Making a Rotary Engine Better

Mazzei is up to making a 5 rotor engine

Having replicated the sound of the 787B with his turbocharged Formula Mazzei RX-7, he now wants to make one that sounds like a V10 F1 engine.

In reply to j_tso :

I'm a fan of any odd number arrangement because they all tend to sound good: 3 cylinder triumphs, 5 cylinder Audis, Mazda's 20b. A 5 rotor should have quite an interesting sound throughout the rev range 

Pete. (l33t FS) said:

In reply to aircooled :

The thermal expansion is why Mazda used a lot of fairly thin tension bolts at fairly low torque.  Spec is 23-29ft-lb on undercut 10x1.0 bolts, so they are more like springs.  ("Tension bolt" isn't just a clever name )  I usually torque to 23 because I am more concerned about the engine being able to tolerate high coolant temps.  Super tight bolts (some drag racers torque the OE bolts to 50!) means there is no "give" to them if the engine gets hot, so the rotor housings get crushed (aluminum is weaker than 17-19 bolts) and then the coolant seals blow on the next cold start because now the whole assembly is relaxed and not under tension...

Interesting.  You might also notice the guy in the video has a mysterious coolant seal almost failure in the video above.  Maybe the result of some of his strengthening.

I suspect there is a good reason why the rotor housings and side housings are not either all aluminum or steel?  Of course the rotors are steel.  I suspect the apex seal lands would not last long on an aluminum rotor. I am sure Wankel and  the engineers that followed him have thoroughly considered all these things long ago of course.

The more I learn about these motors the weirder they get, which I think make them even cooler.

While not cheap, I wonder if titanium would be a better choice for bolts?
 

Good strength and the ability to stretch without permanent deformation. 

In reply to aircooled :

In the early 2000s Mazda showed off a 16X engine that was going to have aluminum side housings. I think the NSU or Citroen rotaries used aluminum.

Racing Beat used to sell them at about $1500 each, and now there's a New Zealand company that makes them out of billet aluminum with replaceable inserts. The clientele for that appears to be mostly drag racers.

I'd take everything Rob Dahm says with a grain of salt. He claims all these "fixes" have been held secret but anyone that's been remotely involved with these engines in the part 20 years knows about all of them. I've never met the guy in person and while he generally seems well-intentioned he has a bit of "I'm internet famous, listen to me" going on.

When it comes to these engines my line of thinking is Mazda has spent 50+ years refining the design so they have a pretty good idea of what they're doing. Yes, the mass-produced engines have lots of concessions for things like emissions, cost, ease of manufacturing, etc. but they generally know what's best.

In reply to infernosg :

Although I've seen a fair number of complaints about the build quality of Mazda remans, especially Renesis ones. There is definitely room for improvement.

infernosg said:

I'd take everything Rob Dahm says with a grain of salt. He claims all these "fixes" have been held secret but anyone that's been remotely involved with these engines in the part 20 years knows about all of them. I've never met the guy in person and while he generally seems well-intentioned he has a bit of "I'm internet famous, listen to me" going on.

I didn't watch the video, so I didn't say anything about it, but a lot of the things people are discussing in this thread have been common knowledge online in places like rx7club and nopistons for fifteen-twenty years.  The rotary community, by and large, is very free and open with their expertise.

When it comes to these engines my line of thinking is Mazda has spent 50+ years refining the design so they have a pretty good idea of what they're doing. Yes, the mass-produced engines have lots of concessions for things like emissions, cost, ease of manufacturing, etc. but they generally know what's best.

I am sure the MSP engine could have been better if they weren't stuck with legacy engine dimensions for cost of manufacture reasons.  The tension bolt locations were determined in 1973, for instance, and the spark plugs have been moved around as the design got more refined.  The result is the 13B-REW and the 13B-MSP get shortchanged for coolant around the hottest point of the engine.  If the tension bolts got moved at the same time the spark plugs did, this wouldn't be a problem, but now you're dealing with redoing some very expensive gang-drilling machinery, casting molds, etc.

I am sure the MSP could have been better if the intermediate housing was 30-50mm wider, but that would require even more tooling changes.  And the engines would probably last longer if they had reverted to 1mm side seals.

No Time said:

While not cheap, I wonder if titanium would be a better choice for bolts?
 

Good strength and the ability to stretch without permanent deformation. 

You'd want to go with Inconel over titanium. 

Pete. (l33t FS) said:

dean1484 said:

I also was fascinated with the loading on the bolts that hold the motor together. The need to make those fit tight makes perfect sense.

It's also wrong, though. Mazda's research showed that there needed to be some disconnect between the combustion area of the rotor housing and the rest of the engine, because of thermal expansion effects.  So they use coolant seals that can move around a bit, and they dowel the engine together bracketing the combustion chamber, so that area can float around.  On race engines they would even cut the ribs on the inner side of the rotor housing for maximum thermal flexibility.

Do you have any writeups on this from Mazda?  Or from other people that have tested researched this?  I am not saying you are wrong I just get skeptical when one person says the answer is black and the other person says the answer is white.  This usually means I need to do my own research.

dean1484 said:

Pete. (l33t FS) said:

dean1484 said:

I also was fascinated with the loading on the bolts that hold the motor together. The need to make those fit tight makes perfect sense.

It's also wrong, though. Mazda's research showed that there needed to be some disconnect between the combustion area of the rotor housing and the rest of the engine, because of thermal expansion effects.  So they use coolant seals that can move around a bit, and they dowel the engine together bracketing the combustion chamber, so that area can float around.  On race engines they would even cut the ribs on the inner side of the rotor housing for maximum thermal flexibility.

Do you have any writeups on this from Mazda?  Or from other people that have tested researched this?  I am not saying you are wrong I just get skeptical when one person says the answer is black and the other person says the answer is white.  This usually means I need to do my own research.

It is from one of the many SAE papers that Mazda wrote on the subject.  They did the research for you, you would just need to track down the papers.  A lot of the papers, unfortunately, have been unpersoned from online.  (well, it IS copyright infringement to post them for free...)

It's kind of important to note that Mazda is mainly concerned with longevity and durability, not high BMEP strength over a very short span of time.  I would expect that most GRM types want an engine that can handle back to back 60 second autocross runs or 20 minute track sessions, not hurling a Corolla or RX-3 down a dragstrip.

it's also interesting to note that the factory endurance racing cars that had dominated for so long in IMSA were largely stock internally, most mods were to the effect of making more power, and cooling.  More power, we might not need, cooling is fairly easy.  They suck on the street but make awesomely reliable racing engines.

Not finding papers but found a pic of what a regarded engine builder who builds high power turbo engines does for increased cooling around the spark plugs, along with a chart from an SAE paper that Mazda put out.

If image doesn't show I'll rehost instead of hotlinking.

The problem he is attacking is the rotor housings warping with heat, which raises the area around the spark plug, which puts a cyclical bending load on the apex seals, which leads to strange wear and loss of compression at best, or stress fracture at worst.

Also, look at those temperatures.  Any ethylene glycol/water based coolant is going to boil there even with a stellar water pump, you'd need about 200-250psi in the block to keep it from film boiling.  Film boiling which makes the engine run hotter even if the water temp is "okay", because the water isn't pulling heat from the hot areas anymore as it is insulated by a layer of steam.  No wonder Evans NPG was/is so popular for FD people.

It also makes me wonder about the Chinese Wankels that they used in some military vehicles.  Information about them is slim at best other than knowing that they existed, and that they used IRON rotor housings.  That would grow with heat a lot less and thus have less of a compression loss issue!

Heh... peripheral port 2.2 truck engine.