Brettus turbo 111 (the ultimate Renesis turbo ?)
 

As the title states ............ this is my attempt at making the ultimate Renesis turbo setup .

I've been running the Greddy system for about 6 years now and learned a heap from maximising that setup (See Brettus turbo and Brettus turbo 11 threads) but figured it was time to do something of my own design.

The engine : will be using my current engine . This engine has been doweled and had rx7 deep apex seals fitted along with increased oil pressure. However , once the turbo system is sorted I plan to refreash this engine as it is down on compression .

The Turbo :
This is a Garrett GT3582r unit . I chose this turbo after considering all the major brands . The main reason I went with it is because of it's relative compactness plus it had the options I was looking for . Other Garrett turbos I looked at ,that would also have fitted, were the GTX3076r, GTX3576r , GTX3582r . I settled on the standard 3582 for a few reasons :

*Discarded the 30 series when I realised they used the same turbine castings and CRHA as the 35 series so size is the same.
*It's more efficient than the smaller X76mm wheel in the range I want to operate it .
*With the design i'm looking at spoolup should be ultra fast , and I figured the smaller 76mm wheel would certainly go into surge without the benefit of a ported shroud. A ported shroud wont fit in a low mount design.
*The GTX3582r ,I felt, was just overkill for my whp goal .
*The cast compressor wheel is said to be more durable than billet when using pre turbo W/M injection.
*It's a lot cheaper than the GTX series

The Brettspeed manifold design :
This is where I'm hoping to make some inroads into making the Renesis turbo a viable alternative to an REW swap .
*Lowmount
*The turbine housing is a 1.01A/R twin scroll fed from the end ports.
*Wastegating will be a single 38mm wastegate through the siamese port .

I'm sure this will raise some eyebrows and create some scepticism . The system has been designed around making this work and I'm excited to see if it does .

Goals :
Reliable 400-450whp
16 psi by at least 3500rpm
Factory drivability


https://cimg0.ibsrv.net/gimg/www.rx8...15a7e564fe.jpg


Turbo on dummy engine .



https://cimg1.ibsrv.net/gimg/www.rx8...916b34c790.jpg
Edited for 4 port solenoid valve


A little more on why i went down this path ...

The 'Achilles heel' of the Renesis when turboed has always been the poor exhaust port flow. The siamese port is particularly poor and I have thought hard about how to improve on that . I even did tests on my last turbo setup where I removed most of the divider in the port. That test proved fruitless.
It occurred to me , why not take the siamese negative and turn it into a positive. The design above wasn't the first one i came up with by any means (and might not be the last) .
I see it as having huge potential it if does work , for the following reasons :

*unparalleled spoolup/response due to very small air volume in the manifold and the twin scroll turbine.
*large turbine housing to reduce backpressure and minimise the amount of air that needs to flow through the wastegate.
*Ability to get more flow out of the siamese ports than previously achieved due to the fact that backpressure at the wastegate can be close to zero if necessary, thereby allowing the siamese port to evacuate gas more easily . It should flow what it needs to flow (up to a point) to achieve the necessary pressure at the turbine.
* The wastegate should flow really well as it has perfectly timed pulses from either rotor and is 100% prioritised. Any restriction to flow should be at the siamese port itself , not the wastegate , even though the WG chosen is small by normal standards .
*no energy sapping major changes of direction for exhaust gases.
*Tubing to turbine does not have to carry gases that go through the wastegate so don't need to be large diameter .

2nd Edit : .Final design after a few iterations :
https://cimg4.ibsrv.net/gimg/www.rx8...77ea1cd838.png





Pasting this summary for those who don't want to go through 85 pages to get to the crux of what was done here ....

Siamese wastegate/twin scroll turbo (edit June 2020)
From data i have I now believe that the siamese wastegate design when used with a high flow twin scroll turbine has several advantages over a more traditional design.
1/The top end power the engine makes is better than anything I can compare it to using a similar turbo and a well setup system . I believe this power advantage is due to there being a lower pressure in front of the wastegate which allows the siamese ports to evacuate gases more effectively. Especially during the last few degrees of exhaust port closing.
2/ Low to mid range power is better than any other Renesis that has a similar top end , and also better than most with less top end potential as well. Mostly due to good spoolup.
3/Improvements in boost control were an added unexpected benefit .
Overall , I would say this design experiment is a success. Major drawback is the difficulty of replicating this result in other setups as it also requires engine port timing modifications to work well.

Bridgeport : (Updated june 2020)
As mentioned above , this modification was very disappointing . I believe I lost somewhere around 20whp across the entire rev range by doing this . A slight reduction in spoolup/response being an added downside. I have subsequently tuned other bridgeported turbo renesis engines and found their performance even worse than I experienced (due to higher exhaust backpressures). My advice is FFS ............................ DON'T BRIDGEPORT YOUR RENESIS !
One surprising side effect that has been beneficial is fuel economy . I get really good fuel economy with a 30% ethanol blend . Mileage being as good if not better than any stock NA Renesis on 100% petrol.

Turbo : (updated June 2020)
I have tried a few different combinations of turbine housing and compressor wheel in the GT3582r base turbo .
In the final analysis I saw only minor improvements going from the stock GT 35r wheel to a billet GTX3576r wheel and then to the GTX3582r.
I found the larger 1.06 AR twinscroll turbine housing to be superior to the 1.01 and the 0.83 in almost all respects
The slightly better spoolup of the 0.83/1.01 did not translate to more power at the wheels in lower rpm ranges as you would expect in a more traditional manifold design.

To date (June 2020) there have been not been any faster spool (specifically BW EFR series) top mount turbos that have exceeded the power band of this setup.

E30 :
Running this compromise blend was a biggie and I would recommend it for anyone running similar boost in a Renesis . I think it's been a great compromise between fuel consumption , fuel system requirements and octane for a street driven Renesis running over 14psi boost.

Two stage exhaust system :
After my latest modification to this I would have to say that I would definitely do it differently if I was starting over . The simplest and cheapest thing I could have done would have been to just go straight to a 31/2" exhaust ...perhaps split into 2x 3" mufflers at the rear .
That said , the setup does produce a unique sound for a Renesis that I really like . It is no longer obnoxiously loud nor is there any drone at highway cruise . Not worth the expense though , I have to say .

Great stuff.

What I'm afraid of it that exhaust is going to be too much of an restriction. At intake side, you'll have pressurized air, making it more dense, but after combustion pressure is lower, and volume is around double. Exhaust bends inside block, if I was to guess, does not look like its made to take massive amounts of flow, with its angle.

I'll definitively stay tuned to see how this goes. It's fantastic that you put so much time and effort into this!

Have thought about this a lot and initially had the same concern . However , there are a couple of things about the design that have led me to think it will flow way more than in any other design .

Oh yeah baby another Brettspeed project.

Not sure if I'm clear on your intent.

Are you sending only the end ports to the turbo and dedicating the siamese port for wastegate flow only?

Yes ... that's the idea .

hmmmm ... will have to mull that one over some. Not so sure that's a good idea under some conditions.

I've done a shit ton of mulling and the more I mull the more confident I get ... heh.

have heard your hubris before, lol ... :p:

It will probably be fine. Only worry is the slightly higher back pressure causing hot spots, but WI will take away that danger. It also lets you use a divided housing.... yeah I love the concept. Make sure to use lots of water. ;)

Well put it this way ........
I wouldn't spend money on it if I didn't think there was a good chance it will work ,and if it does , it could be significant . So I'm prepared to take the risk .

Do you mean from the restriction of the siamese ?

if it stays closed for too long i.e. no flow, I'd be concerned about the temps the seals traveling that path will be abused by

It only needs to be closed while the turbo is spooling up ;)

I can control the WG with vacuum ,to stay open under cruise conditions .

Yeah it's the restriction from closing the siamese. I wouldn't be worried about it, but I would keep a heightened awareness of it while tuning.
Have you tried it yet? It's lovely.

I'm confused.

If the siamese can't flow enough , my concern is that I'll get boost creep as the turbine needs what it needs and will speed up the turbo if it gets too much back-pressure.
My hope is that , as there is no back-pressure on the siamese side , those ports will flow more . The WG can almost provide zero back-pressure if that is needed to allow the siamese to flow enough.


Have not tried opening the WG with vacuum yet ..... the trick will be in using a small spring and whether the EBC will get enough control with that spring in place.

Hopefully it can flow enough to prevent overboosting.

otherwise kudos for the novel idea.

and ... :yelrotflm

Is this the first application of a divided housing turbo on a renesis?

I believe so .

Very very cool. The benefits....:D:

I take it this is a contributing factor to your goal of 16 psi by 3500 rpms? (And carefully selected parts inside that monster of a snail)

Absolutely ... The divided housing (albeit a large A/R )and small volume manifold should make it spool like no other Renesis .

Brettspeed needs to start production of this little beauty.

Been crawling through your 11 thread, and the subsequent fuel pump thread. Thanks for all the great info. Your trial and error helps the rookies (me!) learn.

I'm running a 4psi spring and force it closed to as high as I want with the EBC. Not a problem, just involves some creative plumbing. The siamese port should flow enough to prevent boost creep if you run a EBC with a low pressure spring. The real problems come from full pneumatic boost control systems with bad plumbing, closed loop with electronics makes most of that disappear.

Posted a picture of my WG plumbing on my build thread. Dunno if it's how you're doing it Brettus, but I like how it's worked out for me.

I think I need a visual, I don't see how you will control boost effectively.

I'll do a diagram later . But yeah -poor boost control is the what could sink the idea . I think I have it covered but there is still plenty of risk there .

With a 4psi spring and push pull you can set the controller at 10psi, and when it hits 10 psi it can go all the way full open with the WG if that's what it needs. If the center port WG full open can't control your boost pressure then you have way undersized your turbo. Just think about how small the WG hole is on a stock Greddy, and that tiny thing controls boost...

The problem with plumbing and WG sizing is that a full pneumatic system can't fully open the wastegate at a set pressure. If it's set to 10psi it will crack sometime before 10psi and then get some fraction open at 10psi to maintain pressure. You have to oversize the wastegate because it can never fully open where you need it to.

It's gunna be a 'suck it and see' exercise .It would seem starting with a small spring will be the way to go to avoid boost spikes . I have a sneaky feeling boost control will actually be better than a conventional setup though.

Just tried the vac. line on it with a 6psi spring in there . Almost opens it but not quite at idle . Opens at anything more than idle vac. so looks like a good starting point as exhaust pulses should be enough to open it at idle as well.

With such a quick spool, are you worried about a massive power "hit" like on the original Greddy kit? The dyno charts I have seen made me concerned for the life of my transmission.

That's not going to work very well. With a 6 psi spring you would normally be limited to around 12psi when exhaust pressure will start forcing open the valve, but with vacuum locked in on the top it will drop the max pressure you can run down to almost nothing. You need a 4 way solenoid to do this right.

yeah ... I can see that now . What I might try is a modified check valve that doesn't close till some boost starts to flow passed it . I would prefer something quite simple and seemless rather than an on/off solenoid doing this job.

Possibly not with this setup as pressure at WG may well be way less than on the turbine side.

You're making the same incorrect assumption as the Renesis header theorists. I had it all typed out and then f'ing lost it. Will post the explanation later when I have more time.

Please do. Considering the WG does not need to handle anything close to full exhaust flow (depending on turbo sizing) I don't see a problem with the use of the siamese port for the WG.

It's not an on off solenoid, it's a boost control solenoid. It happens to have 4 ports, but it can handle PWM. The only complicated part is a little extra tubing and in my case a couple check valves with a tee. Look up push pull boost control.

Would you mind going into how this functions a little more please . I'm struggling to get my head around it .
I presume when it's off you have vacuum to top port of WG but what is happening when it's cycling ?


Edit :
I'm sold ... got one on order

The function is pretty simple. You put high pressure on the top of the valve and low pressure on the bottom and it stays closed regardless of boost pressure. You put low pressure on the top and high pressure on the bottom and it stays open regardless of boost pressure.

All the check valves do is allow the low pressure to be at a vacuum (when the throttle is shut) which is enough to overcome around 7lbs spring pressure. It works just as well with the low pressure side directly to metered air, or even a non metered atmospheric vent.

Oh and to have it open the WG during cruise you need a boost controller that can do that, or a relay wired into the controller. Most boost controllers will try to keep the WG closed until they are at set-point.

Checked my Greddy EBC ... It doesn't activate the valve closed till it sees a little over 1psi so it will work without any extra relays etc.

isn't it good for spool to keep the wastegate closed at light load.

I very much doubt it would be noticeable

When Mazda first tried the side port design it was only with the two end ports. What they discovered was gasses from the center iron side were not evacuating effectively. So they added the center port for this purpose. It was never intended to be a full flow port like on the end irons, but rather an additional relief port. They also had to add in the divider plate to minimize cross flow & heat between the two rotors. As a result, the center port flows considerably less than the end ports and this has pretty much been overlooked in every header design; equal length, long tube, etc. Even if you could build a header capable of scavenging on a zero overlap Renesis (you can't, dummies) if you use the same diameter tube on the center port as the end ports then you are effing up the timing of your scavenging design right off the bat. It doesn't flow the same as the end ports, not even with the combined gasses from both rotors.

The question here is will it be capable of enough flow to properly control boost as a dedicated wastgate port. Brettus has wisely chosen a 1.06 AR divided turbine housing which will help flow at the expense of low rpm response. I'll be surprised if it has full boost by 3500 rpm. Partially because of the AR even if divided, but partially because of why the center port exists; it won't be clearing out gasses on the center iron side during spool-up because the wastegate will be closed and not allowing any center port flow. There will be an efficiency issue with the center port closed off. The magnitude is difficult to assess though.

I'm skeptical that the center port has enough flow potential to effectively control boost on it's own. As a restictive port it will potentially be self defeating for wastegate control. More backpressure can force more flow through it, but that same backpressure is also spooling the compressor wheel to pump out more boost. To be effective it must offer the path of least resistance *for both rotors*, which IMO is questionable. So potentially you are in an escalating loop that may prove difficult to control.

I could be wrong though. I only call you out because you presented it almost as a fact that it can flow enough without any room for doubt or error. That's pretty typical for you (Harlan) when making theoretical assertions. The path in reality is not always so clear.... (insert excuses & work-arounds here)

Well thanks for calling me out personalty. I do appreciate the attention.

Now I don't know where the disagreement here started... IF the center port is to minimize EGR on an NA engine which makes a lot of sense then how does that change this discussion? Sure this design might have a bit more EGR when the center port is closed, and sure the WG won't be able to flow as much as it would off a normal port, but in the end does that even matter? I still don't think so, and I'll wait for it to be proven either way.

You accuse me of making assumptions, but right now you are stepping pretty far out with yours. At least this time you are bringing your assumptions to the table, albeit without actual facts.

Subtle difference -->
Otherwise sorry, I generally ignore same like in your own thread but this time I did not choose to do so ...

You didn't leave much wiggle room in this statement, and you must have read my mind to know what I was assuming and at the same time know I was wrong.

I'm more than willing to debate the merits of an idea on theory or on actual testing, but please stop mind reading.

guess I hit your tender spot to the point you're not thinking clearly, ok then ...

How do you know this ?

Team , I think (and hope) the center port flows somewhere between 34-40% of the total flow from an NA engine . Here is my logic ........

Many years ago Hymee showed a renesis on the dyno and the only tube that was glowing red was the center port tube . That led me to make one (fairly big) assumption : that the center port tube flows more in total than either of the side port tubes.
For that to happen ... the individual Siamese port on one rotor must flow more than 34% of the total for that rotor . Or in other words the two ports combined are more than 34% of the total for the engine which leaves 33% for each outer tube .

It is also safe to assume that the individual Siamese port flows less than the other port. Only 5% less would seem unrealistic ...............

Hence my guess ........ 34-40%

If I'm right on that ... it is right in the ballpark for what I expect the wastegate to flow.
X fingers!

This is just a shot in the dark, I know the Siamese port has been fiddled with before. Could you fab up a full divider? I know it would play hell with header/manifold design, and that "divider" piece would get hot as hell, because it's constantly exposed to exhaust gases, and never gets a chance to cool?

The first design I thought of had a fully divided tube and even a balance tube from the other port. I went away from that idea when I realised that the amount of back pressure from the wastegate and downstream of it should be only a few psi . Plus there is more flow potential by opening up the port rather than closing it in.

Ok how about a little problem.

Lets say exhaust pressure in the housing is 2psig, and going out both exhaust ports it drops to 0psig. The pressure drop across each port (and the associated exhaust pipe) is 2psi. At that 2psi there will be a certain amount of flow through each path.
Lets say an 80/20 split between the main and siamese port.

Lets say you mount a device (*cough* turbo *cough*)that creates backpressure on the main exhaust port, and now instead of 2psi in the housing you have 8-10psi. If the siames port is fully open will that create more or less flow through that port in relation to total flow?

If we had real numbers for exhaust housing pressure and turbo backpressure we could even figure out how much different the flow rate would be.


Oh and the reason why the center exhaust pipe glowed red when the others don't is there is no real time for it to cool off between exhaust pulses like the other two ports.

Don't know if this was supposed to be a rhetorical question or not.... but of course it will increase the flow through the siamese port as flow will direct more at the path of least resistance. Theoretically with more pressure in the housing it will increase exhaust velocity through the siamese port equaling greater mass flow rate.

You think 0.0067 seconds is long enough to cool down ?

two pipes the same diameter and wall thickness are flowing hot gas from the same source ... one is glowing red one isn't .
The possibilities:

*the gas is hotter in one than the other
*one is flowing more gas than the other


Pick one ....