Welding twin entry manifolds

[Crucible]

Whats the correct way to merge and weld twin entry/split pulse manifolds?
I have seen 4a turbo manifolds where 1-4 then 2-3 mege into the separate dividers and a differant manifold/turbo on differant 4agte where they ran 1-3 then 2-4.

I would of thought itl go by firing order, for example 1342 you would run 1-4 then 3-2 seperate, because you will get equal pulsing. pulse on one side then next exhaust pulse would be on opposite side.

any thoughts?

[gasman]

1-4 and 2-3 to get the equal pulsing.

like you assumed.

[Crucible]

ok, thats what I thought, depends entirely on firing order.

so for example on 6cyl with firing order 153624, you would have 1-3-2 and 5-6-4 separate? and also keep pipes running to wastegate totally separate with a divider where they merge..

[Crucible]

can anyone confirm that they should be made to suit firing order as Ive mentioned above?

have asked a few guys in the know and it has always been a vague area, so just wanting to confirm for future referance

[touge_ae101]

yeah if you google it all the exhaust/header makers say they should be collected according to the firing order.

[TRD_ZERO]

What happens if the ports are matched 1-3, 2-4? Will if be worse off than a single scroll turbo?

[touge_ae101]

from this article here: http://www.ssheaders.com/header.htm

The optimum cure to this whole problem is to correct the exhaust firing cycle. The two cylinders that fire in succession into each collector have to be separated. This can be done partially by a "Tri-Y" header, where the four primary tubes from each bank merge into two secondary tubes (separating the two pulses firing in succession) and finally collect into a single collector. This type of header helps, but the two pulses are still coming back together at the collector.

The second optimum cure is to cross the two center tubes from each bank, across the engine running them into the collector on the opposite side. This makes the firing cycle in each collector 180 degrees apart, the same as a four cylinder engine. Once this firing order is achieved, the small collector outlet diameter can be used and the "High Velocity Scavenging" at low R.P.M.s cures the reversion problems and eliminates the need for extreme cam duration.

mainly has reference to V8's but the basics of the collection is still the same.

there has got a better article on google somewhere that relates to 4 cyl engines and is very helpful.

also there is a supercharger header calculator on the interweb that calculates what power you can get out of a specific sized header and diff lengths of primaries and secondaries. for both 4-1 and 4-2-1.

hope this helps.

[Crucible]

^^ i read that article, isnt that more related to scavenging though?

there has got a better article on google somewhere that relates to 4 cyl engines and is very helpful.

have done a few searches on google, there seems to be lots of discussion but they all have differant theorys. I guess as a rule on 4,6,8 cylinder engines that keeping the next firing cyl in opposite divider is the way to go then.

it makes sense but hmmm I dunno?

[evil_si]

i normally do my 4 cyl twin scroll manifolds in firing order,
1 & 3 together and 2 & 4

[Crucible]

Link to good article from garret site. article on split pulse about half way down the page.

http://www.turbobygarrett.com/turbobyga ... ch102.html

Seems that its the exhaust pressure from a cylinder beginning its exhaust cycle to another in overlap you try and avoid as thats pressure & heat loss wasted instead of it all going through the exhaust turbine.

makes sense.

i normally do my 4 cyl twin scroll manifolds in firing order,
1 & 3 together and 2 & 4

I think that would be fine arcording to this as when No1 is beginning exhaust at BDC No 3 cylinder valves would be closed as on Comp/power stroke so wouldnt affect the pulse. you just wouldnt merge 1-2 then 3-4 together, correct me if Im wrong

I dunno, how much differance does it really make lol

[evil_si]

there are many conflicting articles out there,


on the 2j altezza the manifolds are 123 456 and each feed a seperate turbo,
being such big turbos we had trouble with them fighting each other,
but I soon resolved that with a balance pipe between the manifolds.

[Bazda]

1-4 & 2-3 on a 4 cylinder.

So you get a pulse into each scroll sequentially.

If you do 1-3, 2-4 you will get 2 pulses into 1 side of the scroll at a time.
Therefore not efficient and will increase the spool time.
So many times i've seen people make them like this and wonder why they dont get good spool running the twinscroll turbos with big a/r's.

Look at all the full race manifolds made for the honda guys they are 1-4 & 2-3.

[XSVWGN]

1-4 & 2-3 on a 4 cylinder.

So you get a pulse into each scroll sequentially.

If you do 1-3, 2-4 you will get 2 pulses into 1 side of the scroll at a time.
Therefore not efficient and will increase the spool time.
So many times i've seen people make them like this and wonder why they dont get good spool running the twinscroll turbos with big a/r's.

Look at all the full race manifolds made for the honda guys they are 1-4 & 2-3.

hmm well given that simon (evil_si) built my manifold the way he does i should expect average spool ...


Matt

[Bazda]

1-4 & 2-3 on a 4 cylinder.

So you get a pulse into each scroll sequentially.

If you do 1-3, 2-4 you will get 2 pulses into 1 side of the scroll at a time.
Therefore not efficient and will increase the spool time.
So many times i've seen people make them like this and wonder why they dont get good spool running the twinscroll turbos with big a/r's.

Look at all the full race manifolds made for the honda guys they are 1-4 & 2-3.

hmm well given that simon (evil_si) built my manifold the way he does i should expect average spool ...


Matt

You will have to wait and see, but im guessing spool wont be optimised. What turbo are you running?

The guy who now welds all my stuff, made all of zohabs evo basically etc, he hadn't made them in ages and when he went to do my mates he thought of the firing order and made my mates ones 1-3 2-4.
Had a big discussion with my dad and another turbo guru and found out it was definatly wrong.

Again this is for turbos, for N/A applications its totally different.

Anyhow think of it like Kyaking, you dont put the paddle in on 1 side twice then the other side twice to go fowards fast.

[XSVWGN]

running an evo7 td05hR

Matt

021 297 1754

[sergei]

Factory ST205 manifold has got the following configuration:
1-4
2-3

Here:

[Crucible]

1-4 & 2-3 on a 4 cylinder.

So you get a pulse into each scroll sequentially.

If you do 1-3, 2-4 you will get 2 pulses into 1 side of the scroll at a time.

This is the theory Ive always gone by.

But according to the garrett article the main aim is keeping the two opposing cylinders which both have exhaust valves open "together" at the "same time" for a short period seperate.

When the exhaust valves on each cylinder are "both" open the pulse from the cylinder beginning exhaust is lost into the cylinder just ending. you want the full energy of the pulse through the turbine, thats why they are divided.

so if this is the main concern, having 1-3 and 2-4 wouldnt be a problem. maybe you are getting two pulses on one divider then a following two on the other but the cylinders are not crossing "exhaust paths" so full pulse pressure is still entering the turbine.

If the firing order was differant on a 4 cyl the grouping would be differant again.

[frost]

factory ct26 is 1-4 2-3
factory td05hr is 1-4 2-3
seems silly for two main car makers to get it wrong,

[sergei]

Most articles about this stuff are for V8s. V8s have completely different timing to an I4, which means their grouping will be different as well.

[iOnic]

The idea is just to keep spent exhaust gases from passing into an adjacent cylinder that's inhaling fresh air during overlap.

Having pulses go into "one side then the other" isn't the main reason behind the better performance. In fact you'll be doing well to measure any noticeable difference between having "2 pulses per side" over "1 pulse per side".

They work well because the exhaust housing has 2 smaller entries instead of 1 big entry. This keeps the velocity and pressure of the exhaust gas up right up to the point where it hits the turbine wheel blades. This gives more energy on the turbine wheel which = quicker spool. Single scroll exhaust housing = pressure drop right before the turbine wheel. I always make mine 1-4/2-3 but I've never noticed any measurable difference if you go 1-3/2-4.