Close

Page 1 of 2 12 LastLast
Results 1 to 25 of 33
  1. #1
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2


    1 out of 1 members liked this post. Yes Reputation No

    Hey DBFIU, got a question about Y pipe design for exhaust systems

    Hey guys, I`m working on a custom Y pipe for my rb26 and having trouble finding the correct info in my various text books. I`m trying to determine the relationship between inlet and outlet sizes of a Y pipe in an exhaust system (post turbo). Its a twin turbo set up where each turbo exits into a pipe and then those 2 pipes merge.

    So far, I`m finding that backpressure is the most influential variable on power for this application. thus, I`m tempted to ignore the physics and just fit the largest size piping I can.

    But, I`d love to know the reasonable assumptions I can make for this application. Can I reasonably assume:
    -I`m working with an ideal gas?
    -I can ignore compressibility effects?
    -I can ignore friction?
    -I can ignore heat loss to the environment?
    -I can assume the same density before and after the split in flows?
    -I can work with the law of conservation of mass to determine sizes?

  2. #2
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    1 out of 1 members liked this post. Yes Reputation No
    Yes you can assume an ideal gas, you will need the correct gamma (Cp/Cv) which can be found online for exhaust gas. ( I believe it is 1.33 instead of 1.4 for air)

    You cannot ignore compressibility if your mach number is greater than 0.25. The way to calculate your mach number is from your mass flow out the exhaust, your pipe size (flow area), and your static and total pressures. Chances are you cannot ignore it.

    You cannot ignore friction, it will play a huge role on your pressure drop especially since the flow is swirling (swirling is when flow has a tangential component to flow, like a corkscrew type of flow).

    You can ignore heat loss to the environment for now, isothermal hand calculations are much easier and the gas is moving so fast that it wont appreciably lose that much temperature from turbo outlet to the Y pipe.

    You can assume densities remain the same before and after the split flows. The pressure will drop after the Y pipe but not much such that you need to globally change your density in your calc. Remember this is a hand calculation and there is no need to over analyze the component. Lets get a rough esimate for now and then we can get into more detail after we have a good baseline design

    I think I can help you, what kind of boost are you looking at?

    Tell me these things and I will design your system.

    Engine speed: RPM
    Engine Displacement
    Boost pressure
    Engine VE, assumed. If you dont give me this I will default to 85%
    Ambient temp
    Intercooler effectiveness I will assume 65% unless otherwise noted
    What kind of compressor are you using? If you have a map that would be helpful.

    Hope it helps.
    Last edited by DBFIU; 02-05-2012 at 01:59 PM.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  3. #3
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No
    Engine speed: 8500RPM
    Engine Displacement 2.6l
    Boost pressure 22psi
    Engine VE, assumed. If you dont give me this I will default to 85% - sure
    Ambient temp 25*C
    Intercooler effectiveness I will assume 65% unless otherwise noted - sure
    What kind of compressor are you using? If you have a map that would be helpful. gt2860rs x2 - http://www.turbobygarrett.com/turbob...t/turbocharger

    I`m likely limited to 2.75" ID inlets due to space constraints. Ideally, I`d use a 3.5" ID outlet.

  4. #4
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    2 out of 2 members liked this post. Yes Reputation No
    Motor will be moving about 1 lb/s through the exhaust.

    I bumped up your VE to 88% because i am assuming you are building this motor from scratch and you will use good flowing heads/valves.
    Also assumed 75% for intercooler effectiveness.

    Power comes out to about 580 crank HP. Not sure if this is right at all, could be totally wrong.


    Anyway, with an exit mass flow of 1 lbm/s you are looking at average exhaust velocity of 158 ft/s off each turbo with 2.75" pipes. When it hits the Y pipe if you want to keep the areas the same you'll need at least a 3.75" merge pipe. Your gas velocity with the 3.5" pipe will be close to 200 ft/s which is in fact speeding the gas up after the merge which will just add more backpressure.

    Do you plan on boosting more than 22 psi in the future?
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  5. #5
    Join Date
    Feb 2010
    Location
    Corpus Christi, Texas, United States
    Posts
    3,106
    Rep Points
    1,236.6
    Mentioned
    6 Post(s)
    Rep Power
    13


    1 out of 1 members liked this post. Yes Reputation No
    With the right injectors and fuel, he could for sure push more than 22psi with those turbos. I have similar on my Z. Click here to enlarge

    DFBIU, you're awesome man. I feel I would have some awesome conversations with you (for me at least) if we actually hung out in person.

  6. #6
    Join Date
    Nov 2010
    Location
    le Paris
    Posts
    6,655
    Rep Points
    -195.0
    Mentioned
    48 Post(s)
    Rep Power
    0


    Yes Reputation No
    Click here to enlarge Originally Posted by SlicktopTTZ Click here to enlarge
    DFBIU, you're awesome man. I feel I would have some awesome conversations with you (for me at least) if we actually hung out in person.
    +1
    It would be really cool to have the opportunity to have a conversation with him.

  7. #7
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    1 out of 1 members liked this post. Yes Reputation No
    You guys can call me anytime you want. Just let me know beforehand when you call so I can put my wife and kids back in the cage.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  8. #8
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No
    Click here to enlarge Originally Posted by DBFIU Click here to enlarge
    Motor will be moving about 1 lb/s through the exhaust.

    I bumped up your VE to 88% because i am assuming you are building this motor from scratch and you will use good flowing heads/valves.
    Also assumed 75% for intercooler effectiveness.

    Power comes out to about 580 crank HP. Not sure if this is right at all, could be totally wrong. That sounds very close to what I expect.


    Anyway, with an exit mass flow of 1 lbm/s you are looking at average exhaust velocity of 158 ft/s off each turbo with 2.75" pipes. When it hits the Y pipe if you want to keep the areas the same you'll need at least a 3.75" merge pipe. Your gas velocity with the 3.5" pipe will be close to 200 ft/s which is in fact speeding the gas up after the merge which will just add more backpressure.

    Do you plan on boosting more than 22 psi in the future? No definite plans.
    Thanks a bunch. I see you mentioned that you want to maintain total area of the pipe before and after the split. How did you reach that conclusion? At first, I thought it was a reasonable assumption to make for this 'just for fun' application, but taking compressibility and friction into consideration, I wasn't so sure. I see our Mach number is about 0.14, so we can safely ignore compressibility (thank god!). How about friction? What laws can I look into to understand its effects (don't have my text infront of me atm)?

  9. #9
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    The reason I wanted to maintain constant areas is to keep the flow from undergoing expansion or contraction losses.

    But, just because the areas are the same does not necessarily mean there wont be 'other' types of losses in the Y pipe. Whenever you have to join two flows, you will be exchanging momentum. As the angle of the branches of the Y pipe approaches zero, the amount of momentum exchanged (assuming the inlet flows are identical) will approach zero. Since your Y pipe angle is not zero, there will be some losses associated with it, but dont worry about those losses they are small compared to a contraction due to an area change.

    Area change and friction are the major players in the losses. I also want to point out, since both flows are swirling like corkscrews, they have to travel a longer path along the inside walls of the pipes which increases the EFFECTIVE length of the pipe since the flow is flowing around the inside at some helix angle. Because of this, my calculated pressure drop may have to be augmented due to the swirl angle, which I cannot assume unless I knew the geometry of your turbine wheel.

    Now in a Y pipe with two flows merging to one, where the inlet flows are both swirling, there will be EVEN MORE momentum exchange such that either the swirls will cancel eachother out or they will be additive (depending on if your turbos are both spinning in the same direction or not).

    Anyway, there is no need to turn this into a total science project. We can go on crazy tangents and try and squeeze out every horsepower possible from the Y pipe, but for your application. You should be fine with the 2.75" pipes into a 3.75". If you cant fit a single 3.75" try and go to dual 3" pipes.

    Does your application force you to use a Y pipe? It would be better to keep the flows separate and just have an H pipe or X pipe in there for pressure balance.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  10. #10
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No

    Thumbs up

    Not forced to use a Y pipe, but i'd like to use single rather than dual for the weight savings.

    Conclusion: for recreational use, a Y pipe can be designed such that the inlets and outlet maintain a constant cross sectional area. Click here to enlarge

  11. #11
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Click here to enlarge Originally Posted by black bnr32 Click here to enlarge
    Not forced to use a Y pipe, but i'd like to use single rather than dual for the weight savings.

    Conclusion: for recreational use, a Y pipe can be designed such that the inlets and outlet maintain a constant cross sectional area. Click here to enlarge
    That will be your best bet.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  12. #12
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Hey @black bnr32 i am going to simulate this Ypipe with swirl components...

    I am kinda bored so this is just for fun and learning.

    Click here to enlarge
    Last edited by DBFIU; 02-08-2012 at 02:27 PM.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  13. #13
    Join Date
    Nov 2010
    Location
    le Paris
    Posts
    6,655
    Rep Points
    -195.0
    Mentioned
    48 Post(s)
    Rep Power
    0


    Yes Reputation No
    Dan i assume you are a mechanical engineer, right?

  14. #14
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Click here to enlarge Originally Posted by Sorena Click here to enlarge
    Dan i assume you are a mechanical engineer, right?

    Yes..
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  15. #15
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No
    Neat...I was looking for some cheap/free cfd software for recreational use, but didn't get far, or I would have done the same.

    so far I am weighing a few options with this project, but no decisions yet.

  16. #16
    Join Date
    Feb 2010
    Location
    NJ
    Posts
    13,485
    Rep Points
    58.0
    Mentioned
    318 Post(s)
    Rep Power
    0


    Yes Reputation No
    Click here to enlarge Originally Posted by DBFIU Click here to enlarge
    Yes..
    devise a plan for me to get a TS on this sucker then Click here to enlarge

  17. #17
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Click here to enlarge Originally Posted by black bnr32 Click here to enlarge
    Neat...I was looking for some cheap/free cfd software for recreational use, but didn't get far, or I would have done the same.

    so far I am weighing a few options with this project, but no decisions yet.
    Doing CFD is totally moot IMO. If you want the 5 hp back from the better Ypipe design just run 0.25 psi more boost. Click here to enlarge

    The CFD is best when used for tuning big power adders like manifolds, turbos, intercoolers, piping, intake/exhaust ports, in cylinder combustion etc...

    There is free CFD out there, google search the following:

    openFOAM
    Salome post processing
    LISA FEA
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  18. #18
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    1 out of 1 members liked this post. Yes Reputation No
    Click here to enlarge Originally Posted by LostMarine Click here to enlarge
    devise a plan for me to get a TS on this sucker then Click here to enlarge
    A TS on an S65 would be so nice. I think it needs to be done, the motor is begging for it.

    For you to do this one would obviously try and mount it directly on top of the engine in place of the manifold, just like any TS on a V8.

    One would quickly realize that because of the location of the discharge port on the blower, and the fact that the S65 has ITBs with dedicated butterfly valves, that the blower will bias a ton of flow to one portion of the engine (the rear) and this bias is not negligible.

    The solution will have to come with some type of manifold design or blower discharge port that allows for even feed of the ITBs. The blower itself is the easy part.

    In fact, I would design the manifold such that the discharge port sits directly centered over the ITBs, this may force your blower snout to be sticking forward too far, or something along those lines wherein you would need custom work. If I had to choose the custom work to be done, it would be there.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  19. #19
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No
    Hey DBFIU, what do you make of the 'coke bottle' shape transitions that Burns Stainless use in their designs? I can't find info for a post-turbo application. I assume its just to reduce expansion losses at the merge.

    http://www.burnsstainless.com/crossoversmain.aspx

  20. #20
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Burns makes some very good off the shelf parts. The expansion joint looks good, just as you said, it is to reduce expansion losses.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  21. #21
    Join Date
    May 2010
    Location
    canada
    Posts
    522
    Rep Points
    122.9
    Mentioned
    2 Post(s)
    Rep Power
    2



    Yes Reputation No
    Same theory for NA and post-turbo applications?

  22. #22
    Join Date
    Apr 2010
    Posts
    4,616
    Rep Points
    3,236.6
    Mentioned
    24 Post(s)
    Rep Power
    33


    Yes Reputation No
    Yes except post turbo you have swirl flow and you typically dont have to worry too much about pulse tuning your post turbo pipes. Pulse tuning on a turbo occurs between the turbo and the head, where in an NA car it occurs in the exhaust header all the way to the tip of the tailpipe. The turbine acts like a big blockage, so when pulse tuning the exhaust manifold on a turbo car, it's usually only done when you need every ounce of power. Most aftermarket companies just make the primaries equal length, which does not constitute pulse tuning, it just means the primary pressure waves dont overlap.

    In NA you kind of have to be careful with pulse tuning your exhaust pipes so the shape of the Y is much more critical. For your application dont even worry about it, just get one of those nice burns Y pipes.

    Sorryfor going off topic, but anyway because you need the Y pipe post turbo you can get away with almost any decent looking Y pipe.
    Some people live long, meaningful lives.

    Other people eat shit and die.

    I'm not racist, I hate everybody equally; especially fat people.


    Click here to enlarge

  23. #23
    Join Date
    Aug 2010
    Location
    Valley Stream NY
    Posts
    3,348
    Rep Points
    2,757.4
    Mentioned
    85 Post(s)
    Rep Power
    28


    Yes Reputation No
    So how does speeding exhaust velocity up attribute to higher back pressure? I thought that as whatever fluid in motion increases in speed, pressure drops? How would keeping to a 3.5" pipe hurt over a 3.75?

  24. #24
    Join Date
    Nov 2009
    Location
    Miami, FL.
    Posts
    1,892
    Rep Points
    771.0
    Mentioned
    9 Post(s)
    Rep Power
    8


    Yes Reputation No
    This thread is $#@!ing awesome.
    Click here to enlarge
    2007 335i Coupe
    Mods: Check the Garage

  25. #25
    Join Date
    Aug 2010
    Location
    over there
    Posts
    288
    Rep Points
    481.1
    Mentioned
    14 Post(s)
    Rep Power
    5


    Yes Reputation No
    DBFIU what would the ideal exhaust system be for the n54 with stock turbos and would it change with rb's?
    Turbo lag is the on ramp to the highway which is power.

Page 1 of 2 12 LastLast

Tags for this Thread

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may post replies
  • You may not post attachments
  • You may not edit your posts
  •