That carb sizing formula is practically useless. It is simply a volumetric
relationship. As with any equation, you need to understand how it was derived
to understand it's limitations and how to properly interpret the results. The
formula is quite easy to derive since it's only a displacement relationship
(see attached derivation and commentary below). One needs to understand that
a carb's flow rating is defined at a specific test pressure drop. The industry
standard for four barrel carbs is a pressure differential equal to 1.5 inches
of mercury (Hg). What this means is a 4 barrel carb rated at 600 CFM will
flow 600 CFM of air, at wide open throttle, when a pressure differential of
1.5 In Hg is applied across it. This is just one point on a curve. When
installed on an engine, this same carb may flow more or less. If the pressure
drop is larger, the carb will flow more. If the pressure drop is smaller, the
carb will flow less. The 1.5 inches of mercury test pressure standard has
little to do with the pressure drop at which a carb will function reasonably.
A good carb will function well on quite a bit less than 1 inch of Hg drop, a
bad one may need two or three inches. 1.5" Hg pressure drop is too much for
a highly tuned engine. 0.7" is considered non-restrictive, though racers may
shoot for 0.5" Hg. Two barrel carbs are usually rated at a different pressure
differential (3.0 In Hg). The reason for the 1.5 In Hg pressure drop is mostly
historical. When 4 barrel carbs first came into popular use, the vacuum pumps
used to rate 2 barrel carbs were unable to pull the same pressure differential
across a 4 barrel carb, so 4 barrels were rated at a lower pressure drop.
There's no such thing as a carb that flows too much, just carbs that don't
atomize well or have a poor air-fuel curve. What's really important is the
atomization the carb provides to the fuel-air mixture and the restriction
(pressure drop) it imparts to the incoming airflow and how well the carb
controls the air-fuel ratio across the rev range. By paying attention to the
aerodynamics of a carb, you can increase it's flow without decreasing the
atomization. Such a carb will increase power (assuming the engine can utilize
the extra flow) without hurting throttle response or fuel economy. Simple
things like a dyno stack or a K&N stub stack can increase carb flow too.
Atomization is a strong function of the booster design and the venturi diameter.
The larger the diameter, the slower the flow and the poorer the mixture.
If you can increase the flow without increasing the minimum area, you get
a higher flowing carb without the down-sides. That's what the tuner carbs
do. For carbs like that, the standard carb sizing formulas just do not apply.
We run a Holley HP950 on the dyno that works better across the rev range on
a 351C than a standard vacuum secondary Holley 4150. Another thing to keep
in mind is that it's important to route cool air to the carb. 1.5 inches of
Hg pressure drop across a carb represents a 5% loss in air density (and
therefore horsepower) at sea level. That's equivalent to going from 80° F
inlet air up to 107° F inlet air.
Be aware that some carb manufacturers and tuners may refer to a carb as a
750 if it started with a 750 main body. After streamlining (by milling the
choke tower off, smoothing the main casting, narrowing the booster legs,
replacing the straight-leg boosters with dog legs, thinning the throttle
shafts, countersinking any screw heads, fitting a different baseplate, etc.),
that 750 body may flow 830 or more CFM, when tested at 1.5" Hg pressure drop.
Here's the carb sizing relationship and a brief discussion I wrote a while
back to try to illustrate the pitfalls of the sizing equation. The forum
software will likely remove my spaces and screw up the formatting but here
goes:
DISP RPM
CFM = ---- * ---- * VE
2 1728
where:
DISP = engine displacement in cubic inches
CFM = required carb flow in cubic feet per minute
RPM = maximum engine speed in revolutions per minute
VE = volumetric efficiency (dimensionless, 1.0 = 100%)
1728 = conversion factor between cubic inches and cubic feet
= 12*12*12
2 = conversion factor for four stroke engine
This equation can be simplified to:
DISP * RPM * VE
CFM = ---------------
3456
Note this sizing formula is simply a relationship between cylinder volume and
the flow required to fill that volume at a given engine speed. Also note, for
a four stroke engine, displacement is divided by two because an intake stroke
occurs every other revolution. While, it's easy to determine displacement and
maximum rpm, you'll probably have to guess at the third variable, volumetric
efficiency (VE), unless you have access to a dyno. Volumetric efficiency is
simply a measure of how efficiently an engine fills its cylinders. A stock,
low performance, street engine may have a VE between 0.7 and 0.8. High
performance street engines may fall between 0.8 and 1.0, while highly tuned
race engines can have VE's exceeding 1.0, perhaps as high as 1.25.
One other thing to understand when using the formula above is that a carb
will only flow in the presence of a pressure differential. On one side of
the carb there is atmospheric pressure and on the other side is manifold
pressure (usually referred to as manifold vacuum since it is typically lower
than atmospheric pressure). Since engines vary in their manifold vacuum
characteristics, a standardized pressure differential was established to
provide a meaningful comparison for different carbs. Before this standard,
venturi size was used for comparison. The standard for four barrel carbs is
a pressure differential equal to 1.5 inches of mercury (Hg). What this means
is a 4 barrel carb rated at 500 CFM will flow 500 CFM of air, at wide open
throttle, when a pressure differential of 1.5 In Hg is applied across it.
When installed on an engine, this same carb may flow more or less. Two barrel
carbs are usually rated at a different pressure differential (3.0 In Hg).
The reason for this is primarily historical. When 4 barrel carbs first came
into popular use, the vacuum pumps used to rate 2 barrel carbs were unable to
pull the same pressure differential across a 4 barrel carb, so 4 barrels were
rated at a lower pressure drop.
Flow ratings from one standard can be related to flow ratings from another
standard. For 2 and 4 barrel carbs:
Flow @ 1.5 In Hg = (CFM Rating @ 3.0 In Hg)/SQRT(3.0/1.5)
Which is approximately:
Flow @ 1.5 In Hg = (CFM Rating @ 3.0 In Hg)/1.414
This relationship is derived from the fact that, for incompressible flow
(assuming subsonic flow... flow will choke at Mach one), the volumetric flow
rate through a venturi is proportional to the square root of the pressure
differential:
Q = K1*A2*SQRT(2*Gc/Rho)*SQRT(P1-P2)
or more simply:
Q = K2*SQRT(P1-P2)
where:
Q = volumetric flow rate
K1 = flow coefficient
A2 = downstream area of the venturi
Gc = gravitational constant
Rho = density
P1 = inlet pressure
P2 = pressure at venturi minimum area
K2 = K1*A2*SQRT(2*Gc/Rho)
Computing the relationship for volumetric flow rate at the two flow
differentials and equating yields the conversion formula.
As an example of using the sizing formula, let's say we have a modified 4.1
liter (252 cubic inches) Buick V6 with a VE of 0.9 and we plan to turn no
more than 6400 rpm. Plugging our numbers into the formula yields a
theoretical estimate of:
252 * 6400 * 0.9
CFM = ----------------
3456
= 420 CFM
In practice, Joe Murawski of the Wedge list runs a 4.1L Buick in his Triumph
TR7 and has tried a variety of carbs, in sizes ranging from a Holley 390 to a
785 CFM Quadrajet, settling on a 500 CFM Edelbrock/Carter AFB as providing the
best power and driveability. His carb choice is somewhat larger than that
predicted. For reasons discussed below, we'll see this is not unusual.
While the formula above may yield useful estimates, it is not necessarily the
ideal it is often portrayed to be. If you have a carb that can flow 500 CFM
in the same application and still properly atomize the fuel, it should make
more power than the 400 CFM carb. From this perspective, larger is better.
Ideally, a carb would present zero restriction to the intake stroke. Such a
carb would have an infinite flow rating. Unfortunately, carbs require a
pressure differential to properly mix fuel with air, which is why carb sizing
is important (and why the above formula is useful, if used in a modified
form). Keep increasing the size of a carb and, at some point, the booster
venturis will not properly atomize the fuel/air mixture and droplets of liquid
fuel will be pulled into the cylinders. Not only is this bad for performance,
it's also hard on the engine. The liquid fuel tends to wash oil off the
cylinder walls, increasing ring and bore wear. This is a particular problem
with engines using large overlap cams, since they provide lower vacuum levels.
When using a larger carb and cam, proper tuning (carb and ignition) becomes
more important.
It's important to understand the basic sizing formula is just a guideline.
It ignores a number of important factors such as manifold design, cam timing,
vehicle weight, gearing, transmission type, intended usage, etc. Furthermore,
it ignores important differences in carb design like venturi efficiency, bore
layout, and secondary style and method of actuation. In practice, I have found
that the above formula applies mainly to square bore carbs with non-air valve
secondaries (Holleys, Autolites), and even then it can be conservative for a
performance application. It typically yields a compromise of fuel efficiency
and power.
Using a dual plane, divided plenum, intake usually allows the use of a carb
with a larger CFM rating than with a single plane, open plenum, intake. This
is because the divider cuts the effective plenum volume in half, increasing the
signal to the boosters. Because of this fact, Edelbrock suggests multiplying
the CFM predicted by the basic sizing formula by 1.1 to 1.3 for single plane
manifolds and by 1.2 to 1.5 for dual planes.
As another example, consider the engine I used to run in my Detomaso Pantera.
It's a 351C Ford with Aussie 2V quench heads, 1 3/4" headers, and a single
plane, open plenum, Weiand Xcelerator intake manifold. Since I retained the
stock cast pistons, I chose a cam with a shift point of 6000 rpm. As a guess,
pick 0.9 for the VE. Since the Pantera is relatively light with short gearing,
pick the high side of the range for K (the intake factor):
K*DISP * RPM * VE 1.3*351*6000*0.9
CFM = ----------------- = ----------------
3456 3456
= 713 CFM
This agrees with my real world experience with Holleys on street modified 351C's.
600 CFM carbs provide good throttle response and fuel economy but can up 20+ peak
horsepower to 750 carbs. On the downside, the usual vacuum seconday Holley 750
with it's lame straight leg boosters hurts fuel economy and has poor throttle
response. The 735 Holley I ran was a happy medium with good throttle response,
power and 20+ MPG on the highway. The big difference with the 735 Holley is it
uses the excellent Ford-designed skirted truck boosters. Note we're referring
to stock Holley carbs here, not custom models with milled choke horns, thinned
booster legs and cross shafts, knife-edged butterflies, streamlined main bodies
and improved booster designs. Those modified carbs can flow more mixture, while
providing adequate atomization. Examples are the various tuner carbs, Holley's
HP series, Quick Fuel technologies, the Demon series etc. For those carbs, the
sizing formulas don't apply very well.
I chose a Holley 735 from a 428CJ application which seems to work well. This
carb, while flowing nearly as much as a 750, has a venturi cluster design that
provides a stronger signal. Throttle response and fuel economy are relatively
good (20+ mpg on the highway), without incurring a noticeable power penalty.
Two other important considerations are bore layout and method of secondary
actuation. Carbs with air valve secondaries (Rochester Quadrajets and
Carters), especially those with spread bore layouts (Thermo Quads, Quadrajets),
can usually be sized larger than square bore Holleys and Autolites. This is
because the smaller primaries increase the flow speed through the boosters,
providing better atomization, while the air valve secondaries passively
restrict air flow until the engine can handle it. Taking these two factors
into consideration, Vizard suggests the following two rules of thumb for
street performance engines where power is more important than fuel economy.
For air valve secondary carbs with an upper rpm limit of 6000 rpm, use:
CFM = 2.3 * DISP
For square bore non-air valve secondary carbs use:
CFM = 2.0 * DISP
For engine speeds above 6000 rpm, multiply by the ratio of maximum rpm to 6000
rpm. Note the second formula yields 702 CFM for my 351C example, which is
close to the basic sizing formula with the intake manifold correction factor
applied. Note that the stock 4300D Motorcraft spreadbore used on 1972 to 1974
351C-4V's was rated at 715 CFM (or 750 CFM, depending upon who you believe).
As an extreme example, I've successfully used a 750 CFM Quadrajet on a
relatively stock 231 cubic inch Buick V6. With the Qjet, it got slightly
better fuel economy than the previous 2 barrel carb (due to the small
primaries) and had noticeably more power (due to the huge secondaries). The
driveability of the carb was fine with no bogs or flat spots. On the V6, I'm
sure it never pulled anywhere near the 750 CFM rating but it did pull what it
required. You could never put a Holley 750 on a little low compression V6 and
expect to make it work. The air valve secondaries allow the use of much larger
CFM ratings without incurring driveability problems. There is a price to be
paid however. Even when wide open, air valve secondaries are slightly more
restrictive to airflow than non-air valve secondaries.
While these formulas should help you choose a carb flow rating, nothing beats
trial and error. Also, once you have a carb installed, you can determine how
restrictive it is by using a vacuum gauge to measure the difference between
atmospheric pressure and the pressure under the carb. With the air cleaner
removed, the air above the carb will be essentially atmospheric. If there's
any difference between it and the pressure sensed under the carb, it's due
to the carb. The higher the difference, the greater the restriction.
Measurements should be made at wide open throttle and 0.7 inches of mercury
is considered non-restrictive.
An alternative to the carb sizing formulas is to realize 100 HP requires 140
CFM based upon a reasonable assumption for Brake Specific Fuel Consumption
(BSFC). The BFSC assumption keeps us from having to guess at volumetric
efficiency. A 550 hp engine uses an actual 770 CFM but you need to convert
the pressure drops. 4 barrel carbs are rated at 1.5" but that is too
restricitive. 0.7" is more reasonable for a tuned engine to keep the carb
from being overly restrictive.
Flow @ 0.7 In Hg = (CFM Rating @ 1.5 In Hg)/SQRT(1.5/0.7)
770 = X / 1.46385
X = 1127 CFM flow rating required
A 350 hp engine uses an actual 490 CFM but that doesn't mean a 490 CFM rated
4 barrel carb will provide the required flow. It takes 717 CFM at 1.5" Hg to
equal that 490 CFM at 0.7" Hg pressure drop:
Flow @ 0.7 In Hg = (CFM Rating @ 1.5 In Hg)/SQRT(1.5/0.7)
490 = X / 1.46385
X = 717 CFM flow rating required
Note the sizing formulas are derived with an implicit assumption of a plenum
manifold and do not apply to independent runner manifolds. Completely
different duty cycle. A plenum manifold allows multiple cylinders to share
the total flow of the carb. On a plenum intake, each cylinder gets to draw
from each barrel (if single plane) or half the barrels (if dual plane) but,
on an independent runner intake, it's one barrel to one cylinder. Independent
runner applications require much larger total CFM because of this. Also, 4
barrel carbs are sequential in that only the front 2 barrels operate at low
demand so 4 barrel carbs can be sized larger without much adverse effect on
low RPM torque.
Dan Jones