I have a shop built 2.5" exhaust system and flowmaster muffler on my isuzu pickup, and I'm building the front half of the system from scratch from mandrel bends bought at summit. The front half was butchered due to my motor swap and I'm piecing it back together. I am using a 02 N/A camaro 3.8L engine through stock tubular manifolds to downpipes into a y pipe and 2.5" all the way out to the tailpipe.

I measured the ID of the stock manifolds incorrectly when placing my order, and bought 1.75" bends. The ID is actually 1.875".

I thought then I must make a return for the right sized pipe.

After some calculations, I found that (2) 1.75" pipes make up the same volume of one 2.5" pipe. Now I guess I am correct in my downpipe size.

Then I thought to myself, If modern exhaust theory is based on a model that describes an efficient exhaust system as "scavenging" exhaust pulses, why does it not take into account the volume of gas at its temperature during cooling?

If your exhaust is say 600F when entering your manifold or header and cools to the point of say 100F when exiting the tailpipe, shouldnt exhaust pipe diameter be reflective of the gas "shrinkage" as it cools?

Here are some numbers assuming exhaust behaves as an ideal gas and the pressure inside an exhaust system is 1 atm due to it having an open end.


degrees F-------gas volume (1 mol)
100F-------------25.51L
300F-------------34.63L
600F-------------48.32L

Exhaust gasses at 600F are almost double the volume as they cool and reach the tailpipe. If exhaust efficiency is to keep velocity constant, wouldn't it not make sense to keep the relative diameter of the pipe consistent with temperature of the exhaust in the system? Why aren't racing exhausts pipes tapered to maintain this difference in volume?

Is this why 2 stroke engines have that curved exhaust shape (like a snowmobile or sportbike)????

(Also, anyone have a good downpipe diameter suggestion for my motor?, since it is for offroad use, I am looking for low rpm torque)

 

Exhaust valve diameter diveded by 3 and multiplied by 5 gets you in the ballpark.

Where does that come from?

I dunno. It was handed down to me years ago. But crunch a few numbers for well known combos and you'll see... ballpark.

As long as we're on it, multiplying your exhaust valve diameter by 1.1 gets you in the ballpark for primary size.

Finally, you almost always need less diameter EVERYWHERE in your exhaust sytem than you think you do. Unless you're using a power adder. Then those guesstimates don't work right anymore.

 

You'll need to use 2 inch downpipes to keep from causing a restriction.

Compare inside diameters (ID), or actually the square of the ID to compare flow area rather than outside diameters (OD). You can figure the actual areas, but PI/4 is linear and cancels out, so you can just use the square of the IDs for comparison of area.

If you assume 1/16 (.0625) wall thickness, the ID will be 1/8 (.125) smaller than the OD. So, one 2-1/2 OD tube has the flow area of 2 tubes 1.83 OD, or nearly the same a two 1-7/8 tubes. However because there is almost 50% more surface area inside the two 1-7/8 tubes, flow inside one 2-1/2 OD tube will be fairly similar to the flow of two 1-7/8 OD tubes.

So, should you really have a single tube after the "Y" equivalent to two 2 inch tubes? That would be about 2-3/4 OD. I don't think you need more than the 2-1/2 OD you already have due to what I said above and what you said below...

I like how you think. Scavenging takes place primarily in the headers, and is based more on resonance than flow, but it gets lots more complicated.

Your best point is that flow (volume of gas) decreases as it passes from the engine to the tailpipe exactly because it cools. Your temps are a little off but the idea is sound. At high power, high rpm exhaust gas temperature (EGT) is above 1200 F at the exhaust port, and maybe a couple hundred F at the tailpipe behind the muffler. The largest relative drop in the shortest diatance after the headers/exhaust maniflod is probably across the muffler, which may be a 300+F EGT drop.

So, yes, we can/should reduce the pipe diameter as we get farther from the engine and therefore cooler gas. However, as we put bends or other restrictions in the system, we need to keep the pipe size up to prevent flow restriction. Theoretically you might want to increase the diameter for the approx. 180 degree bend over an axle. That's been done by some, BTW.

Does this say that placing the muffler as far from the engine as possible makes if effectively "flow" more or at least cause less restriction? I think so.

Does this say that the muffler outlet pipe, if it doesn't have to make many bends, can be smaller than the muffler inlet, not larger? Sure. Ricers won't like this.

Does this say that "header mufflers" might be in the highest flow area and cause the most restriction. Probably yes.

Bottom line, I'd use 2 inch downpipes if the manifolds are 1-7/8 ID and your mandrel-bent 2-1/2 pipe the rest of the way. I'm not a big fan of Flowmaster's ability to flow air, so I'd also use something like a Dynomax Ultraflow welded muffler like P/N 17218, 17219 or 17222 which flow plenty of wind.

http://www.dynomax.com/documents/ultrafloweld_specs.pdf

My highly-opinionated $.02

 

its not just about cross section, keep that in mind.

2 x 2'' square tube will not flow the same as 1x 4'' tube.

A circlular tube with a cross section of 5'' will flow more than a squart with a cross section of 5''

fluids isn't intuitive

 

Ah, but the fluid molecules themselves are smart enough to figure out the surface area of the conduit where the lazy molecules like to rest instead of moving with the crowd. The fast movers get as far away from the wall as possible. That's why they like circular cross sections; it's easy to find the fast lane.

Of course you are correct, Boost. One should adjust the size of the "flat tube" to allow for that. You might compare the boom tube dimensions in the link to the round tubes that they connect to for guidance.

 

Hmmm...
I wonder, if the temperature of the gas is going down and the volume of the pipe is the same, does this help flow by creating a slight vacuum (relative to the higher temp area)? Maybe by keeping the constant diameter piping you not only minimize flow restrictions/head loss around the bends but may also help to draw the gases out further by dropping pressure as temperature drops...

If you dropped the volume of the exhaust pipe as the temperature drops, you would be keeping the pressure the same right? That doesn't help flow...

I think that while the speed of the exhaust gases are slowing down as temp drops, which you mentioned, the mass flow rate is probably relatively constant.

Just my 2 cents...

 

If the temp drops, the volume decreases, it doesn't remain the same. So less volume per minute thru the same size hole means less velocity. Now velocity and pressure are inversely related: as the velocity decreases, the static pressure (pushing in all directions) increases.

Yeah, mass flow rate is pretty much constant, but we're more concerned with cfm, not pounds/min.

Question: Why do we step headers to a larger size in the primaries? Cetainly not to increase backpressure. Maybe there is more than one reason.

 

How big of a role does the expansion of the metal affect this flow, especially since headers can be made of different materials than the rest of the exhaust system.

I know this sounds kind of rediculous, but has anyone ever tried to run duals with one side being a bigger dia. than the other with an x-pipe for vehicles that need high and low rpm power?
OR how about running a lined steel braided hose that is braided to flex in dia. as opposed to wanting to bend?

 

Virtually none, IMO. Thermal expansion might tighten up some chassis clearances, but most exhaust system materials, like steel, stainless, titanium or Inconel have expansion rates that are reasonably similar. On the plus side, as you get into the the power range of the engine, there is more exhaust flow which is hotter so the expanded header tubes should flow better. I'm just kidding...sorta. A 2 inch OD tube heated to 1275F from room temp shoud increase it's flow area about 1-1/2%. BFD, huh?


If you were looking for the last fraction of a % hp, you might size each cylinder's primary based upon it's particular inlet tract length and flow. Of course you'd also have a specific cam grind for each cylinder also. I suspect the cam would be more important. Has this been done? Sure. Is it worth the cost? How fast do you want to go?

Why bother. Braided flex connectors on transverse FWD cars are to allow the engine to rock (for/aft) in it's mounts under torque loads.

You do come up with some interesting ideas, 77, but we've been there before, right?

 

Virtually none, IMO...A 2 inch OD tube heated to 1275F from room temp shoud increase it's flow area about 1-1/2%. BFD, huh?
WOW, i thought that there would be small increase in dia. but that it would add up over the length of the exhuast system, but i didn't know that it increased so little. Thats amazing

If you were looking for the last fraction of a % hp, you might size each cylinder's primary based upon it's particular inlet tract length and flow. Of course you'd also have a specific cam grind for each cylinder also. I suspect the cam would be more important. Has this been done? Sure. Is it worth the cost? How fast do you want to go?
I've always wandered about that to. I just thought that since headers of f1 cars had so many bends, that the flow, and susequently, each pulse of exhaust gas could not be predicted. However, I guess you could take 4 (4 for one header on a v8) of those old wind speed devices

and completely synchronize them so that not only their speed, but the position of the spinning cups are the same. It would tell you whether to increase exhaust speed(dia.), and whether to increase the length (pulse) But thats a cheap lowtech solution to what sounds like a very complex issue.

Why bother. Braided flex connectors on transverse FWD cars are to allow the engine to rock (for/aft) in it's mounts under torque loads.
I was thinking along the lines of allowing the diameter of the exhaust system to expand, so that as the RPMs increased, there would be more room for the increased flow. I guess it would be a moot point though since racing cars only use a specific RPM band, as opposed to running up and down all RPMs, all the time.
You do come up with some interesting ideas, 77, but we've been there before, right?
Thanks!

 

As Temperature drops, the density of the gas goes up so the volume the gas would occupy drops, but the volume of the exhaust system is constant, it didn't change. That is the what I referred to earlier, the constant volume of the pipe. That is why I think pressure drops as temperature drops. (Not to mention the headloss and energy lost in flow work).

Again, the volume through the hole is constant, but the density of the gas is not. Since the mass flowrate throught the system is maintained, as the density goes up, the velocity goes down.

I don't know...that's a good question.

 

Doesn't pressure increase as velocity decreases in gas flow?

Hood scoops with a fairly small opening (Pro Stock?) with a larger volume behind them cause pressure in the plenum above the carbs to be above atmospheric. The 200 mph vehicle speed air entering the scoop slows down in the big volume, and pressure rises. The same thing happens in most cooling systems; there is a large volume or plenum in front of the radiator, so the air getting into this area thru the smaller grille opening loses velocity and static pressure rises and forces the air thru the radiator.

A carb venturi is just the opposite; velocity increases in the venturi so the pressure drops and "sucks" fuel in. OK, atmospheric pressure in the float bowl pushes it into the low pressure area in the venturi.

What am I missing?

 

The difference between the two concepts is that the pipe is stationary in the exhaust and the gas is flowing. In the hood scoop or radiator, the air is stationary and the object is moving, scooping the air...

The air in the hood scoop or in front of the radiator is above atmospheric pressure because the car is moving through the atmosphere, colliding with the molecules. That is like the pressure you feel on your hand when you put it out the window, the faster you go, the higher the pressure/drag/force...If you slow down the vehicle, the pressure goes down too.

There are a couple of principles you can use to figure out generally what is going on in gas/fluid flow. Again, these are not down to the gnat's **s because the are for "ideal" circumstances, but can be followed for ideas as to what is going on.
This is the ideal gas law:
(P*V)/T = (P*V)/T
where P is pressure, V is volume, and T is temparature.
With this, you can see that as temperature drops, either pressure or volume has to go down to maintain the relationship. With the exhaust pipe staying a constant volume, pressure has to drop.

The fluid flow equation I was thinking of for this is mass flow rate is equal to (Density of fluid)*(Cross sectional Area of pipe)*(Speed of flow). (This formula looks like M = P*A*V, that's the best I can do with my keyboard).
As the mass flow rate doesn't change in the system, kind of like current in a series circuit, you can use this formula to figure that as density of the exhaust goes up, and the area and mass flow are constant, the speed of the fluid goes down.

I don't know any equations for the hood scoop pressure increases but I am sure someone a lot more knowledgeable than I am has some formulas relating the vehicle speed, air temp/pressure, scoop size/shape and all that...

I hope that makes sense above, sorry if I made it more confusing...

 

This is not a static box of gas. Rather it is a dynamic flow of gas. It's the volume of the GAS that changes. Now because that lower volume gas is moving thru the constant section pipe, there is less volume per unit of time (cubic ft per min, or Liters per second aka flow). If you decrease the flow thru a constant section, you either decrease the velocity or increase the density. Bernoulli's equation relates velocity and pressure. (pressure/density) + velocity squared/2 = constant. IOW, higher velocity = lower pressure and vice versa.

So, assuming Bernoulli was correct, pressure increases, not decreases as the gas slows down.

You really have to keep Bernoulli's principle in mind. Compressible flow is a much more extensive topic, probably not for this thread.

BTW, if you stick your flat palm out the car window, give it a little angle of attack; tilt the leading edge up 10-15 degrees or so. The lift you feel isn't from molecules impacting the palm of you hand hard enough to raise it. It's from the increased velocity over the top of your 'wing/hand' reducing the pressure on the back of your hand. You can do this without raising the front of your hand if you cup your hand (camber your wing). Agree? If so, you are buying Bernoulli's idea. If you are of the "impact causes lift" faction, we really can't carry on a discussion.

At least we agree that the exhaust gas slows down as it cools, or at least has less volume. Maybe that means that we could/should use smaller pipes as we get cooler gasses.

 

I totally agree with Bernoulli's principle, it applies to nozzles/venturis though. Not straight piping.

If the pressure in your exhaust goes up as the temperature goes down and gas speed goes down, then the pressure at your tailpipe is higher than the pressure at your headers, right? Wouldn't that make flow go the wrong way?

I never tried cupping my hand when out the window, I am going to have to try that.....