Karack

i wish there was a way to revolutionize the rotary engine, it might still be alive.

true direct injection? ceramic seals? too expensive for a production car. it makes me think back to the times of felix wankel, when he had the spark plugs installed INSIDE the rotors... maybe he was smarter than mazda, like george lucas before he had the bright idea to actually make star wars series 1-3.

maybe technology will come to the point of making that actually possible without a rebuild every 10k miles for a tune up and nightmares of ignition diagnosis. a reverse george lucas perhaps.

Brettus

iTrader: (2)

what if you changed the triangles into circles and had them go up and down instead of round and round ?

RIWWP

iTrader: (2)

(Hmm, a pair of ***** bouncing would work for that too....)

Karack

i'm fairly sure even that has been done, by a few MIT students but it's still a rotary engine design and not a boinger..

A New Rotary Engine Design | MIT Technology Review


maybe someone else will pick up where mazda left off, but i'm not foreseeing the original "breakthrough" in this thread.

insane<se3p>

I started to wonder, if the engine can use a process to skip every other power stroke or something? You know like how v8 can use 4 cylinders then switch to all 8. Run off one rotor cruising then both during hard acceleration.

RX what?

Sure, but it would need to close off intake and exhaust ports for that housing, turn off ignition and injectors, turn off oil metering pump or reduce quantity injected to not foul plugs but prevent dry seals from scoring the housing (i would assume anyways).

Carbon8

iTrader: (4)

You would only be able to cut spark, the same amount of fuel would be injected but would just not be burnt. Rendering the idea useless.

A better idea would be to just kill one rotor, but then you would get into how to keep that rotor from destroying itself from oil starvation.

RIWWP

iTrader: (2)

You could cut fuel per face, not a challenge for the manufacturer. You could even ensure lubrication without fouling. The real problem is pumping losses. The shut off rotor would cause huge drag on the working one, losing far more than half the power and killing efficiency more than just optimum 2 rotor firing is already at.

Cylinder deactivation disconnects at the crank, so the dead pistons aren't pumping at all. Doing that on a rotary has a huge number of problems to solve, with balance being a big one. Piston deactivation can deactivate in balanced sets.

Mazmart

iTrader: (12)

You are partially correct: the de-activation is not at the crank, so the pistons are still moving, but, their valves are not functioning (They remain shut) so no air is being pumped.

The Wankel in it's simplicity has no way (So far) of shutting off air in similar fashion.

Paul.

RIWWP

iTrader: (2)

Hmm, I thought I saw a tech breakdown once of one version that had the front pair of cylinders in an inline 6 that physically separated from the rear 2/3rds of the crank via a gear/clutch mechanism. V configurations would be much harder to accomplish this on. Agreed that shutting valves, or having shutter valves close, would be simpiler, though the pistons would still be doing work unless they could produce a vacuum in the cylinder. Not likely given piston ring tolerances, and trying could result in some messy oil results.

Most solutions I see to some of the core issues with the current rotary add complexity and parts. Like variable valve timing, any deactivation method, etc... Part of me thinks that eventually we are going to pick up some more moving parts.

RX what?

Like an ssv but its sole purpose is to close off flow to a housing. And another to close off flow in the exhaust. It would just need a modified lim and header and electronically controlled valves. As far as the "huge drag" of a shut off rotor, seeing that it doesn't take much force to manually turn this engine i would assume momentum and the one rotor can handle it no problem. Its intention is cruising under basically no throttle.

RIWWP

iTrader: (2)

If you don't think that there is huge drag on the engine when there isn't fuel being injected and then sparked, let off the gas in 1st gear at 8,000rpm at some point. The whole concept of truck jake brakes is founded on exactly how much drag an un-fueled un-sparked combustion chamber can produce. If you need ~40hp for cruising on the highway with both rotors firing (which is about right for the RX-8), you will need the active rotor producing something more like 60hp to provide that 40hp plus powering the substantial air pump in the next rotor. The rotary flows a LOT of air, and if it isn't using it to burn fuel, it's going to be a lot of work.

Brettus

iTrader: (2)

It's not an air pump if you shut off the air to it ...................

RIWWP

iTrader: (2)

Unless you can render it a complete vacuum, it will always still be compressing air and/or decompressing it. The more vacuum-ish you can get it, the better, but it will always produce some drag. A minimal amount of air will mean most of the energy cost will come on the expansion side, twice per rotation. A lot of air means the cost is on the compression side. It will always cost energy do to either. Balancing the amount of air to the average volume will be costing less for both, but neither will be zero. And given the rotary's sealing problems, I doubt a complete vacuum is possible. Likely, it would be constantly transfering air from one face to another, at even more energy cost.

Karack

if they ever went far enough with true direct injection then they could cut out whole firing sequences per rotor. alternating 1 firing sequence per rotor so that all 3 faces are still used but cutting power needed by 25%.

RX what?

Are you taking into account the ratio in 1st gear? Do the same 'experiment' in 6th. Worlds apart. Besides one rotor wouldn't be operating at 8k rpm. The whole point of the discussion is for cruising so say 4-5k in 6th gear. And like mentioned before, if air flow and exhaust flow are cut for the deactivated rotor there is no air pump.

RX what?

16x incorporates direct injection. So it might have an eco mode that does just that.

RIWWP

iTrader: (2)

No, I didn't forget the gear ratio. That gear ratio won't be helping any in 6th, because your torque factor for overcoming the same drag at the same RPM is much worse in 6th than in 1st.

RX what?

That's true, although I was pointing out the difference in deceleration from 1st and 6th.

RIWWP

iTrader: (2)

Yup, i'm aware of the difference in decel. My example was countering your point about not producing any drag when it's not firing. It does produce drag, alot of it.

Brettus

iTrader: (2)

But shutting off the air to a rotor ..... isn't that the same as not opening the valves on a piston engine ?
Sure you get the rotor having to work against vacuum on the intake stroke but then that works for you on the compression and exhaust stroke ... it kinda balances out .... doesn't it ?

RIWWP

iTrader: (2)

On a piston engine, they have to figure out how to prevent the camshaft lobes from pushing on the valves in order to prevent them from opening. An alternate cam profile with similar theory of variable valve timing would accomplish it. Alternatively, they could use actually shutter valves in the intake and exhaust runners to shut down flow that way too. When I swapped a 2001 engine into my 99 miata a while back, i had a problem off the start because the variable tumble control valves on the 2001 are against the head, and almost completely seal off the intake when closed, but my 99 ECU thought it was controlling an alternate version and left them closed till 5,500rpm. The tiny hole in each plate was good for about 25g/s at full throttle...talk about dead on it's face.

But, regardless of the method to seal off the combustion chamber (a rotary would need some sort of blocking plate/valve across all intake ports and all exhaust ports), it still wouldn't be a perfect seal, either piston or rotary. Piston rings aren't perfect in their sealing, and it would be pulling and pushing air from/to the crankcase through a rather small tolerance. For the rotary, it would be pushing/pulling past the minor tolerance gaps between the apex seals and housing, corner seals and side seals, and side seals themselves, taking/giving from other faces.

Even if you were able to obtain a complete seal, yes, compressing the air present would then help you on expansion, and expanding with minimal air would then help you on compression. But again, there is always a parasitic energy cost, it is never net-zero-loss.


Thinking about it, on a V configuration engine, shutting off 1 bank wouldn't actually have to be "shut off", it could just run as if it's idling, with extremely minimal fuel injection and appropriately restricted airflow, to prevent a pumping loss. You could even use an EGR type connection to fill the combustion chambers with inert gasses rather than trying to pump small amounts of air at high speed or allowing too much air leading to detonation/no power anyway. They typically have a different set of O2 sensors for each bank, so the ECU wouldn't confuse the results. In theory, you could do this on a rotary.

RX what?

Woah now, never said not producing any drag. However I see your point.

Karack

the direct injection i saw on their test model on the stand was not what an engineer would call "direct injection" which is placed in the combustion chamber, not the intake stroke. i suppose it might still work but the injector placement still mixes with the open port before it closes all the way, so i wouldn't assume it would work that well.

RX what?

Not having piston bores makes it more challenging to spray into the "combustion chamber". On a piston engine the DI timing changes depending on conditions so it might spray at TDC or during the intake stroke. If mazda wanted to follow suit, it would need an injector for the intake stoke, and one at compression phase, say between the plugs.

Still, its an improvement over spraying that almost 90 degree bend in the lim.