well just going to state it again since maybe it wasn’t as clear in the previous discussion here; due to the new technology the newest Garrett G-series line can’t be directly related to the previous Garret GTX and earlier turbos by impeller size. Because the technology on the G-series impellers is higher flow in a smaller size. Yet the technology is such that you won’t get the same response from a G-series by matching it up to the impeller size of the earlier models either. That will just get you a one size larger turbo and the expected response/output potential outcome of doing that.

The general issue is people have become trained to look at impeller size as implied to mass flow, but with the size disparity of the new G-series turbo they now need to re-educate their mind to comprehend actual mass flow rates instead. Which effectively is what a turbo is all about at the most basic level; mass flow balancing.

The new G35 line is effectively equivalent to the next size up from the previous GTX35. To match up to the previous GTX35 or earlier turbos then you need to be looking at the Garrett G30 models. Because when you go compare maps for mass flow between the G30 and previous GTX35 models then it can be seen that they are closely equivalent to each other despite the impeller size differences.

The other thing to be aware of is the RPM limits stated on the far right side of the G-series compressor flow map. If the turbo exceeds that limit much then things aren’t going to end well for the compressor impeller. Ideally a turbo speed sensor and software protection strategy for an rpm limit is the ideal setup.
.

 

just to add, this includes flow through the wastegate. so if anyone was wondering how the compressor can flow ~50lbs a minute and the exhaust only flows ~26, the other ~24 is going out the wastegate.
i haven't done the maths, but it is also true that the different temperatures between the intake and exhaust effect the numbers, so there is that part too.

i'm still trying to figure out how to read the turbine map.

 

not much to read really; when the turbine flow map line of a given impeller for a given A/R goes horizontal it’s indicating the chokepoint for that particular impeller/housing; no matter how high the PR on the X-Axis goes from there the mass flow on the Y-axis is essentially peaked out and not changing.

so where are those choke points on various turbos?

Garrett
GT35R 1.06 T4 - 28 lb/min
GTX30 Gen2 1.06 T4 - 27 lb/min
GTX35 Gen2 1.06 T4 - 32 lb/min
G30 1.01 T3 - 27 lb/min (1.06 T4 ime ~30 lb/min)
G35 1.01 T3 - 32 lb/min (1.06 T4 ime ~35 lb/min)

BorgWarner EFR
7670 1.05 T4 - 25.5 lb/min
8374 0.92 T4 - 28.5 lb/min
8374 1.05 T4 - 30 lb/min

now start breaking down results against these and the picture should hopefully be a bit less cloudy …
.

 

Not really less cloudy for me.

*Stating obvious alert*

I want to know the work required to move the compressor section at the point of the compressor map it is on and how much of the max exhaust turbine flow is being required to meet that work demand.

Because the rest of the exhaust mass flow should be wastegated.

It is only when the compressor work demand is higher than available work output from the exhaust side that a well fabricated turbo set-up will meet exhaust choke flow in the turbo exhaust housing.

________
In practice, most turbo rotaries have boost creep issues (compressor overdriven by an excess of turbine work achieved) when wastegate priority and or area is insufficient.

This is especially an issue with large compressor sections on smaller exhaust sections like in the case of my experience with FC stock hybrid turbos.

So, in actual applications- do we know what turbo exhaust housing choke flow looks like on a 2 rotor?

Its 100% not the boost creep...

Turbo exhaust housing choke flow is what what happens when a given turbo boost creeps above target boost and then drops below TARGET boost in the higher rpms.

A rare beast on 2 rotor because we typically operate our turbos at low compressor pressure ratios.

Actually, in practice this boost creep and the exhaust housing choke flow/boost drop on a 2 rotor can often be corrected by INCREASING wastegate priority and or WG area to lower exhaust manifold pressure.

This is because increasing wastegate flow decreases exhaust manifold pressure which increases flow through the motor lowering the compressor pressure ratio (which self corrects emp back up, but increases exhaust velocity) required for a given engine flow rate, which in turn requires less turbo exhaust work output (which will likewise self correct if under target boost).

Well, its actually WAY more complex than that due to the dynamic nature of peripheral exhaust port rotary overlap. Even if exhaust manifold pressure increases , if emp increase tracks exhaust port velocity increase the exhaust flowing over the rotor tip out the exhaust port entrains the intake air on the adjacent rotor face simulating a lower emp in terms of overlap dynamic (rotary peripheral exhaust port is literally a siphon during overlap).

Now we must examine compressor mass flow and what % of that flow supports internal combustion on a rotary versus external combustion on a rotary.

The external combustion does not directly affect combustion derived power output (indirectly through the dynamic overlap described above), but does directly support turbo exhaust work output.

A 400rwhp 2 rotor is a far different dynamic than 650rwhp 2.0L piston engine, yet they are sharing a turbo sizing.

Why?

We have an exhaust port that never closes on the 1967-2002 rotaries.

Do turbo sizing calculations account for these rotary dynamics correctly?

 

Nobody is talking about a 650 whp application in this thread. And if you went to those successful results and applied what was previously posted then you won’t be looking at any of the turbo examples I listed.

What we have here in this thread is somebody with an old 35R turbo looking to upgrade it to a newer model with a 350 - 450 whp goal. Yet due to technology changes it’s not just an impeller size comparison wrt to the Garrett G-series, which came up in the discussion. That’s all I’m attempting to address.

What seems obvious is that you don’t understand that this thread isn’t about any of that or other advanced considerations that have no relevance at the moment. So why would I bother going into any of it then? It’s not even a one post subject. It’s a novel length book.

dang forum parakeets …
.

 

dang forum parakeets …


Sorry couldn’t resist

 

Exactly!

Take your time, re-read my response.

 

its a line that slants over to the right hand side, usually black, but sometimes red or green....

basically no. question number 2 is the PR on the compressor map the same as PR on the turbine map? or are they two separate PR's for the two separate housings?

for example, if you're running 15psi on your GT35, we can plot the compressor map pretty easily, and say it flows whatever it flows at whatever rpm it is
(lmao just looking there are like 3-4 different GT35R maps, so getting the right one would be a start...)

then do you look at the turbine map and say at 2pr the 1.06 housing is flowing about 32lbs min

OR

is the pre turbo backpressure, at peak rpm, like 17psi, and post turbo BP is about 2, and that is 8.5 pr? *edit its 1.9, my maths are wrong*
flow is the same, but its a very different part of the map!

OR

do you find peak RPM on the compressor map, and just assume the furthest right on the turbine map is also peak rpm?

 

yeah that one got muddled, but for the money you have, or the OP has two choices. the G30 is ~$2k, but that is only $200 more than a new GT35, so its the obvious choice.
or you just buy a replacement CHRA for $1000 and forget this thread ever happened, lol

 

My response was to TeamRX8 listing turbos by peak turbine flow in this thread.

Important data that isnt the most relevant factor to the original poster with the size turbos he was considering.

I attempted to point out where/when peak turbo exhaust flow is relevant and why.

All the turbos suggested by people in this thread are good turbos for ~400rwhp at a low compressor pressure ratio and the OP just has to balance the cost he is willing to spend versus the turbo response he is wanting.

Another GT3582R, but this time in a divided T4 exhaust housing will be an incremental improvement at a relatively low cost.

I believe going down to a GT3576R or equivalent sized turbo in a divided T4 housing will meet his goals if his priority is response.

Goals as I understood- 350rwhp with possibility of 450rwhp on E85.

 

I mean, do we know what a dyno of this looks like?

If you dyno your car without the wastegate(s) you will see what the dyno graph limited by peak turbo exhaust housing flow looks like.

I did not do that,

but I did dyno my car without the wastegates boost reference line attached, so the results were the same until exhaust manifold pressure overcame the wastegates springs pressure holding them closed.

I will try to get the hp/tq files from the dyno owner.

Here is that un-wategated pull in Blue along with the wastegated pulls.



from memory, peak torque was the same on the un-wastegated Blue line and the other 3 high boost wastegated runs.

To note the relevance to this thread regarding turbo sizing-
​​​​​​​
power figures were Dyno Dynamics being run backwards, so that 369rwhp equaled 420rwhp/420rwtq Dynojet the next weekend. Car was on gasoline (race gas).

-edit- added dynojet graph



​​​​​​​

 

You just don’t understand because you never looked at or ever considered it the way it’s presented. You never knew what the peak turbine flow limit was on the EFR7670 and is why you also didn’t readily recognize that the newest EFR8370 is not going to pan out well on a rotary without a larger A/R turbine housing per other thread discussions. You couldn’t peak out the EFR7670 compressor flow to the rated 64 lb/min limit even up near 30 psig boost range because again, the flow limit of the 1.05 A/R housing is holding it back.

The same reason this 13B with EFR7670 only can achieve the same basic output as you did, except it’s bridgeported which then accomplishes it at 1/2 the boost pressure you were at:




https://www.rx7club.com/time-slips-d...5-psi-1122867/


There’s all kinds of data and discussions about compressor housing flow in here, but essentially zilch on turbine housing flow and the relationship. Most people just think a 1.06 A/R applies across the board. I understand why it might seem that way because if you go back to the list I presented, it presents a false appearance that way with all the 1.05/1.06 listings. However, if you go back earlier in this thread where I compared the GT35R to the G30-660 and then the next post shows why the larger G35 model should then move to a lower A/R turbine housing to be in the same ballpark wrt the same output goal.

Because as the turbos move up in size the flow through the turbine wrt response, output, and limit is not going to linearly translate to using a 1.06 A/R. All the A/R represents is a dimensional relationship. The flow through it is what really matters and that flow is not linear as the turbo/impeller sizes change.

The perfect example of that are the various Rob Dahm threads I’ve been posting in. I predicted he made a good choice on the 20b turbo, others said it was too small. It blew through 1000 whp and exceeded the operational limit. On the other 13B selections though I noted he was making bad choices and those all fell flat. A Garrett G40 with 1.06 AR housing that has a 40 lb/min chokepoint is a 700+ whp turbine selection on a 13B and putting it on a unported T2 13B for autox/track use is going to yield exactly what they got; sub 400 whp output and poor response. The 0.85 A/R @ 35 lb/min would be better, but is still a 600 whp 13B turbine housing. It really shouldn’t be any larger than the G35-900 and ultimately a G30 will yield the best response.

Maybe you think my reference to the turbine chokepoint means I don’t understand that the actual flow through the housing will be lower under various circumstances. It’s not that at all. What this accomplishes is sizing the turbo turbine to neither be too big or too small. It’s based on studying and assessing 100s of results. With those results you typically also get what porting was being used, what wastegate size(s), exhaust size, and so on. When you go through all those gyrations then patterns begin to emerge.

What I’m suggesting in the turbine housing selection process is one of those patterns. That’s why I can see that a Garrett G30-660 matches up to GT35R and am not just blindly grabbing a G35-900 instead. And so on. Because I’ve also spent countless hours running stuff through programs like Marchbot. Which is why I readily recognize that their turbine Phi chart is the same as a Garrett turbine map, just that they’re all stacked together on one big sheet rather by turbo model size. And also why I recognize what the flow limit is for each one of them even though it’s not listed.

What you need to do is go spend some time running the models. Because once you see where the data points lay out consistently on the turbine map flow line over and over again maybe you’ll cut out the theoretical rhetoric and come around to a better understanding of what’s really going on.
.

 

i went down the rabbit hole and found the equations Garrett uses. once you correct for temperature the mass flow of the compressor and mass flow of the turbine are equal.
the air that goes into the turbo goes through the engine and then the turbine. The exhaust being hotter makes a big difference, its mass flow corrected for temp and pressure


part of the problem is that Garrett is annoying. not only is this buried, but they also use different terms for the same thing, and they have some terms that are not defined. if they were in school and turned this in its not A+ work....

here are the compressor side maths





put it in excel, be somebody....

nobody i'm aware of has pre turbo AND post turbo pressures, but i built a spread sheet (if i can do it so can you!) and put some reasonable guesses in there, and we are in a very different part of the turbine map than we thought.
the way the maths work is that higher PR = higher backpressure, and the Rotary doesn't like that very much.

it does raise some questions though.
1. its not defined, but the turbine map looks like the choke flow line of a compressor map, but is it?
2. it shows flow at its rated efficiency, what does it do at lower efficiency?
3. Garrett doesn't mention the wastegate at all, except that they sell them. the BW matchbot does have wastegate info, and i think if you do the calculations, and then look at an actual turbo, you could come up with a WG flow, but i'm not sure.
3a, for example if we use Garretts examples from above corrected compressor flow is 46.3 and the turbine flow is 36, does that mean we need a wastegate that flows the difference? ~10lbs min?
4. it would be nice if they had peak rpm on the turbine map, but they don't

 

Your listing the turbine flow limits made me think that even your lowest flowing example was a good fit for what I understood the original poster wants out of the turbo

And I had real world experience with it and could show it.

I specifically showed and stated I showed the turbo limited by peak exhaust section flow (non wategated pull) and then showed the same exact with turbine flow plus wastegated flow.

I believe that set-up would be great for 350rwhp on pump gas and being pushed to 450rwhp on e40/85 and great turbo response.

I dont mind that you disagree or think another turbo would be better- it may be.

_______

Unfortunately, we dont know where I was on the compressor map without turbo shaft speed.
Ans a we have seen from past threads even then a simple boost leak can throw even competent tuners for a loop on that one.

I dont know if 368rwhp on the LCR dyno dynamics which was 420rwhp on Jefferson State Diesel mobile dynojet is similar to 450rwhp on another dynojet.

 

Please stop reading what I write with an implicit bias.

You know well the "efr 8370" that I didnt know even existed was a typo for efr 8374 and I am a proponent for both larger diameter exhaust wheels and higher flowing exhaust sides relative to compressor sides on a rotary.

In this thread I saw a match relevant to what I thought the poster wanted and shared my experience with it.

 

Semantics alert-
​​​​​​​Air mass out of the turbo compressor section plus the very very small mass of fuel (or fuel & additional water) added is what goes through the turbo exhaust section.

 

true, on a rotary we know the volume of fuel, its about 1/10th the volume of air (10:1 AFR), we'd just need to turn volume into mass, which should be easy enough. although i'm not good at math

also Garrett is using SAE STD corrections, 60f, 0 altitude and humidity, and it looked like BW is using SAE, 75f. it makes me think Garretts equations are OLD like 1960's old
anyways if you run the same numbers through garretts equations and the matchbot they are really close (totally close enough), but not the same, so there is a difference

and speaking of the maths to get the flow equations to work in excel i had to put each operation on its own line, so it shows the work. i'm sure someone could get it to work in one cell, but i'm not good at math

 

I wasn’t referencing the thread with the 8370/8374 typo, but the discussion of the 8370 in Blue’s 7670 thread. I possibly may have confused someone else’s comments, but thought it went down that way there. Just the best of my recollection, even if flawed, but no malice or deception is intended.
.

 

I'm a little hesitant about reviving this thread since it was getting a little heated, but here it goes.

The old GT35 turbo ended up being OK as it was (nick was very small) so I've been running it while breaking the car in. Now that its broken in, it runs and boost well but I'm wanting something that spools better, so we're back to the original question.

Quick recap of setup: mid-large street ported 13BT, 9.4:1 rotors, V-mount intercooler, flex fuel setup with plenty of pump and injector for whatever. I plan to run it at wastegate pressure on 93 octane (currently ~ 10 PSI), and raise it with e85. Currently have a GT35r with 1.06 T4 undivided (but I have the divided manifold). My goal is a max of 400-450 at the wheels, with an emphasis on spool time as this is mostly something for fun around the street or country roads, with an occasional track day.

I've been eyeing the G30-770. Its still a stretch, but I could possible wrangle it. However, I just saw the GBC line. Is there anyone running one on an RX7 yet? The gbc35-700 looks VERY interesting, and its something like 1/3 the cost. I realize it won't be as good as a G-series, but I can make some small concessions for that level of savings.

 

Find a cheap aftermarket split pulse housing for your current turbo and use the solenoid controlled secondary runner throttles to stage them in. Use some form of boost control solenoid or at the very least one of those ball seat check type arrangements so your gate doesn't creep open on spool.

 

In the times between the latest thread revival I have bought a brand new EFR 7670 1.05AR divided T4 turbo for $1,500 with aftermarket BOV on it and a BW refurbished EFR 8374 supercore (turbo minus $400 exhaust housing) for $1,000.

So, the deals are out there- you just have to check Craigslist, FB marketplace and Ebay a couple times a week.

 

Already have the solenoid control.
​I've thought about just putting on a divided manifold to my current turbo, but I'm wondering if the damage to the turbine wheel is holding it back. It's possible it's due to the current heat, but my spool seems to be getting worse recently, even in the mornings. That's why I'm looking at a new turbo.

The gbc 35/37 looks very interesting, there's just so little real world info out there.

 

it only emphasizes my comments about an overall lack of understanding on the turbine flow requirement by some people. Which is why I keep bringing it up.

https://www.rx7club.com/single-turbo...turbos-1160788

as previously stated wrt a 13B rotary application; the overall issue on the GBC turbos is that the largest exhaust housing they offer to date is a 0.95 A/R div T4. On the GBC35-770 that’s going to be equivalent to a much tighter GT35R turbine A/R, down in the low-mid 0.8X range. Which imo will require E85 to handle the response and high emap with some margin of engine safety.

You’’re really going to need the GBC37-900, as the the 0.95 A/R turbine for it peaks out at 30 lbs/min @ 2.5 PR.

It can meet your stated 400-450 whp goal. It will be equal in flow capacity on both hot & cold sides to the Garrett G30-900 1.06 A/R. Compared to the GT35R 1.06, a good bit more cold side flow, just a bit less hot side. Or it could be a bit more exhaust flow.

It just depends on which GT35R you have, because it’s often used in a generic sense but their were variations over time. I have a turbine map verified from 2006 that the 1.06 A/R peaks 28 lbs/min, but later 1.06 A/R turbines from around 2013 to current are 32 lbs/min peak. E85 still might be a better choice to maximize response and output.

alternatively you could consider a Pulsar PTG series turbo, which is their “similar duplicate” low cost version of the Garrett G-series; the turbine housings are directly interchangeable between them.
.

 

It's a genuine Gt3582r purchased new in 2017 through Rotary Performance, so that sounds like the higher flow version.

So what you're saying is that it's going to be slightly lower flow in the exhaust, but more on the cold side vs my current setup, right? Big thing I'm curious about is what the spool will be like since it's journel bearing. I guess since it's a little smaller, plus I'm sure a better turbine design, it should be at least a little better?

G series is off the table, just found out my daughter + 6mo grandbaby are moving back in later this month, so I'm not going to strain myself financially right now. I'm wary of the Pulsar stuff, to be honest, so it's really down to gbc37 vs just getting a new chra for my current one.

Also, I'm going to be running lower boost on 93 (spring is 10 psi right now, but I can change it to 8 easily). My ecu can vary boost vs e85 content, so I don't have to worry about changing the settings around.

 

duplicate post
.