Dual Port Internal WG Actuator for Bewstzzz

[culberro]

Does anyone have a favorite? A quick google search brings up Turbonetics, Turbosmart, and Ebay.

The GTX2860 turbski with ATP 0.63 housing and ultimate wastegate is blowing the flapper open and limiting boost to about 20psi, I'd like to get up to 30psi if possible. It's also killing the spool time.

I preloaded the garret actuator that's on there (15psi), and that worked for a bit. With a MBC and I was able to get to ~ 28psi... but I think the spring is getting hot and softening up as it won't get over 20psi now. Harlard also had this exact same issue when he was running the turbo



Thanks!

[RvolvoR]

you need to either just try an actuator with a higher baser spring rate or go to a larger hot-side as you might just have enough exhaust backpressure to blow the gate open or you can only make 28 psi with a .63 hotside on that turbo as it is probaly choking down exhaust flow at that point. I would guess that you probably have 50 or more psi of exhaust pressure.

[VB242]

I bought the kinugawa universal and it has a spring that will get you 24-25 psi, maybe that plus the mbc will get you to 30, or maybe inquire if they have a 30psi spring?

[t8fanning]

I'd get a bigger exhaust housing like Asher said.

[culberro]

I'm very hesitant to get a larger hot side, as this is on the rally car and reducing the boost threshold is a goal. There are a few 2.5L subarus that are using a 0.52 hotside with this turbo to get it spinning quicker. Granted they run a restrictor (32mm?) to limit peak power, but the torque is still there.

And yes, I'm sure the drive pressure is absolutely astronomically high with this turbo.

[Harlard]

I'd try running an equivalent sized T25 Garrett turbine housing and go from there. The ATP one in there isn't doing that turbo any favors, sadly.

[RvolvoR]

Originally Posted by culberro

I'm very hesitant to get a larger hot side, as this is on the rally car and reducing the boost threshold is a goal. There are a few 2.5L subarus that are using a 0.52 hotside with this turbo to get it spinning quicker. Granted they run a restrictor (32mm?) to limit peak power, but the torque is still there.

And yes, I'm sure the drive pressure is absolutely astronomically high with this turbo.

Increase rev limiter if the boost threshold is slightly higher. Problem solved.

Running super high drive pressures is just going to beat parts up faster and in a way, make it less reliable.

Also, there is a possibility of having equal boost threshold with a slightly larger hotside due to more flow...in theory.

[culberro]

Originally Posted by Harlard

I'd try running an equivalent sized T25 Garrett turbine housing and go from there. The ATP one in there isn't doing that turbo any favors, sadly.

I'd consider going to a v-band t25 housing and running and external gate. Packaging might be a bit tricky with a t3 -> v-band adapter on there.

[Duder]

Good advice here. The T25 style OE Garrett turbine housings are a better match aerodynamically to the GT28 turbine wheel. The ATP housing, while convenient on a Redblock, isn't optimal. I did some testing many years back that helped illustrate this but not sure where that data went.

Are you running a 90+ manifold? What kind of power are you making at 28psi?

When I do turbo matches from scratch I always look at engine deltaP, which is intake manifold pressure minus exhaust manifold pressure. Lots of purpose built racing engines are able to keep this positive, i.e. exhaust manifold backpressure is always lower than boost as measured in the intake manifold. With a small turbine housing, high boost, and big power you will more than likely drive yourself into a very very negative deltaP - which effectively chokes up your head. It lowers engine VE and increases pumping losses.

And as Asher is saying, because a larger turbine A/R will reduce exhaust backpressure for the same flow, your wastegate valve will have less opening force on it due to exhaust backpressure and the actuator spring won't need to be preloaded as much to hold it closed.

[Harlard]

Originally Posted by culberro

I'd consider going to a v-band t25 housing and running and external gate. Packaging might be a bit tricky with a t3 -> v-band adapter on there.

The outlet on a stock 90+ is tiny. I'm sure a man of your talents can weld a v band ring on a glowing hot manifold!

Plus it's wet and cold out now so it will make your shop an order of magnitude more comfortable while you do it.

Or you could mill another manifold flat, fill the holes, and drill/tap in the T25 pattern...

Originally Posted by Duder

Good advice here. The T25 style OE Garrett turbine housings are a better match aerodynamically to the GT28 turbine wheel. The ATP housing, while convenient on a Redblock, isn't optimal. I did some testing many years back that helped illustrate this but not sure where that data went.

Are you running a 90+ manifold? What kind of power are you making at 28psi?

When I do turbo matches from scratch I always look at engine deltaP, which is intake manifold pressure minus exhaust manifold pressure. Lots of purpose built racing engines are able to keep this positive, i.e. exhaust manifold backpressure is always lower than boost as measured in the intake manifold. With a small turbine housing, high boost, and big power you will more than likely drive yourself into a very very negative deltaP - which effectively chokes up your head. It lowers engine VE and increases pumping losses.

And as Asher is saying, because a larger turbine A/R will reduce exhaust backpressure for the same flow, your wastegate valve will have less opening force on it due to exhaust backpressure and the actuator spring won't need to be preloaded as much to hold it closed.

I should have listened to you when I had the chance

[culberro]

Originally Posted by Duder

Good advice here. The T25 style OE Garrett turbine housings are a better match aerodynamically to the GT28 turbine wheel. The ATP housing, while convenient on a Redblock, isn't optimal. I did some testing many years back that helped illustrate this but not sure where that data went.

Are you running a 90+ manifold? What kind of power are you making at 28psi?

When I do turbo matches from scratch I always look at engine deltaP, which is intake manifold pressure minus exhaust manifold pressure. Lots of purpose built racing engines are able to keep this positive, i.e. exhaust manifold backpressure is always lower than boost as measured in the intake manifold. With a small turbine housing, high boost, and big power you will more than likely drive yourself into a very very negative deltaP - which effectively chokes up your head. It lowers engine VE and increases pumping losses.

And as Asher is saying, because a larger turbine A/R will reduce exhaust backpressure for the same flow, your wastegate valve will have less opening force on it due to exhaust backpressure and the actuator spring won't need to be preloaded as much to hold it closed.

I'd love to see the testing/data on the Garret housing vs ATP housing if you can find it.

The engine is currently a mostly stock b230ft with IPD turbo cam and a slightly shaved head, and a 90+ that's been machined to a T3 flange. No idea on the power, and the car usually sits around 18psi due to the WG issues... I think that with that boost pressure it's in the ~250 crank HP range. Getting closer to 300hp would be great.

I'm all for going to a larger turbo to decrease drive pressure, but the boost threshold is key here. On the rally car, the engine is going from 2800-7.5k rpm every few seconds, for 5-15min at time.

[culberro]

Originally Posted by Harlard

I'm sure a man of your talents can weld a v band ring on a glowing hot manifold!
Plus it's wet and cold out now so it will make your shop an order of magnitude more comfortable while you do it.

This is very true!

[Duder]

Originally Posted by culberro

I'd love to see the testing/data on the Garret housing vs ATP housing if you can find it.

The engine is currently a mostly stock b230ft with IPD turbo cam and a slightly shaved head, and a 90+ that's been machined to a T3 flange. No idea on the power, and the car usually sits around 18psi due to the WG issues... I think that with that boost pressure it's in the ~250 crank HP range. Getting closer to 300hp would be great.

I'm all for going to a larger turbo to decrease drive pressure, but the boost threshold is key here. On the rally car, the engine is going from 2800-7.5k rpm every few seconds, for 5-15min at time.

"Drive pressure" is a misnomer from the diesel truck world - it's not the pressure that drives the turbine, it's momentum transfer from the exhaust gas molecules physically hitting the turbine blades. The whole system will benefit from lowering pre-turbine exhaust pressure (aka backpressure or manifold pressure).

Let's assume you keep the rest of the turbo unchanged, but simply increase turbine housing A/R. The compressor needs the same amount of shaft power to run at your target boost level; that doesn't change. But the larger A/R housing allows the turbine to make that same amount of shaft power at a lower pressure ratio, and higher flow rate. It's a tradeoff between turbine flow and pressure.

It's possible to achieve the same or better boost threshold with a larger A/R, depending on how terminally undersized your current housing is now.

[2manyturbos]

Tyler has a Volvo 940 SE T-25 core.

[sbabbs]

Fun fact 940SE volvo t25 turbos, .48 a/r exhaust housing. 1990 or 1991 740 turbo with stock m46 has .64 a/r t25 volvo exhaust housing. Both have volvo flanged t3 flange with the step.

I put the .64 one on my 95 940.

[culberro]

Originally Posted by Duder

"Drive pressure" is a misnomer from the diesel truck world - it's not the pressure that drives the turbine, it's momentum transfer from the exhaust gas molecules physically hitting the turbine blades. The whole system will benefit from lowering pre-turbine exhaust pressure (aka backpressure or manifold pressure).

Let's assume you keep the rest of the turbo unchanged, but simply increase turbine housing A/R. The compressor needs the same amount of shaft power to run at your target boost level; that doesn't change. But the larger A/R housing allows the turbine to make that same amount of shaft power at a lower pressure ratio, and higher flow rate. It's a tradeoff between turbine flow and pressure.

It's possible to achieve the same or better boost threshold with a larger A/R, depending on how terminally undersized your current housing is now.

I'll use manifold pressure from now on

So this brings up another question I have:
How does one go about comparing turbine housings of different A/R ratios ,if they are not in the frame size. In this instance the housing I have is a T3 0.63, what would that be in a T25 size?
Can you calculate inlet area and use that as a comparison? If the T3 is 10% larger in inlet area, would a 10% larger T25 A/R housing have the same flow?

[2manyturbos]

Originally Posted by sbabbs

Fun fact 940SE volvo t25 turbos, .48 a/r exhaust housing. 1990 or 1991 740 turbo with stock m46 has .64 a/r t25 volvo exhaust housing. Both have volvo flanged t3 flange with the step.

I put the .64 one on my 95 940.

Good to know. The T25 that was on Tyler's 940 SE came off a 1990 740 Turbo with M46. The entire engine came out of that 740 along with the T25.

[Duder]

Originally Posted by sbabbs

Fun fact 940SE volvo t25 turbos, .48 a/r exhaust housing. 1990 or 1991 740 turbo with stock m46 has .64 a/r t25 volvo exhaust housing. Both have volvo flanged t3 flange with the step.

That turbine housing won't fit on a GT28R or GTX28R though. Different turbine wheel and shroud contour machining.

Originally Posted by culberro

I'll use manifold pressure from now on

So this brings up another question I have:
How does one go about comparing turbine housings of different A/R ratios ,if they are not in the frame size. In this instance the housing I have is a T3 0.63, what would that be in a T25 size?
Can you calculate inlet area and use that as a comparison? If the T3 is 10% larger in inlet area, would a 10% larger T25 A/R housing have the same flow?

You can use the choke flow curves that are published for each wheel/housing combination, from Garrett anyway...I think others publish those too. The swallowing curve or choke curve represents max flow for the housing vs. pressure ratio.

Turbine flow is corrected back to standard conditions before plotting, which makes the flow values look small compared to the compressor. Don't worry about that; compare turbine flow to turbine flow. The full turbine map includes efficiency data as well but that doesn't get published, as it's a bit more sensitive and fairly useless to the average consumer unless you are running a full 1D engine/turbo matching simulation.

Anyway - in your example, step 1 would be to find a map for the 0.63 A/R T3 in question. I should point out that the turbine flow depends on the entire stage - housing and wheel together. Depends on turbine trim, wheel aero design, and a few other factors too. For the sake of argument let's say your 0.63 A/R T3 stage has a choke flow of 20 lb/min corrected. Step 2 is to check the GT28R turbine maps and find the one that flows closest to that value. In this case it would be the 0.86 A/R on a GT28R or GTX28R (which share the same turbine wheel). That 0.86 A/R housing only comes in the bolted variety, with a T25 inlet flange and 5-bolt Nissan pattern outlet. ATP and others do modify them but that's how they are made initially, anyway.

0.72 A/R is the largest v-band in/out housing for the GT28R turbine wheel and it flows about 17.5 lb/min corrected. So at a very high pressure ratio where the curve goes horizontal, you'd be down about 13% less flow capacity vs. your T3 with 0.63 A/R... assuming that 20 lb/min example is accurate. What is commonly called a "T3" turbine is actually several different wheels, so you'd have to check on that based on your exact hardware.

[Duder]

I used this GT/GTX30R turbine flow chart as a reference for your T3:




And here's the GT/GTX28R:

[culberro]

Originally Posted by Duder

*Knowledge bomb dropped*

Thanks for that. I was just comparing those exact graphs when I checked back to see you post that up!

Comparing exhaust flow calculations for the ATP turbine housing is all a guess at this point, as they don't publish any data (maybe don't even have it...)
If we make the assumption that the ATP housing for the GTX28R has the same flow characteristics as the GT30R then it appears that the 0.86 T25 housing would be the best match.
If the ATP housing doesn't flow as well, then the 0.86 T25 housing would be better option.

Well, looks like I have some work to do.

Thank you for that!

[Duder]

No worries. Radial turbines are fascinating and frustrating. They don't behave in an easy to grasp intuitive way...and to get that intuitive grasp you have to play around with the full maps (including efficiency) and real test data for the engine you're working with. Even then it's still easy to make mistakes with turbine wheel and housing selection.

It's rare to find any published turbo test data, turbine or compressor, from any non-OEM turbo company. I am not trying to knock ATP at all; they do great work with the resources they have. It takes mega millions to set up a lab that can easily test turbos on gas stands reliably and generate maps from data. That's why you rarely see any maps at all from the smaller companies that aren't Tier I automotive suppliers.

It's kind of a moot point anyway since you'd need a full engine dyno cell to get all the data you need from your engine to do a proper turbo match even with the exact maps. That's why turbine housing and A/R selection is more of an empirical thing. Make some educated guesses, choose a housing, test it out on your engine, then evaluate and iterate if you're not happy...

[Duder]

It may also be academic, but your engine will flow more than the choke flow curve from the turbine map, even after un-correcting back to physical flow. Any wastegated exhaust flow adds to the turbine flow. So at any operating condition where the wastegate is open, the turbine should be right at choke, and the total exhaust flow will actually be up above that curve.

[culberro]

We've been pretty happy with the turbo, except for the WG issue. How is the flow through the stock Garret t25 internal WGs, what is the diameter on the WG opening?

I guess the other option is to use an external WG somewhere in the system if there's boost creep issues.

[Duder]

Originally Posted by culberro

We've been pretty happy with the turbo, except for the WG issue. How is the flow through the stock Garret t25 internal WGs, what is the diameter on the WG opening?

I guess the other option is to use an external WG somewhere in the system if there's boost creep issues.

The as-cast port diameter is just under 25mm standard, and it's at 90 degrees to the flow. This housing dates back to the SR20DET and RB26DETT OEM applications so likely undersized for what you'll be doing with it. You could port it out as is commonly done with MHI turbos.

The v-band 0.57 and 0.72 A/R housings are nice and modern, non wastegated, but flow with the 0.72 is lower than the older 0.86 A/R wastegated housing as you've seen. If you need the flow of the 0.86 you could always remove the swing valve, weld in a plug, and run a larger external gate.

[Harlard]

This would be neat, so long as ATP makes them dimensionally similar to the original.

https://www.atpturbo.com/mm5/merchan...y_Code=GTHGT28