Twin Scroll Turbo

[gmreider]

Dennis , notice you are going from 3 inch downpipe to a 2 1/2 inch exhaust system. I have debated 3 inch all the way back vs. stepping down to 2 1/2 inch after the down pipe. Your thoughts and opinions Please??
Your work is exquisite !! One thing I must say is : it won't be fun bolting all that exhaust manifold on and off. Lots of tight spaces.

[squaab99t]

What are the current lengths of the runners? Looks good, BTW.

Thanks Jon, that is quite a complement coming from you. I remember back in the day your fab run on c900 headers.
Burnstainless has given me a target length of xx". Cyl 4 I have that and I am making that the driver shape since it is the closest to the turbo. aka most difficult to get that length packed in. For reference: OEM log header the distance 1 to the turbo flange is 14". Cyl 4 to the flange is about 3-4"

Not currently set on a number for 2 and 3. Need to do some math with the speed of sound in the equation. It will be some ratio"

[squaab99t]

Dennis , notice you are going from 3 inch downpipe to a 2 1/2 inch exhaust system. I have debated 3 inch all the way back vs. stepping down to 2 1/2 inch after the down pipe. Your thoughts and opinions Please??
Your work is exquisite !! One thing I must say is : it won't be fun bolting all that exhaust manifold on and off. Lots of tight spaces.

IMO 3" all the way back is overkill. I ran a 2 1/2" DP on my old setup and made 305WHP, so to me that is not the choke point. I'm using the 3" DP this time is because that is the V band Dia that is coming off the turbine housing. Now, will the 3" post the turbo help the spool? I'm thinking so as it will allow a greater volume to exit the turbo. So why not 3" all the way? The exhaust gases' heat will be harvested mainly by the turbine, post that it will loose even more heat via conduction and radiation via the downpipe itself. As the heat drops, the pressure will drop, thus not requiring the "extra" volume a 3" dia would provide. At least that is my reasoning...

"exquisite" have not heard that expression in a while. Thank you. Practice, practice, and more practice. Research too; which filler rod to use, argon flow rate and back purging. I believe the back purging helps on the front side too. The metal is not getting sucked away and keeps the puddle happy.

[KoogiA17]

This is fascinating stuff.

You are pairing the complementary cylinders to each side of the twin scroll. With the twin scroll your hoping to see benefits with lower cylinder temps, EG temps, leaner air fuel ratio and ultimately greater turbine efficiency?

Interesting to compare the different approaches one can take when designing a exhaust manifold. I'm not going to be using a twin scroll turbo in my 99 which is one major difference. Am focusing on minimising the number or bends in each pipe and therefore attempting to minimise energy losses due to turbulence rather than trying to achieve equal length runners.

[jdwertz]

Aluminum foil for back purging?

[squaab99t]

Aluminum foil for back purging?

Yes, works pretty well for one off shapes. Conforms well and you can peel the area you are about to weld. The other option is masking tape on the non hot end. Punch holes in it to bleed off. The input I use a filter to dispurse the argon.

[Jon1]

Burnstainless has given me a target length of 27". Cyl 4 I have at 27" and I am making that the driver shape since it is the closest to the turbo. aka most difficult to get that length packed in. For reference: OEM log header the distance 1 to the turbo flange is 14". Cyl 4 to the flange is about 3-4"

The headers that I built are 16" from head flange to collector, for your reference. Did Burnsstainless give you any documentation to support the 27" length? Is that for turbo only?
thanks,

[squaab99t]

Burnstainless has given me a target length. Cyl 4 I have at, and I am making that the driver shape since it is the closest to the turbo. aka most difficult to get that length packed in. For reference: OEM log header the distance 1 to the turbo flange is 14". Cyl 4 to the flange is about 3-4"

The headers that I built are 16" from head flange to collector, for your reference. Did Burnsstainless give you any documentation to support the length? Is that for turbo only?
thanks,

Burns did not supply a print out since he just ran the numbers in their program while I was on the phone with their head engineer while I ordered. He asked me a series of questions and out came the target length for my application. I'm pretty sure was for a turbo app since a n/a you would want to go with a tri Y setup to make the most usable power.

[KoogiA17]

Burnstainless has given me a target length of 27". Cyl 4 I have at 27" and I am making that the driver shape since it is the closest to the turbo. aka most difficult to get that length packed in. For reference: OEM log header the distance 1 to the turbo flange is 14". Cyl 4 to the flange is about 3-4"

The headers that I built are 16" from head flange to collector, for your reference. Did Burnsstainless give you any documentation to support the 27" length? Is that for turbo only?
thanks,

Burns did not supply a print out since he just ran the numbers in their program while I was on the phone with their head engineer while I ordered. He asked me a series of questions and out came the target length for my application. I'm pretty sure 27" was for a turbo app since a n/a you would want to go with a tri Y setup to make the most usable power.

The calcs for my application give a length of around 30". Header length determination is not an exact science though. Had to design a new turbo support bracket as I couldn't get a design to work with the current one.

[squaab99t]

The calcs for my application give a length of around 30". Header length determination is not an exact science though. Had to design a new turbo support bracket as I couldn't get a design to work with the current one.

30"? Well that adds a bit more confidence to Burns' number.
You are correct. You most likely will not nail it straight away with just calcs but it will get you in the ballpark. From there Dyno testing will be required to prove or disprove changes in primary or secondary length. For the hobby / enthusiasts this likely is a little over the top.

Do you have a pic of the design or what you have accomplished? Nice to share ideas and thoughts. I'm and sure others are interested. Maybe in your own thread or just a snip it in this one?

[KoogiA17]

If only everyone had their own engine dyno at home! Just think of the fumes and cacophony My neighbours would love me, as obviously I would have to get in the middle of the night just to run the engine dyno.

Have modelled the manifold in solidworks and have got a render somewhere, will have to dig it out. As I said above, my approach is to try and minimise turbulence and hence energy loss before the exhaust gases arrive at the turbo so have focused on reduce the number of bends rather than attempting to achieve similar length primaries. At the moment planning on using a BorgWarner EFR Turbo with internal wastegate.

[squaab99t]

If only everyone had their own engine dyno at home! Just think of the fumes and cacophony My neighbours would love me, as obviously I would have to get in the middle of the night just to run

Well you don't have to own one, but have access to one with a good Dyno operator and cold hard cash.
http://www.saabrally.com/forums/viewtopic.php?f=2&t=3285&p=40177&hilit=Header#p40177
The Dyno operator is an ex nascar engine builder of who has ran 100 hp to 1600 hp engines. The trick is to have your shit together to do the tuning and operator runs the Dyno to do back to back repeatable runs for real comparisons. Ended up doing 2 separate 8hr sessions.

[KoogiA17]

If only everyone had their own engine dyno at home! Just think of the fumes and cacophony My neighbours would love me, as obviously I would have to get in the middle of the night just to run

Well you don't have to own one, but have access to one with a good Dyno operator and cold hard cash.
http://www.saabrally.com/forums/viewtopic.php?f=2&t=3285&p=40177&hilit=Header#p40177
The Dyno operator is an ex nascar engine builder of who has ran 100 hp to 1600 hp engines. The trick is to have your shit together to do the tuning and operator runs the Dyno to do back to back repeatable runs for real comparisons. Ended up doing 2 separate 8hr sessions.

Enjoyed the thread you linked to above.

Unfortunately, I don't know a friendy dyno owner.

Nice to see your using double slip collectors. What is your thinking behind utilising an exhaust manifold flange which isn't 1 single piece?

[squaab99t]

Nice to see your using double slip collectors. What is your thinking behind utilizing an exhaust manifold flange which isn't 1 single piece?

A couple things come to mind. In no particular order:
1) The head and flange/runners can independently move with thermal contraction and expansion. This should put less stress on the welds
2) As you tack the pieces in place you can remove the adjacent runners getting them out of the way to get a better shot at placing those tacks evenly around the tube.
3) Likewise it will give better access while welding on the bench. You won't be cock blocked by the neighboring tube, keeping that torch angle right.
4) Having individual primaries will help with install and removal. Free up and provide some wiggle room to get that 10 lbs into the 5 lbs sack. As you pointed out I will be using double slips on the collector, so the tubes can be independent. This will work, only if the turbo is supported with the hanger and not solely by the manifold itself.

I am actually plan on cutting 2-3 apart on center-line of the common bolt hole. Custom "fender" washer or shoulder washer to join/clamp back up.

Full Race has been around awhile. As you can see they purposely cut the flange into 4 separate runners

[Geoff]

When you're done with this job, instead of the turbo being an engine accessory, the engine is going to start looking like an accessory for the turbo.

You won't be cock blocked by the neighboring tube