projekt

This isn't an ad, just a conglomeration of ideas i've had over the years to modernize the rotary. please don't flame, if i'm unclear, i'll attempt to clarify.

1. weight: switch over to aluminum. this one's pretty simple.

2. longevity: switch from a 5 piece block using rubber seals to a 3 piece block with composite gaskets. this one take some imagination to picture. first, imagine the rear plate now attach half of the rear rotor housing to it to make one piece.
assemble and install the rear rotor into this new rear piece.
next take the center iron and attach half of the front and rear rotor housings. bolt that down to the new rear piece. install front rotor.
finally take the front plate with half the front rotor housing and bolt that down to the new middle piece.
the purpose of this is to reduce the amount of seals as well as taking the combustion pressure away from a corner.

3. economy: implement a gas saving feature so that under light load (cruising) fuel is injected on every other rotor face. each seal will still be getting lubed, and each rotor face would still be used, but half the time. this one is pure theory since i'm not even sure the rotary would maintain enough torque in this mode to propel a 3000lb car.

questions and comments are very welcome.

RacerXtreme7

iTrader: (2)

Manufacturing a side housing with an integrated rotor housing is going to be extremly difficult. I don't see a way to add the steel liner, machine it, chrome it, while its got a side housing attached to it. The junction were those meet will always be fuct. Also, using a gasket instead of an "o" ring will leave the gasket exposed to all the cycles of the engine not to mention leave a gap were the side and rotor housings meet. You couldn't get the apex seals to seal this area. Also it would be difficult to clearance the width of the rotor/seals because the distance would change more with a gasket in there dependant on temp and torque applied to the tension bolts.

Other then those issues, good theories, just hard to manufacture.

I do however support all aluminum but Racing Beat already offeres the side housings (or at least used to). Aluminum rotors would be nice.

~Mike...........

projekt

couldn't the steel liner be a single piece too? sort of a pot shape? it could be machined with a tool that runs the epitroid shape, and chroming could be done electro magnetically since the alu won't conduct as well.

i see the point on the plates coming together. would an offset gasket work? by this i mean have the steel press against the steel with a gasket (or o-ring) behind it?

RacerXtreme7

iTrader: (2)

"couldn't the steel liner be a single piece too? sort of a pot shape? it could be machined with a tool that runs the epitroid shape, and chroming could be done electro magnetically since the alu won't conduct as well. "

I'm pretty sure its done like this already. The steel liner starts off as flat sheet stock with teeth, is rolled to form the epitroid shape, then pressued into the rotor housing then machine, then chromed. Or maybe the rotor housings are cast with the liners in them, then machined then chromed. I don't know for sure, I'm sure someone here knows and would chime in. I'm not totally sure on the process of how these are manufactured, I guess it could work. one down side is not being able to lap the side housings during a re-build. but I guess there can always be tooling made for it with ultra presice CNC machining.

As far as sealing from rotor housing to side plate, "o" rings have always worked great. the only failures I've seen are when the side housings have a weak casting in this area due to core shifting during the casting process, or corrosion either on the rotor housing or side housing and thats due to neglect. I've also seen "o" rings get pinched during assembly process, but thats justengine builder error. I see no need to use anything other then the standard "o" ring that MAZDA has used other then possibly making the meat around the "o" ring groove meatier.

I say keep the design as is, make the sides and rotors out of aluminum alloys, add more meat to various areas ("o" ring, and rear plate were these tent to crack) and call it good along with the use of ceramic seals.

~Mike................

projekt

i recall reading that o rings like to die under boost or higher than normal temps. i might be wrong.

for the offset it would look like this: (| = one housing, 1 = other housing)
||||||||||
|||O111 [Rotor]
111111

left is outside, right is inside.

wouldn't ceramic seals be very costly?

any comment on the economy part?

RacerXtreme7

iTrader: (2)

I don't know if there would be enough power to maintain cruising speeds to justify every other cycle being dead, not to mention pumping looses. the rotary has low torque values do to the small moment are on the eccentric shaft (but still compairable with small 4 banger). The engine will still have to suck air in, compress it, then force it out with no benefit of it having a expansion cycle. I have no idea.... It does work on the Cadillac Northstar V8's however, (turning off 4 cyclinders of 8 cyclinders at cruise) but they generate nice torque numbers. There could be a port in compression cycle that would bleed off air, sorta like a miller cycle type of thing. Then it could close off during a real combustion event. This will give the rotary a complex valve system like a boinger though.

For economy and power, I think the next step for the rotary is direct gasoline injection. I see a treand developing in 4 cyclinder piston engines. Were now in the early stages, but theres at least 4 different engines running direct gas injection in full production now from different manufactures (VW/Audi, Mistu, etc.). You get ultra lean light load burning for cruise, and benefit of a more complete burn. The rotary could benefit from this greatly by keeping the fuel away from the narrow ends of the rotors were the flame front likes to die out. Place a plasma chamber in between the current spark plugs in the rotor housings with a ultra high pressure fuel injector shooting directly into this small chamber for idle and light loads (there would need to be a spark here also). For WOT and heavy loads have an injector placed directly into the comp chamber. You also get the benefit of VE going up because your not dissplacing the incomming air with fuel, it'll all be air then intruduce the fuel once compression happens.

I could go on for hours about this stuff, but I'm at work...LOL

~Mike...........

RacerXtreme7

iTrader: (2)

For n/a and moderate turbo applications the OEM steel seals perform well. And yes ceramics are expensive, but that price would drop significantly if they were produced in large automotive type numbers. I guess it doesn't make sence for MAZDA to use them when the steel ones hold up, but if they were to build an engine making more power to be competative with other current sports cars (EVO, STI, etc. making 300 or more hp) turbo charging is a must and at those power levels if MAZDA used ceramics, the engines would beable to get the 200k miles people expect out of modern cars.

OK, back to work I go........ for now

~Mike.........

projekt

ya i just realized a problem with my center O ring idea. spark plugs. you'd have to accomdate for them because having two half threads isn't exactly intelligent. the idea can be scrappe for a simple 3 piece setup still though. just add the read rotor housing to the rear plate and the front housing to the center plate. one less seal to fail per rotor.

i agree on the DI system, but is the top of the engine too hot for an injector? i suspect that was part of the problem for DI systems for piston engines since the head is a relatively hot area. something that would help the combustion itself would be to redesign the motor (since we're doing it anyway and make the rotors smaller in height but deeper. gives us a more square surface to work with.

for the alternating cycle system, i agree that pumping loss would hurt a lot, as well as the rotary's design torque deficiency. as far as having a complexity of vavlves, that's no biggy, look at the RX-8's intake system. it's retarded complicated. that's acceptable for a production car. the tough part would be getting the nonbreathing steps to moderate by using the intake manifold.

and still: ALUMINUM! (very long overdue)

RacerXtreme7

iTrader: (2)

I had to re-read your orginal post.... I didn't realise you meant split the rotor housings in two and make half on each side housing. Bad idea, you'd never get the apex to seal worth a pooh amongst other problems. but yeah, the rea side housing with a whole rotor housing, then standard center plate, then whole housing part of the front side housing could work. But it sounds awfully expensive to manufacture just to eliminate 4 O-rings that really don't fail too often as is.

As far as the bleeder valve system, it would require a camshaft and popit valves (YUK!). I don't know if valve train drag would exceed pumping effiency gains.

~Mike..........

projekt

so then really all the improvement i could hope for is Alu, DI, and optomized engine dimensions?

i also thought about making the apex seal more of a pyramidal shape. so the cross cut would look like /_\ but it seems that the tip of it is the weakest part and this would do very little to address that.

fwiw, i'm not an engineer.

Tom93R1

iTrader: (2)

Mazdaspeed 6 is direct injection, I couldnt imagine it would be that difficult for them to eventually port it over to a rotary application.

Not sure I understand the point of doing a "pyramidal shape" apex seal though. How would it "float" in the groove to remain pressed against the housing? I could imagine it falling over and the spring popping out.

projekt

have the grove be a pyramid as well with the srping underneath it.


edit: i hate the inability to draw with ascii in these forums.

RR/A\RR

RR/ and \RR is the rotor tip. A is the pyramidal seal.

Tom93R1

iTrader: (2)

So how does the seal move up and down like that? Normally the rotor housing is what limits the up and down movement of the seal, in that configuration the rotor tips would stop its movement.

projekt

it doesn't have to touch it, just keep it inserted. base of apex seal is wider than the gap in the top.

remember it's just an idear.

RacerXtreme7

iTrader: (2)

It would rock back and forth like MAD and wear itself, or the groove out rather quickly (then probly fold over). Think, while one side of a rotor face is compressing, the other is under suction, or even if its turbo charged, theres still less pressure on one side then the other. If the seal were sloppy, it would rock back and forth every single phase of the combustion process (several events per RPM). Its fun to brain pic, but this idea won't work.

~Mike.........

Ronald E. Jacques

Aluminum is much lighter, but not nearly as strong as steel. It weighs about forty percent of what steel weighs but has only about sixty percent the strength. Maybe that is strong enough, I don't know. I think titanium would be better to make the rotors out of. It weighs about 70 percent of steel, which would still be a serious savings, but it is much stronger. It is impervious to corrosion.
"Fatigue cycle" is how far you can flex a piece of metal before the cumulative effect will eventually wear it out. Steel has a good fatigue cycle which is why springs are made of it. Titanium is much much better than steel in its ability to resist fatigue, which is why most "cost no object" situations like F1, jets, and the space shuttle use it wherever possible. Aluminum has a fatigue cycle of ZERO. It is very stiff but any bending whatsoever will eventually cause failure. Look at aluminum connecting rods compared to the steel ones for the same application. They make them super beefy to be sure they can resist the stress. They are still lighter but not one to one compared to the weight difference in the material alone.
Now I don't know, maybe an aluminum rotor spinning at 8, 9, 10,000rpm or higher would be just fine, but I'd bet that Mazda would want to beef it up wherever possible, filling in all the hollow cavities and such. It seems that from a cost-benefit point Mazda made their factory rotors as light as was practical for their strength, for steel that is. But with Titanium just naturally being lighter AND stronger, more durable, more heat resistant, etc. they could just cast and machine the same exact rotor out of the new material and be done with it, minimal more R+D necessary.
Titanium is alot more expensive than either steel or aluminum, but how much could it possible cost on a new car to upgrade? 500 bucks? who knows. I'd pay the diffference.

Tom93R1

iTrader: (2)

Titanium is really big $$ to cast. I would bet more than $500 cost difference in titanium vs. steel. Titanium needs to be cast in a vaccuum because it will very readily react with the oxygen in the air. To mass produce parts in a vaccuum is not easy or cheap to do. Maybe at some time in the future but as of now I could imagine several thousand $ in cost difference between a rotary engine with steel vs. titanium rotors.

Ronald E. Jacques

^^^Now that I had forgotten about. Maybe just machining them from solid billet. Who knows. It would be awesome though.

Gag_Halfront

I think the "dead cylinder" idea is a good one. Perhaps some of those hardcore guys who program their own ECUs and stuff can look at this. I think the compression could be bled off by a solenoid that was controlled by the ECU rather than a mechanical valve actuation. that would prevent adding load to the engine right when you're trying to reduce power output and would also maintain the simplicity of the overall engine design. If you fed the compressed air back into the intake you may even see your losses reduced further due to a mild supercharging effect. You would have to compensate by addingmore fuel, but that's ok because you're injecting this thing anyway, not carburating it so you can just shoot more juice when the blow-off solenoid is open.

Now, if you shoot enough gas in to maintain the same A/F when you've just pumped all your air back over from the previous rotor face then you're not really saving anything. Also if you blow off the entire intake charge you will have a low pressure in the chamber when you come around to the exhaust port so you will risk sucking exhaust back into the chamber. This wouldn't be too big of a deal for the next intake cycle because you would end up expelling it all again anyway, but you would destroy your exhaust flow. So your two problems then are that if you pump all the air from the compression cycle back into the intake and compensate with more fuel, you aren't really saving anything, and you don't want to destroy exhaust scavenging by going into the exhaust stroke with low pressure in the system. So here's the compromise. Only hold the blow-off open for the first 30% of compression. Then you're reducing fuel by 60% on that cycle (probably about 45% overall), reducing pumping losses, regaining some of the power lost to the "dead cylinder" and maintaining a little bit of pressure to keep from backflowing when you get to the exhaust port.

I think the only requirements here are fuel injection, a blow-off port controlled by an electric solenoid, and custom ECU software. Oh, and a very accurate crank position sensor.

I think I just spilled out another one of those patentable ideas. *sigh* It's no wonder I'm not a go-zillionaire yet. And another one's gone. and another one's gone. Another one bites the dust!

Zero R

iTrader: (1)

If memory serves correct the housings are cast around the sleeve it was a process mazda patented at the time.

Snrub

iTrader: (1)

Here's some info on direct injection rotaries and miller cycle rotaries and some of my thoughts on the issue. http://www.geocities.com/jeffguilfoil/discmiller.html

The following is from Matras, J. Mazda RX-7: Sports Car Color History. pg 118 MBI Publishing Company. 1994. Osceola WI.

DISC, for Direct-Injection Stratified Charge, evolved in the late 1980s. Two low-pressure injectors were sited on either side of the rotor chamber close to the trailing spark plug. High-speeed air injected in the nozzles' sockets aided fuel atomization. Engine speed was controlled by fuel injection, which reduced pumping losses by the absence of a throttle. The engine also required less cooling as combustion took place in a more localized area, which also reduced emissions of unburned fuel (hydrocarbons). But the lower exhaust temperature caused another problem: It wouldn't light up the catalytic converter!

DISC-II solves that problem with an elaborate fuel/air management system. It involves a pilot injector and spark plug in a small subchamber peripheral to the main chamber. The pilot injector sprays through the subchamber to the main chamber, formnig a pocket of rich mixture in the main chamber. This pocket is ignited by a flame front from the stable mixture in the subchamber. Meanwhile, the main injector shoots fuel into the main chamber past a conventionally (for a rotary) located spark plug.

And while Mazda put Miller-cycle technology into a production piston engein in the 1995 Mazda Millenia in the spring of 1994, it was also working with the pumping-loss philosophy with the rotary engine. "Pumping loss" is the energy an engine wastes at closed or small throttle operation, when trying to draw in air. The Miller-cycle rotary, called "near-future" in Mazda press material, allows a great volume of air to be supplied to the rotor chamber. Any excess of air is released via a "late" side port controlled by a "pumping-loss control valve" which recirculates the excess air to the intake port behind the throttle. Direct fuel injection solves the problem of mixture control with a single injector later than the "exit port."

My comments:
Direct injection would appear to solve a number of the rotary's problems. For one, while the rotary has superior mechanical efficency to a piston engine, it has poor thermal efficiency. This fault reduces the potential power and efficiency of the engine. Second, the rotary engine is inefficient with fuel and has poor emissions. Direct injection produces more precise measuring and allows for stratafied charges (gas clouds of varying density of fuel) to improve fuel efficiency and ensure more complete combustion occurs, meaning improved emissions. The problem mentioned with respect to DISC-I is that the catalytic converter did not receive enough heat. I have trouble believing that the thermally inefficent rotary with direct injection cannot produce enough heat for a modern catalytic converter, but a 1L 3-cyl engine can. My point is that I imagine that this problem is now more or less solved.

DISC-II seems overly complicated and probably no longer necessarly. One point of interest is that the rotors used in the DISC-II engine are of the trailing recess variety. Leading recess rotors have a deeper dish towards the front of the direction of rotor travel, which I believe were found in some FB rotors. The DISC-II rotors remind me of the Curtis Wright rotors that triangular shaped tips with that extended out to create a deeper dish. Picture adding a triangle over a rotor tip from the side profile, with the edges protruding out, it looks almost like an arrow head. The DISC-II rotors have this arrow head profile on at the front of the flame front, but at the back it actually does the opposite and atttachs ot the tip tip of the rotor at a sharper angle.

I really like the concept of the Miller Cycle engine and I do not understand why more of these engines have not been produced. Essentially all Miller Cycle engines are supercharged, which to me seems more efficient than being naturally aspirated. Second the engine's efficiency is improved meaning more fuel economy and potentially more power. The rotary engine needs all the fuel economy improvements it can get. Rotaries are typically lacking in low-end power, so a supercharger would improve that situation. In piston direct injection forced induction engines they are able to run high compression ratios in boosted applications, without causing detonation. Presumably this trait would be true in a rotary, meaning that power and efficiency could occur at the same time. Mazda has produced a direct injection engine for the Mazdaspeed 6 and some JDM vechiles. I think that the rotary despirately needs some improvement to keep pace with modern piston engines and these solutions seem ideal.