Latest Project - Headers

[RarlyL8]

Thank you Goran and yes the collectors will undergo some major changes. Design of the primaries was the focus for the header pictured. That collector is a starting point and used to locate the distal bundle of primaries. The end product will be a merge collector with a reduction cone that transitions the secondaries.
All the while we must keep cost in mind. In the end cost will be the compromise but if we can retain 90+% of the design efficiency while meeting the price point the project will be considered a sucess.
Speaking of secondaries, for a sub 375WHP CIS 930, what would be the optimum size OD for the secondary pipes in terms of velocity? For the sake of math the secondary pipes will be equal length 12-14" long each and boost would be at 1.0bar by 2600rpm through 6500rpm. (I have a number but would like to verify it with a few experts).

[beepbeep]

Originally Posted by RarlyL8

Thank you Goran and yes the collectors will undergo some major changes. Design of the primaries was the focus for the header pictured. That collector is a starting point and used to locate the distal bundle of primaries. The end product will be a merge collector with a reduction cone that transitions the secondaries.
All the while we must keep cost in mind. In the end cost will be the compromise but if we can retain 90+% of the design efficiency while meeting the price point the project will be considered a sucess.
Speaking of secondaries, for a sub 375WHP CIS 930, what would be the optimum size OD for the secondary pipes in terms of velocity? For the sake of math the secondary pipes will be equal length 12-14" long each and boost would be at 1.0bar by 2600rpm through 6500rpm. (I have a number but would like to verify it with a few experts).

Can't you use some pre-fabricated 3-in-1 collectors in order to make manufacturing easier?

Burns makes collectors but they are a bit pricey. Maybe you can give them a call and check if they can provide a discount?


http://www.burnsstainless.com/MergeCollectors/3into1base/3into1base.html



Regarding exhaust pulse velocity, you have two components: one is actual speed of sound in media (shock propagation speed) and other is actual media speed (how quickly media flows). Generally, I would keep headers at roughly same diameter as exhaust ports, maybe a tad bigger (in order to obtain "anti-reversion" lip).

Speed of sound in media is a bit tricky to calculate as it depends on pressure and EGT. Roughly, EGT of 800 deg. C will give you M1 at 6300 ft/sec at 1 bar. But then you also have a backpressure so speed of sound will be higher than that.

There are some good calculation tools around: http://hyperphysics.phy-astr.gsu.edu/hbase/sound/souspe3.html#c1

I believe that speed of media itself is more important in this case. It's a bit trickier to calculate as it will depend on plethora of things.

I believe it's best to read on existing research...like here:

http://www.automedia.com/Turbo_Exhaust_Headers/pht20020501te/2

Regards,

[DDDD]

I have been wondering how the +/- 1.5% deviation in length plays into the magnitude or duration of the pulses. Technically having no deviation at all would provide a stronger pulse but having a slight deviation will provide a longer pulse. Is it possible that the best solution is to have the pulses arrive 'almost' but not exactly at the same time to provide a longer pulse, or is the idea to simply have the biggest and most concentrated pulse possible.


I realize this is a pretty hypothetical question, just curious.

[Miguel Antonett]

You guys know your stuff... I luv this forum.

But keep in mind that pulsing is directly proportional to the rpms; therefor there is an ideal rpms range for optimal performance for every primary lengths you use. I believe that we should be looking for that "sweetspot" range; which might not be equal lenghts... to be able to get torque sooner and a wider range. M2c.

[RarlyL8]

I ordered a set of Burns collectors for a separate project. Are you sitting down?
$920 for a set of 2, no discount.
That is a pretty big budget buster. We'll make our own version of a merge collector.
Yes there is a ton of math and theory that results in an unanswerable question due to lack of hard numbers to plug into the various formulas. What I am shooting for is an increase in off-boost torque combined with quicker spool time. I really think this design is the right compromise. You must also consider the vastly different engine configurations that these headers will be a part of. Many different cam profiles, turbochargers, tuning, CIS vs EFI, boost levels, etc. etc. You cannot over-think it as that is a waste of time. It is a generalized approach.

[911st]

I am not and exhaust expert but have some thoughts to share.

I understand having a perfect merge collector is not critical on turbo motors. Further, the back pressure that will always be present with the turbo in place slows things down so normal aspirated principles may not be a good fit. Further, with the non overlap cams like most of us run on a turbo we are not going to benifit from scavenging anyway no mater what the design.

The goals for a turbo exhaust might better be though of as achiving minimum volume and minimum restrictions. The equal length design should be valid as it allows the exhaust to exit in an orderly fashion and thus minimize restriction. The draw back with equal length is it inceases volume some.

To me a big reasion for equal length headers besides reducing restriction is it helps equalize temps at each cylinder. Ever seen the video of a 935 log style header on a dyno. The shortest primary gets so much hotter and turns red first.

For a turbo race motor an equal length system is the best practices system.

Best pre boost response is probably better found in areas like compression, ideal AFR & ignition settings. Other approaches are very effective like keeping the compressor bypass valve open in reserve for acceleration and or converting the WUR to vacuum sensing so it reduces the metering plate restriction have been found to work well.

The only thing we can do with the exhaust to ingress boost onset is keep volume as small as possible which takes us back to a log style or unequal length header. Still, the 930 system is small in volume anyway (early US system being an exception) and small differences between systems is not going to make much of a difference in boost onset anyway.

The best.

[911st]

There are some exaust shots in this vid.

http://www.youtube.com/watch?v=wPfK_BispVo

[RarlyL8]

Yes reducing hot spots is one problem minimized by using equal length primaries. The hot spot in this design is at the merge collector which is exactly where you want it to be.
It is desired to gain more benefit from this system than just getting the pulse to the turbo as quickly as possible. With a street car you spend the majority of the time off boost. N/A theory does not directly apply but many side benefits are realized. I have kept the equal length primaries as short as possible while using gentle bends. Anxious to see how it performs when complete.

[911st]

I love what your are doing Brian.

I am not an exhaust expert, just a guy that loves this stuff and wants to learn more. I believe there has not been any economic choices for an equal length 930 system to date and it is about time for something like this.

I suspect trying to find improvements in pre boost operation on a 930 from the exhaust will not yield much if anything. From the first 930's everyone has been trying to figure ways to improve pre-boost operation and reduce lag. As known, the CIS air metering design and the very low compression are the main issues.

I suspect just building the best turbo exhaust possible (short tube equal length) will ensure nothing is sacrificed in pre boost operation. Boost onset may be a different story.

A more significant factor might be the primary and secondary tube size choices.

Do we want smaller tubes to minimize volume and encourage faster boost onset or larger tubes in support of more efficient operation at HP peak. Is it worth the extra 27% volume stepping up from a 1 1/2" tube to a 1 7/8" tube?

When thinking about turbo header volume, many do not seem to think about the WG side. Most the existing shorties systems add back a lot of volume to the header manifold in this area.

Keep up the good work.

[RarlyL8]

Yes there is more to be gained in the area of lag reduction if you look at tuning the ignition and AFRs. Turbo design has the most impact. Another huge area is intercooler shrouding and design. That is a must.
I have participated in engine builds that reduced lag to the point you could not detect it. The engine was "on" instantly! This was a single turbo 930.
Header design was just one aspect of that formula and the virtual elimination of lag was not possible without headers. The stock Euro pipes couldn't get us that last few hundred rpm's of boost threshold.
So headers are one small piece of the formula and if designing them you should optimize what you can.

[RarlyL8]

WhoooHoooo! I get to play with all the cool stuff.

These arrived today and will be used as part of a custom build for a fellow Pelicanhead. Should let that baby breath!

[Chuck Jones]

Pure mechanical art......those collectors are surgical!

[les_garten]

Originally Posted by RarlyL8

WhoooHoooo! I get to play with all the cool stuff.

These arrived today and will be used as part of a custom build for a fellow Pelicanhead. Should let that baby breath!

Tasty Bits!!

.

[Reaper930]

Now THAT'S art...

[RarlyL8]

Making progress. Here is some less expensive art:







Once again cudos to the amaizing welding of M&K.

Here is what we have so far. Stay tuned for building the secondary pipes ...

[gogomobil]

Brian, this setup looks very much like the Holcomb headers!
Brilliant!

[RarlyL8]

Yes, it is convergent evolution. There are so many design constraints that is is not surprising that I matched an existing design. A couple of years ago when I designed these I had never seen the Holcombe headers. Also don't know what the Holcombe collectors look like but I wouldn't think they are any better than these. The secondary pipe is not the same. Should have that completed and produced in a few weeks.

[Dai]

Awesome job Brian. Love your craftsmanship.

[Miguel Antonett]

Hey Brian, as offered to you before, let me know when you're ready for testing...

Keep up the good work and see you at the Dragon next weekend!

[RarlyL8]

Accolades for the craftmanship go to Ben. I supplied the design and some very bad spot welds, ha!

Yes, looking forward to the Dragon. Probably have to drive the truck and hop in with someone as my car is down for maintenance and I am quickly running out of time to finish and do the shake down.