Displacement on demand
 

So today I was waiting for some parts, so I decided to test a theory that I have had for a while.

My theory is that if I am pulling 45 kPA of vacuum in my intake manifold while driving down the road I either need new gearing or some other technique for increasing the load on the engine (so that less energy is wasted pumping air across the throttle plates). The engine only pulls 71 kPA of vacuum to stay at 2,700 RPM (cruising speed).

Due to the complete lack of taller gearing for my transmission and/or differential. I opted to test plan B - Displacement on Demand (DOD).

I did my before testing on the same stretch of road about 2 hours earlier in the day and datalogged the trip to compute fuel economy.

My test setup was somewhat simple - I simply turned the idle up to 1600 rpm and then unplugged the power wires to the coil packs for the front rotor. This resulted in an immediate drop to about 900 RPM idle (which was fine).

My testing began at this point. I put my gauge page up on my computer, plugged in the inverter and went for a drive.

After about 10 miles on the interstate (+6 miles to get there), I decided to switch which rotor I was running on, so I pulled over, plugged in the front rotor coil packs and unplugged the rear rotor coil packs. I drove the 10 miles back home on the interstate and decided that all was good, so I plugged the rear rotor back in and drove the 6 miles home.

I had two problems: The first was in the driveway while adjusting the idle - the engine stalled on the first try (rotaries don't start as easily on one rotor as they do on 2). The second was that I was idling at a stop sign and my electric fan kicked on, causing the engine to stall (equally annoying to get running again).

Obviously, for this test I got terrible fuel economy, because I was still pumping fuel through the unfiring rotor, but the numbers listed below are indicative of instantaneous fuel economy at steady-state as calculated from injector energized duration (Assuming that the fuel was not flowing to the inactive rotor).

Baseline Average:
Traveling East: 35.3 mpg instantaneous
Traveling West: 29.0 mpg instantaneous

With DOD:
Traveling East: 45.2 mpg instantaneous (ran about 70 mph consistently)
Traveling West: 37 mpg instantaneous (I lost speed on a couple hills with the headwind)

My vacuum levels were:
Baseline Average:
Traveling East: -53 kPA Average
Traveling West: -38 kPA Average

With DOD:
Traveling East: -25 kPA Average
Traveling West -5 kPA Average

Now I am thinking about building an external controller that switches between the rotor at approximately a 30 second per rotor rate with a controlled transition. DOD mode would be activated by the throttle being below a certain level for 15 seconds or so. I would need some idle logic too (otherwise the e-fan kills the engine).

What do you all think?

The problem with D.O.D. engines is the nasty harmonics it induces due to the "misfires".
I predict your engine is going to eat itself by the end of the year.
More accurately, it's going to kill all it's bearings and - in extreme - the e-shaft too.


-Ted

wow... what if you produced a logic that literally just cut out half the injector pulses, alternating each rotor and half the spark (hopefully correlating ;-) )

interesting idea. even if you made a logic circuit to cut out certain injection pulses, i think you would still have the vibration problems reted mentioned. but it would be a cool experiment to find out the effects

as far as economy mods, the other thing I want to try is modifying the throttle body. If you look at the way the plates work, you can barely open the primary before the secondary plates start cracking open. I think there are some mpg to be gained by keeping the secondary plates closed like cruising

I'm going to be frank here. I think the gains in MPG to be had are going to come in the transmission, differential, and wheels. Increasing the transmission and differential alone will yield a large percentage of the MPG increase. Thinner wheels (reducing rolling resistance) will also increase your gas savings quite largely.

The engine itself could save on MPG by cheating it a little bit. Mixing of various fuels (IE LPG) to decrease the required fuel from the normal fuel source, but that's just ignoring the other fuel costs.

For increasing engine efficiency look at the intake itself. If you lower the air intake via a restrictor you lower the required fuel. This of course comes at a cost of a less horsepower, but you aren't going for hp when you're thinking of MPG.

To regain the hp you can install a bypass which will allow the addition of more air when the TPS is increased beyond a certain rate. This should remedy the lack of power, but again at the cost of MPG. I personally think that's the best bet beyond running on one rotor. Additional saving could be had from ensuring the intake adheres to the hemholtz equation throughout the RPM band. Thereby increasing volumetric efficiency just by running the engine.

two things;

1. PJ showed me this years and years ago, but it might work, http://yarchive.net/car/air_induction.html

2. second i've noticed tuning a few cars that going from say 12:1 afr at cruise to 15:1 at cruise "looses power", my gsl-se particularly used to do this when the sensor in the radiator switched it into closed loop, the car would feel torquey, and then when it switched to closed loop, "loose power".

but if you think about it backwards, you say it takes X amount of fuel to go 25mph, and @12:1 afr, you have to open the throttle Y percent. when you switch to 15:1 afr, the amount of fuel you need to go 25mph is THE SAME, but the amount of air is bigger so you need to open the accelerator MORE. y+(15:1/12:1)

i hope that makes sense, its easy just to think of the gas pedal as a linear power delivery lever, but with a carb particularly, you can get fuel and air to be separately metered (weather you want it or not!).

a lot of new cars in europe have a "stratified charge" mode, where i think the throttle opens, and its fuel controlled like a diesel. i don't know exactly how you'd implement it on a rotary...

mike

The same way the NSU people did it? :lol:

i had an NSU spider and the ports are really weird!

its a 1 rotor P port, with similar dimensions to a 10A rotary actually.

it actually has 2 intake ports, one TEENY primary port, that opens late, and closes late, and then its got one bigger port under it.

and a 2 barrel carb bolts right to the rotor housing.

hell this literally is the car right here... http://www.rotarycarclub.com/rotary_...ead.php?t=1670

I agree that it would be nice to get the revs down on the highway, and I wish that there was a good solution (transmission or differential) but I don't know of one. I am still hoping to get a set of 3.73:1 gears made for the rear differential, that change alone should drop 250 rpm on the interstate and increase manifold pressure to -37 kPA average (a significant improvement in thermal efficiency).

As to the intake idea, I already have a restrictor (my throttle plate) and it is the very element that is destroying my fuel economy. My bypass is activated when I open my throttle plates. Think about it this way: Based on rolldown analysis, it requires 12 Hp to go 60 mph in my RX7. At 60 mph, I am running 2400 rpm and -48kPA MAP. The engine is turning 40 revolutions/second - amounting to 1.3X40 or 52L of air per second being pumped against a 48 kPA head (plus exhaust backpressure - which I am ignoring for this calculation). Since pump horsepower is simply W = QP, I can convert the units to m3/s and N/m2 and I get 0.000052 m3/s * 48000 N/m2 = 2.5 kW = 3.5 Hp. This is just the flow work to pump air across the throttle plates, and it amounts to 30% as much power as is required to drive down the road. Intake restrictions (whatever they are) do not really help fuel economy.

To add to the above calculation, let me do some quick math on the thermal efficiency of my engine. at 2400 RPM, the engine puts 100 Lb-ft of torque to the wheels. This is with an injector on-time of 4.8 ms (primaries, per rotor) with no intake vacuum. At 48 kPA, I am running 2.6 ms injection time and putting 26 Lb-Ft of torque to the wheels. This is more than half the fuel and about 1/4 as much torque. (Efficiency of less than 50% WOT efficiency).

This is the big reason that I tried DOD, unfortunately, I have no valves to hold open, so I still do a lot of work moving the air through the second rotor while generating no power with that rotor. The increase in efficiency is the only reason that DOD makes sense.

As to the tires, I am running Bridgestone Ecopia 195/65/R15s (Low rolling resistance). They improved my 30 mph and below fuel economy substantially (about 10 mpg) over the 245/40/R17 Goodyear Eagles that I was running before, but at 70 mph, there is about 1 mpg difference (the overwhelming majority of the drag at 70 mph is aerodynamic). I am running dyno fluids in my transmission and differential, I picked up almost 1 mpg average by using synthetic in my NA drivetrain (so that is still on the table).

Have you seen misfire-related bearing failures Ted?
I didn't even consider this, now you have me a little bit worried.

However, let me list my experiences:
I knew a guy who purchased an FC (this was about 10 years ago when they were $100 each) with a 'blown engine' we got it running (on one rotor) and disconnected the injectors for the dead rotor (to keep from damaging the exhaust) and he drove the car for 5,000+ miles just like that. He loaned the car to a friend who smashed it up a little and decided that he would fix the damage really quickly (headlight doors, etc.) He opened the hood and observed that the secondary injector was unplugged and he thought that he had figured out why the car was so gutless. He took the manifold apart, plugged the primary back in, put it all back together and 'fixed' the problem. The other rotor had apparently started making compression again some time during that 5,000 miles.

The point is that the car ran great on both rotors, he ran it for a while after that and never had any issues. This was with continued operation on the same rotor for 5,000 miles. I never saw this engine apart so I can't attest to the bearing condition, but the engine made good oil pressure and ran well.

I also had a friend who blew a rotor in New York while on his way to Nebraska. He unplugged the injectors to the dead rotor and drove all the way to Nebraska and back on one rotor. I helped him rebuild his engine and the good rotor looked like brand-new (not even any carbon) his bearings had no damage at all, we re-used all of them. In fact, I put the good rotor in a different engine after his blew up a second time.

The bearing problem might be a turbo-only failure (I hope).

I think you're missing my point. I do not believe your throttle body is doing what you think it is doing. The throttle body is altering the flow itself just by being the stream. Reducing the intake diameter or further 'restricting' it will cause an increase in intake velocity (hurrah for area flow rates). This will do two specific things. One it will increase the vacuum on the manifold, and two the intake velocity will speed up. Since we're doing this the ideal fuel combustion will amount to a slightly better ratio. The harmonic tuning will further increase your volumetric efficiency. This in turn means increased 'bang' for your buck.
Hence why a restrictor plate (and not a throttle body) will cause you to increase efficiency.
If you're serious about doing this I wish you the best of luck. I personally feel however you'll find more gains in a properly designed intake manifold, restrictive set up, and harmonic tuning. That said those will only net you so much. Consequently your low rolling resistance wheels aren't really all that great for improved gas mileage but they're probably the best balance for the buck. When I say low rolling resistance I'm talking about solar power car wheels a-la:
http://www.solarquotes.co.za/wp-cont.../solar-car.jpg
Granted those aren't going to work for you, but those will reduce the friction acting on the car dramatically (which technically speaking is a hell of a trade off). As for the aerodynamic effects I completely agree. If you really wanted to improve your MPG at speeds over 60 MPH you'd want to install a couple of these:
http://www.aerospaceweb.org/question.../generator.jpg

The ideal location I would think would be directly in the front of the hood... but that's just a guess. Possibly putting them on the rear roof line might actually prove more beneficial, but both might be tested on the highway for verification of greatest increase.

Drag reduction by use of the vortex generators will have significant effect on the Cd. The idea is that it sort of 'trips' the boundary layer causing it to not necessarily reattach to the surface but follow the contours still. In essence rolling along on bearing of air over the skin of the car. If you don't care too much about looks you might be able to try some artificial 'shark skin' which is used on marine hulls to reduce drag. The effect is going to be similar... but I don't know any place that would sell 'em.

I tried this on a motorcycle engine once. It doesn't work. The reason is that the fuel will not all make it into the chamber on that single event, it will give a very lean stroke when you start injecting and then an even leaner stroke after you stop injecting. I think that 30 seconds is a better option, as it will give the opportunity for the fuel mix in the chamber to stabilize without allowing the chamber temperatures to reach the WOT levels that they would otherwise while running or the cold temperatures while not running.

It is a good idea, I would definitely give it a try if I hadn't already tried it.

You are right, an engine does not make quite as much power at 15:1 as it does at 12:1 (unless you are running a tumble head design, which would actually do better at 15:1). However, the BSFC for the engine is much better at 15:1.

You are correct, and opening the throttle plates increases the engine efficiency even more, so it is a win/win. The typically accepted optimum Lambda is 1.1 (per the Bosch handbook) this works out to about 16.2:1. Believe it or not, the amount of air is more, but the amount of fuel is actually less (considerably). On my engine 12:1 is FILTHY RICH, almost into the black smoke region. Peak power is lean of that for me (about 13.2:1). 12:1 gives better throttle response, so the internal accelerometer senses the jerk of the throttle response and *thinks* that the engine is making more power (I though 12:1 was good until my first dyno session).

That would be cool, as the efficiency would be good and the losses across the throttle plates would not exist. We could never get away with that here (NOx emission regulations kill any attempts at combustion-related fuel efficiency gains).

You are coming at this problem from your aeronautics background. I am attacking this problem using thermodynamics and engine controls experience. Call it a restrictor or throttle (which means narrowing or restriction), it still creates a pressure drop that I have to pump air across to get it to the exhaust. This means that I do more work to get the air that the engine needs for combustion, while decreasing the temperature ratio of the thermal cycle (lowering efficiency).

Higher vacuum in the intake manifold is all bad (think about the flow work to pump air through the engine from that pressure to the exhaust manifold pressure, without even considering the chemical and thermodynamic implications).

Helmholtz tuning could possibly help, but the required intake runner size to get enough air velocity to see any appreciable benefits from Helmholtz would give up way more power than I am willing to give up. If the intake runner can provide even 0.1 Mach at 2400 rpm and 45 kPA pressure drop, They would have a tiny cross-section and would be something like 5 feet long. On top of that Helmholtz would only help for a very limited speed range.

I really should have done a better job of explaining my goals for this plan, I want my cake and I want to eat it too. I want a 250+ Hp rotary that will still give 35 or even 40 mpg.

I wonder how those wheels do on the corners? :)
I really do sound like I am being unrealistic, but I have a Metro if I want to drive for cheap, I want the RX7 to get decent mileage to see if I can. I also want it to retain some semblance of pleasure to drive, otherwise I might as well drive the Metro.

These basically perform the same function as a spoiler and are a very good idea. I might look into this (I really couldn't care much less what the car looks like).

Unfortunately, this sounds like more work than I am willing to go to. Still not a bad idea.

No. Vortex generators do not perform the same task as a spoiler. Spoilers behave like an upside down wing (unless they're non-functional then they're just extra drag). These will literally trip the boundary layer reducing friction of the car as it goes through the air.


Nah, vortex generators aren't hard to implement at all. Just some adhesive tape and you're golden, or you could build 'em yourself a-la:
http://www.zenith.aero/profiles/blog...g-micro-vortex

I am not sure which type of spoiler you are speaking of, but i am referring to the type which reduces the aerodynamic drag on an automobile. A true spoiler is intended solely to improve fuel economy as it reduces the negative pressure behind the vehicle (similarly to a vortex generator. Downforce is generated by a wing (in the opposite direction as on an airplane).

Note that on the FC, the Aero package reduced drag (CD from 0.31 to 0.29) and included a spoiler.

Maybe they do something different on an airplane.

You don't have to take my word for it, please do some research:

This will get you started:
Spoiler (Automotive)

This is the only spoiler I am familiar with. Since no citation is given for this derivation I am hesitant to believe wiki on this.

Interesting discussion, but NoDOHC is right. Even diesel engines without throttle plates are giving lowest BSFC at almost full load. Load is essential.

Very late closing timing of intake ports could push quite large portion of fresh mixture back to intake allowing for more load and effectively make atkinson cycle with uneven compression - expansion, but we know that piston engines doing this are using high static CR, not option for rotary...

The only option for this on a Rotary would be to add a roots-type supercharger. I thought about adding a supercharger and converting my rotary to diesel a while back, but then I realized that the fuel system development alone would break my budget.

I actually gave thought to the late port close while I was doing my 4-port, I was hoping that the later close that the 4-port has over the 6-port would work to my advantage in the mileage department, but I think that whatever I gained there I gave up in 35% intake runner size increase, polished intake runners, swirl-ported irons and polished rotors.

Ted and i both know a guy who built a bridgeported Rx3, that did 230rwhp, and would get high 20's mpg on the freeway. with a carb! and an exhaust quiet enough that it was streetable. 10 more MPG would be tough though...

actually #2 thing i've noticed is that my other car, a 1958 Tr3, gets 35mpg. other british cars of the era, XK120, MGB, etc actually get decent mileage too, especially compared to "new" cars and they do this with carburated, tractor motors.

how? well the Tr2/3's are 2000lbs, with a 2L 4 cylinder. my S4 Fc weighs in @2880. that's a HUGE difference. and in 1952 the Tr2 was faster in the 1/4 mile than anything you could buy from an American car company, so its slow now, but it wasn't at the time.

oh and #3 back to the mazda rotary. we make the 1000mile round trip to sevenstock every year, and its almost the only drive i get that's almost 100% freeway, and having taken every car from a stock 79, to a 20B FC, to an FD, i've noticed that the FD and the SA both had an rpm range where they got better mileage. the SA gets the best mileage (23-25) in the 3500-4000rpm band. which in 4th puts you right around 65, and in 5th is like 80. driving @3000rpms, actually gets WORSE mileage, although not by a lot, and going over 4k, it gets bad quickly.

the FD also did the same thing. the FD pops out of closed loop @3200rpm, and best mileage is at 3199rpms, which is around 80mph. on that trip my stock FD got 23mpg, my friends ported car got 25. 23mpg + 22 gallon tank = you have to pee really bad.

the 3 rotor FC was plagued with ecu troubles, so it got tuned on the back of a trailer for 45 minutes and returned 19.9mpg, which is totally acceptable

i haven't checked the P port yet, i haven't put enough gas or miles on it to tell.

anyways, i hope it shows that i'm on board with better mileage too!
mike

Another data point...

Several Sac to Vegas and back trips netted me 400 miles between gas fill-up's.
1987 Turbo II on a (reprogrammed) stock ECU
16.6 gallon full capacity fuel tank, but filling up at 14 gallons each time.
That makes it 28.5 miles per gallon.
Key is to drive run under the secondary injector crossover point at 3,800RPM, which is about 75mph - 80mph on my car.


-Ted

I also noticed that with a Rtek ecu on my way to DGRR this year.... I would be around 15.0 afr until the secondaries kicked in, and then it dropped to 12.1 afr. I was alone so I couldn't tune the Rtek, but this is something that people with stock ecus are just stuck with

I got 23ish mpg while hovering around that point with probably spending most of my time about that

yes, it seems modded t2's get excellent mileage, because it runs 14.7:1 out of boost, and its easy to stay out of boost...

-mike

Has anyone here experimented with running negative split at low load? What are your observations? Has it allowed you to maintain stable combustion at leaner AFRs?

I noticed quite a difference between +5 deg split and 0 split, about 1-2 mpg gain at 0 split. im running e6k, so i cant do negative split though

Sounds promising. Did you have to pull any leading timing when you reduced the split, or did you just leave leading as it was?

Sorry, I'm a noob to EFI tuning. :)

i just left leading where it was, around 30 deg at cruise

0 split definitely helps high vacuum stability. It also helps when running lean.

I want to justify driving my RX7 and it is hard to do when I get below 30 mpg around town with it and my Metro gives 50 mpg around town. If I could get DOD operating, I think that I could get 40 ish out of the RX7 which would make the difference livable.

RETed almost has be scared off because of his predictions of engine failure.

I don't run 14.7:1 while cruising, I run 16.2:1. I am running about 60 degrees of advance while cruising at -45kPA. You have to light the lean mix a lot earlier as it burns more slowly.

How is the driveability with such lean mixture?

There is interesting discussion over rx8club about negative split and lean burn during cruising but concensus backed by injector duty and actual gas mileage is, that best gas mileage, driveability and overal responsiveness is best with slightly rich mixture 0.92 - 0.93 Lambda. And this is with high CR, no overlap engine.

I agree with you that conventional rotary engine very same like piston engine has highest thermal efficiency at around 1.15 Lambda? But this is measured at WOT and practical parameters mentioned above aren´t investigated.

You should try it:001_005:

My experience with negative split hasn't been very positive.
At idle, it does drop the pulsewidths just about 10%, which coincides with your mileage #'s.
When trying negative split at high vacuum, low load cruising driving, the engine is "unstable" - it's almost like lean surge but slightly different resonance...
I didn't further the experiment because, IMO, it's a waste of time trying to wring out that last 1% of gas mileage on a rotary engine.
I drive and build my rotary engines for power - not fuel economy.
Building and tuning an RX-7 to pump out 400hp and getting 25mpg is a perfectly fine compromise in my book.
I have a daily driver other than the RX-7 for fuel economy - that's how I solve the gas mileage issue.


-Ted

Honestly, I have adapted to drive the car with the lean mixture - therefore I consider it to be fine. most of my friends consider the car to be 'undriveable'. As long as I hold the throttle at a consistent position while cruising and don't mind some hesitation if I get into it to pass someone, it drives fine.
I have 9.2:1 polished rotors (originally 9.4:1) and no overlap (or only stock overlap - like 8 degrees).
.93 L gives better throttle response - hands down, but my research has shown the engine to use considerably less fuel at stoic. than at 0.93L and slightly better mileage at 1.1 L than at stoic.
The efficiency difference is higher at partial throttles because the BMEP is lower, thus the throttle must be more open (lower vacuum). This contributes to better efficiency.
I have :)

I have datalogs of the same stretch of road on the same day at the same speed while varying ignition timing, fueling and ignition spilt. I then used Excel to average the injection time and determine the average fuel economy (for that stretch of road) obtained at each setting. I probably drove that stretch of road 30 times in one evening.

during one of the Rx8 recalls, we got to go drive em around with the laptop hooked up. the IDC/ dealership laptop does logs like the haltech e8/e11's. so i got a chance to go drive an Rx8 and watch what it the timing did.

compared to our "normal" haltech maps, Mazda runs more advance at low rpm cruise (36BTDC L @2500 in 6th, which is like 35mph), but hit the gas, and it'll go to negative numbers, so WOT @2500rpm is like -5BTDC L, this is a taller peak than i'm used to.

decel is where it starts negative split, take your foot off the gas, and the leading timing is -5 BTDC L, and the trailing will stay high and go down with rpm.

at idle leading is almost always -5 BTDC L, and the trailing actually moves around, so it can be like 5-15 BTDC. the trailing stays more advanced than the leading until maybe 1200rpm, or if you give it any throttle whatsoever. ANY load, and the negative split is gone.

WOT over 6k its running 30BTDC L and 15 BTDC T

the second part, i also agree, 25mpg is pretty good for a heavy car like the FC. to substantially improve it, the car needs to be lighter, smaller tires etc etc. the fundamentals of the car matter a lot for mileage, my 58 Tr3 gets nearly 30mpg, and it has nothing on it that could be considered technology, but its a 2000lbs car, with skinny tires, and a decently tuned small displacement low revving engine.

i had pauls Rx8 for months, and the BEST tank of gas was like 21 or something, with 19ish being average, the FC beats that. Rx8 = 3200lbs with 225's

-mike

My thought on this was that we need a compromise of lean AFR with highest HP.

A simple way to do this is use a constant of 20% injector duty (in 5th gear, level ground) and then change the AFR to accomplish highest cruise speed.

This method would avoid reaching diminishing returns by going too lean.

Barry