Jump to content
SAU Community

Garrett G Series


LaurelPWR

Recommended Posts

  • 2 weeks later...
On 06/10/2020 at 5:08 PM, IB_8229 said:

Folks, reading through the last few pages of posts here, let me chime in with my personal experience on not one, but two Garrett G35-900 turbos i have been through. The first failed at ~2030 km. the second failed at ~2300 km.

The nature of failure was very gradual, with a faint, low frequency "Police siren" or "wooooo" noise accompanying spool, that gradually got louder to the point where the compressor inducer was seen to make contact with the compressor housing, and the shaft was bent. 

Oil feed lines on both were brand new. No oiling issues, as all Oil parameters are monitored. No coolant supply problems either. 

Tune was carried out by the most respected JDM Builder and tuner here in Dubai, and happens to receive, build and tune vehicles from all over the middle east and north Africa.

The first turbo was sent to Garrett for inspection, and they responded with the following: 

Claim.jpg.f41bf22239d9621cef8165578cd4eaac.jpg

880695-5001S

Analysis Findings : Rotor damage - Overspeed -
Technician Conclusion: Turbocharger returned with rubbing wheels vs. contours. Damage of compressor piston rings and overheated of ball bearing as a consequence of Turbo overspeed. During overspeeding cycles, the excessive growth of the compressor wheel may cause a permanent deformation (growth) of the central bore of the wheel, which forms into a barrel shape and is characteristic of a wheel which has been taken beyond its normal operational speed limit. Recommendations: The reason for turbocharger overspeeding needs to be investigated on the vehicle. Possible contributing factors: performance tuning, fuelling problems, poor combustion causing high level of deposits in VNT (vanes blocked). Actuator not controlled correctly: ECU problems, command from ECU not reaching actuator, wiring/connection faults, actuator hose leakage, PWM valve faulty, hose blow-off etc may be contributing factors. Blocked/restricted air filter or hoses.

The turbos failed due to overspeed, and were not happy pushing 2.5 bar of boost for extended durations. Both failures were consistent in nature with respect to how they occurred as well as the time period/mileage as the tunes and driving style were more or less identical in nature. 

I did not bother sending in the second unit for a claim. 

I would recommend Either Garrett GTX Gen 2 as those are extremely robust, (I have run a GTX3071r , and GTX3076r in the past) or Precision CEA to avoid going through similar hassles.

I am now swapping in a Precision 6466 CEA Gen 2.

Hope this helps some of you decide. If you do purchase a G Series Garrett turbo, don't make the mistake i made; do buy a turbo speed sensor, and follow it religiously as part of the tuning process. 

anyone else had G35 900s or similar fail from running 2bar of boost? 

Link to comment
Share on other sites

42 minutes ago, hattori hanzo said:

anyone else had G35 900s or similar fail from running 2bar of boost? 

I'm not aware of any "random" failures.  I feel like I heard of people running into issues when going on a mismatch combo but it may be from this thread, or associated with the guys posting it - the general agreement at the time is that it due to overspeed which thanks to the more efficient G-series wheels etc it becomes a lot more likely than earlier Garretts so the idea would be to keep within the recommended "area" of these things.

To get an idea of if 2bar-ish is a risk, the max wheel speed published for the G35 900 at PR3.0 lines up with around 79lb/min

image.thumb.png.e15cce032a01d6ecbd96ac1de578faf1.png

I'd not necessarily expect that line to be a "it asplode when you go past this point" level, but realistically the further you go from that line the risk climbs exponentially.   Ironically the anti-Borg Warner silly people chuff at the idea of wheel speed sensors being a good idea as though they're the first turbos to be efficient enough to be able to blow themselves up if you let them keep pumping air, but this is why it's not a bad idea.  It's worth noting at that map that it looks like compressor efficiency is actually QUITE solid still at the maxrpm which means it'll be super easy to overspeed one of these things, the common method of waiting until it stops making power easily is likely to get you way off the map if you're going for >2bar.

The "next best" idea is estimating your airflow and picking where on the compressor map you land based off that and your intake manifold gauge pressure.   

My super bro-math way of estimating required airflow with a 2wd car on a hub dyno is relating lb/min to kw with a factor of 7.1.... so if you wanted to know what is a "safe max" point to push a G35 900 on E85 on a hub dyno at bar would be 79*7.1 = 561kw.  If you pushed one to say 600kw you could also go 600/7.1 and find you're likely pushing in the territory of 85lb/min of air which is starting to get a bit up there but shouldn't be COMPLETELY insane.   Ymmv as different dynos, different tuners etc will give different results

A lot of modern ECUs support VE based fuel equations, and if the tuning is done competently (injector/fuel/charge temp etc settings all calibrated accurately) then the ECU will have a pretty good idea of the airmass getting into the engine as otherwise they'd not be able to command the correct injector flow to hit target AFR.    If you have a Link ECU for example you can check a value call "airflow per cyl (estimated)" or something like that.   

If you get that value and throw it at an equation like "(percyl * 6 * rpm / 2) / 453.6" which takes the g/s airmass the ECU is assuming each cylinder it's getting and converts that to lbs of airflow consumed by the entire engine over a minute.  I've found it to be relatively on point for cars I've tuned so far... this again should be <80lb/min at 200kpa MGP if you want to be safe.

 

 

Edited by Lithium
  • Like 3
Link to comment
Share on other sites

On 10/10/2023 at 3:19 PM, hattori hanzo said:

anyone else had G35 900s or similar fail from running 2bar of boost? 

Can you please elaborate on your full setup? Im interested to see how the match plots

-altitude
-fuel used
-exact engine displacement
-max engine speed (RPM)
-max boost target (2 bar or 2.5 bar?)
-what rpm are you fully spooled and WG open
-cyl head and cam spec
-downpipe / exhaust size and config, muffler used, air filter setup, etc.  
 

. The g series are optimized for response and low inertia, while I have not blown up any of the G35-900 and i absolutely love the turbo, I saw (2) g25's die from high altitude overspeed many years ago and learned to respect turbo speed the same as any other high end piece of equipment. 

edit:
@lithium - funny that the big drag race precision turbos have speed sensor ports now 😁

 

Edited by Full-Race Geoff
  • Like 1
Link to comment
Share on other sites

9 hours ago, Full-Race Geoff said:

I saw (2) g25's die from high altitude overspeed many years ago and learned to respect turbo speed the same as any other high end piece of equipment. 

edit:
@lithium - funny that the big drag race precision turbos have speed sensor ports now 😁

 

Good call on the altitude, I even thought briefly about that when I posted the comp map but so used to people living close to sea level that I didn't dwell on it...  I guess there are a lot of areas people live that are high enough that the relatively "not high boost friendly" shape of the G35 900 map would mean it'd not be hard to over speed them when trying to make their rated power even at not a particularly high altitude.

To add to that, there are a LOT of people likely running around out there with boost leaks.   Like people think they've gone above and beyond if they've done a solid spanner check, or even smoke tested their pressure sides of turbo setups - before tuning where possible we'll try and pump 20+psi through the intercooler piping and the amount of times we've heard rushing air on a "nah it's definitely all sealed" setups is solid, potentially in the area of 50%+. I'd not be surprised if the reason a lot of the car's my little group have done that have ended up with "better than people expect" for the setups are if anything an indication of how many people are running around with undiagnosed issues.   

If you're one of the MANY people who are running around with a boost leak then yeah, turbine speed is the only way to be sure unless you pressure test it - I am certain some of the EFR horror stories come from boost leaks, they were basically booby trapped by having a selling point of coming with an internal BOV that just happened to often leak at high boost.  Perfect.

Re: Precision, yeah I've seen some data from people testing big Gen2 and NextGen Precisions and while the Precisions aren't the most "low inertia" turbos it becomes clear from the turbine speed stuff that some of the "magic" with them comes down to the fact they can actually tolerate a hell of a punishing.   Like its not unheard of to have 76mm Precision turbos being run at speeds (as in turbine rpm, not tip speed....) that you should be starting to think about calling it a day with a G35 900 hahaha.

 

 

 

 

  • Like 1
Link to comment
Share on other sites

On 12/10/2023 at 10:18 PM, Full-Race Geoff said:

Can you please elaborate on your full setup? Im interested to see how the match plots

-altitude
-fuel used
-exact engine displacement
-max engine speed (RPM)
-max boost target (2 bar or 2.5 bar?)
-what rpm are you fully spooled and WG open
-cyl head and cam spec
-downpipe / exhaust size and config, muffler used, air filter setup, etc.  
 

. The g series are optimized for response and low inertia, while I have not blown up any of the G35-900 and i absolutely love the turbo, I saw (2) g25's die from high altitude overspeed many years ago and learned to respect turbo speed the same as any other high end piece of equipment. 

edit:
@lithium - funny that the big drag race precision turbos have speed sensor ports now 😁

 

- Perth WA is at sea level 

E85 

2760cc 

7700red line, however peak power is circ a7250 

2bar 

Kelford 182-c - 272 cams 

4" dump, 4" exhaust, 100 cell cat

air filter is a pod on a intake tube like the below 

 

 

 

 

WhatsAppImage2023-10-03at15_21_09.jpg.239db4bcf2e87e9f6333337b683e48dc.jpg

Edited by hattori hanzo
Link to comment
Share on other sites

Just now, hattori hanzo said:

- Perth WA is at sea level 

E85 

2760cc 

7700red line, however peak power is circ a7250 

2bar 

Kelford 182-c - 272 cams 

4" dump, 4" exhaust, 100 cell cat

air filter is a pod on a intake tube like the below 

 

 

Just to clarify, has yours failed or are you worrying about how risky you're playing it?

Link to comment
Share on other sites

1 minute ago, Lithium said:

Just to clarify, has yours failed or are you worrying about how risky you're playing it?

No I am just wondering on how risky I am, I did a quick calc and believe i am sitting at around 2.3 pressure ratio so a bit of room up the sleeve. 

Link to comment
Share on other sites

2 minutes ago, hattori hanzo said:

No I am just wondering on how risky I am, I did a quick calc and believe i am sitting at around 2.3 pressure ratio so a bit of room up the sleeve. 

2.3PR is around 19psi boost at sea level if you have zero intake restriction - you sure you aren't missing a PR there?

Link to comment
Share on other sites

14 minutes ago, hattori hanzo said:

possibly calculated incorrectly. by no means an expert hence asking 

K, I've thrown some numbers loosely based off your dyno plot and what you've said into BW Matchbot.  I've selected the EFR8474 compressor map because it happens to be an excellent match for what you're actually doing and also funnily enough Borg Warner doesn't have the Garrett data ;) I used turbine flow data for the G35 900 with 1.01 hotside (as the 1.06 T4 isn't listen) to loosely match the turbine data.    

Very loose but should give a general idea of what EMAP and pressure ratio etc might be doing.

https://www.borgwarner.com/go/3I1H4G

 

  • Like 1
Link to comment
Share on other sites

2 hours ago, hattori hanzo said:

possibly calculated incorrectly. by no means an expert hence asking 

PS part of why I did the above is so you can see your setup is likely reaching as high as PR 3.3

 

image.thumb.png.f55463309630aa461c67744e9d3126fb.png

Edited by Lithium
Link to comment
Share on other sites

  • 2 months later...

Wow, it's been 2 years since my last post, but I finally have good news to share about my G30-770! Last comment is here back on page 19.

 

TL;DR: twinscroll 1.06 with reputable manifold is better than single-scroll 0.83 with chinese manifold (shocker). Around 500rpm faster spool and +30hp everywhere (ish).

Changes:

The blue line is where we ended; red is just a random run that hit 22psi. At the beginning of the day, O2 correction was up to 14% at the top end, which fairly directly indicated a power/efficiency increase compared to the old setup. At equal boost, its about +30hp/30ftlb (ish), but the most noticeable change is the 500rpm faster spool making it way more drivable in the mid-range, no longer feeling like a pig. Do notice, we settled in at lower average boost and flattened the torque curve for drivability, where on the old sheet we held boost and rode the torque wave. Along with not-quite smoothed boost, not having the old run saved and different correction factor (oops), it's a bit difficult to compare old/new dyno sheets but the real world feeling is drastic and can be observed on the dyno. I hope this is still a helpful conclusion for on-lookers.

Conclusion: good parts are good, twin-scroll works, and I'm pretty happy about the response now, resolving my dyno sheet envy.

While I'm sure there's still room for improvement, I'm happy enough to not want to make changes. I'm sure a G30-660 would respond better and be a better fit at this power level. Surely a rebuilt and high-comp motor would respond better (and could take more boost/rpm). Oxygenated E98 would also come on boost better (this was pump E60). Maybe an EFR would still do better. Single-gate for TS seems fine in my application, but maybe there's a subtle advantage to true dual-wg (i.e. scavenging). Is the smaller runner size compared to 6boost beneficial in the mid-range? Hard to quantify without spending a crap-load more money for diminishing returns, but at least the low-hanging fruit was successful.

rb240_dyno_12-21-2023_censored.JPG

  • Like 3
Link to comment
Share on other sites

  • 2 weeks later...
On 12/30/2023 at 12:42 AM, gik0geck0 said:

TL;DR: twinscroll 1.06 with reputable manifold is better than single-scroll 0.83 with chinese manifold (shocker). Around 500rpm faster spool and +30hp everywhere (ish).

Conclusion: good parts are good, twin-scroll works, and I'm pretty happy about the response now, resolving my dyno sheet envy.

rb240_dyno_12-21-2023_censored.JPG

Convert single scroll to bigger twin scroll, yet spools 500rpm quicker and more hp everywhere. Imo, twin scroll works great.

  • Like 1
Link to comment
Share on other sites

  • 2 months later...

I know this thread is quite old but I wanted to chime in. 
 

I currently have a rb25 neo with a g30-660 and I am getting some wonky results. I’m hoping someone smarter than myself can offer some new info.

car made 373whp/300tq at 19psi on 93 octane. Tuner said the vehicle suffered from “too much back pressure” and was causing low numbers. This is peculiar considering the car is a 3” system and that should be plenty for that power level.

Boost comes on exceptionally quick and the throttle feels more like an on/off switch. I feel virtually no spool. 

 

Stock Comp/stock cam rb25neo

Plazmaman Plenum

DBW

ID1050x

Haltech 2500

6boost manifold

g30-660 1.01 A/R

93 octane


Thoughts?

9D56FBD0-E142-4B1B-9167-0391AD61F8E7.jpeg

Link to comment
Share on other sites

Also yeah, that is a reasonably good result for petrol at 19psi. I'd say this is working as one would expect. If you dialled in 25psi into it it may be interesting to see what it'd do, but because you're on 93 (hello USA) the answer is probably knock.

Almost all of the results you read here are on E85, which makes dramatic differences.

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now



×
×
  • Create New...