# Mazda Rotary - what are the issues to be overcome?

### Help Support HomeBuiltAirplanes.com:

#### rv7charlie

##### Well-Known Member
Hi thjakits, welcome to the conversation!

Just to address the latter parts of your post, the rotary just doesn't have a lot of torque (power basically comes from rpm in the rotary's case). There's no practical way to get low rpm torque out of a rotary, largely because of the very short 'stroke', for lack of a better term. High torque piston engines tend to have relatively long stroke cranks; something that's almost impossible to do with a Wankel style rotary. Mazda does have a prototype engine (the 16X) that has a bigger eccentric (stroke), as well as all aluminum construction, that rotary flyers have been salivating over for a long time, but there's no real evidence we'll ever see it in production.
I owned a great 'automatic pitch adjustment' prop about 25 years ago (on a Lyc); Bernie Warnke was famous for his 'almost constant speed' wood props back in the 1980s-1990s. But that wouldn't buy you much with a rotary; there's no real power until you get up in the 5k rpm range and above.

I don't recall any emphasis on ceramic side housings, unless you're talking about aluminum with a high tech wear surface. You can buy those now, but they're very pricey, and I don't know anyone who's used them in an a/c designed to be actually used. (There's the time-to-climb record holder, but I doubt that engine ever ran longer than 20 minutes at a time, and probably has fewer than 100 hrs total on the engine (with a couple of rebuilds in the mix for various reasons).

Using automotive engine control (if you mean 'from the car') is a very dicey proposition, because modern car controllers have literally hundreds of thousands of lines of code running their cpu's, most of which ties the engine to the rest of the car in ways that are hard to 'hack' around, and since the code isn't open source, there's no way to know whether we've hacked around every transmission shift point code, limp home mode, etc. People have been killed trying. There are several controller sources still available that have been extensively debugged on 'hot rod' ground vehicles that are much better (safer) options. The big hangup at the moment is the gear drive, now that Tracy is retired.

Belt drives: It's been done; there was a toothed belt version flying successfully on an amphib in the USA northwest for a number of years. But they tend to be bulky, require cooling air, and aren't really any lighter than a properly designed gearbox. Most people just won't trust a V-belt array on a high HP engine in a fast a/c, whether it's justified or not.

Charlie

#### thjakits

##### Well-Known Member
Hi Charlie,
...thanx for the welcome!

b) ..."ACS"-prop from Bernie - that's the one!
b1) ...I didn't intend that one for the slow rotary, but for any Elippse application!

c) ...yes, Aluminum sidehousings (for Peripheral Ports) with a ceramic coating...

d) ...Automotive engine control, no - NOT "taken from any car"!! But aftermarket units that can take as few or many inputs as you wish to apply - I would think something on the lines of Megasquirt or similar! FULLY customer programable...
[I'd be scared using a newer "computerized" car controller - way too many variables to watch out for!! I am "fighting" frequently with my VW-CanBus-controlled vehicle: Windows not closing anymore, indicator lights going berserk, ...stop, shut down, pull the key - count to 10 - start - all well again ....or not!]

e) ...actually - I would NOT use a TOOTHED belt in a redrive - one slip or something getting into the teeth and that belt will disintegrate fairly quickly! I'd use the ribbed belts! on conforming pulleys with a smooth tension roller! The power shouldn't be transmitted by the teeth of the belt (...not a "rubber-chain"), but by the friction forces maintained on the pulleys. Robinson even uses the belts as a friction/sliding/rubbing clutch - usually takes about 10-12 seconds to engage and stop slipping. From there it takes about another minute to fully tension the belts - and usually there will be another 4-10 seconds of retensioning once the whole system is warmend up (usually in flight...) - Obviously on a prop that wouldn't be necessary. Calculate needed friction values tension to that and then see if you got vibrations in the belts (visual) and tension some more to eliminate (of course within manufacturers recommendations). The one thing is to GET the belts for aviation application! Most manufacturers will hang up the phone IF you dare to mentioned anything to do with aviation!!
BUT - Having followed much of Tracy's evolution of his redrive - having followed 2 Hi-Vo-chain drive PSRUs - having followed a Spur-Drive intent - and having a couple of thousand hours on belt-reduction/clutch drive helicopters - I am more convinced than ever about belt-drives!!
[As a note to that - in the 20 years on Robinson products (...Nowaday very little, I am on other products now..) - we had one belt-failure! We were unware that on a brand-new R22 a limiting switch in the clutch/tensioner was dead (from the factory), so it would always go to the very end of the jack-screw - severely over-tensioning the belts!! The engagement time was within paramenters, so it was a seriously tough one to spot! Finally one of the 2 belts snapped (...on the ground while tensioning...luckily the attending mechanic saw it - as the 2nd belt is just fine to transmit ALL the power) - after 234 hrs (or so...) of overtensioned work!!!
Usually the blets last the full TBO of 2200 hrs....

f) A note to the direct drive V8 Long-Ez - iirc, it did get a proper bearing on the drive end, to cope with the prop-loads!
[I think the gentleman had a bearing-housing made to bolt to the engine and an adapter to mount to propshaft to the crankshaft. Axial and Radial forces where taken care of properly. I do not recall how the engine was mounted, but I suppose the aftmount was integrated with that bearing housing....

Cheers,

thjakits

#### rv7charlie

##### Well-Known Member
One clarification on rotary housings. The proper term for the housing that contains the spark plugs, and the 'peripheral' (if installed) intake port and the peripheral exhaust port (in pre-Renesis Mazda rotaries) is properly called the side housing. That housing has always been aluminum in the Mazda engines. I mis-used the term earlier (duh) speaking about the *end housings* and center housings, which touch the *sides* of the rotors. Confusing enough that I get sloppy with the terms sometimes, and I've watched rotaries since before my father bought a new 1974 RX4. Anyway, end and center housings are what are available from aftermarket vendors; the side (rotor) housings have chrome wear surfaces, and people rarely have issues with them in aviation applications, unless there's FOD, or damage from broken apex seals (rare except in boosted engines).

I can kinda see why the V belt system might be chosen for a helicopter, with the need for a clutch in addition to the reduction aspect, and the maintenance-intensive nature and complexity of the tensioning system would be a minor addition in a heli. Given your experience with the Robinson system, you could likely make a V-belt system work if you wanted. But I think you will find it quite a bit heavier than other choices, and very few homebuilders willing to deal with that level of complexity and maintenance intensity in a fixed wing a/c.

Charlie

#### thjakits

##### Well-Known Member
In a PSRU there is not much do it - to pulleys (or I think the correct term is sheaves) and a tensioner - NO clutch needed...
I think weight wise there would be not much difference to any other system - in any case less weight.
Though it is VERY important that the sheaves be positioned sturdy!

thjakits

#### kbd512

##### Member
I'm just thinking out loud here, but feel free to jump all over me and tell me how stupid this would be. I'm just trying to work out what a simplified variant of this existing design might look like. Yes, it's long and rambling, much like my thoughts on this. Feel free to skip over it if that suits you.

A properly designed and tested PSRU seems to be the most significant issue preventing more people from even considering rotary engines. If that's not addressed first, then I don't see any path forward to a functional aircraft engine.

Are there any competent PSRU designers out there who are interested in something like this (as a side project to make extra money or to stave off abject boredom after retirement, not something they're depending on for their next meal- unless you're really into risky business)?

Is this just a matter of throwing enough money at the problem, or is there something uniquely challenging about designing a reliable PSRU for rotary engines (above and beyond whatever challenges typically exist in designing reliable re-drives)?

I have to ask because it seems as if nearly every small company that sells one of these things goes out of business in the span of a few years. Maybe there's just not enough money to be made here, but that doesn't mean the plans for the parts couldn't be distributed through some sort of CAD package that's purchased online and then taken to local machine shops who have the tooling required to make the parts. If suitable PSRU gears and bearings could be sourced from racing engines or transmissions to try to limit the number of aviation-specific parts to perhaps just the shaft and housing, that'd be an added bonus.

Once that's been taken care of, there would seem to be the always-present cooling issue to address.

Using 13b type rotors and shafts, assuming an output somewhere between 150hp and 200hp, is it possible to make an air and oil cooled variant of this engine run reliably for any reasonable weight?

For example, could another 50 pounds of Aluminum, along with several gallons of oil and a fan at the front of the case to blow cooling air over the fins, assure sufficient cooling?

Would a heat pipe system used to evenly distribute heat throughout the block substantially reduce the amount of added mass for cooling?

That all-Aluminum short block that Paul Lamar posted about on his website, or whomever is running it now, weighed 109 pounds. However, the actual weight difference of the much stronger and beefier billet plates over cast iron was only 10 to 15 pounds per plate, so 152 pounds or so instead of 197 pounds for the short block without a water pump. If the final installed weight of a 200hp engine and PSRU combination is coming in around 350 pounds, that's not markedly different than an IO-360. If we "only" had a 150hp air cooled engine with approximate O-320 weight-equivalence, would anyone else be happy with that result?

For people who are concerned with weight, all those extra parts bolted onto the 13b bring the dry weight of the long block up from 247 pounds to around 350 pounds. I think the guy from Mazdatrix told me that an all-aluminum short block would only save about 60 pounds over iron, assuming all of it was cast A356. Then we have to add a 45 to 50 pound PSRU and coolant to that, so weight-wise we're more or less right back where we started. That's remarkably similar to an O-360 making the same power. So... For those people considering a replacement for their old four-banger, they're going to end up with an engine that's nearly the same weight, makes nearly the same power, and is considerably more complicated, not in terms of the mechanical operation of the engine, which is obviously much simpler with the rotary, but in terms of everything else that they have to get working correctly to come up with a functional aircraft engine.

Obviously the rotor housings and side irons would have to be redesigned, but drag racers have already gone to that extreme by completely replacing stock Mazda parts with machined billet parts (and they subsequently replaced pretty much every component in the engine with something that was significantly stronger or otherwise more durable than the stock parts). So, is there anything infeasible about doing this (again, assuming that the primary reason for doing this is to achieve something we can't achieve with stock parts or traditional aviation engines, not in some mis-guided and likely fruitless effort to save a little money)?

Speaking of cost, the list price for a brand new, or at least zero-timed, 180hp O-360 from Vans Aircraft is stated as $27.5k. The list price for the 260hp IO-540 is$47.7k. Neither of those prices are anywhere near $120k. I'm sure that more powerful engines cost even more, maybe a lot more, but how many home builders not operating cross-country machines have 260hp+ engines under their cowlings? I only fly myself, never intend to fly anyone but myself in an experimental aircraft, and have never met anyone with any intention of ever becoming a glorified air taxi service except on someone else's dime, so I'm guess that there's not that many. Dick doesn't typically mess around when it comes to his money, so I'm taking him and his people at their word. Viking sells their own 180hp engine for only$17k if $27k is still too rich for your blood and their engine has every bell, whistle, and doo-dad imaginable for people who like that stuff. If people really don't like what either of them are selling, then it's not apparent to me from the sheer numbers sold. With a 2,000 hr TBO, which it might make if we treat it well- but let's be honest with ourselves and assume we're gonna thrash on that thing a bit / skip some oil changes / shorten warm-up periods / etc, that still works out to$13.75 per flight hour. For an extra $10.10 more, you can get your own six shooter if you really love paying for gas. The IO-360 I fly behind drinks about 12 gallons per hour when I'm just tooling around. I'm sure there are professional pilots out there who fly more than that in a year, just for giggles, but for we weekend warriors who maybe spend 4 hours flying on the weekend, that works out to around 500 weekends, or almost 10 years, of fun and adventure before we have to pay the piper for having our fun and overhaul that thing. Maybe one or two jugs doesn't make TBO, but we're still not talking about a lot of cost if we're repairing it ourselves.$2.5k for 2 jugs. Serviceable old pistons are $25 to$50, so... We wouldn't be saving a dime on this project, so that's not a good reason to fly a rotary.

#### narfi

##### Well-Known Member
Log Member
Are you people talking rotaries like a Gnome or a reciprocating (and rotating) Reuleaux triangle in an epitrochoid housing like a Wankel?

From the title of the thread I assume only wankel, specificly those taken from mazda cars.

#### rv7charlie

##### Well-Known Member
Aero,

My experience running Lycs is that up to 8.5-1 compression (most 160 HP 320s, 180 HP 360s), you can run premium mogas. But without 'extraordinary measures', you must use ethanol-free fuel to avoid damage, primarily to the 'soft parts' in older mechanical fuel pumps, and the same for many gascolators. E-free *premium* mogas is available, but often very hard to find, and a lot more expensive than commonly available premium E-gas.

I've flown a couple of ~400 mile cross countries in my 160hp Lyc powered RV-4, with Tracy Crook's Mazda Renesis powered RV-4. In both cases, I was burning E-free premium and he was burning 87 octane E-mogas. Unlike many Lyc drivers, I aggressively lean my Lyc. On both these flights, Tracy burned slightly less than 10% more fuel than my Lyc (same flight profiles). But. Remember that he could burn 87 octane E-gas, while I was forced to burn premium E-free gas. My cost per gallon was around 50% higher than his cost.

Charlie

Last edited:

#### Lendo

##### Well-Known Member
A lot of discussion on Aluminium end and intermediate housings. Does anyone recall the light steel housings developed by Powersport. Using steel immediately eliminates the issues with Aluminium substrate failure. Mistral found out to their sorrow, they tried every surface treatment known to man and everything failed. They may have ended up with an Aluminium and Ceramic particle mix, however I'm only speculating. You see the Aluminium has to have a mix of particles to a depth to support the high side loads. Detonation Gun treatment does the trick, but is quite expensive and there is the final grinding cost as well.

The Light Steel housings Powersport produced were Oven Brazed and also very expensive to produce, hence the reason they never considered them commercially viable. I am led to believe they were only 1 lb. heaver than Al. per housing - I for one can see benefits here.

Technology has become available, like Laser Sintering and other 3D printing of Sintered (powered) metals, but the technologies aren't so plentiful to get prices down - at this point in time anyway.

So the evidence is people are working hard to utilize those technologies when it's feasible to do so. I believe that the size of the housings may be an issue with the current crop of machines of this technology - not to mention of their cost of equipment. When there is sufficient technology out there to create competition, prices will come down.

Billrsv4, asked for responses from people interested in a well designed PSRU, he's a well respected mechanical Engineer, I'm not sure what responses he received from that question on this site, but I do know he's still refining the design, (mainly for best Ratios) and finding quality parts suppliers.

Naturally everyone wants everything now, they want it light and they want it as cheap as possible - can't say as I blame you!
George

#### kbd512

##### Member
Lendo,

The Aluminum / Alumina Oxide CMC alloys that Gamma Alloys sells are stronger than 6061 and 7075 at elevated temperatures (going off of memory, somewhere between 3 and 4 times stronger, though it still melts near the same temperature as ordinary Aluminum because it's still mostly just pure Aluminum), which is one of the reasons that I suggested that anyone serious about an all-Aluminum 13b talk to them about obtaining some of their CMC materials. I believe they have various alloys that are otherwise mechanical equivalents, meaning other mechanical properties being otherwise similar to A356 / 2024 / 6061 / 7075. The alloying metals are partially or completely replaced with Alumina Oxide nanoparticles. The more nano that the material is loaded with, the greater the cost, which was on par with 304 stainless (not cheap, but not as expensive as Titanium). Those mechanical properties were obtained without any heat treatment / tempering at all, which was supposed to be one of its major selling points (meaning you cast / extrude / forge it and then that's it, no subsequent heat treatment). Their head salesman is also an aviation guy, so there's that. He told me when we last spoke about a year ago that they were going to start experimenting with nanoparticle-enhanced Magnesium alloys as well. The material is also very hard, so machining it will require carbide or coated cutting tools. It's also slightly heavier, so something around 0.1g/cm^3 heavier than 6061 or 7075. So, apart from hardness, strength at elevated temperatures, somewhat better corrosion resistance before surface treatments are applied, a slight weight increase, and a significant cost increase, otherwise much like those aforementioned conventional alloys.

#### aeromomentum

##### Well-Known Member
Aero,

My experience running Lycs is that up to 8.5-1 compression (most 160 HP 320s, 180 HP 360s), you can run premium mogas. But without 'extraordinary measures', you must use ethanol-free fuel to avoid damage, primarily to the 'soft parts' in older mechanical fuel pumps, and the same for many gascolators. E-free *premium* mogas is available, but often very hard to find, and a lot more expensive than commonly available premium E-gas.

I've flown a couple of ~400 mile cross countries in my 160hp Lyc powered RV-4, with Tracy Crook's Mazda Renesis powered RV-4. In both cases, I was burning E-free premium and he was burning 87 octane E-mogas. Unlike many Lyc drivers, I aggressively lean my Lyc. On both these flights, Tracy burned slightly less than 10% more fuel than my Lyc (same flight profiles). But. Remember that he could burn 87 octane E-gas, while I was forced to burn premium E-free gas. My cost per gallon was around 50% higher than his cost.

Charlie
While EAB's do not need an STC check out Petersons and EAA auto fuel STC's for more information. There are many engines listed and most can run 87 octane. Yes some like the high compression 160hp O-320 require 91 octane. Certified aircraft do require ethanol free via the STC but not EAB. Keep in mind that ethanol can destroy your fuel tanks, spars, etc in any case. You will need to make sure your carb, FI, fuel lines, etc can take the ethanol but most newer parts can even for a Lycoming.

#### rv7charlie

##### Well-Known Member
Aero, Russel,
I agree with all points; my previous post didn't mention every detail, for brevity. I'm aware that the lower compression engines can run 87; I learned to fly feeding 87 octane mogas to a Luscombe 8A. I did try a little bit of 87 octane in my 1st 160 HP (8.5-1) RV-4 circa 1995, but neither it nor I were fond of the results. Been running E-free premium ever since, in 160 & 180 Lycs. I know that the newer fuel pumps are E-proof, but my current -4's pump was on the plane when I bought it in 2003; likely since before the mid-time engine was installed back in '91. Not going to risk E-gas in that plane due both to that pump, and having no idea how E-proof the ~30 year old sealant in the tanks is. I test every batch of E-free I purchase for ethanol, though I probably don't have to worry about it as much as Russell; my tanks are aluminum.

I know of some guys running E-premium in 180 Lycs, but I haven't made the decision yet to try it in my just-purchased RV-6's 180 Lyc. Can't wait to get the Mazda rotary project in the air, then I can go to E-whatever, octane whatever, with no worries.

Charlie

#### GESchwarz

##### Well-Known Member
My build of over 10 years is designed around the rotary engine. It has gotten to the point where I am just about to start working on the engine installation. I have the engine and nothing else. Who do I talk to about getting outfitted with the PSRU, ECU, Harmonic Damper, etc.?

#### Lendo

##### Well-Known Member
GESchwarz, Being it's for a Rotary, I would be speaking to BillRsv4 and seeing if he's ready to start manufacturing, but he may well be still in the design stage and seeking quality parts manufacture.
It's relatively easy enough to make a PSRU, but it's much more difficult to design a QUALITY PSRU, with Quality components - while keeping costs reasonable to make the product Value for Money for the consumer.
BTW I have no financial association with Bill or his design.
George

#### Urquiola

##### Well-Known Member
Hi thjakits, welcome to the conversation!

Just to address the latter parts of your post, the rotary just doesn't have a lot of torque (power basically comes from rpm in the rotary's case). There's no practical way to get low rpm torque out of a rotary, largely because of the very short 'stroke', for lack of a better term. High torque piston engines tend to have relatively long stroke cranks; something that's almost impossible to do with a Wankel style rotary. Mazda does have a prototype engine (the 16X) that has a bigger eccentric (stroke), as well as all aluminum construction, that rotary flyers have been salivating over for a long time, but there's no real evidence we'll ever see it in production.
I owned a great 'automatic pitch adjustment' prop about 25 years ago (on a Lyc); Bernie Warnke was famous for his 'almost constant speed' wood props back in the 1980s-1990s. But that wouldn't buy you much with a rotary; there's no real power until you get up in the 5k rpm range and above.

I don't recall any emphasis on ceramic side housings, unless you're talking about aluminum with a high tech wear surface. You can buy those now, but they're very pricey, and I don't know anyone who's used them in an a/c designed to be actually used. (There's the time-to-climb record holder, but I doubt that engine ever ran longer than 20 minutes at a time, and probably has fewer than 100 hrs total on the engine (with a couple of rebuilds in the mix for various reasons).

Using automotive engine control (if you mean 'from the car') is a very dicey proposition, because modern car controllers have literally hundreds of thousands of lines of code running their cpu's, most of which ties the engine to the rest of the car in ways that are hard to 'hack' around, and since the code isn't open source, there's no way to know whether we've hacked around every transmission shift point code, limp home mode, etc. People have been killed trying. There are several controller sources still available that have been extensively debugged on 'hot rod' ground vehicles that are much better (safer) options. The big hangup at the moment is the gear drive, now that Tracy is retired.

Belt drives: It's been done; there was a toothed belt version flying successfully on an amphib in the USA northwest for a number of years. But they tend to be bulky, require cooling air, and aren't really any lighter than a properly designed gearbox. Most people just won't trust a V-belt array on a high HP engine in a fast a/c, whether it's justified or not.

Charlie
Yes, Eccentricity in Wankel is roughly same as stroke in a Reciprocating engine, making it an extreme example of 'over-squared' engine, British cars had long stroke engines that gave good performance, good torque, at low rpm end, low load. When raced, piston speed of long stroke engines is higher, increasing wear and mechanical losses.

This arrangement affects emissions in Reciprocating engines, but in Wankel, at low rpm, emissions come from seal leaks, at high rpm, from quenching, flame extinction next to cold or narrow surfaces.

Air cooled housing RCEs work slightly hotter than liquid cooled counterparts, being advantageous from thermodynamic point of view, not so from thermal dilatation and engine life.
About use of alcohol mixes, you'll be surprised from SAE paper 2014-01-2611,'The Influence of Some Synthetic Fuels on the Performance and Emissions in a Wankel Engine', Ksenia Siadkowska et al; and SAE paper 840237, 'The Effect of Alcohol Blends on the Performance of an Air Cooled Rotary Trochoidal Engine'; Marcel Gutman, Izu Iuster. Methanol seems cheaper to make, but is extremely toxic, can turn blind or dead, propanol has the best weight-energy ratio, but the industrial aspects of alcohol production remain a mistery to me. Ethanol is widespread, among other reasons, because there is a worldwide grain surplus, even if in the Chicago futures market, grain is traded every year for a value 40 times the total world grain production of same year. Thanks. Salut +

#### AIRCAB

##### Well-Known Member
One clarification on rotary housings. The proper term for the housing that contains the spark plugs, and the 'peripheral' (if installed) intake port and the peripheral exhaust port (in pre-Renesis Mazda rotaries) is properly called the side housing. That housing has always been aluminum in the Mazda engines. I mis-used the term earlier (duh) speaking about the *end housings* and center housings, which touch the *sides* of the rotors. Confusing enough that I get sloppy with the terms sometimes, and I've watched rotaries since before my father bought a new 1974 RX4. Anyway, end and center housings are what are available from aftermarket vendors; the side (rotor) housings have chrome wear surfaces, and people rarely have issues with them in aviation applications, unless there's FOD, or damage from broken apex seals (rare except in boosted engines).

I can kinda see why the V belt system might be chosen for a helicopter, with the need for a clutch in addition to the reduction aspect, and the maintenance-intensive nature and complexity of the tensioning system would be a minor addition in a heli. Given your experience with the Robinson system, you could likely make a V-belt system work if you wanted. But I think you will find it quite a bit heavier than other choices, and very few homebuilders willing to deal with that level of complexity and maintenance intensity in a fixed wing a/c.

Charlie
Enjoy reading this post. Wankel world refers to the housings with plugs as rotor housings, and the other as plates, side or center.

HBA Supporter

#### Lendo

##### Well-Known Member
Urqiola,
I would have thought Butanol, had a better energy rating, I don't know about weight.
The narrow surfaces you refer to are referred to a SQUISH areas and the do (as you say) extinguish the flame front. The result is fuel burned in the exhaust.This has been overcome in the RX8 with side exhausts, as the Apex seal carries unburnt fuel into the next combustion event. However side inlets and side exhausts are more restrictive than the Peripheral Ports.
George