Sorry what i meant to ask was have you ever taken time based logging and applied a correction factor to it. In other words if you do a dyno pull and have a 0.98 SAE correction you are taking 2% off the recorded figures. So if you take some logging in realtime what are you thoughts about applying the same correction methods? As you quite rightly said doing testing on the road isn't perfectly repeatable due to varying weather conditions. Hope that makes sense this time.

 

Which is exactly the point we are trying to convey over the last few pages. So many times i have found the best dyno engine is not the fastest on track. Eventually i learnt to dyno in a way that matched logged data.
Thanks to all for the discussion so far, its pretty rare to find a forum with so many experienced guys on it most are full of knob heads.

 

It looks like the EFR is catching up at first glance, but I think its one of those graphs that when it comes down to it they are very similar turbos and you can read from it what you want to.

The EFR datalog is constantly at higher rpm, has an unstable level of acceleration leading into the "snap open" in comparison with the Garrett, which has a very very stable level of throttle opening and has very quick throttle opening, again from lower rpm.

The test is obviously done in a higher gear, which allows both turbos to load up a bit - and its pretty evident that the rpm that both turbos reach their full boost is similar... the fact that you can see the manifold pressure gaining on the Garrett as time passes is just an illusion, they are past the point of throttle response and are now in the territory of boost threshold which it is already known is comparable between the two turbos.

If you look at the intersection points of the manifold pressure and throttle position between the two, its pretty obvious that they've both come onto positive pressure with similar delay, despite the fact the test on GTX started with lower rpm. If the test was from equivalent rpm etc and that the EFR was that much quicker building boost, you'd see the the red speed curve gaining on the green one. The only way its possible for it to look how it does is if the two turbos spool similar but one has been asked to do it from lower rpm, and as such needs a little more time elapsed to reach the target boost.

I'd take that plot with a grain of salt for the most part, its hard to tell what would happen if all things were completely equal.

 

I now understand what you're asking. I have not applied a correction factor to it. I don't know about the accuracy of correlating the SAE correction factor (CF) to response time. I would think you would need to validate the time differential at several points (at altitude, at temperature, etc) to find a way to apply a correction factor to it accurately. You may find that you need to apply a "time" CF to the SAE CF used on the dyno.

We did a complete round of testing in a few hours at one location. Helps when you have a 2 mile high speed oval right behind the test cells When we did testing in Colorado, I found a straight section of road by Kremling and we beat that up for a few hours.

 

What a fabulous wealth of knowledge and information on this thread. There are a lot of great heads in here pushing the envelope to develop better products and faster cars.

The only issue I have with the GTX vs EFR line is that the GTX doesn't appear to be a turbo designed for the transient needs being hyped by the EFR line. Garrett themselves have come out saying the GTX was designed to support the drag racing classes and allow for more power using hte same basic inducer sizes. I believe the HTA turbos provide a better comparison when looking at transient turbo response and flow capability. I keep seeing the EFR8374 being compared to a plain 3582r. I don't know of anyone serious about making power running the 13 year old gt35r.

Here is a test I did on a customer car. I dump it in here as another piece of data. Variables are that the header was built equal length on a 2.3L 5 cylinder 20vt with a built bottom end and a stock head, 8.5:1. Fuel was California 91octane. This test was mostly just to look at spool, not max power. The tune was not changed between the 4 tests. The hotsides were TIAL .82 for both the 30r turbos and 35r turbos. All the tests were done within 2 hours from start to finish, same day, same temperature, same Land and Sea dyno.



So what is the GT results doing in the EFR page? Well, I think these EFR8374 graphs lined up with the GT35r are close enough to realize the HTA3582r is going to outspool the EFR8374. I would love to put the two on the dyno against eachother, but a year later, I still have no turbos!! Guess we will have to continue to wait I wish I could hold my customer's money for a year while I continue to do product development!!

And one last thing! I really wish BW would develop a v-band in/out housing that was a drop in for the Tial and PTE housings. While TS is awesome, 5 cylinders do not split well

 

 

034 Motorsports,
Why don't you get a hold of Geoff @ F-R and see if it's on the horizon for BW? If not you could twist his arm to start putting that idea in the mover-shakers at BW. I remember seeing your GTX vs. GT results when they originally came out. Thank you. I like your perspective on the GTX vs. BW... I hadn't really thought about it like that. I guess it shouldn't really be a surprise to find that certain turbochargers perform better at different disciplines.

You make an interesting point about all the contenders in the performance turbocharger industry. I think the people machining compressor wheels and putting them on old turbine wheels will soon find that running steel-nickel alloy turbine wheels is holding them back. I see BW improving upon their already great design and really going Audi 5000 on the competition.

 

Just to be clear I am not 034 motorsport. I work with them on products and I happened to be in California for a wedding to coordinate the tests. I am just a poor graduating engineer welding my way through college with a fabrication shop!!

I asked for a V-band option at Sema 2010(not to be confused with SEMA 2011 when they should have released the product). The engineers said they would not because the turbine was designed around a TS configuration. I asked again this year and rumor has it that Treadstone is working on something, but I have not seen anything on it for a while. If I had to guess, I would say the GTR crowd will eventually talk BW into a single scroll tial copy so they can support the space constraints on the v6.





The GT mismatched ratios do look terrible on paper, but the 650whp HTA3582r, 730whp HTA 3586r and 930whp HTA3794r sure do work well on the dyno, on the street and especially on the track!!

Hank

 

We've made 690whp (520kwatw) on a plain GT3582R on E85 and 35psi boost dropping to 29psi at redline.
And that's with a 1.06 T3 Twin scroll housing.

 

With all due respect, that is just an arbitrary number and dyno variance. We have run the 3582r all out(ie with the wastegate line disconnected, 36psi at TP down to 27 by 8500) back to back with the HTA counterpart all out(ie with the wastegate line disconnected, 39psi TP down to 32) on e85 with identical Tial .82 housings. The HTA wheels do end up making 9-12% more. For relativity, our Mustang dyno locally showed 220WHP for a stock EVO9 the same day we put down 612whp/520wtq with the hta82 on an original Quattro. 3 2007 RS4(factory rated 420hp) put down 280-305 with exhaust and chip that same day.

 

When you return, could you provide a source for your "rule of thumb" statements?

I think back and I realize you are the only person I've seen make this claim like there was data behind it. I've read a ton of data on turbo design and have never came across anything similar, I'd be very interested in reading this data if it is available.

Hydra, I've been paralleling your thought pattern for the last month or two. While I agree with your general approach, you are basing it off some major assumptions, that while are reasonable, they leave a lot open. It would be worthwhile for somebody to model up the wheels though to get a realistic idea of inertia.

Just FWIW, I think that there isn’t a one size fits all solution here on which brand to buy. I think Garrett will continue to perform well in some racing and BW may take over in other forms of racing. I don’t see either dominating the other though in all brackets, or even a majority. I’ve given up on the EFR stuff for now simply because of the size of turbo I’m interested in. It’s small enough that either Garrett’s or BW’s offering will be good for my needs.

 

The engineer in me likes all this techno talk, but until there is more feedback from installed EFRs: really simple dyno numbers, dyno charts, butt-dyno response, 1320' and circuit times, we are just hopping it to be what it claims and I really wish them to be. To the people with EFR, wish you the best to finish your projects and give the community your opinions.

 

03whitegsr,
You're right, I think the best conclusion we can draw is that the EFR is another market offering that may slot somewhere in between certain garrett (and other) products. But I very much doubt its the revolutionary new envelope-pushing product we were led to believe it was. I guess time will tell where the truth lies...

 

Agreed, it started off looking good but now there is nothing left other than undeniable proof being shared with the public to make these things seem anywhere near good as purported. Look forward to hopefully seeing a few leaned on and experiences shared from neutral parties

 

whoa! this thread got off track. hydra, you are close on some things and way off on others.. please consider what i mentioned about match ratios and then realize the biggest benefit of the gamma-ti is that we can use a lighter turbine wheel without compromising great ratios. if a small turbine wheel is used with a large compressor wheel, that by definition means the turbine wheel will be spinning slower for a given unit of airflow. the B2 turbines work well in twinscroll as perrin's test shows - however they are optimized for twinscroll use.

ps 03whitegsr - the person who originally clued me into match ratios ~6years ago was a garrett engineer. Then it was validated by bw engineers about 4 years ago. I will check my stuff to see if i have any sae journals mentioning