Username or Email Address
Do you already have an account?
Forgot your password?
  • Log in or Sign up

    VWVortex


    Results 1 to 28 of 28

    Thread: 1.8t with ITB's, COMMON?

    1. 04-20-2014 11:05 PM #1
      Hi everyone.

      Soon Ill be building an AEB intake manifold. Ill be running a Borg Warner EFR 6758 Twin scroll.. equal length tubular manifold. Im going for quick spool over max hp since this will be a road race car. I know ITBs can drasticly improve throttle response so im seriously considering building a 1 off set for my manifold. It will be a lot of work designing them and actually making them so Id like to make sure its worth doing. The plenum itself will be a center entrance dual plenum design. Intercooler will be mounted horizontal directly in front of the engine with the shortest air path possible. Radiator will be in the trunk.

      So has anyone seen this done? Think Its a good idea? worth the effort?
      Last edited by Truckinduc; 04-20-2014 at 11:29 PM.

    2. Member G60 Carat's Avatar
      Join Date
      May 14th, 2004
      Location
      Regina, SK
      Posts
      5,628
      Vehicles
      1982 BMW 320i, 1986 Jetta Carat 1.8/ABA Turbo
      04-26-2014 12:16 PM #2
      Hmm Curious,

      I'm not 100% sure how much faster the response time will be vs a well designed SRI, and carefully sized, well thought out intercooler/pipe layout. Corky bell talks about air velocity of IC pipes vs the point they become a restriction in his book. It's really a balance of trying to keep velocity up, but restriction down. If you go along the thought of "less restriction, less restriction is everything!" You end up with a lot of power on a dyno sheet to show your friends, but low-mid rpm power and part throttle power will suffer. Not so important on a drag car or dyno queen, but very very important factors on a road race car.

      However I do know the SR20DET used it in the GTI-R so there must be something there. Nissan guys say it gives a flatter mid range, and slightly faster spool time.

      In to see where it goes for sure!
      Last edited by G60 Carat; 04-26-2014 at 12:36 PM.
      "Criminally insane, with a few brief lucid intervals."
      | Rust | Zip-Ties | JB Weld | Bad Idle | Scrapes Ground | Rubs when Turning | Busted Ass Door handles |

    3. 04-26-2014 12:55 PM #3
      It is very possible, just a PITA to get the tuning right...
      http://www.hulsenturbotuning.nl/proj...35pk-2008.html

    4. 04-27-2014 09:12 PM #4
      Quote Originally Posted by G60 Carat View Post
      Hmm Curious,

      I'm not 100% sure how much faster the response time will be vs a well designed SRI, and carefully sized, well thought out intercooler/pipe layout. Corky bell talks about air velocity of IC pipes vs the point they become a restriction in his book. It's really a balance of trying to keep velocity up, but restriction down. If you go along the thought of "less restriction, less restriction is everything!" You end up with a lot of power on a dyno sheet to show your friends, but low-mid rpm power and part throttle power will suffer. Not so important on a drag car or dyno queen, but very very important factors on a road race car.

      However I do know the SR20DET used it in the GTI-R so there must be something there. Nissan guys say it gives a flatter mid range, and slightly faster spool time.

      In to see where it goes for sure!
      Quote Originally Posted by CorrieG60 View Post
      It is very possible, just a PITA to get the tuning right...
      http://www.hulsenturbotuning.nl/proj...35pk-2008.html

      Thanks for the response guys, I was beginning to think no one would touch this subject.

      I know how important velocity VS restriction is. This car will be purpose built for usable midrange power And quick low rpm spool. Autocross - trackday - time attack. The decision to move the radiator to the rear and horizontally mount the intercooler is to minimize piping volume and therefore minimize pressure drop - lag. The cold side intercooler tank will be directly under the intake manifold and feed straight up into the manifold, approximately 14" total. The turbo will mount to the right of the cylinder head above the transmission, approx. 20" of pipe to the intercooler. I have not decided on an actual intercooler core yet. The hope is for the engine to produce 400 to 450 crank hp. The intercooler will be sized appropriately.

      All this is true regardless if i use one throttle body or 4. I can make the throttle body’s from scratch, that’s no problem. The major issue Ill have will be tuning, as I am 95% in the dark on the matter. The entire point of individual throttle body’s would be to pressurize the plenum during short throttle closings like gear shifts. When the throttle plates close the pressure backs up in the plenum momentarily until the plates re open (depending on how quick i can shift). The idea would be to gain instant positive pressure instead of waiting for the relatively high volume of the plenum to build back up to full boost. The effect would be maximized if I did not use a blow off valve (like alot of rally and hill climb cars), But I cant afford the risk of destroying the bearings of a 2000$ turbo due to thrust loading of compressor surge.

      Id love to hear your thoughts on this subject.

    5. 04-27-2014 09:15 PM #5
      Here is the start of the plenum backing plate. It will be 2 piece bolt together style.


    6. 04-27-2014 09:39 PM #6
      This is the last intercooler setup I built where efficiency and minimizing pressure drop were of top priority.







      Last edited by Truckinduc; 04-28-2014 at 04:33 AM.


    7. 04-29-2014 02:29 AM #8
      got a little bit more work done on the plenum




    8. 04-29-2014 09:23 AM #9
      Have you found a set if itb's yet? Looks awesome so far!

    9. 04-30-2014 02:05 AM #10
      Quote Originally Posted by slowgti View Post
      Have you found a set if itb's yet? Looks awesome so far!
      I planned on making them not buying them

    10. Member G60 Carat's Avatar
      Join Date
      May 14th, 2004
      Location
      Regina, SK
      Posts
      5,628
      Vehicles
      1982 BMW 320i, 1986 Jetta Carat 1.8/ABA Turbo
      04-30-2014 09:41 PM #11
      On the topic of BOV's since several, like Greddy RS are adjustable. Could you not dial it in to vent just enough to keep the turbo safe, but save you the spool up time?
      "Criminally insane, with a few brief lucid intervals."
      | Rust | Zip-Ties | JB Weld | Bad Idle | Scrapes Ground | Rubs when Turning | Busted Ass Door handles |

    11. 05-01-2014 03:12 AM #12
      Quote Originally Posted by G60 Carat View Post
      On the topic of BOV's since several, like Greddy RS are adjustable. Could you not dial it in to vent just enough to keep the turbo safe, but save you the spool up time?
      The turbo I'm running, EFR 6857 has a built in diverter that basically does that. The air that escapes actually hits the compressor wheel to keep it spooled, pretty trick.

      I guess it would be difficult to determine what pressure would keep the turbo "safe"

    12. 05-01-2014 03:28 AM #13
      Plenum plate is about 90% done. Just need to drill it and give it an O ring seal. Then I get to decide to build the rest of the plenum from aluminum or carbon fiber.





      And cut some out of the back side to shed some weight


    13. Member G60 Carat's Avatar
      Join Date
      May 14th, 2004
      Location
      Regina, SK
      Posts
      5,628
      Vehicles
      1982 BMW 320i, 1986 Jetta Carat 1.8/ABA Turbo
      05-01-2014 11:24 AM #14
      Holy,

      This might just be the sleeper thread of the year!
      "Criminally insane, with a few brief lucid intervals."
      | Rust | Zip-Ties | JB Weld | Bad Idle | Scrapes Ground | Rubs when Turning | Busted Ass Door handles |

    14. 05-01-2014 12:49 PM #15
      Looking good. Have you sized the throttle blades yet? Seems like you would want them slightly larger than the runner cross section to make up for the volume of the shaft and blade.

    15. 05-01-2014 03:25 PM #16
      Quote Originally Posted by slowgti View Post
      Looking good. Have you sized the throttle blades yet? Seems like you would want them slightly larger than the runner cross section to make up for the volume of the shaft and blade.
      Im staying single TB right now for tuning simplicity. I am building a carbon plenum though.


    16. 05-02-2014 01:36 AM #18
      Well I finished it up by adding the bolt holes and o ring slot. Now all thats left is to weld it to the runners.











    17. 05-02-2014 10:50 AM #19
      Your fabrication/machining skills are impressive!

      I did a CFD study a while back on runner inlet radius/profile shaping that I think would be applicable here. I'll have to finish that up and share at some point.

      For what it's worth, my opinion is that you made the right choice with a single throttle body. I'm sure you are definitely capable of making ITBs happen, but in a turbocharged car I don't think you'd see the same benefits (at least relative to the additional effort).

    18. 05-02-2014 03:12 PM #20
      Quote Originally Posted by leebro61 View Post
      Your fabrication/machining skills are impressive!

      I did a CFD study a while back on runner inlet radius/profile shaping that I think would be applicable here. I'll have to finish that up and share at some point.

      For what it's worth, my opinion is that you made the right choice with a single throttle body. I'm sure you are definitely capable of making ITBs happen, but in a turbocharged car I don't think you'd see the same benefits (at least relative to the additional effort).
      Thanks. I'd love to hear what you think about my design.

      The runners open up into a .5" radius that blend into a .25" radius.

    19. 05-02-2014 08:35 PM #21
      Quote Originally Posted by leebro61 View Post
      Your fabrication/machining skills are impressive!

      I did a CFD study a while back on runner inlet radius/profile shaping that I think would be applicable here. I'll have to finish that up and share at some point.

      For what it's worth, my opinion is that you made the right choice with a single throttle body. I'm sure you are definitely capable of making ITBs happen, but in a turbocharged car I don't think you'd see the same benefits (at least relative to the additional effort).
      Very curious as well. Is there anyway you can play with the surface finish and see how that affects flow?

    20. 05-03-2014 03:02 AM #22
      went ahead and made the plenum flange. This one is for a sheetmetal plenum. I havent 100% decided on carbon fiber for the plenum yet, but If i do Ill make another flange to suit it.



      its hard to see but the inner wall has a radius to it



      and an artsy fartsy photo i took with instagram


    21. 05-03-2014 06:23 PM #23
      Quote Originally Posted by Truckinduc View Post
      Thanks. I'd love to hear what you think about my design.

      The runners open up into a .5" radius that blend into a .25" radius.
      If i get home at a reasonable hour tonight i will host some pics and share my conclusions from my quick study.

    22. 05-04-2014 12:35 AM #24
      Quote Originally Posted by slowgti View Post
      Very curious as well. Is there anyway you can play with the surface finish and see how that affects flow?
      Yes and no (although it's more of a no than a yes ). I'll start with the "no", then work my way to the "yes".

      In CFD we discretize (sub-divide) the physical geometry into small cells where we solve continuum equations for flow field properties. The cells are typically much, much larger in the streamwise direction than whatever geometric variations you would normally expect to see because of surface finish; so in a way the CFD would not even know that the surface imperfections were there. Here is an example below.



      The solid black line is a NACA airfoil that is being gridded. The equation of that line is a smooth, continuous shape. The CFD grid of that shape would look similar to the red or the blue line(s)... which are just clusters of points that the code would ~interpret as being connected by straight lines (pending the order of accuracy with which the code computes surface curvature), which in reality is of course not the case. If the variation due to surface finish is smaller than the distance between the grid points (which it almost certainly would be), it would be neglected in the computation. I can elaborate if this is still unclear.

      Now, for the "yes" answer. Most CFD turbulence models solve equations for both the turbulent kinetic energy and one of either the turbulent dissipation rate or the specific turbulent dissipation rate. Some studies have shown that you can emulate surface roughness effects by specifying particular values of the specific dissipation rate along solid, viscous surfaces. I don't have any experience with that, so I can't comment on it's success first hand.

      I can say that in most industries the effect of surface finish is measured experimentally. Once you get to "high enough" velocities and reasonable surfaces finishes things quickly become "hydraulically smooth", meaning surface roughness effects stop being important.

    23. 05-04-2014 01:22 AM #25
      So, now on to my "study". Let me preface this by explaining my purpose behind the work. I have been writing, developing and validating my own CFD code for the past few years. I use the code for automotive design work and to further my understanding of CFD/aerodynamics (which is also my full time profession). I am gearing up to design a new intake manifold for our company (JDL Auto Design) and in doing so I wanted to run test cases through my code that I could compare back against "known" data. One really well validated case is pressure loss vs. entrance radius/pipe diameter [see reference below, or find a similar plot in ANY fluid mechanics text book].



      So, I created grids and analyzed flow for a constant radius inlet of 0.090", 0.188", 0.375" and 0.750" on a 2.00" diameter pipe. Shown below are contours of velocity magnitude on a scale of 0-150 meters per second. All cases are ran to the same inlet to exit pressure ratio.

      0.090" / 2.00" = 0.045



      0.188" / 2.00" = 0.094



      0.375" / 2.00" = 0.188



      0.750" / 2.00" = 0.375



      I'll be brief with my explanation of the physics...

      Surface static pressure variations are caused by (among other things) local changes in surface curvature. Convex surfaces create low pressure regions/accelerating flows and concave shapes create high(er) pressure regions/decelerating flows. Surface curvature is inversely proportional to radius, so really tight radius inlets have really high surface curvature. This causes the flow to rapidly accelerate to high velocities, which by itself is fine. The problem is, when this curvature is removed, the pressure returns to normal (rises), so the flow must decelerate. If the deceleration is too abrupt, the flow seperates and large pressure losses follow. Two plots follow below:

      [1] My CFD results are plotted against three different textbook references I found. You can see that the absolute level of loss is different between the cfd and most of the textbook data. There are many reasons for this. If my inlet and exit domain are different than where the measurements were taken, I may be accounting for additional loss that was not present in the measurements. Or, most likely, the test data was done using a different fluid at a different velocity and viscosity than I simulated, so that can definitely shift the results. MOST IMPORTANTLY, the cfd captures the sharp increase in loss going below a r/d ratio of 0.10. The cfd also recognizes that there is little/no benefit to increasing r/d beyond 0.20. Both of these conclusions agree closely with the data sets I found.




      [2] Surface "ideal" Mach Number distributions are plotted for all the cases I ran. You can clearly see that the green and purple dots (0.188" and 0.090") accelerate over a short distance to the highest Mach Numbers and then must decelerate. You can also see that the red and blue dots (0.375" and 0.750") show similar levels of acceleration and peak Mach Number, so based on this it's not surprising that there isn't a benefit between the two radius'.



      I typically don't share stuff like this, but I welcome questions for those that are interested...

    24. 05-04-2014 02:22 AM #26
      Well I've been patiently waiting your write up and I must say I wasn't expecting that much. I appreciate the effort you have put into this brief summary of your work. While most of it is over my head, (I'm a mechanical engineering student but math isn't my strong point) I can use this information. I'm using my very limited and crude understanding of aerodynamics and thermodynamics to design this intake manifold, intercooler and exhaust manifold.

      I'd love to hear your opinion on the work I've put into this manifold so far and if you think I'm headed in the right direction. Ill post some rough sketches tomorrow of the plenum design.

    25. 05-04-2014 01:01 PM #27
      Quote Originally Posted by leebro61 View Post
      So, now on to my "study". Let me preface this by explaining my purpose behind the work. I have been writing, developing and validating my own CFD code for the past few years. I use the code for automotive design work and to further my understanding of CFD/aerodynamics (which is also my full time profession). I am gearing up to design a new intake manifold for our company (JDL Auto Design) and in doing so I wanted to run test cases through my code that I could compare back against "known" data. One really well validated case is pressure loss vs. entrance radius/pipe diameter [see reference below, or find a similar plot in ANY fluid mechanics text book].



      So, I created grids and analyzed flow for a constant radius inlet of 0.090", 0.188", 0.375" and 0.750" on a 2.00" diameter pipe. Shown below are contours of velocity magnitude on a scale of 0-150 meters per second. All cases are ran to the same inlet to exit pressure ratio.

      0.090" / 2.00" = 0.045



      0.188" / 2.00" = 0.094



      0.375" / 2.00" = 0.188



      0.750" / 2.00" = 0.375



      I'll be brief with my explanation of the physics...

      Surface static pressure variations are caused by (among other things) local changes in surface curvature. Convex surfaces create low pressure regions/accelerating flows and concave shapes create high(er) pressure regions/decelerating flows. Surface curvature is inversely proportional to radius, so really tight radius inlets have really high surface curvature. This causes the flow to rapidly accelerate to high velocities, which by itself is fine. The problem is, when this curvature is removed, the pressure returns to normal (rises), so the flow must decelerate. If the deceleration is too abrupt, the flow seperates and large pressure losses follow. Two plots follow below:

      [1] My CFD results are plotted against three different textbook references I found. You can see that the absolute level of loss is different between the cfd and most of the textbook data. There are many reasons for this. If my inlet and exit domain are different than where the measurements were taken, I may be accounting for additional loss that was not present in the measurements. Or, most likely, the test data was done using a different fluid at a different velocity and viscosity than I simulated, so that can definitely shift the results. MOST IMPORTANTLY, the cfd captures the sharp increase in loss going below a r/d ratio of 0.10. The cfd also recognizes that there is little/no benefit to increasing r/d beyond 0.20. Both of these conclusions agree closely with the data sets I found.




      [2] Surface "ideal" Mach Number distributions are plotted for all the cases I ran. You can clearly see that the green and purple dots (0.188" and 0.090") accelerate over a short distance to the highest Mach Numbers and then must decelerate. You can also see that the red and blue dots (0.375" and 0.750") show similar levels of acceleration and peak Mach Number, so based on this it's not surprising that there isn't a benefit between the two radius'.



      I typically don't share stuff like this, but I welcome questions for those that are interested...
      I owe you a beer. Thank you, time to break out the calculator

    26. Member vwking's Avatar
      Join Date
      Dec 14th, 2001
      Location
      NORTH JERSEY
      Posts
      1,277
      Vehicles
      too many
      05-07-2014 02:17 PM #28
      i love your work i want one

    Posting Permissions

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