Borg Warner EFR Series Turbo's V 2.0

[burn4005]

nice to see a tuner ramping in power instead of having the 5000rpm torque mound.

 

would be a hoot to drive.

For context, I have 490rwkw at 27psi on a 2.6 8374 and that was doing 126,000rpm on the turbo (maxed) so there would still be a bit left in the 8474 for sure.

Edited April 19, 2024 by burn4005

[TurboTapin]

On 1/18/2024 at 8:10 PM, Komdotkom said:

Thanks gents, your input is very much appreciated

Clearly I'm not smart enough to work this out, who do you engage as a consultant to spec a turbo for something like this? I wouldn't know where to start looking for the right kind of boffin for this.

Turbomachinery engineers. They're just mechanical engineers who generally studied in universities focused on oil and gas, aviation or that had a formula SAE competition team. We have an entire team of them in our engineering department... they geek out over data and performance testing. I worked with them a lot when we developed an anti surge system for centrifugal compressors ran by electrical motors (~20 000-50 000HP) to avoid surging on accidental power loss. 

They laugh when I mention automotive turbochargers.... it seems they're lightyear's behind in technology due to poor funding. If you think formula 1 spends a lot of money, a single turbine/centrifugal compressor package in my line of work including the balance of plant equipment generally ranges in the 150-250Million range depending on size. 
 

[Borci88]

Just trying to revive this thread... Probably a few months off completion though. 

[Beelzebub89]

Has anyone seen any results with an 8370 on an rb26 yet? Seems like it could be a good option sitting between 7670 and 8374 

 

https://www.full-race.com/borgwarner-efr-8370-turbo

[burn4005]

8370 Only makes sense to me at altitude where you need the extra PR on the compressor and gain it on The turbine. The 8374 turbine is restrictive enough at sea level.

[Beelzebub89]

Restrictive when flowing close to 500kw though yeah? just thinking if 400 to 450kw is the goal then the 8370 fits in that spot. Is the 70mm rear going to be restrictive at that power level? On the 7670s isn't it the compressor running out of puff at 400kw rather than the rear being restrictive?

[Borci88]

On 27/08/2024 at 7:49 PM, Borci88 said:

Just trying to revive this thread... Probably a few months off completion though. 

Bit of an update to this one.

Having some issues on the dyno that held us back (boost spiking) and I want to pass some info over you guys and see what you think is wrong with my setup. The current readout on this dyno is 462rwkw on a low reading dyno so keep in mind it is a real world 500rwkw setup on a hub dyno. Don't read into the power figure too much as a sign of the issue.

The short and curly of it is:

2.8 Litre Racepace build
RB25 NEO N/A Head with VCT (internally standard however )
Borgwarner EFR 8474 
Turbosmart 50mm Straight Gate + Mac valve
6Boost Manifold
4" dump to full 4" exhaust (nil restrictions) Wastegate plumbed back in and all angles in the exhaust system are acceptable and not too sharp.
GFB SV52 BOV in cooler piping 
Turbosmart BOV in EFR Housing

 

The issue we are having is it comes onto full boost for example at 4000rpm and spikes to 24/25psi, before dropping down to 17psi before slowly rising back up to the target boost of 23psi. It was extremely uncontrollable and the tuner actually had to ramp in boost progrssively with each 1000rpm on each boost setting we selected to try and reduce the amount of spiking. Sometimes we would see a drop of 10psi from the peak at the beginning of the run, to the low, until it took the next 500-1000rpm to stabilise back up to the target boost.

The tuner is pretty confident that the straight gate is just a poorly designed product and leaks too much boost upon cracking the gate open and theres no way to fix it other than going to a poppet valve. He's also confient theres no ignition breakdown or floating valves. The fueling is extremely stable as well. Turbo speed is somewhere around the 109,000rpm area.

The spanner in the works for me is that prior to this Borgwarner and StraightGate, the car was tuned on -5 twins at a diferent tuner, and he also had issues controlling the boost with it spiking around the same rpm range, so to me this sounds like the same issue and it can't be anything on the turbo side as this was all changed and I think the behaviour is extremely similar, if not the same.

We also removed the mac valve and did a run on wastegate pressure and it still spiked and had the same behaviour.

My thoughts on possibilities are:
Boost Leak
VCT Cam Gear isn't reliably activating consistently - (On this however, we did a run with the VCT disabled and the boost still spiked)
Turbosmart BOV is not handling the boost? However this seems unlikely to not be able to handle 20psi.

I have a couple of logs that I can't make sense of if anybody knows how to read them and can obtain further logs of other parameters if they are not enough, happy to pay for anyones time.

The dyno readout with the power figure is the most recent last week. The other picture is from two weeks prior to that where we couldn't break 400kw (we removed the cat), however the issue of the boost control persisted.

@Lithium @Piggaz @burn4005 @GTSBoy @discopotato03
I've tagged those that were quite active in recent pages here, no disrespect to those that know turbos well but I missed tagging.

Cheers 

[GTSBoy]

Photo of manifold showing gate location? I mean, it's 6Boost, so we probably shouldn't be worrying, but always wroth knowing what the layout is.

Plumbed back to atmosphere? Or into the dump?

[Piggaz]

A few things that seem a bit off here.

- why is there 2 BOV’s? 
- the turbo smart BOV on the compressor housing, is it turned up ALL the way? I have seen this become an issue on old man Pete’s car. It would push open and recirc, turbo speed would rise and the boost pressure would do weird things.

- stock head? Does that include springs?

- tried a different MAC valve? Is it plumbed correctly?

[Lithium]

What transmission are you running?  It's a bit tricky with the scaling, but at face value the power "curve" looks more like a "line" which is a bit odd... basically a lot more like a dyno plot I'd expect with a highish (compared to a factory auto) stall torque converter type setup.

If this is running an auto then this kind of boost control challenge is definitely a thing, the rpm scale on the dyno doesn't reflect what the engine is actually doing (unless the dyno has access to the engine's ACTUAL speed electronically) and what you'll get is a big rpm flare up as the engine torque launches past the converter pump's ability to resist torque at that rpm, then as the converter starts picking up rpm it will kinda even out again and the engine rpm will pick up more steadily.

The trick with this "flare up" is if it's kinda near the boost threshold for the turbo then the engine's airflow requirements to maintain the previous boost level will outrun the turbo's ability to supply that boost - so you end up with a natural flattening off, if not dip when that happens.   If you are running closed loop, or even tune the "feed forward" wastegate duty cycle to deal with that rpm spike then when the engine starts settling to a more typical climb you'll actually have a situation where the gate is "too closed" and boost will run away for a bit, then have to pull down again.     

It's not trivial to get this perfect as most boost control systems are generally expecting more predictable engine rpm rates of change, but if you *know* that's whats going on then you can at least "accept your fate" and realise getting that area perfect is kinda chasing your tail a bit, and assume that if the rest is working sensibly and the spike/dip isn't completely uncontrolled then you should be good.

Sorry if I've gone off on a tangent, but the dyno plot and boost control behaviour look a LOT like what I've seen tuning autos in the past.

What ECU are you running? Could possibly be convinced into looking at logs if I get too bored this weekend haha.

Edited May 24 by Lithium

[Borci88]

3 hours ago, GTSBoy said:

Photo of manifold showing gate location? I mean, it's 6Boost, so we probably shouldn't be worrying, but always wroth knowing what the layout is.

Plumbed back to atmosphere? Or into the dump?

Thanks for the quick replies guys its appreciated. A small extension was welded onto the standard 6boost external gate pipe which you can see where the pipe goes from black to stainless just below and to the right of the rear housing in the first picture. Overall I would say the flow is pretty good other than 6boosts choice to come straight off the collector at a decent angle..

2 hours ago, Piggaz said:

A few things that seem a bit off here.

- why is there 2 BOV’s? 
- the turbo smart BOV on the compressor housing, is it turned up ALL the way? I have seen this become an issue on old man Pete’s car. It would push open and recirc, turbo speed would rise and the boost pressure would do weird things.

- stock head? Does that include springs?

- tried a different MAC valve? Is it plumbed correctly?

Not sure why I went with two valves, I originally replaced the stock twin bovs with the GFB when I had the twins on. When I purchased the EFR it came with the Turbosmart Kompact BOV so I figured that would be a better option than the stock EFR Bov. I don't believe the Turbosmart BOV is adjustable? When I get the spike and then sudden dip in boost pressure, the turbo speed does drop as well.

Stock head size wise however I believe it has Neo Turbo springs and a Neo Turbo intake camshaft and an aftermarket exhaust camshaft in the vicinity of 260 degrees.

We didn't try a different MAC valve, we tried two different ways of plumbing it and we also tried removing the mac valve entirely and just having the boost source from the turbo directly connected to the wastegate and it still spiked / dropped and exhibited the same behaviour.

45 minutes ago, Lithium said:

What transmission are you running?  It's a bit tricky with the scaling, but at face value the power "curve" looks more like a "line" which is a bit odd... basically a lot more like a dyno plot I'd expect with a highish (compared to a factory auto) stall torque converter type setup.

If this is running an auto then this kind of boost control challenge is definitely a thing, the rpm scale on the dyno doesn't reflect what the engine is actually doing (unless the dyno has access to the engine's ACTUAL speed electronically) and what you'll get is a big rpm flare up as the engine torque launches past the converter pump's ability to resist torque at that rpm, then as the converter starts picking up rpm it will kinda even out again and the engine rpm will pick up more steadily.

The trick with this "flare up" is if it's kinda near the boost threshold for the turbo then the engine's airflow requirements to maintain the previous boost level will outrun the turbo's ability to supply that boost - so you end up with a natural flattening off, if not dip when that happens.   If you are running closed loop, or even tune the "feed forward" wastegate duty cycle to deal with that rpm spike then when the engine starts settling to a more typical climb you'll actually have a situation where the gate is "too closed" and boost will run away for a bit, then have to pull down again.     

It's not trivial to get this perfect as most boost control systems are generally expecting more predictable engine rpm rates of change, but if you *know* that's whats going on then you can at least "accept your fate" and realise getting that area perfect is kinda chasing your tail a bit, and assume that if the rest is working sensibly and the spike/dip isn't completely uncontrolled then you should be good.

Sorry if I've gone off on a tangent, but the dyno plot and boost control behaviour look a LOT like what I've seen tuning autos in the past.

What ECU are you running? Could possibly be convinced into looking at logs if I get too bored this weekend haha.

Standard R33 GTR 5 speed tansmission. I'm running a Haltech Elite 2500 and can provide some logs if you. I understand what you're saying in that it looks like an auto plot however no, it's still a manual and it just has a lot of torque down low, for all intents and purposes it's a very impressive street car.

I've attached a photo of the quickbitz dyno plot which was when the only difference is I was running -5 twin turbos with a mac valve. As you can see theres a decent dip in AFRs between 125kmh and 135kmh. Our problem now is not that the AFRs are dropping, just the boost pressure is dropping, however it is evident in the same RPM range of the map, coincidentally or not.

[Piggaz]

Big question… have you decided on a gearbox? Because it’s going to come… and a gearset ain’t going to do the job. 
 

save your pennies 🤪

[Borci88]

3 minutes ago, Piggaz said:

Big question… have you decided on a gearbox? Because it’s going to come… and a gearset ain’t going to do the job. 
 

save your pennies 🤪

I've got a spare standard box sitting in the garage 

I know it will come one day though, but it's driven pretty tamely on the four times a year it gets out... 

[Borci88]

 

Heres a dyno log comparing

MAP
TPS
RPM
TURBO SPEED

Each hump is 2nd, then 3rd then 4th gear. This particular log is using gate pressure with the mac valve completely removed.
As you can see it spikes to 23psi, and then drops to 20psi over the course of approx 700rpm, this is the minimum pressure the system will run using a 12psi spring. I think my next set of logs will be changing the spring to the 6psi spring and going out and doing another set of logs again and see if the spike is as dramatic.

As you can imagine if it spikes 3psi on wastegate pressure, that increases to 5-6psi of difference in the higher boost levels approaching the 28psi target with the mac valve.

I even have a few logs here where with the mac valve targetting 28psi, the haltech is seeing 38.6psi.... 

[acsplit]

Can you vent the straight gate to atmosphere easily for testing?

 

[Borci88]

Just now, acsplit said:

Can you vent the straight gate to atmosphere easily for testing?

 

Not without making up a new screamer pipe from scatch I don't think based upon the angles of the current return.

Backpressure shouldn't be an issue though as the straight gate isn't affected by backpressure in the same way that a poppet valve is, and on top of that the exhast system is 4" all the way through so there should be minimal pressure if any to begin with.

[Piggaz]

-5’s did it.

8474 did it.

Not hot side related.

ECU related? 🤨🤪

[GTSBoy]

I'm thinking it is skinny NA Neo port sizing and cams. Somehow.

[Borci88]

18 minutes ago, Piggaz said:

-5’s did it.

8474 did it.

Not hot side related.

ECU related? 🤨🤪

It could be in the tune but in both cases it couldn't be completely solved or eliminated by some of the best tuners in victoria so I'm DOUBTFUL that it's in the tune or ECU but stranger things have happened of course.

3 minutes ago, GTSBoy said:

I'm thinking it is skinny NA Neo port sizing and cams. Somehow.

I've pondered that theory too but is there any theories that thinner and smaller ports overall would cause a turbo to spike? I would've thought it would just restrict it in general if it was small enough to make a difference. I wouldn't expect it to make a different at what is essentially sub 300kw at 4000rpm anyway.


I think the next step is going to have to be the 6psi spring to rule out the idea that the gate is cracking open far too wide initially. At least that is the cheapest (free) thing to check initially.

[GTSBoy]

14 minutes ago, Borci88 said:

I've pondered that theory too but is there any theories that thinner and smaller ports overall would cause a turbo to spike? I would've thought it would just restrict it in general if it was small enough to make a difference. I wouldn't expect it to make a different at what is essentially sub 300kw at 4000rpm anyway.

My theory would revolve around the ~4000rpm point being peak torque territory on an RB, where you're supposed to be looking at max efficiency of the head, etc, and....just not having it because the skinny ports are doing something to f**k it up.

I'll admit, it's a loose theory.