Recirculated Wastegate vs. Open Dumptube Back-to-back Dyno

[MorrisonFab]

Not a DSM, but very relevant for anyone wondering about the power differences between the two when boost is cranked and the turbo is near maxed out!

Back-to-back dyno: Recirculated wastegate vs open dumptube, highest boost setting
The goal was to see if a well sorted recirculated setup will make the same all-in power across the board, highest boost being a key factor.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

Tylor’s 2.0L Evo 9 with a HPT 6466, 1.07A/R v-band, MF v-band manifold with either recirculated O2 housing or O2 housing with dumptube, 44mm MVR, 3” MF downpipe, 3” to 3.5” ETS catback exhaust. E85, tested and tuned on Haltech by Roni at SpeedLab.Inc
All pulls were at the highest boost setting in the 850whp range (~46psi).

You must be logged in to view this image or video.

3 first runs/traces with the recirculated O2 housing/bend off the turbo, and 2 last traces with the O2 housing/bend using a separate dumptube.
Regardless of the setup, each trace tracks with the rest; some a little more, some a little less, but mostly run-to-run variance from slight boost differences or time sitting between pulls (which affects how much energy goes into heating the turbine housing back up vs driving the turbine etc). The dumptube setup showed a little quicker spool in one pull, and the recirc had the highest power pull, but likely mostly due to the factors above in this case since recirc isn't going to increase power, only be quieter under full boost.

Boost traces:

You must be logged in to view this image or video.

Ultimately, when the setup was being pushed, there was no discernable difference in the dyno graphs. When you want to max the turbo out, the power was there. More boost= less wastegate opening= less potential difference.

This was only the initial test though, added in just after dialing in the new setup on the dyno, with more back-to-back testing planned at lower boost levels where differences between the two configurations are more likely, due to the larger proportion of exhaust gas being bypassed through the wastegate. When the wastegate opens near target boost on a recirculated setup, all of the bypassed exhaust gets re-entered into the main exhaust whereas the open dumptube takes a little load off the main exhaust, essentially acting like a larger exhaust overall. These tests at the highest boost setting have the highest total exhaust flow, but the smallest proportion of wastegated exhaust flow, keeping the differences between where the bypassed exhaust goes negligible. The lower amount of exhaust being bypassed/wastegated around the turbo is exactly what increases/maintains boost to begin with (wastegate nearly closed).

The setup made 900whp with some more boost, so near maxed out in this comparison with boost still falling at high rpm. Some of the main factors are exhaust size (post turbo and if undersized for the potential power/flow), engine vs turbo size/efficiency which greatly affects the amount of exhaust gas that needs to be bypassed at a given power/boost level, and how the re-entry is handled into the main exhaust (the more potential losses if not done well when recirculated, mostly at lower boost).

A huge thank you to Roni and SpeedLab.Inc for doing the dyno tuning/testing and Tylor for coming to us to being with and offering his setup for testing!

So damn cool to see direct overlays and more to come!

 

[Stapl3]

As you said, the wastegate is barely used when running at max boost. I don't understand the point of this thread.

Do you have a test where boost was low and the wastegate was actually doing something?

 

[MorrisonFab]

As you said, the wastegate is barely used when running at max boost. I don't understand the point of this thread.

Do you have a test where boost was low and the wastegate was actually doing something?

It wasn't a given if even a relatively small amount of exhaust being re-introduced on top of 850whp+ through a 3" exhaust would have a detrimental effect or not, and we weren't 100% confident to say the difference would reach/approach zero at high boost levels either.

I think being unable to tell which trace belongs to which setup would be a surprise to most, including the person doing the testing. Gave a perfect opportunity to kick things off and talk about what to expect in future planned testing, which we mentioned.

It's one thing to say it, another to prove it. We really like this side of things, as well as sharing the results along the way.

 

[Canadian_CD9A]

As you said, the wastegate is barely used when running at max boost. I don't understand the point of this thread.

Do you have a test where boost was low and the wastegate was actually doing something?

The air is going somewhere - either through the turbine or the wastegate, and they're measuring power difference caused by effective backpressure at the merge. If the dyno doesn't show a difference at max effort, why would it show a difference when there is less overall exhaust volume to move at that lower boost level? Neither port is insufficient to do its job and the exhaust can flow it.

 

[CanadianTalonTC]

The air is going somewhere - either through the turbine or the wastegate, and they're measuring power difference caused by effective backpressure at the merge. If the dyno doesn't show a difference at max effort, why would it show a difference when there is less overall exhaust volume to move at that lower boost level? Neither port is insufficient to do its job and the exhaust can flow it.

The only other way I can view this is if the boost level is lower then the wastegate will be flowing more vs at higher boost causing turbulence (maybe? thats why they are doing this I guess) at the collector. I'd be curious to see differences at different boost levels.

Cool to see the experimentation. I personally run a recirculated housing due to noise and such and had no idea there was even an argument to be made that one was possibly more efficient.

 

[MorrisonFab]

As far as we understand it, the additional backpressure at full boost will stem at the turbine outlet, and be caused by the additional flow demands downstream (recirculated all going through the full exhaust vs some relieved beforehand by the dumptube), and made worse if the recirc re-entry/angle causes additional turbulence in the main exhaust flow (especially if close to the turbine outlet and entering sideways, etc). Additional backpressure at the turbine outlet will require more drive pressure at the turbine inlet to maintain the same pressure ratio across the turbine itself, which is then seen by the engine as more total backpressure/drive pressure and hurt VE and power per psi. The potential effect would be similar to an exhaust system that is too small, and open dumptube configuration is an extra exhaust path that also has no chance of disturbing the main flow.

These runs with the boost cranked up has very little exhaust flow through the wastegate to make much difference where it ends up going, and testing at low boost levels will mean a much more significant portion of the exhaust flow is coming from the recirculated wastegate (now open more to bypass exhaust and keep boost down) and a much higher chance to disturb the main exhaust flow when re-entering (and cause the above effects).

If you have to split hairs to find differences even at lower boost levels that will be great, since it means you can have a quieter setup at full boost without a compromise elsewhere, which is a common concern. Lots of instances where other folks have said switching to open dump improved power and spool, which we don't doubt it can, but wanted to dive into why/when/how much. Unaware of any other back-to-back testing to truly see

The original test was planned with different boost levels, but the shop doing it wasn't able to fit it in this time

 

[CanadianTalonTC]

As far as we understand it, the additional backpressure at full boost will stem at the turbine outlet, and be caused by the additional flow demands downstream (recirculated all going through the full exhaust vs some relieved beforehand by the dumptube), and made worse if the recirc re-entry/angle causes additional turbulence in the main exhaust flow (especially if close to the turbine outlet and entering sideways, etc). Additional backpressure at the turbine outlet will require more drive pressure at the turbine inlet to maintain the same pressure ratio across the turbine itself, which is then seen by the engine as more total backpressure/drive pressure and hurt VE and power per psi. The potential effect would be similar to an exhaust system that is too small, and open dumptube configuration is an extra exhaust path that also has no chance of disturbing the main flow.

These runs with the boost cranked up has very little exhaust flow through the wastegate to make much difference where it ends up going, and testing at low boost levels will mean a much more significant portion of the exhaust flow is coming from the recirculated wastegate (now open more to bypass exhaust and keep boost down) and a much higher chance to disturb the main exhaust flow when re-entering (and cause the above effects).

If you have to split hairs to find differences even at lower boost levels that will be great, since it means you can have a quieter setup at full boost without a compromise elsewhere, which is a common concern. Lots of instances where other folks have said switching to open dump improved power and spool, which we don't doubt it can, but wanted to dive into why/when/how much. Unaware of any other back-to-back testing to truly see

The original test was planned with different boost levels, but the shop doing it wasn't able to fit it in this time

This is exactly what I meant. The prior guy's post talking about how it makes no difference is since deleted for some reason but as far as my brain can comprehend, exhaust gasses coming in at an angle to others has to cause some sort of change within the flow and as you stated, the more is bypassed, the more turbulent it should be in theory.

Sorry if I've missed it, but do you plan to retest to see effects then?

 

[DreamChaser7]

This is one area I am personally still on the fence about which direction to go, and far less educated on the matter than most here. I appreciated the testing and exploration on the topic myself. Thanks you two, keep up the excellent work and support for this community!

 

[MorrisonFab]

This is exactly what I meant. The prior guy's post talking about how it makes no difference is since deleted for some reason but as far as my brain can comprehend, exhaust gasses coming in at an angle to others has to cause some sort of change within the flow and as you stated, the more is bypassed, the more turbulent it should be in theory.

Sorry if I've missed it, but do you plan to retest to see effects then?

Correct, and will get more interesting

They will be going back in the near future for lower boost runs, yes. Lots lined up actually, unfortunately we don't have any control over exactly when it can be done but extremely grateful when it does work out.

This is one area I am personally still on the fence about which direction to go, and far less educated on the matter than most here. I appreciated the testing and exploration on the topic myself. Thanks you two, keep up the excellent work and support for this community!

Thank you! We still believe it will mostly come down to personal preference rather than performance

 

[Canadian_CD9A]

Agreed that added wastegate flow entering the exhaust at an angle will disrupt the overall flow of the exhaust, but wasting exhaust pressure (and thereby slowing down the compressor and lowering boost) reduces the overall volume of air being shoved through the exhaust manifold, which should reduce the overall effect of the wastegate flow disruption. I'm very curious to see what that balance is throughout the boost range. I've seen tests on SRT-4s showing significant gains by dumping (38HP in one case), but I can't imagine that those recirculation systems were optimal. I'm also trying to find a very oldschool Australian Skyline R34 build thread in which an F1 powertrain engineer built an exhaust that swirled the wastegate flow into the exhaust by fabricating a helix over a ~10" section of the downpipe to minimize the losses of recirculation, but I don't think the results were ever shared. I can't believe it's 2024 and so little testing has been done on the subject.

 

[bastarddsm]

I'm fairly certain I've seen the same results back in the day with my HY35. When I went to a dumped WG, it didn't pick up anything that I could be certain of.

It's really not magic, and honestly quite trivial that it doesn't effect power if you know about this stuff. For the most part the engines airflow goes proportional to the exhaust pressure ^-.1, or something like that, maybe a little more sensitive, but close enough. . The compressor work and turbine shaft power is roughly proportional to mass flow * PR^1.4. And the pressure through the exhaust pipe is some sqrt function of the mass flow. That all gets stacked up you end up with the engines mass flow is proportional to mass flow through the exhaust ^-0.01 So basically if you double the mass flow through the exhaust pipe (recirc the WG) you still get 99% of the mass flow. Funny enough the increase in pumping loss during the exhaust stroke likely far far exceeds the power loss from the loss in mass flow.

Another point, high exhaust/boost pressure ratios favor high compression ratio, in terms of flow.

Of course there are outliers. Tractor on a 76mm inducer that makes about 1500hp (that would be 2000+ on methanol) They run all in, no WG. Usually a 5" elbow pointing up off the TH. On the dyno one was way off like -250hp - had a fancy pie cut ti elbow on it. Swapped it out for a mendrel bend, and all the power came back. Wild thing is it had no change in boost/exhaust/ts/tuneup. Put the piecuts back on, and it backed right off. Figure that one out.

 

[MorrisonFab]

I'm fairly certain I've seen the same results back in the day with my HY35. When I went to a dumped WG, it didn't pick up anything that I could be certain of.

Saw the same on the HX35, exact same trap speed

It's really not magic, and honestly quite trivial that it doesn't effect power if you know about this stuff. For the most part the engines airflow goes proportional to the exhaust pressure ^-.1, or something like that, maybe a little more sensitive, but close enough. . The compressor work and turbine shaft power is roughly proportional to mass flow * PR^1.4. And the pressure through the exhaust pipe is some sqrt function of the mass flow. That all gets stacked up you end up with the engines mass flow is proportional to mass flow through the exhaust ^-0.01 So basically if you double the mass flow through the exhaust pipe (recirc the WG) you still get 99% of the mass flow. Funny enough the increase in pumping loss during the exhaust stroke likely far far exceeds the power loss from the loss in mass flow.

Not as familiar with those relationships/fomulas, but seem to notice the opposite effect with mass flow changes exceeding pumping loss when it comes to increased drive pressure/backpressure. Increased drive pressure coming with a decrease in VE, that reduces mass flow at the same psi and rpm. And with the turbine working off a pressure ratio, additional backpressure after the turbine will seemingly be amplified by that pressure ratio on the inlet side. Mostly looking at engines with decent amounts of duration and overlap that will tend to become more sensitive to exhaust backpressure though.

Another point, high exhaust/boost pressure ratios favor high compression ratio, in terms of flow.

High compression ratio having less chamber volume near TDC for residual exhaust gas? Brute forcing it out vs relying on a more favorable EMAP:MAP ratio to help scavenge?

Of course there are outliers. Tractor on a 76mm inducer that makes about 1500hp (that would be 2000+ on methanol) They run all in, no WG. Usually a 5" elbow pointing up off the TH. On the dyno one was way off like -250hp - had a fancy pie cut ti elbow on it. Swapped it out for a mendrel bend, and all the power came back. Wild thing is it had no change in boost/exhaust/ts/tuneup. Put the piecuts back on, and it backed right off. Figure that one out.

Still makes no sense

 

[bastarddsm]

The formulas are just some emperical relationships from the data I've seen over the years. It's probably more sensative than I indicated, but the point is that it's far less sensative than 1:1. Meaning halving the "drive" pressure will not double the mass flow - It'll return like 5% or something on that level. The VGT on my car offered a lot of room to experiment and collect data.

I think if you look in the heywood he has some relationships of this sort as well, but his data will probably be based on old timey stuff.

On the turbine stuff, ideally it operates on pressure ratio for a constant flow incompressible application, but we are compressible pulse flow, and there is that pesky efficiency that gets in the way of making it easy. So I don't think it's quite as sensative to back pressure as strictly looking at the pressure ratio - but that's speculation. Real testing of this would be hard and very dependant on the individual setup. What we measure as exhaust pressures really has no firm relation to the PR that the turbine actually see's. Yes, there is some correlation, but we measure it because it's what we can measure, not because it's what we should measure.

Also remember that the mass flow through the turbine is almost always choked (PR>2). Once it hits that point volume flow(velocity) through the nozzle is constant, and increases in mass flow are from increases in density, which is almost linear with pressure.

When we learn about turbines in introduction the thermo we learn that the shaft power is related to enthalpy change across the turbine which is essentially PR^gamma (1.2-1.4ish for exhaust gasses). What we don't learn is that turbines really rely on compressible flow which is a whole nother thing, and simple things like PR end up being about meaningless for trying to get a good calculation. The guys who have a good understanding of this stuff have like CEO salaries at Boeing/GE/NASA/SPACEX/ECT, not us peons f***in with cars.

So we do the best we can, and fit a curve to some data and say, well it works except when it doesn't. And we recognize anyone talking in certanties and saying they can calculate (or simulate)this or that, is probably a moron.

Anyway keep up the good work, thanks for the data! Also I did not mean to marginalize your work by saying it's trvial, more that the theory that it shouldn't matter is fairly trivial. Physical testing is always the best.

See you guys in a month!

 

[DreamChaser7]

Still trying to be a sponge here Kurt, but you’re making my brain hurt.