Maximizing airflow to the engine helps ensure your engine produces the maximum horsepower and torque. Selecting a carb too big for the job is really all about making sure that whatever it flows in the way of CFM is done while meeting the demands for a sufficiently strong booster signal. Even without having the entire range of rebuild/modification equipment of a full-blown carb shop there is a lot you can do to a stock Holley to increase airflow without incurring any penalties. Even simple mods produce increased airflow as well as enhance the allimportant booster gain/signal. In this chapter I use an 850 vacuum secondary Holley to show how to bump it to almost 960 cfm, and actually increase the booster signal in a proportion slightly greater than the airflow increase. In practice, this means the low-speed output of this carb is as per its original CFM rating, but the high-speed output is as per the modified CFM rating.
This Tech Tip is From the Full Book, DAVID VIZARD’S HOW TO SUPER TUNE AND MODIFY HOLLEY CARBURETORS. For a comprehensive guide on this entire subject you can visit this link:
SHARE THIS ARTICLE: Please feel free to share this article on Facebook, in Forums, or with any Clubs you participate in. You can copy and paste this link to share: http://musclecardiy.com/performance/how-to-improve-airflow-in-holley-carburetors/
Now you may ask, If I can do this in my shop at home, why can’t Holley engineers with all their resources do it? The answer is, They can. Everything I discuss here is known to Holley engineers. The reason they don’t incorporate most of what I am about to show you is simply a matter of production cost.
The Source of all Induction
Whether induction is brought about by the exhaust pulse or the downward motion of the piston during the induction stroke, its source is the cylinder. When selecting a carb, make sure that the induction demand originating at the cylinder is allowed to communicate as effectively as possible with the carb. This means that whatever intake manifold you choose needs to be as flow efficient as possible.
ficient as possible. Runners must be the right size for good port velocity and they must be shaped to be efficient in terms of flow. A runner too big for the job does not produce anywhere near the best torque curve. The better the intake, the better your engine responds to a thoughtfully selected carb CFM. Also bear in mind that as far as the boosters are concerned, their job of atomizing the fuel is better served by a pulsating flow as each individual cylinder draws. An inefficiently flowing manifold or one with too much volume (due to big runners or big plenum or a combination of both) damps the pulsing flow seen by the boosters and low-speed output suffers while showing no gain at high speed. Part of the reason a large carb can work so well on a good (emphasize good) dual-plane is the fact that the boosters see a much stronger induction pulse from any given cylinder.
Making More Air
There are plenty of reasons for wanting more air from a Holley carb without spoiling its low-speed capability. Just as I was about to write this chapter, I prepped a carb for a couple of big-block Chevy test applications. Both were true street-performance engines, and as such, a vacuum secondary carb was appropriate. One was a low-buck 468 engine built to demonstrate the results from a set of Edelbrock’s entry-level E heads. The other engine was for a GM 572 update project.
The goal was to retain total streetability while extracting maximum output within a certain cost range. Any decent-size big-block really deserves the services of a Dominator in terms of WOT output, but in this case cost and fuel economy meant utilizing a 4150-platform carb. Because a big-block needs Dominator airflow you can see the need to make the most of any increased airflow potential a 4150-style carb may have. Because a choke was a prime ingredient, an 850 vacuum secondary carb was chosen for the series of flow tests. These tests also help you better understand and appreciate where major flow losses occur and where they might look as if they occur but don’t.
Flow Mods: Phase One
Mounting the carb directly into the flow bench and blowing air through the carb is no different than mounting it the other way around and drawing air through it. The advantage of blowing air is that the exiting flow path can be easily plotted with the aid of a pitot tube velocity probe. Refer to Phase One shown in “Airflow Test Results” on page 134. Test 1 is the stock carb’s flow capability. Test 2 is about as simple as it gets for more airflow through a Holley carb, especially one that must have a choke and, consequently, a choke horn. Test 2 is also with a K&N Stub Stack installed. Note the baseline airflow went from 849 to 872 cfm and it took much longer to unpack the Stub Stack than it did to install it.
Incidentally, I have had a number of successful engine builders express surprise when they see a Stub Stack on one of my Holley-equipped engines. The usual comment is along the lines of, “They don’t actually work, do they?” That is sort of a statement and a question all rolled into one. Well, here is the bottom line: The reason I designed the original Stub Stack and took it to K&N was because I felt there was a need for such a system. My business is making more horsepower, so if it did not work, I would have dropped the idea. When I demonstrated a prototype to K&N engineers on my dyno, it showed a 5- to 6-hp increase on a 400-horse small-block Chevy. On the popular 496 (1/4-stroked 454 with 0.060- inch oversize bore) a big-block Chevy Stub Stack can be worth up to 9 hp, depending on the rest of the engine’s spec.
Flow Mods: Phase Two
On the subject of Stub Stacks and spacers and how they affect flow and power potential is where mounting the carb upside down and blowing through it shows just what a spacer can do for flow. In Test 2 here, the baseline is once again 849 cfm for a bare carb. The first spacer to be flow tested is of a not-so-common 2-inch-thick, “four round to two oval” hole design. Note the flow went from 849 to 937 for a whopping 88 cfm increase. If that sounds almost unbelievable, let me explain why the gain was so substantial.
It’s far more important how the air leaves an object in its flow path than how the air arrives at it, unless supersonic speeds are involved. In this instance the spacer allows the discharge from the base plate to be tidied up along its exit path. That is good, but what you see here is something of an artificial situation. With no manifold downstream of the spacer, the spacer itself acts as if the manifold is perfect in terms of allowing an uninterrupted flow. In practice the presence of the intake manifold can seriously interrupt the discharge flow pattern out of the spacer. With a well-designed single-plane intake having steeply inclined ports for a more direct carbto-runner path, things are not too bad. Where problems arise is when the hood-to-carb clearance may be an issue, especially with a dual-plane intake and, specifically, with the higher of the two plenums.
On the high plenum of a dualplane without a spacer, the air exiting the carb has to make an almost immediate right-angle turn upon exiting the carb (see Figure 14.10). That, as you can imagine, is not good for flow. By using a “four round to two oval” hole spacer as shown in Figure 14.11, you can dress out the typically tight radius seen at the top turn of the high plenum on a dualplane to good effect.
With this as part of the power recipe, I have seen as much as 570 hp from a dual-plane, bored and stroked small-block Chevy. This design of spacer maintains the integrity of the dual-plane 180-degree induction pulse separation concept. In other words, you are not making a poor single-plane intake from a dual-plane concept.
Test 3 shows the result of using the most common of all spacers: the 2-inch open design. In this instance flow increased from 849 to 954, a 105-cfm increase. Test 4 shows the effect of a fourhole tapered spacer. This type of spacer (sometimes referred to as a “super sucker”) seeks to streamline the exit as far as possible within the 2 inches it occupies between the carb and the intake. On single-plane race intakes these are the spacers most likely to work.
On some occasions race rules or race conditions may call for a spacer whose sole purpose is to reduce engine output to one more appropriate to the track. Test 5 and Test 6 show the results of using two reverse-tapered spacers (one is 1 inch thick and one is 2 inches thick) and a restrictor plate.
The interesting factor here is that when flow is reduced like this on the bench, you see it mirrored when the engine is on the dyno, unless the intake manifold is incredibly bad. The point is that you can destroy flow potential a lot more easily than you can generate it.
Flow Mods: Phase Three
The next step is to physically modify the carb in an effort to obtain extra flow. It took me about a day to modify the 850 unit in Figure 14.24, and the only tools I used were a set of needle files, a die grinder with a speed reducer for about 2,000 rpm, and some nearly worn-out 100-grit emery rolls.
In Test 1, the baseline flow is once again 849. For Test 2, and to demonstrate a point, I removed the base plate completely. This resulted in 943 cfm. This big increase indicates that the butterflies in the base plate constitute a major flow restriction. I so often hear, ”Well why didn’t you simply bore out the venturis?” The answer is that with any flow development project you start by improving the worst flow restrictions. The venturis of a carb are already super-efficient in terms of flow for a given area. The butterflies, on the other hand, are not as efficient. The butterflies and shaft assembly cause the 144-cfm flow loss.
For Test 3, I reinstalled the base plate and performed the first simple flow mod. This entailed removing the choke plate and its shaft from the choke horn. This move provided a 3.5-cfm flow increase (but because I am rounding to the nearest whole number it is recorded as 4 cfm). That is probably a lot less than you might have thought.
In just a few minutes working with a file, you can augment this mod by applying a radius to the top edges of the choke horn. This increases the flow about 3 to 5 more cfm. You may see carbs that have been reworked for a higher performance by milling off the choke horn. If the edges of the milled surface have been given only a cursory de-burr, these sharp edges actually detract from the carb’s flow potential.
If you have the choke horn milled from your carb, be sure to install a K&N Stub Stack made specifically for milled horns (see Figure 14.9). For Test 4, the form of the flowrestrictive butterfly and shaft assembly is addressed. The first move is to simply file off the excess and aerodynamically disastrous material from the ends of the butterfly securing screws. Note that as the carb leaves the factory these screw ends are staked to spread them so they don’t back out and get eaten by the engine. To guard against this after the screws have been filed off, they need to be removed one at a time and “locked in” with blue Loc-Tite.
The airflow test results on page 134 shows that simple move was worth 19 cfm. A point to note here is that each time you increase airflow by using a more efficient butterfly and shaft assembly you actually increase the booster activity. That means all this extra airflow has no downside as far as low-speed output is concerned. Test 5 involves slabbing the throttle shafts to reduce the crosssectional area presented to the incoming airstream. This is a popular move and is done on many highperformance carbs, either built by Holly or an aftermarket carb company. As the airflow tests show a loss of 13 cfm, you can safely conclude it is not all it’s supposed to be. I have reduced the area presented to the airstream but the introduction of corners has increased the shaft’s drag coefficient by a greater amount. The net result is a loss.
In Test 6, the shafts are aero formed (see Figure 14.18 and Figure 14.19). In doing so we are not only reducing the area presented to the airstream but also decreasing its drag coefficient. The numbers tell the story: an increase over Test 4 (stock shafts) of no less than 32 cfm!
Holley employs a number of different styles of butterfly securing screws. On the more basic, streetoriented carbs (such as our 850) the Phillips screw head is not particularly streamlined. Loading them in a lathe (or a securely held drill gun) and filing the heads to a dome helps. Reshaping the screw heads in Test 7 resulted in another 6 cfm of airflow.
Can you streamline the butterflies as done with the shafts? In this context, it’s called knife edging, which can make a big difference depending on the type of butterfly. Early Holleys had 1/16-inch-thick brass butterflies but the current models mostly have a steel butterfly about 0.030 inch thick. Knife edging is worth about 6 cfm on the earlier, thicker butterflies.
For Test 8, the butterflies were knife edged, which left about 0.010 of the original edge so as to still fit the bore. Fit is not an issue if you have time with a needle file so you can cut the knife edge right to the minimum. Test 8 shows that the half hour or so I spent knife edging the butterflies only paid off to the tune of 2 cfm. The only consolation is that it is 2 cfm in the right direction. It is clear that we have pretty much exhausted any gains to be had by simple grinding and/or filing. But I’ve shown the potential for improvement in flow without touching the boosters or enlarging the venturi.
The next question is, How much more can you can get if machining is an option? If you have a lathe or small mill, it is possible to machine the 1.75-inch butterflies on our 850 to take a 1.80-inch butterfly. On the 850 this can add about another 10 to 15 cfm. Let’s also consider using a smaller carb, such as a 750 with a 111⁄16-inch throttle blade bore instead of the 850’s 13 ⁄4-inch bore. By taking an 850 base plate and adapting it to a 750 it is possible, with nothing more than some cleanup on the venturis and boosters, to have a 750 flowing about 920 cfm. In so doing you can have all the low speed a 750 can deliver but with the top end of a 920. This is a move that many carb specialists make in the bid for more air without incurring a compromised low-speed output.
Once you have reduced restrictions caused by the butterfly/shaft assembly you can look at the venturis and boosters to see what improvements can be made. I am not talking about a big enlargement of the venturis because the 850’s 1.5-inch venturi is almost as big as you can go in a conventional 4150 casting. Any bigger and the signal (which is already marginal with a 1.5-inch venturi) just goes completely down the drain. Test 9 was confined to executing a good blend and cleanup job on both the venturis and the boosters. The results show what can be done without removing the boosters from the body. This is a situation that most people can relate to. Sure, removing the boosters can produce better results, but the tool to reinstall them costs about $300.
Figure 14.22 shows a steady hand with a die grinder can eliminate any venturi casting mismatch and clean up the booster venturis to produce a sharp edge at both leading and trailing edges. The result is 935 cfm. That’s an overall increase of 10 percent. The booster signal also rose by more than 5 percent. The installation of a Stub Stack for Test 10 pushed the flow to 956 cfm. That is an increase in flow of more than 12.5 percent. Test 11 and Test 2 are of the body alone. If you compare them you see that the gains made were very small.
That alone is a prime indicator that the hardware in the airflow is the prime restriction, not the venturi size.
So how much power increase can we expect from an increase in airflow with the modifications to our 850 carb? The answer to that depends almost entirely on how starved of air the engine is. I found dyno test results for our 850 that were also dependent on modifying a regular dog-leg booster to a stepped dog-leg version. On 450– to 550-hp small-blocks using a good dual-plane and making gains of between 5 and 15 hp are common. A big-block with a dualplane intake can show gains as much as 30 hp but 15 hp is more common. If the manifold is a single-plane, however, the effect of a carb that is too small is a little less pronounced. After a rework of a stepped dog-leg booster, it produces better results as proven in Figure 14.24.
Figure 14.24 shows a stock 850 versus a modified 850 (956 cfm in this particular case). Peak power rose by about 17.6 hp but the power hung on longer so there was an increase of almost 24 hp at 6,200 rpm. All the additional power came with no loss at any other part of the RPM range.
Written by David Vizard and Posted with Permission of CarTechBooks