I have tried searching but I'm not finding exactly what I'm trying to figure out. Let me get it out of the way at the beginning though, I am NOT talking about compound turbos and this is for a gas engine.

I have an idea for a build but haven't found much information about it, which leads me to believe one of two things; either is doesn't work or nobody wants to deal with making it work.

The thought is to set up turbos in what would otherwise work as compound turbos;
hot side: engine>small turbo>large turbo
cold side: large turbo>small turbo>intercooler>etc.

The major difference being to install a wastegate equal in size to the collector feeding the small turbo (i.e. 50mm) to completely bypass the small turbo hot side once the large turbo is spooled. I would leave the factory boost control to operate this turbo as it will be the most used (street car) but also add a Hobbs switch to basically make the solenoid 100% duty cycle once a set boost is achieved by the large turbo. The solenoid is rated to work at 100% duty cycle so no issue there. Then I would have an aftermarket boost controller to control the large turbo.

So here's the question I can't find a good answer to: Can a turbo compressor have compressed air blown through it and be allowed to free spin? In my mind I picture it as a turbo that is operating off boost and just has the engine pulling air through it. I imaging the large wastegate wouldn't be 100% effective and there would be some amount of exhaust still going through the small turbine, but not enough to create additional boost. This would just allow the turbo to spin freely as to not create a major restriction to the air being pushed through it. Looking at compressor maps for the two turbos I have in mind, if you were to extend the shaft speed lines down to a PR of 1.0, the volumetric flow through the small turbo should not exceed a safe shaft speed at the required airflow. I'm not sure if that even matters considering I am not trying to drive the turbo at this point.

Does anyone have any input or experience on this?

 

I didn't rewatch the video, so I think this is the one. I think that's what Grannas is doing.

 

Thanks for sharing! At first I was thinking, "I just said I wasn't talking about compound!" But after watching, it's really close to what I'm thinking. It looks like they have the ability to bypass the small turbo, but I think in another video they said they hit 58psi which means they have to be using it to compound at some point. It seems me idea may not be totally crazy.

 

You cant completely bypass the small turbo hot side .

My buddy did similar to what you are wanting.
s366 and s480

engine to s366, wastegate #1 between engine and s366 routed to turbine of s480, s366 turbine outlet also fed to s480 turbine inlet, wastegate #2 between s366 turbine outlet and s480 turbine inlet.
At approx. 7 psi wg #1 would open to control s366. At lil over 20psi wastegate #2 would vent to atmosphere to control s480.

Compressor side was s480 feeding s366 to air to air ic to engine

This is a stick car and s366 drove like n/a but just couldnt make the power. With the compound kept the driveability but carried out the power. It was so impressive. It just made power everywhere.

 

That sounds awesome. Thanks. What is the reasoning for not being able to completely bypass the first turbo? The pipe between the 2 hot sides would only have the pressure from the large turbo restriction, but the small turbo would have that plus its own restriction. With the right sized wastegate it would seem like most of the flow would bypass the small turbo and take the path of least resistance. Is this not what happened in your friend's case?

 

Correction, #2 wg was on the turbine housing of s366.

 

Not sure why this has been tied up waiting for a moderator for almost a day when everything else has gone through just fine, so I'm trying again.

That sounds awesome. Thanks. What is the reasoning for not being able to completely bypass the first turbo? The pipe between the 2 hot sides would only have the pressure from the large turbo restriction, but the small turbo would have that plus its own restriction. With the right sized wastegate it would seem like most of the flow would bypass the small turbo and take the path of least resistance. Is this not what happened in your friend's case?

 

I guess I didnt read your post thoroughly.

One of the higher end manuf. (bmw?) had a setup like this but had ecu controlled setup. Valves that would bypass the low speed turbo all together.

I have no 1st hand knowledge on what you want to do. I would make the argument that the non spinning low speed turbo could be a flow restriction. And it probably wouldnt just sit idle. The incoming air flowing to it would make it spin just like exhaust flow makes the turbine side spin.

 

where are the turbo guru's around here? LOL

 

I am no guru, I assumed a set up like this is there to widen the power band in a street car. Small turbo for quick spool. Large turbo for top end power. The small turbo would become a restriction at big power so the gates by passes exhaust to the larger one. This is probably wrong, but I think keeping exhaust back pressure in check is more important than getting hung up on which turbo is making the boost.

 

Yeah exactly. I want to rev to 8-9k but not wait til 5k to be in boost. The reason I am focusing on which turbo is doing the work is to make sure I am operating with reasonable efficiency of both compressor maps and I want to avoid compounding because I need to stay under 30 psi to stay within the limits of the block.

 

Did you ever tell us what engine you are working on? If it's common then there may be enough experience for someone to recommend the turbo sizes.

Our of curiosity what ECU do you plan to run?

 

I believe I have the sizing figured out, but I'm always open to more input, as most of my knowledge is working with single or parallel turbos. There could be some completely different variables in play here that I may not have considered. I would also like to stick with Garrett since I have good pricing through the shop I used to work at.

My plan is to de-stoke a Subaru Ej25 down to 2.33L so I can have a wider rev range. It'll be run off the factory ECU, but probably have the addition of a separate boost controller for the large turbo. I'll build a custom header to run a low mount Garrett G25-550 with the 0.72 A/R hot side for quick spool down low to make up for the lost torque of the shorter stroke. Then I would mount a Garrett GBC37-900 with the 0.95 A/R in the factory turbo location to hold boost up to 8k or 9k redline (where ever it stops making power). I'm thinking I need about 66lb/min at 3.0 PR to get the 550whp that I am limiting myself to to preserve the block. From what I've seen around the Subaru forums, this concept is way over most people's heads which is why I thought I would ask around here.

 

If you want sequential turbos, then it has to be set up that way. You can't blow thru another turbo and say you don't want to compound. It's still going to multiply. To do it sequentially you have to physically seperate flow to/ from each turbo, exhaust and boost.

You'll have to pick where you want the switchover at, and then we can go from there. DO NOT pick a/r "because you want spool" a/r is picked according to exhaust flow.

 

Thank you for the input. I don't doubt what you say, but I do have some follow up questions. Just trying to get a full understanding.

-If the small turbo wastegate is equal in size to the header collector and positioned in a way where it would have flow priority when it is open, how much is the small turbo really doing once the wastegate is fully open? Of course, it's spinning. I assume no matter what some amount of exhaust is going to go through it, but how much could it actually be multiplying at such a low flow?

-Honestly, I am not opposed to trying to run this as a compound setup, however I don’t feel like Garrett has a combination of turbos that would pair well to reach the goals I want and stay in a good efficiency range for both turbos. It is totally possible I am approaching the high-pressure turbo math all wrong though. I haven’t found much reliable information out there to go off. I am calculating the pressure and density of the low-pressure outlet and using that to get the volumetric flow rate. Then I am converting the mass flow rate axis on the compressor maps to volumetric since the high pressure isn’t referencing ambient pressure. Am I on the right track, or should I be doing this differently?

-Last question goes back to the wastegate for the small turbo. And this one is more hypothetical. I understand what your are saying about not using A/R to choose a turbine housing, and in a normal setup I absolutely agree, however if the idea was just to spin it up quickly to get the large turbo on line and then bypass it completely (assuming that was even possible) as long as it's capable of flowing the necessary volume required to switch to the large turbo, would the smaller A/R really be an issue?

 

Ok, I'll answer these in order...

To the exhaust flow, there is no difference between a turbine and wastegate, both are a pressure drop and will therefore have an equal pressure drop. Meaning the turbine will still be providing power to the compressor. Even if it's only doing a measly 1.2 pressure ratio, it multiplies whatever pressure ratio is coming out of the low pressure turbo. If your making 15psi (2.0pr) out of the large turbo, and the high pressure is only doing 1.2pr (3psi if it was a single) then your now making 2.4pr/20psi. It's always going to multiply, and making it go through a wastegate first isn't the cleanest/ most efficient way to do it, since it expands the exhaust and will require a larger turbine stage in the low pressure turbo than what normally would be required.

You have to correct the compressor flow to standard conditions so that it can be relevant to the compressor map for the high pressure turbo. Find the compressor outlet temp in absolute temp. Divide that by standard absolute temp (288k metric, 520R standard) square root that, multiply that by the lbs/min flow rate, then divide that by the pressure ratio.

Say the big turbo is doing 60lbs/min at 2.0 pressure ratio and 75% efficiency. Comp outlet temp is 372°K. 372÷288=1.3. Square root of 1.3 is 1.14. 1.14×60lbs/min= 68. Finally 68÷2.0pr= 34lbs/min. So compressor is sucking in 60lbs/min, compressing it and is coming out at 34lbs/min.

Garrett doesn't make really good pressure capable turbos until around the gt47 sized turbos. Holsets are really good at high pressure, but aren't ball bearing. Ball bearing handles the thrust load changes from compounds better. Garretts can definitely be made for the part....

Hope that helps.

 

First of all, no apology necessary. I hope for a speedy recovery for both you and your wife.

This information is a huge help! It is definitely a different formula than I have been using. I recalculated a few points to test and it brings me back to a good efficiency range on both turbos as a compound set up. I'm going to take some time to re-crunch all my numbers and see where that lands me now. I'll definitely have a few more questions.

 

What you should notice, is despite breathing in 60lbs/min, the outlet flow should match the breathing capability of the engine at whatever rpm your figuring.

Say 2.3L at 9000rpms, at 100% ve that's 364cfm, or 27lbs/min. You'll see that as boost rises, pre compressor flow increases, but the outlet flow always stays the same at the same rpm. Even if you did 100lbs/min at 60psi, the airflow leaving the turbo will still be 27lbs/min

 

Okay I think you lost me there. I feel like something in there should have switched to volumetric flow.

If I'm interpreting this right it seems you are saying the compressor outlet should match the flow of the engine naturally aspirated. If that's the case I'm definitely doing something wrong. Here's an example of what I'm seeing:

ambient: 70 degF and 14.7psi
disp: 2,332 cc or 143 ci
NA airflow at 8k rpm with 95% VE 314 cfm or 22lb/min
62lb/min at 3.0 PR needed to make 630 crank hp under boost

For low pressure turbo (GTX-3584RS) the inlet must pull in the full 62 lb/min
The LP outlet is at 1.76 PR at comp efficiency 0f 68% making 377K air temp

The high pressure turbo is still going to ingest 62lb/min because the mass isn't changing, just the volume and density. Now using the formula you gave before, assuming you meant corrected mass flow to work with a compressor map, I come up with 38lb/min entering the HP turbo. Since that is nowhere near the 22lb/min NA air requirement I tried running everything again thinking you probably meant what the engine sees.

So, HP turbo adds another 1.76 PR at 68% efficiency with an outlet temp of 461K. If I use that correction formula again I come up with 29lb/min into the intercooler. That's 25% higher than the 22lb/min I came up with before.

Either I'm misunderstanding or my math is way off. I really don't understand how mass is changing.

 

You have to figure the temp and pressure drop through the intercooler, and then correct that.

Like you said, the mass actually isn't changing, correcting it is just taking into account the new density of the air and making it relatable on the compressor maps. The volume is what is actually changing but the process to account for the density change is alot harder to do.