posted on August 06, 2012 02:01
Rotary Radness, Mazdatrix's 13B REW/MSP Hybrid!
By Mike Kojima
So far we have not focused too much attention on the Mazda Rotary engine on the pages of MotoIQ. This is not because we haven't wanted to but because no one on our staff has much expertise on the engine except for Eric and he has been pretty busy lately doing other things in the world of internal combustion besides messing on rotaries.
However rotary experts are around, some right in our back yard. Mazdatrix in Signal Hill California is just freeway minutes from the palatial MotoIQ HQ. Mazdatrix is a rotary specialist shop owned by Dave Lemon, a long time SCCA, IMSA and NASA racer/champion. You might have seen Dave's Turbo RX-8 race car before gracing our pages as a MPTCC and Time Attack racer. Mazdatrix is also the employer and sponsor of Formula Drift ace Kyle Mohan, who's RX-8 has been a top 10 contender in the rough world of pro drifting .
|Mazdatrix Boss, Dave Lemon can often be found tearing it up in MotoIQ's MPTCC TO class.
|Kyle Mohan, Formula D and Time Attack pilot is one of Madzatrix's chief wrenches during the day.
Building Rotaries is pretty tricky, although they are deceivingly simple, there are many tricks to building an engine that will have a good power output and a long service life. Rotaries are sensitive to tuning and detonation, something every turbo rotary fan needs to be aware of.
This leads us to our feature engine. The original turbo Rotaries available in the USA are the 13B T and the 13B REW. These engines used a port configuration of intake ports in the side housings with a peripheral exhaust port in the rotor housing. These engines achieved pretty decent power outputs, with good volumetric efficiency. However these engines were pretty poor in the emissions department mostly due to the fact that it was difficult to get rid of all the port overlap with peripheral exhaust ports. This allowed unburned hydrocarbons to blow out the exhaust on overlap.
Basically the rotary is like a two stroke and the "valve" timing is controlled by the shape and location of the ports in the housings. Port location and size is like camshaft duration and centerline in a conventional piston engine.
In an effort to clean up the Rotaries noxious exhaust, the 13B MSP Renesis or Multiple Side Port engine was introduced in the RX-8. The Renesis engine did away with the peripheral exhaust ports, instead putting them into the side housings. This did away with port overlap which cleaned up emissions but also limited the power potential of the Renesis engine.
Well along came Mazdatrix to find a power solution to wake up the 13B MSP. For their performance build, Mazdatrix keeps the multiported side housings of the Renesis engine and adds the peripheral ported exhaust rotor housings of the REW. This gives the power producing overlap of the REW with additional exhaust port area and generously sized intake ports from the MSP side housings!
The result is a turbo rotary with better VE due to increased intake and exhaust port area and reasonable port timing for turbocharged use. A win win situation for everything but emissions. This is a race motor though and really no dirtier than the 13B REW.
Check it out as we take a look at how one of Mazdatrix's hybrids are built.
|Dave Lemon with his creation, the 13B REW/MSP Hybrid. The engine requires a pretty elaborate exhaust manifold to accommodate the side and peripheral exhaust ports. This engine can really breathe and has a wide powerband. It is sort of like having a high lift but moderate duration camshaft in your naturally aspirated engine.
Monday, August 06, 2012 6:17 AM
Nice! Just two quickie comments though. First off, that intake port configuration is called a 6-port, not a bridge port. http://www.mazdarotary.net/porting.htm has an illustration of what a bridgeport is. The difference is that the stock 6-port engines, like the Renesis, have the upper port opened or closed depending on RPM, varying the effective port area and timing - on the older motors it moved the port closing from 30 degrees ABDC to 70 degrees ABDC. Same idea as VTEC, etc. A bridgeport adds a fixed "eyebrow" right by the edge of the rotor housing; this leaves a bridge of material for the size seals to ride on, while having the port actually opening sooner than is otherwise possible for side ports. Some bridge ports include cuts into the rotor housing too, making them open as early as peripheral ports - they came out of roadracing classes where bridgeports got a different weight break than full out peripheral port engines.
On the exhaust side... the interesting thing is that, at least according to Mazda's SAE papers, it's not mostly about removing overlap, though that did help misfire issues. But because the way the rotors spin around, the mixture charge gets concentrated by the trailing apex seal, and forms a layer that's too rich to ignite. On the peripheral exhaust engines, this just gets dumped out the exhaust port every cycle, which isn't great for emissions. With the Renesis style engines, because the mixture of unburned stuff is pressed right against the rotor housing, and the exhaust ports don't extend all the way to the rotor housing, it just keeps getting recirculated - some burns the next cycle, some gets added to, but it stays in the engine.
One of the things I find interesting here is how relatively stock the shortblock (as it were) is. There's various oiling mods that I do, and I know Dave Lemon does on some engines... although in fairness, maybe they're just not needed as much on the Renesis. Mazda tends to do various tweaks every new generation of engines.
Rotary exhaust is also a pain in the ass! If you do the geometry, the rotary exhaust port gains port area a lot faster than most piston engines (possibly it's possible to have cams with a higher ramp rate, but I can't say for sure) and there's nothing in the exhaust path to remove energy from it - no valves, no sharp turns, etc. So you end up with really really hot, fast, sharp pulses. Works great for turbos, but kind of a bitch to quiet down on naturally aspirated engines.
Monday, August 06, 2012 6:36 AM
That turned into more than two quick comments. Oh well.
Monday, August 06, 2012 6:47 AM
Very interesting article, and great reply by Kenku.
Anyone know how this hybrid setup would work on a NA engine? Not enough "lift" on the intake compared to a periphal port? Wider powerband?
Monday, August 06, 2012 8:25 AM
Well, in specific or in general? I'll speculate on both because, well, what the hell. ;) Also, for terminology, when I talk about a "port" I'm refering to the hole that determines the timing and flow area, not the rest of the runner.
In specific, this engine looks to have stock intake ports. On the other hand, for the vague class of port that they are (I think of things as either "street", "bridge" or "peripheral" with variations within each) they're pretty huge. I have not examined the Renesis's implemenation of the ports firsthand but I've built streetported racing engines (class rules) based on prior generation implementations of it... suffice to say that it takes a lot of work to efficiently go larger. If you were to make an engine that converts the two separate intake ports in the end housings into one big port, you would actually have to fill in runner area to have gas velocity be reasonable. So all that said... I think that an NA engine built like this could generate some pretty impressive numbers, especially if not hindered by OEM intake manifolding. I wouldn't be surprised by something close to 300hp at the flywheel if it was built to spin high enough, it was tuned right, and the intake manifold was designed properly.
In general, which is to say with lots of intake porting... it gets interesting. See, in theory if you have unrestricted porting on the old peripheral exhaust style engines, you have as much exhaust flow as you want and there's no reason for this sort of thing. In practice though, what would be called the runner on a piston engine has a steel insert to blend the port shape into the header shape on the street engines. It has two negative effects - it restricts what the port shape can look like, and more importantly it rapidly expands in area. This kills exhaust gas velocity (relatively speaking) and hurts flow for a given port size, but gives air injection into the exhaust stream time to mix, and makes things a bit less loud. So there's compromise... the steel insert shape sets an upper bound on flow through the exhaust port, but if you remove it the timing has to get *huge* to not have even worse expansion. Of note, some people (myself included) have done experiments making new steel inserts retaining the same area throughout, and the actual factory race housings had port area of similar size to the street engines, but without the steel inserts and with much smaller area exhaust ports... but those have been out of production for a while.
At any rate, in light of the actual practical limitations, there's some interesting potential with using the renesis and peripheral style exhaust together on a heavily ported engine in that now you can make your exhaust flow do things at points in the engine cycle that it couldn't before. I'm not really certain if those things would be useful or not though; theory vs. practice.
What really honestly bothers me is that (at least to my knowledge) nobody outside Mazda has a computational model to look at gas exchange and such in these suckers; everyone is operating by (sometimes educated, in some cases very educated) guesswork on timing and area, with the best people checking their guesses on dynos and being honest enough to not chalk it up to other factors. A couple builders have flowbenches to look at port and runner design, but far from most. There's not much hardcore development on NA power anymore; people running RX-8s (okay, they're tube chassis cars with peripheral ported 3 rotors, but anyway) in Grand Am, and drag guys. Can you imagine if piston engine development was done like this nowadays?
Monday, August 06, 2012 10:52 AM
Kenku, I called it a bridge port because this engine isn't exactly like the older six port motors that had a spool valve in the housings to control the opening of the ports. I should change it to a "bridged intake port". This keeps the side seals from snagging vs the bridge port NA style porting that is done to keep the corner seal in place.
Monday, August 06, 2012 11:15 AM
Kenku, do you have a list of specific oil system mods? I don't remember any besides clearances and staking the oil pressure relief valve.
Are there common tricks you do on older engines that have already been applied on the later ones?
Thank you for the comments! I do not have much rotary experience. I built a mild streetport motor about 20 years ago just out of curiosity so I could take it apart and see how it worked. I used a Racing Beat template for the porting. Dave helped me with parts (I don't think he remembers me from back then) The funny thing is he looks exactly the same.
I remember that motor as having a nice wide powerband, a decent idle and emissions so bad that people driving near my car would complain. I tried to smog test it and it made something like 10,000 times over the limit!
My old RX-7 mule ran the quarter mile in the low 14's which was pretty fast for back then, it was slightly slower than a stock twin turbo z for instance.
I sold the thing for a nice profit before I had to smog it! It was just a learning experience and I was not in love with the car.
Other than that I have zero experience with the engines other than helping my friends tune one or two.
Monday, August 06, 2012 11:29 AM
Ok, dumb questions from a rotary-illiterate moron. What materials are used for the rotor, side housings and rotor housings? The rotor and rotor housings look to be aluminum with cast iron side housings. amirite?
Monday, August 06, 2012 11:39 AM
The rotor is cast steel, side housings are nitrided iron and the rotor housings are aluminum with some plasma sprayed or chrome plated (not sure exact process) liners.
Monday, August 06, 2012 11:47 AM
Am I way off base here, or is assembling a rotary motor a lot easier than a piston one?
Monday, August 06, 2012 11:49 AM
I checked, it's a plating not a spray. A rotary has some easy part and hard parts. For full details I suggest buying Mazdatrix DVD's on how to assemble engines.
Monday, August 06, 2012 11:54 AM
No worries; never sure where the line is on "inform" vs. "telling people what they already know" so there goes the wall of text, heh.
Side seal life is actually one of those fun, non-obvious things you have to contend with on things with big side intake ports, even "street" ports. You end up wanting to open the port as early as possible for breathing, but there's a hard limit on what you can do without one of the side seals dropping into the port, and if you go that far you have to shape the port closing edge to bring the other side seal back up gradually enough you're not breaking bits off of it. It takes a little artistry, but done right you can have a fair amount of the corner seal hanging in space over the port and still have a reliable engine.
http://www.mazdatrix.com/sccaportrule.htm is actually neat to look at; it was put together for the aid of SCCA tech inspectors on a class that only allows "street" ports. It's not confirmed (no dynos required) but it's believed Dave had the highest power RX-7 in the class.
Oil system mods... the first and most obvious is that the later engines (FD and RX-8) have a groove around main bearings to feed multiple holes or windows in the bearing shells themselves; everyone uses those; the Renesis stationary gears are also more hardened compared to the stock ones, apparently. Also sometimes the rear portion of the rear main bearing seizes to the eccentric shaft - actually lost an engine on a dyno to that. Mazda published a competition prep manual for the IMSA racers in the late 70s/early 80s that has a lot of little things, including grinding that part of the eccentric shaft down by, off the top of my head, something like 20 microns. That was hard to come up with a way to do, and actually the Renesis eccentric shaft now has it stock. The later engines (FD) got a larger oil pump and a higher pressure rear regulator; they're largely identical to the old race parts but a lot cheaper. Unfortunately the 2nd gen or later oil pump doesn't bolt to the 12A or 1st gen 13B, so we're stuck with the GSL-SE oil pump. The 2nd gen and later eccentric shafts have a wax bypass thingie on the front that... does some emissions function or something by venting oil. Mazdatrix has a plug, and the earlier eccentric shafts just had metal there, so they used to be the preferred bits by dint of being a little bit stronger. Also, the eccentric shaft has jets to spray the inside of the rotors - in stock form they have check valves and pressure regulators and crap, but those can plug or fail, so the racing fix is to fit fixed jets with just a 2mm hole. Mazda sells those as is now, but the old comp manual tells you to cut some bits off of the stock parts and put a Weber carb air corrector in it. I also deburr the castings where oil exits on general principle.
Getting more exotic... the oil system moves through some really weird channels on the engine. On the first pic of Page 12 you can see where the oil pump is... the pickup at the bottom, the machined part with epitrochoid entrances where the pump sits, and the cast runner where the pressurized oil exits the engine. In the stock form, the oil then goes up and forwards out that machined part to the front cover, where there's an oil relief valve and a bunch more 90 degree turns. What you can do though is drill out the brass plug staked into the oil gallery, thread the outside of the iron where that meets the outside, and put an AN fitting adapter there. I also ream out the oil gallery a bit oversize and because I want to be 100% sure what the finish inside it is like. You can also port the oil pump ports a bit too to increase flow, but that's chancy and it looks like the Renesis improved those.
Next there's all the external stuff before the oil comes back into the engine. I use external filters; Fram HP1. What the hell, it's cheap insurance. Now, on the old engines, oil used to come in at the bottom of the rear iron, split between the pressure regulator and the oil filter pedestal, then split from there to go to both the rear main bearing and to the front main bearing via a route I'll get to. Since the filter's outboard, an aluminum block goes there that feeds both passages. The passage to the rear main bearing gets reamed oversize in my standard paranoia.
Now, to feed the front main bearing, oil normally goes forwards through a series of dowels in the rotor housings, takes a 90 degree turn down a drilled passage in the front iron and then another 90 degree turn to a drilled passage to the front main. On page 12, note the dowel to the right of the eccentric shaft - that's not really related in function, but happens to be where the last 90 degree turn is before the front main. So, same procedure as the front for the oil pump exit - OEM blockoff plug drilled out, threaded, AN fitting put in. A -8 or so (it's been a while since I made a new one) line connects the oil distribution block at the rear to the front main bearing passage.
That there, in synopsis, is about every oil modification I've seen in pro built rotaries, or thought up myself, excepting a full dry sump system or an accusump. Unless I forgot something. It may seem like I have a pathological hatred for 90 degree turns in drilled passages... well I do. And in all honesty, I'm not sure that all of this is 100% necessary... but I do know that at some point in the time my dad and I have been doing this, we stopped having bearing failures completely aside from the time on the dyno.
Monday, August 06, 2012 12:06 PM
Materials... the rotors are actually cast iron; I've machined them and tried to weld to them. Side housings are nitrided cast iron and the rotor housings actually have a sheetmetal insert with a complicated sorta-sawtooth profile to lock it to the aluminum part of the housing when it's cast - the really old ones, before they did that, sometimes dropped out after it was heat cycled enough and that hurt cooling. The inside coating is just a "simple" hard chrome.
And yes, they're dead simple to rebuild. You have to check various clearances obviously and file stuff to fit, but there's not too much of that. In my experience, the eccentric shaft and stuff does not have appreciable tolerances such that I have to measure bearings - I do anyway most of the time, but that just means throwing out an e-shaft. You put vaseline (or assembly lube I guess, whatever) on the seals as you put them into the engine to keep stuff from dropping out and... well yeah. Only tricky part is the center iron - you have to thread it around the eccentric shaft, and to get it in place the eccentric shaft actually has to be lifted a little. The trick comes in that there's a little corner section of the apex seals that are under spring pressure - if you lift the eccentric shaft crooked, the rotor comes up too and they go flying across the shop, and off the top of my head you can only buy them separate from the apex seals for the 12A. You can do it as a one man operation with a floor jack under the e-shaft front though, or if you're kind of limber, brace your foot on top of the engine stand and lift the shaft with your thigh/knee. Fundamentally though, on something where the clearances are set, it's all clean and so on, I can go from a pile of parts to an engine in a couple hours by myself without opening any manuals - I do cheat and write torque sequence and amount on the rear iron with paint pen though.
Monday, August 06, 2012 12:19 PM
Thats a nice list Kenku
I've only replaced the wax pellet with a solid one, FD OPR, and 3 window bearings on the stationary gear for my oil mods since its also a street car. I always tell people if they are getting an rx7, to always replace the thermal pellets. A failed pellet will cause you to lose 15-30 psi in oil pressure and will only squirt oil to the rear rotors if any.
Monday, August 06, 2012 1:06 PM
Kenku, you need to do some rotary articles for us!
I sort of remember that the corners of the apex seals are glued in place as an assembly aid and snap off after assembly?
I remember the old motors, particularly the 13B, you had to get hardened stationary gears and put snap rings on the gears on the rotors to prevent them from walking.
I think the new motors are not so prone to these problems?
Then there is the whole subject on the different apex seals with the pros and cons.
You really should write something for us! This article isn't intended as a how to in building a rotary, just as a guide for what Mazdatrix does.
Monday, August 06, 2012 1:48 PM
The pictures of my shop are terrible though! Actually I'm flattered, though right at the moment I'm having all my time and money disappear into wedding stuff. I was going to do a turbo motor in a tube chassis Miata as a next project or something, but then *someone* started a series on building the GD chassis Impreza (which I'm using for my DD) and now it's tempting to throw money at it first, heh heh. Although if I get something sold, I really could pick up a turbo engine core for reasonable money...
Moving on... yeah, the corners of the seals are glued from the factory. The theory is that they come detached at startup, but sometimes they come off before that. And you can reuse seals most of the time, but CA glue doesn't necessarily always stick to used cast iron as well as people might like. There's nothing like having a little piece of cast iron skitter across the floor and having to call Mazda at the last minute because after a couple hours of looking you can't find it.
Honestly, of all the things I've seen go wrong (and there's a lot - dad and I do this for fun, but for a while he had trouble with watching gauges) I've never seen the rotor gears walk out. One of the probuilt engines I tore down had some of the spring pins locating the stationary gear replaced with cap screws, which I thought was kind of clever. Of course, now I'm going to have one fail. The big reason why I haven't bothered of yet is that so far I haven't come up with a good way to get the rotor gears out. I have an idea on how to do so; it just means spending a lot of time on the lathe making a tool.
Apex seals... for NA usage, I've always seen the selection as pretty easy. There's the carbon/aluminum ones for more frequent rebuilds but the ability to survive higher RPM, OEM steel ones for longer life span but some worries about killing housings at high RPM, or ceramics if I'm certain everything is going to be okay indefinitely and I'm not going to be opening the engine. I may be biased by Mazda's contingency/support program though, which gets dealer pricing on the OEM seals.
Monday, August 06, 2012 3:24 PM
Great article, and I'd love to see more rotary stuff from motoIQ. I'd really love to know the real performance advantage of running this exhaust port layout. While people are pointing things out I think there's a few things I can add.
Big one, those are not REW rotor housings and are instead 84-85 GSL-SE housings. The only somewhat modern 13b rotor housings made with the water seals cut into the rotor housings instead of the iron side housings were on the GSL-SE (Besides the 13B Renesis of course). These housings are hard to find and did not have the modern chroming process applied to them. This makes finding usable one's even harder to find and even in the build they are using used housings.
As for gluing seals, you only need to glue the corners if you are assembling a engine with a half bridge. This is because you have to put the apex seal in upside down so that the corner is on the opposite side of the bridge port.
Locking the stationary gear is only necessary if you plan to rev the engine at or above 10K RPM. This is usually only necessary with NA peripheral port engines as they are the only motors with a power band that can extend that high. This is to prevent the excess vibration from destroying the stationary gear.
As for Apex seals, everyone has a preference, but the more common seals being used today are Rotary Aviation Classic and Super Seals, Goopy performance seals, and ALS ceramic seals. It really depends on what you're looking to do with your engine.
A softer metal seal (like RA Classic, Atkins, and Mazda) will take less time to break in and will generally produce better compression which leads to more power. These seals are also not as abrasive on the rotor housing which will lead to long life. Problem is that there is a greater risk of the seal shattering during detonation.
A harder metal seal like the RA Super Seal is more resistant to detonation and therefore will take more abuse without breaking. The drawback is that they will generally eat the chrome on your rotor housings over time and take a much longer time to build compression during break-in.
The Goopy seals seem to be a option between the two and have become very popular in the community.
Cermamic seals are the best of both worlds. They build great compression, last long, and don't eat the chrome off your housings. Problem is that they are usually 3-4 times the cost of steel apex seals, which depending on your budget might not be a option. Typically, if you're looking to replace apex seals your engine is blown and you are going to be spending a good amount of money to rebuild it. Asking someone to cough up a extra grand just for the ceramic apex seals doesn't usually go over well.
Monday, August 06, 2012 3:40 PM
GSL-SE housings are still available new. I have no doubt whatsoever that the new production ones are using the same chroming process as everything else. They're cheaper too! The exhaust ports are a bit different than the later engines though; one engine I'm building I had to hit the exhaust area with a boring bar to fit some inserts intended for a REW. In theory you could just cut water seals in REW housings, but why bother?
Perfectly possible to have a street port or bridgeport making power up in the 10k RPM range, just not many racing classes where you get to that you can't just run a peripheral port, at least in this country. There's a class in Aussieland called Improved Production that allows bridgeports where supposedly they're spinning to the 13k range. Depends more on intake and exhaust tuning than anything. I still have never seen a rotor gear walking out.
Monday, August 06, 2012 9:17 PM
this article satisfied the rotary fan in me. though i still scratch my head at them, i find them very intriguing and enticing. i hope to see more rotary content here too.
Monday, August 06, 2012 10:38 PM
What do you guys think about the ceramic rotor housing liners?
Tuesday, August 07, 2012 6:35 AM
If you're referring to replacing the inner layer with a cermet instead of the hard chrome... in theory I like it a lot, especially since it means that now the rotor housings are repairable. Mazda stopped making 12A rotor housings so that's the only way to get replacements I know of if they wear out, aside from scouring ebay or finding a lucky junkyard motor... and they're bound to stop making 13B housings eventually too. Mazda also had great results with it at LeMans in terms of reduced friction and wear.
In practice, last time I looked into it, the only company I could find doing it was just relatively unproven, and looking at it right now they charge more to replate the housings than Mazda does for brand new ones. If I had factory race housings or used up my store of 12A ones I'd be thinking about trying it though. Actually, I wonder if Dave Lemon has any experience with them; Mazdatrix is one of the few places that actually seems to have the will and ability to test out weird things, like the titanium rotors they're working on...
Tuesday, August 07, 2012 11:05 AM
i love reading about rotaries. the combination of side and peripheral ports seems pretty neat.
what's the small hole above the intake in this picture? http://photos.motoiq.com/MotoIQ/Tech/Mazdatrix-Rotary-Build/i-XqmB4jV/0/L/jsn-046-L.jpg
Tuesday, August 07, 2012 11:17 AM
Oil injection port. Engine oil is injected in very small quantities to lube the apex seals; it's one of the reasons why people advise against synthetic oils in rotaries, as it's believed to not burn off very well. Some synthetics also don't transfer heat very well, and the rotaries do a lot of heat transfer via the oil.
Tuesday, August 07, 2012 11:30 AM
Ok. I knew about the oil injection but i always thought it was mixed with the intake charge.
about the synthetic/dyno juice argument. would it not be possible to split what the injected oil is fed from? there are some seriously nice synthetic oils now that would probably help the motor if it wasn't forced to be lubed from the same stuff as the apex seals are?
Tuesday, August 07, 2012 11:36 AM
The older engines it was just mixed in with the intake charge... but Mazda started putting it into the rotor housings too. Seems to work.
There's lots of ways to split where the injected oil comes from, yeah... don't know how many are off the shelf though. The easiest method is to just run 2-stroke oil premixed in the gas. I personally am just... cautious about synthetic oils at least in the race motors I'm in charge of. We've tried it once or twice, and lost the engine due to bearing failure the same race weekend the synthetic went in. Correlation is not causation, there may be other factors at work, etcetera... but it's put me off the idea of experimenting with oil.
Tuesday, August 07, 2012 11:40 AM
yeah, it's hard to pin the down the cause.
personally i landed on the other side of the fence with our lemons racing. we haven't blown a single motor since we switched to Klotz. this kind of track record is pretty hard to ignore but you're right, that does not mean that it's the oil that made the difference.
Tuesday, August 07, 2012 11:47 AM
Besides, at least in my case... 5-6 quarts of GTX 20-50 dino oil vs. 5-6 quarts of superfancy synthetic every race weekend... it's a little bit of cost savings. Okay, possibly I don't have to change the oil every race weekend, but... enh, have you gotten the idea that maybe I'm being a bit conservative?
Tuesday, August 07, 2012 11:50 AM
see, it's all perspective, $7/quart * 6 quarts = $42 for a race weekend. much cheaper than swapping and/or rebuilding the motor. even a cheap lemons motor.
anyways. this is off topic, bring on more wankel magic!
Tuesday, August 07, 2012 3:00 PM
I've never seen more rotary information in one place before.
Im going to have to devote some time to all this wealth.
Tuesday, August 07, 2012 9:56 PM
GREAT STUFF Mike!
Haha, I suspected those ports were shaped that way for timing purposes! But there was still stuff that blew my mind so I'm only partially rotarded (wah, wah, wah) :-p
Wednesday, August 08, 2012 9:30 AM
does anyone know how much this hybrid combination would produce in naturally aspirated trim with no porting? just the 6 intake ports and the 5 exhaust ports?
Wednesday, August 08, 2012 9:41 AM
Precisely? Nope. Maybe around 280 at the flywheel with OEM intake manifolds; Mazda puts a lot of effort into tuning the NA manifolds and as a result they end up pretty optimized for the engine configuration they come on. That's also a pulled-out-of-my-ass guesstimate.
Non-OEM would let you retune the intake system for higher RPMs, which the intake ports should be able to handle... but that gets into a "how much do you want to spend to spin at high RPM" thing, as safely going past something like 10k adds the need for some bits to support it. With everything else in the engine optimized for it, you could theoretically make power out to very high RPMs, but in practice I don't know of anyone who has done it who will talk about numbers.
Wednesday, August 08, 2012 9:48 AM
i guess i wasn't too worried about intake and exhaust manifolds. it just seems that this hybrid gives a whole lot of port area without requiring porting and would be a good way for someone that wants high N/A power to get started without knowing all the ins and outs of porting a wankel.
obviously that exhaust manifold will be tricky, but i'd feel way more confident doing that than shaping my own ports in a 13b
the intake obviously hasn't changed from stock so any extra power would just be from spinning faster or a better intake manifold. but the power already gotten from this motor in stock trim makes you wonder how much extra is available with just the extra exhaust ports.
Wednesday, August 08, 2012 10:00 AM
It all depends on what you call "high N/A power". If you're thinking about it from a "get started" sense, this setup isn't great - there's not even bad off-the-shelf headers for it, and I can't really call a stainless header (mild steel dies pretty fast at rotary EGTs) a DIY candidate that's likely to turn out well.
Honestly, for something other than the car it came in, the Renesis is a pretty high power N/A engine. They can reliably do 250ish HP at the flywheel with no emissions and a standalone, and do so in Star Mazda trim. That's pretty hard to do with an older 13B with just a street port.
Wednesday, August 08, 2012 10:05 AM
"getting started" as in i have a bunch of experience with piston engines. tigging a stainless exhaust manifold is pretty trivial for me and so is diy fuel injection. (I'm the idiot that shoved a WWII radial engine in a 2nd gen MR2) but the wobbly triangle engine insides scare me a bit.
Wednesday, August 08, 2012 10:20 AM
Aha, we're dealing with specifics instead of generalities - that certainly changes things a bit. ;) No offense intended, especially with as terrible as my first try at a header was.
A thing to watch out for is that the RX-8 apex seals were redesigned with less height - they didn't have to go over any holes in the rotor housings, so they could eliminate some beam stiffness. And the RX-7 rotors probably will have the side seals drop into the RX-8 side ports... so you'll have to have someone cut the RX-8 rotors for the older style apex seals, but that's pretty common anyway as the RX-8 rotors are the lightest and highest compression OEM bits. Using the GSL-SE (1st gen RX-7) rotor housings like Mazdatrix should work, but there's a rumor Mazda changed dimensions of the rotor housings very slightly at some point, so checking clearances is important. You can google for the mazda comp manual if you want an idea on what clearances go into the engine - the exact specs will be different for a Renesis based engine but the specs remain.
Other than that... I'd really want to tune it on an engine dyno and play with header lengths and configurations, but y'know.
Wednesday, August 08, 2012 10:26 AM
on the apex seals. it seems they make 1 piece, 2 piece and 3 piece units. i can see why street applications would get the units that could seal better but aren't the dropping seal issues solved with the single piece units? seems like the only drawback would be a slightly worse seal due to the lack of tight seal to the side plates which may affect emissions but i can't see it affecting power significantly.
yeah, header configurations could get tricky. since the peripheral port opens first you may be able to induce a bit of a vacuum on the side ports with the proper configuration but then what do you do with that middle exhaust port. it seems like with the merged port from both sides it significantly complicates exhaust shape tuning possibilities.
Wednesday, August 08, 2012 10:36 AM
3 piece seals were something Mazda tried to do in the past - the main seal is split into two parts, and the end is the normal triangular part. They're not available anymore, so it's just 1 vs 2 piece, with the advantage/disadvantage you guessed... well, there's a little less low-end torque and stuff too, but big deal. One piece seals are handy in bridgeport engines because the triangular part can fall out otherwise.
The issue is that the side seals (on the flanks of the rotor) were moved on the RX-8 rotors, which is one of the reasons they could make the ports a bit bigger. Did you see that page I linked a bit before showing view out through the ports of the side seals? The RX-8 side irons were developed with the RX-8 rotors in mind, so the edge of the ports can be a touch closer to the rotor housing than you can get away with on the RX-7.
Maybe. I admit I haven't intensively checked compatability between the things.
Wednesday, August 08, 2012 10:47 AM
oh very neat. i hadn't looked at that link before.
some of those ports look seriously sketchy from a thermal expansion point of view.
Thanks for all the info!
Wednesday, August 08, 2012 10:54 AM
Not to be blase about thermal expansion, but there's water passages through the irons, and they're cast iron. On something ground by hand (the ports that is to say) I can't see tolerances being held well enough that a (off the top of my head) 0.7mm thick seal will stay or drop out depending on if the iron is at operating temperature or not.
And sure; I enjoy this stuff, have thought about it a lot, and have learned a lot the hard way. I'm not anywhere near as expert as some of the big name builders (Dave qualifies in a major way here) but I don't have anything to be hurt from sharing knowledge - besides, jawing about this stuff is fun!
Wednesday, August 08, 2012 10:57 AM
i was referring to the last picture: http://www.mazdatrix.com/scca/prodportrulespics/Bridge.jpg
where i could see the tab that sticks out curving inwards or outwards depending on the temperature differential of the tab itself.
which part of the country are you in? do you race lemons?
Wednesday, August 08, 2012 11:00 AM
Being rotarded, how does the 787B make so much power? I realize it's a quad rotor and has variable length runners, probably running some crazy fuel. Oh, and the fact that it's a full on race motor. But are there other tricks in there that could be used on a street motor?
I wish I could have torn that thing apart while I was in Hiroshima...
Wednesday, August 08, 2012 11:20 AM
Marc: Oh yeah that... well, all of those are for SCCA EProd, which is a reasonably large budget class. So for the national championships, it's far from unknown for people to have qualifying motors that resemble hand grenades in concept. Come to mention it, Dave Lemon's personal motor had that reputation, even it if wasn't extra-legal. As for location, I'm in Wisconsin. Haven't done LeMons - dad's been doing SCCA-type stuff and I don't have the time budget for a LeMons car. Although, come to think of it, I've been staring at Subaru WRC style antilag stuff for a while now, and I do have a $50 Holset HT60 kicking around... that's a terrible idea.
czubaka: The 787B used a 4-rotor using peripheral intake ports and exhaust ports. Peripheral intake ports are actually hellaciously efficient - they're a big damn hole with no bends or changes in area or nothin'. If you look at how the seal moves over a peripheral port too, they never actually close (they just start feeding into the next combustion chamber) so the air in the intake tract never has to completely stop. The only reason they're not used on every single rotary engine ever is that they are not so great at part throttle efficiency and low load situations... like pretty much all normal driving. They're also rather intolerant of backpressure. Fuel was reputedly just unleaded pump gas on the race stuff - I think the ACO controlled that relatively tightly.
It did have a few tricks. The wear surfaces were cermet coated for less friction and better wear, it had a 3rd spark plug per rotor for a bit more combustion efficiency, very optimized exhaust (I don't know if they had to run mufflers) and the intake trumpets could vary in length to acoustically tune across about a 5k RPM window. Really though, some drag guys have bettered it in power-per-rotor.
Wednesday, August 08, 2012 11:27 AM
you're pretty close, i'm in Indiana. we have a race in joliet, IL in about 10 weeks.
we'll be running an MR2 and an accord but I've got a chassis that would love to see some wankel love if i had some help.
maybe i can convince you to just come down for the race to have a fun weekend even if you don't drive? or we do have a spare driver's spot open if you're interested.
Wednesday, August 08, 2012 11:36 AM
... LOL. The Autobahn event the weekend of the 12-14th of October? I'm getting married / going on my honeymoon that weekend... thus why the "not much time/money budget this year". I'd sure as hell be game to at least spectate otherwise.
Hmmm. I'll have to go through my spares/not-quite-junk pile. I'm pretty sure I have enough sitting around for a hot 12A if I cut a few corners.
Wednesday, August 08, 2012 11:47 AM
yes, that event. congrats on getting married!
get a hold of me if you're interested in mediocre lemons fame with your rotary knowledge (Marc at marclabranche dot com)
Wednesday, August 08, 2012 2:08 PM
I wonder what would happen if you put reed valves like a two stroke on a peripheral intake?
Wednesday, August 08, 2012 2:27 PM
I'm not an expert on reed valves, but ignoring flow considerations, wouldn't they shut at negative pressure waves and kill acoustic tuning on the intake side? That'd seem like a pretty big hit to swallow at high RPM, although it would kill off reversion and stuff at the low end of the powerband, wouldn't it.
Makes me wish I had an engine dyno; I actually have a reed valve assembly that flows enough to be worth trying (it's from a post-WW2 US Navy copy of the German V-1 pulsejet) and a peripheral port engine sitting around.
Monday, September 17, 2012 1:52 PM
Kenku, when you tried synthetic oil, what were you using to lubricate internally? I see a lot of the FD guys converting the OMP lines to a separate reservoir so they can use cheap ashless 2 stroke oil, which is actually designed to burn in a combustion chamber. Then you are a bit more free to select a crankcase oil. I dont know this for a fact, but I have been told at that point you can use a synthetic in the engine, and that will generally help with fuel dilution issues.
Sorry to come into this topic so late.
Monday, September 17, 2012 2:08 PM
No worries; 'tis why I have email notification on. ;)
On the race car in question, and frankly all that I've done lately, we just mix 2-stroke oil in with the fuel. Honestly probably way more of it than is necessary, but hey, doesn't seem to hurt anything except for occasionally being a little smoky. Like I said before, I can't definitively blame synthetic for the bearings going away, I just don't want to mess with it when the non-synthetic keeps working. Maybe when we're to the point of having a decent datalogger watching oil temperature and pressure I'll be a little more willing to try thinning the oil or running synthetic or something but in the mean time there's bigger fish to fry.
MotoIQ Proudly Presents Our Partners: