Gtx Compressors - Did Garrett Set The Bars Right ?

[discopotato03]

This is probably going to be a theoretical thread because none of us work in the development halls at Honeywell .

A search I did came up with this thread that Mike Kojima did a while back .

http://www.motoiq.com/magazine_articles/id/1802/pageid/2596/an-inside-look-at-the-new-garrett-gtx-turbochargers.aspx

Note this is the last of a five page article but it mentions that Garrett have ways of adjusting things to get the compressor airflow results they want - or setting the bar for each at a given point if you like . It also mentions the type of milling process they use to control cost and speed up production .

If you search this stuff you come across other machined wheel suppliers and they make a few sizes inbetween the ones Garrett does . For example Garrett do 60 63 67 71 76 82mm etc versions of their GTX wheels , they get larger but are beyond what most of us consider usable . FPs HTA range inclide a 68 71 73 76 82 and more so you can see a few inbetween sizes appearing .

I will go back and correct the GTX sizes later because many are actually slightly larger than the advertised major wheel diametre .

You have to give the turbo manufacturers credit for continuing development though much of this is being driven by more moden engines and the regulators emissions and consumption demands . Note Mike mentions that typical boost pressures have been rising and some like the Evo 10 use something like 21 pounds of boost out of the box .

The aftermarket portion of Garrett is small fry compared to OE production and they obviously have peformance targets for their aftermarket turbochargers . And they have a market to chase and you'd think the volume sales would be largely in North America .

I think its fair to say that there are some gaps in the currently available GTX wheel range and my gut feeling is that they could use something in the 68-69mm area and also the 73mm area . Its been seen here that the GTX3071R can push real good performance but people are doubting the GTX3076R and it may be a case of overkill for the GT30 UHP turbine .

No ones done much with the GTX3067R yet and I have to wonder if it could have had a slightly larger 68 or 69mm wheel .

With a few more sizes like say 69 and 73mm I think Garrett could safely cover most peoples needs of power vs "response" .

Thoughts ? Cheers A .

[GTScotT]

Realistically a TD06SL2 20G covers the 68mm gap, and that is exactly what FP sell. The 68mm wheel is from their MHI range of turbos and replaces an OE 68mm wheel on the said OE MHI turbo.

There are results on this thread already for what a said SL2 20G can do, they are in the upper 200kw bracket on pump and an easy 300 on E85.

Stao at hypergear has also redeveloped his own 20.5G billet item which recently maxed out at 350kw with amazing response. My thoughts are that people should explore the left field options also before consistently relying on garrett results.

[discopotato03]

Well thats one opinion but you can also look at the other end of the equation and explore where these things start to work as well as when they're going off the scale . The question I'm asking is did they get the maximums right to start with as where those numbers fit in with what the market wants - realistically speaking . Its obvious that they were aiming to have more out of each size in the GT range ie GT60mm through GT82mm which is where much of the market is anyway . We all know having excessive compressor capacity WILL add turbine lag because the things trying to churn more air than the user needs , someone here recently said that their GTX3076R is a bit disappointing as in that its lazy to get going . If it had been a 71 or maybe 73mm X wheel the story WOULD have been different to some degree . In the 60mm GT30R group compressors now range from 67mm to 82mm .

20G is basically the big trim version of the 16G 18G 20G group . From memory the 20G was native to some OE TD06 turbos and the reason they sometimes end up in TD05 "Hi Flows" is because its physically an easy change to make , basically swap wheels and remachine the compressor housing to suit . Laggy conversion on a TD05 turbine which is why the Evo crowd think they are dogs .

A .

[GTScotT]

Yep, we certainly arent suggesting a TD05H 20g.

However a TD06SL2 turbine is something you yourself have referred to as being wanted in the market - 61mm turbine with a mid 70s trim.

Hypergear took that and added his billet compressor to the following E85 based effect:

http://www.skylinesaustralia.com/forums/topic/261613-hypergear-hiflow-service-continued/page-345

[discopotato03]

GTX2860R 60.0 mm 58T 2.50 Bar - 40.0 lbs - 400 Hp .

GTX2863R 63.4 mm 56T 2.50 Bar - 43.0 lbs - 430 Hp .

GTX2867R 67.0 mm 55T 2.50 Bar - 47.0 lbs - 480 Hp .

GTX3067R 67.0 mm 55T 2.50 Bar - 47.0 lbs - 500 Hp .

GT2871R ..................48T 2.25 Bar - 39.0 lbs - 460 Hp .

GT2871R ..................52T 2.25 Bar - 45.0 lbs - 475 Hp .

GT3071R ..................56T 2.50 Bar - 49.0 lbs - 460 Hp .

GTX3071R 71.4 mm 58T 2.50 Bar - 55.0 lbs - 560 hp .

GT3076R ..................56T 2.50 Bar - 52.5 lbs - 525 Hp .

GTX3076R 76.6 mm 58T 2.50 Bar - 62.0 lbs - 640 Hp .

GT3582R ..................56T 2.50 Bar - 61.0 lbs - 675 Hp .

GTX3582R 82.5mm 58T 2.50 Bar - 67.5 lbs - 750 Hp .

Some of these numbers , from Garrett Vol 5 Catalog , don't always add up and in some cases they pushed the Hp up depending on hot side improvements .

Note the GT3582Rs Hp looks high on the pounds flow number and in the catalog they show different maps depending on if the comp housing is port shrouded or not . I think 61 pounds for the std and 67 for the PS'd one .

I threw in the GT2871Rs for comparison and you'd think 39 vs 45 pounds flow would net more than a 15 Hp difference . Maybe the 0.64 AR turbine housing is the Achilles heel of the GTRS and not the cold side .

Cheers A .

Edited February 26, 2013 by discopotato03

[Dale FZ1]

I seem to recall the 2871 range going 390hp, 420hp, 450hp for the 48T, 52T, 56T.

Most of Garrett's ratings seem to factor use of the biggest A/R turbine

[discopotato03]

I agree , its just what they mention in there Vol 5 Catalog

http://www.turbobygarrett.com/turbobygarrett/productcatalog

A .

[GTSBoy]

Is thsi not just a case of GTX compressors, being optimised for high PR, not liking to work at lower PRs as they spool up. These are, for want of a better term, "ultra-high-performance" compressors, and you'd somewhat expect them to work well on a race engine that you kept spooled up most of the time.

Complaining that they don't have as good response as other less over the top compressor designs (plain GT being the obvious comparison there) seems like a waste of time. At the same time, asking for a smaller compressor size than any given available size is only asking to cut off top end flow capacity for improved response.

If the engine is built needing to run to high PR in order to get flow through it, then the smaller wheel probably won't be a good match. If it could work, then it probably could have worked without a GTX compressor, no?

[discopotato03]

No its not a waste of time at all , its wanting the range of operation to suit a reasonably conservative ask .

The sizes I think could be usefull are in betweens and larger than the smallest current GT30R GTX compressor . I very much doubt many here are going to use 3 1/2 bar of boost very often if ever . I also don't see a manufacturers race team lugging a GTX box from the local vendor and tossing on their race engine .

More likely they would use something hand built like say a TR30R like the Audi and Rally teams used . To them 10Gs is so what .

A .

[GTScotT]

Where can a laymen like us get the fabled TR30R disco?

This I would love to know.

[GTScotT]

You know for a thumb suck you have done really well as a guesstimate on how these turbos will match.

A 3076R generally DOES want to sit at about 18psi and Mick_O definitely did hit a wall at 25! Generally 14psi is also all that you need from a GTRS also, and 17psi is over stressing the thing. Pretty sure garden variety 3071s also seldom see anywhere near 20psi also.

Given there are other variables to be considered I think that plot is a damn fine representation for 'rule of thumb'.

[AngryRB]

Really good post, can we add a rough guide to a relationship between the lbs/min and rwkw

[GTSBoy]

Nevermind. Wrong thought.

Edited February 27, 2013 by GTSBoy

[wolverine]

I appreciate the time and effort you put in on these topics Lith.

[GTScotT]

I appreciate the time and effort you put in on these topics Lith.

+11ty

[AngryRB]

+1

[discopotato03]

I can't fathom the reasoning in that graph . What I do know is that it takes X pounds per minute to support Y horsepower and boost pressure is how much extra it takes above atmospheric to drive it into the cylinders .

I agree that GT28 turbines and housings are lacking on an RB25 but the issue is its hard to find an acceptable middle ground turbocharger .

I was speaking to Scott today and he said he finds exactly the same thing . Good small turbos good bigish power turbos but SFA in the middle .

I reckon what this illustrates the gap in the GT28 to GT30 turbine/housing area area and the band aid fixes aren't cutting it .

Yes Lith we know GTX3071Rs work fine and the GT30 hot side is a good match for the 71X cold side . If a 0.82 T housing suits you along with the 55 lb/min cold side . Tough shit if you want a 50 lb/min cold side and the up to date compressor family . Currently the choices are 47 lb/min or 55lb/min and while its only a 8lb/min difference its right through the area that f**ks everyone up . This is why I think Garrett undershot and overshot at this point in the GTX line up and hence the "setting the bars right" question .

And its only half the problem . There is still no answer to the turbine families void and the cropped GT30 turbine is one a half hearted compromise and two only usable (properly) in HKS turbine housings - namely the 0.68 AR GT Pro S one .

I will be brave and say there is little you can do to increase the flow of a GT28 NS111 turbine with currently available bolt on Garrett turbine housings .

There is SFA you can do to a 60mm 84T GT UHP turbine that doesn't choke exhaust flow up high engine rev wise .

They can screw around all they like with the aluminium bits but in the end to fix the problem we need a better turbine solution . Lith you are probably thinking suck it up man there is no answer and thats correct - with Garretts currently available turbochargers .

Have been shown another path that doesn't involve Honeywell anything so thats where my searching is going now .

Andrew I'll be in touch soon to hear the fine details of yours vs the 71X car .

A .

[Nee-san]

I can't fathom the reasoning in that graph . What I do know is that it takes X pounds per minute to support Y horsepower and boost pressure is how much extra it takes above atmospheric to drive it into the cylinders .

I agree that GT28 turbines and housings are lacking on an RB25 but the issue is its hard to find an acceptable middle ground turbocharger .

I was speaking to Scott today and he said he finds exactly the same thing . Good small turbos good bigish power turbos but SFA in the middle .

I reckon what this illustrates the gap in the GT28 to GT30 turbine/housing area area and the band aid fixes aren't cutting it .

Yes Lith we know GTX3071Rs work fine and the GT30 hot side is a good match for the 71X cold side . If a 0.82 T housing suits you along with the 55 lb/min cold side . Tough shit if you want a 50 lb/min cold side and the up to date compressor family . Currently the choices are 47 lb/min or 55lb/min and while its only a 8lb/min difference its right through the area that f**ks everyone up . This is why I think Garrett undershot and overshot at this point in the GTX line up and hence the "setting the bars right" question .

And its only half the problem . There is still no answer to the turbine families void and the cropped GT30 turbine is one a half hearted compromise and two only usable (properly) in HKS turbine housings - namely the 0.68 AR GT Pro S one .

I will be brave and say there is little you can do to increase the flow of a GT28 NS111 turbine with currently available bolt on Garrett turbine housings .

There is SFA you can do to a 60mm 84T GT UHP turbine that doesn't choke exhaust flow up high engine rev wise .

They can screw around all they like with the aluminium bits but in the end to fix the problem we need a better turbine solution . Lith you are probably thinking suck it up man there is no answer and thats correct - with Garretts currently available turbochargers .

Have been shown another path that doesn't involve Honeywell anything so thats where my searching is going now .

Andrew I'll be in touch soon to hear the fine details of yours vs the 71X car .

A .

I agree, and the common twin turbo pairings also have this problem.

-5 straight to -10.

This is why I will be using a Borg Warner turbo.

Correct turbo selection is probably something people spend the least amount of time on, which is insane.

[GTScotT]

That's a pretty serious suggestion imo, and shouldn't be overlooked. To arm people with facts this could easily be built by hypergear.

In fact the HG highflow profile is actually a small trim 76mm compressor on a T3 turbine of some sort. Something im also going to be asking him about.

the said highflow unit is said to be nonsensical in stock housings for response, and is track/hill climb proven. Apparently they breeze to full boost on the highway with the lightest throttle.

I've also been out on the road with one, chase cam from a T28 SR.... The particular model was a PU highflow which has the .82 rear housing on the same unit. The car was tuned @ 250kw by jem but had some issues, vct wasn't working for starters and it was running insanely rich. But... No sign of resistance. The T28 SR with its OEM response had nothing over it, and 250kw is obviously out of ballpark for a T28.. so the speed was immense.

I am confident in saying an OP6 variant would meet and exceed Disco's demands, if he can stomach the thought of a homebrew turbo.

[discopotato03]

No not really .

What people need to understand is reasoning behind Garretts aging UHP turbines in big trim form , which is most of them .

The whole point of getting away from ancient T series turbos like from T3 T4 , I'm talking real T3 not just any olld T3 flanged dryer BTW , was to make the units smaller lighter and cheaper to produce . Also to make them faster responding things that put less heat into the air under boost .

Three important things to think about .

Firstly smaller lighter higher speed wheels can often pump as much air as the old "T" bone wheels .

Secondly compact rolling element bearings shit all over bushes and thrust plates friction and load carrying wise , they can take significantly higher shaft speeds and use less oil doing it . BTW think annular contact not just "ball bearing" or the "roller bearing" which is not a "ball" bearing at all .

Thirdly for their flow range GT UHP turbines are smaller diametre and larger trim things than their T Series equivalents . A good example is a GT35 turbine which is approximately 68mm 84 trim . The T04 museum piece is 74.2mm and available in three trims or N O and P , From memory the P was the largest at I think 76T and the O around 69 trim and can't remember what the N was - guesses about 63T .

The GT35 is going to be lighter and more compact than a P trim and is supposed to perform about the same flow wise - in a GT30 turbine housing .

This is why when the T04Z (T04 R actually) turned up people started high flowing GT3582Rs with the T04R compressor wheel .

You gets the Inco turbine with the Ni Resist housing plus ball bearings and machine out your own compressor housing . And you don't get bent over by HKS in the process .

The GT30 UHP turbine was a really good thing in its day and to crank 540 odd horse power with a 60mm diametre turbine was pretty flash - in 1989-1990 . Wide exducer tips schmich looking paddle shaped blades and that big trim size to make a medium sized turbine flow 500+ worth of exhaust gas .

You bloody beauty finally something less laggy than an O or P and this time - its got balls !

So time goes by and road performance engines are getting smaller or more performance is needed from the smaller capacity existing ones , basically people want the performance over a broader range starting earlier .

The manufacturers race teams also wanted better and had the bucks and the drive to win at international levels . Garrett had reason to develop turbochargers that wiped their arse all over what HKS though was good "power and response" . The ask for a true competition turbocharger is a lot higher than any OE or aftermarket unit because so much time effort and money goes into competing at the top levels . Also reputations are at stake and winning means selling your product .

Through the 90s tarmack racing was on the nose so competition turbocharger development was being aimed at rally cars ie Grp A .

The ask was for a lot of boost soonest but with a restrictor limiting maximum flow . It forced Garrett to think very had about turbo size weight and innertia and they had to come up with wheels that worked in this environment . You can throw all the T and GT UHP turbines in the bin here because they don't work , too big too heavy not efficient .

Then they think yeah this NS111 thing works well so why not make it a bit bigger , erm - maaayyyybe 60 mm but in smaller trims - like say 63 and 76 .

Now if you remember the 53.84mm NS111 is available in 62 and 76 trims , don't think this was by accident ...

Its now history that the TR30R series turbochargers worked well and put many cars in the winners circle .

Great , but what about the aftermarket turbocharger scene . Well there was that time when Garrett wouldn't sell the HKS spec GT BB turbos and the just different enough pretenders were a bit yawnsville . Also you had problems if you wanted a T3 flanged IW BB turbo because the IW GT30 turbine housings didn't exist in those days and have not been around too many years in fact .

So you get to the stage where you can have a few T3 flanged GT28 based BB turbos and Garrett rocks up the the GT30 IW ones so you can use GT30 whatevers on your 280ZXT/VLT/FJ20/RB20-25-30 . Alls good till you work out that she isn't as lively under 3000-3500 than you like it to be and suddenly the gap becomes obvious in turbine family performance . The easy fixes are tried like compressor wheel trims and port shroud housings - even a drop in turbine housing AR . All have compromises and none really fix the problem .

Lith theres nothing stopping you from using those T series BB hybrids as Garrett calls them but the TA/TB turbines are nothing like a GT28 or GT30 - bigger diametre and narrower , nossle shape in a GT turbine housing is wrong for them as well . Could use a TB31 turbine housing I suppose .

Yep that T04E 50 trim compressor with the nice map . I've got one of those here Brett brough me back from Sema too many years ago and woud have had it in some crazy hybrid until I started working out what was what with Garretts GT BB range . A lot of research went into that from countless sources and that GT30 turbo/cartridge list I posted here ages ago was the result of much of it . I wanted people to know what was out there and how not to get fobbed off by turbo agents etc , how to get into the HKS spec gear and not get raped for it .

Now to the GTX era . Currently nothing has changed turbine wise and I reckon that while there are some minor efficiency and not so minor capacity gains to be had the two major improvements have been port shrouded comp covers and that they are QUIET , more an OE gain than a petrol heads one .

The answer Lith is to give us the turbine or one like it thats known to work and maybe even think about a GT35 sized one . People work out eventually that torque is more important than horsepower and as long as the torque ox revs reasonably freely its all good .

Me ? No I don't think GT28 GTRS performance is sparkling . It is to a degree a step on the path to learning what I want at around town speeds and realising that you can have that along with larger turbo potential performance . I know for a fact you are all over EFR turbos and some of them can do truly amasing things . A couple of the latest ones and some due out soon give staggering performance for their size and are going to be a very hard act to beat .

Do I think Garrett people sniff around these forums for feedback ? Do I think they keep an eye on those pointing out the gaps in their ranges . Am I bitching about these things in the hopes that others start noticing the gaps and demanding fixes or shopping elsewhere . You tell me .

I think better turbines are in the pipe ATM but unfortunately these things take time . Really unless there is some OE app needing a responsive 4-500 Hp capable turbine it won't get much priority .

A .