an impressive V2 using WR250 parts. If you haven't, checked it yet, please do. The Mahle engine was definitely not impressive, seemed unnecessarily complicated and a bit heavy. Do you have any performance/weight figures for the AMG engine? Judging from their initial presentation it seemed nice. Performancewise I would vote Auckland too...

[Markus]

Do you have a link or something for the Auckland V2, couldn't find it from their website?

I don't have any figures from the AMG engine but we lifted it at FSG this year and after seeing UAS Graz at FSG and FSH I'd estimate:

Around 60kg dry (including turbo and aftercooler) and around 45kW at FSH (they had problems in FSG).

[Jan_Dressler]

Originally posted by mech5496:
Jan, wow mate! Thank you! http://fsae.com/groupee_common/emoticons/icon_smile.gif
I chose a 3 cylinder instead of a single because the turbo (-compound). No CVT (efficiency, weight). Roller-bearing camshaft? Oh yeah, and -crankshaft! http://fsae.com/groupee_common/emoticons/icon_smile.gif


Originally posted by Jan_Dressler:
300cc W-3 (30°+90°, gearbox in the 90°-angle), main bearings on the webs, Bishop-type rotary valves, supercharged, turbocompound for fuel efficiency, 3-speed epicyclic gearbox (a la Wilson preselector), one clutch for each wheel both for starting and torque vectoring (no diff).
So, this is mine! (After 20h of quick'n'dirty Pro/E-ing, no, the valves dont't have the correct shape (although the timing should be quite ok), the gears have no teeth, and so on... Just a beginning.)
http://i50.tinypic.com/205tnqq.jpg
http://i45.tinypic.com/2dsofwn.jpg
And a little "family shot", the W-3 in front of an advanced 600 cc-4 cylinder-concept (based on a PC37 Honda engine) in front of a "conventional" dry sumped PC37 package (our 2010 one, in fact):
http://i47.tinypic.com/2dqlylf.jpg

Happy new year! http://fsae.com/groupee_common/emoticons/icon_smile.gif

[mech5496]

Markus,
you can find it on Aucklands' facebook page. 60kg is dissapointing at least. Similar weight with a 4cyl and less power. Graz in fact had a more powerful and less heavy engine with their forced induction Rotax.

Jan,
WOW mate! I like how it came along, and it looks really compact (impessed by the size comparison between the engine and the tripod housings). However, your arrangement puts the engine almost at the rear axle (bad for weight distribjtion) and leaves a huge space between the drivers' back amd the enhine (bad for MoI). Any weight/size estimations?

BTW I would love to hear Z's view on the subject (although I am almost sure about it). So Z, if you had infinite budget, what would you build?

[Jan_Dressler]

Originally posted by mech5496:
Jan,
WOW mate! I like how it came along, and it looks really compact (impessed by the size comparison between the engine and the tripod housings). However, your arrangement puts the engine almost at the rear axle (bad for weight distribjtion) and leaves a huge space between the drivers' back amd the enhine (bad for MoI). Any weight/size estimations?
Yeah, I know, weight distribution... But at least the engine would be very light. Allows for a bit lower position for the driver (although this is limited by template rules). You can shift weight more forward (radiators, battery, etc.), but I agree with your MoI-argument. Should not be so bad if the car is at minimum wheelbase, though (and, as already said, the engine should be fairly light).

Weight/size estimations: Sorry, I just have what you see in the pictures http://fsae.com/groupee_common/emoticons/icon_smile.gif So, the only thing I can tell you is that the engine is 296,88cc (60 mm bore * 35 mm stroke). With proper supercharging (the rotary valves should help there, better VE, less prone to knocking, dual cross-tumble flow with that bore/stroke-ratio) one can get all the power you want (or all the power that is possible due to the restrictor) out of a 300 cc engine.
Dragsters, and F1 in the 80ies, got/get 1000 hp/liter (albeit only for short runs, but F1 today achieves 300+ hp/liter / the equivalent BMEP. With turbo-/supercharging we can have all the efficiency...). Combine that with turbocompounding (get the lost power back to the crankshaft, the rules practically demand this?! Why is nobody doing this??).

And yeah, I would really like to hear Z's opinion on these matters too http://fsae.com/groupee_common/emoticons/icon_smile.gif

[AxelRipper]

For me, I'd go with a 440cc, 180 degree V-Twin. Tilt the longitudinal output shaft at about 30 degrees above horizontal to give some room for the exhaust packaging underneath the motor, and allow room to keep the drivetrain low. Integrate the gearbox off the output shaft with either a 3 speed or e-CVT with an active differential.

I'm gonna take a guess on what Z will say: Jawa motor, CVT, either spool or active diff.

[TurboTom]

I agree with Jan_dressler, 300cc turbo sounds good. Husky have the te310 which is 2 3kg dry, with a cr of 13.1:1.
Bang a turbo on it, e85 to reduce knock and small intake/ exhaust assembly to minimize loss and increase throttle response, you've probably got something that will put out as much as a 450 single.
Garretts gt06 weighs in at just over 4kg, and including intake and exhaust assembly, you have an engine package that weighs in at under 30kg. Given the wr450 weighs in at 33kg dry ( our muffler weighs about 5kg) you can save yourself 10kg+.
That and the engine efficiency will be through the roof would see you win fuel everytime and what should be a 145kg 450 single 10" wheel concept goes to 135kg with similar power, go 8" rims you're down to 130kg and all of a sudden the 300cc turbo single doesn't seem so underpowered.

[mech5496]

TurboTom, I like your thinking...

Jan, IMO energy recovery from a turbocompound would be minimal, and most probably dependent on the load/RPM, as it is a mechanical system. What about using it in combination with an alternator? That way you could use exhaust gas energy to produce electricity. Get rid of (mechanical) water and oil pumps and replace them with electric ones should free some power. I know, conversion of mechanical to electric energy would lower the efficiency a bit, but IMO you would have greater control over recovered energy. Alternatively, you can do the above powered by recovered energy from braking (small front hub motors) plus turbocompound.

What about an electric turbo? You could run it entirely on recovered energy, no backpressure, complete control over boost, zero lag etc..Thoughts?

P.S. Happy new year everyone!

[Z]

Originally posted by Jan_Dressler:
... I would really like to hear Z's opinion on these matters too http://fsae.com/groupee_common/emoticons/icon_smile.gif
Sorry guys, but several weeks of enforced feasting and jolity are killing me (ie. visiting relatives, etc...). And more next week... http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif
~~~o0o~~~

Disappointing start to this thread with suggestions of V6/8/10 engines and 8 speed gearboxes. That is not "fantasy", it is just mindless copying of the big boys! Mumble, grumble ... you must fantasize harder! http://fsae.com/groupee_common/emoticons/icon_smile.gif

But good to see many suggestions that "fantastic" in FSAE can be small, light, and relatively simple. I'd go for something like Harry's single (post half-way down page 1) with horizontal cylinder pointing rearward, and lateral crank tucked under driver's back.
~o0o~

For first year I'd try aircooled, turboed, ~450cc, ~120hp (max available through restrictor), and weighing ~20kg. I figure this is possible because it was possible 40 years ago (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=54520089051#54520089051). I would have the crank reverse-rotating (opposite to wheels), and possibly even add ~5 kg to it to give beneficial anti-roll gyro couple when cornering.

First year transmission would be a beefy clutch, open diff (because 60% rear weight and max aero), and NO gears! Just a simple two stage drive with chain and gear (because reverse rotation). Or, if good control-system students are available, then spool-diff and clutch on each driveshaft for torque vectoring, as suggested by Jan.

All packaged into a compact unit about 0.5m long (front bevelled at 45 degrees for driver's back), 0.25m wide, and 0.25m high (I/E pipes are extra). The CG is close to front (near crank), and nothing much behind rear axle line. Engine plugs into simple chassis from above/behind. Unstressed engine with NO suspension connections, so easily swapped if problems. Engines are only "necessary ballast", but unfortunately rather unreliable ballast!
~o0o~

In later years I might add more mass to the crank so it can act as a flywheel energy store. This stored energy would be utilised through an Infinitely Variable Transmission, built from a single planetary (epicyclic) gearset with hydrostatic control. Engine runs mostly at WOT. Accelerator pedal linked to IVT so engine/flywheel speeds up during no-thrust cornering (storing energy), and slows down during max-thrust acceleration on straights (releasing the flywheel energy).

I figure this small energy store would allow the engine to be downsized to about 200cc (~60mm B&S). The smaller piston, cylinder, etc., offsets any extra mass in the flywheel. Small engine running constantly at WOT gives very good fuel efficiency (compared with big engine sucking air through half closed throttle, and hence low real CR).

The IVT has crank driving the sun gear as "input", either the planets or ring gear are the "output" shaft, and the hydrostatics control the third "control" shaft. The hydrostatics use a fixed displacement pump (eg. gear, roller-vane, etc.) connected to the planetary control shaft, and variable displacement pump (eg. swashplate) connected back to the crank/sun-gear, giving the variable ratio and torque multiplication (like on many tractors these days! http://fsae.com/groupee_common/emoticons/icon_smile.gif).

The various gear ratios are set so there is no oil flow at a "sweet spot" of about 60kph, so no viscous friction losses. Ie., the swashplate is set for zero flow at 60kph, thus "locking" the planetary control shaft and giving gear-only transmission. The engine/IVT package has to work either side of this sweet spot (for flywheel energy storage and release), but the oil flow rates are low so only small friction losses.

Optionally, the hydraulic pressure can also drive the slightly unconventional open-diff for torque vectoring (hint: the diff's planet gears act as a hydraulic gear motor). Can explain in more detail later...
~o0o~

Now, if I really start fantasizing, then I might dream of enlightened rule-makers who allow a greater range of IC engine types. If so, then I would go for a turbo-supercharged, TWO-STROKE DIESEL. This would look very similar to the above four-stroke engine, but with double the number of power pulses it could be downsized to maybe 100cc (~50mm B&S). Less is more!

Gas exchange would be via two conventional poppet valves in the head, set at a wide-ish V-angle. The turbo-supercharger pressurises air in the intake pipe, which runs between the valves (ie. parallel with cylinder axis). Towards the end of each power stroke the exhaust valve opens (say, 45 degrees before BDC). Cylinder pressure drops, then the intake opens, still before BDC.

Presurized intake air rushes down the intake side of the cylinder, across the piston top (now near BDC), and up the exhaust side and out. It is an upside-down "loop scavenge" system, with no need for ports in the cylinder walls. When the cylinder is flushed the exhaust closes, then intake closes (maybe ~60 deg after BDC?). Piston travels upward compressing air, then fuel injection around TDC. Bang, power stroke, and repeat.

To extract the last drops of efficiency this engine could be "internally air-cooled". The turbo is oversized to pump extra air through the cylinder during gas exchange (remember this is a diesel, so no fuel or oil in the air). This sucks heat out of the hot surfaces directly, rather than the heat having to flow through the walls. This heat energy further expands the exhaust gases providing extra drive to the turbo. The turbo can be "compounded" to send this energy back to the crank (simple way is to "blow" through a Roots type SC).
~~~~~o0o~~~~~

Jan,

BTW, here is a "roller bearing mains on the webs" crank from WWII Tatra diesel V12. The crank is built up from short sections (one per V-twin) that are bolted together at the mains/webs. Tatra still build similar V8s for their trucks.
http://autospace.co/wp-content/uploads/2012/03/TatraV910.jpg

Z

[murpia]

2 questions:

Is there any tuning advantage from intake / exhaust interlinking of an inline twin or triple? (vs. a V-twin which makes that hard on the exhaust side).

Does anyone have reliable data on the efficiency of a CVT / IVT (any technology), including the parasitic losses of pump drives, etc.? A dry-sumped 4-speed (i.e. 2 fork) traditional cluster should be >97% engine -> wheels. I doubt any CVT will exceed 85%. That's a reasonably significant loss to make up.

Regards, Ian

[mech5496]

Z, I was thinking about the IVT concept. By having the sun as input and outer ring as output and allowing some slip on the planet cage you can vary the ratio (dependant on slip). I was thinking of some sort of mechanical brake, but the losses during slip would be huge. What if someone used an elecric "brake" instead to recover all that energy? Could be used to power pumps, ditch the alternator etc.
Alternatively, a 2-gear system could be built by designing a geared system which would rotate the planet carrier at a given speed and lock it (or reverse rotation) to change ratio. This would lower the losses significally and I could live with two or three gears....

[DannytheRadomski]

A few questions:
First, in a V engine, would increasing the angle to greater than 180 allow you to lower the center of gravity at all?
Second, can you route the exhaust in a flat engine through the top (or through the center in a V engine)?
Third, what's with the low gear count? I understand it means less weight and complexity, but adding gears can increase fuel efficiency.

This might sound stupid, but its a differential really needed? Karts dont usually use them, and while it would hurt autocross, the cost, loss of weight, and less time spent could help overall.

Please keep in mind that I am not actually a participant of FSAE yet, but I am trying to prepare before entering college. I don't want to go in blind and naive, so I'm getting an idea of what would and wouldn't work now.

[Cardriverx]

First - That would just be taking a less than 180 degree V and mounting it upside down. A lot of the mass is in the crankshaft area and since mounting it upside down would raise the crank it would defeat the purpose.

Second - Yes you could if you custom designed it. I think it was BMW that recently came out with a turbo V8 with the turbo in the V of the engine to reduce heat loss.

Third - Gears add weight and it takes time to shift. We only do maybe 65 mph at the most, so as long as your engine has a somewhat wide powerband there is no need for the extra gears. Thats also why some of the guys are talking about a IVT/CVT as that could get rid of the efficiency loss from having a low gear count (assuming the IVT/CVT is a really efficient one).

Last - No, its not. Some teams do well with a spool, but your car must be designed with it in mind. It really depends on your goal, I mean if you want to win acceleration than a spool might be the best way to go.



Originally posted by DannytheRadomski:
A few questions:
First, in a V engine, would increasing the angle to greater than 180 allow you to lower the center of gravity at all?
Second, can you route the exhaust in a flat engine through the top (or through the center in a V engine)?
Third, what's with the low gear count? I understand it means less weight and complexity, but adding gears can increase fuel efficiency.

This might sound stupid, but its a differential really needed? Karts dont usually use them, and while it would hurt autocross, the cost, loss of weight, and less time spent could help overall.

Please keep in mind that I am not actually a participant of FSAE yet, but I am trying to prepare before entering college. I don't want to go in blind and naive, so I'm getting an idea of what would and wouldn't work now.

[DannytheRadomski]

Sorry about the first question. I should rephrase that as: What is the heaviest part of the engine, and is there a specific setup that gets that part as low as possible?

[Racer-X]

The crank and rotating assembly are usually the heaviest. The head is probably the second heaviest. How to get it low is something we are all working on.

[Cardriverx]

What racer-X said. Really the only way to get everything as low as possible is to put it all on one plane, AKA a boxer engine. But that kind of packaging has its own issues.



Originally posted by DannytheRadomski:
Sorry about the first question. I should rephrase that as: What is the heaviest part of the engine, and is there a specific setup that gets that part as low as possible?

[Dunk Mckay]

I'd go for a longitudinal 3-cyl 609cc, either a planetary or CVT gearbox, something automatic and reliable to keep things simple for the driver. Probably a built in diff and susp mounts. I don't normally like this because it means your susp geometry is frozen with the engine so can't change year on year or to tweak it for design iterations. But assuming a fantasy case where you can that' what I'd do for stiffness.
Some sort of charging, probably just go for the most reliable system you can think of as power differences between different types aren't gonna matter for squat.
Possibly Direct Injection, and laser honed...at this point I'm just trying to make it interesting, but after all that's what I'd want out of a fantasy engine, to make working with it interesting (not in a "oh I keep having problems to solve" way tho, more "oh wow, we can do this too" way).

[Z]

Originally posted by mech5496:
... about the IVT concept. By having the sun as input and outer ring as output and allowing some slip on the planet cage you can vary the ratio (dependant on slip).
Harry,

Yes, the "slip on the planet cage" controls the IVT's ratio. Unlimited slip = neutral, or "infinite" ratio, and no drive. Zero slip = fixed reduction according to number of teeth on sun and ring gears. Using a brake on the control shaft (planet cage) is quite common and acts just like a clutched gearbox (eg. Model T Ford, ++).

The idea of the hydrostatic control is to allow controlled slip, but not throw the energy away. So the control shaft (planet cage) drives a hydraulic pump which delivers a fixed quantity of oil per rev. This oil is sent to the variable displacement swashplate motor which adds extra torque to the sun gear.

When the swashplate is set for large volume per rev it delivers large torque to the sun gear for a "low" gear (high numerical ratio, including "infinite" ratio, and even reverse!). This also implies a lot of slip at the control shaft, so low output shaft speed.

When the swashplate is set for low volume per rev there is little extra torque to the sun gear (= "high" gear), and little slip of the control shaft so high output speed. The swashplate can also be set for negative volume per rev (pumps oil in opposite direction) which gives extra high "overdrive" gear.

When the swashplate is set for zero volume per rev (swashplate perpendicular to shaft), then no oil flow, no extra torque to sun gear, no slip of the control shaft (planet cage), and the whole assembly acts as a simple planetary reduction. So at this "sweet spot" there is almost the same high efficiency as a single stage gear reduction.

Also, your suggestion of an "electric brake" for the control shaft is how many hybrid cars marry an IC engine to an electric drive these days (not sure exactly which models, they all look the same to me http://fsae.com/groupee_common/emoticons/icon_smile.gif). An electric motor and generator replace the two hydraulic pump/motors in the above description.
~~~~~o0o~~~~~

Ian,

The only extra losses of the above IVT's "sweet spot", above those of a single planetary reduction, is the drive to/from the swashplate, and any "control" flows. But there are no oil flow losses when at this "gears-only" ratio. However, the hydrostatic IVT has high losses (low efficiency) when operating well away from its gears-only ratio. So high losses in low gears (ie. low vehicle speeds), and VERY high losses in reverse gear, because of the high oil flows.

But how much time do you spend in low or reverse? And how much power do you use there? Well, on tractors it is a lot of time and power. So these typically have a range of "overall" gear ratios that set their IVT's sweet spot at different vehicle speeds. So one ratio might have the sweet spot at zero speed for a lot of forward/backward work. Another ratio has it at low forward speed for ploughing. Then another at high forward speed for driving home.

I don't have any hard figures for these efficiencies (maybe Google?). But I know that tractors, along with jumbo jets, are bought with fuel economy strongly in mind. The all-mechanical gearboxes still seem to have a slight edge, but are more work for the driver...

Z

[Jan_Dressler]

Originally posted by mech5496:
Jan, IMO energy recovery from a turbocompound would be minimal, and most probably dependent on the load/RPM, as it is a mechanical system. What about using it in combination with an alternator? That way you could use exhaust gas energy to produce electricity. Get rid of (mechanical) water and oil pumps and replace them with electric ones should free some power. I know, conversion of mechanical to electric energy would lower the efficiency a bit, but IMO you would have greater control over recovered energy. Alternatively, you can do the above powered by recovered energy from braking (small front hub motors) plus turbocompound.

What about an electric turbo? You could run it entirely on recovered energy, no backpressure, complete control over boost, zero lag etc..Thoughts?

P.S. Happy new year everyone!
Minimal? What is minmal for you?
Dependent on load/rpm - sure. You can do it electric, as you suggested, but I think the mechanical way has a better efficiency. Electric turbo is another idea, but again...
You can have all the good transient behavior from a mechanical supercharger, boost control - well, controlled bypass/backflow valve. Also more efficient. Brake energy recovery could supply your electric system with power, as you also suggested.

@Z: Thanks for the nice picture, I was inspired by some prewar Mercedes engines (although these did not have the roller bearings on the webs), a WWII Maybach engine used in tanks:
http://i45.tinypic.com/2d83spl.jpg
and this kind of smaller V-12:
http://mperniss.de/v12/images/Bild%20182.jpg

@murpia: The intake / exhaust thing is even more interesting if you look at supercharged engines (or my proposed combination of supercharged and turbocompound). That's why I proposed a triple (with a strange configuration, but still even ignition sequence every 240 degrees).

@Harry, electric brake... VW developed an electromagnetic gearbox in the 90ies, also quite interesting http://fsae.com/groupee_common/emoticons/icon_wink.gif Here (http://miredaktion.sv-www.de/imperia/md/content/ap/charts/steiger.pdf). Sorry, it's german, but look at the drawings...

@Danny, about the more-than-180°-thing... Think about your intake/exhaust-packaging, if the engine is more than 2 cylinders. Ferrari apparently had some problems with packaging their exhaust system when they went for their 180° V-12 in the 70ies. Exhaust on the inner side of the cylinder V: Old story:
http://encarsglobe.com/data_images/models/ferrari-312/ferrari-312-06.jpg

@Dunk Mckay: Why direct injection?

[Gruntguru]

Originally posted by Z:
Gas exchange would be via two conventional poppet valves in the head, set at a wide-ish V-angle. The turbo-supercharger pressurises air in the intake pipe, which runs between the valves (ie. parallel with cylinder axis). Towards the end of each power stroke the exhaust valve opens (say, 45 degrees before BDC). Cylinder pressure drops, then the intake opens, still before BDC.

Presurized intake air rushes down the intake side of the cylinder, across the piston top (now near BDC), and up the exhaust side and out. It is an upside-down "loop scavenge" system, with no need for ports in the cylinder walls. When the cylinder is flushed the exhaust closes, then intake closes (maybe ~60 deg after BDC?). Piston travels upward compressing air, then fuel injection around TDC. Bang, power stroke, and repeat.

To extract the last drops of efficiency this engine could be "internally air-cooled". The turbo is oversized to pump extra air through the cylinder during gas exchange (remember this is a diesel, so no fuel or oil in the air). This sucks heat out of the hot surfaces directly, rather than the heat having to flow through the walls. This heat energy further expands the exhaust gases providing extra drive to the turbo. The turbo can be "compounded" to send this energy back to the crank (simple way is to "blow" through a Roots type SC). Z
That would require the Roots to be sized/geared to less-than engine displacement ie negative boost when turbo is not helping - not so good for scavenging. BTW, the engine would need a scavenge pump for starting and other off-boost running.

Flat twin has a number of benefits. Perhaps run the intake and exhaust parallel to the crankshaft. This would require some innovative valve gear but the CG could be ultra low.
DannytheRadomski -
The >180 deg concept could then come into play allowing the cylinder heads to "droop" down to the same ground clearance as the crankcase.

My favourite is probably the inline 3, especially if you are turbocharging. Low inertial torque. Even power pulses. Good mechanical balance. Good turbine performance (blowdown with simple manifold). Good restrictor performance without huge plenum.

[rjwoods77]

Originally posted by Jon Burford:
Racer-X,
I like your common sense.
More people should use snowmobile engines.
An electronically controlled CVT, Enter Rob...

If I did it again and used a CVT and had a sharp team of guys I would go with this unit. I'll save the discussion of the benefits and just give you some things to ponder on...

http://burgmanusa.com/bkb/650+...Info#Burgman_650_CVT (http://burgmanusa.com/bkb/650+CVT+Info#Burgman_650_CVT)

http://www.contitech.de/pages/...5_hybridring_en.html (http://www.contitech.de/pages/news/aktuelles/080325_hybridring_en.html)

http://www.mate.tue.nl/mate/pdfs/10396.pdf

Also my fantasy engine would be a dry sumped flat twin with the crank snout pointed straight up in the air that connects to a Burgman 650 cvt that runs the secondary toward the back of the car the inputs into a spiral bevel geared differential that has the pinion snout pointed straight up and is mounted to the engine block. Maybe even a M-291 3.5 Flat 12 head design. It would package the tightest/serviceable/accessable of any engine and drivetrain package in FSAE. Also it would be very cool in in its 917-esqueness since the cooling fan for the cvt is on the primary pulley.You can't really have a fantasy engine discussion unless in includes the transmission and final drive setups with it and I think for all concerns of such this is the best overall. It would be a complete unit that is shaped as close to a "box" that is possible to minimize space waste of leaned banks of cylinders at the same time keeping the major masses as low as possible.

[Charles Kaneb]

Hmm... I see a lot of very complicated and heavy contraptions being wheeled out in this thread. Especially by Z, of "16 identical frame mounts and adding about three tubes to your frame represents a huge amount of weight and complication" fame.

The engine I think I could build the best FSAE car around is as follows.

Past experience shows little advantage to a multicylinder engine in FSAE, and in motorcycles singles are getting more and more popular. The vibration's bad, and they idle at seven zillion RPM, but in exchange you get a physically small engine with a big, efficient combustion chamber at the top of the only hole.

I want to build a tiny, light car. Based on previous attempts to do this, I think the total length of the engine package matters, as does its weight.

Taking a look at virtually every FSAE car, the intake and exhaust take up a lot of room, and almost all of it longitudinal. Air comes in at the front of the engine, exhaust leaves at the back. Yamaha's new no-self-starter racing dirtbike engine does the opposite - but it doesn't shorten the car much, because now the intake goes where you'd like the driver's head to be. Instead, I'm going to relocate the ports to have the intake come in on the left and the exhaust leave on the right.

There are two good ways to do this. One is to go with a SOHC arrangement, and have dual intake ports to the outside, and either dual exhaust ports or an exhaust port beyond the end of the cam. The other is to go with widely-spaced dual cams, fairly long followers, a single intake port beyond the end of the cams and a single exhaust port between them. Each cam would control one intake valve and one exhaust valve. If my calculations and CFD show that I can get enough flow to make good power with a two-valve head that makes it even easier thermally and simpler/smaller mechanically.

For bore and stroke, my answer is "A lot of both". I don't see any efficiency advantage to a smaller-displacement engine that has to turn faster and have wilder cams and bigger ports. Locomotives and tractors have few, big, cylinders. So will this engine. 599cc just in case someone wants to use it in a class with a strict 600 limit.

My power goal's modest. 35kW. That's what a good single can flow through the restrictor, and it's also "enough" in a 150kg car. This allows me to pick up a lot of energy efficiency by specifying low-lift, short-duration cams, nice little ports, and lots of compression.

Now, on to the transmission. I'm expecting a very broad operating range where we'll make enough power to shove the car down the straights very quickly. I'd like simple controls for my driver, and to not add much length, width, or weight.

So instead of a traditional multi-speed gearbox, I'm going to do something fairly similar to a Rotax DD2. On each end of the crankshaft is a clutch connecting the crankshaft to a gear in constant mesh with one on the differential.

Engage one clutch for low gear. Engage the other clutch for high gear. The two will be connected by a shaft that when pulled to the left engages low gear, and when pulled to the right engages high gear, and in between doesn't engage either. The spread between the two gears will be fairly large - one gear to go about 35 MPH for jack-rabbit starts, tight turns, and the skidpad, and the other gear to go about 75 MPH.

Depending on the speed of the engine, its direction of rotation, and the desired axle speeds, I may need to have a pair of very short jackshafts to get enough reduction and have the car go forward rather than backwards. These would be mounted behind the cylinder and above the diff.

Having the diff integrated into the engine case solves a lot of problems. It gets rid of some big external loads (chain or gear tension), it removes the limitations of minimum length and clearance for chains or belts, and gets rid of a lot of exposed mechanical parts. It will also save a lot of weight no matter how bad and beefy my castings have to be. Pick up your chainguard sometime.

On the inside of the driven gears, there could be broad, flat, cylindrical surfaces for you to slide your differential into. The gears already have to be supported by bearings, so all these have to do is spin the diff of your choice. Join them to CV or tripod housings by a tube and have a spool; slide the outer housing of a clutch-type or Torsen between them and slide the splined tripod housings inside it and have a differential.

Wah-lah! About a foot (30cm) saved off the average FSAE car, about twenty-five pounds (10kg) saved in gearbox and drivetrain parts alone, a bunch of unnecessary gearchanges saved, and probably pretty good energy efficiency in the end.

[mech5496]

Right now I' confused between CVT/epicyclic IVT. (Not that any of this will ever happen for my team but more as amental excercise) The Continental claimed efficiency is impressive and so is the control you can have over an E-CVT, and that's why it was my initial choice; however, for efficieny reasons I'd might use a single CVT with twin clutches on the final drive for vectoring.
IVT seems far more complicated and hard to build, but lighter, less bulky and more efficient. Operation should be similar, bjt IMO the CVT is more applicable (easier) for an FSAE team. I'd love to have a comparison betweenN IVT, CVT and a traditional 3 or 4 gear gb in terms of weight, efficiency and packaging.

[Homemade WRX]

As much as I love my boxer engines; I'd go with a longitudinally mounted, heavily canted, parrallel twin here. I'd probably keep it in the 400cc range so that a turbo application still works.

Kicking around a few numbers on a 67mm bore; I'd run 'around' a 63.8mm crank (450cc), allowing 25 m/s at 11,745 RPM (might go with a bit more stroke to reduce lag). With a 1.7:1 rod ratio and about a 20mm compression height, deck height would be around 160.5 mm. So it's still a low CG and block length would be in the neighborhood of 175mm, leaving room for a chain and bellhousing face.

Dry sump and turbo would be mounted on the low-side of course. Dry-sump would be an integral part of the sump, which would house stressed/splayed main caps.

Oh, I could just ramble on and I haven't even touched the heads LOL This would be for a manual car, probably 4-spd.

If going CVT, I'd just make it a narrow RPM torque monster. Small bore, large stroke, short rod ratio...

[Bus_Lengths]

If going CVT, I'd just make it a narrow RPM torque monster. Small bore, large stroke, short rod ratio...

I imagine this is what you were implying, but I believe you mean 'narrow RPM power monster'.



There's a lot of neat ideas in this thread. I'm still disappointed however by the overall trend of wanting to make these 60 hp, 300 lb, single cylinder go-karts - the car would be lighter if we just got rid of the suspension and used super tiny tires too.....

I would only support downsizing the engine IF you can keep the restrictor maxed out (which not many actually do) and thus have a low weight AND 100+ hp. I see so many teams eager to give up 50 lbs of engine weight at the expense of 50% of their power. Given the relative insensitivity of the tires to load at these low weights, it has just never made a lot of sense to me.

Not to mention, I'm not sure it's possible for a single or twin to compete with the symphony a 4 cylinder at 15,000 RPM makes.

I guess you do get one main advantage of the lighter engine. You can place an extra 50 lbs of ballast in the car to try and get that ideal 30/70 weight distribution. I say 'ideal' because you won't have a prayer of lifting the front tires in the acceleration event unless you have at least 70% of the static weight on the rear tires.

[Racer-X]

@Bus_Lengths

I agree with you on some level. A well executed 4 cylinder car can and often does well at comp. At our level just building a "good" car is enough of a challenge. A 600cc N/A 4 cylinder is the easiest way to make 85 horsepower. If a team is willing to build a car around it and deal with the extra mass it allows the team to focus on everything else that it takes to build a relatively easy to drive fast car. I'm sure I'm preaching to the choir when I say a chain is only as strong as its weakest link and building a solid car trumps building an "awesome" one. No point in going to a single and saving 60 pounds if your suspension geometry is a cluster, or worse, the car doesn't run...

The counterpoint to this is if a single is done correctly and teams can build a good car around one it has a slightly higher scoring potential with fuel econ and all, look at GFR. When teams are able to build a 320 pound RELABLE car with a single they can have a well balanced car. I love power but I'll be the first to admit an FSAE car does not need 100 horsepower, not even 85, most "good" 4 cylinder cars are overkill. The tires can't handle 1500+N of thrust a piece in a 550lb car+driver (our car will spin the tires in 2rd gear any time). With that every pound you save is less mass that needs to be accelerated, both down a straight and in a corner. In theory a single looks pretty good, and some teams have pulled it off, and that makes it easy to get caught up with the idea.

[L B0MB]

@Racer-X
I agree that the 600cc 4cyl is the best choice for amateur teams. It is strong, reliable and has plenty of power to make up for driver mistakes. The only places it suffers are fuel economy and grip in the wet.

As for a fantasy engine... How about a 600cc inline 3 cylinder, miller cycle, turbo, intercooled, direct injection, cooled EGR, rearwards canted cylinders, front facing intake, integral diff, flat shifting 4 speed sequential, centrifugal clutch

[Z]

Jan,

I wasn't aware of that Maybach engine. Its crank is remarkably similar to the Tatra's, yet quite different to all other engines (even the Hirth jointed, roller bearing cranks common at the time).

I wonder who was looking over whose shoulder? http://fsae.com/groupee_common/emoticons/icon_smile.gif
~~~~~o0o~~~~~

Gruntguru,

I would size the Roots SC to about 1 to 2 x swept volume for good low speed flushing and internal air-cooling. The turbo wouldn't need a wastegate. Any turbo overboost would spin the SC, sending power to the crank.

Remember, both intake and exhaust valves are open around piston BDC (for ~ one quarter of each rev). So overboost from the turbo compressor spins the SC, then blows through the cylinder and back to the turbine, picking up heat energy on the way. Easier to explain with pictures...
~~~~~o0o~~~~~

Charles,

"I see a lot of very complicated and heavy contraptions being wheeled out in this thread. Especially by Z..."

Hmmm, methinks this is a leg pull, but nevertheless...

I started with a 20kg, 120hp, 450cc single, with clutch and direct drive to an open diff. Then I downsized it to 100cc, added some mass to the flywheel, added two small hydraulic pumps, and deleted the (quite big and heavy) clutch and rear brakes, for similar power, perhaps less weight, and much better fuel economy. Err, ... "complicated and heavy"...???

"I want to build a tiny, light car."

I do like that part, and what followed... http://fsae.com/groupee_common/emoticons/icon_smile.gif
~~~~~o0o~~~~~

Harry,

An IC engined car with a CVT needs a clutch, but no clutch if it has an IVT. The CVT can do any ratio between low and high forward gears, so needs the clutch to get started. An IVT can do any ratio between reverse and high forward gear, including the "infinitely" low ratios (numerically high) around zero vehicle speed.

Most IVTs use some sort of CVT together with a planetary gearset to shift the CVT's ratio range. The hydraulic gearpump + swashplate is itself an IVT. But by combining it with planetary gears it is possible to get a useful ratio where the hydraulics are stationary, thus giving better efficiency overall (hydraulics lose efficiency at high oil flow rates).

BTW, all those "Zero Turn Radius" mowers currently flooding the market use gearpump + swashplate type drives. About 30 years ago I was told by very experienced engineers that such hydraulic systems would be "impossibly" expensive for ride-on mowers. Engineers! Hmphh!!! http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif
~~~~~o0o~~~~~

Bus-Lengths, Racer-X,

The main performance advantage from a lighter car comes when you add aero downforce. The downforce achievable is pretty much the same for heavy or light cars (depends mainly on plan area), but you get much greater cornering Gs with a lightweight car.

In fact, IMO, 3G cornering should be quite easy with a simple, light car! http://fsae.com/groupee_common/emoticons/icon_smile.gif

Oh, and as mentioned before, the maximum power through the restrictor of ~120hp was possible 40 years ago from an air-cooled, turboed, 450cc single.
~~~~~o0o~~~~~

Well, going "bush" tomorrow for ~2 weeks...

So happy fantasizing ... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[mech5496]

Z,

As mentioned, I would go with twin 'final drive' clutches, one for each rear wheel for vectoring, so the transmission would spin all the time. The E-CVT control unit could let the engine rev up to a certain limit (lets say 500rpm above the max power RPM) when the driver pulls a clutch paddle, while at the same time would disengage both clutches. Full throttle, engine revs up, let go the paddle, ECU slips the clutches (to avoid belt slip, otherwise just dump them), car starts accelerating, ECU reads front wheel speed so it starts to shift the CVT to keep the engine on the max power RPM, off you go. Not a very common CVT launch... http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Homemade WRX]

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

If going CVT, I'd just make it a narrow RPM torque monster. Small bore, large stroke, short rod ratio...

I imagine this is what you were implying, but I believe you mean 'narrow RPM power monster'.
</div></BLOCKQUOTE>

No, I meant torque but failed to mention 'relative' to FSAE engines.

If I wanted power, I'd go with my turbo CBR250RR engine again.

[MCoach]

Fantasy engine for FSAE?

Simple, stupid, simple.

Single cylinder, 250cc, 4 valve, Crank masses kept low in weight and radius to enable lower CG.
Two speed transmission, integrated.
Salisbury differential, also integrated.
Supercharger mounted on the back of the engine between cylinder jug and diff.
Aluminum or magnesium block.
Carb that thing, let's go racing.



Just to add more fuel to this:
Wheels and tires: 8" in-house customs.
Shifter: on steering column (a la shifter kart)
Chassis: Minimum template fit
Electronics: Spark only (no driver signals, extraneous sensors). Coil/points run off flywheel, single wire ignition.
Driver: 85lb sickly horse jockey
Miscellaneous: Reduction of part count focused on the entire car. Lots of aluminum and lightweight.


In a game of no minimum weight, I don't care how much power is available. If anything can be smaller, lighter, more efficient, do it. It's a race to the bottom, anything otherwise is wasting time. This I would only put secondary to driver development in the game of FSAE. Want to pull 1-2 seconds faster in the same car without changing anything? Let the driver play around with the car, learn it, anticipate it.

[Will M]

Well if we expand the idea to include set ups for a formula car that we would want to own rather than a set up that we think would be best for winning a formula competition I would go electric.
I confess that I have never work on an electric team and don’t know too much about them but this is the design I see in my head.
--------------------------------------------
I would have one motor driving each rear wheel with a single gear reduction (if a reduction is needed).
And as I would have the car set up for real autocrosses instead of FSAE events the battery would be much smaller (roughly 10% the size of a typical formula battery pack).
The battery would be set up for quick swaps; i.e. a fresh battery for every run.

The current electric cars can match the pace of many ICE cars and if you remove 90% of their battery weight I think you’d have a killer combination.
Lower cooling requirements would help with aero designs.
A smaller drive train would help with packaging.
Independent electric motors lead to traction control and torque vectoring.

Of course it wouldn’t have the energy to finish the endurance event but the ‘weekend autocrosser’ goes to autocrosses not to endurance races.

Just my 2 cents.

-William

[JulianH]

Will, electric is a nice choice (Both the Delft and the AMZ car of 2012 weighted around 110kg without batteries (and without aero on our side) and had about 85kW).

Take enough energy for 2 AutoX laps and you are still in the "lightest FSAE car ever" region.

But, electric cars are sooo expensive, I think therefore the "weekend autocrosser" would still choose a combustion car.

[mech5496]

I thought this was an IC-only thread. IMO endurance is an integral part of the comp and cannot be ignored, but since this is a fantasy thead...

Given the 85kw limit, take a look at Delfts' DUT12; 85kw, 145kg, 4WD. Now cut about 15kg from batteries (their battery pack was already really small) and add aero. What you have is what I expect DUT 13 to be, but lighter.

If you forget about the 85kw limit, take a look at our car. Ditch 75% of battery capacity (and about 30kg) and move what is left on both sides of the cockpit for less MoI and less rearward weight distribution Add a second motor, and of course, lots and efficient aero. In direct drive configuration and 10", the car has a top speed of 172kph, just right for autox, aint it? Boom, 220kg, 200kw (272hp), 1500Nm torque at the rear axle.

Wanna get completely nuts? Ditch the voltage limit and substitute the above motors with two Yasa 400... Two one-stage gearboxes (one per motor) with a reduction of 3,75:1 to get the same top speed as above. Being 3kg/motor lighter would keep the weight about the same, while placing the motors lower and a bit forward would account for better GC/MoI. OK, numbers! 220kg, 330kw (449hp!) and 3000Nm torque on the rear axle... thats more than 2000hp/tonne! Plus the massive torque! I know, it is a complete overkill, but definitely impressive, aint it? http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Jon Burford]

But, electric cars are sooo expensive, I think therefore the "weekend autocrosser" would still choose a combustion car.

I thought that was the whole point of Formula Student?

[MCoach]

Having worked on robots that come to a halt without warning nor signs of what's wrong, electric is definitely not something for a weekend autocross project. Things just stop and everything looks ok. Could it be a loose wire? A cracked solder joint? A misplaced connection? Maybe that NPN transistor we never tested before installation decided to short itself. With so many things to go wrong, it can become very tricky and time consuming to track down the problem. Ask GFRe about that.

When was the last time you took apart your computer because it wouldn't start properly? Did you fix it? You'll find that it comes down to something stupid like corrupted code or a blown capacitor on something else obscure like the graphics card. What typically happens? The part or system is diagnosed on a bench top and then replaced. Fine. That's easy, swap out audio card, swap in new one.

...But what if it's your car?

If these cars were meant to race in a series where the focus was on pure winning in an autocross setting and the project was able to be supported by professionals, then sure, I'll concede the advantage electric has on being able to point the car exactly where you want it.

[Jon Burford]

@M
I don't really mind the complexity and as I have said previously, I have a complete love of the AMZ car from 2012 and I am massively impressed by the tech and effort that it takes to make them work.
Despite that I would never buy an electric car to go racing in, I just don't see it for myself, the cost being one and the experience another.

The IC engine is an ancient mechanised wonder, an almost laughable method of turning a shaft. Why is it then that not even modern tech can drag me away from it?? It's the sense of drama, the vibration, the noise, the smell. There is something about that. This isn't just me going on, being a potential costumer for a weekend race car myself, that's what I want, not an electric powered anything and I'm sure I'm not alone. That has to count for something.

[Owen Thomas]

I can agree with the sentiments posted by MCoach and Jon. It's fair to expect that your typical weekend warrior can check suspension bolts and change the oil, and generally troubleshoot regular automotive problems with the (vast) resources available for IC engines. Is it equally fair to expect that they can check for electromagnetic interference in a sensor wire? Or stare down miles of C++ to fix a problem with one of the control systems? Probably not.

At the same time, I do not think these opinions are particularily valid for this competition. Sure, the idea of making a weekend autocross car for retail promotes making a robust, well rounded design, but the reality is that the majority of competitive FSAE vehicles are not this way. A car made to be sold in bulk (>1) would be far more driven by the marketing needs, and take multiple years and iterations. We have inexperienced students making a prototype in (typically) a year or less. We also have the unique advantage that the designer of any given system is usually around to troubleshoot it when something goes wrong - something not available to the consumer.

This argument comes up all the time around here, but I think it's important to remember that the "weekend autocrosser" statement in the rules is outlying a premise for the competition. It is not the goal. That said, some very competitive teams build the car to clean house at normal autocross events (eg Jayhawks, Maryland), but the difference is still that these are 1-off cars, and would be fundamentally different if they were made to be sold.

Anyways, this topic was about fantasy "engines", but I also agree that the potential advantages of electric drive are pretty tempting. Additionally, I think it's interesting that this thread has evolved more into "what is your fantasy engine package", instead of just "engine". You can make almost any size and power work (as shown by the many ideas posted), but in the end it really just comes down to packaging.

[BluSTi]

You guys sure like V's here. Not really sure why http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif.

My ideal?

450cc turbo boxer.

[BluSTi]

There's some other stuff I want too, but I won't explain it here since I want to see if we can build it first. If not, then I'll explain the whole thing. http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Markus]

Originally posted by Jon Burford:
The IC engine is an ancient mechanised wonder, an almost laughable method of turning a shaft. Why is it then that not even modern tech can drag me away from it?? It's the sense of drama, the vibration, the noise, the smell. There is something about that. This isn't just me going on, being a potential costumer for a weekend race car myself, that's what I want, not an electric powered anything and I'm sure I'm not alone. That has to count for something.
Remember, IC engines are the modern choice in automotives. It was preceeded by electric motors and steam power, so actually you don't want to go with the "ancient" method... http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Will M]

What you would want for a FSAE event car, a personal FSAE car, and an OEM level car would be quite different.

For and event car I would go super simple like MCoach.

For a personal car I would want I described or the monster mech5496 described.
Keeping in mind that it is a ‘fantasy engine’ I would fantasize it to be cheap and reliable, so there http://fsae.com/groupee_common/emoticons/icon_wink.gif
And if the fantasy included super energy dense batteries then it could be an event car.

And an OEM car to sell…. well that is a completely different beast but safety, reliability, and cost would rule your design.

-William

[Homemade WRX]

And an OEM car to sell…. well that is a completely different beast but safety, reliability, and cost would rule your design.

-William

Well that's my everyday, consulting, real world. Penny-pinchers don't exist in my 'fantasy' http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Jon Burford]

Originally posted by Markus:
Remember, IC engines are the modern choice in automotives. It was preceeded by electric motors and steam power, so actually you don't want to go with the "ancient" method... http://fsae.com/groupee_common/emoticons/icon_wink.gif
Granted that they were using electricity before, but I am talking about the new modern "electric revolution"
I have been to Finland a couple of times, I didn't see much demand for electric racing http://fsae.com/groupee_common/emoticons/icon_wink.gif ...

Anybody think there should be an event where the team tries to sell the car to a potential buyer rather than a design/business judge, a real world sort of test?
where actual cost may come into it, along with what they might think of the car?

[BluSTi]

Originally posted by Jon Burford:

Anybody think there should be an event where the team tries to sell the car to a potential buyer rather than a design/business judge, a real world sort of test?
where actual cost may come into it, along with what they might think of the car?

That would make for an interesting exercise. Perhaps as an addition to the "cost" event.

[MCoach]

Jon, I also hold an appreciation to what electricity can do. I'm trying to do some modelling currently to possibly roll out a system control on our car sometime in 2014. The part that worries me most is the fail safe of electronic systems and what happens when there is a partial or full system failure. Is the car going to be able to continue on by falling back on a mechanical system or if the computer locks up on some segment of code during a particularly bad time, will it be recoverable? It's just those extra worries that have to go with it of the unseen things that might happen.

Fortunately, I've got some time to put towards this and hopefully work out the kinks -- some of them.

[Jon Burford]

Originally posted by BluSTi:

That would make for an interesting exercise. Perhaps as an addition to the "cost" event.

I think it's more worthy than the current cost event in it's current form. unfortunately it requires a accurate car cost. enter cost report (but one where you cost your car by actual costs i.e with NON DISCOUNTED invoices, records of machine time used, records of build time. sounds long winded, I know I would rather have done than than a normal cost report though.)

MCoach,
My experiences don't really lie in the making working of computer things. my experience stops at wiring things together, pluggin a laptop in and filling in boxes http://fsae.com/groupee_common/emoticons/icon_wink.gif although I understand what your saying and I guess it comes down to having some kind of foolproof, system that you test the hell out of. A brief look at the rules shows me various safety things, multiple go-pedal sensors etc. But I guess if whatever is monitoring them has a fit you could be in trouble.
I hope you have fun with it http://fsae.com/groupee_common/emoticons/icon_smile.gif

[murpia]

Originally posted by BluSTi:
You guys sure like V's here. Not really sure why http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif.

My ideal?

450cc turbo boxer.
V's package well (longitudinally).

What's a boxer if not a 180deg V? Doesn't that result in too high a crank height, cramped exhaust routing and no space for accessories along the sump?

What port geometry do you propose for your boxer turbo?

A turbo V works quite nicely: Put the plenum in the V and the turbo behind the engine above the gear cluster. Oil pumps down one side of the sump, alternator down the other. Shallow out the V as far as you can without cramping the exhaust exits (120deg?).

Regards, Ian

[Mbirt]

Originally posted by murpia:

What port geometry do you propose for your boxer turbo? Personally, I propose an air-cooled, splash-lubricated flathead boxer twin with as many off-the-shelf parts from the Briggs and Stratton Blockzilla jr. dragster platform as possible. A horizontal crank would place all four ports vertical, a vertical crank would place all ports pointing rearward. The single camshaft actuating all four valves means no special second timing chain with custom drive sprocket at the crank. If you really think you need more than 50 hp, you could run an aerocharger and give the restrictor a good choking.

[mech5496]

Ian,

a boxer could package neatly like Rob suggested. You can have the intake ports pointing front, exhaust pointing rearwards. Plenum can take up the empty space between drivers back and engine.

[BluSTi]

Originally posted by mech5496:
Ian,

a boxer could package neatly like Rob suggested. You can have the intake ports pointing front, exhaust pointing rearwards. Plenum can take up the empty space between drivers back and engine.

Now you're on to something...

[murpia]

Originally posted by mech5496:
Ian,

a boxer could package neatly like Rob suggested. You can have the intake ports pointing front, exhaust pointing rearwards. Plenum can take up the empty space between drivers back and engine.
OK, yes, I can see this setup packaging nicely, but I have a hard time imagining (fantasising?) the camshaft drive(s).

http://www.subaru.com/content/engineering/boxer_displacement_25l/Turbo%20Boxer%20Engine_282x160.png

Assuming that image link works, it highlights my CoG & crank height concerns ref a flat engine vs. a V...

Regards, Ian

[Simon Dingle]

Originally posted by mech5496:
Ian,

a boxer could package neatly like Rob suggested. You can have the intake ports pointing front, exhaust pointing rearwards. Plenum can take up the empty space between drivers back and engine.

Why would you want the empty space between the engine and the driver's back to be filled with air? I would rather that empty space be filled with a crank.

My favourite so far is the idea of a 450cc single pushing 25bar BMEP. Small, light, low-speed and high torque. If it could be air-cooled then that would be amazing, but I can't find the engine/car that Z's talking about and I'd worry that you couldn't get sufficient air-flow at the speeds FSAE is conducted at.

I quite like the v-twin as it does look pretty and seems to package nicely. But as far as I can see a v-twin (or flat twin) has double the heat loss of a single and double the frictional losses (not actually double but you get my point) and double the drive-belts and double the camshafts etc, etc, and for what benefit? Improved air-flow conditions through the restrictor is the only useful one I can think of. EDIT: I forgot about the improved balance of a twin.

[mech5496]

Ian,

If you take a look at Rob's post, he suggested an upward pointing output shaft.Just rotate the boxer by 90deg (possible only with 2cyl)

Simon,

I certainly do not. Roll back a couple of pages and you will spot my favorite... http://fsae.com/groupee_common/emoticons/icon_wink.gif

[rjwoods77]

1) I would consider a flat twin over a single due to having an engine that meets a happy medium of having a smaller downsized engine that would allow for the flexibility of learning and power levels that you get from a multi-cylinder engine, have improved VNH characteristic which are in realty a selling point and just for the symmetry of it all.

2) I would consider a flat twin over a v twin because everything would package better, have lower cg, waste less space and also wouldnt have an odd firing.

I have used a 90 degree v-twin in my car so I have been intimately involved with the packaging of one and many 2am conversations with myself in the shop about how it was 90 degrees short of where it would be best.

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

...and as far as I have seen might possibly be the tightest packaging of a drivetrain vs rear suspension I have ever seen as the main roll hoop plane to rear axle center line was 19 inches. As you can see the rear axle sat underneath cylinder number two and there was room for any of the FSAE sized diffs to sit under there.

Murpia:

Due to the length of runners and plenum size required for a v-twin you cant really package the system between the V. It actually ends up being above the V which does take up a fair bit of space. Also the exhaust routing gets really pesky due to the longitudinal versus transverse nature of the exhaust ports. The Briggs v-twin ports are 90 degrees out of plane of the connecting rods unlike most bike engines and actually helped with the packaging constraints. Also the sump on that Subaru engine is big and the mounting of engine height wise in those cars is based of the flywheel/clutch diameter and not that sump size. If they used a small clutch, like one that you would use in FSAE you can effectively hit the deck with it. Also if you had the exhaust ports pointing down the size of that 450cc to 600cc port wouldn't be that big and thus using the correctly sized 304SS 90 degree bend coming out of that port would be a non issue given the small radius it would possess. Another option could be to a 90 degree flow path versus a 180 crossflow head but you will sacrifice for the packaging. Another thing is with the spacing a flat twins intake and exhaust ports it works out just about perfect for the runner lengths required.

Mbirt:

In my experience with the Briggs V twins you could use the pushrod heads off of the engine I used to make what you are describing pretty easy but if you were to do an engine without water cooling you might as well make your own heads since there are no cavities inside them. Maybe even do a Dominion 32v head style!!! The only bitch about an aircooled engine is knowing how hot it is even with CHT's and the respective issues of tuning under such uncertainties. You would have to rig up a ton of thermocouples on your first engine to determine the wheres and hows of how to get effective cooling on it. I know Briggs already made flat twins before but with how super simplistic the engines are I would just go and make my own. To add to that I would definitely build it with a vertical shaft with also vertical intake and exhaust ports. I would slam the engine directly against the main roll hoop plane and manifold toward the back of the car. I sketched up a ton of prelim stuff a number of years ago and it works out perfectly with the cvt and everything.Here is some redneck fun with an old one...

http://www.youtube.com/watch?v=Fg3nZMdAbP0

...also I would entertain the thought of doing an open source design and manufacture of an engine and drivetrain for a team(s) to use if the right people were to get together on it and there was no rules infractions by making this package and letting a team roll with it.

[AxelRipper]

I see the lawnmower crew has beat me to the punch in suggesting that the best way to package a flat motor is to make it pushrod for the port design (or a flathead as MBirt suggested)

And as for the 450cc/cyl air-cooled super-engine from 1972 that Z was talking about: http://youtu.be/nsuQo3dYPYI

[L B0MB]

For those wanting a boxer engine with front / rear ports, BMW has been doing them for years.

Here is the older R1200:
http://www.blogobike.it/wp-content/uploads/2011/01/lg+2005_bmw_r1200gs+engine_diagram_view.jpg

Another pic of the older head showing valve actuation:
http://www.rtrider.co.uk/12%20engine.jpg

Here is the revised R1200 for 2010:
http://www.allaboutbikes.com/picture3/2010bmw/Mergded%20copy.jpg

And the more traditional layout with watercooling for 2013:
http://bikeadvice.s3.amazonaws.com/wp-content/uploads/2012/10/BMW-Intermot2012-R1200GS-engine.jpg

[mech5496]

OR you can have a 'vertical' crankshaft and normal head design...

[Will M]

@ L BOMB

So at least for the 2010 model each cam shaft actuates one intake valve and one exhaust value, right?

-William

[L B0MB]

Originally posted by Will M:
@ L BOMB

So at least for the 2010 model each cam shaft actuates one intake valve and one exhaust value, right?

-William

Correct.

I've added another pic to my original post showing the valve actuation for the earlier model too.

[murpia]

Originally posted by Z:
In later years I might add more mass to the crank so it can act as a flywheel energy store. This stored energy would be utilised through an Infinitely Variable Transmission, built from a single planetary (epicyclic) gearset with hydrostatic control. Engine runs mostly at WOT. Accelerator pedal linked to IVT so engine/flywheel speeds up during no-thrust cornering (storing energy), and slows down during max-thrust acceleration on straights (releasing the flywheel energy).

I figure this small energy store would allow the engine to be downsized to about 200cc (~60mm B&S). The smaller piston, cylinder, etc., offsets any extra mass in the flywheel. Small engine running constantly at WOT gives very good fuel efficiency (compared with big engine sucking air through half closed throttle, and hence low real CR).

The IVT has crank driving the sun gear as "input", either the planets or ring gear are the "output" shaft, and the hydrostatics control the third "control" shaft. The hydrostatics use a fixed displacement pump (eg. gear, roller-vane, etc.) connected to the planetary control shaft, and variable displacement pump (eg. swashplate) connected back to the crank/sun-gear, giving the variable ratio and torque multiplication (like on many tractors these days! http://fsae.com/groupee_common/emoticons/icon_smile.gif).

The various gear ratios are set so there is no oil flow at a "sweet spot" of about 60kph, so no viscous friction losses. Ie., the swashplate is set for zero flow at 60kph, thus "locking" the planetary control shaft and giving gear-only transmission. The engine/IVT package has to work either side of this sweet spot (for flywheel energy storage and release), but the oil flow rates are low so only small friction losses.

Optionally, the hydraulic pressure can also drive the slightly unconventional open-diff for torque vectoring (hint: the diff's planet gears act as a hydraulic gear motor). Can explain in more detail later...
Z
Hi Z,

I was thinking a bit more about this concept, and I have a question about control:

Is the hydrostatic IVT you describe ratio controlled, or torque controlled?

A push-belt CVT is ratio controlled, i.e. the actuator selects the ratio and barring compliance & traction-slip effects this relationship holds. So, if as an example we select a constant CVT ratio of 1.5, and the engine is at 3000rpm and accelerates to 6000rpm, the car is going twice as fast. The amount of time that takes is down to the engine torque output, which is under completely independent driver control (via the throttle pedal).

For your high-inertia engine crankshaft energy store, accelerating the engine is going to waste power that should be used to accelerate the car. So what you need is that when the driver asks for torque, the inertia of the car and the inertia of the engine are 'pulled' together by a torque actuator (the IVT). They can also be 'pushed' apart again during braking, to implement a kind of KERS.

That would not work if the CVT element of the IVT is ratio controlled. In that case the CVT torque is proportional to rate-of-change of CVT ratio, which needs to be closed loop controlled to generate the target torque, in the presence of compliance & slip. At the very least a decent torque sensor will be needed.

Regards, Ian

Edit: This issue is also present in some concepts for torque-vectoring differentials I have seen presented. They talk about speed control of the rear wheels as if that means torque control. With a non-linear tyre slip ratio vs. contact patch force relationship, making that assumption is pretty flaky.

[Will M]

@ L BOMB

So one side would be just under 600cc...
Could be an option for a large air-cooled single cylinder engine...


-William

[mech5496]

If you want a large, air-cooled single, go with a Yamaha XT600! We did back in 2007/2008... http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

[Mbirt]

Originally posted by mech5496:
If you want a large, air-cooled single, go with a Yamaha XT600! We did back in 2007/2008... http://fsae.com/groupee_common/emoticons/icon_biggrin.gif Harry, the XT600 is relatively rare in the US, but the Raptor 660/700 ATV is quite popular with a killer aftermarket. Do you know if anything is compatible between the two motors? I know the Raptor is water-cooled, but if certain components are shared, you could really turn an XT600 into a performer. What was your team's experience with it like?

[mech5496]

Matt,

The XT600 is a completely different engine from the 660. Actually, newer generation XTs come with the water-cooled 660cc motor, which is almost identical to the 700cc Raptor. The 600 is a massively overdesigned engine, so a bit on the heavy side. It is also quite bulky. With a 9:1 CR it is quite a sleeper in its' stock form too... Ours had extensive modifications on CR, piston, conrod, crank, valvetrain, ports and camshaft, but I cannot recall power/torque figures. We also had issues to make it run properly on our Haltech E6X, but I suppose this was due to being our first ever single. We also noted some cooling issues. Another important aspect is that the head has twin intake and exhaust ports, which makes intake/exhaust design a bit tricky for a single (but no match to a 4pot though...). Our engine leaderhasd great plans for the poor XT, but thed the team decided to go the WR route. Nevertheless, the XT is one of the best sounding singles ever IMO, whichmade us saying 'it has huge square steel balls' http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Back on topic, I believe there was an Italian team running the same engine back in 2006. Same team also ran a destroked, watercooled 660 a couple of years later, no idea if they still run it though.

[DannytheRadomski]

Is there any advantage of a Boxer over a 180 degree (flat) V-twin engine?

[rjwoods77]

Quick question for the "couple gears people". What would be your ratio spread on those gears?

[MCoach]

The "couple of gears people', such as myself, base that on how many gears are actually currently used our cars. I can say that we typically only use two gears regularly, 2nd and 3rd.

1st is typically only for launches, which can typically be replaced with a higher rev drop from 2nd. I mean, most of that is based on the torque curve from the engine. I would pick just our 2nd and 3rd ratios and fit a sprocket to compliment the choice. I've only ever used 4th once and that was on a track completely non-representative of an FSAE track.

That gearing would be well representative of a single cylinder engine.

[Racer-X]

Like what MCoach said you really only need 2nd and 3rd for most courses. On our car we used a large sprocket and removed 5th and 6th. They were simple to take out and we wanted to play it on the safe side and leave 1st and 4th in just in case.

[mech5496]

I will double MCoach on this. Actually we made it through the endurance in FATA with 2nd gear only, and pretty competitive laptimes. Also, we have never used 1st gear (too much wheelspin)

[JWard]

Agreed, especially these singles with their lovely wide torque bands, the twins too to a slightly lesser extent.

[rjwoods77]

I was asking for some rough numbers in order to compare the range of a couple CVTS versus the an equivalent gear spread of a fantasy engine. If you have actually used them in competition then even better. Please fill out the below as an addendum to your comments...

Redline RPM:
Tire Size:
Final Drive Ratio:
Gear Ratios 1,2,3,etc:

[AxelRipper]

Rob,

Our current setup is a WR450, and if what I have is right (spreadsheet kinda blown up and can't find the good one right now)

Redline: 11,000
Tire size: 20.5x6x13
Final Drive: 2.76
1: 2.417
2: 1.733
3: 1.313
4: 1.050
5: .840
Primary reduction: 2.652

If any of these are wrong, I'm sure MBirt will correct them.

I've also done some work looking at it, since we only use 2nd and 3rd right now for most of our driving, and with some final drive changes we'd be able to run just 1st and 2nd with almost the same results as 2nd and 3rd. In reality, we could probably get by with only 1 gear and a centrifugal clutch on most courses, but our test lot is primarily 2nd gear courses, and at comp we may be nearing 4th in spots.

[mech5496]

WR450 here too, so primary and ratios should be the same... 10500 redline, final drive: 13t front 46-52t rear sprockets, pretty sure we use 48t at comp. 10" tires, I think they are 460mm dia.

[Rob Giovenale]

The WWU V8 drivetrain served as a fully stressed member which provided great packaging and weight with very low cg.
The gen 2 drivetrain weighed 131 lbs which was 12 lbs lighter than the prototype. Crank c/l is ~75mm above bottom plane. CG of engine is approximately 75mm above crank c/l. Drive output is 18.7 inches from front mounting.
Here is an ancient drawing with some dimensions. http://i48.tinypic.com/2hmd5sj.jpg
My fantasy engine would be a modern evolution of this package with features like crank c/l under 57mm, improved robustness, packaging, calibration, and other great gains.

[Mbirt]

That's awesome, Rob! Thanks for sharing. The Viking 30 captured my interest when I was still in high school and I'm sure it's at least partially responsible for getting others involved in FSAE too.

[Z]

Ian,

"Is the hydrostatic IVT [plus flywheel energy store] you describe ratio controlled, or torque controlled?"
-----

It would be "torque controlled", as you rightly suggest here;

-----
"... So what you need is that when the driver asks for torque, the inertia of the car and the inertia of the engine are 'pulled' together by a torque actuator (the IVT). They can also be 'pushed' apart again during braking, to implement a kind of KERS.

That would not work if the CVT element of the IVT is ratio controlled. In that case the CVT torque is proportional to rate-of-change of CVT ratio, which needs to be closed loop controlled to generate the target torque, in the presence of compliance & slip. At the very least a decent torque sensor will be needed."
-----

The physical implementation of the "controller" could have electronics in it, or be entirely hydraulic/mechanical. I haven't fantasized too hard about this yet, but I feel the "decent torque sensor" should be the driver. Essentially, the driver's right foot demands some rate-of-change of IVT ratio, and his/her bum decides if that gives enough rear wheel torque. Ie., it is a "driver in the feedback loop" controller.

I should note that this "engine-flywheel as energy store" would only work well in a narrow range of situations, such as FSAE Enduros that can be predicted reasonably accurately beforehand. It would not be suitable for more general purpose driving, such as with passenger cars.

This is because the goal would be to have the engine running constantly at WOT. At each corner the car slows down and the flywheel speeds up, acting as a KERS. During the next straight ALL that stored flywheel energy, and NO MORE, must be returned as car KE. If not, then the engine/flywheel either progressively slows down or speeds up too much. The output of the 100cc engine at WOT would have to be accurately matched to the average frictional power loss of the car over the course of its journey.

A better general purpose system would have an IVT between engine and Flywheel Energy Store, and another CVT between engine and FES. The FES would operate at much higher revs than a typical engine (because KE=1/2 x I x W-SQUARED), and would probably be in a vacuum container (Google "Flybrid"). This way the engine can always run at WOT for good efficiency (not being strangled), and its power output can be modulated by running it at different speeds.

But also note (!) that the main reason I suggested a heavy, REVERSE ROTATING engine flywheel for FSAE use is that is gives a very useful anti-roll gyro couple during cornering. It helps keep the inside wheels planted! So you can have a narrower car, open-diff, etc. This gyro effect is dependent on angular momentum (= I x W), not energy (with W^2), so MoI of the flywheel is relatively more important than its speed.
~~~~~~~~~~o0o~~~~~~~~~~

Danny the R,

"Is there any advantage of a Boxer over a 180 degree (flat) V-twin engine?"

Well, the Boxer and 180-deg-V-twin share similar overall shape and I/E piping, so the main differences are mechanical balance, crankshaft shape/complexity, and smoothness of power delivery.

The Boxer (as in L BOMB's pic below) has good natural balance without any extra balance shafts. But it does require an "S-shaped" crank with two Big-Ends set 180 degrees apart. This makes the crank somewhat flexible. This can be fixed by putting an extra main bearing between the two BEs (eg. Porsche and Subaru flat engines). But now the crank webs must be very thin (P and S), or there must be greater offset between the cylinders.

But greater offset between the cylinders gives a greater "couple unbalance" (from all harmonics), messing up the initially good balance. But even without the central main bearing (just a thick web) the Boxer still has some couple unbalance. And lo-and-behold, looking at L BOMB's pic it seems that BMW have decided that a balance shaft to eliminate this couple is necessary! See the counter-crank-speed, gear driven shaft running through the chain driven cam-drive shaft directly under the crank. Geez, those Bimmer riders must be soft!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif

http://www.blogobike.it/wp-content/uploads/2011/01/lg+2005_bmw_r1200gs+engine_diagram_view.jpg

A 180-deg-V-twin would look like the pic, except that its crank would be like a single's, with only one wider BE with both conrods on it, either side-by-side or with one forked. This makes for a simpler, stiffer crank. But with both pistons moving together it has almost double the "shake" unbalance of each single cylinder. I say "almost" because only the primary unbalances add up, whereas ALL EVEN ORDER HARMONICS CANCEL! This feature of 180 deg V's is a good thing.

So fitting this sort of engine with two counter-crank-speed balance weights gives "perfect" mechanical balance in all respects. Well, only for the forked-rod layout, the side-by-side rod layout still has a very small even-order couple unbalance, about 15% of the Boxer's unbalance.

That just leaves the power pulses. The Boxer has 360-360 power delivery, whereas the 180 deg V-twin has 180-540. The V-twin's power delivery will naturally be "lumpier" at very low revs, but at anything above a few thousand rpm you should only notice a difference in sound.
~~~~~o0o~~~~~

Mbirt,

Perhaps your flat-twin, flat-head Briggs could be done as a 180 degree V?

You have to add some balance weights and two cam lobes, but the crankshaft becomes much simpler and stiffer and the engine has even less vibration.

Z

[DannytheRadomski]

Like Mbirt, the Viking 30 is the reason I'm here, and I intend to build something similar to it when I get into FSAE (I know why that's ridiculous, but I want to try it). That engine is even what inspired this thread. Thanks, Rob, for putting those up.

[Mbirt]

With my praise for the Viking 30 stated, I would also like to share my love for the WR450f single. A few weeks ago I found the power curve from our 2007 car with the Mahle motor. Its peak power was only 5 hp greater than the peak sprocket output of our current WR450f setup and the single has the smoother, wider racing power band to boot. The Mahle made more torque "off-tune" at low speeds, but not 33% more as the displacement difference might lead one to believe.

So maybe my fantasy engine is the WR450 then. I've really enjoyed working with it. The basic engine design has been around since the late 90's. The latest model is fuel injected with a 12-1 crank trigger and a charging system that can now support EFI. The stock 12.3:1 CR piston has a nice design for overlap cross flow. There is a huge aftermarket with eager sponsors. Engine development can be treated as unrestricted with sufficient plenum volume. Only one intake exhaust tract to develop and package. In my opinion, it's balanced quite well and doesn't vibration the car objectionably. All of this goodness at a weight of only 69 lb.

Seriously, the only change I would make to the WR450f is an upgrade to a larger, tougher starter found on a 450-class ATV.

[AxelRipper]

Originally posted by DannytheRadomski:
Like Mbirt, the Viking 30 is the reason I'm here, and I intend to build something similar to it when I get into FSAE (I know why that's ridiculous, but I want to try it). That engine is even what inspired this thread. Thanks, Rob, for putting those up.

When you actually get to FSAE, the dreams stay alive but the reality pushes them to the back of your mind. I'd love to take an F4i and make a boxer motor out of it, and I may do it once I graduate (and finish my project truck, and win LeMons overall with a Chevette/Olds 350D combo) but the reality is, there is so much more to the sport than trying to make the shiniest toy. The biggest limiting factors: time and money. As we tell people who have overly crazy ideas: "If you can do it without using any of the team's time or money, and still get your other work done, go for it" generally meaning if you've finished everything that there is for you to do (waiting for someone else's design/something else to do) and you want to work on it, fine. But get your work done first (and generally there is also school work to be done, other parts to be made, or beer to be drank).

Still fun to dream though.

[MCoach]

Originally posted by Mbirt:
With my praise for the Viking 30 stated, I would also like to share my love for the WR450f single. A few weeks ago I found the power curve from our 2007 car with the Mahle motor. Its peak power was only 5 hp greater than the peak sprocket output of our current WR450f setup and the single has the smoother, wider racing power band to boot. The Mahle made more torque "off-tune" at low speeds, but not 33% more as the displacement difference might lead one to believe.

So maybe my fantasy engine is the WR450 then. I've really enjoyed working with it. The basic engine design has been around since the late 90's. The latest model is fuel injected with a 12-1 crank trigger and a charging system that can now support EFI. The stock 12.3:1 CR piston has a nice design for overlap cross flow. There is a huge aftermarket with eager sponsors. Engine development can be treated as unrestricted with sufficient plenum volume. Only one intake exhaust tract to develop and package. In my opinion, it's balanced quite well and doesn't vibration the car objectionably. All of this goodness at a weight of only 69 lb.

Seriously, the only change I would make to the WR450f is an upgrade to a larger, tougher starter found on a 450-class ATV.

I have to second this, +1 absolutely so far.


From having to work with the Yamaha compared to other single cylinders I've worked with (particularly a couple two strokes, Suzuki 250, 450, and the Honda CRF 450), the Yamaha WR450 just continues to amaze me. It's stupid simple, doesn't destroy starter clutches by looking at it improperly, while the CRF 450 ate those things for breakfast. It is less finicky than any two stroke that I've worked with as well. Packaging the damn thing? Easy. You have a small box, intake bolts (yes, bolts!) to one side and the exhaust bolts on to the other side. Crank trigger is pretty standard for everything other than a Bosch unit. The sky really is the limit with it. I have a feeling that power-wise, it could keep up with several of the big 600cc motors with some focus really poured into it. :P

For those that don't feel so EFI savvy, it comes in a carbed model before 2012. Oh darn, you blew up the engine? Ebay has an extensive list of parts that are cross compatible between the motorcross bike, the enduro bike (WR), and the ATV. Valve cover? $5. Side covers? $10. Radiators? $10. Everything seems to be readily available.
Just about all parts are able to swap back and forth between the EFI model and the carbed model.


The only thing that comes to mind that could be a potential competitors that are currently on the market would be the Rotax DS450, a few of the KTM engines, and the LTR 450. But, the downside to each of these engines is cost and availability. The LTR is probably the closest to the Yamaha in the best aspects. KTM engines are expensive as well as all parts to boot.

The best part I would say that it has over the 600cc motors is that it is dry sumped to start and development is not usually throttled by the restrictor too far into the diminishing returns range.

So, while I stated my previous answer, if it were to be something that currently exists, I would have to fully agree with MBirt.

[Charles Kaneb]

Would it be worth starting a petition to up the displacement limit to 750cc, without increasing the restrictor size?

Right now, we have a broad variety of motocross-derived singles that are light and simple to package - but are racing engines from Day 1, eat ancillaries, and don't last long even at the stock 40 hp. We have a v-twin that my team's advisor once famously compared to a supermodel on this forum. We have a parallel twin with a built-in CVT whose odd crank arrangement limits its power significantly with the restrictor. We have the industrial v-twins that rookie and second year teams should be using to rack up overall-dynamic-event placings in the low teens, but instead make so little power that only experienced guys are tempted to use them. We have the venerable fours that are boat-anchor heavy, require intense design of the intake and exhaust to fit and make power, make great, broadband power, require a two-gallon fuel tank to go sixteen miles, and last till the cows come home.

If we raise the displacement limit to 750, we add:

1) The Suzuki SV650 engine - a v-twin that weighs 110 lbs, makes seventy horsepower stock, and lasts tens of thousands of street miles.
2) The Yamaha XT660 engine - 90 lb aircooled enduro engine with a more durable starter and stator/reg-rec than a dirtbike engine.
3) The big Honda Foreman and Kawasaki Mule engines - I don't know much about them other than that they have to be abuse-tolerant.
4) The Triumph Daytona 675 engine - probably a lot like the 600-4 motorcycle engines, but with one fewer exhaust pipe to package. Might be good for the British teams.
5) The Suzuki GSX-R-750 engine - like the 600-4s but capable of pulling from 2500 rather than 3000?
6)?

[Markus]

6) The soon-to-be published Yamaha Triple range, which will most likely be 675cc's because of racing

I have to second Charles: with the restrictor the displacement limit is kindoff useless (well it prevents people from being stupid, 1k breathing through 20mm would be useless but somebody would do it) but it could certainly be raised to 700 or 750cc because of the numerous options it would allow...

[Charles Kaneb]

With no displacement limit I'd drop a Suzuki Bandit 1250 engine in there. Air/oil cooled, and I could saw the gearbox off the back and just use a centrifugal clutch as it would probably pull from about 2 rpm.

[MCoach]

Originally posted by Charles Kaneb:
Would it be worth starting a petition to up the displacement limit to 750cc, without increasing the restrictor size?

610cc gas, 910cc diesel?

[Z]

Originally posted by Markus:
I have to second Charles: with the restrictor the displacement limit is kindoff useless (well it prevents people from being stupid...
I also agree that the displacement limit is useless. But, geez, if people want to be stupid, then let them! http://fsae.com/groupee_common/emoticons/icon_smile.gif

I would love to see someone, perhaps an Indian team, turn up at comp with a REAL RACING ENGINE. Maybe a restored Formula 1 Cosworth DFV that they found somewhere. Not only would said DFV weigh a lot more than, say, an old Royal Enfield 500, but breathing through the 20mm restrictor it would also likely have less power! A lot of friction from all those pistons!!! http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Z

[Pete Marsh]

7) Kawasaki EX650/Ninja 650 - in line twin road bike, semi dry sump.

[Racer-X]

If you were to remove the engine limitations it would make sense to go to a larger motor to get all the torque at once down low with the same amount of power. Kind of like a FSAE big block. The cars would be faster and of course that is something the rules comity wants to stop.

I would like to see the displacement limit moved up to 651cc though, there would be so many more options, and so many twins.

[Tom Wettenhall]

Opening up the displacement limit would allow the XR650R or KTM 690 motors as well, for what it's worth. (Maybe less highly stressed than 450 MX motors?)

I could see a lot of teams using the Triumph, MV or upcoming Yam triples as an alternative to an Aprilia or Honda Four. I'd personally love to run a triple - lower RPM power than the Honda without the historically rough starting and plenum volume vs. throttle response issues of the wee singles.

Also, engines with many cylinders break in more entertaining ways. None of this cooking starters, overheating in the driver change or getting stuck in third crap. We do engine failures properly, with red flags and bits of crankcase all over the countryside.

[mech5496]

On a side note, there is the new Kawasaki Ninja 300 twin, which seems interesting for FSAE.

[MCoach]

Well, to note highly stressed engines...

We're making 60hp per cylinder, NA.

[Markus]

Originally posted by Racer-X:
If you were to remove the engine limitations it would make sense to go to a larger motor to get all the torque at once down low with the same amount of power. Kind of like a FSAE big block. The cars would be faster and of course that is something the rules comity wants to stop.

I would like to see the displacement limit moved up to 651cc though, there would be so many more options, and so many twins.
Power and weight are what limits the speed of these cars, using a lot bigger engine would most likely make the car go slower on track. And from safety point of view you can compare it to the current electric cars, they have all that torque down low and they don't seem that much more dangerous in a controlled environment.


Originally posted by Tom Wettenhall:
Also, engines with many cylinders break in more entertaining ways. None of this cooking starters, overheating in the driver change or getting stuck in third crap. We do engine failures properly, with red flags and bits of crankcase all over the countryside.
Very nicely spoken, sir.

[Racer-X]

Markus, the E-cars weigh significantly more than the IC cars. Whereas a liter bike motor isn't much more than a current 4.

Power determines speed, but having a flatter torque curve would help with acceleration and make it easier to achieve those speeds. I don't think a little more torque, and hence acceleration, would make FSAE dangerous but I don't think the rules comity would see it that way.

[Markus]

Says who?
DUT Racing (TU Delft): ~150kg, 4wd, 85kW(?)
Acceleration time: 3,45s (!!!)

No matter what engine you're going to choose, you will never achieve that performance with a restricted IC engine. Engine weight is going to supress the torque benefit way before you get that kindoff acceleration...

[Z]

Originally posted by Markus:
DUT Racing (TU Delft): ~150kg, 4wd, 85kW(?)
Acceleration time: 3,45s (!!!)

No matter what engine you're going to choose, you will never achieve that performance with a restricted IC engine.
Markus,

Says who?

The DUT spec sounds pretty much like the "brown go-kart" I've been pushing for some time now. Except 150kg is a bit overweight (maybe extra FoS?). And 4wd replaced with 60+% rear weight (easier all round, and similar peak acceleration). A 20mm restricted IC engine can flow ~90kW peak, and as pointed out earlier, so can an air-cooled, 20kg, 450cc single. And a simple IVT gives the same or better torque characteristics as an electric motor (actually a moot point, because said motor could spin the wheels almost anywhere).

C'mon guys... Fantasize harder! http://fsae.com/groupee_common/emoticons/icon_mad.gif

Or do some engineering??? http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[Markus]

Says the current results. Feel free to put your money where your mouth is, before that it's just bits in the cyberspace...

Z, batteries weight. Most likely they have extra FoS too but it's insignificant compared to accus. They're design philosophy has been "oversized go-kart" as long as I can remember but it's not actually your fantasy car with the double a-arms, fancy dampers and other complex solutions...

[MCoach]

Z, I think you underestimate the weight of a single cylinder engine. The weight of most singles is actually closer to 30kg (66lbs) and even that is just under normal.

Coming from an area where I actually had to figure out how to manufacture an IVT (for who will not be named), the required tolerances are outside of what I would consider....reachable without professional dedication. Not just support, but dedication. It's simple to call something infinitely variable, a completely different animal to manufacture it. From the group was was requesting it, I also would call it less efficient that a gearbox, based on their data. Besides that point, I would argue that I don't need 90kW, nor an IVT for a "fantasy engine".

Shift times have been recorded down to 50ms from teams that I've talked to personally. as noted earlier, most teams don't need much more than two gears. Most of this is because the cars are traction limited, never mind speed limited overall. A lot of time is spent at lateral accelerations, making longitudinal accelerations not as important. This isn't Can-Am.

IVTs are not very efficient, but uesful for things such as tractors.

450cc motors making 90kW may be possible, however, unreasonable considering the engine you are referencing is an absolute race motor that was held back by it's own design of being air-cooled. The Porsche 917 may have had an undeniable knack for being victorious, they were also notable devils. During it's natural aspiration days it trailed behind it's water cooled counter parts by about 100hp. To put some numbers in context:

Yamaha 2012 Enduro WR450 stock -- 39 hp, 450cc, 85hp/L
Suzuki 2012 MotoX RM-Z450 stock -- 55hp, 450cc, 122hp/L
FSAE 2012 NA Endurance Spec WR540 -- 60hp, 450cc, 133hp/L
Porsche 1969 NA, Aircooled 917 H12 -- 542hp, 4.5L, 120hp/L
Ferrari 1971 NA,Watercooled 712 -- 680hp, 6.8L, 100hp/L
Porsche 1971 NA, Aircooled 917 H12 -- 630hp, 5.0L, 126hp/L
Porsche 1972 Turbo, Aircooled 917 H12 -- 1100hp, 5.4L, 203hp/L
FSAE Estimated Turbo Spec WR540 -- 93hp, 450cc, 207hp/L
NASCAR 2012 Fuel Injectected Cup Motor -- 850hp, 5.7L, 149hp/L
Formula 1 2007 NA Ford Cosworth DFV V8 -- 465HP, 3.0L, 155hp/L
Formula 1 2007 NA V10 -- 1000HP, 3.0L, 333hp/L
Formula 1 2012 NA V8 -- 800HP, 2.4L, 333hp/L
Formula 1 1986 NA V6 -- 1500HP, 1.5L, 1000hp/L

Unfortunately the hp/L stats are much easier to dig up than BMEP, but I'll see how quickly I can find those as well (only need the RPM from here for those engines). Not until Porsche turbocharged their engine, did they finally get a 'respectable' number. It is worthy to note that the hp value used for the FSAE turbo spec motor are taken from a similar motor we are testing and I would consider realistic. It is also worthy to note that the engines listed that surpass that spec have a much larger team behind their engine development (more than 1 http://fsae.com/groupee_common/emoticons/icon_razz.gif) and better luck with fuel selection.


To just hit things with a final note of electric motors, they are damned effective. DUT is far from a 'brown go-kart'. It's closer to a professional cost-no-option race product than most vehicles in this competition. I've left enough words for headaches and rambling, so I'll skip to this last bit. The gains in electric motors is CONTROL. Just like anything, you can have a 20ft tall ogre swinging his arms into battle, but without control, he's still useless. It works 60% of the time, everytime! 4WD I would opt to keep, because peak accelerations may be the same, but with proper control and usage, as Delft has shown that they are capable of, average accelerations are what make the lap times. Still, the issue of something electrical breaking makes it a bigger gamble.


Am I fantasizing too hard? http://fsae.com/groupee_common/emoticons/icon_rolleyes.gif

Come on Z, I've now suggested something simpler than you. Carbed single, two speed trans, 'let's go racing' type of engine. That' brown go-kart quality, and you want active diffs and all the nonsense.

Or maybe it's just that I've spent most of my life trying to make things work in real life to get an idea of what is right in fantasy land. Coworth DFV isn't all that far off from where we are already, we just keep our power down to make sure we don't blow it. Engines blowing up are frowned upon in a uni setting. http://fsae.com/groupee_common/emoticons/icon_razz.gif

[mech5496]

Besides control, 4WD in EVs has an equal, if not greater advantage; it allows you to recover much more energy while braking than you would normally do with an RWD car. This in turn allows for a smaller capacity and thus ligjter battery pack (and increased efficiency aswell). I bet that DUT battery is about 20 to 25 kg lighter than ours, and thats due to 4WD. Ireally doubt if they use (or need) mechanical brakes during endurance. Another interesting technology for use on FSAE EVs is supercapacitors. A really small battery could be added to aid capacity, while supercapacitors could be used to store regenerated energy. Right now, regenerated energy is somewhat limited from the max charging current of most available battery cells out there. This could make EVs even lighter, but even more complicated... And yes, DUT is far more complicated than a brown go kart!

[Racer-X]

Originally posted by Markus:
Says who?
DUT Racing (TU Delft): ~150kg, 4wd, 85kW(?)
Acceleration time: 3,45s (!!!)

No matter what engine you're going to choose, you will never achieve that performance with a restricted IC engine. Engine weight is going to supress the torque benefit way before you get that kindoff acceleration...

For teams already going the 4 cyl big wings route the 5-10 pounds that a liter bike engine weighs over a 600 would be worth it for the increase in toque. 30 foot pounds for 5 pounds is a trade I would take all day. A liter bike is capable of 80 foot pounds and 120 horsepower with the restrictor. What the electric cars have going for them is the four wheel drive, can't beat traction like that.

[Markus]

What electric cars have going for them is the instant max torque from > 0rpm and precise control.

A liter bike engine faces the same challenges or even more than a 600cc in breathing through that restrictor. Simplified that means it's just as easy to push out 120hp of a 600cc than a 1000cc.

Quick googling estimate on R1 engine came out ~175 pounds. R6 engine weights some 130 pounds.

Now the 1000cc engine can be lightened more but that 45 pounds is a sheit-load of extra weight for the same power and small torque benefit...

[RenM]

Originally posted by Racer-X:


For teams already going the 4 cyl big wings route the 5-10 pounds that a liter bike engine weighs over a 600 would be worth it for the increase in toque. 30 foot pounds for 5 pounds is a trade I would take all day. A liter bike is capable of 80 foot pounds and 120 horsepower with the restrictor. What the electric cars have going for them is the four wheel drive, can't beat traction like that.

going for the 1l engine with a 20 mm restrictor is just useless in any way.
The restrictor will start to choke at around 6500 rpm so revving higher will be useless and your torque advantage is compromised by a longer gear ratio.
You will have higher Heat transfer, mechanical and throttle losses.
There is no way any engine will reach 120hp with a 20 mm restrictor...

[MCoach]

For teams already going the 4 cyl big wings route the 5-10 pounds that a liter bike engine weighs over a 600 would be worth it for the increase in toque. 30 foot pounds for 5 pounds is a trade I would take all day. A liter bike is capable of 80 foot pounds and 120 horsepower with the restrictor.


Some of the 450cc engines are capable of hitting choke flow with some coaxing (read turbo), 600cc engines are all capable of hitting choke flow and 1000cc engines are 'betta for all dat damn low down torque!'

So, anyone see the success of the Dodge Viper in Petite Le Mans this year? Oh wait....

[Racer-X]

The CBR1000rr and the 600rr were the motors that I was comparing, with those I figured there was a 5-10 pound difference like Coleasterling said. If we say add another 10 pounds for a bigger fuel tank, more fuel, and a larger radiator it still comes in at 15-20 pounds of added weight for a whole lot of torque.

A 78% increase in torque is not a small benefit, I'd say that's huge.

Engines have reached 110 horsepower on chassis dynos before and the theoretical maximum power with the restrictor is around 120. It is reasonable to think that if someone wanted to they could reach those numbers with a liter bike.

As for the torque not being useful due to gearing I disagree. Lets say with the 600 you reach choke flow at 11,000 RPM and you make 45 foot pounds of torque for the liter bike you choke at 7,000 and make 80 foot pounds. This means your shift points will be 12,000 and 8,000 respectively. The liter bike would have to run a ratio 78% higher (numerically lower)than the 600 before they break even. I calculate that the total reduction would need to be 50% taller to maintain the same top speed in first gear based on the shift points. I am assuming the liter bike has the same ratios as the 600 just for comparisons sake. This results in a net rear wheel torque gain of 28% and weight gain of 5% by my math.


The liter bike is an extreme example, most of that torque gain would be wasted as it is converted into smoke and black stripes on the pavement. If I were allowed to exceed 600cc for an FSAE engine I would go with a 650 twin. Way better for packaging and would deliver all the torque we could ever want.

[RenM]

I wouldn't trust some random chassis dyno too much.
People can easily claim some performance figure, but without knowing the circumstances i don't believe them.

A higher displacement engine will always have higher losses as i stated before. There is no way you will reach a higher maximum power.

Your torque comparison is flawed. In the end its all about the power (not maximum power). As long as you have the same Volumetric Efficiency for both engines, which can be assumed, you won't be able to see a torque advantadge. With the same fuel, same efficiency and same VE, the power output is only determined by your displacement * rpm and such is your maximum useful rpm, limited by the restrictor. You will always end up with the same power at your wheel. Your engine torque changes but so does your gear ratio.
I guess you are comparing stock torque figures, but the cbr1000rr has its highest torque at 8500 rpm, way beyond the choking limit of the restrictor. Plus: You can influence your VE with your intake/exhaust headers and cam timings and give a 600cc 4 cylinder engine a very good low and mid range torque.

[Mbirt]

Originally posted by Racer-X:
Engines have reached 110 horsepower on chassis dynos before and the theoretical maximum power with the restrictor is around 120. I wish we would have attended MIS 2011 to get a dyno sheet showing our n/a 450 just put 75 hp to the rollers...

[mech5496]

75hp on a na 450?! You got my attention now...

[Racer-X]

It's not only about power, its about torque at the wheels. If you have two engines making the same power and they are geared appropriately they will be making the same max torque at the wheels and the same max "thrust" or force to accelerate the car at that point in the rpm range. If you have a CVT or only plan on driving at one point in the RPM range that's great. Now if one makes more net torque after gearing at a lower point in the RPM range it will be easier to drive and accelerate quicker. This engine will deliver more torque to the rear wheels through the range of each gear though the max rear wheel torque will be the same. Get what I'm saying? If that's hard to follow I apologize.

If you think of forced induction as increased displacement then you can see how going to a larger engine would have a similar effect. Though the larger engine is not as efficient as a turbo when it has to breathe through a 20mm restrictor. That said you would see a similar results with the torque building early followed by the restrictor choking and power leveling off.

I imagine the fastest FSAE engine with a conventional transmission as being something that is capable of choking the restrictor almost off idle allowing for realistic gearing*. The thing would have gobs torque and the power "curve" would be a flat line. In a car like that you would be able to deliver the maximum amount of torque possible given our situation. Increasing displacement or going forced induction are the only ways to get close to doing that.

My point is that it would be possible to increase the area under the torque curve at the rear wheels with a larger engine. Just like people do with the turbos except the larger engine wouldn't have the lag. The liter bike is an extreme example of that.

*Choking the restrictor for some of the usable range is doable now with a 600cc but it either requires an impossibly large rear sprocket or puts the car out of the power in slow corners in the lowest gear.

[Mbirt]

Originally posted by mech5496:
75hp on a na 450?! You got my attention now... Sorry to disappoint, Harry, but I'm just joking about the ridiculous numbers that the MIS 2011 dyno spat out, echoing RenM's concern. Google "MIS 2011 dyno" and you'll see.

As for the upper limit of true 450cc singles, I've seen sheets showing mid-60's to the rear tire of some flat track racing dirt bikes and some sand drag ATV guys claim to be into the 60's at the rear tires as well. Our Mustang ATV dyno says we're putting 54 hp to the rollers.

[RenM]

You are right when you say that a high torque at low / mid range rpm is important for a good driveability and a fast acceleration out of small corners, but thats because it means that you will have a lot of power in the lower rpm range.
It does not matter how much absolute torque you have. If the power output is the same (again i am not talking about max power, but power over the rpm range), there will be no difference in torque at the wheel after appropriate gearing.

A completely flat power line is not desirable. Simplified your tire can always transfer a fixed amount of torque. The transferable power increases with the revolutions. So for your lowest gear you will want a flat torque curve and a gearing ratio, so that your wheel torque is slightly higher then your transferable torque. In the next gears you will want a flat power curve, so you have the maximum power the engine can deliver, all the time.

As i said before you can of course also increase the Volumetric Efficiency (which is what a turbo does), but still, you wont be able to deliver more power/torque. It only helps to reduce the rev range to increase your engines efficiency.
You can not directly transfer some street tuning stuff to a restricted engine like in FSAE.

I don't really get your last sentence about the 600cc. A 600cc 4 cylinder engine can easily choke the restrictor, have more torque then needed (you can get almost 70 Nm) and have a large usable rev range from 5000 to 11.000 rpm.

[Tom Wettenhall]

I'm not sure what the fuss is about with choking the restrictor. Surely the aim is to make enough power to be traction limited for most (define most according to team goals) of the lap. That shouldn't require that much power - we never had any problem achieving it with a basically stock-internals 600.

I'd say the main challenge would be to deliver that power in such a way that the driver can easily keep the tyres on the edge of grip, which means a fairly linear torque curve and fast throttle response at all RPM. Things like turbos and large, slow revving engines detract from both of those to my mind. An engine with multiple small cylinders, like a 600RR, fulfills the requirement for adequate power and sharp throttle response, leaving only the powerband problem. I'd argue that a 1000 is likely to have too much torque - one inch of throttle travel corresponds to 50Nm rather than 20, so a normal human would have to drive 10Nm from the limit rather than 5 as a tiny movement is more likely to break the tyres loose. Functionally, the 600 is delivering more torque.

[Racer-X]

I think everyone is taking the liter bike engine too literally. I agree it is overkill. My point is that the greater displacement means you will reach the maximum amount of power and therefore torque allowed by the restrictor sooner. To what extent one needs that power is up for debate.

RenM power doesn't do work, that's torque. Power is the measure of how fast you can do that work and is one way of looking at how an powertrain package will perform. I like to look at rear wheel torque because it is directly translatable to force, and F=MA. With a conventional transmission you are stuck with one ratio for a part of the rev range and you want to maximize torque in that range.

I agree with you that max power through an RPM range can be geared to do the maximum amount work given the restrictor. That is what I'm saying about the larger engine. It will choke the restrictor and make max power at a realistic RPM.

That brings me to what I was saying about the 600cc engine. A 600cc engine can choke the restrictor for about 2,000 rpm right at the end of its rev range. Without expensive internal modification you can't get one to spin fast enough to choke the restrictor through a whole gear. On top of that you would need a huge final reduction to make that useable, like 4-4.5:1, and that doesn't fit.

I don't get what you are saying about the power and torque curves though. In a perfect world you would have a huge flat torque curve and the driver can use a higher gear or the right pedal less if they want less torque at a speed. We can't have that flat torque curve because of the restrictor, so the next best thing is to have a flat power curve that maximizes torque. Are you suggesting running different maps for different gears to change power output?

[Will M]

The conversation has drifted toward the limitations of real engines.
More make believe please.

For a fantasy engine I would want an OEM set up with:
A 600cc parallel twin
Air cooling
A dry sump
An intake already with a 20mm restrictor
Two speed transaxle

And the whole thing packaged to be as narrow as possible.

For the first year you could use the OEM intake and tune.
Later you could redo the intake to better fit your needs.
Or if you screw up your design you could go back to the OEM intake and tune.

I feel that for a first year team this would be perfect.
Ordered directly from the OEM and it is ready for plug and play.

-William

[Tom Wettenhall]

Okay:

600cc SOHC 12V triple
Transverse crankshaft
Liquid cooling
Low height wet sump with accumulator
Integral two speed transaxle (spool diff)
Cylinders angled rearward and reversed (intake goes straight up from the front of the motor)
~11k redline

Why narrow? I'd would have thought short, for mass centralisation.

Liquid cooling means you can bury it under the seat, three cylinders gives you the balance of throttle response and linear power I was talking about, and SOHC is fine as you won't make much more power revving above ~10k RPM anyway because restrictor. Wet sump for simplicity and packaging. Two speeds is enough, as has been said quite a lot on this thread.

It's heavier and not as fuel efficient as a 'perfect' engine, but it should be more reliable and driveable, which means heapo testing, which means you go faster at the end of the day anyway.

[RenM]

Originally posted by Racer-X:
RenM power doesn't do work, that's torque. Power is the measure of how fast you can do that work and is one way of looking at how an powertrain package will perform. I like to look at rear wheel torque because it is directly translatable to force, and F=MA. With a conventional transmission you are stuck with one ratio for a part of the rev range and you want to maximize torque in that range.

I agree with you that max power through an RPM range can be geared to do the maximum amount work given the restrictor. That is what I'm saying about the larger engine. It will choke the restrictor and make max power at a realistic RPM.

That brings me to what I was saying about the 600cc engine. A 600cc engine can choke the restrictor for about 2,000 rpm right at the end of its rev range. Without expensive internal modification you can't get one to spin fast enough to choke the restrictor through a whole gear. On top of that you would need a huge final reduction to make that useable, like 4-4.5:1, and that doesn't fit.

I don't get what you are saying about the power and torque curves though. In a perfect world you would have a huge flat torque curve and the driver can use a higher gear or the right pedal less if they want less torque at a speed. We can't have that flat torque curve because of the restrictor, so the next best thing is to have a flat power curve that maximizes torque. Are you suggesting running different maps for different gears to change power output?

oh come on. Talking about power simply takes out the gear ratios out of the debate. Torque can be altered by gear ratios, power cant, so its much easier and much more feasible to talk about power and not about engine torque. If you insist so much on using torque, the torque at the wheel multiplied with its speed gives you the power at the wheel, which is the power the engine can deliver.

you can have a 600c engine choke at 9.000 rpm and have a nice power plateau for 2,500 rpm without expensive modifications. All it takes is some thorough engineering and knowledge about engine tuning.

We can have a flat torque curve until the restrictor chokes. In the lowest gear used for driving you can start at a low rev and at the end of that gear you reach the higher revs and your power plateau and stay there with your next gears.

[Will M]

@Tom Wettenhall

I would want it to be narrowed for latter aero implementation.
I always struggled to get a decent diffuser to fit around our F4i.
Not that it can't be don't; a narrow engine would just make it easier.

-William

[Racer-X]

Power only takes the ratios out if you run some sort of cvt or other non conventional transmission. Power does not do work, I think you will agree that a car geared to do 200mph with 2 gears is going to have poor acceleration times even if it has the same power as an appropriately geared car.

Even at that my argument for a larger motor is that you can reach choke flow and peak power at a reasonable RPM. On our car we run the f4i and we are maxed out on rear sprocket size and we still have issues with falling out of the power in slow corners. Without increasing the half shaft angles or lengthening the wheelbase we can't go any lower. I calculate that in a 25mph corner we are at about 5,500 rpm in first.

If you are choking at 9,000 rpm you have poor restrictor design or you are using forced induction. The flow bench data off our intake showed that we choke just before 11,000 RPM, I'd believe 10,000 if you guys did something very right, but not 9k. Even then you would need to be able to choke the restrictor from 7,000 to 12,000 rpm if you want to maximize rear wheel torque. I say the only ways to do that are with a larger engine or forced induction.

Maybe you guys have a better time packaging your engine or something and can run more friendly ratios. With our car we are always looking for ways to maximize low end torque. Its a great safety net for the drivers and helps with acceleration from a dead stop.

[Homemade WRX]

Originally posted by murpia:
V's package well (longitudinally).

What's a boxer if not a 180deg V? Doesn't that result in too high a crank height, cramped exhaust routing and no space for accessories along the sump?

A turbo V works quite nicely: Put the plenum in the V and the turbo behind the engine above the gear cluster. Oil pumps down one side of the sump, alternator down the other. Shallow out the V as far as you can without cramping the exhaust exits (120deg?).

Regards, Ian
Well in the same light, what is an inline engine other than a zero degree V?

As for a V, it might be beneficial to keep the heat in the exhaust and reduced weight of your exhaust manifold by keep it in the valley and keep your light manifolds outside of the V.

I imagine any arrangement would really depend on layout/packaging.


Originally posted by murpia:
OK, yes, I can see this setup packaging nicely, but I have a hard time imagining (fantasising?) the camshaft drive(s).

http://www.subaru.com/content/engineering/boxer_displacement_25l/Turbo%20Boxer%20Engine_282x160.png

Assuming that image link works, it highlights my CoG & crank height concerns ref a flat engine vs. a V...

Regards, Ian

Well considering that is with the factories deep oil sump, of course it seems tall. Now layer the image of a 209mm deck height V8 with an equally deep oil pan on top of that.

Keep it apples to apples...

[RenM]

we are running in circles. As i said i am not talking about the maximum power, but about the power over the rev range.
Once again: you will not gain a torque advantage at the rear wheels with a higher displacement, because you will have to change your gear ratio, because of the restrictor choking. You can of course decide to run in high rev ranges, but you will get a lot of throttle and mechanical losses which lead to a bad efficiency.

We neither have a poor restrictor design nor use forced induction. We have a very good designed gas exchanged cycle, with properly tuned intake and exhaust runner lengths and adjusted cam timings and profiles. You do realize that you can reduce the rpm at which you get the maximum possible flow by increasing your VE?

It is not a good target to run your engine choked all the time. As i said, you need a flat torque curve at the lower rpms where you are traction limited and you need a power plateau of not more then 2,500 rpm for your higher gears.

[MCoach]

If you want an example of an engine that primarily operates at constant power due to choke flow, rally cars basically hit the restrictor wall at ~3400 and continue all the way out to ~6500.

http://i267.photobucket.com/albums/ii284/OkdF/Suzuki_SL_torque_power.jpg

[Z]

Some comments regarding the above (many) posts:

ENGINES.
========
Firstly, chasing maximum engine power only wins the pissing contest. Yes, it's fun, but FSAE is won with lateral Gs (ie. AERO! So who'll be the first to consistent 3G-lateral!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif ). Nevertheless, ...

Maximum Power.
-----------------------
From observations of several decades worth of restricted engines I have concluded that about 30kW (40hp) per square centimeter of restrictor area is the most anyone can get. Figures for current Le Mans cars are reasonably reliable and are about 35hp/sq.cm for these 5,000+km engines. Doing the theoretical calculation (air mass flow rate, fuel energy content, thermal efficiency, etc.) suggests a high energy fuel and very high efficiency engine (high CR, low friction) is needed to reach 30kW/sq.cm (can post calcs if anyone wants).

Porsche 917.
------------------
For anyone who missed it earlier, here is a link to the 917's flat-12 specs. (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=54520089051#54520089051) Also my long-winded ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=86720419051#86720419051) about the car and engine. (BTW, MCoach, why do you think this engine "was held back by it's own design of being air-cooled" given that it was so dominant in European endurance racing that it was banned, and then in North America it was "the car that killed Can-Am"??? http://fsae.com/groupee_common/emoticons/icon_smile.gif)

Notable for this engine is that with its turbo it only needed a 2-valve head for 120hp/cylinder. The 2-valve head also made air cooling easier. In the late 1970s Porsche built a similar engine for use in a broader range of racing series, and this had a 4-valve head which needed water cooling (barrel was still air-cooled).

Other Engine Stuff.
-------------------------
In FSAE maximum power is only used for a handful of seconds per minute. Top fuel dragsters put out ~8,000hp, yet they have no cooling systems whatsoever! That is because they are only at full power for 5 seconds, then switch off (err, and the methanol helps...). So cooling a max power FSAE engine should not be a problem. If it is, then you are doing something wrong.

On the other hand, if you are aiming for the maximum 90kW from a turboed single, then I reckon you DO NEED a crank with ~50mm diameter mains and big ends. IMO the skinny little cranks on standard MX bike engines are NOT big enough. I reckon off-the-shelf aftermarket small-block-V8 parts like conrods, slipper bearings, valves, etc., would be the best way to go. 2-valve NA sbV8s make 100hp/cylinder, so just add that turbo...

The various Jawa speedway engines weigh about 28kg (admittedly without starter, clutch, etc++.). But I figure the crank alone is at least 10kg, maybe 15kg (this probably for smoother throttle response, because less likely to suddenly over rev when breaking traction). So, all things considered, I reckon a realistic FSAE total engine/transmission weight of about 30kg is feasible.

In fact, my weight budget for the whole car would be;

1. engine = 30kgs,
2. chassis = 30kgs,
3. front-axle (= 2 x wheels, brakes, suspension, etc.) = 30kgs,
4. rear-axle (as above) = 30kgs,
5. everything-else (= IA, seat, electrics, aero, +++) = 30kgs
So Total Car Weight = 150kg.

Add a driver = 50kgs (or 30kgs if they have "the operation", ie., "You don't NEED legs!!! We'll just sew your feet to your butt!" http://fsae.com/groupee_common/emoticons/icon_smile.gif) gives total mass = less than 200kgs.
~~~~~o0o~~~~~

TORQUE & TRANSMISSION.
=======================
Regarding the earlier posts between Racer-X and RenM. As they both said (I think?), it is the engine's maximum power that determines the car's maximum acceleration (given enough tyre grip). Having a torquey, large capacity engine is only a benefit if the car is disadvantaged by a gearbox with a limited number of ratios. A very peaky, but higher power engine will win if it can always operate at its peak power revs, perhaps by driving through a CVT or IVT.

But in FSAE it is even simpler. As I explained in this previous ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=56320469051#56320469051), if you have a maximum power engine and a lightweight car, then you should be able to spin your wheels all the way down the longest straight (ie. traction limited everywhere!). So,

"The important point is, if you can nearly spin the wheels [everywhere] in top gear, then there is really no need for a lower gear. All that a lower gear does is allow you to, perhaps inadvertently, "smoke the tyres", and then possibly spin the whole car just when you want to be going straight."

As RenM said, you DO NOT WANT a flat power curve everywhere (ie. with restrictor choked). What you DO WANT is a flat torque curve everywhere, preferably with the torque just enough to give about 10-30% tyre slip-ratio (= max thrust). This flat torque curve gives a linearly rising power curve, with low power at low speeds, because that is all you need.

The only problem here is that IC engines don't run well at VERY low speeds. So it is only at low car speeds that the engine should be allowed to run a bit faster, albeit still with the same torque at the rear wheels. This can be done by slipping a clutch, or by using an IVT together with reduced throttle opening.

I repeat, the aim is NOT to increase rear wheel torque, but simply to let the engine run more smoothly by running at higher revs. Earlier I suggested the use of an IVT so the flywheel can be used as an energy store, but that is a different matter. Here an IVT would be used as a smoother, more efficient clutch.

It is worth noting almost every FSAE car has some sort of conventional friction plate clutch. These are quite large and heavy. Worse yet, they have the unfortunate habit of turning into a lump of coal at the most innappropriate times. Swapping the clutch for a similar sized (or smaller?) IVT would have significant advantages. There are also other "hydraulic" options that would help with low speed running (eg. auto-box type converter with "lock-up", and others http://fsae.com/groupee_common/emoticons/icon_wink.gif ).
~~~~~o0o~~~~~

Ooops! Another too long-winded ramble... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[rjwoods77]

Hello all,

So I took some time to throw some CAD down on the idea I mentioned in a couple posts on this thread before...

http://fsae.com/eve/forums/a/t...20697151#48320697151 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/67020597151?r=48320697151#48320697151)

http://fsae.com/eve/forums/a/t...20897151#94320897151 (http://fsae.com/eve/forums/a/tpc/f/125607348/m/67020597151?r=94320897151#94320897151)


...since I was interested to see if my hunch looked like I had envisioned and was pretty shocked at how well the drivetrain "box" fit together. I took the Briggs and Stratton 90 degree V engine model (590cc)that UB used...

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

...and flattened the engine out to 180 and then positioned the cylinders coaxial to each other (more on that later). In this pushrod configuration and with the crankshaft in a vertical orientation it left the ports in a vertical orientation as well. I added the fantastic idea that Doug Millikan pointed out utilizing the 2:1 cam gear reduction to drive the primary clutch (the reason you will see the primary clutch not on the crank snout but offset from it) so that all would be needed is anywhere from a 3:1 to 3.5:1 final drive reduction(Gaged CVT is 4:1 low/ 1:1 high and a Burgman CVT is 1.9:1 low/ 0.45 high but only going to 1:1 on the Burgman) which is covered in the range of ring and pinons in BWM motorcycle 146mm OD ring and pinion from their older R75 and R100 bikes that have similar power and weight as FSAE....

http://bmwmotorcycletech.info/ringgears.htm

...which after machining the center section out would fit with room over a Drexler (represented by the dual circle sketch in the side view). It is also worth noting using the stock BMW pinion and bearings since it would be easy to machine splines on the secondary clutch to mate to the pinion...

http://bavariancycleworks.com/...2011/12/P1020525.jpg (http://bavariancycleworks.com/wp-content/uploads/2011/12/P1020525.jpg)
http://bavariancycleworks.com/...2011/12/P1020524.jpg (http://bavariancycleworks.com/wp-content/uploads/2011/12/P1020524.jpg)

I also threw a GT12 in the area under the drivers back where commonly is a bunch of room and also to get the weight away from the rear axle centerline. I modeled (but not shown in these pictures) the shortest possible exhaust entry with available elbows and straights (Burns Stainless)into the turbo in the vertical port configuration and the primary length from port to flange was 13 to 14 inches depending a couple factors. I also tried with the same turbo positioning and longitudinal ports and that value went down to 11 to 12 inches depending on a couple of factors. It mounts low and gets the exhaust coming out of the car in the mid position instead of trying to cram it at the end of the car. Also you can put your restrictor and throttle down there. If there were no turbo and you were going into a 2 into 1 collector then the values would be the same which is ballpark where you want to be for an engine of that size and rpm and of course longer if need be. That effectively cancels the manifold concerns about a flat twin being disadvantages when actually in my experience with the 90 degree V is a pain because the ports are too close to work around what is needed in FSAE. I did come rough sketching with the intake lengths too and it pans out to be an advantage again to have the cylinders spread at 180 degrees. If turboed then you could mount the intercooler right to the compressor housing and then duct into the two intake ports. Comparable if not shorter routing than a parallel or V engine.

The pictures with and without a WR450 layed over using crank and cylinder centerlines as reference match point...

http://imageshack.us/g/1/9977715/ (Thank you Mbirt for hosting)


...and you can see how tight this thing would fit. The transverse area that the twin takes up I rarely if ever see anyone have anything in that space. The round object attached to the case it the huge stock Briggs and Stratton starter.

On a final note, having run the packaging exercise and knowing the general requirements of rpm, ancillaries, team needs, etc I would go with a OHC design that would place the ports longitudinally instead of vertically. I would also water cool it having dealt with air cooled mysteries, tuning issues, duty cycle, etc in our engine program. I would still have the 2:1 reduction in the case for the primary clutch in such a case as well. It is worth noting with a vertical crankshaft there are multiple drivetrains that can be tried out. My CVT and 90 degree diff, Z's kart clutch and 90 degree diff, etc. I would be interested to see how this fits into peoples envelopes.

[murpia]

Originally posted by Homemade WRX:
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by murpia:
OK, yes, I can see this setup packaging nicely, but I have a hard time imagining (fantasising?) the camshaft drive(s).

http://www.subaru.com/content/engineering/boxer_displacement_25l/Turbo%20Boxer%20Engine_282x160.png

Assuming that image link works, it highlights my CoG & crank height concerns ref a flat engine vs. a V...

Regards, Ian

Well considering that is with the factories deep oil sump, of course it seems tall. Now layer the image of a 209mm deck height V8 with an equally deep oil pan on top of that.

Keep it apples to apples... </div></BLOCKQUOTE>
My point was with reference to the exhausts. You can only drop the crank height of a 180deg engine so far before the lowest thing is the exhausts. With a V you can go lower.

Regards, Ian

[mech5496]

Just remembered that the new Honda CB500 uses a 471cc twin....

[MCoach]

Maximum Power.
-----------------------
From observations of several decades worth of restricted engines I have concluded that about 30kW (40hp) per square centimeter of restrictor area is the most anyone can get. Figures for current Le Mans cars are reasonably reliable and are about 35hp/sq.cm for these 5,000+km engines. Doing the theoretical calculation (air mass flow rate, fuel energy content, thermal efficiency, etc.) suggests a high energy fuel and very high efficiency engine (high CR, low friction) is needed to reach 30kW/sq.cm (can post calcs if anyone wants).

The only problem I have with comparing FSAE used engines to Le Mans engines is the amount of fancy extras (DLC on everything) that go into those engines and the fact that minimum weight plays into this. For this reason, teams like Audi chose to go with a V-10 rather than the lighter, more compact V-6 that could have been capable of producing the same power. They wouldn't have gained anything because the weight would have had to been made up elsewhere. Le Mans engines also run through there own restrictor. Which makes them more reasonable to look at then F1 or old Can-Am cars.



Porsche 917.
------------------
For anyone who missed it earlier, here is a link to the 917's flat-12 specs. (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=54520089051#54520089051) Also my long-winded ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=86720419051#86720419051) about the car and engine. (BTW, MCoach, why do you think this engine "was held back by it's own design of being air-cooled" given that it was so dominant in European endurance racing that it was banned, and then in North America it was "the car that killed Can-Am"??? http://fsae.com/groupee_common/emoticons/icon_smile.gif)

Notable for this engine is that with its turbo it only needed a 2-valve head for 120hp/cylinder. The 2-valve head also made air cooling easier. In the late 1970s Porsche built a similar engine for use in a broader range of racing series, and this had a 4-valve head which needed water cooling (barrel was still air-cooled).


The Porsche 917 was known as the car that killed Can-Am, yes this is true. But it's not like it was a very reliable or dominant car for very long. The car may have not been watercooled, but it was certainly oil and fuel cooled. At the time there were not power restrictions as is typical to see today, along with this was no fuel consumption minimum. Looking back through history, you can also note quite a few other cars that were able to break away from the rest of the group by having a breakthrough in technology or application. We could look at the Hudson Hornet for all I care. Using this reasoning why don't we run flat 16 engines or straight 8s? http://fsae.com/groupee_common/emoticons/icon_biggrin.gif
The undoing of the Porsche was relying on fuel cooling and lots of downforce when the fuel consumption rules caught up with them.



Other Engine Stuff.
-------------------------
In FSAE maximum power is only used for a handful of seconds per minute. Top fuel dragsters put out ~8,000hp, yet they have no cooling systems whatsoever! That is because they are only at full power for 5 seconds, then switch off (err, and the methanol helps...). So cooling a max power FSAE engine should not be a problem. If it is, then you are doing something wrong.

On the other hand, if you are aiming for the maximum 90kW from a turboed single, then I reckon you DO NEED a crank with ~50mm diameter mains and big ends. IMO the skinny little cranks on standard MX bike engines are NOT big enough. I reckon off-the-shelf aftermarket small-block-V8 parts like conrods, slipper bearings, valves, etc., would be the best way to go. 2-valve NA sbV8s make 100hp/cylinder, so just add that turbo...

The various Jawa speedway engines weigh about 28kg (admittedly without starter, clutch, etc++.). But I figure the crank alone is at least 10kg, maybe 15kg (this probably for smoother throttle response, because less likely to suddenly over rev when breaking traction). So, all things considered, I reckon a realistic FSAE total engine/transmission weight of about 30kg is feasible.


Cooling is required for how long these things stand in line waiting to take there runs around the courses. 90kW is a nice lofty goal for a nice air cooled adiabatic engine. Unfortunately, another thing that comes in when comparing to something like the Porsche 917. The Porsche never had to deal with the issue of choke flow through an orfice. The 600 engines that teams typically use make about 90kW stock, but the restrictor plays havoc with trying to flow air properly, and lots of time and resources are spent trying to just scramble back to the starting point.



In fact, my weight budget for the whole car would be;

1. engine = 30kgs,
2. chassis = 30kgs,
3. front-axle (= 2 x wheels, brakes, suspension, etc.) = 30kgs,
4. rear-axle (as above) = 30kgs,
5. everything-else (= IA, seat, electrics, aero, +++) = 30kgs
So Total Car Weight = 150kg.

Add a driver = 50kgs (or 30kgs if they have "the operation", ie., "You don't NEED legs!!! We'll just sew your feet to your butt!" http://fsae.com/groupee_common/emoticons/icon_smile.gif) gives total mass = less than 200kgs.


This as a target I can somewhat agree with... it is inline with my own targets and Although, the driver is a little small...even for my tastes.
Maybe more a kg around here and there for some +/- interactions, but I can agree.



~~~~~o0o~~~~~

TORQUE & TRANSMISSION.
=======================
Regarding the earlier posts between Racer-X and RenM. As they both said (I think?), it is the engine's maximum power that determines the car's maximum acceleration (given enough tyre grip). Having a torquey, large capacity engine is only a benefit if the car is disadvantaged by a gearbox with a limited number of ratios. A very peaky, but higher power engine will win if it can always operate at its peak power revs, perhaps by driving through a CVT or IVT.

But in FSAE it is even simpler. As I explained in this previous ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=56320469051#56320469051), if you have a maximum power engine and a lightweight car, then you should be able to spin your wheels all the way down the longest straight (ie. traction limited everywhere!). So,

"The important point is, if you can nearly spin the wheels [everywhere] in top gear, then there is really no need for a lower gear. All that a lower gear does is allow you to, perhaps inadvertently, "smoke the tyres", and then possibly spin the whole car just when you want to be going straight."

As RenM said, you DO NOT WANT a flat power curve everywhere (ie. with restrictor choked). What you DO WANT is a flat torque curve everywhere, preferably with the torque just enough to give about 10-30% tyre slip-ratio (= max thrust). This flat torque curve gives a linearly rising power curve, with low power at low speeds, because that is all you need.

The only problem here is that IC engines don't run well at VERY low speeds. So it is only at low car speeds that the engine should be allowed to run a bit faster, albeit still with the same torque at the rear wheels. This can be done by slipping a clutch, or by using an IVT together with reduced throttle opening.

I repeat, the aim is NOT to increase rear wheel torque, but simply to let the engine run more smoothly by running at higher revs. Earlier I suggested the use of an IVT so the flywheel can be used as an energy store, but that is a different matter. Here an IVT would be used as a smoother, more efficient clutch.

It is worth noting almost every FSAE car has some sort of conventional friction plate clutch. These are quite large and heavy. Worse yet, they have the unfortunate habit of turning into a lump of coal at the most innappropriate times. Swapping the clutch for a similar sized (or smaller?) IVT would have significant advantages. There are also other "hydraulic" options that would help with low speed running (eg. auto-box type converter with "lock-up", and others http://fsae.com/groupee_common/emoticons/icon_wink.gif ).
~~~~~o0o~~~~~

Ooops! Another too long-winded ramble... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

Flat power is bad, yup. If were to realistically adapt an IVT to a motorcycle engine, how would you go about it? I have a feeling that there would a long process that magically happens, but the surface finishes required would also not be that hard for this magical little contraption. I'm not here to insult or say you are wrong, just to let you know that we operate in a real world where things are a little more complicated than, 'slap a custom made, smallest ever IVT onto a FSAE car, with a 3G aero package, with a turbo single making 90kW through a restrictor that weighs 200kg with driver, and hurry up and win already with that darn beam axle.' Now that was long-windeded http://fsae.com/groupee_common/emoticons/icon_razz.gif

[Mbirt]

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

Maximum Power.
-----------------------
From observations of several decades worth of restricted engines I have concluded that about 30kW (40hp) per square centimeter of restrictor area is the most anyone can get. Figures for current Le Mans cars are reasonably reliable and are about 35hp/sq.cm for these 5,000+km engines. Doing the theoretical calculation (air mass flow rate, fuel energy content, thermal efficiency, etc.) suggests a high energy fuel and very high efficiency engine (high CR, low friction) is needed to reach 30kW/sq.cm (can post calcs if anyone wants).

The only problem I have with comparing FSAE used engines to Le Mans engines is the amount of fancy extras (DLC on everything) that go into those engines and the fact that minimum weight plays into this. For this reason, teams like Audi chose to go with a V-10 rather than the lighter, more compact V-6 that could have been capable of producing the same power. They wouldn't have gained anything because the weight would have had to been made up elsewhere. Le Mans engines also run through there own restrictor. Which makes them more reasonable to look at then F1 or old Can-Am cars.



Porsche 917.
------------------
For anyone who missed it earlier, here is a link to the 917's flat-12 specs. (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=54520089051#54520089051) Also my long-winded ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=86720419051#86720419051) about the car and engine. (BTW, MCoach, why do you think this engine "was held back by it's own design of being air-cooled" given that it was so dominant in European endurance racing that it was banned, and then in North America it was "the car that killed Can-Am"??? http://fsae.com/groupee_common/emoticons/icon_smile.gif)

Notable for this engine is that with its turbo it only needed a 2-valve head for 120hp/cylinder. The 2-valve head also made air cooling easier. In the late 1970s Porsche built a similar engine for use in a broader range of racing series, and this had a 4-valve head which needed water cooling (barrel was still air-cooled).


The Porsche 917 was known as the car that killed Can-Am, yes this is true. But it's not like it was a very reliable or dominant car for very long. The car may have not been watercooled, but it was certainly oil and fuel cooled. At the time there were not power restrictions as is typical to see today, along with this was no fuel consumption minimum. Looking back through history, you can also note quite a few other cars that were able to break away from the rest of the group by having a breakthrough in technology or application. We could look at the Hudson Hornet for all I care. Using this reasoning why don't we run flat 16 engines or straight 8s? http://fsae.com/groupee_common/emoticons/icon_biggrin.gif
The undoing of the Porsche was relying on fuel cooling and lots of downforce when the fuel consumption rules caught up with them.



Other Engine Stuff.
-------------------------
In FSAE maximum power is only used for a handful of seconds per minute. Top fuel dragsters put out ~8,000hp, yet they have no cooling systems whatsoever! That is because they are only at full power for 5 seconds, then switch off (err, and the methanol helps...). So cooling a max power FSAE engine should not be a problem. If it is, then you are doing something wrong.

On the other hand, if you are aiming for the maximum 90kW from a turboed single, then I reckon you DO NEED a crank with ~50mm diameter mains and big ends. IMO the skinny little cranks on standard MX bike engines are NOT big enough. I reckon off-the-shelf aftermarket small-block-V8 parts like conrods, slipper bearings, valves, etc., would be the best way to go. 2-valve NA sbV8s make 100hp/cylinder, so just add that turbo...

The various Jawa speedway engines weigh about 28kg (admittedly without starter, clutch, etc++.). But I figure the crank alone is at least 10kg, maybe 15kg (this probably for smoother throttle response, because less likely to suddenly over rev when breaking traction). So, all things considered, I reckon a realistic FSAE total engine/transmission weight of about 30kg is feasible.


Cooling is required for how long these things stand in line waiting to take there runs around the courses. 90kW is a nice lofty goal for a nice air cooled adiabatic engine. Unfortunately, another thing that comes in when comparing to something like the Porsche 917. The Porsche never had to deal with the issue of choke flow through an orfice. The 600 engines that teams typically use make about 90kW stock, but the restrictor plays havoc with trying to flow air properly, and lots of time and resources are spent trying to just scramble back to the starting point.



In fact, my weight budget for the whole car would be;

1. engine = 30kgs,
2. chassis = 30kgs,
3. front-axle (= 2 x wheels, brakes, suspension, etc.) = 30kgs,
4. rear-axle (as above) = 30kgs,
5. everything-else (= IA, seat, electrics, aero, +++) = 30kgs
So Total Car Weight = 150kg.

Add a driver = 50kgs (or 30kgs if they have "the operation", ie., "You don't NEED legs!!! We'll just sew your feet to your butt!" http://fsae.com/groupee_common/emoticons/icon_smile.gif) gives total mass = less than 200kgs.


This as a target I can somewhat agree with... it is inline with my own targets and Although, the driver is a little small...even for my tastes.
Maybe more a kg around here and there for some +/- interactions, but I can agree.



~~~~~o0o~~~~~

TORQUE & TRANSMISSION.
=======================
Regarding the earlier posts between Racer-X and RenM. As they both said (I think?), it is the engine's maximum power that determines the car's maximum acceleration (given enough tyre grip). Having a torquey, large capacity engine is only a benefit if the car is disadvantaged by a gearbox with a limited number of ratios. A very peaky, but higher power engine will win if it can always operate at its peak power revs, perhaps by driving through a CVT or IVT.

But in FSAE it is even simpler. As I explained in this previous ramble (http://fsae.com/eve/forums/a/tpc/f/125607348/m/824105905?r=56320469051#56320469051), if you have a maximum power engine and a lightweight car, then you should be able to spin your wheels all the way down the longest straight (ie. traction limited everywhere!). So,

"The important point is, if you can nearly spin the wheels [everywhere] in top gear, then there is really no need for a lower gear. All that a lower gear does is allow you to, perhaps inadvertently, "smoke the tyres", and then possibly spin the whole car just when you want to be going straight."

As RenM said, you DO NOT WANT a flat power curve everywhere (ie. with restrictor choked). What you DO WANT is a flat torque curve everywhere, preferably with the torque just enough to give about 10-30% tyre slip-ratio (= max thrust). This flat torque curve gives a linearly rising power curve, with low power at low speeds, because that is all you need.

The only problem here is that IC engines don't run well at VERY low speeds. So it is only at low car speeds that the engine should be allowed to run a bit faster, albeit still with the same torque at the rear wheels. This can be done by slipping a clutch, or by using an IVT together with reduced throttle opening.

I repeat, the aim is NOT to increase rear wheel torque, but simply to let the engine run more smoothly by running at higher revs. Earlier I suggested the use of an IVT so the flywheel can be used as an energy store, but that is a different matter. Here an IVT would be used as a smoother, more efficient clutch.

It is worth noting almost every FSAE car has some sort of conventional friction plate clutch. These are quite large and heavy. Worse yet, they have the unfortunate habit of turning into a lump of coal at the most innappropriate times. Swapping the clutch for a similar sized (or smaller?) IVT would have significant advantages. There are also other "hydraulic" options that would help with low speed running (eg. auto-box type converter with "lock-up", and others http://fsae.com/groupee_common/emoticons/icon_wink.gif ).
~~~~~o0o~~~~~

Ooops! Another too long-winded ramble... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

Flat power is bad, yup. If were to realistically adapt an IVT to a motorcycle engine, how would you go about it? I have a feeling that there would a long process that magically happens, but the surface finishes required would also not be that hard for this magical little contraption. I'm not here to insult or say you are wrong, just to let you know that we operate in a real world where things are a little more complicated than, 'slap a custom made, smallest ever IVT onto a FSAE car, with a 3G aero package, with a turbo single making 90kW through a restrictor that weighs 200kg with driver, and hurry up and win already with that darn beam axle.' Now that was long-windeded http://fsae.com/groupee_common/emoticons/icon_razz.gif </div></BLOCKQUOTE>You missed free lunch to type this.

[Will M]

Originally posted by MCoach:



... that we operate in a real world ... http://fsae.com/groupee_common/emoticons/icon_razz.gif

As a counter point this is a Fantasy Engine thread, so if it is diffuclt and expensive to make then Z is right on target.

-William

[Z]

Hey, you guys are misquoting me!

I started with...

Originally posted by Z:
Firstly, chasing maximum engine power only wins the pissing contest. Yes, it's fun, but FSAE is won with lateral Gs.
...
Nevertheless, ...
http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif
~~~o0o~~~

Now if I really wanted a fantastic engine, as I have fantasized in the past for my hyper-supercar, then I would go for something really RADICAL ..... and use Rob's vertical crank. Yep, just drop the letter "C" ......

So, say a nine cylinder RADIAL, using perhaps 9 x 660cc MX cylinders and heads, all bolted to a bespoke, circular crankcase. Overall size about 80cm diameter by 30cm high (including bevel-gear PTO). Easy 600hp off-the-shelf. Very smooth running. Very low CG.

Or, double it up for a two-row 18 cylinder radial? That'd be 1,200hp and still only about 40cm high.....

Z

[Will M]

Z,

If you wanted to stay in the FSAE cc limit (see OP) then this would work.

http://www.troybuiltmodels.com/items/RCS400R.html

-Will;iam

[mech5496]

I assume you are all familiar with the legendary Honda C50... Those engines are reallyp opular in Greece, with a huge drag racing scene and lots of aftermarket parts. What's more, they are extremely simple and there are big bore/stroke kits from 72 up to 178cc. Tagekawa is a pretty popular engine builder for such bikes, offering everything from internals to custom 4V heads (http://www.takegawa.co.jp/2008_takegawa/flash/index_e.html). Their 178cc engine makes about 25hp @10000RPM. A radial engine out of them should be tiny (and legal), and so will a 2x178cc flat twin..

[AxelRipper]

Originally posted by Z:
Hey, you guys are misquoting me!

I started with...
<BLOCKQUOTE class="ip-ubbcode-quote"><div class="ip-ubbcode-quote-title">quote:</div><div class="ip-ubbcode-quote-content">Originally posted by Z:
Firstly, chasing maximum engine power only wins the pissing contest. Yes, it's fun, but FSAE is won with lateral Gs.
...
Nevertheless, ...
http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif http://fsae.com/groupee_common/emoticons/icon_smile.gif
~~~o0o~~~

Now if I really wanted a fantastic engine, as I have fantasized in the past for my hyper-supercar, then I would go for something really RADICAL ..... and use Rob's vertical crank. Yep, just drop the letter "C" ......

So, say a nine cylinder RADIAL, using perhaps 9 x 660cc MX cylinders and heads, all bolted to a bespoke, circular crankcase. Overall size about 80cm diameter by 30cm high (including bevel-gear PTO). Easy 600hp off-the-shelf. Very smooth running. Very low CG.

Or, double it up for a two-row 18 cylinder radial? That'd be 1,200hp and still only about 40cm high.....

Z </div></BLOCKQUOTE>

So.... This? http://youtu.be/un9BY0F9x-o

[Timmey]

More like this, http://www.youtube.com/watch?v=f2V7B7-gdRA

[Homemade WRX]

Originally posted by RenM:
As i said before you can of course also increase the Volumetric Efficiency (which is what a turbo does), but still, you wont be able to deliver more power/torque. It only helps to reduce the rev range to increase your engines efficiency.

...or to choke a 4-cylinder 250cc and make more power/torque than the engine could have otherwise put down.


Originally posted by murpia:
My point was with reference to the exhausts. You can only drop the crank height of a 180deg engine so far before the lowest thing is the exhausts. With a V you can go lower.

Regards, Ian
Oh, agreed. How we usually do on performance EJ engines is slap the drysump where the pan is, and you still have room for the exhaust. For my time attack cars, we dump the turbo right in front of the engine. So, if you picture this in a chassis, like a Saker and put it on the ground, CG is still quite low. Yes, a crank centerline could be lower on the 'V' but is the CG of the engine lower?

V in chassis:
http://www.constructorscarclub...s/Saker-Cooley37.jpg (http://www.constructorscarclub.org.nz/Resources/Saker-Cooley37.jpg)

Still a wetsumped EJ with a hi-mount turbo and top mounted intercooler:
http://sakerracing.files.wordp...09/sakerracing38.jpg (http://sakerracing.files.wordpress.com/2008/09/sakerracing38.jpg)

[Z]

Love those radials!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif

I can see myself on this one - Radial Engined Motorbike! (http://www.youtube.com/watch?v=fU_HYXOgbRw) http://fsae.com/groupee_common/emoticons/icon_biggrin.gif http://fsae.com/groupee_common/emoticons/icon_biggrin.gif http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Or, more seriously, there is the Zoche Aero Diesel. (http://www.zoche.de/specs.html) This two-stroke diesel has been around for a long time. It has many interesting features (eg. crank to conrod connection), and IMO has great potential. I suspect there must be some "competitive pressures" that are preventing sales, perhaps until the patents run out??? Such is progress these days....
~~~o0o~~~

Will,

The TroyBuilt looks good, but I don't think it is suitable for FSAE (ie. fuel/oil lube of internals, 12mm PTO (IIRC), price, etc.). However, it does show how easy it is to make a small FSAE sized engine. Mbirt's flat-head, flat-twin B&S is definitely feasible.
~~~o0o~~~

Axel,

Russell rocks! Ah, men in sheds....

That's pretty much what I had in mind. Russ's radial is a long stroke (120mm?), low revver to drive a prop. The barrels look about 100mm diameter, but I don't know what from? Anyway, a shorter stroke version would make the whole engine more compact and higher revving. Lay it down vertical-crank style, add a turbo, and that's a real supercar engine!
~~~o0o~~~

Timmey,

As above.

I have often thought of using just two of the cylinders/heads from a WWII radial aeroengine. Assemble as a flat-twin with longitudinal crank, big clutch (or torque convertor), then direct drive into a heavy duty differential (and big driveshafts!). Compact and low CG. Mount in the back of a lightweight sports car, like Micah's Saker.

About two litres per cylinder, so that's four litres. ... Then add a turbo and watch that Goggomobil go-go-go ... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[mech5496]

On my latest post, check this out:
http://www.j-mart.biz/j-parts/...7_26&products_id=358 (http://www.j-mart.biz/j-parts/index.php?main_page=product_info&cPath=1_8_77_26&products_id=358)

Shipping weight is 6.2kg with carbs, so a total weight of 11kg is possible. Add two conrods, crankshaft, flywheel and crankcase, you are still around 20kg and around 50hp from 356cc in a tiny package...http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Z]

Harry,

Hmmmmm, ... how about three of those barrels/heads on a 3-cylinder radial! Conrods side-by-side on a simple, single-throw crank inside fairly simple crankcase. ~75hp from 534cc and maybe 25kg, with very smooth running.

Maybe with a transverse crank and the cylinders in an inverted "Y" (should package well against the seat back). Or with radial laid down (with vertical crank), with chain or rubber-belt CVT to a bevel gear diff (as per Rob's post).

Z

[mech5496]

I prefer the lay down approach...actually I still prefer my inline twin! Just wanted to show that you can go a bit crazy on engines with off-the-shelf parts. Rob, would you mind giving it a go on CAD to compare with youer Briggs derived flat twin? L

[jlangholzj]

Originally posted by Z:
Love those radials!!! http://fsae.com/groupee_common/emoticons/icon_smile.gif

I can see myself on this one - Radial Engined Motorbike! (http://www.youtube.com/watch?v=fU_HYXOgbRw) http://fsae.com/groupee_common/emoticons/icon_biggrin.gif http://fsae.com/groupee_common/emoticons/icon_biggrin.gif http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

Or, more seriously, there is the Zoche Aero Diesel. (http://www.zoche.de/specs.html) This two-stroke diesel has been around for a long time. It has many interesting features (eg. crank to conrod connection), and IMO has great potential. I suspect there must be some "competitive pressures" that are preventing sales, perhaps until the patents run out??? Such is progress these days....
~~~o0o~~~

Will,

The TroyBuilt looks good, but I don't think it is suitable for FSAE (ie. fuel/oil lube of internals, 12mm PTO (IIRC), price, etc.). However, it does show how easy it is to make a small FSAE sized engine. Mbirt's flat-head, flat-twin B&S is definitely feasible.
~~~o0o~~~

Axel,

Russell rocks! Ah, men in sheds....

That's pretty much what I had in mind. Russ's radial is a long stroke (120mm?), low revver to drive a prop. The barrels look about 100mm diameter, but I don't know what from? Anyway, a shorter stroke version would make the whole engine more compact and higher revving. Lay it down vertical-crank style, add a turbo, and that's a real supercar engine!
~~~o0o~~~

Timmey,

As above.

I have often thought of using just two of the cylinders/heads from a WWII radial aeroengine. Assemble as a flat-twin with longitudinal crank, big clutch (or torque convertor), then direct drive into a heavy duty differential (and big driveshafts!). Compact and low CG. Mount in the back of a lightweight sports car, like Micah's Saker.

About two litres per cylinder, so that's four litres. ... Then add a turbo and watch that Goggomobil go-go-go ... http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

basically you're thinking of a variant of a A-65 or C90 continental that are common in something like a champ. Problem is they're pretty heavy compared to what we've already got availability to. I think we're along the right line though, I've always fantasized about a flat-4 or boxer style engine is FSAE.

I think the trick would be incorporating it into a transaxle from the engine though....think something like a smaller FSAE spec porsche....

mmmm....slant nose p-cars....

[rjwoods77]

Originally posted by mech5496:
I prefer the lay down approach...actually I still prefer my inline twin! Just wanted to show that you can go a bit crazy on engines with off-the-shelf parts. Rob, would you mind giving it a go on CAD to compare with youer Briggs derived flat twin? L

I did a whole bunch of work on replacing the Briggs V twin with a Jawa in many different positions (some of which are in previous posts with pictures) and it works nice compared to a v twin. A parallel twin would just be the same in the side view albeit shorter and in the top view a wider. It won't package as tight,low and be manifold friendly as a flat twin setup and on top of that you will get even pulses and better NVH. I really think in a parallel twin cylinder config the lengths of the primaries to the plenum for intake and collector for exhaust is actually a packaging deficiency in comparison to the separated ports of the flat twin. Whether vertical or longitudinal ports you pretty much package the intake and exhaust below the side impact height rule especially if a turbo setup.

[MCoach]

NVH is overrated.

[rjwoods77]

Originally posted by MCoach:
NVH is overrated.

Being around engineers of various discipline in engine technologies makes me wonder how much that is true. Triboligists with their obsessive talk about oil films and pump engineers talk of cavitation and valvetrain engineers talk of harmonics for example. NVH seems to be very important to them and their related subsystems. For what FSAE does...the possible benefits I don't know...but considering such things might be important. BMW went through an awful amount of rigmarole on their newer parallel twin...

http://www.ashonbikes.com/bmw_engine_balancer

[mech5496]

Rob, I meant ginving the Tagekawa-parts flat twin a go, not the I2...http://fsae.com/groupee_common/emoticons/icon_wink.gif

[MCoach]

Right, to a point, it can do some devastating things. Steel pistons have the tendency to cause the coolant in the water jackets to cavitate and eat away at the walls, shaking off oil films cause increased wear, etc. Things like that are a real concern and need to be taken care of. I won't be completely ignorant on the subject.


But, it's the things like filling the valleys in the intake of V style engines with dense foam because the companies are concerned about drivers hearing the clicking from injectors or specifying an exhaust system with a muffler so big it's known as a 'suitcase'. Stuff like companies complaining that the sound of the hydraulic unit (made by a third party) needs to be redesigned because it can be heard when the radio and AC are off, and travelling at a speed which wind noise is not heard. Complaining that a quiter, although weaker belt design is needed because you can hear a slight hum from the engine running in the same condition. That's the part that gets me ruffled about NVH...

[rjwoods77]

Originally posted by mech5496:
Rob, I meant ginving the Tagekawa-parts flat twin a go, not the I2...http://fsae.com/groupee_common/emoticons/icon_wink.gif

No real need to make anything different that what I have. The bore and stroke on that B&S is 75.5x66 and the Tagekawa is 67x50.5. Assuming rough parts size similarities the deck to deck height should be about 66(stroke)-50.5(stoke)+10(approx connecting rod delta)= 25.5mm x 2 = 51mm. But the size of the OHC heads are much bigger length wise, 25mm taller would be a fair hip shot, so the overall length of your 356cc engine would probably be the same as the B&S flat twin. Overall all cooling fin dimensions in the side view you could say would be minus 8mm in height and width which is really negligible. So if you use my B&S flat twin model it should be right around the correct envelope of space it would take up. I would be more than happy to provide you or anyone a parasolid to insert to see what the comparative packaging differences would be versus current engine and chassis geometry.

[Z]

ATTENTION ALL AUSTRALIAN TEAMS !
================================

I think I may have found the "Eureka Engine". (From Greek = "I have found it!", and reference to Oz's colonial past...)

The Radial Engined Motorbike I linked to above uses the Rotec R2800 (http://www.rotecradialengines.com/0RotecR2800/R2800.htm) 7-cylinder radial aero engine. And Rotec just happen to be a small and growing family business based in Melbourne, Oz!

Some background from their website. The company is run by two brothers, one an Auto Electrician, and the other a Toolmaker/Machinist. Dad does the books. Fortunately, not a qualified Engineer in sight, else they mightn't have got off the ground!

Their endeavour started a little over ten years ago with a 350cc 7-cylinder prototype engine (ie. 7 x 50cc) intended for the bigger model airplanes (similar to the TroyBuilt in earlier post). Customer interest drove them to building the 7 x 400cc (=2,800cc) R2800, and later the bigger 9 x 400cc R3600, for full-sized, homebuilt aeroplanes. All these engines are closely modelled on WWII radial aero engines. Particularly obvious is the similarity to Pratt & Whitney's mass produced R2800, although the P&W has 2 rows x 9 = 18 cylinders, and displaces 2,800 cubic inches (= ~16 x the Rotec's 2.8 litres).

Of interest to FSAE is that the Rotec's engines have a "square" 80mm B&S, and a fully machined and nitrided 4140 steel cylinder that is screwed into the aluminium head. Early heads were fully machined 6xxx billet aluminium, although current heads are castings. Specs are roughly 15+hp and 15kg per cylinder for these low revving (3,600 rpm) aero engines where reliability and economy are priorities.

See these Rotec Engine Photos (http://www.rotecradialengines.com/photo.htm) to see how "men in sheds" can build beautiful engines, without ever asking "pls urgently post equations for optimal engine design...".
~o0o~

Anyway, here is my idea:

You approach Rotec for some sponsorship in the form of some of their head/barrels, plus miscellaneous small parts, and advice on how to make stuff. You then design and build a laydown single-cylinder, 400cc, air-cooled engine, which is not far removed from 1/12th of the Porsche 917's fearsome flat-12.

Initially you just make the crankcase, maybe use the existing crankshaft, buy a suitable conrod, make a small camshaft and its drive (the radial's "ring" cam is unnecessary), maybe add a balance shaft or two, plus all the odds and ends like EFI, fan, etc. This very lightweight NA engine should comfortably rev to ~7,000rpm, and give 30kW (40hp). Just right for a simple, lightweight, FSAE car.

Next add a turbo. The barrel-screwed-into-head is perfectly suited to high boost pressures. With 1 bar boost (2 bar absolute inlet pressure) you should see 60kW. At 2 bar boost (ie. 30 psi gauge) you should be able to get the maximum 90kW available through the restrictor.

Or, then again, you could ask some trade school students to do it for you? http://fsae.com/groupee_common/emoticons/icon_smile.gif

Z

[raitinger]

It looks like different aspects of my fantasy engine have been mentioned, but I'll add my two cents.

Mine would be a 610cc boxer twin. Direct injection, prolly N/A?, incorporated dry-sump, and I can't imagine it would be physically possible for our application but air-cooled. Also, with unlimited funds you can throw dream materials at the entire thing to Be pistons, Ti, etc. Roller and ceramic bearings all around.

I guess what I haven't seen that much discussion on, is the benefits of starting from scratch regarding the torque curve of your choosing and general nature of the performance. I'd say a big disadvantage of the 600cc 4cyls vs the singles is that the 4s aren't developed for the demands of autocross as much as the singles. The 4s are much better suited for a road race type application where the durations of acceleration are much longer and high RPM HP is important, not really all that important for SAE. Moving to the designed purpose of the singles of MotoX or Super Moto, and more responsive shorter bursts of acceleration become very important. Finally for SAE, we are lucky to see a long uninterrupted opportunity for acceleration. For one the longest the rules allow is 250' with hairpins at both ends, and it's rare for that pure straight to make it into an SAE course. Although we are planning on making it a staple at Lincoln http://fsae.com/groupee_common/emoticons/icon_smile.gif

Long story short with unlimited resources you could make a very responsive torque monster, hell your new driving range could be a Big Block V8 style 1500-7000 rpm. Doing this you would negate some of the downside of a large displacement engine with a restrictor.

I chose the boxer layout because after overall weight and power, Cg height and affect on yaw moment would be the next biggest factors for contributing to overall vehicle performance. With the opposed twin the engine could be essentially right against the driver's lower back, and the driver could be reclined over top of the engine. I don't think it gets much better than that before you integrate the driver into the engine.

[rjwoods77]

Raitinger,

Forced air cooling on our B&S was achieved with a simple squirrel cage blower fan attached to the flywheel and some sheet metal guards. All of the air cooling subsystem parts weighed next to nothing and didn't take up much space...

http://i.ebayimg.com/t/20-Hp-V...P3(kmyP5w~~60_12.JPG (http://i.ebayimg.com/t/20-Hp-V-Twin-Briggs-And-Stratton-Flywheel-Fan-for-Motor-407777-Part-791236-/00/s/NDgwWDY0MA==/$(KGrHqFHJBcE+N1!PWidBP3(kmyP5w%7E%7E60_12.JPG)

http://i.ebayimg.com/t/garden-...RDTmE6yYQ~~60_12.JPG (http://i.ebayimg.com/t/garden-tractor-16hp-briggs-stratton-vanguard-v-twin-AIR-SHROUD-HORIZONTAL-/00/s/MTIwMFgxNjAw/z/VRQAAMXQH-pRDTmF/$T2eC16F,!zEE9s3!(JrCBRDTmE6yYQ%7E%7E60_12.JPG)

...E85 would help with cooling even more. I have a parasolid file of a 590cc air cooled flat twin you could throw in your model and see what it looks like if you are interested in comparing.

[rjwoods77]

Bump with a delicious air cooled clicky clack cacophony...

http://www.youtube.com/watch?v=mGaGIwdp-f0

[mech5496]

On Rob's idea with bevel gear final drive, here's a list of commercial available motorcycle RnPs, together with their ratios as well as rated power and peak shaft torque. Might be useful to some of us. If you have more info, please add them up.

http://www.datafilehost.com/download-6d749f37.html

[Jan_Dressler]

Originally posted by Z:
Now if I really wanted a fantastic engine, as I have fantasized in the past for my hyper-supercar, then I would go for something really RADICAL .....
Me too... With the letter "C" http://fsae.com/groupee_common/emoticons/icon_biggrin.gif
http://i45.tinypic.com/1zbdilh.jpg
http://i49.tinypic.com/2r5qfdk.jpg
Sorry for the late answer http://fsae.com/groupee_common/emoticons/icon_wink.gif And yes, it is only a compilation of ideas, and not "complete" yet, still working on it.

[mech5496]

Jan,

Care to add a few more details? (Impressive CAD nonetheless)

[Jan_Dressler]

Thank you http://fsae.com/groupee_common/emoticons/icon_smile.gif Not really that impressive, only basic CAD functions and quite a lot of time...

Details:

- V6
- 2.4 l (bore 100 mm, stroke 50.9 mm)
- 60°, 90°, 120° V-angle (several reasons:
- shorter engine (about -50 mm compared to a V with one angle);
- low CoG on one end, possibility of raising the floor (and creating a good cokebottle shaped bodywork) on the other end (when installed in mid-rear position. If installed in mid-front position the other way round, it is possible to create a raised nose in a modern Super 7-interpretation and so on).)
- Rotary Valves (at 30° angle) à la Bishop (search for the patents, and Horrock's "A Numerical Study of a Rotary Valve Internal Combustion Engine" (and Hunter's "Rotary Valve Engines"))
- Valve drive with helical gears (the blue ones)
- Built crankshaft (Hirth joints) with rolling-element bearings (main and conrod bearings), one piece conrods, 0° - 90° - 210° --> TDCs in 0° - 60° - 90° - 180° - 210° - 330°
- Dry sump lubrication with centrifugal air-oil-separator, piston cooling

Features still to add:
- Intake system (on the outside of the V)
- exhaust system with 2 turbogenerators / 2 turbines working on one generator
- 1 electric supercharger
- hybrid assisted drivetrain

- the other parts of the car http://fsae.com/groupee_common/emoticons/icon_biggrin.gif

But since this not an FSAE engine, this is quite off-topic...