Iron Melting Cupola Furnace by Stephen D.Chastain Text

Iron Melting Cupola Furnaces
For the Small Foundry
Written and Illustrated by Steve Chastain

Steve Chastain
By Stephen D. Chastain
Copyright © 2000 by Stephen D. Chastain.

Jacksonville, FL all rights reserved
Printed in USA
ISBN 0-9702203-0-8
234567890
Table of Contents
Purpose & Introduction 8
I. Cupola Theory and Design Considerations 11
II. Building a 10 inch Cupola 28
III. Cupola Operation 49
IV. Air Supply and Blowers 59
V. Designing Centrifugal Fans 67
VI. Construction of Centrifugal Fans 77
VII. Construction of a Pitot Tube and Manometer 95
VIII. Calculation of Air Flow 101
IX. Additional Cupolas based on the 15” Shell 107
X Oxygen Enrichment 111
XI. Purchase of Coke 115
XII. Conclusion 116
Appendix: Suppliers 118
Airflow through pipe 120
Index 122
WARNING - DISCLAMER
This book is to provide information on the methods used by the

author to develop a cupola. The author is neither an engineer
nor scientist. No liability is assumed for the use of the
information contained in this book. If you do not wish to be
bound by the above, you may return the book for a full refund.
Warning: As with any fire, the incomplete combustion of coke

produces carbon monoxide (CO), a poisonous gas. Carbon
monoxide bums off harmlessly to carbon dioxide in air. The
flame at the top of the cupola is the burning carbon monoxide.
The auto-ignition temperature of CO is about 1200 degrees F. If
you use a higher stack for better charge preheat, the discharge
temperature may be as low as 400 degrees. In such case it may
be desirable to install a gas pilot to be sure the gas ignites upon
leaving the cupola, and or you may maintain a boxer fan in the
work area to provide additional ventilation. A regenerative blast
preheater was deliberately left out of the book because of the
hazards of pumping CO about the work area.
MATH: The math used in this book is no more complicated

than basic high school algebra. Any one should be able to do the
computations with a $9 calculator and a half-hour of practice
(read the instructions for the calculator). As you read the book
you should work the examples on the calculator. You will see
just how easy it is. You will be missing out on the ability to
customize your cupola if you don’t. Get over your math fears
and do it. Good luck!
FOREWORD
Since I wrote “BUILDING SMALL CUPOLA FURNACES"

five years ago, I have heard back from many people building
them in both this country and other places around the world.
Probably the most successful of any of these was Steve
Chastain. He approached this project in a very scientific manner
and his first results more than justified his great efforts; he had
perfectly hot usable iron on his very first heat! I was so
impressed with Steve’s results that I encouraged him to put his
work on paper for the benefit of the rest of us. He has now done
so, and I know you will be very pleased with his thorough book
and drawings. If you are building a cupola of about 10 in bore,
you will find meticulous instructions that allow you to duplicate
Steve’s own amazing furnace; and I am sure you will agree this
book is a terrific bargain. Happy and safe melting to all.
Stewart Marshall
17th May, 2000
About the Author:
I grew up in central Florida, and have always been considered

a creative person. While attending Florida Southern College, I
studied economics, music, and theater. Later, I started my own
band that toured nationally for several years. While on tour, I
noticed the inadequacy of the commercially available
production equipment. I designed and built my own line of
staging equipment, later forming SDC Electronics to
manufacture and sell the equipment. I left the music and staging
businesses to pursue a career in engineering. At the time of this
writing, I am entering my junior year as a mechanical
engineering student. While in school, I have been working as a
lighting consultant, rebuilding industrial lift trucks, and running
a small foundry and machine shop that produces parts for
antique engines and equipment.
One of my theatrical
dimmers. This rack
Contains thirty-six
2.4 kW dimmers
ACKNOWLEDGMENTS
I would like to thank Dave Gingery and Stewart Marshall

for convincing me I had something to say. They provided
support during the writing process. I want to thank my
neighbors, especially Randy Smith and his family for putting
up with the perpetual experiment in my yard. Jim Brown,
Ph.D. always comes over when I need a hand with the
operation of the cupola. Peter Reich introduced me to true
cupola authorities, Norm Lilleybeck and Rod Schueler, who
reviewed an early manuscript and encouraged me to
continue. William Provis of Modem Equipment Company
provided additional research material, which also helped
confirm my figures. Most of all, I would like to thank my
wife, Gina, who endures my long hours and keeps a positive
attitude.
PURPOSE:
The purpose of this book is to provide a set of plans for a

cupola that will produce 35 - 55 lbs. of iron every 8 to 12
minutes. The cupola will be built from scrap. The cost will be
$50 to $200 depending on the type of refractory used. The book
will describe construction, setup, and basic operational
parameters. Several factors affect operation. Height and size of
tuyeres, size of windbelt, location of tap and slag holes, and
minimum height of stack are critical for proper operation. Other
critical parameters include size and weight of iron and coke
charges. The formula must be followed exactly or you will have
poor operation. Brief theory and calculations describe larger
cupolas, up to 18” in diameter.
This book is not intended to be a foundry manual. The
minimum foundry experience required for this project should
include competence pouring non-ferrous metals. There are
several good books available from Lindsay publications and
Stewart Marshall regarding foundry work.
INTRODUCTION:
Old engines fascinate me and I love to restore old machines.

Many engines and machines are scrapped because of a broken
casting; parts are extremely expensive, or nonexistent. I had
several pieces that could be fixed if I could cast the parts. My
first project was a set of pistons for an antique four-cylinder
generator set. Then came manifolds, waterpumps, magnetos,
and cylinder sleeves. As friends found out I could pour iron,
several tractor and machine projects followed. Lathe accessories
followed by a cam grinder were designed and built. Now there
seemed to be no limit to what I could do. I have always felt I
could have anything I wanted, I just had to build it. The cupola
lets you restore old machines or design new ones. You have a
8
sense of freedom when you can design, cast, and build any
piece of equipment you can dream up. Man has been pouring
iron for thousands of years. It is not nearly as difficult as you
might think. Good luck, I hope you enjoy your cupola as much
as I enjoy mine.
Machining my first Piston Casting
This piston went in a 10 kW generator set that powers my house
9
Stack
Cutaway of Cupola
10
L
CUPOLA THEORY AND DESIGN CONSIDERATIONS:
Cupola Zones:
There are five areas in a cupola. These areas are the well,
combustion, reduction, melting, and the preheating areas.
The well is at the bottom of the cupola. It stores the liquid
iron until the cupola is tapped. Because of the close contact
with the coke and exposure time, most of the iron’s carbon pick
up occurs in the well. The depth of the well affects the
temperature of the tapped iron. Hotter iron comes from a
shallow well. Slag separates and floats on top of the iron in the
well.
The combustion zone is where the blast enters the cupola
and reacts with the coke to form carbon monoxide and carbon
dioxide. The heat of reaction superheats the iron and generates
hot gases, which melt and preheat the charge. This is the area
where the thermal energy or heat is generated.
The reduction zone is outside the combustion or oxidation
zone. The iron is superheated in this zone. Oxides of iron are
also reduced here.
The melting zone is the area from the top of the coke bed to
where the iron actually melts. The melting zone is located
between the reduction zone and the preheat zone.
The preheat zone is above the melting area and extends to
the top of the charged material. The temperature of the charge
must rise from ambient or room temperature to melting
temperature. The charge receives its largest amount of heat
gain in this area. Gases entering this region are approximately
2200 degrees Fahrenheit, however after giving up their heat to
the charge; they exit at 400 to 800 degrees Fahrenheit. Do not
underestimate the importance of the preheat area. Good melting
11
depends on adequate preheat. Other functions of the preheat
include the drying of the charge materials.
The cupola stack should be kept full so that charge materials
can be adequately preheated. The lower in the stack the charges
sit, the lower the melting zone and cooler the metal. The height
of the stack is limited by the ability of the coke bed to support
the weight of the charges and the ability of the blast and hot
gases to penetrate the charges. The efficiency of the cupola is
largely dependent on the heat transfer between the rising hot
gases and the descending charges. The longer the contact time,
the better the heat transfer. Thus a higher bed and stack are
superior to a low bed and short stack. Short stacks do not allow
sufficient preheat.
Pouring a 60 pound faceplate for a lathe
12
Cupola Zones
13
MATERIAL SIZE:
The size of the metal making up the charge is critical to

maintaining efficient preheating and melting. If the pieces are
too large, they will not melt in the melt zone but will work their
way down to the tuyere level. The iron will not superheat and
this could lead to a frozen taphole. If small pieces are charged
without some type of grouping, they often burn up or will clog
up the voids in the stack. This could lead to bridging in the
cupola. Small and thin pieces should be gathered and pressed
into a briquette. Borings may be melted by filling short lengths
of pipe and plugging the ends A good size for the 10 inch
cupola is 2 by 3 by !4 inch.
Coke must be screened to proper size, being 1/10 to 1/12 the
inside diameter of the cupola. The space between larger pieces
allows the gasses to rise faster is the stack. This raises the
combustion zone thereby reducing the preheat of the charge.
Smaller pieces could pack too tightly, increasing resistance to
the gas flow. Fines must be eliminated as they end up being
blown out of the stack as a shower of sparks (sometimes rising
60 feet or more) or come down as dust covering the whole area.
If limestone is used as a desulfurizing flux, it should be in
pebble form. Larger pieces will not decompose properly.
However, if it is too fine, it will clog the furnace or be blown
out of the stack.
By varying the iron to coke ratio and the amount of air, any
cupola is capable of a wide range of melting rates and melting
temperatures. In all cases, the air and coke must be balanced to
maintain consistent melting.
When the coke and air are unbalanced, melting problems
arise. Too much coke results in a rising bed, wasted coke, slow
melting, and lower temperatures. Too much air burns away the
bed, lowering the temperature and leaving free oxygen in the
cupola to bum the iron. A “sparky tap” indicates too much air.
14
Ideally, stack gases should contain 13 percent carbon dioxide,
13.2 percent carbon monoxide, and 73.8 percent nitrogen. The
balance of coke and air is critical for proper operation. Metal
temperature is a direct result of a properly maintained bed
height. A proper bed will allow taps at 2750 degrees. However,
imbalance in the cupola will lead to lower temperatures.
COMMON CUPOLA RATIOS:
Conventional industrial cupolas are built to melt from 2 to

32 tons per hour. Their ratios offer a place to start the design
process. However, the surface area to volume ratio AJV is
higher for our small cupolas, putting the 10-inch cupola at the
higher end of the thermal loss scale. Small cupolas are not as
efficient as large ones.
Thermal Balance Sheet:
Of all the energy contained in the coke only 25% to 65% is

used to heat the iron charge. The remaining 35% to 75% is lost.
The losses are as follows:
% total heat available in charge
Lining and Tuyeres 4 to 20
Stack Discharge 6 to 30
COproduction 17 17
(incomplete combustion)
Slag_8 to 8
TOTAL LOSSES: 35% to 75%
HEAT LEFT FOR MELTING: 65% to 25%
15
Conventional cupolas melt 10 pounds of iron per square inch
of area between the tuyeres per hour. The 10-inch cupola could
theoretically melt 785 pounds of iron per hour.
Melt rate in pounds = 10 7t (D/2)2 pounds / hour

hr
D= diameter
7t = 3.14
D = 10 inches
Melt rate = 107t(10/2)2 = 785 pounds / hour
The maximum ratio of iron to coke is 10 to 1; ten pounds of

iron per pound of coke. Our 10-inch cupola is melting at a 6 to
1 ratio. Best cupola operation occurs with incomplete
combustion. The stack gasses should contain 13% C02 (carbon
dioxide) 13.2% CO (carbon monoxide) and 73.8 % N
(nitrogen). The CO burns to C02 as it is discharged from the
stack, giving a large visible flame if melting at night.
If our 10-inch cupola is melting 330 pounds per hour, we can
calculate the amount of air required. One pound of carbon
requires 113 cubic feet of air to provide the 13% CO2 - 13.2%
CO ratio. Coke contains approximately 90% carbon. What are
the air requirements for the 10-inch cupola?
330 pounds iron X 1 pound coke X .9 pound carbon X lhour =.,825. pound C
hour 6 pounds iron 1 pound coke 60 min min
(113 cubic feet air / pound coke) X .825 pound coke = 93.2 cubic feet air
The shop vac used on our cupola is rated at 90 cubic feet per
minute. The calculations appear realistic for our situation.
16
TUYERE AREA
The tuyere area is based on the inside diameter of the cupola

at the tuyere level. Standard ratios for smaller cupolas range
from Vi to 1/6 the cross sectional area of the cupola at the
tuyeres. Earlier practice favored larger tuyeres that increase in
size as they approach the inside of the cupola. The coefficient of
discharge is much higher for this type of nozzle. However,
modern practice favors smaller tuyeres and higher blast
pressures. The theory is, as pressure increases, discharge
velocity at the tuyeres increases. Blast penetration of the bed
will increase thereby lowering the combustion zone.
Combustion Moves Toward Moves Away

Zone Center From Walls
And Lower And Lower
Large Smaller Angled

Tuyeres Tuyeres Tuyeres
Higher Higher
Pressure Pressure
Theory of Blast
17
As the blast increases the combustion zone will move closer to
the center of the cupola, reducing the heat loss through the
walls. It should also be possible to further lower the combustion
zone by angling the tuyeres downward.
Calculate the tuyere area for the 10-inch cupola using the ratio:
1/6 the sectional area at tuyere level.
D = 10 inches A= tt(D/2)2
A= area
7t = 3.14 A= 7t(10-inches/2)2 = 78.5 inches2
Tuyere area = 1/6 A = 1 X 78.5 inches2 = 13.1 inches2

6
13.1 inches is the required area. Now to find a suitable diameter

pipe-
area of 2-inch pipe = 7i(D/2)2 = iz(2/2)2 = 3.14 inches2
Number of tuyeres = (13.1 inches2/3.14 inches2) = 4.17
Area of 2 Vi-inch pipe = 7t(2.5 inches / 2) = 4.9 inches
Number of tuyeres = 13.1 inches2 = 2.66

4.9 inches2
4 tuyeres at 2-inches diameter are closer to the desired ratio than

3 tuyeres at 2 Vi-inches diameter. 2-inch pipe will be used.
18
CUPOLA HEIGHT:
Cupola height (before legs) is determined in two steps. The
stack height above tuyeres and the well depth + constant make
up the two calculations. For small cupolas a constant of 5-
inches will be used. Larger cupolas use 10 inches. The well
depth will be considered first. Deep wells increase the difficulty
of maintaining hot iron; therefore a conservative (shallow) well
will be used for our experimental cupola. The theoretical
capacity of the cupola is 1 pound per square inch of sectional
area at the tuyeres. For a 10-inch cupola this tap would be
approximately 78 pounds. A tap of 35 pounds should certainly
be attainable, therefore we will calculate for a 35-pound tap.
The density of iron is approximately 4.4 ounces / cubic inch.
Ounces of iron = 35 pounds iron X 16 ounces = 560 ounces iron

1 pound
Cubic inches of iron = 560 ounces of iron = 127 inches3

4.4 ounce / inch3
Well volume available for iron = .46 total well area
Total well volume = 127 inches3 = 277 inches3

.46
Height to slag hole (from bottom of well) = total well volume

area of well
Area of well = 78.5 inches2
Height to slag hole = 277 inches3 =3.52 inches

(from bottom of well) 78.5 inches2
19
Height to tuyeres includes the (height of the 2-inch sand
bottom)+ (3.52 inch height of the slag hole) + (5-inches
(constant))
Height to tuyeres = 2 inches + 3.52 inches + 5 inches = 10.52

inches.
HEIGHT ABOVE TUYERES:
Height above tuyeres is a simple calculation. As shown earlier

higher stacks are superior to shorter stacks. The minimum stack
height is 4.5D at tuyere level. Between 6.2 to 6.5 is considered
optimum.
Calculate the height above tuyeres using the 4.5D ratio.
4.5 X 10-inches = 45-inches
The total height of the cupola less legs = sum of both heights +
tuyere diameter.
Sum = height to tuyeres + height above tuyeres + tuyeres
Z = 10.52 inches + 45 inches + 2 inches = 57.52 inches
Height of cupola without legs = 57.5 inches
Minimum leg height = length of door + 6 inches + depth of sand

bed covering the foundation. Leg height may be adjusted
upward from this figure to increase operator comfort.
The ratio (1.3 X total tuyere area) gives minimum sectional area
of the windbelt, (1.75 X tuyere area) would be the maximum for
this type of cupola.
20
CALCULATION OF THE BED HEIGHT:
Because any cupola can be operated over a wide melting

rate, actual bed height depends upon the operation of the
cupola. For larger cupolas the height is given by:
6 + 10.
The constant of 6-inches is added to the minimum height and

represents the maximum height of the bed. The constant may be
adjusted; however a higher bed usually will give hotter iron.
The minimum height of the bed for the 10” cupola using 5”
blast pressure is calculated as follows:
10.5 15 inches = 17.85 inches

V 1.73
CALCULATION OF CHARGE WEIGHTS:

The weight of coke used per charge is estimated by:
.0787 lbs./inch 2 x Sectional Area inch'
The weight of iron charges is proportional to the weight of the

coke charges. Common ratios are from 6:1 to 10:1.
Calculate the charge weights for the 10” cupola:
Weight of coke: .0787 lb./inch2 x 78.5 inch2= 6.18 lbs.
Weight of iron at a 6:1 ratio: 6 x 6.181bs. = 37 lbs. iron
CALUCLATIONS FOR 18 INCH CUPOLA:

D = 18 inches
2
Area = 254 inches
Maximum theoretical* tap = 254 pounds
21
163 pounds = 64% of theoretical tap
163 pounds = 2608 ounces
Cubic inches iron = 2608 ounces = 592 inches3

4.4 ounce/inch3
Total well volume = 592 inches3 = 1288 inches3
.46
Height to slaghole = total well volume
sectional area
Height to slaghole = 1288 inches3 = 5.07 inches

254 inches2
Height to tuyeres = X bottom, tuyeres, slaghole, 5 inches
Sand depth on bottom = 3 inches
Height to tuyeres = (3 + 5 + 5.07) inches = 13 inches

Tuyere area = .2 sectional area = 51 inches2
2-Vz inch pipe has approximately 4.9 inches2 area therefore 10
tuyeres would give 19.3 % of sectional area.
Windbelt sectional area = 1.3 to 1.6 tuyere area
1:1.3 = 58 square inches, dimensions 6 by 9 5/8 inches
Height above tuyeres = 4.5 D at tuyeres
4.5 X 18 inches = 81 inches
Total height without legs = 81 inches + 2 Zi inches + well
Height (163 lb. well) = (81 + 2.5 + 13.07) inches = 97 inches
*in modem commercial practice well capacities may be up to

400 lbs.
22
_Well Depths for 18 inch Cupola_
18 inch cupola 3 inch sand bottom
height to slaghole height to tuyeres
from sand bottom_from base plate
150 pound well = 4.67 inches 12.67 inches

Use the 6.5 D height ratio and angle the tuyeres downward 15
degrees for the larger wells.
73
sUi 91
* -
ibi-
'Wl
Dcc-Ot 1998
n.
BUILDING A 10 INCH CUPOLA
MAKE A CIRCLE CUTTING ATTACHMENT:
This project requires several precision circles to be cut with a

torch. The first thing to do is make a circle cutting attachment.
Each torch is different so the drawing is not dimensioned.
The body is made from 1-inch square aluminum stock. I
Circle Cu++i
cement
28
poured my stock while making other aluminum castings. Square
the stock and drill the holes for your particular torch, the hinge
and screws. Mill or saw out a slot to receive the hinge block.
Cut the hinge block from Vi inch steel plate and drill it to
receive the hinge pins. Clamp the body in a vise and cut it open
with a hacksaw. Tap the screw holes Vi - 20 and install the
Completed Cutting Attachment
hinge. Clean up the assembly with a file for a snug fit on your
torch. Various radius arms are made from drill rod. Heat the end
until red, then bend the right angle. Heat again until non¬
magnetic, then quench to harden. Clean the tip with emery cloth
and dip in oil. Heat the tip until straw brown and quench again,
this tempers the tool. When cool, grind the tip to a point. I used
Vi” drill rod, however it is not as rigid as I would like; 5/16” rod
might be better.
29
THE CUPOLA SHELL:
The shell should be the first thing you get. It needs to be

aired out before you start cutting, so get it first then start
looking for other materials.
Cut the shell from discarded 100 lb. propane cylinders.
These can be found at any propane supplier. Once the propane
cylinders’ bottoms become pitted by rust, they are removed
from service. Most gas dealers are glad to get rid of them once
they become rusty. While you are there, have them remove the
valve. You can also do it at home with a pipe wrench that has a
3-ft. pipe over the handle for leverage.
Cylinders come in 2 types. One is rolled from a single
3-ft. sheet with the top and bottom welded on. (I prefer this
type for the bottom) The other type is pressed out of 2 tubes
that are then welded together in the middle. I used this type for
the top part of the cupola. It took me several months to collect
all of the materials from the scrap heap, so my tanks were well
ventilated by then.
Any welder knows cutting into a tank can cause an
explosion, so always fill the tank with water before cutting it.
Cut the top and bottom from one cylinder. Cut the other
cylinder around the middle. Save the top half of the second
cylinder for the top of the cupola. Plug the valve hole in the
short top section and cut a 10-inch hole in the lid using the
circle cutting attachment. Clean up the edges with a hand
grinder (I use a 4 inch and 9 inch grinder). The 9-inch grinder
makes fast work of this job.
To remove the paint from the cylinders, paint them with
paint stripper. A single edge razorblade used in a gasket
h- H
removing tool makes this messy job go much faster. Once these
are clean, drill the tap and slag holes 1 inch in diameter.
Tuyeres are the openings in the sides of a cupola that admit the
blast of air. Locate the tuyere holes so that the bottom of each
hole is 11 inches from the base plate. Cut the holes using a hole
saw. It will be slightly larger than the 2-inch pipe diameter.
That is fine since you will need the room to angle the tuyeres
down 10 degrees. Next, build and install the wind belt. When
the wind belt is completed, weld the lugs on the bottom of the
shell that are used to bolt on the base plate. The six lugs used
are made from 1 inch long, 1 inch in diameter rod chucked in
the lathe and drilled through the center with a 9/16-inch bit.
Weld three more lugs on the top of the shell. Stack the short
section on the shell. Bolt another lug separated by washers to
each of the 3 lugs you just welded. Weld the new lugs to the
top half; these will be used to draw the top half down to form a
good seal in the uncured refractory.
Bend rebar handles using a torch to heat the bent-area red
hot, quickly clamp in a vise and bend. These were sized to be a
comfortable fit to my hand while wearing welding gloves.
Weld the handles to the top of the shells. Weld an extra set on
the bottom of the lower shell.
THE WIND BELT:
The wind belt wraps around the cupola to deliver air to the
tuyeres. Cut the top and bottom of the wind belt from Vi inch
plate using the circle cutting attachment. You could use thinner
material as long as you are able to weld it without burning it up.
I find myself standing on the wind belt during set up, so I am
happier with the thicker stock used here. Once cut, clean them
up with a hand grinder. Mark the shell for the top section of the
wind belt then turn the shell upside down. Put the top ring of
the wind belt over the shell and check the fit. Mark the high
spots on the ring, grind them and check the fit again. This could
take several Trials. Once fitted, you should be able to pound it
32
into place with a 2-lb. hammer. Once the first ring is in
position, tack weld it into place. The tuyeres should be cut 2 Vz
inches long.
Install the tuyeres maintaining the 10-degree angle towards the

base plate. The innermost bottoms of the tuyeres should now
be 10 Vi inches from the base plate. Be sure to maintain the 10-
inch diameter between the tuyeres or you will not be able to fit
the lining form in place. You will want to radius the inner edge
of each tuyere with a hand grinder before you install them. You
can do your final fitting after they are welded in place. Fit the
remaining ring and pound it into position maintaining 6 inches
between the top and bottom plates. Tack it into place. Measure
and mark the locations of the tuyeres on the ring so you can find
and locate the peepholes later. The belt is wrapped with strips of
33
16-gage sheet steel. It is difficult to maintain a perfectly straight
strip while welding the wind belt so cut four strips 17 inches by
7 inches. These will be easy to weld and the excess can quickly
be removed with a hand grinder. Finally, locate and drill the
peepholes with a hole cutting saw. Install the pipe nipples into
the peepholes and weld into place. Install the throttle and
pressure tap. When you are satisfied and everything is lined up,
weld completely around both rings. Clean the job up with a
hand grinder.
Make the throttle body from a 6-inch length of 2 '/2-inch
tubing. Tubing is measured by the outside diameter making it
different than pipe. Bore one end to accept the shop vac hose.
Alternately, the shop vac hose could be attached with a strip of
sheet metal wound around the pipe and hose, the assembly
being secured with a pair of hose clamps. Substitute 2-inch pipe
with the sheet metal arrangement if tubing can not be found.
Later I modified the assembly by welding a flange to the throttle
input. I made a flanged adapter for the shop vac that allowed me
to attach the blower directly to the cupola, bypassing the hose
and eliminating the vacuum’s canister. A 15% increase in
airflow was recorded by eliminating the hose and canister. I
would strongly encourage use of the flanged adapter. You
should also be aware that considerable heat is radiated from the
cupola, the vacuum will melt if it is too close to the stack.
Make a spacer 12 to 18 inches long. Remove the blower from
the adapter if you turn it off for more than 1 minute.
The throttle rod is made from !4 inch by 5-'/2 inch steel rod.
Mill, grind or file a flat on the throttle rod to accept the throttle
plate, which is made from thin sheet metal and attached with 2
#6-32 screws. The throttle plate is cut in the shape of a disk that
just clears the inside diameter of the throttle body. The action of
the throttle is like that of one in a carburator. A spring and a pair
of Vi-20 nuts go the bottom of the throttle rod to provide
resistance to throttle movement.
34
Flanged Adapter
SLAG DOOR:
A 5-inch by 6-inch rectangle is now cut around the slag hole.

Four sections of 13/4-inch by 16-gauge strip are bent around the
edges of the door and welded into place. Round the comers
35
a«j iofi *-J strtr'
SU^ Door
with a grinder. Several 1-inch spikes are made from stiff steel
wire and welded on the inside of the door. These spikes help
hold the uncured fireclay in place. The slag door is held in
place with a 20-inch by 3/8-inch steel rod passing through 2
hooks also made from 3/8-inch steel rod welded to the shell.
36
BASE PLATE AND DOOR:
Cut the base plate and door from 3/8-inch plate. This is the
minimum thickness. It can be thicker but don’t make it thinner.
The base plate is cut 24 inches by 24 inches square. This is a
change from the original design, so you can catch the drops on a
steel sheet. A 10-inch circle is cut in the center of the base plate.
The shell is centered over the cut out and the location of the
mounting holes are marked and drilled. Cut the door 11 inches
in diameter. Drill with !4 inch holes spaced 1 inch apart, as
shown in the photo. These holes are to vent the steam coming
from the sand bed. On cold days you can see the steam shooting
out of these holes!
J
Ef*' -."i
P»in by tui
o cl
V v(u an Vs in si de
w al 1 Vi inc d ro<i
Detail of Modified Baseplate
38
for the outer portion of the hinge. Be sure to keep the clearance
specified so that the hinge will not bind when hot. The hinge
must support the weight of all the charges, be sure it is sturdy.
Align the door and hinge, then weld it all together.
Cut a wedge with a 5-degree taper from Vi by 3-inch stock.
Weld it opposite the hinge. Weld the feet to the legs. Square and
weld the legs to the baseplate. At this point you can layout the
latch. The angle between the arms will be approximately 100 to
120 degrees. It should latch without hitting the legs. The pivot
pin is made from a 4-inch length of V* inch screw stock. Make
the pivot from a 2-inch piece of 2-inch round stock. Drill the
hole for the latch pin and weld the pin in place. Double nuts on
the bottom of the latch pin make the assembly adjustable for a
perfect fit.
Flip the base plate over and attach the shell with the six Vi-
inch bolts.
39
SPOUT:
A spout is made from a 5-inch length of 3-inch channel iron.

Hold it in position on the shell and mark the base plate for a
relief cut. The cut is made, cleaned up, and the spout is welded
in place on the shell. The spout, when lined with sand and
fireclay should be below the taphole. This helps keep iron from
freezing in the taphole. Ideally you should get a 1-inch drop
between the taphole and the spout lining.
LINING THE FURNACE:
The furnace is lined by ramming refractory around a form.

The form is located relative to the shell by using four wood
strips at 90 degrees to each other. The strips are 2 1/2 inches
wide and 36 inches long. Ram the refractory around the form
and raise the wood strips as each layer of refractory is laid.
Fireclay and sand are fairly light when compared to
commercial refractory. A fireclay mixture can be rammed with
a heavy stick or rod, however you will have to use a 1-inch steel
rod and a 2-pound hammer to ram some of the commercial
refractory. Commercial refractory is expensive but holds up
very well. I use a product made by A.P. Green called “Green
Pac.” It is rated for 3000 degrees. They make “Blue Pac which
is rated for 3200 degrees; it is a little more expensive. Others
make a similar product. You are looking for 85% alumina phos
bonded plastic. This refractory cures extremely hard. I found
my refractory under “Firebrick” in the phone book.
Because the top of the furnace is subjected to much lower
temperatures it is lined with a sand and fire clay mixture. The
mixture is one 50-pound bag of coarse sand, one 50-pound bag
of fine sand and one 50-pound bag of fireclay. Pour half a bag
of each into a wheelbarrow. Turn it over until mixed then add
the rest and turn it over for quite awhile. It must be well mixed,
40
so be patient. When it is well mixed start adding water by
misting with a garden hose. Don’t let the water puddle in the
mixture or the clay will separate from the sand. This process
takes a while, so be prepared to do some shoveling. When the
mixture has the consistency of stiff mortar it is about right.
Don’t let the mixture become too wet, it will not cure well and
will slough off when fired. When you are satisfied with the
consistency, cover it with plastic and take a break for a few
hours.
Stewart Marshall Describes is his book Building Small
Cupolas an alternate way of mixing refractory. He mixes his
sand and then adds about half the water. He mixes the sand and
water very well and slowly adds the clay while constantly
turning. Eventually he works the remaining water into the mix.
He claims this method causes the clay to stick to the sand better
than the dry mix.
MAKING THE FORM:
Make the form by wrapping 24-gauge sheet metal around two

10-inch plywood discs. These can be cut with a saw, however
turning them on a lathe is better. The top disk has a large hole
drilled for your fingers; a wood screw is placed in one side of
the disc and a 1-inch deep slot is cut in one edge. The bottom
disk has a stout screw put in one side. When removing the disk
from the cupola you will grab this screw with a pair of pliers
and pull it up. Cut a piece of sheet metal 36-inches long. The
width of the piece is the circumference of the disc plus 1-inch.
For a 10-inch disc, this would be 32 Vi inches wide. Make a
right angle bend 1-inch deep and 36-inches long. This will hook
into the saw slot of each disk. Tape the sheet metal in place
with duct tape. If the metal is springy and difficult to work, roll
it up to a diameter less than 10 inches and try taping it again.
41
After ramming the lining, peel the tape off the top disk, grab the
screw with pliers and remove the disk. Remove the bottom disk
in a similar manner. Collapse the sheet metal and remove it.
Don’t be concerned if the form has collapsed a little out of
round while ramming the lining. Set the bottom wooden disc
back in the top of the furnace and carefully hammer it back
down through the bore using the steel rod. This will trim the
excess refractory from the high spots. This excess material can
42
be used to fill any voids. Using a piece of welding rod, poke
holes every inch or two around the lining. This is to vent the
lining during the curing process. Set the form up in the top half
of the cupola shell and repeat the process using the fireclay and
sand mixture. Cost savings are the reason for using the fireclay
mixture here.
Before ramming the lining, the tap hole void is made by
opening the bottom door and inserting a wooden form. Bolt it in
place through the tap hole. Ram the lining around the form to
leave a void for the fireclay “taphole.” After ramming, the

wooden form is removed. Alternately, you could cut the
refractory with a knife and a hammer. During operation, this
space will be packed with the fireclay mixture. Pushing a piece
of '/2-inch diameter rebar through from the inside of the cupola
will make the taphole. From the outside of the cupola, using a
wet finger, the hole is then smoothed to the conical shape as
shown. After each heat, the fireclay section is broken out. A
new taphole will be made for each heat. This eliminates any
iron buildup in the taphole. An iron-coated taphole may cause
the taphole to freeze up at the beginning of the melt. It is best to
start with a fresh taphole each heat.
43
Sand Bottom Taphole
The slag hole is made by opening the slag door and driving a
knife through the refractory with a hammer. Cut an opening as
shown in the photo.
44
Slag Door
FIRE THE LINING:
As soon as the lining is in spray the furnace with high

temperature grill paint. Both the lining and the paint must cure
45
at high temperature. Pour enough damp sand into the cupola to
make a bed 2 inches deep. Open all the tuyeres. Soak several
wads of newspaper in starter fluid and drop them down the
shaft. Adding 1 cup of gasoline to 1 gallon of diesel can make a
satisfactory starter fluid. Soak enough charcoal in starter fluid to
make three layers in the cupola. Light the furnace by placing a
lit propane torch through the taphole. When the charcoal is
burning well, add more charcoal. This could take several hours,
but you will want to have the whole cupola full of burning
charcoal. You may want to put a blow drier or other small
blower up to the tap hole to help the fire along. After about 5
hours, start adding coke and turn on a light blast. After another
hour increase the blast and add more coke. Let it cook another
hour then turn on full blast. The cupola will be extremely hot by
now. After 30 minutes turn off the blast, open the tuyeres, and
cover the top with one of the 15-inch circles you cut when
making the wind belt. Let the fire burn out overnight. Drop the
bottom and inspect the lining. Patch any holes. Your furnace is
now finished.
Firing the lining is a slow process. If you get the lining too
hot too fast, the outer surface of the lining will vitrify but the
inside will still be wet. When you turn the heat up, steam will
build up in the lining causing it to crack and slough off. Be
patient and take your time.
46
1 / •
1
n
/1 •
py. / 1/1,
L*7< ^
III. OPERATION OF THE CUPOLA
SEQUENCE OF OPERATION:
1. Set Up Work Area

2. Weigh Charges
3. Make the Taphole
4. Make the Bottom
5. Make Slaghole
6. Lighting and Burning in the Bed
7. Charge
8. Soak or Preheat
9. Melting
10. Tapping
11. Dropping the Bottom
12. Cleaning the Cupola
Before the first melt you will have to mark a rod to

the proper bed height. The intial bed height should be 18
inches above the tuyeres. Use a bod rod to measure
down from the top of the cupola to the top of the coke
bed, mark the handle. By using a bod rod you will
always have the measuring rod available.
Because Cupolas are fast melters actual melting time
will probably not exceed an hour. However preparation
of the cupola for melting could exceed three or four
hours. Each step should be carefully carried out. Once
melting starts you will not have time to prepare charges,
search for tapping bars, hammers or bod clay. Keep
everything organized as shown in the site layout.
49
buckets with 35 pounds of iron in each i cutting torch
SITE LAYOUT
I. SETUPTHECUPOLA:
Set up the cupola on a firm foundation away from trees and

houses. A firm foundation could be a concrete slab or any
footing that will not allow the fully charged cupola to tip
wobble or lean.
The area immediately around the cupola should be dry.
When hot iron contacts a wet surface steam is instantly formed.
This sends a shower of molten iron around the whole area. Keep
your work area dry and free of combustible materials. However,
because of flying sparks, wet down the area outside the work
area with a hose.
A 6-inch deep bed of sand should be built up around the
cupola. It should be sloped slightly towards the center so any
molten iron spilled will not run out of the bed area. A sheet of V*
inch plate should be cut to fit between the legs. Attach at least
six feet of chain to the plate. This plate will catch the drop at the
end of the melt. If the drop falls into the sand bed it becomes a
real mess, this makes it difficult to sort out the coke and iron
from the sand. Having the iron and coke drop on the plate saves
much time sorting after the heat. Use the chain to drag the plate
from under the cupola.
2. WEIGH CHARGES:
Proper preparation of the charges is critical for cupola

operation. Coke must be screened to proper size. The ideal size
is 1/12 the inside diameter of the cupola. This is .83-inch for the
10 in cupola. Since you will not find screens of this size you
must use 2 screens. Choose screens with Vi inch squares, the
other with 1-inch squares. Sift the coke through the small screen
to separate and discard the fines (half-inch coke could be used
in a 7” cupola). Then run the screened coke through the large
52
screen to separate the large pieces of coke. Coke between Vi and
1 inch in diameter will give the best results. You can get good
results with coke between Vi inch and W2 inches in diameter. I
purchase “nut coke” it is IV2 inches and smaller. Large coke can
be crushed in a coke crusher as described by Stewart Marshall.
Once you have filled several 5-gallon buckets with screened
coke, start weighing the charges. Ten paper grocery bags are
each filled with 6 pounds of coke. Ten five-gallon buckets are
filled with 35 pounds of iron each. The iron charge can be 10
pounds of large iron, and measure V2 x 3 x 4 inches. The
remaining 25 pounds should be smaller, thinner iron.
When all the coke and iron charges are weighed, place one
folded bag of coke in each iron bucket. Place the charge buckets
next to the ladder as shown in the site layout.
3. MAKE THE TAPHOLE:
The fireclay mixture that was used for the lining can now be
rammed into the void left around the taphole. Once the mixture
is in place pierce the taphole from the inside surface of the
cupola with a piece of rebar. Move to the front and smooth the
hole with a wet finger to form a taper towards the spout. Next,
line the spout with the fireclay mixture. Cut a trough in the clay
to carry the iron.
4. MAKE THE BOTTOM:
The bottom is made from a mixture of % sand and !4 Georgia

clay. Any clay should work, if you find a baseball field, that
clay will work also. Bottom sand should have a slight bond, but
not the bond that fireclay would give. The bottom will set fairly
hard under the heat of the cupola. A bottom made with fireclay
would setup much too hard for the bottom to drop. You would
probably have to chisel it out.
53
The bottom of the cupola is rammed with the weaker sand
and clay mixture. It must be uniformly rammed so no iron leaks
through. The bottom should be 2 inches deep and sloped
towards the taphole Vi inch per foot. Fillet the edges where it
joins the lining. Pack the joint between the top and bottom
cupola shells with the mixture used on the bottom.
5. MAKE THE SLAG HOLE:
Build a mound of the fireclay mixture upon the back door

and pierce it with a piece of rebar similar to the taphole. The
door is now locked in place and the edges are carefully packed
with the fireclay mixture to ensure no iron leaks develop around
the door.
6. LIGHTING AND BURNING IN THE BED:
Open the tuyeres and light the coke fire using paper,
charcoal and starting fluid. Place a small blower at the taphole
to help the bed bum. Add coke, then wait for each charge to
fully ignite before adding the next. Continue this process until
the height of the burning bed is over the tuyeres. When the
height of the burning bed is over the tuyeres, close them. Now
use the exhaust from a shop-vac and apply a light blast. Remove
the filter from the vacuum to get the maximum airflow.
Frequently check the bed with the bod rod, marked earlier.
When the burning bed is 18 inches above the tuyeres (calculated
in chapter 1, it may be slightly higher, but no lower) add an
additional 2 to 3 inches of coke. Apply full blast to preheat the
furnace and bum the bed back to 18 inches above the tuyeres.
The cupola will be very hot at this point. Now cut off the blast
and open the tuyeres.
54
7. CHARGE:
Charge the cupola with iron and coke (iron first) in layers
until the cupola is full. Usually 3 charges will fit.
8. SOAKING:
Let the cupola sit with no blast, tuyeres, slaghole and tap
hole open for 45 minutes. This is to allow the charges to
preheat. Rehearse a tapping and bodding operation during the
preheat phase. Carry the empty ladle to each mold to ensure you
can easily reach each mold and practice a pour. This rehearsal is
important. You don’t want to find out you can’t reach a mold
when you have 55 pounds of white hot iron in your hands.
9. MELTING:
Apply a 3-inch blast after the soaking period. During the

next few minutes the charges will begin to settle and the
indicated pressure may increase. The final blast pressure should
be between 4 and 5 inches.
After 8 to 10 minutes iron should be visible at the taphole. If
it is seen before this, the bed is too low. Decrease the blast. If
more than 12 minutes pass the bed is too high. Increase the blast
to bum away the excess coke.
The first iron will be cold. It may start to freeze in the spout
and taphole. Work quickly to knock it out of the way with a
piece of rebar. Be careful not to let the rebar freeze in the tap
hole or you will have to cut it out with a torch. You must work
very quickly here! Once the stream of iron is steady, bod it off.
Wait 2 to 3 minutes to let the iron build up in the well and heat
the sand bottom. Tap the cupola. This first iron will be cold and
55
should be poured into an ingot mold. Charge the cupola with
iron and coke and prepare for the next tap.
10. TAPPING THE CUPOLA:
Tap the cupola by picking out the bod with a tapping bar. A
tapping bar is a 3-foot section of rebar with a chisel point on
one end and a T handle on the other. The iron may start flowing
with a little poking or it may require breaking through a frozen
iron plug. Hammering the tapping bar through with a 2-pound
hammer can usually break plugs. Be quick to extract the tapping
bar so the bar does not freeze in the taphole. At the first sign of
slag close the taphole with a bod. Frozen slag in the taphole
makes tapping very difficult. After tapping the cupola the
taphole must be closed with a cone shaped plug called a bod.
The bod is roughly the size of a golf ball and is made of the
same sand and fireclay mixture that was used for the spout. This
mixture sets up fairly hard and might be better is some sawdust
were added. The sawdust would make the bod softer and easier
to pick out. A bod rod is made of a 3-foot section of rebar with
a T handle on one end, and a 1 */2-inch disc welded to the other.
I cut my disc from a rod, however Va inch plate also works well.
The disc end of the bod rod is placed in a bucket of water, then
a wad of clay is formed into a cone over the wet tip. This rod is
placed, standing up, in the concrete blocks as to be ready when
needed. It is a good idea to keep at least 2 bod rods and tapping
bars ready incase a bod falls off, or the tapping bar becomes
coated with iron.
For larger taps, up to 55 pounds, the slaghole is bodded off
when iron reaches it. Wait 4 to 5 minutes, then tap the cupola.
56
Clay Bod
11. DROPPING THE BOTTOM:
Tap all the liquid iron remaining in the cupola at he end of

the melt. Turn off the blast and open the door latch. A properly
constructed sand bottom will support the mass of unbumed
coke, but will rupture readily when picked with the tapping bar.
The cupola’s contents will be extremely hot. It is only a matter
of a few seconds before you will have to step away from the
drop. After you have cooled down, prod the bottom again to be
sure nothing is lodged in the cupola. When you are confident
the contents have fully dropped, spray the drop with a hose.
Keep a safe distance from the drop when spraying it with water.
Steam will be violently liberated and you could easily be
burned.
11. CLEANING THE CUPOLA

The next day, when the cupola has cooled, it may be cleaned
and patched (if any refractory has cracked and sloughed off)- A
pointed steel bar and hammer may be used here. A pneumatic
57
•1 )U • s'Zr
1 I'Vr
ifl mW' A
■vnw r 1 m. * mn
It A V* 4 1 L KB
Pj 1 F» »
I *F T
1 1
IMll
I# .
f £
lL*-
• [ i ffr
!1 ^
wdrWx
HJ ■
r 1 1 1
i f A .v fl
I • *A
‘nlF
■V •
W< i 4r 9
Cleaning the cupola with a pneumatic chisel
remove the smooth glaze that has formed on the lining. Knock
out the tap hole with a 1-inch steel rod and the 2-pound
hammer. Be sure the lining is smooth and free of large lumps.
After several heats I prefer an alternative to dropping the
bottom. At the end of the heat all of the iron is tapped out and
blast is turned off. The cupola is allowed to cool overnight. The
next day the cupola is inverted and the bottom is knocked out
with a sledgehammer. The cleanup is finished with a pointed
bar and hammer.
58
IV. AIR SUPPLY:
The balance of air and coke is critical for proper cupola

operation. Cupola blowers should be sized at 2.5 CFM per
square inch of sectional area between the tuyeres. Besides CFM,
static pressure is also important. Static pressure is the pressure a
blower develops if it is connected to a sealed container. As
static pressure in the cupola windbelt increases, the CFM
decreases. Static pressure is related to the physical size of the
coke and iron pieces that make up the charges, how tightly
packed the charges are, and the height of the stack. In other
words, resistance to airflow increases the static pressure. Fans
are rated by CFM and by static pressure. It does no good to
have the proper CFM if there is not enough pressure to push it
through the cupola’s coke bed and stack.
Pressure blowers supply the combustion air for cupolas.
Pressure blowers come in two types, positive displacement, and
centrifugal fans. Roots type blowers are the positive
displacement type. With each revolution of the blower lobes, a
known quantity of air is compressed. Roots blowers are capable
of pressures over 25 PSI. Maximum rated output of common
roots blowers is shown in the chart below. Roots blowers are
commonly used on mosquito spraying equipment. Many city
and county pest control departments routinely junk such
blowers. Dresser Industries manufactures roots blowers and has
distributors in most states. You may be able to find a used
blower through one of their pump dealers.
Centrifugal fans are more commonly used on cupolas. The
six types of centrifugal fans are classified by blade type. In
order of decreasing efficiency they are as follows: 1) airfoil
blades 92% efficient, 2) backward curved blades 85%, 3)
backward inclined blades 78%, 4) radial tipped blades 70% 5)
forward curved blades 65% and 6) radial blades 50%. We are
only concerned with radial blade high-pressure blowers A high-
59
pressure blower’s output is measured in ounces or in inches
water column. 1 ounce of pressure will raise a column of water
1.73 inches. Water column is measured by a manometer.
Inlet Vac. Temp. Rise Press. Rise

Frame Speed Inches Hg. Fahr. Deg. PSI
Size RPM (kPa) (C°> (kPa)
22
24
5275
5275
Era
Era 225
210
(125)
(117)
12
7
( 82)
( 47)
32 3600 15 (50) 225 (126) 15 (101)
33 3600 226 (126) 12 ( 82)
36 3600 225 (125) 7 ( 47)
42 3600 240 (133) 15 (101)
45 3600 15 (50) 225 (125) 10 ( 68)
47 3600 15 (50) 226 (125) 7 ( 47)
53 2850 15 (50) 225 (125) 15 (101)
56 2850 15 (60) 225 (125) 10 ( 68)
59 2850 15 (50) 225 (125) 7 ( 47)
65 2350 16 (53) 250 (139) 15 (101)
68 2350 16 (53) 240 (133) 12 ( 82)
615 2350 12 (40) 130 ( 72) 6 ( 40)
76 2050 16 (53) 250 (139) 15 (101)
711 2050 16 (53) 210 (117) 10 ( 68)
718 2050 12 (40) 130 ( 72) 6 ( 40)
Operating Characteristics of Small Roots Blowers
A manometer is a glass or plastic tube in the shape of a “U” that

is partially filled with colored water. One side of the tube is
connected to the windbelt; the other side is left open to the air.
The pressure in the windbelt forces the one side of the water
column down and the other side up. The difference in height of
the two water columns is measured in inches with a ruler.
60
SELECTION OF CUPOLA BLOWERS:
Cupolas will normally melt 4000 pounds of iron for every

1000 cfm of air supplied. The 10 inch cupola is melting 330
pounds / hour at 90 cfm. It is operating at approximately 92% of
the above general rule. Usually cupola blowers are sized at 2.5
cfm per square inch of sectional area between the tuyeres. The
recommend blower for the 10-inch cupola is:
Area of the cupola = 78.5 inches2
2.5 cfm X 78.5 = 196 cfm, approximately 200 cfm
Blowers are frequently sized up 10% to make up for leaks in the

system and variations in temperature. However they are often
run at 80% to 90% of the calculated value.
Operating pressure is related to system resistance.
Recommended fan pressures are shown on the graph. The
cupola wind belt will probably not reach these pressures; they
are only recommended blower capacity.
Blower Sizes for Cupola Operation
Inside Diameter Area Actual cfm Recommended Blower size

in inches inches2 cfm discharge pressure oz.
10 78.5 165 216 8

18 254 510 700 16
23 415 830 1140 20
The graph below illustrates suggested blower pressure capacity

for sectional area between the tuyeres. To use the graph find
the section area on the bottom, draw a line straight up until you
61
OZ PRESSURE
00
vO
>raiHWVia
2
«"
SECTIONAL AREA BETWEEN TUYERES IN INCHES

o
in
o oo vo rr o
^HMOdHS-aoH awaa
66
V. DESIGNING CENTRIFUGAL FANS:
Design of propeller type fans is fairly straightforward. A

satisfactory prototype may be built from the equations for
propeller type fans. Formulas for centrifugal type fans are not as
accurate. Approximate dimensions may be calculated, however
actual performance must be measured after assembly. The fan
speed or dimensions may be adjusted to bring the fan into
compliance with the original design goals.
After commercial fans are built, they are tested. The
performance is plotted on a graph called a fan output curve, or
performance graph. Fan curves are very useful tools to the fan
designer. If you know the actual dimensions and rpm of the fan
plotted on the fan curve you can use formulas called the fan
laws to convert fan performance of a known fan to meet new
requirements. Once you build and test one fan you can size it up
or down and, or change the speed and have predictable results,
provided all dimensions remain geometrically proportional.
This includes such factors as number of blades, front and rear
plates, inlet and outlet area. In other words, the new fan must be
an exact scaled copy of the original.
DESIGNING WITHOUT PERFORMANCE CURVES:
Fan output is related to wheel diameter, width, speed,

number of blades, inlet and outlet areas. Once you know the
cfm and static pressure fan must produce, you may start with
the formula below.
The minimum wheel diameter:
Outside diameter inches = (15500/rpm)VSP
SP = static pressure
67
The constant 15500 may increase for less efficient radial blade
designs, or decrease for the more efficient designs.
The minimum inside diameter, or inlet size:

3_
inside diameter = 1 lV (cfm/rpm)
width of the wheel = (550 cfm)/(D x tipspeed) ft/min
D = diameter in inches
Tip speed is the speed of the outside edge of the wheel.
Tip speed = (JtD^^rpmyU
Fan efficiency is assumed to be 50%
Horsepower required by fan:
Hp = (cfm x inches wc) / 3178 wc = water column
DESIGNING A WHEEL FOR THE 10 INCH CUPOLA:
Parameters:
cfm = 220
static pressure = 8 oz
convert oz to inches wc: (8 oz) x 1.73 inches/oz = 13.84 inches
Since motors run approximately 3500 rpm we will start here.

This may or may not provide us with a workable design.
outside diameter = (15500/3500)Vl3.84 = 16.475 inches
68
tip speed = (tt x 16.475 x 3500)/ 12 = 15096 ft/min
Width of wheel = (550 x 220) / (16.475 x 15096) = .487 inches
The 487inch wide blade will be difficult to assemble, lets

increase the rpm.
rpm = 5700
od = (15500/5700)Vl3.84 = 10.1 inches
rounding to 10 inches, tip speed is 14,922 ft/min
width of wheel = (550 x 220) / (10 x 14,922) = .81 inches
.81 inches will be easier to manufacture and a 10-inch wheel

requires less horsepower, therefore we will use this wheel.
Calculate the inlet diameter:

3_
Minimum inlet diameter = 11 V (220 / 5700) = 3.72 inches
Calculate the horsepower required by the fan:
hp = (220 x 13.84)/ 3178 = 96 hp
MECHANICAL STRESSES DEVELOPED IN

CENTRIFUGAL FANS:
Balance of centrifugal fans is critical. If the wheel of our 10-

inch fan is out of balance by the weight of a sheet of paper , a
rotating force of 30 pounds will be developed. The vibration
caused by such an imbalance would soon cause the fan to fly
apart. Fortunately balancing a fan is not difficult if you build a
set of balancing ways. Construction of balancing ways will be
described later.
69
The centrifugal force formula:
Force = (W x R x rpm2) / 2933
W = weight in pounds
R = distance in feet from the center of the blade to the center of

the fan wheel.
Lets calculate the centrifugal force on one of the fan blades of

the 10-inch fan. If the blade is made from 16-gage steel it would
weigh approximately 1.45 oz.
Convert ounces to pounds: 1.45oz/16oz/pound = .091 pound
Force = ((.091pound) x (,348foot) x (57002)) / 2933 = 3 50 foot-pounds
If the fan were to come apart the blades would have a force of
350 foot-pounds and a speed of 170 mph. As you can see the
forces generated in this small fan are considerable. Although
fan construction is not difficult it does require precision work
and precision balancing. A poorly constructed fan will at best
explode and wreck the afternoon’s melt. It could also
dismember or kill someone. A properly constructed fan is a
precision tool. Treat it as such and you will have no problems.
SCROLL HOUSING:
The individual streams of air leaving the blade tips are

collected and formed into a single air stream that is discharged
through the housing outlet. Three circular sections form the
spiral. The radii of these sections are 5D, 7D, and 9D. The
radii are off set from the center of the wheel by .ID.
Calculate the scroll radii for the 10-inch wheel:
70
R, = .5 x 10 inches = 5 inches
Rj = .7 x 10 inches = 7 inches
Rj = .9 x 10 inches = 9 inches
Calculate the offset:
Offset = .1x10=1 inch
Discharge opening is calculated by: 2(D x W) / 3
Discharge area = 2(10 x .82) / 3 = 5.46 inches2
Cut off point: the cutoff point is where the scroll is closest to
the wheel. The closer the cutoff point is to the wheel the noisier
the fan will be. A good figure is 5% of the wheel diameter.
Cutoff point = .05D
Cutoff point = .05 x 10 inches = .5 inches
DESIGNING FROM AN EXISTING FAN:
After you have designed and tested a fan, or have a known

design with a fan performance graph, you may calculate the
performance of another geometrically similar fan, i.e. “an exact
scaled copy.” To convert fan performance, use the following
rules:
To convert the performance of a fan from one speed to another:
cfm varies directly with speed:
(cfm2/ cfm, ) = (rpm 2 /rpm,)
71
Calculate the new cfm if the speed is increased:
cfm, =200
rpm, = 1800
rpmj = 3600
cfmj = (cfm, rpmj) / rpm,
cfm2 =(200x3600)/ 1800
cfm2 = 400
Pressure varies as the square of the speed.
(SP2 / SP, ) = (rpmj / rpm ,)2
SP2 = SP, (rpmj / rpm ,)2
SP, = 3 inches wc
rpm, = 1800
rprr^ = 3600
SP2 = 3 inches (3600 / 1800)2
SP2 = 12 inches
Calculate new rpm if SP is known:
rpm 2 = rpm,V SP2 /SP,
72
rpm , = 3600
SP, = 5.2
SP2 = 13.84
rpm 2 = 3600V 13.8/5.2
rpmj = 5865
Horsepower varies as the cube of the speed:
(hp2 / hp, ) = (rpm, /rpm, )3
hp2 = hp, (rpmj /rpm, )3
hp, = .125
rpm , = 1800
rpr^ = 3600
hp2= . 125( 3600 / 1800)3
hp2 = l
Changing the wheel diameter:
cfm varies as the cube of the size.
(cfm2 /cfm,) = ( D2 / D, )3
73
cfirij = cfm, (D2 / D, )3
cfm, = 200
D, = 10 inches
D2 = 14 inches
cfmj = 200cfm (14 inches /10 inches)3
cfhij = 549
Static pressure varies as the square of the diameter.
(SP2 / SP, ) = ( D2/ Dt)J
SP2 =SP,(D2 /D,)2
D, = lOinches
D2 = 14 inches
SPt = 5.2 inches
SP2 = 5.2 inches(14inches /10 inches)2
SP2 = 10.2 inches
Horsepower varies as the fifth power of the diameter.
(hp, /hp,) = (DJ/D,)J
hp2 = hp,(D2/D,)s
74
D, =10 inches
D2 = 20 inches
hp, =*/2
hp2 = Zi hp( 20 inches / 10 inches)5
hp2 = 16
In order to design the best fan for your application you must
know which parameter is the most important, static pressure,
cfm or horsepower input. The following conclusions may be
helpful:
Increasing wheel speed increases the SP faster than the cfm

because SP is a squared factor while cfm is directly proportional
to the wheel speed.
Increasing the wheel diameter increases the cfm faster than the
SP because the cfm varies as the cube of the diameter, and SP
varies as the square of the diameter.
Power consumption increases very rapidly for an increase in

diameter because the horsepower varies as the fifth power of the
diameter. For example, if a 1 horsepower 12-inch wheel is
increased to 24 inches, the new power required will be 32
horsepower.
CALCULATING PULLEY SIZES:
Obtaining the correct fan speed from the motor speed is

calculated from the ratio:
75
RPM IP = RPM IP
12 2 1
RPM( = Motor speed P( = Motor Pulley Diameter
RPM 2 = Fan speed P2 = Fan Pulley Diameter
The formula may be easily manipulated to solve for either

pulley size if one pulley size and the speeds of the motor and
fan are known.
P} = (P2 x rpm2 )/RPM]
P2 = (P( x RPM () / RPM2
Calculate the pulley diameter for a 3600-RPM motor if the

desired fan speed is 5700 RPM and the fan pulley is 2 inches in
diameter:
P( = (P2xRPM2)/RPMi
P( =(2x5700)/3600
Pt = 3.167 inches diameter
76
VI. CONSTRUCTION OF CENTRIFUGAL FANS:
Fan construction is at least a full day’s work, and most likely

a two-day project. Centrifugal fans may be built of sheet metal,
or steel plate. The 10-inch wheel may be constructed of 20-
gauge steel, however a sturdier metal such as 16 gauge should
be used for the scroll housing. The thinner metal, when used for
the scroll housing, tends to collapse toward the wheel during
operation. For the larger or direct drive fans, a scroll housing
back plate might be built of 14 inch plate, drilled to receive a C-
faced motor, such as the motors used on pumps.
Vanes may be held on the wheel by steel pop rivets. As we
have seen earlier, the forces generated in a fan are considerable,
ruling out aluminum rivets. Larger wheels may have welded
vanes.
Depending on the thickness of the material chosen, the wheel
may be cut out with tin snips, a jig saw, band saw, or torch.
With much effort, 20-gauge material may be cut with
compound snips; 16 gauge may be cut with a jigsaw, or
preferably a band saw; 14 inch plate requires a torch. My first
several wheels were cut with hand tools, however a rotary table
and a mill make the whole fan project much faster and easier.
Hand construction methods are described here.
Using dividers, draw two wheels (the front and rear plates)
on the sheetmetal. Carefully cut the wheels out without bending
or distorting them. You may cut the wheels a little larger than
the scribed line and clean them up by grinding to scribed mark.
Draw a horizontal line through the center of one circle. Using a
protractor, carefully scribe the locations of the blades. To find
degrees between the blades, divide 360° by the number of
blades. For the twelve blades used on the 10-inch wheel there
are 30° between each blade. Scribe two circles to locate the
rivets for the blades. On the remaining wheel draw the circle
for the inlet area.
77
outer circle 4.215 +
radius \
Plate layout for 10” cupola fan
8
Center punch the hole in the center of the wheels and all the
rivets holes on one of the back plates. Locate and punch the
holes for the mounting hub. All of these holes must be
.785”
90° bend
N
b 3.14’ End view
Fan Vane
accurately located, or the wheel will be impossible to balance.

Clamp the wheels together and drill two rivet holes on both
halves of the wheel. Rivet the wheels together and finish
drilling the rivet holes. Remove the rivets, drill the center hole
the size of the shaft you are using (motor shaft or mandrill).
This is best accomplished by screwing the wheel to a section of
plywood and step drilling, increasing the drill diameter by no
more than 1/8 inch each pass. Sheet metal has the tendency to
climb up the bit, so securely attach the wheel to the plywood.
If you are lucky enough to have a trepanning tool and drill
press, the inlet hole is easily drilled. Otherwise you will have to
cut it with snips and grind it to final size, or drill a series of
small holes just inside of the scribed line. The inner circle is
removed and discarded. The hole is ground and /or filed to final
size.
79
The vanes are cut out, although the plan calls for .81 inch, I
increased the width to 1.1 inches so that I might use the fan on a
12-inch cupola. This worked out well as the fan can be throttled
down for the smaller cupola. The mounting tabs are scribed and
bent. Make the bends by clamping in a vise and using a hammer
to pound the bend flat. If you have a triple beam balance, now is
the time to weigh and grind the vanes until their weight is
within 1/10 of a gram of each other. This cuts your balancing
time down to a minimum. If you do not have a balance, grind
them until they are as identical in size as possible.
The centerline of each mounting tab is marked. The location
of one of the rivet holes is drilled on each mounting tab. The
second set of holes will be drilled after the vane has been
riveted in place with one set of rivets.
Once the vanes are riveted in place install the hub. A hub
may be turned on the lathe, salvaged from an old fan, or made
from an old pulley. Once the hub is installed, balance the wheel
on a set of balancing ways. The balancing ways require the
wheel be mounted on a polished shaft to work properly. Drill
rod is preferred. The slightest surface imperfection will ruin the
accuracy of the balancing ways. This means galvanized shafting
is useless for balancing.
Balance the wheel by removing material from the heavy side
of the wheel. When placed on a set of balancing ways the wheel
will roll until the heavy side is down. Test the wheel and reverse
the wheel on the ways, if they are truly level you will get the
same result (heavy spot) each time. If the wheel has been well
constructed, it may take only light sanding of the heavy edge
with a belt sander. Otherwise grinding and drilling the heavy
side should work. Continue testing and trimming until the wheel
will rest in any position on the ways without moving. Balancing
can be a time consuming process. Care used in building the
wheel will minimize the amount of time spent balancing.
80
Rear Plate with Vanes Attached
BEARINGS:
Pillow block bearings are poured from aluminum. Bore the

pillow blocks by using an angle plate on the faceplate of the
lathe. Number 6203 ball bearings are fitted into the castings.
These bearings are rated to run at 17,000 rpm, so 5800 rpm used
in the fan should be acceptable.
Purchasing the pillow blocks will be easier; however 6203
bearings are so cheap, $2.95, I decided to make my pillow
blocks.
81
Shop made fan joined with sheet metal screws
BALANCING WAYS:
Balancing ways must be nearly frictionless and absolutely

level. A good set of balancing ways may be easily built using
precision steel rules for the frictionless edge. A frame is built to
support and level the rules. The rules are attached with cap
screws. The screws should be snug but not overly tight. This
may distort the rules giving in accurate results. Because the
rules do not have holes drilled in them, but rest on top of the cap
screw threads, shims may be needed under the bottom edge of
each the washers to maintain a flat surface for the cap screws,
and prevent distortion of the rules.
The frame is welded up from 1-inch angle. My angle was
salvaged from an old bed frame. The hinge was made from 1/8
by 1-inch strap. All of the bolts are V* - 20. The base of the
balancing ways is constructed of a 6 inch by 10-inch section of
Masonite. The distance between the ways is kept small to
minimize the inertia of the shaft. Later, substituting a wider
82
section of Masonite will allow you to balance wider wheels and
motor rotors.
6-inch ruler
Balancing Ways Side View
83
Set the ways up on a level surface and adjust the feet until
there is no wobble. After each of the ways is leveled, check
between both of the ways. If your level is not long enough to
reach across the distance between the ways, use a piece of
polished shaft or drill rod to span the distance. Recheck the
ways until they are absolutely level.
6-inch ruler
Balancing Ways End View

Shop made Balancing Ways
SCROLL HOUSING
Layout of the fan scroll is done using the offset and three
radii calculated in the fan designing section. For a 10-inch
wheel the offset is 1-inch, the three radii are 5-inches, 7-inches,
and 9-inches, the cutoff point is 1/2-inch. When doing the
layout you must add 3/8 of an inch to each radii to account for
the flange that will extend around the edge for joining the
housing together. Welding is the preferred method of joining
the scroll housing sides together; however, sheet-metal screws
will work. To begin the layout, locate and center punch the
center of the wheel on the scroll housing. Scribe the wheel
around the center point. Scribe the horizontal and vertical center
85
lines. Locate and center punch the offsets to the right and left of
the vertical center line. The layout of the scroll housing is
shown in the drawings below.
Lay out and drill or cut the holes for the shaft and the air inlet.
Because the spinning fan would quickly remove a finger is it
were stuck into the inlet, it would be a good idea to cut a section
of hardware cloth and a mounting ring to cover the inlet.
Hardware cloth is available at any hardware store.
After the scroll has been cut out, bending the 3/8-inch flange
around the scroll and welded together, or it may be left straight
and joined with screws. Welding is the preferred method. To
form a flange for welding, bend the edge up approximately 1/3
of the way (30°) with a pair of pliers. Make a second pass,
bending the flange up another 30°, make a final pass completing
the bend. The flange will smooth quite well with a hammer and
auto-body dolly. If no dollies are available, a short section of 3-
inch round stock will work. After the scroll has been welded it
is cleaned up with a grinder and given a coat of Bondo. The
Bondo is used to fill any rough spots and dents so the cover of
the fan will have a smooth surface on which to seat.
If you choose to use screws to join your housing, you should
scribe a line 3/16-inch in from the scroll edge and locate your
screw holes. The center strip that joins the two scrolls together
is cut with V* -inch extra width. Two edges are bent up as was
done on the scrolls. The strip is bent around the scroll and
drilled for the screws. Using vise grip pliers to help hold the
strip in place, drill a screw hole, insert a screw, move the pliers
down and repeat the process until all the screws are in. When
the assembly is completely finished, seal it with caulking or
silicone sealant.
Generation of a scroll housing is show in the four steps
illustrated below.
86
radius 3
VENTURI INLET:
The venturi inlet serves two purposes. First it increases the

efficiency of the inlet from 15 to 28%. Second, it minimizes the
recirculation of air within the fan. Tight clearance between the
venturi and the inlet of the fan wheel will prevent the output air
from being sucked back into the inlet.
Make a ring shaped pattern with the 2 Zi0 draft from a 2 x 4.
Cut 6 sections with 60° sides, these will glue up to form a
hexagon shaped ring. The rounded ring shape and the proper
draft are cut in the lathe. The pattern is sanded smooth and
painted.
89
Pour the venturi inlet from aluminum. Finish the ring by
turning in the lathe. The venturi is roughed out at a 45° angle.
Make the final shape by forming with a wood turning chisel.
Check the shape as you turn by using a cardboard template with
the proper radius. The radius should be a minimum of 14% of
the wheel diameter. This is 1.4-inches for the 10-inch fan wheel.
Sand and polish the venturi after the radius is formed. This is a
quick job in the lathe. Three holes are located 120°apart on the
backside of the venturi. The ring is drilled and tapped for 6-32
screws, which attach it to the blower housing.
TRANSITION PIECE:
The transition piece is also poured from aluminum. Make a

split pattern and a core box. Both the pattern and the core box
are sanded smooth and filled with Bondo autobody putty. The
transitional radius on the inside of the core box is also formed
with Bondo and sanded smooth. Cut the flange from Vi inch
thick plywood. The flange is given draft by building up Bondo
and sanding the taper.
The core mix is made from sand, wheat paste (wallpaper
paste) and molasses water. The water mixture is roughly 10%
to 20% molasses and the balance is water. Mix a ratio of 90%
sand to 10% wheat paste (more may be added if needed). Add
the molasses water until the mass will hold a shape. Press it into
the corebox and add two reinforcing wires. Flip the corebox
over on a section of sheet steel for baking. Make two more
cores (you will have an extra incase one breaks). Bake the cores
at 350 for about a half an hour. The cores should be completely
dry and brick hard. The cores are removed from the oven and
glued together with a thick mixture of wheat paste and molasses
water.
90
3.5
Fan End of Transition Piece
The core is placed in the mold and the mold is immediately

poured. A core left in the mold will absorb water and make a
gassy casting.
Clean up the casting and drill the flange with three equally
spaced 9/32-inch holes. Later, the flange will be attached to the
blast main with V* inch bolts. A 1/2-inch lip is bent out around
the discharge opening of the fan scroll housing. Attach the
casting to the fan with four 10/32-inch screws. Drill and tap the
casting to accept the screws. Make a cork gasket to seal the area
between the casting and the fan housing. Cork gasket material is
readily available at auto part stores.
91
Layout of Flanged Transition Piece
92
Split Pattern and Core Box
Tight Clearance of the Venturi Inlet
93
Completed Cupola Blower
If properly constructed, these blowers move a considerable

amount of air. You will be pleasantly surprised by the strong
blast they produce. I doubled the output of my cupola by
building a proper blower.
94
VII. CONSTRUCTION OF PITOT TUBE AND
MANOMETER.
Construction of a pitot tube is simple and should not take

more than an afternoon. Start with a section of soft 3/8-inch
diameter copper tubing. Carefully drill the eight holes as shown
in the drawing. Remove any burrs from the cut ends and attach
the Tee coupling using electronic solder and flux (available at
radio shack). Be sure the end is bright and clean before you
solder. A small propane torch makes fast work of this job.
Make the nosepiece, the two brass plugs, and the spacer on the
lathe. Cut a section of 1/8 inch tubing for the center tube. This
tubing is used for automotive oil pressure lines, and can be
found at auto parts stores. Carefully measure the distance for the
spacer and solder into position on the inner tube. Solder the
nosepiece to the inner tube and insert the assembly into the 3/8-
inch tubing. Solder the nose to the 3/8-inch tubing. Do not
solder the spacer to the 3/8 tubing. Fit and solder the remaining
brass plugs in position. Polish the nose piece and tube brightly.
If you are using a buffing wheel, it is much easier to polish the
tube before bending. Using a tubing bender carefully bend the
assembly. Apply a coat of clear lacquer. Although our tube does
not meet laboratory specifications it will still work suprisingly
well. I tested my tube on a still day by riding down the street
and recording the inches pressure vs. speedometer. It was
accurate to V2 mph.
The Air Movement and Control Association (AMCA) states
the pitot tube stem diameter should not exceed 1/30 the test duct
diameter. While our tube may not meet this specification, it
works well enough for our purposes.
95
1 Head is free of nicks and burrs
8 holes see detail

0.5D radius
(.1875)
3 D radius
(1.125)
1/8 inch tubing
D (3/8 inch tubing)

y+.\M
8 holes-0.13 D (#55 Drill)
equally spaced and free from
burrs. Hole depth not less
3/8 inch “T” than hole diameter
Static Pressure
AMCA standards state: D shall be no

greater than .033 diameter of test duct
For our tests we will subtract the pitot

Tube diameter from the duct diameter
Brass plugs
Total Pressure
Pitot - Static Tube (Spherical head)
96
CONSTRUCTION OF THE MANOMETER.
Construction of the manometer should only take an

afternoon. Use a flex rule for the manometer scale. Make the
manometer frame from wood or aluminum. The manometer
seen in the photo of the cupola is constructed of two sheets of 14
inch aluminum plate covered with a sheet of Plexiglas. The
section of Plexiglas is so small it can often be found as a scrap
at the local hardware store.
Screw one aluminum plate to a piece of wood, and mill a
groove for the ruler that is .010 deep and as wide as your ruler.
Mount the form and wooden base to a rotary table and mill the
“U”groove. Drill and tap the screw holes for the complete
assembly. The center section of the U form is held in place from
the rear by 6-32 screws. Assemble the frame (back section that
has only screw holes and two pieces of U form) and Plexiglas
cover. Trim the assembly to size on the mill. Smooth the edges
with a sander. Open the form, press the tube into position, and
replace the Plexiglas cover. Put a slight bend in the flex-rule
and insert into its groove.
Alternately the frame could be cut from Va plywood.
Carefully trim the semicircular section as accurately as possible.
Rout a shallow groove for the ruler then form the inside section
maintaining as smooth a curve as possible. Paint the form white
for visibility. Assemble the form, press the tubing into position
and attach the Plexiglas cover. The tension of the rule can be
adjusted by slightly bending the flex-rule.
The manometer is half filled with a dilute solution of food
coloring and water. Dimensions are given for the manometer
used on the 10-inch cupola. These are not critical, build it
according to materials on hand. Just keep the tubes straight, and
the curved section as smooth as possible. If you are going to use
a higher pressure, you will need a longer manometer, or use a
liquid that is heavier than water.
97
Calculate the new “inches-rise / fpm” of the new material
according to the specific gravity (density of the new liquid
relative to water).
Calculate the new scale for a manometer using mercury:
Specific Gravity of Mercury = 13.6gm/cm3
Specific Gravity of Water = 1 gm/cm3
Inches rise = (1.73 inches) x (SG water / SG mercury)
Inches rise = 1.73 inches x (.0735) = .127 inches
Mercury might be used for a large cupola or a cupola with a hot

blast system because a hot blast system often requires an
additional 12 ounces of pressure above the cold blast pressure.
Glycerin might be used for smaller cupolas. Furthermore, using
a liquid that is lighter than water (lower specific gravity) will
make a more sensitive scale. Mineral spirits combined with an
inclined scale will make a very sensitive scale. Flammable
liquids should only be used when testing a fan and are not to
be used around an operating cupola.
SG of Glycerin = 1.26 gm/cm3

SG of Ethyl Alcohol = .81 gm/cm3
99
Shop made Pitot tube and manometer
100
VIII. CALCULATION OF AIRFLOW.
The optimal operation of the cupola requires the accurate

measurement of the blast. A manometer attached to the
windbelt will provide a good measure of cupola’s operational
condition. Such measurements, when recorded in a notebook,
will provide a basis for future operation and help one achieve
repeatable results. However, the actual volume of airflow must
be measured with a Pitot tube.
Accurate measurement of airflow requires several readings
across the diameter of the duct. Since our duct is so small in
diameter, we will use one reading in the center of the duct
(maximum velocity). Calculate the velocity at the center of the
duct, and assume the average duct velocity to be 90 percent of
this calculated maximum value. The actual velocity is lower
than the maximum or center velocity because the air travels
slower at the walls of the duct than in the center.
Turbulence in the airflow will not give accurate readings.
Therefore the duct being tested should be smooth and free of
burrs. Likewise the entrance and exits of the test duct should be
smooth. Transition pieces or adapters that attach the fan to the
test duct have no more than a 7 degree taper. By using a flow
straightener and a duct that is ten times longer than it is wide
you can attain minimal turbulence. L = 10D. A flow
straightener, can be made from thin sheet metal, while it will
not meet ACMA standards, it might be helpful.
The pitot tube is inserted through a hole in the side of the
duct and pointed squarely into the airflow. Be sure to locate the
hole in the duct so that the tip is 10 diameters from the duct
input. The velocity of airflow will be calculated from the
manometer reading.
Velocity of airflow is calculated by:
V = 4005Vh
V = velocity in feet / minute

h = height in inches water column
101
Calculate the velocity of airflow for a manometer reading of
1.25 inches.
V = 4005V 1.25 inches
V = 4478 feet per minute
Volume of airflow is calculated by:
CFM = A x V
CFM = Area duct x Velocity of airflow
Calculate the volume of airflow for 10-inch cupola with a 2-

inch diameter blast main:
Actual inside diameter of blast main D = 2.067 inches
Area of blast main in square feet = .0218(D / 2)2
A = .0218(2.067 / 2)2 = .0233 foot2
CFM = A x V
CFM = (4478 ft / min) x (.0233 ft) = 104.3
Average airflow is 90% the maximum calculated rate.
Average airflow = (.9) x (104.3) = 93.9 CFM
A more accurate calculation would subtract the area of the pitot

from the test duct.
Area of pitot = .000767 foot2
The same calculations are performed with the new duct area the
final flow being: 90.9 CFM
102
The airflow is a little low for best output; 125 to 165 CFM
would be closer to commercial practice.
The effect of temperature on weight of air calculations: As

the temperature of the surrounding air increases, the density of
the air decreases. Hot air is lighter then cold air. This is obvious
as hot air balloons rise. Our calculations are based on 70
degrees ambient temperature. The weight of 1 cubic foot of air
at 70 degrees (sea level) is .0749 pounds. Air at 0 degrees is
15% heavier; likewise air at 100 degrees is about 6% lighter.
Air at 400 degrees is about 40% lighter. If you wish to correct
for air temperature in your calculations the table of air densities
is listed below.
Density of air at 14.7 psia
Temperature degrees F Density
0 .0863
32 .0807
42 .0791
52 .0776
62 .0760
70 .0749
80 .0735
90 .0722
100 .0709
200 .0602
400 .0461
600 .0400
800 .0337
1000 .0290
The fan output graph shown below is calculated for a 12

blade, 10” wheel, 1.1” wide, running at 5800 rpm.
103
104
Fan Test Duct with Two Manometers
STATIC PRESSURE Vs AVERAGE CFM
10x1.1 INCH WHEEL @ 5800 RPM
BASED ON INLET AIR AT 76° F
INCHES W.C.
As you can see, the fan produces 300 cfm at 0 static
pressure. It produces 19 inches water column at 0 cfm. At our
target pressure of 13.84 inches water column it produces 165
cfm. Which is pretty much exactly what we are looking for.
This is a powerful fan, be careful that the cupola is not
overblown. Pay close attention to your manometers. Know your
target values ahead of time. You will need to throttle this fan.
As stated early in the fan design section, actual output must
be tested and the formula adjusted to reach desired output. The
formulas predict the output will be 300 cfm, which is accurate
at 0 static pressure. However fan output is a function of both
static pressure and cfm. The adjusted formula that represents the
output of 165 cfm at 13.84 inches water column is:
Width of wheel = (747 x cfm) / (D inches x tipspeed feet/min)
All other formulas remain the same.
You may use this new formula in your calculations, however it

may be easier, now that the fan has been tested and graphed, to
scale this fan up or down using the “fan laws” to meet the needs
of any sized cupola.
106
IX. ADDITIONAL CUPOLAS BASED ON THE 15
INCH PROPANE CYLINDER SHELL:
Both larger and smaller cupolas may be built using the 15-
inch diameter propane cylinder. The construction techniques are
the same as described earlier. The main difference is the
thickness of the refractory and the length of the wind belt. The
refractory is thinned to 1.5 inches. Because of this, the inner
cylinder will probably glow at red heat. Much heat will be lost
through this wall, however the cooling effect of the blast should
keep the cylinder from becoming too hot. One advantage of
using thinner refractory will be considerable preheat of the
blast. Actually most of the heat lost through the walls will be
recycled back into the cupola. The thinner refractory will also
lower the cost of building the unit.
Another difference is the bolt on well. Several wells, each of
different capacity may be made for cupola. The 7-inch cupola
has a 10” diameter sleeve of 16-gauge steel that holds the
refractory. This may be difficult to roll; however the local
technical collage may be able to do it for you. Thinner metal
may be experimented with. Maybe several 5 gallon metal
buckets could be welded together to form the inner sleeve. A
disk cut from 16 gauge is welded to the top of the cylinder to
secure an airtight connection for the wind belt. A shop vac
should provide ample blast for the 7-inch unit.
The 12-inch unit is built of several cylinders bolted together.
The bottom cylinder is lined with refractory 1.5-inches thick.
The wind belt is constructed as used on the 10” cupola. The
difference is the belt is extended the full length of the bottom
cylinder. The lining of the upper cylinder tapers from 10 inches
diameter at the top to 12 inches diameter at the bottom. This is
to prevent the charge materials from hanging up in the shaft as
they expand during the preheat. With both cupolas sand
backing is specified to cut down on the cost of the units. A
taphole opening is made by forming a strip of 16 gauge around
the wooden form similar to the one used to form the taphole
107
void in the refractory. The strip is welded into place between
the inner and outer shells to keep the sand from falling through
the shell. An alternative would be to use castable refractory to
fill the empty space between the inner and outer shells of the
well. With the proper blast, a tap of 100 pounds or more is
possible with this cupola.
Tapping the Cupola

Note protective gear includes welder’s hat and mask, goggles,
gloves, long sleeves, long pants, and safety shoes
108
1 Vi inch pipe tuyeres angled
down 15°
\**— 21”
X. OXYGEN ENRICHMENT:
Oxygen enrichment of the blast air is used for two main

reasons, to increase the melt rate, or to increase the temperature
of the tap. Other reasons for oxygen enrichment would include
reduction of coke and the reduction of sulfur in the tap.
If a cupola is producing its maximum output, the output
may be further increased by as much as 25% with the addition
of oxygen to the blast air. Production increases relative to the
percent higher oxygen levels are summarized below.
Oxygen Increase in melt rate
1% 8%
2% 15%
3% 21%
4% 25%
TAP TEMPERATURE INCREASES:
If the blast is reduced and oxygen is introduced to the new

lower blast volume, the temperature of the tap will increase.
The increase is largely related to the increase in radiated heat of
combustion. Radiated heat increases rapidly with an increase in
temperature. The increase is on the order of the fourth power as
illustrated below.
R = K(T4- T4)
R = radiated heat K = a constant
Tj = flame temperature T^ = temperature of iron
Air is 21% oxygen and 79% nitrogen. The maximum

temperature of combustion of coke in air is bout 3500° ;
however if the percentage of oxygen is increased to 25% (4%
111
enrichment) the temperature of combustion exceeds 3800°. At
30% oxygen the temperature of combustion approaches 4200°.
At such temperature the refractory disintegrates and the whole
cupola comes apart. Careful calculation and monitoring of the
airflow are critical to prevent a catastrophic failure of the
cupola. A 4% increase in oxygen should increase the tap
temperature about 175°. The reduction of the blast is dependent
upon the percent of oxygen introduced and is summarized in the
table below.
Oxygen Blast Re<
1% 6%
2% 11%
3% 16%
4% 20%
Calculate the cfm of oxygen for a 165-cfm blast:
Cfm oxygen = cfm x (% oxygen / 100)
Cfm oxygen = 165 x ( 4 / 100) = 6.6 cfm
Calculate the new reduced blast volume for a 4% increase in

oxygen if the current blast is 165 cfm:
New cfm = cfm - ( cfm x (% reduction / 100))
New cfm = 165 - ( 165 x (20 / 100)) = 132 cfm
The tuyere diameter may have to be reduced at the lower

volumes in order to increase the blast velocity to provide
sufficient penetration of the bed coke.
Oxygen enrichment may be used on a continuous or an

intermittent basis. For example if you wanted to reduce the
amount of cold iron at startup you could use oxygen
112
enrichment. Likewise, if you needed a particularly hot tap you
could use oxygen enrichment.
COKE AND SULFUR REDUCTIONS:
Production of ductile iron requires very low levels of sulfur.

Sulfur pickup is reduced by raising the temperature and by
reducing the iron’s exposure to sulfur contained in the coke.
Foundry coke contains about .6% sulfur. A 2% oxygen
enrichment with reduced blast enable one to reduce the charge
coke by 15%, likewise a 15% reduction of sulfur is found in the
iron.
METHODS OF OXYGEN INTRODUCTION:
Oxygen is introduced into the cupola by two methods, either

by injection into the blast main or by injection into the tuyeres.
Most enriched cupolas introduce the oxygen into the blast main
through a diffuser. The use of preheated blast may require the
use of tuyere injection.
Oxygen Diffuser
113
Tuyere injection is accomplished by the use of an oxygen
lance. The tip of the lance is kept about 3-inches from the end
of the tuyere to keep it from burning up. Stainless steel is the
preferred material for the lance used in tuyere injection.
The combination of hot blast and oxygen injection may raise
the temperature of the tuyeres to the melting point. This is
controlled by the use of water cooled tuyeres. Water-cooled
tuyeres are made by tightly coiling copper tubing around the
tuyere. Water is circulated through the assembly to keep it cool.
REGULATION OF OXYGEN ENRICHMENT:
Commercial operations use liquid oxygen in thermos bottles

to supply their furnaces. We will use a welding cylinder,
regulator, and a panel mount flow gauge. A suitable gauge is
made by “King” and is available through MSC industrial
supply. The current price for a 0-20-cfm gauge is $72; a smaller
0-4 cfm gauge is $42.
Because cupola operation is fast paced one would be well
advised to calculate the desired inches pressure to be displayed
on the Pitot tube and manometer. Likewise the oxygen flow rate
should be calculated beforehand. A chart is made up and
displayed beside the meters. Your manometer could be
calibrated in cfm instead of inches to eliminate the volume of
airflow calculations during melting.
114
XI. PURCHASE OF COKE:
Coke is sold in several sizes. Most large consumers

purchase coke that is 3 inches in diameter and greater.
Some suppliers sell smaller coke, often called “nut
coke” that is 1 Vi inches and smaller in diameter. This
size is easy to use in the small cupola. If you purchase
the larger coke you will have to crush it to the correct
size.
Coke is usually sold in train car quantities and larger.
Remember this when dealing with your coke supplier,
because he is often busy with several large orders. Most
Suppliers will work you into their schedule, but be
patient.
A ton of coke will fit nicely into a full sized pickup
bed. I use a Chevy half-ton pickup and have no
problems. It usually takes about two scoops from a
front-end loader to fill the truck and the whole process
takes about five minutes. I am currently paying 5 cents a
pound, making the ton of coke $100. Some suppliers sell
a ton bag. They dump the coke on a tarp, then tie it up at
the top. It is put on a pallet and shipped in this fashion.
Often the freight cost more than the coke. A ton of coke
will give about 10 heats in the 10- inch cupola.
Coke may be found locally through the phone book,
but I found my coke through the Internet. Another
source of coke might be a local foundry.
I drive to Alabama to purchase my coke, as there are
several coke companies located there. I found the people
at Empire Coke the most accommodating; they even
took me on a tour of the plant. Another popular
company is ABC coke. There are several coke suppliers
throughout the country and you may be fortunate
enough to find one close to your location.
115
XII. CONCLUSION:
Cupola melting is not difficult. You can successfully melt

without ever performing a single calculation. Do not be
intimidated by the mathematical descriptions provided in the
text. Chapters II, III, and IV certainly contain enough
information to successfully melt iron; however if you want to
design and operate your own cupola, the math is a very valuable
tool that should provide you with good results. Remember every
cupola will operate over a wide range of conditions. The
calculations will put you in the ballpark but you will have to
fine-tune your operating procedure by trial and error.
Your first cupola should be a small one, 10 to 12 inches or
less. The temptation to build a large cupola is great, however
you should resist until you have gained some experience on a
small one. Because cupolas are such fast melters, you will have
to have all your molds ready ahead of time. Preparing such a
large quantity of sand is more than one person can easily
handle. The 10-inch cupola, when operated with proper blast
will melt 660 pounds of iron per hour. A 1 hour heat will
require about 2 tons of sand. The amount of sand used for a
large cupola can be overwhelming. My advice is to start small.
You’ll soon know what you can handle.
For a small operation that occasionally needs a large tap, a
small cupola with a deep well and oxygen injection might
provide you with the extra capacity required. Oxygen injection
is for advanced and experienced cupola work and should not be
attempted by the beginner.
Iron exists in many forms, gray, ductile, white, mottled,
malleable, etc. You should read a book on basic metallurgy.
Initially you might not understand it, but after inspection of the
castings from your first few melts you will. Try putting a few
charges of steel in the cupola and pour some test bars. Fracture
the casting and you will see white iron. It is extremely hard and
brittle. Your first few heats will be experimental so you can
learn the process.
116
Operating a cupola is hard work, but there is nothing like the
excitement of tapping white-hot iron and filling a mold. I
certainly enjoy it and you will too. Now that you can produce
iron castings you are only limited by your imagination. Good
Appendix
SUPPLIERS:
American Foundrymen’s Society
505 State Street
Des Plaines, IL 60016-8399
Foundry books, from basic to advanced

800 537 4237 www.afsinc.com
modemcastaing.com
COKE:
ABC Coke Hickman Williams

800 321-4015 US 800 862-1890
205 849 1300 Canada 800 265 6415
Mexico 011 52 8 363-4041
Empire Coke ASi International

Holt, Alabama Cleveland, OH
205 323-2400 800 860-4766
800 544 8702 216 485-1800
Asbury Cast Metals Group Miller & CO

Sunbury, PA Rosemont, IL
717 286-2176 800727-9847
847 696-2400
St. Louis Coke & Foundry Supply
314 727-7500
Joy-Mark Inc.
2121 E Norse Ave.
Cudahy, Wis. 53110
414 769-8155
Ceramic reinforced ladles.
Part# 407 10x8.5x 11.375
Lindsay Publications 815 935-5353

P.O. Box 538 lindsaybks.com
Bradley, IL 60915 (Foundry books , Technical books)
MSC Industrial Supply Co.

75 Maxess Road
Melville, NY 11747-9415
(800) 645-7270
Request the “Big Book” catalog

118
Budget Casting Supply
San Mateo CA
650 345-3891 budgetcastingsupply.com
Piedmont Foundry Supplies

P.O. Box 87
3191 Rogers Lane Columbus, GA Phone 324-5938
Cloverdale, VA 24077 Phone 992-3911, Greensboro. NC, Phone 294-2661
Porter Warner Foundry Supplies:
Alabama 205 251-8223 Tennessee 423 266-4735

Arizona 602 244-9166
California 213 722-1335 TX 817 633-1767
501 633 0285 713 222 6210
909 822 5591
South Carolina 803 789-3444
Uni. West Foundry Supplies:
Denver,CO 303 388-0922 Washington 206 767-9880
Oregon 503 226-4836
Rice Industries
Minnesota 651 784-1881
Stewart Marshall
P.O. Box 279
Lopez Island, WA 98261
“Building Small Cupola Furnaces" Book, Foundry CD.
Australian Foundry Institute

1 St. Peters way
Glengowrie, SA 5044
61 8 8295 67 12
Canadian Foundry Association

Ste. 1500
Ottawa, Ontario KIN 7B7
Canada
Cast Metals Institute www.castmetals.com

800 537-4237
119
C/>
0)
1079
1119
1403
1435
1466
05
1036
1197
1234
1269
1304
1338
1371
699
co co
518
733
792
846
898
946
992
S cm O) m
c ■o- m
m
t/>
a>
sz in <n o> o T— CO OO CM O' f" r-~ f' co CM o r-~ t—
o S o> a m I-" 3 CM m 8 co m r-~ 3 CO -o 3 CO
c CM CO CO co ■M- ->t ■M- m in m m in co co co co co C" c- r-~-
</>
o 0)
Q. ■C
U in ■o- r- CO CO o o co co CO 05 m o m 05 CO r- o CO
Q_ Sl T" X* r- o CM m r- 05 CM co r" 05 CM CO lO 00 05 8
CM CM CM CO CO co co CO X* 'T 3 X* •o ■'T m m m m 8 m m CO
Q.
»*-
O) o
□
o
l.
S (/>
a> 0)
F T— O m f- co m CM r- CM r- o CO CO o T— CM CO CO xt co CO
■■■ ns € CO CO CO o CM M- CO r- O) O CM CO 3 m r~- co 05 o CM 3 xfr m
I— C T— CM CM CM CM CM CM co co co co co co CO CO m- ■M- M- M- M- M-
Q
CO
5
Li. </>
0)
o .c
o C" CO o> o O) 00 r~- CO ■M- CM 05 r-~ •0- r— CO M- r— C- co
Lm c m xj- o
CM 3 ■O' in co CO CO O) o CM CO co x* m co CO r-~ CO CO 05
oo o
T— T“ M— T“ r— ^— T— T— M— CM CM CM CM CM CM CM CM CM CM CM CM CM
eg
£ T-
o
*E </>
05
sz M-
5
24
30
42
47
57
67
76
80
o> CO CM CM CO CM
56
68
88
< o fs. O 8 m CO 3 CO 05
c 8 y— T— "r- T— T— T— T—
0) T~
CM
o>
m
}v
l_ c
a> F
CO o co o CN CO CM r- CM 05 m o m m
1 8 X* o r- o> 8338 CO CO § m CO r- m- 8 co CM
<D
< >> T“ CO m in CO CO CM y— o CM in d in CO in CO CM CO o
CO CO o r-~ o> ID CO yf 8 05 m 8 05 CM m c- 05 o M—
o CO Tf o 8 CM (ft o CO (D O) a M- r-~ o CM i-~ 05 CO CO
CM CO •O- 5 -a m U*) CD CD CD co r-- r- r- CO CO s oo CO 05 05 8 05
o
5
C/j O m o m o m m o m o m o m o m o m o m o m o
a) m r*» o CM m r- § CM m o CM m r-- o CM in r- o CM m r- o
-C d o CM CM CM CM co co co co xf ■M- M- x* in in in in co
o
c
120
® w m n r ro NOlifOinOtCONIOOinMtNfflr-n
..,
CT) CM cn CO ..
CO ,»i-^M»(M^(OBT-n»(OONlON .
T»-ininif)co(£ic£icoh~r^r'h'Cocococoo>o>3>0)00ooo^-
NMcontortNT-inoirtiooiNm eoocoif>NO)T-N
OlrNjnMOO sas^ssalllliiii5|S
h-COCOCOCOOOCOO)
» O N t MOOlO'r^-CMNNNNN rOOlOlBMO^
t- n 't m SCOOt-CMWÎOCONCO®- OT-cNCMcoojintor^
....
CO CO CO CO (DlOIONNNNNNNNNOOCOCOCO co co oo co
eg t- o o> co
CD S inni-aMONOMONaipnoNt'-N^
NeoeooiOT-NnrtîniDNseooiOJOT-CMNtnwîn
vitTfîninininininininininininincD(0(0(Di0®(D(D
OllOT-SNCOnBVOÔlÔl
oiOr-T-wcNmnTt'tinm®®
tt co co oo cm
o _
So m o) to
T~ T- CM
-* ■o- ^ ^ ^ M-
N N ffl W Ol ffl
(Mnnconnnnnnrtrtnn CO CO CO CO CO CO
CO O ■U- CO CM UO O CM CMCOO>CMmcOCMlOCOT-^COO>CM»OCO
5> o o o coeoco^^vmininoooosNS
T- CM CM CM CM CM CM CM CMCMCMCMCMCMCMCMCMCMCMCMCMCMCMCM
SNmCMSi-M-K^Tf
oiocooi^®ino)(0
CM CO
eoscoeocgmcoo»-
(DCMOr-CpMSinco
CNOirifflncooiNnidfflioo't CO CO T- 05 CO N (J>
rr-00)XMO'fCMOS'M-C»t co co co cm i^- i- m
OCM'CWNfflrCOWOBOt-n CMO-me-CDOl-
O O OOOt-t— t— r- CM CM CM CO (O (O CO CO M M
momoioomomom O If) _ ojng^ O B O B O m

CM B S O CM B
uo o
cm m t'-
CMlONOCMinNONiflN o cm ® s
®®®NSNN®®®MO)Ol COlOOOOr-^T-T-CMCMCM
121
INDEX Clay, 40,53
Cleaning, cupola 57,58
Adapter, Flanged, 34,35 Coke,
Air Fines, 52
Flow in pipe, 120,121 Purchase of, 115
Flow, velocity of 101 Screen, 52
Oxygen in, 111 Size of, 52
Pre-heater, 107-110 Sulfur in, 113
Supply, 39 Weight, 53
To coke ratio, 16 Combustion, zone, 11,13
Weight of, 103 Core sand and making, 90
Cupola zones, 11-13
Balancing ways, 82-85
Base plate, 37,38 Desulfurizing, 14
Bearings, ball 81 Diffuser, oxygen 113
Bed Door
Burning in, 54 Bottom, 37
Coke height, 21 Rear, 35,36
Coke measuring, 49 Slag, 45
Blast
Penetration, 17 Efficiency,
Theory of, 18 Cupola,
Preheated 107-110 Fan, 59,68
Blower
Roots, 59, 60, 64, 65, 66 Fan laws, 71,75
Size, 59,61,62,63 Fan performance graph,
Bod Designing from, 67,71-75
Clay, 56, 57 Graph, 105
Rod,49 Roots, 64,65
Bottom Plotting, 67
Door, 37 Fan plates, 78, 81
Dropping, 57 Flame, discharge, 4
Sand, 44 Frozen taphole, 14
Fines, 14
Carbon dioxide, 4, 15 Firing lining, 45, 46
Carbon monoxide, 4, 15 Form
Circle cutting attachment, 28, 29 Lining, 41
Centrifugal Taphole, 43
Blower, completed, 94,104 Furnace, lining 40-46
Force, calculation of, 69, 70
CFM Gas Discharge,
Calculation of, 101-103 Composition of, 15,16
Ratio, air to coke, 16 Temperature of, 4
122
Green Pac, refractory, 40 Performance curve, fan, 67, 105
Pitot tube, 95, 96
Heat Plate,
Combustion temperature, Bottom, layout, 38
Increase of, 111,112 Fan, 78,81
Loss, 15 Steel, drop, 52
Radiated, 34 Pre-heat,
Formula for, 111 Air, 107, 110
Height of, Charge, 11,12,55
Bed coke, 21 Zone, 11,13
Cupola above tuyeres, 19,20 Pressure, static, 59,60
Legs, 20 Pulley, size calculation, 75, 76
Tuyeres, 19,20
Horsepower, required by fan, Ramming, lining, 41
68,74 Reduction zone, 11,13
Housing, scroll, 70,71, 85 - 88 Refractory, 40,41
Regulator, oxygen, 114
Inlet diameter, fan, 68 Rivets, 77,78
Latch, Sand,
Back door, 36,45 Bed in cupola, 44,46, 53
Bottom door, 37,39, Bed under cupola, 52
Lining, Lining, 40, 107-110
Firing, 45-46 Molding ,115
Form, 41 Scroll housing, 70,71, 85 - 88
Furnace, 40-46 Shell, 30,31, 107, 108
Site layout, 50, 51
Manometer, 60, 95, 97-101,104 Size of charge material, 14
Material, charge, size, 14 Slag door, 45
Mechanical, stress in fans, 69 Soaking, 55
Melt rate, 16 Sparks, 14, 52,
Increase of, 111 Spout, 40
Melting, 55 Starter fluid, 46
Melting zone, 11,13 Stress, mechanical, in fans, 69
Molding, sand 115 Sulfur, in coke 113
Nitrogen, in air, 111 Taphole, 43,44,53

Form, 43,
Oxygen enrichment, 111-114 Frozen, 55, 56
Diffuser, 113 Making, 53
In air, 111 Tapping the cupola, 56
Temperature of combustion, Temperature, increase due to
Increase, 111,112 Oxygen injection, 111,112
123
Test duct, 101,102,
Photo of, 104
Throttle, 34,
Drawing of, 113
Tip speed, fan 68, 69
Transition piece, 90-93, 101
Tuyere
Area calculation, 17, 18
Downward angle. 17,18,23,27
Height of, 19-20
Water cooled, 114
Vane, fan, 79, 80

Venturi, inlet. 89, 93
Weight of air,
Correcting for temperature.
103
Well, 11-13
Capacity of, calculation, 19
Depth of, 19-23
Wheel diameter, fan. 67,68
Wind belt, 32-34
Zones, cupola, 11-13
124
ORDER FORM
Additional copies of “Iron Melting Cupolas for the

Small Foundry” May be ordered for $19.95 plus $3.20
postage. A limited number of videotapes of an early
melt are available. E-mail Steve regarding price and
availability.
Name_
Address_
City_State_
Zip_Phone_
E-mail address_
Mail to: Steve Chastain

2925 Mandarin Meadows Dr.
Jacksonville, FL 32223
125
THIS PAGE IS BLANK
but this is not a printing or scanning
fault and no content is missing.
126

Iron Melting Cupola Furnace by Stephen D.Chastain Text

Uploaded by

Copyright:

Available Formats

Iron Melting Cupola Furnace by Stephen D.Chastain Text

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Iron Melting Cupola Furnace by Stephen D.Chastain Text

Uploaded by

Copyright:

Available Formats

Iron Melting Cupola Furnaces

For the Small Foundry

Written and Illustrated by Steve Chastain

Copyright © 2000 by Stephen D. Chastain.

Purpose & Introduction 8

I. Cupola Theory and Design Considerations 11

II. Building a 10 inch Cupola 28

III. Cupola Operation 49

IV. Air Supply and Blowers 59

V. Designing Centrifugal Fans 67

VI. Construction of Centrifugal Fans 77

VII. Construction of a Pitot Tube and Manometer 95

VIII. Calculation of Air Flow 101

IX. Additional Cupolas based on the 15” Shell 107

X Oxygen Enrichment 111

XI. Purchase of Coke 115

XII. Conclusion 116

Appendix: Suppliers 118

Airflow through pipe 120

This book is to provide information on the methods used by the

Warning: As with any fire, the incomplete combustion of coke

MATH: The math used in this book is no more complicated

Since I wrote “BUILDING SMALL CUPOLA FURNACES"

I grew up in central Florida, and have always been considered

I would like to thank Dave Gingery and Stewart Marshall

The purpose of this book is to provide a set of plans for a

Old engines fascinate me and I love to restore old machines.

Machining my first Piston Casting

This piston went in a 10 kW generator set that powers my house

Pouring a 60 pound faceplate for a lathe

The size of the metal making up the charge is critical to

COMMON CUPOLA RATIOS:

Conventional industrial cupolas are built to melt from 2 to

Thermal Balance Sheet:

Of all the energy contained in the coke only 25% to 65% is

Lining and Tuyeres 4 to 20

TOTAL LOSSES: 35% to 75%

HEAT LEFT FOR MELTING: 65% to 25%

Melt rate in pounds = 10 7t (D/2)2 pounds / hour

Melt rate = 107t(10/2)2 = 785 pounds / hour

The maximum ratio of iron to coke is 10 to 1; ten pounds of

The tuyere area is based on the inside diameter of the cupola

Combustion Moves Toward Moves Away

Large Smaller Angled

7t = 3.14 A= 7t(10-inches/2)2 = 78.5 inches2

Tuyere area = 1/6 A = 1 X 78.5 inches2 = 13.1 inches2

13.1 inches is the required area. Now to find a suitable diameter

area of 2-inch pipe = 7i(D/2)2 = iz(2/2)2 = 3.14 inches2

Number of tuyeres = (13.1 inches2/3.14 inches2) = 4.17

Area of 2 Vi-inch pipe = 7t(2.5 inches / 2) = 4.9 inches

Number of tuyeres = 13.1 inches2 = 2.66

4 tuyeres at 2-inches diameter are closer to the desired ratio than

Ounces of iron = 35 pounds iron X 16 ounces = 560 ounces iron

Cubic inches of iron = 560 ounces of iron = 127 inches3

Well volume available for iron = .46 total well area

Total well volume = 127 inches3 = 277 inches3

Height to slag hole (from bottom of well) = total well volume

Area of well = 78.5 inches2

Height to slag hole = 277 inches3 =3.52 inches