Custom Tuned Turbo Exhaust Systems

[DonWG]

My exhaust system with the heat wrap removed. Surprising how well the material has held up after being wrap for over 3 years.
A good look at the T4 to T6 mounting flange adapter.

[DonWG]

The primaries are stepped starting with 1.625" and increasing to 1.75".

[DonWG]

A look down one of the runners.

[DonWG]

Big HKS 60mm wastegate mounting pad.

[DonWG]

Backside view of the Y-pipe junction. Those are an egt hookup and a quick disconnect hookup for a back pressure gauge.

[DonWG]

A view into the Y-pipe junction.

[DonWG]

Another overall view.

[Boost231]

wow, don those manifolds are a work of art. thanks for the pictures

[Sleeper]

Beautiful Donnie
Thanks for posting

Paul

[DonWG]

Thanks for the compliments.

The primaries are tuned equal length to within 1/4".
Some would think that long tube primaries would give a narrow powerband. On contrar. All of my engine specifications were configured together as a whole. Each component specification supporting another. With the resulting package, these pipes work very well from 4250 to 7800 rpm.

And they sound awesome!

[DonWG]

Does anyone know what the chemical is called for cleaning stainless?

[87natl]

The Hendrix Group - cleaning stainless steels

Originally Posted by DonWG

Does anyone know what the chemical is called for cleaning stainless?

simple cleaning with vineger or club soda but some good stuff on that site

[DonWG]

Originally Posted by 87natl

The Hendrix Group - cleaning stainless steels


simple cleaning with vineger or club soda but some good stuff on that site

Thank you, sir.

[slimtastic]

so what kinda back pressure to boost ratio do you get with those headers? with what housing/turbo?

also have you found a real tuning use for the one egt gauge? ive always been under the assumption that a W/B is used for actual a/f ratio and then you use egt's on each primary for individual cylinder correction( for removing lean/ rich cylinders)?

[DonWG]

Originally Posted by slimtastic

so what kinda back pressure to boost ratio do you get with those headers? with what housing/turbo?

also have you found a real tuning use for the one egt gauge? ive always been under the assumption that a W/B is used for actual a/f ratio and then you use egt's on each primary for individual cylinder correction( for removing lean/ rich cylinders)?

With the old turbo (T76) I had a BP/boost ratio of 1.5something:1. That was with a T4 turbine housing, Q trim, .96 a/r.
The new turbo is a FI91X. T6 split turbine housing, 1.15 a/r. I have not had the opportunity to get a BP/boost ratio yet. I'm going to be mounting an in-car camera to monitor gauges for me. I don't have everthing on the datalogger. I would imagine that with the extreme difference between the T4 and the T6 housing, I'm going to have much better than the past 1.5x number.

I read the plugs for individual cylinder temps. I'm sure it's not as accurate as an egt in each primary, and actually I've been thinking of trying such a system. When I first built the headers I had bought all the egt bungs to weld to the primaries, but have put off welding them in. The single egt at least lets me know if I'm in the ballpark. I primarily use the wide band sensor, plug readings, and timeslips to set the a/f and then check the egt as a secondary reading just to make sure I'm still in a safe egt zone. I don't use the egt as a primary reading.

[Dusty Bradford]

Nice work, especially in the merge area of the crossover.

Have you ever found any data that supports the idea that stepped, equal length headers are a benefit in a turbo application? With the backpressure I have thought there would be no or very little benefit.

[DonWG]

Originally Posted by Dusty Bradford

Nice work, especially in the merge area of the crossover.

Have you ever found any data that supports the idea that stepped, equal length headers are a benefit in a turbo application? With the backpressure I have thought there would be no or very little benefit.

In the studying I did when I was first putting together my engine configuration, I spent a lot of time looking at the high end turbo motorsports classes. Especially F1. Studying the F1 turbo era, I found that tuned length headers were very popular. In fact, I noticed that the era started out with more log style manifolds and evolved into tuned length. Now, these people have R&D resources at their disposal that I can't even begin to imagine. I would not dare to second guess their decisions to use tuned length headers. If they found there was power in that decision, then I was on the band wagon too. Another well known tuner, Steve Kinsler of Kinsler Fuel Injection has documented that a turbocharged engine reacts to cam and manifold tuning much the same way a naturally aspirated engine does. His quote on cam selection for a turbocharged engine is, "Don't choke that engine". Again, who am I to second guess his successful experiences. A lot of times you don't need to go through the expense to do your own testing to find answers to some tuning questions. Just look at what other highly successful people have done. And people that have much more resources for testing than a simple hobbiest racer.

On the question of stepped primaries. Few people realize this, but the perfect primary tube for a 4 stroke engine is actually one that has a continual taper to it, right up to the collector. Of course, shaping a complete header system with tapered primaries is a huge undertaking and is very expensive. I have actually seen a tapered primary system. Pretty cool. Study 2 stroke exhaust systems and you will understand the importance of the tapered primary idea for a 4 stroke application.
Since tapered primaries is beyond the realm of most people to afford, let alone fabricate, the next best solution is to step the primary diameter at calculated intervals. My primaries should have had another step, but I really didn't want to end up with merging three 1 7/8" primaries. As I look back on it, I wish I had gone ahead and put that extra step in it.

The Y-pipe is something I'm very proud of. In fabricating it I found an angle that gave the merge a very interesting quality. If you take the Y-pipe by itself and take an air gun to blow air into one manifold side of the Y-pipe, the air blast will exit through the turbo flange opening. The interesting thing is, none of the air blast will exit the opposite manifold side opening. But rather, a suction would be present at that side. I was very pleased with that result.

[charlief1]

Once again Donnie, you get me thinking. Your work is amazing and if I get to the point you're at I'd be a happy man.

[DonWG]

Originally Posted by charlief1

Once again Donnie, you get me thinking. Your work is amazing and if I get to the point you're at I'd be a happy man.

CharlieF1. I'm glad you're getting something out of all this.

[it's just a six]

Curious on the type of stainless material used? 304 321,etc?

Also,how many miles did you put on your car with the exhaust wrapped?
Thanks
Looks nice! BTW.

[DonWG]

Originally Posted by it's just a six

Curious on the type of stainless material used? 304 321,etc?

Also,how many miles did you put on your car with the exhaust wrapped?
Thanks
Looks nice! BTW.

The stainless material is 304. I didn't have the extra bread to do 321. Flanges and bungs are mild steel.

The car is strictly for drag racing. Probably around 300 passes. Countless testing sessions at the shop.
You have to keep in mind that exhaust temps of an alcohol engine are lower than a gasoline engine, so parts on an alcohol engine see less thermal loading.

[Dusty Bradford]

Originally Posted by DonWG

In the studying I did when I was first putting together my engine configuration, I spent a lot of time looking at the high end turbo motorsports classes. Especially F1. Studying the F1 turbo era, I found that tuned length headers were very popular. In fact, I noticed that the era started out with more log style manifolds and evolved into tuned length. Now, these people have R&D resources at their disposal that I can't even begin to imagine. I would not dare to second guess their decisions to use tuned length headers. If they found there was power in that decision, then I was on the band wagon too. Another well known tuner, Steve Kinsler of Kinsler Fuel Injection has documented that a turbocharged engine reacts to cam and manifold tuning much the same way a naturally aspirated engine does. His quote on cam selection for a turbocharged engine is, "Don't choke that engine". Again, who am I to second guess his successful experiences. A lot of times you don't need to go through the expense to do your own testing to find answers to some tuning questions. Just look at what other highly successful people have done. And people that have much more resources for testing than a simple hobbiest racer.

I have looked at some pics from the F1 stuff and noticed the tuned length. I didn't know enough details of their rules to know why they may have went to the tuned length exhaust. Rules like if they were turbo limited. I had heard at one time they were boost limited but never researched it. I don't have room for tuned length anyway. I figured they were like NASCAR. They will work for 2 years and spend $200,000 to find 4 horsepower. Mentally I put this technology in the same group. They may have found power but how much. 4hp or 40hp?? I thought you may have seen some testing to know what kind of gains. Thanks for the explanation

[DonWG]

Originally Posted by Dusty Bradford

I have looked at some pics from the F1 stuff and noticed the tuned length. I didn't know enough details of their rules to know why they may have went to the tuned length exhaust. Rules like if they were turbo limited. I had heard at one time they were boost limited but never researched it. I don't have room for tuned length anyway. I figured they were like NASCAR. They will work for 2 years and spend $200,000 to find 4 horsepower. Mentally I put this technology in the same group. They may have found power but how much. 4hp or 40hp?? I thought you may have seen some testing to know what kind of gains. Thanks for the explanation

The only testing that I have experience with is on a desktop dyno. Many, many hours on a desktop dyno. Also, a little coaching from a few very good engine tuners. In fact, one of these fellows turned me onto the engine sim I use to this day. An updated version, of course. It's very enlightening what you can learn by tweaking this and that on a sim.

When I was putting together my configuration, the foundation of the combination was going to be the heads. M&As. At the time I was putting together my combination, M&As were the only aftermarket heads available. I had them max ported and flowed, sticking with the original size valves. There is an intake port gas flow velocity target that is optimum for an engine and the sim let me know that to optimize this particular head I would need to destroke the engine. The shorter stroke slows down the velocity through the intake port for a given rpm. Hence, increasing the rpm capability of the engine before reaching the flow limit of the intake port. I picked a stroke of 3.06". The destroked crank also gave me a stronger crank pin arrangement. There is more meat at the joining of the split pin with the shorter stroke. A more durable arrangement. An added benefit.
Next came the camshaft specs. With the loss of displacement with the shorter stroke, I knew I would have to make up for it with rpm.

If you take away from any one of the 4 variables that ultimately produce horsepower in an engine, you can make up for the loss by increasing any one of the other variables. The 4 variables are P.L.A.N. Some may have seen this in a popular turbocharging book.
P = cylinder pressure
L = length of stroke (crankshaft stroke)
A = cylinder piston area (bore diameter)
N = rotations per minute (rpm)

The head and crank combination was now setup for high rpm operation. The next specification to look at was camshaft.
The intake duration is the primary specification that sets where the useful powerband of the engine will be. In what rpm range.
If you want rpm, you must increase durations. There is no other way around that. The sim let me know that my heads, with the given stroke, would be done at 7800 rpm. I next picked an intake duration figure that would give me usefull power to 7800 rpm. The rest of the cam specs were picked to support the intake duration spec and the shortcomings of the head. A little more duration for the exhaust side to balance out the flows.

Whenever you increase durations, you will increase overlap of the intake and exhaust events. If you want rpm, you are going to have overlap no matter how much you play with lobe separation angles. Seeing that I was going to have to deal with a lot of overlap, I decided to pick a lobe separation angle that would be the most efficient for the engine. Not just one that gave me the least overlap.

Will be back.

[DonWG]

Here is the interesting part of the story.
I know I've posted about this before, but I'm still amazed at how many people there are out there that can't grasp this concept. For those that would like to learn more about pressure pulse tuning, I would suggest learning about 2 stroke engine tuning, particularly the exhaust side, first. 2 stroke exhaust tuning is a very interesting subject. The shape of the 2 stroke exhaust pipe is designed to help draw out the exhaust and a portion of the incoming air/fuel charge into the exhaust pipe. As the piston is moving up and compressing the new mixture, and just before the piston closes off the exhaust port in the cylinder, a pressure wave in the exhaust chamber is traveling towards the exhaust port and crams that portion of the intake charge that made it into the exhaust pipe back into the cylinder through the exhaust port. This is known as exhaust side supercharging. This is free energy that is increasing horsepower. Anyone that is familiar with 2 stroke engines will tell you that there is plenty of power to be had by just matching the correct exhaust pipe to a 2 stroke engine. It is one tuning spec that is never ignored with a 2 stroke engine.
With a 4 stroke engine, that is a completely different story. Especially in the turbocharged ranks.

I'll finish this later.

[DonWG]

A quick note.
F1 - 1987 The TAG-PO1 1.5 liter V6. Porsche design. 8:1 CR. Qualifying boost was 41 psi. Peak power at 12,000 rpm.
F1 - 1984 The Renault EF4 1.5 liter V6. 7.5:1 CR. Boost was 32 psi with more available for qualifying.
F1 - 1988 Ferrari Tipo 033B V6. 2.5 bar boost.
F1 - 1988 Honda RA168-E 1.5 liter V6 competing against 3.5L n/a cars. Boost was limited to 2.5 bar. Still they dominated the season.
F1 - Early 1980s BMW M12/13 based on BMW 2002's four cylinder block. 6.7:1 CR. Needless to say, they used high boost numbers for qualifying to obtain nearly 1,000 hp from 1.5 liters at just 9,500 rpm.
F1 - 1988 Sadly, the last season of the turbocharged F1 engine.

Some years of the Honda and particularly the BMW powerplants used long tuned primaries, along with other manufacturers.

[DonWG]

If all the variables are in place, exhaust side supercharging can be had with a 4 stroke engine. Even a turbocharged 4 stroke engine.

[DonWG]

I don't have any real world comparison of my engine with and without tuned exhaust, but what I'll do later is run a sim with a non-tuned exhaust and post the difference in hp between the two.

OK then. The story left off with the choice to pick a camshaft that would put the useable power range of the small engine up to 7500 to 7800 rpm. This choice meant a bunch of overlap. Enough to make most turbo engine tuners in the Buick community cringe with disgust.
Overlap is good, and bad. If an engine is not tuned properly, as far as manifolding goes, overlap can result in wasted fuel and charge air volume out the exhaust and/or exhaust reversion into the combustion chamber, and even up into the inlet tract. Overlap is what gives naturally aspirated engines setup for good mid-range to upper rpm power their nasty, rumbling idle.
You'll notice on high rpm, naturally aspirated 4 stroke engines that are tuned correctly, they will have tuned exhaust manifolding. They are taking advantage of resonance or pressure pulse tuning. Some think that the only purpose of tuned exhaust headers is to better evacuate the chamber of spent exhaust gases by the creation of a low pressure pulse meeting the closing exhaust valve during the overlap period when the intake valve is also opening. The low pressure pulse meeting the exhaust valve and chamber creates a greater pressure differential between the intake tract and the exhaust tract, doing a better job of evacuating the chamber of spent exhaust gases.

[DonWG]

With any appreciable amount of overlap, what will happen during the overlap period is the exhaust will be better evacuated or flushed from the chamber and at the tail end will be some of the incoming charge air and fuel also being drawn into the exhaust pipe. Since pressure pulses are a series of high and low pressure waves, the trick is to tune the exhaust so that the low pressure pulse changes to a high pressure pulse just as the fresh fuel and air is starting to travel past the exhaust valve. If timed right the high pressure pulse will hit just as the exhaust valve is about to close and the air/fuel charge just starting to make its way down the exhaust pipe will be rammed back into the combustion chamber by the high pressure pulse. This is exhaust side supercharging. Although, in a 4 stroke engine the intake valve is opening so the affect is not the same as it would be with a 2 stroke engine. What it does help is the bsfc of the 4 stroke. Instead of air and fuel being wasted to the exhaust, some air and fuel is recovered by it being rammed back into the cylinder before the exhaust valve seats.
A sim that provides a pressure and flow direction display can show you the pressures of the waves and the direction of the waves for the intake and exhaust at various degrees of crank angle, and can show you how the waves relate to the cam timing you've chosen. The timing of the waves will change with rpm. That is why tuned manifolding systems are more efficient at certain rpm ranges that at others. The timing of the pressure waves can be out of sync at certain rpms that will actually hurt power and cause fuel consumption to jump.

[DonWG]

Some would argue that with a turbocharged engine the exhaust is always at a positive pressure and that pressure pulsing or resonance waves are nonexistant because of that.

[DonWG]

I just did a sim with a set of ATR headers in place of the present system. Boost controlled to 24 psi.
It lost 55hp @ 6750 rpm. That's about a 5 to 6 percent reduction.

[Dusty Bradford]

Originally Posted by DonWG

I just did a sim with a set of ATR headers in place of the present system. Boost controlled to 24 psi.
It lost 55hp @ 6750 rpm. That's about a 5 to 6 percent reduction.

What does the sim take into account regarding the turbo system? I'm guessing it uses the airflow capability of your 91mm turbo. The reason I ask is the ATR headers would never be capable of supporting your turbo size IMO so it would not suprise me that it lost hp. What primary size was used in these examples?

[DonWG]

Originally Posted by Dusty Bradford

What does the sim take into account regarding the turbo system? I'm guessing it uses the airflow capability of your 91mm turbo. The reason I ask is the ATR headers would never be capable of supporting your turbo size IMO so it would not suprise me that it lost hp. What primary size was used in these examples?

All I did was replace the primary and collector IDs and lengths. I also set the boost level to 24 psi for both tests.
The ATR headers used a primary ID of 1.4". The length was an average for all the primaries, just under 8". I used a collector length of 30" to account for the crossover piping. The ID for the collector, 2.25".

I wanted to pick a typical and popular aftermarket header system to do the comparison with. Are there some other specs you want to try?

[Chevota]

Years of studying, hard work and data results posted for all to appreciate. What a nice guy!

scott wile

[Dusty Bradford]

Originally Posted by DonWG

All I did was replace the primary and collector IDs and lengths. I also set the boost level to 24 psi for both tests.
The ATR headers used a primary ID of 1.4". The length was an average for all the primaries, just under 8". I used a collector length of 30" to account for the crossover piping. The ID for the collector, 2.25".

I wanted to pick a typical and popular aftermarket header system to do the comparison with. Are there some other specs you want to try?

I was just trying to get an idea of what input the sim needs to spit out the #'s. I thought it may actually have some sort of method to input turbo system measurements. The ATR header design like all stock Buicks would seem nearly impossible to be able to input into a sim. For example, the collector length isn't really 30" but how can you accurately describe the long distance of 2.25 pipe between the collector and the turbo?? All this seems like it would effect the output of the sim but not necessarily the power of the motor. In the real world. I would assume the ATR design would definately cost power on a motor such as yours with a 91mm. Now back up to something like a 67mm TSM style motor where the ATR header can support the airflow of the engine and I wonder if the tuned system would be a benefit.

Without an accurate way of describing the design of the stock style headers it would seem the sim output may be inaccurate.

[DonWG]

Originally Posted by Dusty Bradford

I was just trying to get an idea of what input the sim needs to spit out the #'s. I thought it may actually have some sort of method to input turbo system measurements. The ATR header design like all stock Buicks would seem nearly impossible to be able to input into a sim. For example, the collector length isn't really 30" but how can you accurately describe the long distance of 2.25 pipe between the collector and the turbo?? All this seems like it would effect the output of the sim but not necessarily the power of the motor. In the real world. I would assume the ATR design would definately cost power on a motor such as yours with a 91mm. Now back up to something like a 67mm TSM style motor where the ATR header can support the airflow of the engine and I wonder if the tuned system would be a benefit.

Without an accurate way of describing the design of the stock style headers it would seem the sim output may be inaccurate.

It is quite possible for the sim to be inaccurate. The output of the sim will more closely mirror real world if the data that is being entered more closely follows the actual specification. Even though, it will give an idea of traits or patterns. For instance, if I took that collector length and tried all kinds of different lengths and still came up with a 55hp deficiency, I would have to guess that the problem is not in the collector length dimension. I would then have to look at the primaries.
I noticed in the sim pressure and gas speed readout during the calculation that there was quite a bit of exhaust reversion in the midrange. I would guess that if I were to throw a smaller compressor and turbine housing into the calc that the higher exhaust system back pressure would make the reversion worse. I'm willing to try it for you though.

Do you have a compressor map for the 67mm turbo? Or one that is closely comparible? I need some figures off of the map to properly simulate it.

I thought TSMs were getting close to 900bhp now. That's not too far off from my engine at the limited boost I've been using in the past. And aren't they bigger cubed? They're using better heads by far. Shouldn't that mean they're capable of pushing more airflow than my engine?

I should note at this point that if a camshaft with very little or no overlap is being used, there's really no sense in worrying about using a tuned exhaust system. I can guarantee though, that using a cam with no overlap will limit the engine.