compound turbo design questions
#1
compound turbo design questions
Say we wanted to design a compound turbo setup, is it necessarily the best way to shoot for a horsepower target or is it more effective to define how we want it operate to attempt to size the turbo wheels and housings to achieve desired flow rates and pressure ratios while maintaining favorable drive pressure across the powerband much as possible? Here are the equations for calculating the pressure ratios and actual secondary boost:
1) overall pressure ratio: (total boost + atmospheric pressure)/atm pr
2) primary pressure ratio: (interstage boost + atm pr)/atm pr
3) secondary pressure ratio: (overall pressure ratio)/(primary pressure ratio)
4) secondary boost: [(secondary pressure ratio)(atm pr)] - atm pressure
Note that the atmospheric pressure would be for your location whether in mountains or sea level. Analysis of the pressure ratios can help determine waste gate settings to attempt to get the load on turbos distributed equitably. Still learning about design and operation of compound turbo setups; it is really fascinating to understand how they work.
1) overall pressure ratio: (total boost + atmospheric pressure)/atm pr
2) primary pressure ratio: (interstage boost + atm pr)/atm pr
3) secondary pressure ratio: (overall pressure ratio)/(primary pressure ratio)
4) secondary boost: [(secondary pressure ratio)(atm pr)] - atm pressure
Note that the atmospheric pressure would be for your location whether in mountains or sea level. Analysis of the pressure ratios can help determine waste gate settings to attempt to get the load on turbos distributed equitably. Still learning about design and operation of compound turbo setups; it is really fascinating to understand how they work.
#2
DTR 1st Sergeant
That's a good start. Also there are white papers that can be observed on the Garrett website explaining Diesel turbocharging and compounding turbos.
In the end, the math probably gets you about 80% of the way, in the final analysis two things have stood out to me. First, because every truck is different, there are not always equal outcomes given the same setups. Second, there is some trial and error for the finished product that you can't get away from.
In the end, the math probably gets you about 80% of the way, in the final analysis two things have stood out to me. First, because every truck is different, there are not always equal outcomes given the same setups. Second, there is some trial and error for the finished product that you can't get away from.
#3
DTR 1st Sergeant
Let me state up front that a lot of my learning curve is from direct experience from trial and error and a lot of practical R&D. My experience is what it is, no more no less. I guess what I am trying to say is please don't make me out to be "the expert". I am no Pius Bell.
That said, my first twins were the old Piers HX35/BHT3B. Working with Piers Harry on improving these, he suggested (this was a long time ago) that we give the new Garrett BB GT series of turbos a try. Working first with Piers and then later with his partner Harry Kelleher at Peak Performance and Tom at Turboresource (I have to give a shout out to those guys, you truly won't find better people to work with and I owe huge debts of gratitude to both), we managed to go through 4 different ball bearing turbo configurations (I had 3 of them myself) handling from 450hp through >1000hp on #2.
But that's it. Folks like Nathan at MPI or Rip Rook have done much more than I.
That said, my first twins were the old Piers HX35/BHT3B. Working with Piers Harry on improving these, he suggested (this was a long time ago) that we give the new Garrett BB GT series of turbos a try. Working first with Piers and then later with his partner Harry Kelleher at Peak Performance and Tom at Turboresource (I have to give a shout out to those guys, you truly won't find better people to work with and I owe huge debts of gratitude to both), we managed to go through 4 different ball bearing turbo configurations (I had 3 of them myself) handling from 450hp through >1000hp on #2.
But that's it. Folks like Nathan at MPI or Rip Rook have done much more than I.
#4
DTR 1st Sergeant
Now, from here, I will put things in more general layman's term language. I am afraid the math involved is really no benefit to most folks... it won't provide the salient or useable information the average reader wants/needs.
To start, let's talk about concepts.
I like to talk about windows. Basically, every turbo has a "window" of operation. Otherwise known as its "Map". As a single turbo, every turbo has a relatively small operating window. To illustrate, take a BD super B turbo with its small compressor wheel and turbine with a silver bullet with its relatively large compressor wheel and turbine. The super b is outstanding down low and very responsive but peters out quickly on the top end. The silver bullet on the other hand will scream on the top end, but you really don't want to tow with it!
What compounding allows is a much larger window. The spool of the smaller turbo and the air flow of the larger turbo. It takes care of a much broader application of power. It still has a "window" though. What ball bearing turbos offer, by way of much faster spool, is a yet larger window of operation not possible with journal bearing turbos.
The holy grail of having "everything" is not yet possible-- at least not in the realm of being financially practical. I can't tell you how many times I get the question of "I want instant spool down low with no smoke so I can tow 40,000 lbs, support 10 billion HP with egts no higher than 1100*, and do all this at $4000 or less and I don't want water injection." Not gonna happen.
To start, let's talk about concepts.
I like to talk about windows. Basically, every turbo has a "window" of operation. Otherwise known as its "Map". As a single turbo, every turbo has a relatively small operating window. To illustrate, take a BD super B turbo with its small compressor wheel and turbine with a silver bullet with its relatively large compressor wheel and turbine. The super b is outstanding down low and very responsive but peters out quickly on the top end. The silver bullet on the other hand will scream on the top end, but you really don't want to tow with it!
What compounding allows is a much larger window. The spool of the smaller turbo and the air flow of the larger turbo. It takes care of a much broader application of power. It still has a "window" though. What ball bearing turbos offer, by way of much faster spool, is a yet larger window of operation not possible with journal bearing turbos.
The holy grail of having "everything" is not yet possible-- at least not in the realm of being financially practical. I can't tell you how many times I get the question of "I want instant spool down low with no smoke so I can tow 40,000 lbs, support 10 billion HP with egts no higher than 1100*, and do all this at $4000 or less and I don't want water injection." Not gonna happen.
#5
Beats my experience! 😬 I replaced stock turbo with a GT3782R I got from Tom at Turbo Re-Source and it is really a great little charger. I more than halfway thought about getting a stocker-475 setup from Anarchy Diesel and using the Garrett in place of the HE351. Having spent a couple hundred hours reading everything I can get ahold of on compounds I still have a lot of questions and really like getting knowledge from guys who have a lot of hands-on because that's really the best! School of hard knocks etc.
#6
DTR 1st Sergeant
To the OP and the questions, while pressure ratios and air flow are important, "boost" is a less relevant number. It is different on every truck and only is a relative measurement of restriction. Things like the flow of the CAC, a cam, intake horn, exhaust manifold, turbine selection, head porting, and etc, all play varying degrees of influence.
Equal sharing of turbo loads only begins to be important in the last 1/3 of max flow of the secondary or compounding turbo. Efforts to simply try to get 1:1 ratios all the way through the range is wasted effort. When at max HP, equal distribution of air is desirable... not mandatory. This is where some of the real world results vs math come in. Brown's diesel has turned well over 1k of HP using a 62mm top turbo with a s5xx on the bottom. This means the primary (or atmospheric or bottom turbo) was doing most of the work and the top turbo (compounding or secondary turbo) was just an inline restriction at that point.
It is not correct to say at a certain point you can only get more HP with a larger primary. It can be done with a larger secondary too.
Let's go back and visit the old Piers twins. The 35/3B twins were good to about 550 hp. Going to a HX40 on the top (40/3B) allowed up to 700hp. The bottom turbo was the same. The difference is not simply that the HX 40 provided more air. It does, but not that much. What happens is that by virtue of the larger compressor wheel and shroud, the top turbo is better able to "swallow" and process the air being fed by the primary. There is a law of diminshing returns when there is an attempt to mis-match sizes of turbos. In other words a really small top turbo for spool with a really large bottom turbo is no bueno. There has to be a reasonable size relationship between the two. Garrett recommends two turbo sizes (like a gt 35 with a gt42) although, like the Brown's Diesel example previously, that is not an absolute.
Equal sharing of turbo loads only begins to be important in the last 1/3 of max flow of the secondary or compounding turbo. Efforts to simply try to get 1:1 ratios all the way through the range is wasted effort. When at max HP, equal distribution of air is desirable... not mandatory. This is where some of the real world results vs math come in. Brown's diesel has turned well over 1k of HP using a 62mm top turbo with a s5xx on the bottom. This means the primary (or atmospheric or bottom turbo) was doing most of the work and the top turbo (compounding or secondary turbo) was just an inline restriction at that point.
It is not correct to say at a certain point you can only get more HP with a larger primary. It can be done with a larger secondary too.
Let's go back and visit the old Piers twins. The 35/3B twins were good to about 550 hp. Going to a HX40 on the top (40/3B) allowed up to 700hp. The bottom turbo was the same. The difference is not simply that the HX 40 provided more air. It does, but not that much. What happens is that by virtue of the larger compressor wheel and shroud, the top turbo is better able to "swallow" and process the air being fed by the primary. There is a law of diminshing returns when there is an attempt to mis-match sizes of turbos. In other words a really small top turbo for spool with a really large bottom turbo is no bueno. There has to be a reasonable size relationship between the two. Garrett recommends two turbo sizes (like a gt 35 with a gt42) although, like the Brown's Diesel example previously, that is not an absolute.
#7
DTR 1st Sergeant
So far, we've visited only the airflow end.
Now I want to talk a little about the exhaust end.
The influences on exhaust are first the restrictions. Exhaust houses, turbine sizes, turbine blade design, wastegates, head porting, exahust manifolds, even the downpipe and the rest of the exhaust is important. How many here went from their factory 4" turbo back to a 5" turbo back and all of a sudden experienced a loss of lower end spool and power? While you want as much exhaust evacuation as possible, some back pressure is necessary for it all to work.
Now I want to talk a little about the exhaust end.
The influences on exhaust are first the restrictions. Exhaust houses, turbine sizes, turbine blade design, wastegates, head porting, exahust manifolds, even the downpipe and the rest of the exhaust is important. How many here went from their factory 4" turbo back to a 5" turbo back and all of a sudden experienced a loss of lower end spool and power? While you want as much exhaust evacuation as possible, some back pressure is necessary for it all to work.
Trending Topics
#8
Great points. I visited with Tom about these things and he recommended the GT4202R under the 3782 or even the 3788. And I think that what a lot of guys seem to ignore is the drive pressure and that boost is just a measure of restriction. I do know that with my truck when I installed the 3782 that the engine really ran nicely. It has the turbine wheel out of a 3788and the compressor wheel out of a 3582. I don't think it really moves a lot more air than stock, but that turbine side removes a lot of restriction of the stock turbine so the resulting boost I see is quite a bit lower than it was stock. EGTs are very reasonable and mileage is good too. Towing is pretty good too but obviously limited on the air and still have stock fuel with EFI Live tuning. Anyway I was wondering exactly how to go about sizing the chargers and thought it could be narrowed down somewhat by using the compressor maps to have the HP charger map coinciding with at least half or two thirds of the LP charger map? So that the transition would be more gradual and maybe even not really need to wastegate around the secondary? Don't know if that thinking is accurate but seems plausible. Really just so many questions around the wastegate. And turbine sizing in relation to drive pressure is often overlooked but I suspect may kind of be the hidden secret to building a really awesome set of compound turbos. Thinking of enhancing the engine breathing between the intake stroke to the exhaust stroke means looking at the systemic restriction and making those restrictions low as practical can only help. It's all a balance of air throughput with flow restriction for good spool with adequate top end so how do we go about designing a system for overall efficient air flow through the engine? Is that the best way to begin to do our design work?
#9
Good thing I didn't rush out to buy a 5" exhaust... In fact I still have the factory cat in place but from what I have read those do not create a lot of restriction in the exhaust anyway. Hopefully that is true. I also have an Aeroturbine 4040XL muffler inline just to knock some of the drone out of the cabin.
#10
DTR 1st Sergeant
Who here has had a silver 62 or silver 64 (with its relatively large turbine) and regretted the spool, or was disappointed in the high egts at cruise or down low in the rpm range? But, there was no drive pressure to speak of! Ideally, drive pressure should be 1:1 or slightly higher. If it is inverse (Like what happens with the silver 62 often) then the turbo is not driven hard enough to be of good benefit.
#11
DTR 1st Sergeant
OK, now to get to the meat of the questions.
So the first variable we want to talk about is a HP goal. We want to support the upper end of the goal and work backwards. That will be the window. Sometimes it is the other way around as there are those folks who require the ability to tow heavy (prime requirement) but also want to hit 1000 hp. So the goal is unrealistic for the most part and a compromise ensues. Especially since the cost of supporting all that skyrockets! You would be surprised at the folks who think EFI live with injectors and twins (and nothing else) is all that is needed for huge HP.
So we know that a certain level of HP will produce a given volume of exhaust gas. From there we can do some rough calculations on what turbine sizes/exhaust housing sizes and wastegates are needed. Again compromise ensues.... something that provides low restriction at the upper end of HP is not optimal for low end performance. I choose to be in the middle and split the difference. So to still have respectable low end and good spool, I tighten things up a little and expect higher drive pressures at the upper end.
Now measuring boost vs drive, since I have removed a great deal of restriction on the intake side, my boost levels are much lower than would otherwise be stock-- I've removed about 18 psi of restriction. This allows the turbos to remain closer to the prime efficiency islands in the map. If I am at 72 psi (900hp) and both turbos are equalized, then that is only 36 psi apiece and not really driving either all that hard and are in a good island in the pressure ratio.
So the first variable we want to talk about is a HP goal. We want to support the upper end of the goal and work backwards. That will be the window. Sometimes it is the other way around as there are those folks who require the ability to tow heavy (prime requirement) but also want to hit 1000 hp. So the goal is unrealistic for the most part and a compromise ensues. Especially since the cost of supporting all that skyrockets! You would be surprised at the folks who think EFI live with injectors and twins (and nothing else) is all that is needed for huge HP.
So we know that a certain level of HP will produce a given volume of exhaust gas. From there we can do some rough calculations on what turbine sizes/exhaust housing sizes and wastegates are needed. Again compromise ensues.... something that provides low restriction at the upper end of HP is not optimal for low end performance. I choose to be in the middle and split the difference. So to still have respectable low end and good spool, I tighten things up a little and expect higher drive pressures at the upper end.
Now measuring boost vs drive, since I have removed a great deal of restriction on the intake side, my boost levels are much lower than would otherwise be stock-- I've removed about 18 psi of restriction. This allows the turbos to remain closer to the prime efficiency islands in the map. If I am at 72 psi (900hp) and both turbos are equalized, then that is only 36 psi apiece and not really driving either all that hard and are in a good island in the pressure ratio.
#12
DTR 1st Sergeant
All of this that I am talking about so far is building streetable multi-pupose trucks. Of course if it is a puller, dyno queen. drag racer... then I would look at things differently.
#13
DTR 1st Sergeant
Once we have the HP goal... and have an understanding of exhaust gas flow rates, we then look to the turbos with an understanding that it takes a certain amount of air (or "air molecules" if you will)
to supply that level of HP. I like to use the lbs/min of air as opposed to CFM. While there exists formulas to convert either, it is important to note temperature differences. You can have X CFM, but the amount of "air molecules" supplied will diminish as temperature goes up and air expands. It is not uncommon to see 600*f in air temperature coming out of the secondary turbo with all the compression... a good CAC becomes important as your air supply goes up over stock. Remember that the stock CAC is designed for a stock turbo. It is efficient for that and not a whole lot more.
to supply that level of HP. I like to use the lbs/min of air as opposed to CFM. While there exists formulas to convert either, it is important to note temperature differences. You can have X CFM, but the amount of "air molecules" supplied will diminish as temperature goes up and air expands. It is not uncommon to see 600*f in air temperature coming out of the secondary turbo with all the compression... a good CAC becomes important as your air supply goes up over stock. Remember that the stock CAC is designed for a stock turbo. It is efficient for that and not a whole lot more.
#14
DTR 1st Sergeant
So, to make it simple, we now have a HP goal, and start looking at primary chargers that produce enough lbs/min to support the goal. I like to go a little over than the formula would otherwise state due to inherent losses within the system. Then, we pick a compatable secondary, figure out the EH's and wastegate sizes.
For the most part, all of the pertinent combos have been vetted throughout the HP ranges and I (and others that do this for a living) can pick off the top of my head. In other words, been there done that no need to reinvent the wheel. (enough metaphors?)
That being said, I have seen and heard some industry folks who should know better give some pretty questionable advise, so buyer beware. If it sounds too good to be true... it probably is.
For the most part, all of the pertinent combos have been vetted throughout the HP ranges and I (and others that do this for a living) can pick off the top of my head. In other words, been there done that no need to reinvent the wheel. (enough metaphors?)
That being said, I have seen and heard some industry folks who should know better give some pretty questionable advise, so buyer beware. If it sounds too good to be true... it probably is.
#15
OK, now to get to the meat of the questions.
So the first variable we want to talk about is a HP goal. We want to support the upper end of the goal and work backwards. That will be the window. Sometimes it is the other way around as there are those folks who require the ability to tow heavy (prime requirement) but also want to hit 1000 hp. So the goal is unrealistic for the most part and a compromise ensues. Especially since the cost of supporting all that skyrockets! You would be surprised at the folks who think EFI live with injectors and twins (and nothing else) is all that is needed for huge HP.
So we know that a certain level of HP will produce a given volume of exhaust gas. From there we can do some rough calculations on what turbine sizes/exhaust housing sizes and wastegates are needed. Again compromise ensues.... something that provides low restriction at the upper end of HP is not optimal for low end performance. I choose to be in the middle and split the difference. So to still have respectable low end and good spool, I tighten things up a little and expect higher drive pressures at the upper end.
Now measuring boost vs drive, since I have removed a great deal of restriction on the intake side, my boost levels are much lower than would otherwise be stock-- I've removed about 18 psi of restriction. This allows the turbos to remain closer to the prime efficiency islands in the map. If I am at 72 psi (900hp) and both turbos are equalized, then that is only 36 psi apiece and not really driving either all that hard and are in a good island in the pressure ratio.
So the first variable we want to talk about is a HP goal. We want to support the upper end of the goal and work backwards. That will be the window. Sometimes it is the other way around as there are those folks who require the ability to tow heavy (prime requirement) but also want to hit 1000 hp. So the goal is unrealistic for the most part and a compromise ensues. Especially since the cost of supporting all that skyrockets! You would be surprised at the folks who think EFI live with injectors and twins (and nothing else) is all that is needed for huge HP.
So we know that a certain level of HP will produce a given volume of exhaust gas. From there we can do some rough calculations on what turbine sizes/exhaust housing sizes and wastegates are needed. Again compromise ensues.... something that provides low restriction at the upper end of HP is not optimal for low end performance. I choose to be in the middle and split the difference. So to still have respectable low end and good spool, I tighten things up a little and expect higher drive pressures at the upper end.
Now measuring boost vs drive, since I have removed a great deal of restriction on the intake side, my boost levels are much lower than would otherwise be stock-- I've removed about 18 psi of restriction. This allows the turbos to remain closer to the prime efficiency islands in the map. If I am at 72 psi (900hp) and both turbos are equalized, then that is only 36 psi apiece and not really driving either all that hard and are in a good island in the pressure ratio.
How did you remove 18 psi of restriction? Also I was [blissfully] unaware of the charge air cooler not being able to cool much beyond the stock charger output? My wallet won't like that... Oh well. I have run lots of numbers from examples guys wrote in on compd and it always seems that at the top end the turbos are not balanced (primary doing most of the work) BUT that there can be good linearity displayed on the secondary both before and after the gate setpoint is exceeded. But not a lot of turbo systems seemed to show this. Also have to have multiple data points recorded to see this obviously, and record above and below the psi setpoint of the gate.
It has been pointed out that on a compound system that the ratios between secondary and primary are more important than the actual gauge pressure on the interstage and the overall boost pressure and apparently that is true in that if a person thinks the primary or secondary is doing more than it should the only real way to see that is to do those calculations otherwise can't really tell. Still kind of confused about how to properly size a twin set properly. Say I [think] I would like to have 600 hp and 1200 torque and I mostly use the truck for daily driving and towing a 10000 lb fifth wheel. Also say that I don't ever use the truck for racing or sled pulling. And that I would like to keep the EGTs as cool as possible for engine longevity. And go even further to say that I just bought a set of BBi Stage 1 injectors which are around 90 hp or thereabouts and that I may consider a dual pump setup in the future so that fuel should [theoretically?] not be much of a problem. Also say that I have EFI Live so getting the tuning right should not be too difficult either.
So in this theoretical desired application what would be best selection for the turbos? Assuming BB Garretts, would that be the GT3582R over the GT4202R or the 3788 over the 4202? Or a bigger primary?