Fixing the Rotary Engine

Unreliable this, apex seal that.
We know, it's crazy, but maybe it's worth a shot. An old thread written by Howard Coleman takes an extensive look into the real potential of the Rotary engine, and you may just reconsider what you think about them.

---

Here it is in all its glory:

Making The Case For The <Rotary> Powered FD

“going to the dark side… pursuing other options…“ we see it every day on the Board and we all know it is code for going piston. 

The trend is understandable given the amount of blown rotary motors. 

Plan A, remaining unmodded, becomes more unrealistic as our 13+ year FDs age and the ravages of cumulative thermal meltdown work on the fragile turbo control system and motor.

Plan B, modding for either longevity or more power often misses the mark for lack of a crucial subsystem.

Before we get into the substance of this post let me say, clearly, that if you want to switch your rotary out for an alternative power plant I have zero problem with that. Permit me to state the obvious: it is your car, do what you want with it. Some of my best friends own transplants. i currently have my nephew’s GTO powered FD in my garage. It drives well. Most transplants are born out of frustration with continued FD rotary engine problems.

I wish to make the case for retaining the rotary. This is not to be confused with a diatribe against transplants. Further, it would be great if the thread continues this direction. 

How'd all the rotary reliability issues come to pass and are they really fixable?



Back in 83-89 when i was racing rotaries versus factory backed piston engines in SCCA GT3 class one of the most significant aspects of the rotary was that it was bullet proof. The $20,000 four valve Toyota engines that were the primary competition were running cams that were so radical they had about 4 thou piston to valve clearance so as to make enough hp to run w the rotary. Over-rev 500 rpm on a shift and you had a bushel basket of shrapnel for your $20,000. 

My Roger Mandeville or Daryl Drummond motors ran from 7000 minimum to 10,000 max for the entire 35 minute race. i remember at the 83 Runoffs at Road Atlanta asking Roger after qualifying if he thought we should throw in a fresh motor. The motor in the car had run 12 National races, plus two days of practice and two days of Runoffs qualifying... Roger asked if the motor hot started O K and i answered in the affirmative. He laughed and said that if we liked working we could change it but it wouldn't add another hp. 

Try that w a piston engine.... the valve stems would be loosey goosey leaking oil into the combustion chamber, the valve springs would be shot, the rod bolts would be stretching etc etc.

I ran 72 National races, 5 Road Atlanta Runoffs and had one blown motor. 

The circumstance of my one misfortune was after the first day of qualifying at the Runoffs I was fourth. I asked Daryl if we might put one more tooth (Speedway Eng quick change rear end) in the overall gear ratio for Q2. We were pulling 10,000 at the bottom of the hill after the long back straight… the change would add an additional 400 rpm in 5th. Daryl said it would be 50/50 the motor would live as it had been pulling hard from the keyhole. I gave it a shot and the motor lost a seal at 10,400. Unlike a piston motor, most of it was salvageable. 

Even when wounded, racing rotaries just keep on humming. i remember seeing Chris Dembs at the (approx) 82 Runoffs pouring oil down his Weber carb to get his rotary started and he blew off a very strong field... 

Perhaps the most demonstrative example of rotary reliability was in the 24 Hours of Daytona. During the 80s Mazda racked up a huge number of wins in GTU and GTO over Porsche. 

At the end of the 24 hours all the Mazda racers would make a point of bringing their cars into the pits and letting the engines idle while they leisurely extricated themselves. The engines (huge bridgeports or peripherals) would just brap brap away like they had been on a cruise. 

Bullet Proof.

The piston engine cars still running after 24 hours would idle at about 5000 rpm. They were total junk. I use the J word literally. After a 24 hour race most piston engines are torn down and almost all moving parts are discarded or sold for pennies down the food chain. 

Rotaries ruled the 24 Hours in the 80s. All from a shortblock that is approx 13 inches by 13 inches by 13 inches... w no rods to bend, no valves to crunch, no cams to screw with no windage problems in the crankcase....

So what happened to the FD rotary?

Turbos happened.

A turbo is exactly what the rotary needs to take advantage of the motor’s immense breathing ability. While the rotary doesn’t make as much hp as a piston engine per given unit of airflow it can flow much more air and that’s where the turbo enters the picture.

Mazda threw some air at the FD with a couple of turbos. 

More air, more power more heat... and here is where our journey will take two roads... 

The OEM FD turbo system being Road One and the aftermarket single setup being Road Two.

The OEM Mazda turbo system is a thermal disaster. In an effort to build early spool/torque Mazda created a manifold system that bolts 22 pounds of cast iron, one of the most heat retentive materials, to the aluminum rotor housings. Aluminum is one of the most heat receptive materials. 

The rotary runs around 200-400 degrees F more EGT than a piston engine. This heat is channeled into a huge hunk of cast iron that is highly restrictive (the two turbines flow exhaust directly at each other?) and the whole deal ends up as a very effective heat retention device bolted to the side of the motor.

Then there’s the turbo control system… 72 rubber hoses in an oven. Tick, tick, tick.

I can just see the design engineer feeling all warm and fuzzy looking at the Centrino-like web of hoses on his CAD machine. Maybe he wouldn’t have felt so satisfied had he considered that instead of copper conductors the control paths on his diagram would be composed of rubber hoses that would need to function and maintain integrity in oven-like conditions. 

There's other aspects such as the fact that the turbo compressors cavitate above a wimpy 14 psi superheating the charge air . Raise the boost from that point and you get more pressure and less oxygen into the motor. Lose- lose.

But hey, Penske won the 71 Trans-Am in a Javelin… you can make anything work if you throw enough at it. So we do silicone hoses, pre-cat elimination, the down pipe and settle for modest hp increases and cross our fingers. All due to the turbo system. 

Road One is whistling past the graveyard.

It doesn't have to be that way.

Turbocharging and melting your rotary do not go necessarily hand in hand.

Road Two

A properly fixtured and tuned single turbo system can make lots more power with LESS ENGINE STRESS.

I designed my own single turbo system. Actually, it features a single (TO4) turbo for each rotor. Two rotors, two turbos, two three inch downpipes. One of my design objectives was to increase the heat removal from the engine by decreasing exhaust back pressure.



My twins run stage 5 turbine wheels that are 10.816 square inches in area versus the largest T4 Q wheel at 6.646 square inches. How does it work?

I log/monitor (EBP) exhaust back pressure. At 20 psi boost my EBP is 21 PSI! That's very close to F1 turbo standards. I have run my twins for 5 years.

The engine has been in my car for four years, 14,000 miles, has logged over 100 2000-8000 4th gear dyno pulls and still shows 17 inches of vacuum at idle. On my Sep 28 dyno run we raised the boost to 20 psi and made 498 rwhp. All with a very conservative tune. I expect to make 630 at 27 psi. 

Most of what i have can be obtained with any well designed single turbo system. 

For instance, and this is not inconsequential, a single setup replaces the 22 pounds of heat soaking cast iron with 5 pounds of heat rejecting stainless steel flanges. a huge thermal help for your motor. 

In addition to greatly reducing the heat soaking mass consider the flow enhancement. We all have seen pictures of the major cracks in the OEM manifold. These cracks are caused by poor flow which creates EBP and heat build up. Cast iron never cracks when it is able to efficiently pass heat. 

The poor flow of the OEM manifold is especially important as the rotary needs exhaust flow much more than the piston engine. 

The typical single manifold efficiently transfers exhaust energy to the turbine wheel and directly out the exhaust. Less EBP, less heat retained in the engine. Win win. An aftermarket single turbo system, Plan B, needs to be on your shopping list.

Of course a single turbosystem, in itself does not solve the turbo’d rotary’s reliability issues. Many of the transplanted FDs were built after the owner had gone the Plan B route. 

Here’s where my story gets really interesting…

A Two-Fer.

Howsabout we solve the reliability problem and deliver racegas horsepower 24-7. 

A month or so ago I received a PM and a comment within remained with me…. “Tuned for 500 rwhp but driving around with 300 and something. Racegas on the dyno, but 93 pump on the street. Typical rotary.”

If you stroll thru our board’s archives you will find an almost unending amount of threads where XXX made 500+ RWHP on the dyno. If you read the details you will find the “R” word. Racegas.

Racegas means you really are driving a major discount-to dyno hp FD on the street. “But look at my dyno sheet!”

It doesn’t have to be that way. You can have all the hp you made on the dyno all the time! And you can have enhanced reliability to the point you don’t have to be looking at alternative engine projects.

500+ rwhp on 91-93 octane. All the time with greatly improved reliability. 

After 4 years of nothing but happy thrashing my curiosity finally won out so a week ago I yanked my perfectly running motor to see what was going on inside. 

It looked and spec’d pretty much as it had when I built it four years ago. Zero carbon near any of the seal grooves. Side seal clearance, an under-appreciated but very important engine metric, had increased but one thousandth! 

I was talking to my go-to guy about my motor and just for fun asked him to guess my side seal clearance. He knew my motor was in great shape and guessed 5 thousandths. My side seals were a tight 3 thousandths! This guess comes from someone who for the last 3 years has made within a rounding error of 1000 rwhp with his methanol two rotor that he builds and races! 

My engine will receive new springs for the apex, side and oil control rings, new oil ring O rings and new coolant seals and go back together.

How in a world of carbon’d up rotaries does my motor look like it has barely been around the block?

While the engine has been properly supported with many of the correct systems of which you are familiar there is one system that you probably don’t run and it is the key:

Auxiliary Injection.

It really is all about heat with the rotary. 

A piston engine performs best at approx 1320 F pre-turbo EGT while the rotary is happiest at 1600. When you combine the latent rotary heat with a poorly designed (OEM) manifold you get eventual Armageddon. 

Additionally, when you combine a rich AFR in a losing effort to cool the beast with the factory external oil pump that sprays crankcase oil on the apex seals you are generating carbon deposits that eventually fill the gaps that allow apex and side seal movement, (half of the motors I disassemble failed due to the effects of over carbonization.)

Auxiliary Injection (AI) can deliver solutions to 3 FD rotary problems two of which are fatal:

Failure due to heat/knock.
Failure due to over carbonization
Top tick dyno HP on PumpGas 24-7

Here’s what’s going on within your turbo’d rotary… 

A turbo vastly increases the motor’s ability to make power by doubling, or more, the amount of airflow. When combined with a commensurate amount of fuel up goes the torque and hp, and (CCP) Combustion Chamber Pressure. It isn’t far off the mark to equate CCP with the mechanical compression ratio. So as you up airflow/boost you are upping the compression ratio. Alot.

More air, more HP, more CCP, more heat. Even if your AFR is perfect, your timing is perfect as torque increases so does CCP.

More CCP equals more heat.

Gasoline auto-ignites at 495 degrees F, all gasoline, that includes race gas. Auto-ignites means ignites by itself, without the aid of a spark. If the fuel in the motor ignites at the wrong time bad things happen. Generally it is the the apex seals as they are more fragile than anything in a piston engine.

If you combine the higher rotary heat with the more fragile character the of the apex seals you have a bad situation… all because we are creating more CCP/heat with the turbo.

Fix the heat, you fix the problem.

AI FIXES THE HEAT.

AI can be water or alcohol or both. Each injectant has it’s own characteristics but both work. I run alcohol. There are three subsets of alcohol: Ethanol, Methanol and Isopropyl. I run methanol. 

Alcohol auto ignites at 867 F so it adds 350 degrees of knock protection versus gasoline. Alcohol “flashes” (atomizes) almost instantaneously providing huge cooling to the charge air. Alcohol is a fuel of course and specs out around 110 octane. Being that it is a fuel and I run 1400 CC/Min along with the usual 850/1600 injectors (4900 CC/Min) for my pump gas the addition of methanol has dropped my duty cycle from a close to static near 90% on-time to just under a happy 70%. 

Water of course doesn’t burn so it can’t pre-ignite. Water contains significantly more cooling than alcohol but takes alot longer to atomize so much of it‘s cooling takes place in the combustion process. Since water is not a fuel there is a limit to how much you can stuff in the combustion chamber. Water also require more ignition firepower. Most board members have used water as a cooling aid and have not attempted to “tune up” with it. Tuning up would be turning up the boost using the cooling to make the motor happy. Water is of course more easily handled than alcohol and generally is allowed my many sanctioning bodies where alcohol may or may not be.

Water may be the sleeper with a lot of potential. Rice Racing in Australia has developed a water system that injects water so that it is greatly atomized by the turbo compressor. Rice is after reliability and hp with his system. A client recently made 600 rwhp on water and pump gas. 

Both water and alcohol remove prodigious amounts of carbon from your motor. Think steam cleaning.

The function of all AI injectants is to remove heat from the CCP. Remove the heat and you can turn up the boost or just drive around at normal boost with your motor benefiting from coolness.

If you are somewhat familiar with AI you know it is generally set to initiate under boost and that you can set a “ramp’ as to delivery generally based on the amount of boost. A major problem with this type of system is that if you make target boost at 4500 RPM and hold it to redline the injectant delivered will not vary as the boost isn’t varying. 

We all know that your base fuel map does grow towards peak torque and tapers off after that with the decreasing volumetric efficiency and time per compression event. The system only tracks boost so it is not capable of making adjustments independent of boost. It is RPM Dumb.

The determinant as to how much injectant is delivered is the speed of the pump which is governed by your settings. The injectant enters the motor through a nozzle.

That’s where I started two years ago. After thorough research I bought/installed the Alkycontrol system. I ran it for all of 07 including 3 lengthy dyno sessions. It never failed, it was absolutely constant as to both delivery and timing. The motor loved the alcohol. No knock nowhere on pump and 1200 CC/Min of methanol..

The (old school) nozzle-pump system works great on a dyno and for drag racing. 

It is not so great for road racing or aggressive street where you are constantly modulating your throttle. The pump hasn’t a prayer of keeping up. Think autocross in a Hummer.

Consider if you were to power your car with this system running the fuel delivery! Compare the typical computer driven fuel injected OEM system with tuning in milliseconds of injector on-time being able to deliver exactly what you command to a pump speed nozzle (garden hose) system.

Elegant, precise versus stone age.

I bring good news… 21st century AI is now available.

I call it High Definition AI…. HD-AI.

And the best part? 

It is invisible. 

During my last dyno session (Sep 08) we checked injectant rail pressure initially and never thought about the system again. 23 2000-8000+ 4th gear runs and the only thing I did was to check my methanol level. We just tuned until we ran out of wastegate capability at 20 psi.

Invisible, but not to the motor. 

Power FC knock numbers under 9 on pump gas at 20 psi and 498 rwhp. 


And here is the “money” chart featuring 156 discrete cells for you to tune. Unlike the old school AI notice you can call for any amount of delivery at any cell on the chart. You can of course change the X numbers and Y numbers to whatever you wish. This is pure magic and the answer to the turbo rotaries heat problems.

---

The module on the left makes it all happen. the module on the right is the FJO wideband. they interface so if there is a problem with AI delivery and the motor goes lean the units cut boost or spark etc.

Looking forward if you fry your rotary it is your fault because with HD-AI it doesn’t have to happen.

Currently 3 companies, to my knowledge offer HD AI:

Aquamist
Coolinmist 
FJO 

I run the FJO system. I did have to add a methanol pressure regulator and a return line to the system to get it to work properly with 100% methanol. Details available in the Auxiliary Injection section in my thread: “Forging fearlessly ahead with FJO.” 

FJO believes this may not be necessary should you be using water as injectant. 


Here’s a how I fixtured my two 700 CC/Min FJO injectors. If you run power steering the lower injector won’t fit so I suggest you run one 1200. Actually i suggest you ditch the power steering... manual since 99.

Rather than get too much into the nuts and bolts in this post I suggest you visit the Auxiliary Injection section. While it may not be a high traffic section it has been around for a few years and you will find lots of info amongst the archives and stickies.

Combine HD AI with the proper support systems and a well built motor and your will be close to untouchable. 

Support systems that I find valuable:

Dual oil coolers. (see my thread “dual oil coolers cheap”)
Aluminum radiator
Efficient intercooler

Instrumentation:
If it isn’t loggable & digital it doesn’t count.
Pre turbo EGTs— see my thread “Dual Digital EGTS”
Fuel pressure
AFRs
Oil temperature- loggable not needed. here's how i did it.

Oil pressure- digital, but loggable not needed
All the readouts from the Power FC/Datalogit

Ignition amp and big coils on the lead plugs, required with “tuned up” water
NGK 6725 10.5 plugs in the lead
High quality sparkplug wires
3 inch minimum straight thru exhaust
Battery relocated to rear pass side compartment
Remove/disable exterior oil pump
Premix 2 cycle oil for less carbonization
Mobil one 5-30
Comp Cams ZDDP anti wear w each oil change
Greddy elbow
Relocate air temp sensor to just after IC
Remove double throttle
GT35R or better
Bite the bullet and buy an Exedy twin cerametallic clutch (you’ll need it)
Ditto the 4 wheel RacingBrake package 

All this adds up to lots of $s.

Many of us have already spent most of the money to get (almost) there. Many have completed a 90 yard drive, are on the 5 yard line. At this point no points have been scored. AI-HD is the 5 yards and payoff. Do it. 

The rotary is a truly awesome beast. As of December 08 it is a beast in a thermal cage. HD-AI unlocks the door to the cage. Do HD-AI right and leave all the turbo’d rotary reliability issues in your rear view mirror along with any car you meet on the street or track. 

I am there.

Howard Coleman
They're not unreliable, unless you suck at life/modify them improperly.

http://www.rx7club.com/3rd-generation-specific-1993-2002-16/making-case-rotary-powered-fd-fix-806104

---