Audi A4R4 450HP CBEA (CR) Build

Rub87 said:

LP turbine WGV is only way to regulate the boost once the HP bypass is open.

Click to expand...

Unless the LP is a VGT too...

Eager for updates, Louis!

Updates coming this weekend!

I have been doing a lot of design work over the last two weeks. The goal was to create a hardware package that would operate similar to the image Dave posted earlier. We had a lot of back and forth and are both fairly happy with the final design results. There are still a few assumptions on the table that will get ironed out as I move forward with actual fabrication. I have an assortment of material here already but still have a few more laser profiles to get cut. I will also be using some technology for turning the pipe copes and holes into a flat layout, printing on paper, then wrapping on pipe to create accurate cut lines.
The first layout I created was the easiest to model and also the easiest to fabricate. It also had some fairly large volumes in both the exhaust side and compressor side. Hear you will see the LP is sitting behind the engine and there is a an expanding log style manifold with valve on end, while the HP turbo is mounted tight on the side in a traditional position. For clarification, the HP is the GTD2060VK turbo, and the LP is a GTA4082V. The HP needs modifications to the turbine inlet and outlet. The LP turbo needs modification to the compressor housing.




I will not show all the iterations we went through and cut straight to the final design we have arrived at. While the HP turbo is a bit further from the manifold, we have much smaller volumes in this system. We have also grouped 1-4 and 2-3 primaries together prior to entry into a central plenum. This plenum allows for full 3" feed direct to the LP turbo, while the HP is an efficient branch off the top.






The heart of collecting the exhaust gases and sending them in the right direction is the manifold/plenum design. We have minimized the volumes as much as possible and have tried to keep flows moving in one direction (no reversals).







.

Awesome work, Louis! People cannot get an appreciation of how difficult it is to package the two turbos efficiently. The 2060 is already a pretty decent size turbo on its own right; a 4082 takes it to a whole other level, weighing it at 60 pounds alone, requiring its own mount to the chassis! The pictures don't adequately show that the turbo itself is about the length of the whole engine block!

We tried to package the plumbing with as short, direct gas flow paths as possible, while still being able to fabricate reasonably easily, and also incorporate 2 bypass valves, and separate water-to-air interstage and aftercoolers. Compromises had to be made, but I for one am extremely happy and proud of the work Louis has done!

Thank you Dave.

Yes, the GTA40 does look a little ridiculous next to the engine. lol The additional motor mount will definitely be handy. I will spend a fair bit of time designing a new passenger side engine mount that will reach over and pick-up the LP. Because of the larger moment arm, as you mentioned, the extra motor mount will be positioned as part of this frame.

The good news is that I have another 8 days to work on this stuff (not full time, but pretty close).

cmon guys. do some effort for the packaging.



too bad this design makes for the flange at the head to warp all the time. initially I had the slide piece in the middle but it was too leaky. then I welded it and now it warps. I need to find some short compensator, ideally from BMW x40i single turbo 6 inline engine.

That's a nice plate of spaghetti there, Ruben! What two turbos are you using? We looked at every way we could package the GTA4082V closer, but maybe you'd like to try. We're also constrained by using flapper bypass valves already on hand that is basically a fixed 3" ID X 2.5" long cylinder that has to be packaged inline, and when you're using 3" diameter Schedule 10 in places, you are limited to minimum bend radii.

We could maybe rework the packaging a little tighter on the exhaust side using Louis's initial log manifold that is very similar to yours, but it would have meant a lot more total length, volume and convoluted flow paths when considering both the hot and cold sides and the two charge coolers as a whole system.

gt1444v and efr7064. Now I dont really know what do. my time is so limited and the effort to take turbos off and try to find a solution for the flange is quite big
the efr works totally nice. 1400 kg/h of air at 5400 rpm with boost = emp, low end makes 2 bar boost at 1700 rpm in higher gears

Adapting somehow the setup from 313 bitdi isn't ok? Too complicated to control it?

This is for 2.0, not for 3.0 tdi. Also the 3.0 LP turbo is a gt3067 which is pretty small, topping out at 410-420 hp/1350 kg/h

Rub87 said:

cmon guys. do some effort for the packaging.



too bad this design makes for the flange at the head to warp all the time. initially I had the slide piece in the middle but it was too leaky. then I welded it and now it warps. I need to find some short compensator, ideally from BMW x40i single turbo 6 inline engine.

Click to expand...

Perhaps the trick is to use thicker stock material and mill it for tighter fit, rather than tubing. There is a manifold with this slide connection made for Cummins 5.9. I have never heard of it leaking. I haven't seen the one on the BMW but might be a good option for adaption.

As for packaging, yes, this is 3" pipe feeding to LP and outlet from HP is also 3". If I choke down flow with smaller diameter pipes and valves like yours then I can package it a lot tighter. I do like the use of the flexible couplings in your manifold design.

I am running a log style manifold very similar to the design you show above on my GBT2056V/CR engine. I have not had any issues with warping and leaking and that is with a hot tune. One would think the flexible coupling you installed would alleviate most of your issues.?

we've had 2 failures with these flexible joints on the V6 biturbo project. Maybe its not great quality, for sure they broke.

Macradiators.com said:

we've had 2 failures with these flexible joints on the V6 biturbo project. Maybe its not great quality, for sure they broke.

Click to expand...

good to know. I don't think I am going to change anything at this point. If I have problems, then I will...but I am doubtful I will. I have not had an issue with a single manifold I have built to date. EGT is not going to be any hotter on this build.

TDIsyncro said:

This is what I was thinking about for inputs. Does Raspberry Pi set-up seem like a reasonable approach? Or do you have a different platform in mind?



Is the primary reason for the LP turbine waste gate to reduce the restriction caused by the LP turbine when running primarily in the HP turbo range? Or was it intended as a bypass for big power numbers? I am not planning to be exceeding the LP turbine capacity this time but I can see the possible benefits when trying to retain HP turbo response.

Click to expand...

From some experience - LP turbine's, especially of your sizing are not that restrictive at the flow rates when you are most concerned building boost quickly with the HP side. Once the flows are Higher and the LP turbine restriction would actually make a difference to HP turbine power you want some boost from your LP turbo anyway.

m1ketdi said:

From some experience - LP turbine's, especially of your sizing are not that restrictive at the flow rates when you are most concerned building boost quickly with the HP side. Once the flows are Higher and the LP turbine restriction would actually make a difference to HP turbine power you want some boost from your LP turbo anyway.

Click to expand...

I appreciate your comments.

TDIsyncro said:

Perhaps the trick is to use thicker stock material and mill it for tighter fit, rather than tubing. There is a manifold with this slide connection made for Cummins 5.9. I have never heard of it leaking. I haven't seen the one on the BMW but might be a good option for adaption.

As for packaging, yes, this is 3" pipe feeding to LP and outlet from HP is also 3". If I choke down flow with smaller diameter pipes and valves like yours then I can package it a lot tighter. I do like the use of the flexible couplings in your manifold design.

I am running a log style manifold very similar to the design you show above on my GBT2056V/CR engine. I have not had any issues with warping and leaking and that is with a hot tune. One would think the flexible coupling you installed would alleviate most of your issues.?

Click to expand...

imo 3' is widely too big, for reference. with a 3' exhaust on a 3.0 tdi doing 450hp/1450 kg/h you measure like under 50mbar pressure after turbine.

this is with the gas at lets say 650-700 degC and ambient pressure and lets say 4 meters pipe with some bends

If you will be able to run this kind of airlow on a 2.0, I would suspect the pressure before turbine to be around 4500 mbar abs at minimum, ok, it might be 950 degC. but the air density, and this space it requires will be ~4 times smaller.
I know the example is overly simplistic. but I think you get the point. this is why on a small displacement/high boost engine, you come away with very small intercooler piping, primary runners, throttle body, etc. the only area where you need to use NA logic for sizing is at the piping pre compressor and post turbine. any pipe that has pressurized gas in is, can be way smaller to flow same massflow at fix pressure drop

the manifold I designed was aimed to have the piping 'free' as in not used to support the wieght of either turbo. the flex joints will also live for long if you use them for what are intended, which is to compensate for thermal only expansion/movement. this kind of joint will fail soon if they have too much movement, as for example caused by engine vibrations.

I started with a 12mm plate. then after inital surfacing it was like 11mm. then after I had welded the slip joint I had to straighten it again. now its below 10mm left :/

Ruben, I will explain why I have the main 3" pipe. If you have any thoughts on the issues that favoured me in that direction I would appreciate it. Firstly, my valve is a 3" valve. Second, the hp turbine outlet size tapers out to a 3" pipe at the connection point. Third, the inlet to the LP turbo readily translates into a 3" pipe.. Ie - it takes very little press and die grind work to be a good transition. Keeping everything tight, the 1R elbow off the valve is part of the merge between the hp and LP. If I taper down the hp outlet, it takes length and volume. Smaller pipe from valve would have to be transitioned up in size to match LP inlet. There is a lot going on in that small space. If outlet from hp turbine was a 2" then it would inspire me to make the main pipe to LP smaller. Perhaps also then reducing the valve size.

Much ado about nothing Just keep up the good work, Louis! Head up and not too much time online, it can really bum you out.

TDIMeister said:

Much ado about nothing Just keep up the good work, Louis! Head up and not too much time online, it can really bum you out.

Click to expand...

I have this big pipe, and I know I will use it.

haha.

its not because the housing of HP turbine tapers to 3' you need to use a 3' pipe. afaik a gtb20 or 22 turbine is stock wise only max 46mm exducer, and yes the housing expands as normallly downstream turbine you have near ambient pressure. in you case there will rarely be ambient pressure there at all. so lets say a 60 or 63mm pipe will be more than plenty.

also the inlet of the LP turbine is 3', does not mean that there need to be 2 massive 3' pipe needed to go there. because you will never has all massflow coming either through the HP or the HP bypass alone. it will always come from both at rated power.

imo 3' is lots of heat loss and dead volume that will do nothing but make things slower to respond

Rub87 said:

haha.

its not because the housing of HP turbine tapers to 3' you need to use a 3' pipe. afaik a gtb20 or 22 turbine is stock wise only max 46mm exducer, and yes the housing expands as normallly downstream turbine you have near ambient pressure. in you case there will rarely be ambient pressure there at all. so lets say a 60 or 63mm pipe will be more than plenty.

Click to expand...

Are you suggesting I do this:




If I have to put an extra reducer in there then I end up with same volume as the 3" pipe.

just make the pipe extend past the flange so it somewhere mates on the cone. or do like I did in my HP turbo I just cut the housing off so the cone ID matches the pipe ID I planned to use.

Rub87 said:

just make the pipe extend past the flange so it somewhere mates on the cone. or do like I did in my HP turbo I just cut the housing off so the cone ID matches the pipe ID I planned to use.

Click to expand...

ok. That is simple and good idea

OK, this is 60mm pipe extended into cone, 1R elbows. 30deg merge into collector at LP inlet. It reduced overall length of collector by 40mm so good move. Now, if you have a more compact valve for me I can compressor plenum and collector.

So after consideration of Rubens comments and some further comments from Dave, I have spent a bunch more time on 6 more iterations of the design. The plenum on the manifold has now been centralized and the primary runner lengths are more equal. The primaries are now reduced in size to 1-1/4" SCH10S stainless, which offers a 1.442' ID. The main length from plenum to LP turbo has been compressed a bit more in length and the elbow from control valve to LP inlet was reduced to 2-1/2". I did not update the boost piping but this is final design on hotside. Laser profiles and pipe materials are ordered.






The new plenum is shown with a simple fabrication jig. This jig will allow me to properly place the main elements relative to one another and the rest will fall in place fairly easily. I have looked closely on tolerances of laser profiles and jig parts so expect things to fit very nicely. Everything is in 316L stainless except the main flange mounting to head which is made from 5/8" thick mild steel. I had this flange piece cut already and it fits really nice.

I have also started the modifications required on the turbo chargers.

The LP turbo is better suited with a radial orientation on the compressor outlet, not the front pointing outlet it came with. It was a fairly easy job to cut the neck of the turbo, remove the barbed nipple, and re-weld it back on compressor housing. Die grinder blending was then performed on the inside to allow for smooth flow.



Here is the turbo reassembled after being cleaned.

The HP turbo turbine has this beautiful heat shielding and ceramic insulation. However, the inlet geometry is not very good for my purposes.



I have removed the heat shield and insulation, hopefully I will be able to reinstall it after modification.



The existing inlet must be removed.





Here is the new inlet starting to be fabricated. I was originally going to use a standard GTB flange set I made but decided to make a custom set that better fits the tubing I am using. Bolt up will be much more pleasant.



I am quiet happy how it is turning out so far. The inside surface is already blended together vary nicely.

please try to put the heatshielding back on every bit of heat helps to keep turbos alive and the paint on the hood.

I made this many years ago for a 1.9. 2260vzk with similar shielding.

Rub87 said:

please try to put the heatshielding back on every bit of heat helps to keep turbos alive and the paint on the hood.

I made this many years ago for a 1.9. 2260vzk with similar shielding.

Click to expand...

Reinstall of the shielding looks great. I image the dust coming from the insulation is very unhealthy if breathed in. When I was removing it I wore a respirator. The slightest flex of it would send a tiny puff of microscope particles in air.
It looks like you welded am extra tab on the housing for extra support point. Is this correct? I am also curious what is going on with the oil feed. It appears to enter from the top and side of the turbo body. Did you make internal passage in the plate?

Oil feed and drain on this turbo is bottom side. My assumption is small hole is for feed, large hole is drain, and ignore strange fitting on other side of drain hole.