I have access to a CNC mill, manual mill, and many other machining equipment. I am in school for machining and think this would be a great experiment. Can i just make an exact replica of the stock Mazda seal, but one piece of high grade steel. Then send it out to be heat treated?



I know i can do this and get it to at least .001" or better on a manual mill then clearance it later by hand to fit perfect. But what about the edge of the seal that touches the housing? Is there a certain radius that it should be rounded to?



I wanted to try to make one set out of chromoly then send it out to be heat treated. But i think it might be to heavy. I think aluminum would never withstand the heat, even the high grade stuff. Any reccamendations on a material?

mazda has a special way of doing the sales.

i forget what it is but the seals that dont last long are the kind that are GROUND to size

but your using a cnc machine so its more of a fyi.

also there seals are made of different metals and this maybe a expensive experiment

why dont you try titinum and heat treat it and then cryotreat them as well?

I think I have the SAE paper somewhere. Basically Mazda cryotreated the seals using some sort of laser in order to get the temps down to near absolute zero. Yes I know it sounds counter intuitive, but it has something to do with bouncing the laser to pull heat away from the part.



Making a stronger seal is going to be hard to do, but you can aim to make a lighter seal which will allow you to rev higher.

Titanium is a bad choice. It expands greatly as it heats up, and does not wear well. It just tends to gall.



Sen2two, you wont be able to mill the seals to size and get an acceptable tolerance range. You will need to be in the ten thousandths range, and for that, you will need a surface grinder. Besides, the tooling marks from machining will create localized stresses and failure points.



The material choice and heat treating are what make or break steel seals. Mazda seals are cast iron thats nitrided on the wear surface. The cast iron is dimensionally stable over a broad heat range, and the nitride coating allows the seals to last a long time. If you wear through it, the cast iron is soft enough to not shred rotor housings. They are somewhat brittle though. Aftermarket seals range from soft steel seals that dont break, and harder steel seals that dont wear much, and can be rough on housings, but are very resistant to detonation.



Carbon apex seals are actually a suspension of carbon particles in an aluminum matrix. They are very light, and produce practically no wear, and minimal damage if they break. They are a bit fragile for a turbo engine though.



Piston engine rings are almost always cast iron with a wear coating. Its just proven itself as an effective means of sealing in combustion. Kinda hard to outsmart decades of the smartest engineers coming to the came choice for a sealing material.



Apex seals are not a good starting point on the learning curve.

I agree with the above. You could maybe consider starting with a pair of junked rotor PP housings? I always wanted to make some.

make the shaft!

The seals will be going into a junk motor. housings are on the line of usable and paper weight. plates, are just good enough to use also. Motor will be an experiment for a few things I'm making.If they turn out OK, They will go into my "good" motors.I will be going through with this. I'm not looking to compete in the market. Unless i stumble across an awesome design. These are really just for me. I enjoy making my own parts. There's a certain pride in that.



I have been doing a lot of research lately on what kind of material to use and i have a few ideas, as well as talking to my instructor about this.He knows nothing about rotary or apex seal design, but he has been a machinist for over 40 years. I glued a stock 3 piece seal together today and laid it out. I'm going to make a test one in aluminum just to see if i can get it into spec on a manual mill. I will also make one on the CNC.



Motivefab, I'm curious as to why you would say i cant get it within .010", I have gotten plenty on parts withing .001" with the very same mill i will be using to make these.And with the right tooling, speed, cutter, set-up, there should not be any tooling marks. Especially on a CNC. When done right, a CNC can leave a polished finish even on rounded edges.









My main concern is the contact area on the seal. I am unsure of what radius to give the seal where it meets the housing.

I'm working on something, not a new shaft, but mating the two together. What would that make? https://www.nopistons.com/forums/pub...DIR#>/wink.gif





I'll be starting a new thread on that soon enough. https://www.nopistons.com/forums/pub...IR#>/smile.gif

Not .010 tolerance, I mean closer to .0001"



I like making my own parts too, but some things are too specialized to be pulled off even in a well equipped machine shop.

I wasn't saying you were wrong. Just asking why...



I do have access to a surface grinder, but with my experience, they can't compare to the tight tolerance that a mill can give.I doubt that the seals need to be withing .0001" anyways. Looking at Mazdas tolerances in other areas, There's no reason why that one would be so close. Most people that build full out race motors still have to clearence the apex seals and apex seal slot anyway.



Either way, these will probley only see my motors, and maybe a few friends of mine. But that's really it.



If anyone wants to add any advice on doing this, or not doing this... feel free. I will not take it the wrong way.

um it might make scrap! the two rotor is easy, the rotors are 180 degrees apart. the muti-rotors are different as they actually need to run in a certain firing order to balance out



this info is actually in the rotary engine book.



not a big deal though, the shaft is still very simple, relative to the apex seal, which has to be hard on the wear surfaces, and soft on the sides

You'd also have to try and machine in the convex, thats also what all the aftermarket seals are missing to, without they sag in the middle.

What do you mean by this?



The rounded edge where the seal meets the housing? or something else?

Not if you mate the two together with the the 1st and 3rd rotor identical, making the 2d and 4th also identical. Yeah, it will sound like two 2-rotors, but it will still be a 4-rotor.



Also, it's been done. I will not be the first here. Little info is out there on it. But I'm pretty sure I can pull it off...

If you are hell-bent on making your own parts... why not make your own aluminum side housings? I've held a set of the RB ones... they are SILLY light. I could probably pick up a 3 rotor built from these and walk around with it. (the engine wasn't put together yet) As far as i know, RB does not make the odd housing for a 3 rotor yet either...





Just a thought......

Making plates, e-shafts, ect. will come in time. But for right now I'm more interested in making apex seals, The only other thing I have been working on is making housings. But every design I come up with for making the steel liner and chrome surfacing just dosnt seem right.



There are 2 or 3 companies that make the thick plate for the 3 rotor also.

something else, look in the FSM

I'll take a guess..



while measuring a stock Mazda 3 piece seal, I noticed it gets slimmer towards the center. Not sure if this is what you are talking about, or my seal was in bad shape?

The apex seals are concave in the center. They come to flat at operating temperature. The center of the seal is heated by the exhaust gasses. The Renesis may even be convex due to the exhaust gasses heating the sides of the rotor. But I have read nothing on that yet. The wear surfaces are hardened by exposure to electron beams similar to the welding process.



You could put a segment in a shadowgraph and get the radius.



Milling machines have become more accurate through the years but really close tolorance stuff is ground or lapped to finish size. When you get down to .0001" or less you are grinding or lapping to real close then hand selecting to an air gaged bore size for example. As in small 9 piston hydraulic pump. You measure as close as is possible then install the piston you hand select to fit in the bore and pressurize to say 100 PSI with the oil at 100 degrees, then count the drops of oil that leak past the piston (Called case drain)to calculate the clearance.



So a nine piston pump the size of a .30 pistole cylinder may be produced with 9 holes all different sizes. So you sit at a station and hand fit one of a big selection of pistons graded by diameter sitting in trays. You put in a piston and stick the cylinder into a fixture and pressurize it with oil. Then a laser counts the drops of oil leaking past the piston, and you are looking for 5 to 7 drops in 15 seconds. You get 9 drops. So you pull out that piston and select a larger diameter piston and try again. This is the kind of pump that powers the flight control rams (cylinders) on aircraft. They produce 3,000 to 9,000 pounds of pressure. They run off of the engine accessory section.



You could build pumps for years and never handle a part that was more than .0002" out of spec.



Apex seals have side clearance to prevent sticking in the groove. What happens when you burn sump oil in the combustion chamber. Too much clearance lets the apex seal vibrate across the chrome. Producing evenly spaced lines in the chrome.

A small number of heavy dents below the ramp is just the apex seal ski jumping the ramp from too many revs. Better oils than in the 70s and for those who mix there own top oil allows three piece zero clearance seals with little danger of gumming or sticking.



Lynn E. Hanover

The material I plan on using has a melting point of almost 5000 °F and it has one of the lowest coefficient of thermal expansion of the top metals. It also has a Brinell hardness of 1500 MPa. Which is more than double Titanium.

hows this project coming?