Has cycle reduction or rotor deactivation never been tried?
02-03-17, 03:36 AM
#1
Full Member
Thread Starter
Join Date: Jun 2006
Location: Orlando, FL
Posts: 155
Likes: 0
Received 0 Likes
on
0 Posts
Has cycle reduction or rotor deactivation never been tried?
I've heard rotor deactivation mentioned some, and there are some obvious problems, but it hasn't been tried?
It won't exactly work for a 2 rotor, because the even combustion pressures of both rotors going off balances them out.
Yes, 1 rotor rotary engines are very balanced, but a 2 rotor running just on one rotor is not. (though, it's not quite as bad as running on one rotor because one of them lost compression, as that increases pumping losses. A lot of people wrongly seem to think that rotor deactivation doesn't work because of how weak RX-7s or RX-8s are when one rotor isn't working. That's like comparing cylinder deactivation to blowing a piston in a piston engine car, almost, however...)
But.. there are 3 rotors. Has no one tried it on them? Particularly an NA one, with individual intake runners?
There is the problem of increased EGR, but in theory it should pretty much work "out of the box" simply with the proper ignition and fuel mapping to shut off the front and rear rotor.
Running on the middle rotor only, and the engine otherwise being fine, I would think would show gains if you run that one rotor at 75% load at say 3000RPM or so versus all 3 of them at 25% load. There would be problems, solvable ones in an engine made to run with "rotor deactivation", but I think it'd show noticeably improved break-specific-fuel-consumption(BSFC).
Well, regardless, I've never seen anyone mention something else, which to me seems like it'd work even better:
Cycle reduction.
Each rotor goes through 3 complete intake-compression-ignition-exhaust cycles per revolution.
Why would you not simply do a cycle for every other face? Or every 4th? Or 5th?
You wouldn't do every 3rd, I don't think, since that'd be firing on the same rotor face every time and heat unevenly.
Now, I did the math, too.
See in my head, I figured in an RX-8 you're at about 20-25% load while just cruising at 60MPH. It takes very little HP to go at that speed compared to the much higher amounts the engine can make at that given RPM.
And in my head I was figuring, well at that load, aren't rotaries less efficient comparably to piston engines? Their BSFC is actually fairly decent at high load. I figured it'd probably be in the 1.1-1.3 lb/bhp/hr realm @ 25% load. That's bad.
And people have said they get 28mpg cruising at 60MPH in their RX-8, that that's a very efficient point for them. @2900 RPM or, I believe.
So lets get to the actual math.
There is one variable I'm missing, and that's how much hp the RX-8 actually uses to travel 60mph. But I think 12.5hp is a reasonable guess. Cars usually need around 10-20hp but it varies by drag, rolling resistance of tires, drivetrain losses, and so on.
Now 28mpg at 60mph means 13.24lb of fuel used (6.183lb per US Gallon, for reference). That's straight forward, 1.059 lb/bhp/hr. A little better than I thought, assuming 12.5hp is the number. Either way, that's its usage at 25% load.
But, though I don't have the actual number, I believe the RX-8 would be at around 0.40 lb/bhp/hr at that RPM. I know the FD is about 0.45 BSFC @ 3500, and is about 0.05 higher at peak than the Renesis is at its 0.60 peak. So 0.40 seems like a reasonable number to figure if it was at 100% load @ 2900RPM instead.
1.059 @ 25% load is a 265% increase in fuel usage over 0.40 @ WOT. That means if the engine could instead operate at 100% load at 60MPH, ie if it was a 0.325L engine instead of 1.3L, you would get 74mpg instead of 28mpg.
No, really. 74mpg from a rotary engine. This is why they really make sense as a "range-extender" for EV/hybrids, because they are reasonably efficient at full load and constant RPM. It's partial load and variable RPMs that sink their efficiency.
It is as simple as that, if it were downsized. There's not really any unaccounted for variables that would greatly reduce it. It is totally feasible to make a 0.325L rotary engine that makes 12hp at 0.40 BSFC, thus that is the gas mileage you would get.
The problem however is if you ran a 0.325L engine in an RX-8, it would take forever just to get up to 60MPH to begin with, and that would roughly be its speed limit.
There comes the "cycle reduction" I was thinking about.
If you instead only inject fuel and do ignition every 4th cycle (every 1.33 turns of each rotor) at 100% load instead of 25% load, you increase the efficiency of the combustion by 260%.
Now, it's not totally that simple, as there are losses involved in doing combustion every quarter cycle in a 1.3L engine compared to every cycle in a 0.325L engine. But how much? How big are those losses that cut into that 260% increase? Even if they are 100% total reductions overall, you're still talking a 60% increase in fuel economy (which is in my head, what I figure it'd end up being around. That would give the RX-8 similar fuel economy to a Mazda3)
There's obviously going to be pumping losses induced, but how much? It shouldn't be more than the huge gains. (Think of flooring your RX-7 WOT to 3000 rpm, and letting off the throttle. It revs much faster than pumping losses and friction drop it back down. This is part of why I feel a rotary will benefit more from "cycle reduction" than piston engines do from "cylinder deactivation")
There will of course be a reduction in the effect of exhaust scavenging in the manifold, too, but still, how much would these cut into the big differences in theoretical BSFC?
The pumping losses should be somewhat naturally alleviated on a single rotor by the exhaust gases from combustion traveling out the exhaust helping to pull out the air pumped by the next cycle.
They can also be further alleviated by having a simple EGR system that loops a small amount of exhaust gas back from the exhaust port to a small port on the opposite side of the end housings, so that still expanding exhaust gas is sucked back in and further expands on its way out the side exhaust ports once again. You would ideally probably want an EGR system to loop back into the intake, as well, but... those are things that come in with perfecting the idea, whereas I believe you'd see huge gains with a fairly simple test.
Now to me, this frankly seems very "duh".
It'd even be easy to test.
You'd just need to take a Renesis and make it a single rotor (two rotor is more complicated, but doable and has some problems with EGR and fuel-air looping back through the intake manifold between housings to sold) with a custom intake and exhaust manifold to match the two sets instead of 3 sets of ports, and a custom e-shaft, then custom map to make it skip cycles.
On an engine dyno, it should be very easy to measure the increased BSFC you can get at cruising speeds and light acceleration. It'd be very easy to compare how efficient doing 50% load on all cycles is versus 100% load on every-other cycle, or doing 25% load on all cycles versus 100% load on every forth cycle.
But.. no one has done it and I don't have the money to build a single rotor Renesis and put it on an engine dyno. Am I missing something?
It won't exactly work for a 2 rotor, because the even combustion pressures of both rotors going off balances them out.
Yes, 1 rotor rotary engines are very balanced, but a 2 rotor running just on one rotor is not. (though, it's not quite as bad as running on one rotor because one of them lost compression, as that increases pumping losses. A lot of people wrongly seem to think that rotor deactivation doesn't work because of how weak RX-7s or RX-8s are when one rotor isn't working. That's like comparing cylinder deactivation to blowing a piston in a piston engine car, almost, however...)
But.. there are 3 rotors. Has no one tried it on them? Particularly an NA one, with individual intake runners?
There is the problem of increased EGR, but in theory it should pretty much work "out of the box" simply with the proper ignition and fuel mapping to shut off the front and rear rotor.
Running on the middle rotor only, and the engine otherwise being fine, I would think would show gains if you run that one rotor at 75% load at say 3000RPM or so versus all 3 of them at 25% load. There would be problems, solvable ones in an engine made to run with "rotor deactivation", but I think it'd show noticeably improved break-specific-fuel-consumption(BSFC).
Well, regardless, I've never seen anyone mention something else, which to me seems like it'd work even better:
Cycle reduction.
Each rotor goes through 3 complete intake-compression-ignition-exhaust cycles per revolution.
Why would you not simply do a cycle for every other face? Or every 4th? Or 5th?
You wouldn't do every 3rd, I don't think, since that'd be firing on the same rotor face every time and heat unevenly.
Now, I did the math, too.
See in my head, I figured in an RX-8 you're at about 20-25% load while just cruising at 60MPH. It takes very little HP to go at that speed compared to the much higher amounts the engine can make at that given RPM.
And in my head I was figuring, well at that load, aren't rotaries less efficient comparably to piston engines? Their BSFC is actually fairly decent at high load. I figured it'd probably be in the 1.1-1.3 lb/bhp/hr realm @ 25% load. That's bad.
And people have said they get 28mpg cruising at 60MPH in their RX-8, that that's a very efficient point for them. @2900 RPM or, I believe.
So lets get to the actual math.
There is one variable I'm missing, and that's how much hp the RX-8 actually uses to travel 60mph. But I think 12.5hp is a reasonable guess. Cars usually need around 10-20hp but it varies by drag, rolling resistance of tires, drivetrain losses, and so on.
Now 28mpg at 60mph means 13.24lb of fuel used (6.183lb per US Gallon, for reference). That's straight forward, 1.059 lb/bhp/hr. A little better than I thought, assuming 12.5hp is the number. Either way, that's its usage at 25% load.
But, though I don't have the actual number, I believe the RX-8 would be at around 0.40 lb/bhp/hr at that RPM. I know the FD is about 0.45 BSFC @ 3500, and is about 0.05 higher at peak than the Renesis is at its 0.60 peak. So 0.40 seems like a reasonable number to figure if it was at 100% load @ 2900RPM instead.
1.059 @ 25% load is a 265% increase in fuel usage over 0.40 @ WOT. That means if the engine could instead operate at 100% load at 60MPH, ie if it was a 0.325L engine instead of 1.3L, you would get 74mpg instead of 28mpg.
No, really. 74mpg from a rotary engine. This is why they really make sense as a "range-extender" for EV/hybrids, because they are reasonably efficient at full load and constant RPM. It's partial load and variable RPMs that sink their efficiency.
It is as simple as that, if it were downsized. There's not really any unaccounted for variables that would greatly reduce it. It is totally feasible to make a 0.325L rotary engine that makes 12hp at 0.40 BSFC, thus that is the gas mileage you would get.
The problem however is if you ran a 0.325L engine in an RX-8, it would take forever just to get up to 60MPH to begin with, and that would roughly be its speed limit.
There comes the "cycle reduction" I was thinking about.
If you instead only inject fuel and do ignition every 4th cycle (every 1.33 turns of each rotor) at 100% load instead of 25% load, you increase the efficiency of the combustion by 260%.
Now, it's not totally that simple, as there are losses involved in doing combustion every quarter cycle in a 1.3L engine compared to every cycle in a 0.325L engine. But how much? How big are those losses that cut into that 260% increase? Even if they are 100% total reductions overall, you're still talking a 60% increase in fuel economy (which is in my head, what I figure it'd end up being around. That would give the RX-8 similar fuel economy to a Mazda3)
There's obviously going to be pumping losses induced, but how much? It shouldn't be more than the huge gains. (Think of flooring your RX-7 WOT to 3000 rpm, and letting off the throttle. It revs much faster than pumping losses and friction drop it back down. This is part of why I feel a rotary will benefit more from "cycle reduction" than piston engines do from "cylinder deactivation")
There will of course be a reduction in the effect of exhaust scavenging in the manifold, too, but still, how much would these cut into the big differences in theoretical BSFC?
The pumping losses should be somewhat naturally alleviated on a single rotor by the exhaust gases from combustion traveling out the exhaust helping to pull out the air pumped by the next cycle.
They can also be further alleviated by having a simple EGR system that loops a small amount of exhaust gas back from the exhaust port to a small port on the opposite side of the end housings, so that still expanding exhaust gas is sucked back in and further expands on its way out the side exhaust ports once again. You would ideally probably want an EGR system to loop back into the intake, as well, but... those are things that come in with perfecting the idea, whereas I believe you'd see huge gains with a fairly simple test.
Now to me, this frankly seems very "duh".
It'd even be easy to test.
You'd just need to take a Renesis and make it a single rotor (two rotor is more complicated, but doable and has some problems with EGR and fuel-air looping back through the intake manifold between housings to sold) with a custom intake and exhaust manifold to match the two sets instead of 3 sets of ports, and a custom e-shaft, then custom map to make it skip cycles.
On an engine dyno, it should be very easy to measure the increased BSFC you can get at cruising speeds and light acceleration. It'd be very easy to compare how efficient doing 50% load on all cycles is versus 100% load on every-other cycle, or doing 25% load on all cycles versus 100% load on every forth cycle.
But.. no one has done it and I don't have the money to build a single rotor Renesis and put it on an engine dyno. Am I missing something?
Last edited by zaque; 02-03-17 at 03:43 AM.
03-29-17, 03:39 PM
#3
Yes, as arghx says-
The Norton 2 rotor motorcycle actually interrupted the ignition on one rotor at idle/low load to put more load on the other rotor to raise the throttle opening on the operational rotor so that the effects of exhaust dilution from high overlap peripheral ports was diminished for a smoother idle.
The Norton 2 rotor motorcycle actually interrupted the ignition on one rotor at idle/low load to put more load on the other rotor to raise the throttle opening on the operational rotor so that the effects of exhaust dilution from high overlap peripheral ports was diminished for a smoother idle.