1 5/8” headers and 500hp?

[learner1]

Tom Walker » Thu Aug 27, 2020 10:51 pm
Maybe we get the very best header for our application and now we need to re cam the engine to take advantage of the new exhaust efficiency, now the new cam might benefit from a change in the intake port area, which requires some carb massaging which might lead to some timing finessing which might lead to a header change, which might lead to any and everything. The envelope constantly gets improved and progresses. Just time and money I guess, it makes the world go around.

Yes! Just as modifications of the intake port require cam changes modification of the exhaust also requires cam changes.

[Kraniet]

Should one assume theres a linear relationship between pipe size and HP?

Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

[hoffman900]

Should one assume theres a linear relationship between pipe size and HP?

Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

I’ve seen 1.375” off the head to the first step support more when built by Calvin, but yes, straight pipe and or very large radius bends.

A 1.375” header like 99% of the stuff you see out there won’t be able to pull it off. Same with that 1.625” number. My observation is the packaging and how smooth the whole thing in matters as much or more than the tuned lengths (which optimum may vary cylinder to cylinder anyway, depending on the assymetrical-ness between each cylinder for a given engine / set up).

[Kraniet]

Should one assume theres a linear relationship between pipe size and HP?

Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

I’ve seen 1.375” off the head to the first step support more when built by Calvin, but yes, straight pipe and or very large radius bends.

A 1.375” header like 99% of the stuff you see out there won’t be able to pull it off. Same with that 1.625” number. My observation is the packaging and how smooth the whole thing in matters as much or more than the tuned lengths (which optimum may vary cylinder to cylinder anyway, depending on the assymetrical-ness between each cylinder for a given engine / set up).

I guess one should also assume that the whole exhaust system will affect how well the headers are working. Collectors and secondaries but also the back pressure produced from mufflers and/or cat would change the working conditions for the primaries I would think.

[hoffman900]

Should one assume theres a linear relationship between pipe size and HP?

Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

I’ve seen 1.375” off the head to the first step support more when built by Calvin, but yes, straight pipe and or very large radius bends.

A 1.375” header like 99% of the stuff you see out there won’t be able to pull it off. Same with that 1.625” number. My observation is the packaging and how smooth the whole thing in matters as much or more than the tuned lengths (which optimum may vary cylinder to cylinder anyway, depending on the assymetrical-ness between each cylinder for a given engine / set up).

I guess one should also assume that the whole exhaust system will affect how well the headers are working. Collectors and secondaries but also the back pressure produced from mufflers and/or cat would change the working conditions for the primaries I would think.

100%, it's a system, not a collection of parts. That goes for the entire engine.

See Darin Morgan's post from 2007 (and his comment on the NASCAR port wanting to be a little bigger):

have never stated 133% for an exhaust exit. Its 103% to 105%. NASCAR type heads like it on the larger side about 108% maybe due to the long exhaust system I dont know. The exhaust throat size comes down to the engine design, valve design, its rpm range and engine speed as well as the engines intended purpose. There are simple guidelines to go by but again, as I have stated many times, the dyno and race track tell the whole story. What we say and talk about on this board is no substitute for your own R&D. Its just points you in the general direction. The exhaust throat size debate usually comes down to arguing semantics. Get on the dyno and creep up on it and watch what happens. The point at which you stop gaining fuel signal ( pulling more fuel) and making more power is where you should quit.

(1) With nasty ramp rate, high lift cams the throat likes to be very large but it must be done correctly.
(2) If you try and use large throats with a nail head exhaust valve, its total disaster and you end up with an engine that wont accelerate.
(3) Low lifts, slow opening, large intake to exhaust valve ratios dont need large throats. Since I dont work on such engines I have had to rely on others R&D efforts to validate this theory.
(4) You cant tune an exhaust port by looking at the flow numbers. If its dead ass smooth and quite, you probably hit a home run. If its loud and turbulent, forget it, it wont scatter *^*&^. It wont run and I have never seen a loud nasty turbulent port make power. I have actually hurt the air flow to get the port smooth and the results where, more often that not, positive. If you plan on making an exhaust port flow a lot of air dont plan on making a lot of power. That's not where its at. I can make my exhaust ports flow another 50cfm ( on Comp and Pro Stock heads) with no problem at all. It would take me all of fifteen minutes. I would also murder the top end power and kill the fuel flow at high engine speeds. The only exception to the rule of exhaust port sizing is Super Charged or Nitrous assisted engines. In these cases its hard to get them big enough but the large throat rule still applies.
(5) If you match the exhaust ports to the headers, just cut your losses and throw the heads away. They are now junk.

These are my simple rules.

[pcnsd]

1. Should one assume theres a linear relationship between pipe size and HP?

2. Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

3. I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

1. No. It is not a linear relationship.

2. My estimation is a 1.375: diameter off the head used in a tight power band per Mr. Elston should support somewhere in the area of 44 HP per cylinder. (Numbers and math can give an illusion of precision. Don't fall for it. YMMV.)
See the attached spreadsheet to understand how Mr. Elston's numbers where manipulated to render a constant used the expand the range of suggested tube sizes.

3. Yes and more. My take of Mr. Elston writings is that we are stepping tube size and applying a choke or a AR device somewhere down stream to protect the back door.

Calvin Elston tube sizing.xlsx

[digger]

1. Should one assume theres a linear relationship between pipe size and HP?

2. Calvins 1.625" pipe support 73.7 HP
Assuming linear relation a 1.375" would support 52.9 HP?

3. I also assume that figure from Calvin is for a straight and/or big radius bend pipe?

1. No. It is not a linear relationship.

2. My estimation is a 1.375: diameter off the head used in a tight power band per Mr. Elston should support somewhere in the area of 44 HP per cylinder. (Numbers and math can give an illusion of precision. Don't fall for it. YMMV.)
See the attached spreadsheet to understand how Mr. Elston's numbers where manipulated to render a constant used the expand the range of suggested tube sizes.

3. Yes and more. My take of Mr. Elston writings is that we are stepping tube size and applying a choke or a AR device somewhere down stream to protect the back door.

Calvin Elston tube sizing.xlsx

thanks for capturing that as the site doesn't load anymore.

[rustbucket79]

Just a generic FYI, the shop I work at has 1 3/4” by 3 1/2 headers, and I personally have a a set of Doug’s 1 7/8 by 3 1/2 Camaro headers, and on back to back comparisons on a handful of higher RPM engines, my headers will lose under 5300 and gain over 5600. Typical gains are 8 to 15 HP, the higher the peak rpm, the better chance it’ll hit that 15.

Did a comparison on a big street small block, this time comparing 1 3/4” to 2” because they were in the shop, the graphs overlayed from 3800 to 5500. (HP peak)

No doubt this all goes out the window once you add a proper street exhaust system to the mix.

[Kraniet]

Heres a quote from Calvin regarding his approach to header construction.

1) Stepping tube sizes is to me essentially a flow issue. I never hold a tube diameter more than 12", unless it is already larger than it needs to be for the given power level. Stepping tube diameters also introduces an area change. These do return a negative wave of their own back to the cylinder, but they also slow particle velocities. A single tube diameter step is essentially a non player "tuning" wise compared to the simple need to increase tube diameter as a flow issue with a given length of tubing. Double steps (or more) become increasingly more involved or effective in "tuning", and that is another whole subject. (2 cycle stuff being the complete opposite as completely a "tuning" issue).
I rely on Burns Stainless and their software to tell me how big a tube I need to flow the given power level at the end of the primary tube length called out. Most of the time their "length" calculations are really close. (again another issue)
So, I measure the area of the exhaust port given to me by the cylinder head porter or manufacturer and that determines the tube diameter I use for the first "step" length. (another story as I seldom see port areas close to the diameters of tubing I have to work with...) If that "first" step length happens to be the same diameter as needed at the end of the whole primary tube length, then there is no reason to step as it is already too large.... at which point the urge overcomes me to ask to the "tuner-builder-porter" if they ever thought about not making the port so large... with the resulting deer in the headlight stare...but I digress...
If the port area and resulting first "step" diameter is smaller than the diameter needed at the end of the primary, then I must step. As noted above, I never run them longer than 12" but I am more partial lately to 10" as a rule. So if a header needs to be say 25" long, stepping from 1.75 to 2.0" at the end, (2 steps), there would most likely be 1.75@10", 1.88@10" and 5" of 2.0". Simple. Current PS stuff I do for a 400incher is 4 steps in 16-17" which is in essence a megaphone with steps every 4 inches. Simple.
Note that 10-12" is the longest step length I use, that does not mean they can't be shorter...
Calvin Elston
Elston Exhaust
Matthews, NC 28104
346-704-4430
Blog: www.exhausting101.com

Hes also said that hes never seen an exhaust port that was too small. Its quite interesting

[hoffman900]

Heres a quote from Calvin regarding his approach to header construction.

1) Stepping tube sizes is to me essentially a flow issue. I never hold a tube diameter more than 12", unless it is already larger than it needs to be for the given power level. Stepping tube diameters also introduces an area change. These do return a negative wave of their own back to the cylinder, but they also slow particle velocities. A single tube diameter step is essentially a non player "tuning" wise compared to the simple need to increase tube diameter as a flow issue with a given length of tubing. Double steps (or more) become increasingly more involved or effective in "tuning", and that is another whole subject. (2 cycle stuff being the complete opposite as completely a "tuning" issue).
I rely on Burns Stainless and their software to tell me how big a tube I need to flow the given power level at the end of the primary tube length called out. Most of the time their "length" calculations are really close. (again another issue)
So, I measure the area of the exhaust port given to me by the cylinder head porter or manufacturer and that determines the tube diameter I use for the first "step" length. (another story as I seldom see port areas close to the diameters of tubing I have to work with...) If that "first" step length happens to be the same diameter as needed at the end of the whole primary tube length, then there is no reason to step as it is already too large.... at which point the urge overcomes me to ask to the "tuner-builder-porter" if they ever thought about not making the port so large... with the resulting deer in the headlight stare...but I digress...
If the port area and resulting first "step" diameter is smaller than the diameter needed at the end of the primary, then I must step. As noted above, I never run them longer than 12" but I am more partial lately to 10" as a rule. So if a header needs to be say 25" long, stepping from 1.75 to 2.0" at the end, (2 steps), there would most likely be 1.75@10", 1.88@10" and 5" of 2.0". Simple. Current PS stuff I do for a 400incher is 4 steps in 16-17" which is in essence a megaphone with steps every 4 inches. Simple.
Note that 10-12" is the longest step length I use, that does not mean they can't be shorter...
Calvin Elston
Elston Exhaust
Matthews, NC 28104
346-704-4430
Blog: www.exhausting101.com

Hes also said that hes never seen an exhaust port that was too small. Its quite interesting

I think Calvin was making a dig at:

1) people designing exhaust ports via CFM. Read Darin's post. It's about area, flow stability, and velocity. I think one of the silliest things (and something Darin Morgan has said) is flowbench people porting to maintain some fictitious intake/exhaust flow ratio. Also to quote Warren Johnson, all flowbenches have flaws and are designed wrong, it is useful as a tool, but you need to know its limits. He didn't specifically say it, but the exhaust port is one of those limits.
2) Construction varies so much between headers, plus packaging, and that dictates tube diameter, which in turn dictates exhaust port size. My observation is the header is an after thought and most are super large, but because they have to be due to how they are constructed.

Here is a quote from myself some years ago:

There is more to it than diameters when comparing across.

Nascar Cup headers start at 1 7/8” when they were making north of 850bhp. 358ci, but revving out over 9000rpm, so do the math in terms of airflow.

WoO Sprint cars are using 1 7/8” off the head, and were experimenting with 1 3/4” off the head on applications pushing 1000bhp. Granted they are running alcohol.

Mark Donohue talked about the extensive header testing they / Traco did, trying everything from 1 3/4” to 2” with no changes in power. I believe Hooker was doing all the testing. They just went with whatever tube was cheapest. Fast forward to now, and Calvin Elston has built headers with 1 5/8” off the head for vintage Trans Am engines that are making 150bhp more than Mark’s car had back then and a ton more rpm, for the same displacement.


That said, you can’t compare sizes from a shelf header, a cheap dyno header, or a custom bespoke piece with large bend radii / etc. Calvin can build you a set that starts at 1 5/8” and support more power than a cheap Hooker or Hedman can at 1 3/4” or even 1 7/8”. You might even have to change a few other things to realize the advantage. He has mentioned how saavy engine builders realized the better you make the exhaust, the smaller the exhaust valve, which typically allows a larger intake. That’s not always so obvious. Furthermore, engine power curves matter way more than peak. A $30k+ engine developed by Hendricks, Ilmor, etc. making 500bhp is going to run away from a 500bhp Car Craft magazine special.

That said, there are a few others that can build headers as good as Calvin, but him being in his semi-retirement, is accessible to the rest of us. I do think he opened up a lot of people's eyes to this.

[blackford]

If the exhaust gasket dimensions (Felpro 1487, 1.62 x 1.42) are a good indicator of the exhaust port dimensions for a 210cc Victor Jr, I would think that the port mismatch might be a larger issue than the flow of 1 5/8 headers. I've had difficulties with finding 1 5/8 headers that had large enough exhaust openings for significantly smaller exhaust ports.

[hoffman900]

If the exhaust gasket dimensions (Felpro 1487, 1.62 x 1.42) are a good indicator of the exhaust port dimensions for a 210cc Victor Jr, I would think that the port mismatch might be a larger issue than the flow of 1 5/8 headers. I've had difficulties with finding 1 5/8 headers that had large enough exhaust openings for significantly smaller exhaust ports.

Yeah. I mean, everything I’m talking about is custom territory, and not all custom is created equal.

[BLSTIC]

A random combo i put together
383
11:1
2.05/1.6 valve
320@700/240@700
XR292R 254/260/110/+2
0.66"/0.67"
single plane

some sims (so not an actual dyno test)

the primary 30" and colletor 17" lengths came from pipemax and remain same

some preliminary runs
1.625 to 1.875" and increments
2.5 and 3" collector

all runs (refer to legend)



best and worst 1-875-3" vs 1.625-2.5"



A quick diagnosis comparing at 7,000 rpm where there is ~4% difference 678hp vs 650 hp (my first interpretation atleast)

mass flow (smaller header has more reversion at IVO due to cylinder pressure higher than intake pressure see pressure grpah)



pressure ( cylinder pressure at BDC to IVO much higher with smaller header, despite smaller smaller header making higher peak suction quite a different shape though, note at this rpm the intake pressure is far from ideal at IVO)



this is delivery ratio/VE (so a small 1% difference due to the reversion at start)



so there is a 4% loss of power at 7000 with smaller header, essentially all caused by higher pressure BDC to IVO resulting in 1% loss of intake fill and 3% pumping loss

There are lots of things you could do to improve things, stepped size on the small headers, different length, more/less duration etc, better intake geometry but each and every change that improves at a given rpm the consequence at other rpm need to be considered so it comes down to bunch of compromises.

it illustares that while scavenging is great it needs to be balanced against pumping losses

Do you still have these files? Could you do a torque output graph starting at lower rpm, say 2500 or 3000rpm, and then the mass flow and pressure graphs at 3500rpm (or wherever the smaller header was showing the most gains vs bigger header)?

It would help me understand why smaller headers help at low rpm. I ask for torque output graphs from such low RPM because horsepower curves are multiplied by RPM and so the differences don't show up so much at low RPM where they would show on a torque graph, and I have a suspicion that the smaller headers could be bringing the torque curve in a bit earlier.

I have EA+ myself, but it doesn't output graphs like this, I think I need Pro for that.

What I *expect* to see is the cylinder pressure drop being a lot closer match early in the cycle (without the header limiting flow) and higher exhaust velocity around IVO. I'd expect there to be lower cylinder pressure on average on the overlap period.

Thanks in advance

[digger]

A random combo i put together
383
11:1
2.05/1.6 valve
320@700/240@700
XR292R 254/260/110/+2
0.66"/0.67"
single plane

some sims (so not an actual dyno test)

the primary 30" and colletor 17" lengths came from pipemax and remain same

some preliminary runs
1.625 to 1.875" and increments
2.5 and 3" collector

all runs (refer to legend)



best and worst 1-875-3" vs 1.625-2.5"



A quick diagnosis comparing at 7,000 rpm where there is ~4% difference 678hp vs 650 hp (my first interpretation atleast)

mass flow (smaller header has more reversion at IVO due to cylinder pressure higher than intake pressure see pressure grpah)



pressure ( cylinder pressure at BDC to IVO much higher with smaller header, despite smaller smaller header making higher peak suction quite a different shape though, note at this rpm the intake pressure is far from ideal at IVO)



this is delivery ratio/VE (so a small 1% difference due to the reversion at start)



so there is a 4% loss of power at 7000 with smaller header, essentially all caused by higher pressure BDC to IVO resulting in 1% loss of intake fill and 3% pumping loss

There are lots of things you could do to improve things, stepped size on the small headers, different length, more/less duration etc, better intake geometry but each and every change that improves at a given rpm the consequence at other rpm need to be considered so it comes down to bunch of compromises.

it illustares that while scavenging is great it needs to be balanced against pumping losses

Do you still have these files? Could you do a torque output graph starting at lower rpm, say 2500 or 3000rpm, and then the mass flow and pressure graphs at 3500rpm (or wherever the smaller header was showing the most gains vs bigger header)?

It would help me understand why smaller headers help at low rpm. I ask for torque output graphs from such low RPM because horsepower curves are multiplied by RPM and so the differences don't show up so much at low RPM where they would show on a torque graph, and I have a suspicion that the smaller headers could be bringing the torque curve in a bit earlier.

I have EA+ myself, but it doesn't output graphs like this, I think I need Pro for that.

What I *expect* to see is the cylinder pressure drop being a lot closer match early in the cycle (without the header limiting flow) and higher exhaust velocity around IVO. I'd expect there to be lower cylinder pressure on average on the overlap period.

Thanks in advance

dont have the inputs to run at lower rpm but have results so can show 4500rpm from those previous runs

Torque


Mass Flow


Ceff


Pressure

[BLSTIC]

dont have the inputs to run at lower rpm but have results so can show 4500rpm from those previous runs

Torque


Mass Flow


Ceff


Pressure

Well would you look at that...

There's a positive pressure wave coming up the exhaust to mess with the tail end of the overlap period that the smaller pipe does a slightly better job of delaying and minimizing.

Wasn't expecting that one. It didn't show what I was expecting at all, but I suppose I should have seen it coming by the 4000rpm being basically zero difference.

Thanks again