I tried 50:1 premix (a quart on gas tank fill) along with running factory S4 oil injection when I ran Rotary Aviation classic 3mm seals, 2mm Super Seals and 3mm Super Seals.

These RA apex seals still eat your rotor housings.

Any rotor housing/apex seal will last longer if you keep the RPMs down and oil/coolant temps low to help cool the apex seals and rotor housing surface.

Last several motors I did were 3mm factory apex seals and they are kind to rotor housings except the corner seal wears that groove around the rotor housing edge.

On the last 3mm Mazda seal motor I did I used a turbo that shifted the powerband to the low rpm so I didn't lose anything shifting around 6,500rpm.

I noticed this low rpm engine built and kept noticeably better compression than previous engines. The 3mm seals don't like constantly wanging against the 8,000rpm redline...

 

oh damn ... that kind of sucks .. starting to dislike these apex seals . I was hoping to get a good 40-50 k miles out of this engine since the housings were so mint . looks like I'm SOL

The engine was ported a bit more then I would of liked, hopefully running 1oz per gallon of premix + OMP will help reduce the wear.

 

I was hoping to get a good 40-50 k miles out of this engine since the housings were so mint .

Completely possible. Stock apex seals/rotor housings usually last 150-200,000 miles with normal people driving them (not raced).

40 k miles on a stock motor and you have near mint rotor housings- 40 k miles with these seals and you won't be re-using the rotor housings. That is not to say the motor won't still run...

 

yeah still what can I say kind of disappointing LOL I guess I'll go with ALS or Goopy next time around . goopy are supposed to be easier on the housings then RA so I would rather swap out seals , then swap out housings . every 40k miles.

Well in 40k Miles or so when ever its time for this motor to come out . We will see how it is inside! granted it will of been raced and ran at or near 8k for alot of its life

 

I have been browsing this thread for the last 2 years, and I have a question if someone can answer it.

First question: How does a "soft" seal eat a hardened chromium surface? I would imagine that a hard chrome surface would eat the softer seal.

Second question: How does a "hard" seal not eat the chromium surface? Chrome is some hard stuff, but can be cut (based on the Mohs hardness scale).

Hardened steel is only 7.5 on the scale whereas chrome is 8.5.

 

I have been browsing this thread for the last 2 years, and I have a question if someone can answer it.

1) First question: How does a "soft" seal eat a hardened chromium surface? I would imagine that a hard chrome surface would eat the softer seal.

2) Second question: How does a "hard" seal not eat the chromium surface? Chrome is some hard stuff, but can be cut (based on the Mohs hardness scale).

Hardened steel is only 7.5 on the scale whereas chrome is 8.5.

Your two questions have one answer- galling.

1) I haven't sent any samples off to be tested, but what it looks like to my eye is the soft seals have a higher coefficient of friction against the chrome so build up enormous heat at the wear surface (hence the cracking and warping of the seal's wear surface) and are continuously welding themselves to the chrome housing surface and breaking free tearing off chrome and apex seal and depositing it back onto the rotor housing surface other than where it came from.

We are saying the soft seals are "wearing out" the chrome housing, but its more moving the chrome around into radial ridges and valleys. Instead of worn out perhaps we could say it has been turned into an unusable housing and an unusable seal.

This is called galling and it is from adhesion of incompatible materials sliding against each other.

To quote Wiki
Galling is most commonly found in metal surfaces that are in sliding contact with each other. It is especially common where there is inadequate lubrication between the surfaces. However, certain metals will generally be more prone to galling, due to the atomic structure of their crystals. For example, aluminum is a metal which will gall very easily, whereas annealed (softened) steel is slightly more resistant to galling. Steel that is fully hardened is very resistant to galling.

Now you know why the harder steel "wears" out the chrome less than the softer steel.

2) Hard factory seals have a lower coefficient of friction against the chrome housing so develop less frictional heat at the wear surface. Lower friction is specifically why they are hardened.

Mazda used an electron beam to melt the top ~3mm of the apex seal and then rapidly quenched it to form a fine crystalline carbide ceramic like material at the wear surface and a tougher ductile seal body.

Mazda SAE paper 860560
http://www.rotaryeng.net/material-tech.pdf

 

Hmm, this is an interesting theory that sounds very plausible. But to further your theory, I believe the chatter marks are caused from not only the galling incidences, but the apex seals are binding due to this higher CF value and then floating on the apex seal spring. The seals are reaching a sort of harmonic frequency and continually lift and slam back onto the rotor surface. Maybe, at the point where the materials are starting to gall the exerting forces drive the seal off of the rotor housing which induce the seal lift off?

I imagine lower rpms and high premixing should mitigate the wear rate but in time, with the wrong apex seal material, the damage will still occur.

 

This theory would leave me to believe the knock sensor could/should detect if this chattering is occuring. I have noticed that my knock levels are higher after my rebuild last summer. The only thing I changed in the rebuild was the apex seals (from OEM to ALS/E&J). The louder an engine gets (more knock level reading) could possibly mean the seals are lifting more from the housing? Just a theory.

 

Hmm, this is an interesting theory that sounds very plausible. But to further your theory, I believe the chatter marks are caused from not only the galling incidences, but the apex seals are binding due to this higher CF value and then floating on the apex seal spring. The seals are reaching a sort of harmonic frequency and continually lift and slam back onto the rotor surface. Maybe, at the point where the materials are starting to gall the exerting forces drive the seal off of the rotor housing which induce the seal lift off?

I imagine lower rpms and high premixing should mitigate the wear rate but in time, with the wrong apex seal material, the damage will still occur.

You must divorce axial chatter marks from radial grooving in your thinking.

Chatter marks are a well documented phenomenon in the development of the rotary and are caused by harmonics, binding and floating of the apex seal as you state.

The radial grooving left by galling go around and around the rotor housing, axial chatter marks go from side to side across the width of the rotor housing.

Galling is not well documented because no company (besides aftermarket) would put incompatible seal material into production.

 

This theory would leave me to believe the knock sensor could/should detect if this chattering is occuring. I have noticed that my knock levels are higher after my rebuild last summer. The only thing I changed in the rebuild was the apex seals (from OEM to ALS/E&J). The louder an engine gets (more knock level reading) could possibly mean the seals are lifting more from the housing? Just a theory.

Yes, but not chatter per say, but a related and well documented issue called by Mazda "spit back".

Spit back is likely causing compression pre-ignition knock which probably has a similar frequency as detonation knock.

Spit back is when the apex seal lifts from the rotor housing when crossing the minor axis (the narrow section of the rotor housing) on the compression stroke causing exhaust gasses below the apex seal to ignite the compression mixture above the apex seal prematurely.

Spit back is caused by apex seal inertia (weight of seal and rpm vs gas pressure behind the seal), friction (softer seal has more friction in groove and on housing), chatter (harmonic/friction induced), improper clearances (influenced by seal temperature as well as rotor/oil temperature).

You on the right track thinking about chatter as they are inter-related phenomenon.

 

Ah, galling. Know of it but didnt think of it.

My old 160k mile housings have chatter marks, but nothing out of the ordinary. Even after 160k my housings would have been usable except for the extreme corner seal wear. The corner seals dug a canal through the chrome (didnt get all the way through) but it was like 3mm wide.

I got new rotor housings. Everything was fine except where the triangle corner seal did its damage. Like I said, it was the motor that came with the car, so it had the 3 piece seals so I am not sure if there is a huge difference between that and a 2 piece seal when it comes to wear and tear. I put in the newest FD 2 piece mazda seals and springs for the rebuild. The only thing that I can think of that would preserve rotor housings is a one peice seal, but they dont seal very good.

 

Ok, so I read through the paper you posted (should be working ), thank you btw, was not aware of the term 'spit back' prior to reading that. I don't recall Kenichi Yamamoto's paper ever mentioning that term, but I might have just forgot.

Anyways, according to that document, one would see more wear with boosted engines as FPo would be greater. We can assume F(aftermarket seal) is greater than F(stock seal) so all else being equal, you'll automatically get more spit back using that equation. I suppose aftermarket seals really should require the use of aftermarket seal springs that will balance out that delicate relationship? I know I was never advised to upgrade my apex seal springs when switching to E&J/ALS seals.

On a side note, I think it is a safe assumption that the knock sensor will indeed detect spit-back, as spit back will cause light knock...and thus chatter correct?

 

was not aware of the term 'spit back' prior to reading that. I don't recall Kenichi Yamamoto's paper ever mentioning that term, but I might have just forgot.

First time I read the term "spit-back" was reading Mazda's paper on development of synthetic lubricating oil with Idemitsu for the Lemans engines and increasing high rpm power with increased spark timing/reduced spit back.

SAE paper 922375 "The Development of Lubricating Oils for Rotary Racing Engines"

http://www.rotaryeng.net/rotary-oil.pdf

I think it is a safe assumption that the knock sensor will indeed detect spit-back, as spit back will cause light knock...and thus chatter correct?

I believe so as well. Note the factory knock retard is only operational in the lower rpms to avoid false triggering.

I know I was never advised to upgrade my apex seal springs when switching to E&J/ALS seals.

The springs behind the metal seals in a rotary engine only help seal the engine enough to start. It is combustion gas pressure that pushes the seals against the housings to seal once started.

This is why a "worn out" , "flooded" or "carbon locked" rotary is very hard to start (often requiring a pull start) but does not stall out (runs fine) once running.

It is possible to increase the gas pressure behind the seals through higher clearances.

It is also possible to cut trunnions from on the sides of the apex seal down to the base for a faster response to combustion gas pressure such as Mercedes did with the C-111 4 rotor super car apex seals.
*fun side note- Mercedes also developed the ceramic tip cast iron apex seals in the '60s that Mazda perfected the low cost production of with their chilled ion beam hardening of ductile cast iron in the '70s.

More gas pressure behind the seals will mean more friction/wear/less power. It would probably be the worst thing you can do when your apex seals are an incompatible material with the rotor housing chrome.

 

In that case a boosted motor would compound the friction forces acting against the apex seal. Do you have any experience running a soft seal on a naturally aspirated motor?

 

None

 

I too would like to know if super seals are safe on n/a rotarys? Theres a fc i wanna get that has ra super seals. Its a stock s5 na motor with super seals and omp delete so it needs to be premixxed. Since i dont plan to boost it would the motor be fine so long as a premix?

 

it may be worth pointing this out, but reciprocating piston engines seal the same way.

 

i have a little. i've got a bunch of 73 and older engine cores, and while the seals were always bad, the chrome surface always looks great. i also have a P port with carbon seals that i put together in 08, it doesn't have a lot of miles on it, so i can't really comment on wear, but just from the way it runs you'd never know it has 1 piece seals.

its also a bit hard to separate the effect of the PP from any effect of the seals, as the PP is tamer than the side port in some ways.

i went carbon seals, as the carbon seals are very easy on the housings i can't replace. they are also much much lighter than a steel seal, so they work @high rpm, and i expect to see 9k at some point.

i was cautioned to watch for detonation with the carbon seals, which seems odd for an NA engine, but it will knock @low rpm/high load.

 

Originally Posted by rx7 SE View Post
Do you have any experience running a soft seal on a naturally aspirated motor?

i have a little. i've got a bunch of 73 and older engine cores, and while the seals were always bad, the chrome surface always looks great. i also have a P port with carbon seals that i put together in 08, it doesn't have a lot of miles on it, so i can't really comment on wear, but just from the way it runs you'd never know it has 1 piece seals.

But those soft seals are a material compatible with the chrome housings developed by Mazda- I think RX7 SE specifically meant aftermarket metal apex seals.

I would say why?
You want an apex seal that is indestructible in an NA rotary and will run over 200,000miles look no further than the stock 2 piece apex seals.

The factory 3 piece design offered slightly better sealing for emissions/economy and gave up some durability in the 150,000+ mile range. Once the engine no longer had to meet stringent production standards Mazda reverted to the 2 piece design.

It is only if you need to consistently run at or over 8,000rpm that you need a different apex seal and you can use the well proven carbon/aluminum or ceramic.

Any metal seal will be worse for high rpm than the stock Mazda seal.

it may be worth pointing this out, but reciprocating piston engines seal the same way.

Exactly, but there aren't springs behind piston rings (because it is a spring) so it isn't so common that people get the wrong idea on the sealing forces.

 

i was cautioned to watch for detonation with the carbon seals, which seems odd for an NA engine, but it will knock @low rpm/high load.

Yes, Mazda had to use a distributor with centrifugal and vacuum advance/retard to stop the engine knock that would damage the stock '67-73 engines with Carbon/Aluminum apex seals.

Nsu decided against such complexity and had only a centrifugal advance distributor and used steel apex seals for knock resistance. This caused their production engines to last about 30,000miles...

so they changed the apex seal material to steel with a Titanium Carbide wear surface bonded on like a machine tool, but it was too late for their reputation.

Again, this Titanium Carbide bonded wear surface is basically the same thing as Mazda's much cheaper production process of creating Iron Carbide wear surface on ductile iron seals with chilled electron beam hardening.

 

This makes me think.

If someone actually wanted to create a superior metal apex seal they could work with a modern machine tool manufacturer to make carbide bonded metal alloy body apex seals that should be both tougher than stock to detonation shock and have the same wear properties or slightly better.

The down side is that they would be even more expensive than the simple yet expensive ceramic apex seals that everyone is too cheap to purchase.

We saw with the aerospace grade metal coil clad coolant seals how little demand there is for very expensive solutions to even the most common rotary problems- I am sure this is the same reason no one has made a better apex seal than stock.

 

if you look at the distributor curves for the older engines (they are in the FSM), they hardly run any timing at all, and actually the race engines are interesting here too. as sealing improves, power improves, and timing actually increases. for instance the Rx3 style P port runs 15 degrees BTDC, and the 79 Rx7 style runs 20 BTDC. and the oil study makes clear, the R26B wants to run 25.

anyways data point #2, the NSU spider runs a seal that is like Mazda's carbon seal, its not steel. or maybe the Ro80 was steel, and the spyder wasn't, i've never played with an Ro80

 

we did a track day with the Mclaren club, and they had a few P1's out there. the P1 has a Diamond Like Coating (DLC) on the ceramic disc brake rotor. its really obvious, as the rotor is flat and shiny like a mirror, no grooves, no scratches.

if something like that works on a brake rotor, it might do well on a rotor housing too.

 

anyways data point #2, the NSU spider runs a seal that is like Mazda's carbon seal, its not steel. or maybe the Ro80 was steel, and the spyder wasn't, i've never played with an Ro80

Yes, the low production Spider's (~2,000units) DKM-502 was a problematic engine whose road testing was prematurely foisted off on the public in NSU's rush for the glory/payoff of realizing a production Wankel engine.

It had carbon apex seals with steel corner pieces at each end (like Mercedes instead of 1 end like Mazda). The engines ran fairly reliably in the lab, but once they were in the hands of consumers the engine knock I mentioned broke the carbon apex seals regularly.

Jan P. Norbye "the Wankel engine 1971
chapter "NSU develops the Wankel pg 133-134

(*proceeding my notes)


This engine (*KKM612 of the ~37,000unit in 10 years RO-80) was the first NSU Wankel engine with cast-iron apex seals, replacing the KKM-502's (*~2,000unit Spider's) carbon compound seals which had a tendency to break up under detonation. The carbon seals worked fine in laboratory tests, but on the road, they did not hold up because at extremely high rotational speeds the centrifugal advance mechanism would advance timing to the point of detonation, and this form of abnormal combustion immediately cracked the brittle carbon seals. The improved KKM-612 metallic apex seal was a link-type seal. Unlike the KKM-502 seal, the split between the center and corner pieces was on the tip (*such as in Mazda apex seals corner piece) rather than on the axial face of the seal strip (*as on aftermarket corner seal piece). It was partially self-adjusting to wear on both axial faces, and to conform to any convex or concave deformation of the seal strip groove. The seal was installed oversize (*apex seal long piece standing proud of the corner pieces on each end) and had a break-in period of 30 hours. Small slots in the leading face of the apex seal (*gas transfer ports on the combustion chamber side of the apex seal as I previously described the Mercedes apex seals as having) ensured rapid pressure build-up beneath the seal to prevent spitback (firing across the apex seal from the leading combustion chamber into the following one, i.e., the chamber under compression).
It was found the cast-iron apex seal was incompatible with the chromed working surface used with the previous carbon compound seals: therefore NSU developed an entirely new coating material for the working surface called EINSIL. This material, a nickel plate containing fine particles of silicone-carbide suspended in an electrolyte, is integrated with the surface during the plating process.

....Under deliberate test bed knocking, the old-type carbon seals stayed together for only 10 seconds; under identical conditions the iron seals lived for 10 hours.

....Sill, the NSU engineers were surprised at the amount of wear found in the engines of cars that had been running on a stop-start test cycle. To cure this problem, they twice changed the kind of cast-iron used for the rotor tip seals, each tie gaining a significant improvement in cold-start wear, and they are at present satisfied that the trouble has been cured. (*note as I previously noted, a final post production change was to use Titanium Carbide bonded to the cast iron apex seals as they still wore out within 30,000miles).

 

i had an NSU spider, and it came with a couple of spare engines, and i'm pretty sure it had one piece seals in it.

its pretty neat, it has a 2 barrel carb feeding 2 peripheral ports. the primary port is teeny, and i think was the later opening one, the main port was much bigger, and had all the overlap.

the other interesting thing was that the rotor housing was a closed casting, so no water seals, and the side irons are 30mm thick, as all they do is seal the side, no water jacket.