shelleys_man_06

I was walking towards my computer, and a new question popped into my head. When is appropriate to switch to bigger apex seals? How much chamber heat and pressure can the current seals hold before they are compromised? What about the side seals? I don't have CFD or FEM...yet, so I can't do any nerdy analysis. Also, are there better sealing materials for rotary engines, other than rubber, assuming it is rubber? I fear the broken O-ring if and when I crank up the boost.

86rx7

We'll, upgraded seals are never really required. Detonation breaks apex seals, not chamber pressure. As long as you tune the car right 2mm are fine for all the power you can make. someone calculated once that it would take something like 2500 hp per rotor before the seals were comproimised. 3mm seals on an rx8 would lower the redline to something like 7000 rpm due to seal chatter and seal float because of there increased mass. Some of the top 600-700 HP rx7's use 2mm apex seals. Side seals also never need upgrading. Things that fail due to extreme HP are the rotors, the eshaft and the side plates near the dowl area. No one knows how stong these componets are on the rx8 yet, but i would suspect that the eshaft and housings are stronger the previous cars. the rotors may be weaker due to there thinner lightweight casting, but only time will tell. generally High GP has a tedency to dent rotors, stake e-shafts and the twisting motin between the housings causes dowl pin seat to crack(these things only occurs at ubsurdly high HP's) This is why dowl pinning or tenstion bolt studs are often used on high HP cars. Lastly apex and side seals are not rubber. The only "rubber" seals in the motor are the inner and outer water o-rings (think headgasket) and the o-rings for the oil control rings. These both do not fail unless the car is severly overheated. There are ceramic apex seals avaliable, and ive heard of protoypes from Ianatti, but these are mainly used for racting due to there expence. The benifits of ceramic seals are less engine wear, less friction, Less mass which allows a higher redline ( i think the rx8 seals are allready light enough to spin the mottor much faster thne stock) and the ability to run higher exhaust temps (this only really comes into play on NA race engines where they are leaning them out for every once of power, a turbo cars exhaust temps would be much cooler thne this due to the extra unburnt fuel in the exhaust. They are also more resitant to bening (only helps a car with periphreal exhaust and/or intake as there is no real bending load on the seals in the rx8) and they are a little stronger, but wills till break from detonation.

In short there are no seals immune to detonation. Rx8 seals will take almost anything you can thow at them as long as there is no detonation. There is no need to upgrade, and thicker apex seals would be a redline reducing downgrade. Other parts fail first in ultra high HP operation. Ceramic seals are great, but will not save you from detonation if it occurs( not to mention the 2k a set price tag). If you want the motor to last, tune it right with nice fat AF ratios, down in the 11's.

shelleys_man_06

I had forgotten about the issue of detonation. Sorry I had implied that an apex seal is made of rubber. I was referring to the O-ring seal. So, I suppose the stock seals will suffice for a street/track car? Though the benefits of ceramic apex seals are clear, do they have more of a tendency to wear out? Also, what type of material is a typical apex seal made of? I just wanted to have someone clarify that, since there aren't many threads that discuss the strength of the RENESIS. Thank you for the info 86rx7 .

86rx7

Ceramic seals will outlast stock seals due to there hardness and lower coeficient of friction. Carbon apex seals on the other hand, which are only used for racing, non FI rotaries wear out in ~20k miles on the street, however i am not aware of carbons being avaliable for the rx7 (probably never will be because the stock seals are pretty light/ the only advantage of carbons is ther light weight which alows a higher redline) Stock apex seals (i'm 98% sure rx8 seals are manufactured the same way rx7 apexes are ) are bainitic cast iron with an electron beam chill hardend tip.

shelleys_man_06

I had forgotten that ceramic is quite a durable material.

shelleys_man_06

Alright. I found a link with more information (anecdote) about the usefulness of ceramic apex seals.

http://www.selectmaz.com.au/apex_seals.htm

shelleys_man_06

How many pieces (one-piece, two-piece, etc.) make up the RX-8's apex seal? Also, what type of material is the apex seal spring, and what is the spring constant? What is the purpose of the electron beam-chilled tip for the apex seal?

shelleys_man_06

I wonder if 86rx7 has the answer...

86rx7

2 pieces, a main piece and a corner piece. This is so the seal runs zero end clearence. The springs are some kind of stainless steel I think, there are two of them an inner and an outer. I'm unsure of the spring rate, but in a rotary the seal springs are mainly for starting . Apex corner and side seals work off of gas pressure to press them into there sealing surfaces. the chiill hardend tip is to increase the hardness of the wear surface of the seal for improved life and less friction.

86rx7

pics

shelleys_man_06

Thank you for the information 86rx7. You are a true beacon of knowledge .

shelleys_man_06

What are the differences between a one-piece, and other segmented apex seals? What kind of advantages/disadvantages they have over each other?

86rx7

Well, one piece seals need to have there end clearence set. Because of this the will allways leak a little pas the ends of the seal. This doesnt matter to much if your motor is spinning really high rpms all the time like a race motor (hence why carbon apex seals are one piece) Two piece seals effectivly run zero end clearence because the seal spring pushes the corner piece against the side end plate. There is still a small leak because the seals are machined slightly longer then the rotor width so that the seal doesnt leak from the gap where it is divided. The corner piece is actually slid down slightly away from the rotor housing in operation due to this, but this leak is tiny compared to one piece seals. Some aftermarket apex seals for rx7's have larger corner pieces which leak somewhat more, but it is still much much less then one piece seals. Two piece seals work good for low and high rpm. Three piece seals are what was used on previous rx7's after 86(now replaced by 2 piece seals) and are duvided once again lengthwise across the main part of the seal. These seals effectivly run zero groove clearence as well as zero end clearence. The draw backs to the three piece design have been debated for a long time and i havnt seen enough evidence either way to make a conculssion, among the possible drawbacks are a thermal break in the seal which hurts heat transfer and more apex groove wear. Three piece seals seal even better at low rpm the two piece and prefore equally well at high rpm. I have it on fairly good athority that mazda went with 2 piece seals for the rx8 (and replacment seals for the rx7) because threepiece seals extra machining steps nearly doubled the cost per seal, and the slight bump in low rpm compression wasnt worth the extra cost. I believe they are right,.. the difference in performance between two and three piece is fairly small.

Thoose are the 3 major types of apex seals used today. There was atleast one more type at one time, an early three piece seal composed of a main piece and two corner pieces on each side, but this does exactly the same thing as a two piece seal with twice the leakage, so it evolved into the two piece.

Basicallly in terms of starting ease and low rpm sealing three piece are the best, followed very closly by two piece, then one piece. All seals will have similar performance at high rpm (two/three piece seals still slightly outpreform one piece at high rpm) Mazda used two piece ceramics on the 787b, and they saw a ~2% increase in torque starting from low rpm, peaking at a little over 2% at 6k, and falling of to ~0% by 9k over one piece seals.

thats all i can think of right now.. heres the paper i got the r26b #'s from http://www.mymazdarotary.com/mazda_r...paper_html.htm

shelleys_man_06

Thank you for the info 86rx7. Are the different types of apex seals more than just speed dependent? Which of the three you mentioned is best suited for forced induction?

TRZ750

It is great to see a post like 86rx7 does as he has correct information, not like a number of the so called rotary experts that post on this site.

I will give you a little more background. 1st is any auto manufacture has a few main issues in developing an engine for the street (pure racing is an exception). 1st is cost of manufacture. Just because it is better it may not be used if the improvement is not worth the cost. 2nd is reliability for the general customer. This is where the engineers are constatly making chages because of warranty and customer issues. 3rd is EPA fuel milage. Commonly automatics can get better EPA restults, like the 8 does, because the engineers can design the shift patteren arround the EPA test.

As for the seals. The carbon seal was eliminated in production because of all three reasons. Cost, reliability, and low speed sealing for the EPA. The two piece was much better for all three. The three piece, like stated by 86rx7, benifit in sealing was probably not worth the cost.

Now for a few technical points. As stated by 86rx7, the springs are just for starting, so only issue is that they keep there springyness when hot. I had a rotary pickup. After a real hard 200 full throttle miles pulling my boat back from Lake Havasu with a strong head wind the engine started to have hard starting. When the rotor first started turning it would spinn freely, so the starter was able to spin it fast enough so centrifical force pushed the apex seals out enough fo the engin to start. Once running I still had good power and normal gas milage, so I drove it for a few months untail I had time to take the engine apart. When I did the only thing I found wrong was flat apex seal springs.

The apex sel width is an interesting issue. This is just like piston ring width on a 2 stroke. A wider seal does have more material to resist wear, but is also heavier. On a pistone engine at high rpm the piston ring is forced up durring the up stroke and reaches a high speed when ~ at mid travel. Then the piston and ring must slow down to a stop by the time the piston is at top dead center. The weight of the ring (& piston) has inertia and will want to lift from the bottom of the ring grove to the top. Well the primary sealing surface of the ring is both the face to the cylinder and the bottom of the ring to the piston. Normally compression and combustion pressure pushes the ring down and holds it on the bottom piston ring grove surface. If the ring lifts from the bottom, as it will do at high rpm, the gasses will excape and you will lose power. This is why racing engines have thin rings - so they have less weight and less inertia and seal at high rpms. A similar problem occurs in the rotary. The rotor never comes to a stop, but the tip with the apex seal does slow down and speed up. So a thinner sae is better at high speed. But a thinner seal is weaker and will wear faster, so may not be the best for the street. Now a ceramic seal is lighter and very strong, but very costly to make, so again not for the street.

Mazda did have an issue with the 74 to 84 generation 3 mm two piece seal that is interesting in this discussion. This may be the reason they tried the additional pieces. When new the apex seal and the rotor housing would, at times, not properly seal. (This is one reason the milage increases with break in.) That would cause some blow by gas that would cause localized heating of the apex seal and it would worp! A slight "U" shape that would not seal and the nww customer would have hard starting and other problems. They preventer this problem with a coating on the seal that would last through the break in period, but the coating cost money. SO the multi piece seal was tried to eliminate the problem at a lower cost.

The inner rubber seal, head gasket, was the main reason for engine failures back in the 1970s to 73 models. During warm up the aluminun will expand different than the steal and the rubber was being burned. The 84 and later seals had a teflon coating to resist the busrning. THen they moved the O ring from the AL housing to the Steel housing and is no longer a problem.

Enough.

TRZ750

Please forgive all the bad typing and spelling. I need to go and no time to fix.

shelleys_man_06

Where does one draw the line on picking the right seal for the right application? I hear the stock seals will fare well against, I guess, at the most, 15 psi of boost. I'm not sure; this is the first time I have ever dealt with a rotary engine.

rotarygod

The e-shaft is stronger than the older ones. Mazda actually published that somewhere. I wish I remembered where. No clue about the housings though.

shelleys_man_06

Don't feel like your're hijacking the thread, Charles . I started this because maybe someone here can put some serious input about properly sealing the engine, especially if we want to use FI, nitrous, etc.

I think the myth of nitrous and rotary engines is stemmed from using nitrous on 4-ports like the 13BT, and 13B-REW. I suspect bad tuning. Also, I believe Mazda stopped using carbon seals a long time ago.

I want to learn more about rotary engine seals, from the different designs to materials involved. I'm still curious about the spring rate in the apex seals, side seals, etc.

For nitrous, I think the stock seals will do just fine. I wish I had an FEA program to simulate the stress these seals face. If you wanted some extra insurance, not to mention an empty bank account, you could invest in the Ianetti ceramic apex seals. I hear they are the Rolls Royce of apex seals .

86rx7

Mazda still used teflon linners on the inner combustion o-rings after they moved the grooves. Also, for the rx8 they moved the grooves back to the rotor housing for a reason i'm not sure of..(possibly because of the groove wall corroding away when they were in the side plates on very high mileage cars / cars run without antifreeze or another corrosion preventer?)

Also rotor housings were coated with ptfe to prevent seal scoring during initial startup/breakin. this was switched to a graphite coating on the 93+ engines, the rx8 probably uses the same graphite coating.

The Main thing that let mazda switch from carbon seals was the micro porous chrome plating on the housings. When mazda was first working on the design, they were plauged with seal chatter problems making the housings unusable after a few thousand miles. This was solved by using carbon seals which were self lubricating. The first mazda rotary, the cosmo sport used these 6mm carbon seals. The draw back as TRZ750 stated was wear and low rpm sealing (from being one piece). Mazda tried a few things to use steel seals in the rotary, (seals with holes drilled in various places to change the harmonics of the seal etc) But what finally worked was a coating of chrome on the housings that had microscopic holes and grooves in it to retain lubrication. This let mazda use the steel seals we see today. but i digress..

For a street car i'd stick with two piece seals. All the seals preform similarly under boost. It's not horsepower that breaks seals, its detonation. FI it,.. just tune it properly and you'll be ok. I'd contact a tunner atleast for advice on tunning it. I believe the safe zone for previous FI rotaries under high boost conditions were AFR's around 10.5:1...but i'd talk to a few other people and get more opinions before you settle on a target AFR.

the only other problem i can think of is insuficient output from the OMP. Previous FI rotaries had slightly higher output from the OMP then there NA counterparts. You could probably reprogram the rx8 ecu, or trick it somehow to inject more oil, but the simple solution may be to add a little premix to make up for it. This is only really for long term reliability anyway..

shelleys_man_06

What is PTFE, and why did Mazda make the switch to a graphite coating?

Also, what kind of advantages/disadvantages would one see in single spring, or multiple spring designs (apex seals)?

BTW, Rotary Aviation has a good FAQ for their apex seals .

http://www.rotaryaviation.com/apex_seal_faq.htm

rotarygod

It was really a very simple reason why. They wanted the extra casting thickness between the seal grooves and the exhaust ports. If they kept them in the side housings there would have been less material between the water seals and the exhaust sleeves. Since the exhaust is very hot, unlike the intake, they felt that every last little bit of material was needed.

86rx7

That makes sense Rotarygod, good to know.

I was mistaken on the date of the change to graphite, it appears that it toke place in 1984. This document goes into why more in depth.

http://www.thecarricos.com/ACRE/Docu...tech84-1-6.pdf

Ophitoxaemia

>As far as rpm lomits I can say I have had my RX-8 to 11.5k rpm and it still runs great.

is there an easy way to defeat the rev limiter (which i think is set to 9.5K)? does the motor make power over 10K? and i would rev the heck out of my 1984 rx7 (way past the 8K limit of the tach) during racing and never blew the water pump (which is what i thought was the limit).

if accessory explosion limits the rx8 in revs, can we make more power with underdrives and a raised rev limiter?

thanks,

james

Ophitoxaemia

> It is said there is no substitute for displacement

no, there is, its called revs and this is where the rotary excels. i am pretty sure the renesis will be limited on a street car by the accessories, so a brief blip wont hurt anything. hold it there for long and the water pump will explode or cavitate (and rotaries dont like to overheat at 11K). electric water pump? maybe we have one now, i dont even know!

ok, i understand the 11.5K now. S2000s are pretty easy to blow up with missed downshift to 2nd instead of 4th. since on the freeway you need to downshift at least to 4th to get in the powerband, it is even easier to accidently hit 2nd in the rx8. done it myself :P.

if the intake and exhaust and computer can handle it, raising the rev limiter 1K should be a nice improvement in peak power, if not driveability.

james