TeamRX8 4-Port Renesis Street Turbo Concept Thread
 

So I've been bouncing this idea around for a while about building a turbo RX8 with a single turbo sized to produce a power range skewed mostly for maximum response and torque in the low-med rpm range rather than moderate - big power increase to a 9k redline. Which, this is the opposite of what most people do or want in general with an RX8. Understandable, since it will cost the same either way, so why not go for teh powaaaa ----? I had proposed on the forum here maybe two years ago, and of course all the chicken-little’s screamed the sky was a fallin’… :)

Well, all I'd want to achieve out of this is a super zippy street car that can slice and dice traffic by just stepping on the throttle without having to drop 3 or 4 gears with a high revving scream. Kind of the opposite of what an RX8 is, as supplied from Mazda, but the chassis itself is great for the goal if it just had the oomph. I’m fairly certain that this idea has the potential to offer an engine similar to the 13B-REW that came from Mazda in the 1993 Mazda RX-7 Twin Turbo (FD3), except with faster response, huge increase in low-end torque; both in magnitude and how soon it builds at low rpm, maybe 100 rwhp more (~330 - 350 rwhp is an estimate- only on E85 street fuel, "not a goal"), but limited to a max 7000 -7500 red-line limit. Other advantages are that a single turbo setup eliminates all the complication, transitional issues, and excessive weight associated with the FD3 twin turbo configuration. To me, an RX8 as such a well handling and reasonably accommodating 4-seater chassis, and fitted with such an engine, would have been the car that many people wished Mazda had offered back in 2003/2004.

After crunching a few common numbers, 350 rwhp on a 2-rotor engine works out to an approx. 47 lb/hr mass flow rate. Again, this power point isn’t the real goal, but we have to determine what the max range of the turbo needs to be for sizing it within the necessary power range. Of course the newer Borg Warner EFR series turbos seem to be the rage recently for providing great response and power, especially on rotary engines. To date for a 2-rotor engine most are going with the EFR 7670 or larger frame sizes, which everyone keeps telling me to use instead. Except they recommend this because they never bothered to compare in detail how the two mass flow maps for those turbos compare to each other as indicated below.

https://cimg3.ibsrv.net/gimg/www.rx8...e16cc8be65.jpg

Just the standard 7163 compressor flow map in lb/min mass flow rate

https://cimg3.ibsrv.net/gimg/www.rx8...9a06d9ebfc.jpg


Which when done as above, it becomes obvious that they’re fairly identical across a large part of the mass flow range. However, the EFR 7670 does have a larger turbine which provides the basis for it to also offer much higher boost range exceeding 25 psig or so over the EFR 7163, as indicated the upper LH area of map overlays. The thing is, that’s not where this engine is targeted to perform. So the larger physical size, high boost range, etc. of the EFR 7670 is not necessary because they basically correspond very closely where the actual operating range will be. For a low-mount position the larger size of the EFR 7670 is a big negative (next larger frame size in the EFR lineup), if it even can fit. Also, the smaller sizing of the EFR 7163 is known to have a corresponding increase in response over the EFR 7670, which is one of the primary goals for this concept. Some people said the EFR 7163 is a lot less efficient out on the far RH maximum flow range, which it is, but only by 2% - 4%. Again, we aren't intending to run at redline continuously i.e., not a track car. So this is a moot point to me.

On to the engine choice. While a few people have boosted the lower power 4-port Renesis, most choose the higher power 6-port. One of the primary issues with the 4-port engine is that here in North America it was only available with a junky 4-spd automatic transmission. However, the rest of the world could get it with a 5-spd manual transmission, which is also the same guts as the 5-spd transmission that came in the 1993+ RX7-TT FD3 chassis, but they just swapped out the bellhousing and tailshaft ends for RX8 specific replacements. This transmission is actually considered stronger than either of the 6-spd manual transmissions in the RX-8, and also a much better choice for a turbo engine that can spit out low-end torque to pull the car out of the hole rather than downshift a gear or two.

https://cimg4.ibsrv.net/gimg/www.rx8...4221192a0f.jpg

https://cimg7.ibsrv.net/gimg/www.rx8...5e0e1d3389.jpg


When Mazda designed the 4-port motor they altered the both the intake and exhaust ports location/port timing some to enhance the low end torque for it. The 4-port motor came with a 7500 rpm redline, while the 6-port motor with the two additional APV ports as designed for high rpm power with more aggressive port location/timing for that purpose to achieve a 9000 rpm redline. The 4-port motor actually produced a higher peak torque rating and more low-end torque in general as indicated in the Mazda information below.

https://cimg5.ibsrv.net/gimg/www.rx8...003935c738.jpg


https://cimg9.ibsrv.net/gimg/www.rx8...362a27bb59.jpg


So given the target goals of this concept, I decided that the 4-port motor is the better choice for that purpose. However, the lower-power 4-port motor also requires some part changes if it now needs to handle much higher power levels than it was designed for. As supplied by Mazda, some of the internal parts on the 4-port had their performance design level reduced to lower costs. The stationary gears, the e-shaft, and a few other parts will need to be upgraded to 6-port or higher level aftermarket parts for their higher strength properties. Also, we already had mentioned that the NA 4-port motor was auto-trans only in North America. The rear iron plate where the transmission mounts will also need to be replaced with the manual transmission iron plate, which you’d have to import from overseas. The simple solution to this is to instead buy a imported JDM 4-port engine, which typically can also be purchased intact with the 5-speed transmission; two birds, one purchasing stone. Currently I see them being offered on Fleabay for around $1400 for the combined pair, which isn’t a bad price really, but they also don’t sell very quickly so my suggestion is to instead make a lower offer to try to strike a better deal.

So I’m going to leave this off here for the moment and block a few pages for later revisions, but stay tuned because this might possibly develop beyond the concept stage if the moon and the stars align just right …


.

So if this was my personal car, this is how I envision implementing this concept in the engine bay. I'd want a V-mount setup with the radiator on the bottom V- and the intercooler on the top V- and venting out through a hole in the hood. Most pipe the the turbo to intercooler and then around to the throttle body on the UIM. You can see in the photo below, which is an FD3 RX7 of a racing friend of mine that has the usual V-mount/UIM setup.

https://cimg6.ibsrv.net/gimg/www.rx8...bfdd108c2e.jpg

https://cimg2.ibsrv.net/gimg/www.rx8...ad595b1e07.jpg


However, my approach is to eliminate the UIM, put a plenum & throttle body assembly directly on the LIM where the UIM had previously bolted to it. I'd also reconfigure the IC with the plenum tanks on the top and bottom, plumb from the compressor discharge to one IC intake plenum, then plumb the other IC discharge plenum to the throttle body on the LIM. This shortens up the piping quite a bit, especially when compared to front mount ICs that might be 3x - 4x more if not longer. This supports quick response from the time the compressor starts spinning up to to the to when it reaches the intake ports. There is much less volume to compress between the two end points that count. Some would argue that eliminating the UIM is going to kill low-end TQ. That would be true on an NA engine, but torque from this engine will come from the turbo being responsive and kicking in earlier than some people might suppose possible. That's how I see it anyway. It could prove otherwise, but until the photo below is a rough approximation of how I see the engine bay laying out with this concept.

https://cimg8.ibsrv.net/gimg/www.rx8...77e84ca5d9.jpg


My intention would also be to eliminate the thermostat housing assembly and convert to an electric water pump setup. What this does is to clear away the radiator hoses and T-stat body in that critical space area on the RH side of the engine and provides a freer path for the turbo inlet piping to pass through there. It would also incorporate a custom engine mount/bracket setup to on the RH side eliminate that turbo inlet piping limitation as well. The inlet pipe can more or less run straight up into the bumper opening area for fresh air.


.

Regardless of the negative comments from certain people, the general results have already been proven for a higher level of performance using the EFR 7670 turbo, which we already determined is pretty much the same flow map performance as the EFR 7163 in post#1 above. The difference is the EFR 7670 is the next frame size up for the BW EFR turbo lineup. It has a larger more free flowing turbine than the EFR 7163, which again we see that in the turbo flow map comparison where the EFR 7670 can run much higher boost levels than the EFR 7163, but that range/area of the map isn’t used or needed for what’s being proposed here.

The EFR 7670 results are also made using the earlier 13B with peripheral exhaust ports, which that engine generates much higher exhaust gas energy and can use that to spin up the larger turbine assembly well. The counter to this is the EFR 7163 being a smaller frame size and turbine assembly is going to help make up for the Renesis engines lower exhaust energy to provide, in my estimation, an equivalent response. The EFR 7163 turbine assembly is not only tighter and lighter, it’s a slightly different design than all the other EFR turbos due to it originally being designed and used in the Indy car race series.

It’s not just lighter than the EFR 7670 turbine, but when you compare the BW EFR turbos to an equivalent Garrett turbo, the EFR turbine wheel is generally about 50% lighter than the Garrett turbine wheel. This is one of the main reasons that the EFR turbos pretty much trounce a Garrett turbo in both response and how soon boost builds at lower rpm. The turbine needs to get up to around 120,000 rpm, so it only makes sense that if the weight is cut in half that it can get moving and up to that rotational speed a lot sooner.

If you want to compare the EFR 7163 to a Garrett turbo, it’d be equivalent to a GTX3076. Except they’re not really equivalent as the EFR 7163 will outperform the GTX3076 in just about every regard, including max power potential. You can see that here in his linked webpage where the EFR 7163 was dyno tested against a number of other Garrett turbos on a Subaru piston engine. The EFR 7163 proves to pretty much dominate in every test:

https://www.perrin.com/blog/post/bor...-test-and-tune


And then of course is this well know result of the peri exhaust port 13B using the EFR 7670 on TX7Club by member BLUE TII. Which again, this was using a higher power level target than I’d be going after, but generally speaking the low-mid range torque and power levels indicated here should all be within reach of the EFR 7163 as I’ve proposed. The low rpm response and TQ output is what we’d be looking for with the EFR 7163:

https://www.rx7club.com/single-turbo...esults-1070794


https://cimg3.ibsrv.net/gimg/www.rx8...9d27475ce8.jpg

saved3

saved4

saved5

Which there's a guy that has an EFR 7163 on an RX8 6-port and goes by the name "Punchy Lightbeer", which kind of matches up I think :lol:. He's blown maybe 4 or 5 Renesis engines because he guesses mostly. Before the EFR 7163 he had a EFR 6758 with 0.64 AR, which of course blew the engine. Guy seems to have the money and parts, but maybe not much technical knowledge about what he's doing. Sometimes his car goes full boost at 3400 rpm, other times at 4500 rpm, so something isn't setup properly. I don't like how he piped the turbo manifold either, plus he has a smallish cat converter, probably too restrictive exhaust overall, and lots of messed up stuff like that in general.

Some of his parts are shown in the vid below, but you can see more and some dyno runs with the inconsistent boost up etc. on his youtube home page. He did get about 290 rwhp on 9 - 11 psig boost range. IMO he's doing it all wrong so I personally wouldn't use any of this to judge what this turbo is really capable of or as anything substantial either for or against the proposed concept of this thread.

Here's what he wrote on the vid below, kind of tells the tale if you get the idea:




.

Here's a 4 port that made some power

This can become interesting.

Regarding intercooler, how come you dont look for a liquid intercooler? I'd expect it to be better on a "not race car", due to stored cooling for standstills and less volume between compressor and rotor to improve response. I guess you have considered it, and found it not to be the best option...?

I had forgotten about that one; still had the auto trans. A big Garrett turbo too; boost comes on way late then rockets up all at the end. I seem to recall they cut out the frame/subframe to fit it, but the manifold was decent except I also seem to recall they blocked the center siamese exhaust port and only fed it from the two end exhaust ports only. Makes it a lot easier to pipe that way, but I'm strongly opposed to doing that for a variety of reasons. Interesting, but sort of the opposite of what the concept of this thread is.

Thanks for posting it up.

edit, actually I went back and reviewed the dynos graphs in more detail. It does rocket up from around 3000 rpm, but it seems to be starting from a really low power point there; 35 - 50 rwhp. I can't seem to find any info on what the turbo specifications are?


.

After seeing your plan to use the 4 port the Pac car came to mind.I like your design for the placement of IC and short piping to and from IC. It good to see a new take on turboing the 8.Good luck with the project if you do it.

No, I hadn't considered that option at all. Generally speaking water to air means more weight, plumbing, another heat exchanger & accessories to cool the water down, more complications in general. On the surface it doesn't seem like a good idea to me, but I suppose I can look into it more

Yes, definitively some negative sides. For racing, Id say they weight to much on the negative side, and you usually don't stop a lot for traffic lights:) But for a DD, the energy density in your water gives you slower increase of temp, and the mentioned reduced volume between comp and intake port.

On the plumbing side, there are more components, but it gives advantages like considerably smaller and flexible hoses, so you don't have large pipes stuffing up your space, this mean its easier to give the radiator the best possible placement(V-mount..?), since you only are going to run relatively small hoses between locations. Together, this could more than equalize the extra plumbing.

That was my thoughts, at least:)

Well I didn't detail that there would be dedicated ducting on both the IC inlet and discharge. It's not shown in any of the photos, but essentially it'll be sealed off from the actual engine bay. Otherwise yes, there won't be any air flow through it when the car is stationary. I want to simplify things, not make them more complicated. I don't see myself using a water-air IC. Sorry.

Looks like there was a big thread here on the forum for the Pac Performance 4-port turbo mentioned above:

https://www.rx8club.com/series-i-maj...ackage-246821/


After reviewing the Mazda factory engine dyno graphs that I had previously posted above in the OP, which are flywheel hp, it looks like around 40 - 50 hp at the rear wheels around 3000 rpm is about right for the 4-port engine.





.

most of these are only valid if you have a difficult engine bay. The rx8 has such easy front mount or v-mount piping there is no need for complex piping. The length of piping on a front mount on the rx8 is probably half what you find on most factory turbo cars and has less bends.

Looks ok when compared to some renesis turbo setups but terrible when compared to an rew. Very narrow power band .They say in the thread 19psi to achieve those numbers .... I'm surprised they didn't need more boost than that . Must have been a fairly big turbo.
Was there ever any news after this dyno or did it fade away after an engine or trans failure ?

what makes the 4port a good choice for turbo?

is it more room from the simple inlet manifold or is it the port timing amd more scope to port the motor?

now with the new adaptronic the ease of swapping to a 4 port is now much more viable.

To me it seems a good choice for what Team is trying to achieve . The peak power potential is limited but using that turbo it would be anyway - especially on a 6 port.

dunno they are still making 400hp. from the pac web site "The Pac Performance Stage 4 upgrade is the same package as used on the Mazda Australia Targa Tasmania cars ...". You don't start with the low power engine in racing unless there is some benefit.

Im guessing it is the porting potential. The 6port has never been the choice for turbos for that reason.

the higher torque of the 4port would be in the port timing and the longer runners which will move (disappear?) with turbo porting.

This is going to be an interesting build if it eventuates.

Did they actually say they used a 4 port on the Targa cars ? Info is limited on their site . I get the impression they made that 'project 500' car , realised it was a lemon , and moved on.
If you have a look at a 4 port iron ............ the potential is minimal .

Edit : looks like they used a 6 port for the targa car.

Well again they used (or so it seems to me) a larger Garrett turbo, though I still haven't found the specs for yet either. So I don't think it correlates to this concept much. Other than they made some decent power numbers for some period of time to show my approximation of 350 rwhp is achievable if the rest of the system can support what it takes to get there.

Not sure I agree with the irons comment, but we'll leave that for another day ... just the same, I'm neither looking for 6-port power potential nor want port timing that supports higher rpm power range. I'm fairly set on this and believe it will prove out well if pursued ... we'll see ...

what you think has no relevance for me, but thanks for sharing just the same

Added more info in Post #3 above that was reserved blank previously

love this.

had a long talk with scott, in the early days.. i said if you make a low mount turbo that makes just 300 whp. and semi reliable i would buy one.

and here we are.

the rx8 platform is about as good as it gets for a hpde track car..

really dont think you need 300 whp if you could just make some tq.

have tracked a gti with a diff and 300 whp and tq. it was faster, but the rx8 is still more fun to drive/control..

hey, but what do i know..

beers :beer: