Designing An Exhaust That Pumps Up The Power

If we consider a four-cycle engine in terms of the normal, induction, compression, expansion (power stroke), and exhaust cycles, we also have a fifth element that needs to be added – exhaust scavenging. If the exhaust system is suitably length-tuned, negative pressure waves can scavenge the combustion chamber during the valve overlap period that exists at the end of the exhaust stroke and beginning of the intake.

To understand how this increases an engine’s breathing, let’s consider the cylinder displacement and combustion chamber volumes of a typical high-performance Chevrolet 350ci small-block. When traveling down the bore, the piston of a 350ci engine displaces 727cc. If the engine has a compression ratio of 12.0:1, the total combustion chamber volume above the 727cc will be 63cc. If a negative pressure wave draws out the residual exhaust gases remaining in the combustion chamber at TDC, then the cylinder can draw in 790cc (727 + 63). The result is that this engine now runs like a 385ci engine instead of a 350. But there’s more to this than just scavenging the chamber.

Starting at cycle number five, we see that the exhaust-generated vacuum starts the intake charge moving into the cylinder before the piston starts to go down the bore. As the crankshaft rotates farther we get to cycle number one. This is normally considered the induction stroke that draws the new charge in. In an ideal situation, cycle number five has cleared the combustion chamber and put a considerable amount of kinetic energy into the incoming charge before the piston even starts down the bore. The result is an engine that can achieve a volumetric efficiency well over 100 percent. The bottom line is, a good exhaust system is worth a lot of extra torque, horsepower, and (best of all) extra mileage.

If enough energy is put into the incoming charge by the exhaust, then it’s possible to cause the cylinder to fill to a pressure greater than atmospheric at the time of intake valve closure. To put the importance and capability of the exhaust pressure wave into perspective, let’s consider a few numbers. Air moves from one point to another by virtue of a pressure difference. The charge-inducing pressure difference (suction) is normally associated with the piston traveling down the bore on the intake stroke. The better the head flows, the less suction it takes to fill (or nearly fill) the cylinder.

Because this occurs during the valve overlap period, much of this suction is applied via the open intake valve. The exhaust system can draw on the intake port as much as 500-percent harder than the piston going down the bore. Under these circumstances, it’s the exhaust that is the primary element of induction, not the piston traveling down the bore. With such a system, the charge in the intake port can be traveling into the cylinder at 100 feet-per-second, even though the piston is still parked at TDC! In practice, the exhaust phenomena turns a high performance engine into a five-cycle engine with two consecutive induction events.

If headers are not in the budget, exceptionally-reworked exhaust manifolds can flow “almost” as well as headers.

Low Budget and Basic Iron

If the cost of a set of headers is outside your budget, you’ll have to make the best of a set of iron exhaust manifolds. As has been so well demonstrated by Randy Brzezinski at Brzezinski Racing, these can be modified to achieved power gains equal to about two-thirds of that produced by good headers.

If you are porting a set of iron manifolds, port match the top and sides, but not the bottom. Leaving a step here acts as an anti-reversion dam to cut the reverse flow on the floor of the typical Chevy port. When this reverse flow occurs, it brings about a substantial reduction (20 to 30 lbs./ft.) in low-speed torque.

However, the modifications are often more than just grinding the manifolds to make them flow better. In many cases, the manifolds have deflector plates welded in to cut cylinder-to-cylinder interference. But, replicating Brzezinski’s best efforts could well be as costly as buying an inexpensive brand of headers.

For a performance street motor, having a wide power band produces almost as much driving pleasure as having a lot of horsepower. The best output at any particular RPM is seen when a certain exhaust velocity exists. This means the best-sized pipe for 3,000 rpm will be different than what’s optimal at 5,000 rpm, so some compromises must be struck.

The industry standard for primary pipe sizes for the street-performance small-block has been established at 1 5/8 inches. This works well on motors from 200 to 375 hp. Headers with 1 3/4-inch pipes are made for popular applications, but those are if your power target is above 375 horsepower.

This graph can be used to make a near optimal primary pipe size selection first time around. First, you will need the exhaust port flow for your heads at the valve’s full lift point. Next, determine which of the graph lines fits your needs. The green curve is for best all-around results on the street. The purple line is more for street/strip applications where low-speed torque will suffer marginally, but the drag strip results will look good. The blue line is for high RPM race operation where the lowest RPM seen is about 75 to 80 percent of the maximum. An example pipe selection works like this: the heads have 175 cfm at the valve lift used and the application is street/strip, so we choose the purple line. Find the 175-cfm point along the bottom scale. Next follow this up to the purple line then directly across to the scale on the left hand side. This indicates a primary pipe size of 1.625 ID is needed for the job.

The collector diameter is as critical as the primary pipe’s. For good all-around street performance, a 2 1⁄2-inch collector works well. A 3-inch collector is only of advantage when the horsepower is above the 375 mark. If you have a really high output unit in mind, then the collector size should be one and three-quarter times the diameter of the primary pipes.

Having a high pressure at the port during the last phase of the exhaust stroke means the combustion chamber is at a higher pressure than the intake port. When this is the case, the piston will have to travel a short way down the bore before it can draw in any fresh charge. This means the higher combustion chamber pressure has the same effect as reducing the engine’s cubic-inches, and that’s something we don’t want.

From the graph, it can be seen that headers with around 18 inches do not perform as well as those with longer primaries. However, once the header’s primary length is greater than about 24 inches, the system becomes very insensitive to length changes. This means we do not have to worry about getting all the primary lengths exactly equal, because almost any length between 24 and 42 inches gets the job done nicely.

Installing headers improves the flow and separates ports so there’s less interference. The effect headers have on a short-cammed street engine proves worthwhile even if the exhaust system has limited flow and produces backpressure.

Pipe Lengths

Misconceptions concerning header primary pipe length are widespread. First, let’s take the often-quoted phrase “equal length headers.” Under ideal conditions, it’s entirely practical for an exhaust system to scavenge the cylinder at near maximum intensity over a 4,000-rpm bandwidth. Most race engines use an RPM bandwidth of 3,000 rpm or less. If the primary pipe scavenging effect overlaps this band, then it matters little that one pipe tunes as much as 1,000-rpm different from another. Since this is the case, pipe lengths varying by almost a foot have little effect on power.

A good street engine can have a working RPM range as high as 6,000 rpm. One way to spread the aforementioned 4,000-rpm band is to use differing primary lengths. A positive, power-increasing attribute of differing primary lengths is that it allows larger radius, higher-flowing bends and more convenient pipe routing to the collector in confined engine bays.

Apart from the reasons just stated, there’s another good reason why worrying about equal primary lengths is a waste of time. In practice, a V8 engine such as the small-block Chevy is simply insensitive to substantial primary length changes.

Unlike the primary length, the collector or secondary length is far more critical. What you see here is the difference in rear wheel output on an IMCA dirt car engine. The dark blue curves are for a stub about 1-inch long on the collector. The red curves are for a 10-inch secondary pipe on one side (that’s all there was room for) and a more optimal 14-inch on the other. Adding these secondary lengths cost less than 15 bucks and represents a terrific return in terms of output gain per dollars spent.

Dyno tests with headers having primary lengths adjustable in 3-inch increments show that lengths between 24 and 42 inches have only a minor effect on the power curve, although the longer pipes did favor the low end. Since a typical street header can have pipes ranging in length from 24 to 42 inches, we can safely conclude that each pipe comes into its own somewhere in the used RPM range, thus helping to spread the powerband.

At this point, we can sum up the situation concerning header pipe length by saying that, under most circumstances, it’s not critical. This is fortunate, because it’s difficult to change header primary-pipe length, and custom headers are expensive compared to off-the-shelf items. With that in mind, let’s move on to the secondary or collector length.

A basic rule on collectors is that short, large diameters favor top end, while long, small diameters favor low end. Except for the most highly developed engines, most collectors seen at the track are both too large in diameter and too short. For a 7,500-rpm race-cammed small-block Chevy, a collector length of 8 to 12 inches proves to be about the most effective. If the vehicle concerned has a relatively tight converter and launches at low engine RPM, then a longer secondary pipe can cut e.t.’s by getting the car underway quicker. On a car with a tight converter  (2,000 stall), I have successfully employed a 40-inch collector length.

Be aware that selecting a high-flow muffler will be of little use if the cats used are low flow. Check out what’s available here from Walker and other high-flow cat manufacturers.

Muffled Systems

Tuning the collector length is all well and good, but what if the vehicle has mufflers? The good news is that it’s not necessary to give up or even significantly compromise exhaust system performance. To understand how this can be done, we need to consider the two major aspects of exhaust system performance separately: pressure wave tuning from length selection, and minimizing backpressure by selecting suitably high-flowing mufflers.

Usually, a quiet exhaust system and power are considered mutually exclusive, but this needn’t be the case. It’s just a question of knowing how to select suitable components and putting them all together in an appropriate manner. An important aspect is muffler flow, so we’ll now deal with this in detail.

A muffler test where the muffler’s capability is tested just by attaching it to the end of the collector is a false test. Here’s why: We already know that the engine’s output is sensitive to the secondary lengths. Just adding a muffler without consideration of its type can drastically alter the secondary tuned lengths, and in so doing, completely invalidate the test.

Muffler Flow

Buying a muffler based on pipe diameter has no merit other than it allows the identification of a convenient pipe size. It has little bearing on satisfying the engine’s flow requirements. In case you find that hard to believe, let me tell you that there are, or have been, several well-known brands of mufflers where the only difference between a 2-inch and a 2 1/2-inch muffler is the size of the incoming and outgoing pipe. The flow, which is largely dictated by the design of the innards, was only marginally better with the larger pipe size. Indeed, there are muffler manufacturers who have “tuned up” their products by introducing new models with larger inlet and exit diameters, but the innards stayed about the same, and so did the flow numbers.

So, what was achieved? Just more muffler sales to those who understandably assumed bigger must be better. Just so we know exactly where things stand, let’s make it clear that the engine doesn’t have a tape measure, and is, for all practical purposes, totally insensitive to size. On the other hand it’s sensitive to flow capability, so we should buy our mufflers based on flow, not pipe size.

In the instance just mentioned, the muffler had an increase in the entry and exit pipe diameter, but the apparent size of the muffler core remained unchanged. It makes for an interesting mental experiment to view a muffler installation as three distinct parts. These are: entry pipe, muffler core, and exit pipe. If the muffler core flows significantly less than an equivalent length of pipe the size of the entry and exit pipe, then the engine “sees” the muffler as if it were smaller and consequently more restrictive.

Don’t confuse what it is you are trying to achieve when selecting a muffler. The size of the inlet/outlet pipe has little to do with what the engine “sees.” Take a look at the number one set of pipes. The pipe ends represent the inlet and outlet. If whatever muffler used was 100-percent flow efficient for its size rating, it would appear to the engine as a straight pipe of the same size as the inlet and outlet pipe as per number three. In fact, most mufflers are as shown in number two, and to the engine they appear as per number four. As you can see, the inlet and outlet are hardly restrictions here.
What we need is a muffler with a core flow greater than the inlet and outlet pipe. If this is achieved, we get something that can be represented by number five—and that’s the best for power.

If the core has more flow than the equivalent pipe size, it appears larger than the entry and exit pipe. Result: the muffler is seen by the engine as a near zero restriction. A section of straight pipe the length of a typical muffler flows about 115 cfm per square inch. This means a 2 1/2-inch pipe will flow about 560cfm. If a muffler flowing 400 cfm is attached to a 2 1⁄2-inch pipe, the engine sees the muffler as a pipe having an apparent diameter of only 2.1 inches. I bring up this point because so many hotrodders worry about having a suitably large pipe into and out of the muffler. This concern is totally misplaced, as in almost all cases, the muffler is the point of restriction, not the pipe.

Required Flow

Knowing how much a muffler flows is a good start. In the absence of knowing any better, we could assume that the greater the flow the better. As it happens, in much the same way as air filters, this is one of the areas where too big doesn’t hurt power. Increasing muffler flow unlocks potential engine power. Once all the potential power is unlocked, further increases in exhaust system flow will not produce any further benefits in terms of power. On the other hand, depending on the muffler design, any excess flow capability may lead to an otherwise noisier system. From this we can conclude that too much muffler flow serves no useful purpose and could cost more money than necessary.

The trick is to use just the right muffler at the lowest cost to allow the full power potential of the engine to be realized without undue compromises in terms of noise. Now the question is, how much flow is enough?

Here are the results of the backpressure tests I ran to see just how much flow is needed by an engine so it does not suffer backpressure-induced losses. As you can see, 2.2-cfm per horsepower muffler-flow allows the engine to produce 99.5 percent of its open exhaust output.

Once the flow exceeds about 2.2 cfm per horsepower, the gains from increased muffler capacity drop to about 1 percent or less. In other words, the engine’s demand for flow has been satisfied, and backpressure-induced power losses are contained to within less than 1 percent of open-pipe output. With the availability of this key number, you’re now in position to determine how much muffler flow your engine is likely to need.

It only requires that you make a reasonable estimate of its open exhaust power potential, and then multiply this number by 2.2. For example, a small-block Chevy that made 400 hp with an open exhaust will require 880 cfm (400 x 2.2). Two 440 cfm mufflers will get the job done and contain the loss to four horsepower or less. See how easy making an appropriate choice becomes with mufflers rated in CFM?

Pressure Waves

Many mufflers are made up of inter-connected chambers while others are of the “glass pack” variety. These types represent opposite ends of a spectrum, and have a substantially differing response to arriving pressure waves.

This book is an all-new color edition of a previous best seller. It contains the latest engine-building techniques, profiles current technology, and includes today’s affordable parts and engines. Vizard performs ten engine builds, which include dyno charts and parts lists.

Previously, we emphasized that collector length was, in most cases, more influential than the primary pipes. Adding a muffler to a system with already optimized lengths may produce a pressure wave-induced response that has a far greater effect on power than the change in muffler flow.

Let’s assume that a test muffler is attached directly to the end of the collector. Remember, a pressure wave is reflected when it reaches the end of the exhaust pipe or when a substantial increase in cross-sectional area occurs. Mufflers with chambers, such as a Flowmaster, often appear to the pressure wave much the same as the end of the pipe. This means the pressure waves see no change in length, and reflection occurs largely as it did prior to the fitment of the muffler.

Now let’s look at a high-flow glass pack. If the glass pack is densely packed and the perforation holes are small, the muffler can appear as a considerable extension of the tailpipe length. A 3-inch change in tailpipe length can have a measurable effect on the power curve, so what would you expect if another apparent 18 inches was added? Answer: A big drop in top end power.

Situations involving high compression ratios, long duration cams, and nitrous oxide are usually more demanding in terms of noise suppression. Big cubes, shorter cams, and lower compression ratios are easier to muffle. Unfortunately, aftermarket performance muffler noise levels are rarely comparable to original equipment. Still, a little more noise from a nicely tuned engine is, for a performance enthusiast, just another form of music.

Balance Pipes

Balance pipes have two possible attributes: increased power, and reduced noise. The dimensions of the balance pipes are not overly critical. The only dimension that appears to have measurable influence is the pipe diameter. This requires an area at least equal to that of 2 1/4-inch diameter, with 2 1/2 to 2 3/4 being preferable. Anything above 2 3/4 inches does not appear to deliver any further gains, but I have only conducted tests on engines up to about 600 hp. As for the length of the balance pipe, this appears to be largely immaterial.

Balance pipes hare a benefit because they increased power and reduced noise.

Summary

Although I have not dealt with the theory behind everything covered, you still should be sufficiently informed to assemble a near-zero-loss muffled exhaust system. As long as you don’t lose sight of the essentials and principles involved, good results will be achieved. These guidelines are sufficiently well defined to accomplish the desired goals. Step outside these recommendations and you are on your own!

Randy Bolig

Randy Bolig has been working on cars, and involved in the hobby ever since he bought his first car when he was only 14 years old. His passion got him noticed by many locals, and he began to help them with their own vehicles.