What the EFR turbos should have been

[LuckyX2]

So after a lot of time contemplating whether I would get a GTX2867 or EFR6758 I eventually came across something I hadn't heard of before, the EFR7163...

Now this turbo is what the EFR's should have been from the very beginning!



Think of this as essentially a second generation EFR with a whole new "mixed flow" turbine design. Here's a some reading on what advantages mixed flow turbines bring.

http://www1.eere.energy.gov/vehicles...deer11_sun.pdf

Garrett has been using the same outdated turbine wheel for decades so clearly engineers can come up with something better by now, right? Mixed flow is that better design. So what does this mean for us in terms of end results? Well, it means faster spool on a larger turbo for both low and high rpm power.

Review from Perrin:
Borg Warner EFR-7163 Test and Tune » PERRIN Performance Official Blog

Faster time to torque than the 6758 (Spools about the same but makes more power):


Spools faster than the 7064 (Shouldn't say 6758):


Here's a good quote from a guy running a prototype model on his Mazdaspeed 6:
"The spool is unreal. We went from vacuum to 22 psi in 400-500 RPM ANY time we hit the gas above 2400 RPM."

So what are your thoughts guys? Seems like a pretty big leap in turbo performance to me and the "revolution" the EFR's were supposed to be.

2014 RELEASE UPDATE!

The 7163 is finally available to buy and in new housing options like 0.80 T4 twinscroll and V band IWG or EWG
BorgWarner EFR 7163 Turbo - Full-Race.com

Here's what the compressor map looks like:


Here's it overlayed with GT30 maps. Much broader.


Overlayed with a GTX3071. Still broader.


And here's what it should theoretically do on an SR20 at 25psi and 7500rpm max:
BW Match Bot Map



More map points on a 2L, courtesy of Kingtal0n:


Image of the change in turbine geometry:


Looks like they upped the power estimates by 50hp from their original 225-500hp range:


This R35 is already breaking records with twin 7163s:

What record it is exactly, I'm not sure.

[Croustibat]

I dont believe in miracles, nor crappy comparisons and commercial bullshit. And i would like to see real flowmaps.

From your test :

Not the best comparison because this is the one where the TGV housing was closed on my car. So it was killing the HP up top. Since the smaller turbo spooled slower, its possible that it effect spool a little as well.

There is no miracle, the turbo spools faster than a gtx3076r and makes 20 less HP, and the tester says its test rig limits top end ... so in the end it spools sooner than a gtx3076r, but i guess it also makes 50 less HP at the same pressure. They just are not the same application.

Finally, i would really like to know the price difference, what you actually know about "mixed flow", and how you know garrett is not using that already.

[PoorMans180SX]

Yes, this turbo is pretty awesome. I'm hoping they make it available in twin-scroll T4 housing.

However, it won't be available for quite a while, as not even the regular EFR's are available. Trust me, I'm on the waiting list.

Croustibat, you have to be kidding about Garrett using Mixed-flow technology. They haven't updated about their designs in forever besides the new GTX wheels.

Mixed flow turbines essentially allow the turbine wheel to capture more energy at low exhaust speeds by striking a compromise between radial (typical turbo design) and axial turbine flow. I don't have a degree, so I can't go into all the crazy formulas, but that is the gist of it.

[LuckyX2]

Originally Posted by Croustibat

i would like to see real flowmaps.

This turbo isn't released yet so no flowmaps currently.

Originally Posted by Croustibat

There is no miracle, the turbo spools faster than a gtx3076r and makes 20 less HP, and the tester says its test rig limits top end ... so in the end it spools sooner than a gtx3076r, but i guess it also makes 50 less HP at the same pressure.

So the EFR takes a hit in top end but the GTX doesn't? How do you figure the difference goes from 20Hp to 50Hp? They're both subject to the same disadvantage so I'd say the test is fair.

Originally Posted by Croustibat

Finally, i would really like to know the price difference, what you actually know about "mixed flow", and how you know garrett is not using that already.

They're not released yet so neither are the prices. There really isn't anything new here in terms of materials cost or manufacturing difficulty so I'd say prices will be about the same plus a little extra for R&D for the mixed flow turbine. Did you read the pdf I linked to about turbine efficiency? Mixed flow turbines showed a clear improvement in efficiency. And no, Garrett is not using mixed flow turbines on their GT/GTX series turbos but maybe they are on some diesel or industrial application I am not aware of. They are still using turbines from the 70's in the GT/GTX; it's only the compressor that has been improved. I know at the very least they have some diesel turbos with variable geometry turbines but I'm not sure of any other advancements they are pursuing on the turbine side.


PoorMans180SX, it looks like you've got it. Radial turbines flow perpendicular to the axis of rotation while axial turbines flow parallel. Mixed flow turbines are a combination of the two flowing diagonally I guess you could say. And yes, it's gonna be a bit of a wait. Current estimate on the regular EFR's is August/September while these new ones won't show up until the end of 2012 or early 2013. Also, I'd venture a guess that there will be a twinscroll housing just based off of how they make the twinscrolls only for the larger wheels that this has and also how they state there will be new housings available.

Also, lol at the spammer info...

[cotbu]

I like the testing, it's pretty much the same as I do!
The testing I've done for the GTX2867 vs the GT2871r has been flagged, as they are not comparable. I think that means the gtx2867 is not an upgrade for the 2871r 56 trim.

Garrett does have to step their game up. they only compete with themselves.
While other manufacturers try to out do them at every turn.
This turbo (EFR7163) is a game changer for sure
More power at the same boost, and faster spool that's what we want right!

[LuckyX2]

Originally Posted by cotbu

I like the testing, it's pretty much the same as I do!
The testing I've done for the GTX2867 vs the GT2871r has been flagged, as they are not comparable. I think that means the gtx2867 is not an upgrade for the 2871r 56 trim.

Can you clarify this?


And I found some good info about these on a Mazdaspeed 6:
Full-Race Bound! - Mazdaspeed Forums

Here's an a good quote from the guy who's running it:
"The spool is unreal. We went from vacuum to 22 psi in 400-500 RPM ANY time we hit the gas above 2400 RPM."

That's ridiculous transient response for a turbo that flows this much...

And here's his dyno results from page two:

Keep in mind this is on an AWD Mazdaspeed 6 which is going to have more transmission loss than our RWD cars. It also has an extra 0.3L on us though...

I suggest you guys read through the first few pages too instead of just what I posted here. They talk a lot about their impressions which are more important than dyno numbers imo.

[cotbu]

I was told they are not marketing the gtx2867 turbo as an upgrade from the gt2871r 56 trim. The gtx3071r is an upgrade from the gt3071r.

And I would retune if the graph looked like that on an sr20. Looks good for drifting though.

[LuckyX2]

Originally Posted by cotbu

I was told they are not marketing the gtx2867 turbo as an upgrade from the gt2871r 56 trim. The gtx3071r is an upgrade from the gt3071r.

And I would retune if the graph looked like that on an sr20. Looks good for drifting though.

Did you read through any more of the thread? That was on 87 octane (not sure why that's all he had access to...) and with a restrictive intake/exhaust. Fix those problems and the torque won't fall off like that. There's no reason this turbo should tail off up top with only the 22psi or so he was running on it. Assuming this has a similar compressor map to the 7064, I wouldn't expect it to start tailing off until 25psi or so on a 2.3L and even higher on our 2.0L. That's assuming a 7k redline; take it to 8-9k and I'm sure you'll see it fall a bit.

And why is the GTX2867 not an upgrade vs the GT2871? It flows 4lbs more and with a 4mm smaller compressor which means better response.

[cotbu]

I'm dealing with sr20 engines and real world scenarios.
I've done testing with the 2867 and the 2871, in my case the 2867 needs more boost to achieve the same power output. If I wanted an upgrade from a gt2871r, it would not be the gtx2867. It would be a gtx3071!
Coming from a 2560r or 2860r then yes the 2867 looks much better, but like you stated there are even better options soon to come.

btw 20psi vs 24psi
I don't think my stock engine would stay together running 24psi on my gt2871r est 425-440

[steve shadows]

Very Very Sexy indeed. Borg Warner is starting to become a new favorite of mine too...

[steve shadows]

Originally Posted by cotbu

gt2871r, it would not be the gtx2867. It would be a gtx3071!

GTX3071R is the way to go in my opinion for any 2-3 liter 4 cylinder that is looking for good mid-range power and torque...very little trade off great turbo.

[LuckyX2]

Originally Posted by cotbu

I'm dealing with sr20 engines and real world scenarios.
I've done testing with the 2867 and the 2871, in my case the 2867 needs more boost to achieve the same power output. If I wanted an upgrade from a gt2871r, it would not be the gtx2867. It would be a gtx3071!
Coming from a 2560r or 2860r then yes the 2867 looks much better, but like you stated there are even better options soon to come.

btw 20psi vs 24psi
I don't think my stock engine would stay together running 24psi on my gt2871r est 425-440

I've never seen these results. Can you post them in the GTX2867 thread or here please? I'd like to see.

Originally Posted by steve shadows

Very Very Sexy indeed. Borg Warner is starting to become a new favorite of mine too...

Yes but quite the tease too! I just wanna have at it already!

[Croustibat]

Originally Posted by PoorMans180SX

Yes, this turbo is pretty awesome. I'm hoping they make it available in twin-scroll T4 housing.

However, it won't be available for quite a while, as not even the regular EFR's are available. Trust me, I'm on the waiting list.

Croustibat, you have to be kidding about Garrett using Mixed-flow technology. They haven't updated about their designs in forever besides the new GTX wheels.

Mixed flow turbines essentially allow the turbine wheel to capture more energy at low exhaust speeds by striking a compromise between radial (typical turbo design) and axial turbine flow. I don't have a degree, so I can't go into all the crazy formulas, but that is the gist of it.

I only quoted you but it also applies for the other that answered.

What i meant is simple: the test protocol is bad, so its results go to the bin.

that PDF has nothing but marketing terms on it, and this turbo is not released. It may be good when it is, but as of now we really know nothing. And i kind of become very skeptical when a company claims to have created the holy grail but has nothing right now to backup its claim. Nothing more.

I worked a bit on turbines so i am not a specialist at it, but as far as new technologies go, i know this: when there is a technological breakthrough, people know about it, and everyone follows the trend. So far "Mixed flow" is just a marketing wording.

I am not in any way claiming that turbo is shit. It may be cool. But we know nothing about it, neither its real capabilities, nor its price. Because if it costs twice what a gtx3071r costs, no matter how good it is, i am not getting one.

[Crondy]

Fuggin in.

[gotta240]

Originally Posted by LuckyX2

I've never seen these results. Can you post them in the GTX2867 thread or here please? I'd like to see.

I'm with him...... I've been following the 2867 tests but haven't seen the test you speak of Cotbu...... Can you fill us in? A link? The results? Everything I've seen says the 2867 makes same power as 2871 with faster spool....how isn't that an upgrade(other than them not marketing it as such?)

[cotbu]

No I can't! Do you guys find it even remotely funny that you can't find this information? c'mon!

I'm paraphrasing:
Here is how, I tried to explain it. The gtx2867 does spool faster and it should, It does make the same power at a higher boost level, it does use the same turbine housing.
If, I have to run more boost, is it an upgrade to the GT2871r?

The new forged 11 blade design makes it a more effiecient turbo.
This was probably the response of a lowlevel employee, but whatever!

My hands are tied, just test it for yourself, but be real! MHO, it's not worth the money to upgrade from a gt2871r 56 trim.

PS I had a big whiny, crying post of how, and why I couldn't post the info!

[gotta240]

I'm lost and i suck at reading between the lines......,.... What I gather is....

1. You tested the turbo but the results are NOT posted anywhere....
2. You are "not allowed" to post results because you are somehow linked or obligated to garrett and cant be seen "badmouthing" them???????

Am I at least reading what you're saying somewhat correctly?

[cotbu]

1. yes
2. No obligation to Garrett, but yes "badmouthing" is frowned upon!

Don't forget this turbo is already on the market and this information should be available. I asked, why is it hard to find?

Buy it try it, maybe "return it"!

[LuckyX2]

I spoke with Raffi at Full-Race and he says Borg Warner is still tweaking the aero on their turbine wheels and "They have made some considerable increases from the first gen 7163 that was unveiled at SEMA". They're hoping to finalize the design and start production by September meaning we may see these second gen turbos by the end of 2012. Early 2013 would be a safer bet tho.

Also, I came across this on MotoIQ:
Team America World Time Attack - ARK Design's BNR32 Skyline GT-R Part 7

Looks like Borg Warner is working on a quick spool valve for divided manifolds based on the location of the "secret parts" and the lever arm extending from it.

[PoorMans180SX]

A quick spool valve on a divided manifold would have to be a very complicated design, being that you can't just block off half of the exhaust manifold. Interesting none the less.

I am really looking forward to when Borg finally gets all the kinks worked out and has production levels up to par.

[PoorMans180SX]

A less-covered shot of it:

[Def]

Originally Posted by PoorMans180SX

A quick spool valve on a divided manifold would have to be a very complicated design, being that you can't just block off half of the exhaust manifold. Interesting none the less.

I am really looking forward to when Borg finally gets all the kinks worked out and has production levels up to par.

It's not that complicated. You just flow both entries into one scroll, then open it up to flow each entry into its own scroll.

Basically doubles the A/R but you don't get tons of losses at higher RPM where you're running a divided housing on an open collector.

Really not that hard to make it reliable either, considering internal wastegates put up with the same environment and function. Those go for longer than a turbo will last.

[PoorMans180SX]

Yeah I see what you're saying. I was over-thinking.

[LuckyX2]

Next Gen is here: Full-Race/Borg-Warner EFR7163 in stock location. 440whp/443wtq - NASIOC

And the gradual leak of info continues...

[PoorMans180SX]

That boost response looks crazy. Can't wait till this thing comes out.

[LuckyX2]

Originally Posted by PoorMans180SX

That boost response looks crazy. Can't wait till this thing comes out.

Seriously. I never thought I'd drop the cash to buy a fancy new turbo instead of used but this one is tempting me like no other.

The STI has way too many differences to our engines to compare engine to engine but he gives us a good idea of its relative performance when he says "Damn near stock turbo spool with nearly 200whp more by redline."
The 08' STI came with a VF48 turbo which is around the performance of a GT2860. A GT3076 can make almost 200whp more than a GT2860 just like the EFR7163 makes nearly 200whp more than the VF48.

So in terms of our Garrett turbos, this turbo would make the power of a GT3076 but with the spool of a GT2860.

[PoorMans180SX]

Yeah the 7163 is actually has very similar wheels to the 7064 (obviously from the names) so it will probably flow in the 55lb/min range. Pretty nuts when you consider the spool.

[LuckyX2]

Originally Posted by PoorMans180SX

Yeah the 7163 is actually has very similar wheels to the 7064 (obviously from the names) so it will probably flow in the 55lb/min range. Pretty nuts when you consider the spool.

Is that a safe thing to assume? The mixed flow turbine seems like enough of a departure from convention that a 63mm mixed flow isn't going to consume air in the same way a 64mm radial turbine will. Compressor sides are nearly identical though.

Also, I'd like to see BW take their turbine wheels down in size a little; I'm curious to see how that would affect things. Currently BW has their turbines 5-10mm larger than the comparable Garretts and there are a few things to consider with that. They can make a larger wheel than Garrett and still have the same mass. The problem is that it's inertia that keeps the wheel from moving, not mass and inertia increases linearly with mass but squared with radius. So by increasing the radius by 5-10mm over Garretts they might actually be winding up with a higher inertia even though their wheels weigh less. This however is offset a bit because when you increase the radius of the turbine wheel, you're also increasing the torque arm length for the force of the exhaust acting upon it, making it easier to rotate.

There are a lot of things to consider when sizing the turbine, especially when working with a new material that provides a mass decrease. I'm sure BW has done their homework and are certainly smarter than me but I still think it would be interesting to see something like a 7158 or 7155. That ought to flow like a GT2871 but spool like a T25.

Mind you, this is all in terms of their "old" 1st generation EFR turbines. This new mixed flow wheel changes things.

[PoorMans180SX]

Originally Posted by LuckyX2

Is that a safe thing to assume? The mixed flow turbine seems like enough of a departure from convention that a 63mm mixed flow isn't going to consume air in the same way a 64mm radial turbine will. Compressor sides are nearly identical though.

Also, I'd like to see BW take their turbine wheels down in size a little; I'm curious to see how that would affect things. Currently BW has their turbines 5-10mm larger than the comparable Garretts and there are a few things to consider with that. They can make a larger wheel than Garrett and still have the same mass. The problem is that it's inertia that keeps the wheel from moving, not mass and inertia increases linearly with mass but squared with radius. So by increasing the radius by 5-10mm over Garretts they might actually be winding up with a higher inertia even though their wheels weigh less. This however is offset a bit because when you increase the radius of the turbine wheel, you're also increasing the torque arm length for the force of the exhaust acting upon it, making it easier to rotate.

There are a lot of things to consider when sizing the turbine, especially when working with a new material that provides a mass decrease. I'm sure BW has done their homework and are certainly smarter than me but I still think it would be interesting to see something like a 7158 or 7155. That ought to flow like a GT2871 but spool like a T25.

Mind you, this is all in terms of their "old" 1st generation EFR turbines. This new mixed flow wheel changes things.

The thing is, if you have big differences in turbine and compressor size, you lose turbine effeciency because the tip speeds are different, and power and transient response are sacrificed. Garrett has finally caught on with their new GTX3576. It's a GTX3076 compressor wheel with a 3582 turbine wheel. Spools exactly the same as a X3076 and makes somewhere around 50whp more.

Check out this article:

Turbo Tech: Compressor and Turbine Map Details

And this:

http://forums.evolutionm.net/evo-dyn...-gtx3576r.html

[inopsey]

Originally Posted by LuckyX2

There are a lot of things to consider when sizing the turbine, especially when working with a new material that provides a mass decrease. I'm sure BW has done their homework and are certainly smarter than me but I still think it would be interesting to see something like a 7158 or 7155. That ought to flow like a GT2871 but spool like a T25.

having the turbine and compressor wheels more off balance will not make a better turbo.