Simpson Report Page 1

Reduction in Sand-Related Scrap

Through Effective Sand Cooling
M. J. Aklinski
National Engineering Company
Chicago, Illinois
M. J. Granlund
Foundry Systems Control
Winfield, Illinois
ABSTRACT
Hot sand is sand that has a temperature so high that
difficulties are encountered in preparation of the sand, in the
molding of the sand, and in the casting results. The actual
temperature of hot sand varies according to the authors cited.
Generally, sand of 120F or higher is considered hot.
Schumacher stated that sand over 160F (71C) does not mull
to any consistency in physical properties, but sand below
120F develops uniformity when mulled. Between 120F and
160F mulling produces sand that is inconsistent and difficult
to control.
This paper presents data on sand temperatures taken over a
long period of time before and after the sand cooler
installations. The effects of hot sand on various properties,
and on the scrap rate before and after sand cooler
installations, are discussed. Here again, the time period is in
months of data. A discussion of sand cooling principles is
included as an appendix..
INTRODUCTION
Hot sand is the number one problem in foundry green sand
systems and is the cause of more casting scrap than any other
sand system problem. Hot sand is any sand so high in
temperature that difficulties are encountered in the preparation
of the sand, in the molding of the sand, and/or in the casting
results. The actual temperature of hot sand varies according to
the author cited. Generally sand of 120F or higher is
considered hot. Schumacher stated that sand over 160F (71C)
does not mull to any consistency in physical properties, but
sand below 120F develop uniformity when properly mulled.
Between 120F and 160F, mulling produces sand that is
inconsistent and difficult to control.
The data presented was taken over a long period of time. The
temperature data covers several weeks before and after cooler
installation and the scrap data covers a nine-month period before
cooler installation and nine months after installation. The long
term trends are shown in temperature reduction as well as scrap
reduction.
The foundryman has been aware of the problem of hot sand and
its many ramifications, but has done very little about it. What has
caused the new interest in the subject? The more restrictive
quality requirements that have come with the greater
competitiveness of the world casting market, lower sand-to-metal
ratios, higher line utilization, and insufficient sand storage space
have been large factors in this interest. Casting requirements have
become more stringent and today's customer will not accept the
castings that were sold several years ago. The casting must be
made right the first time, with less rework than in the past.
HOT SAND: WHERE DOES IT COME FROM?
In a regenerative green sand system, the composition of the sand
mix is formulated to meet the requirements of the molding
method and the type of casting produced. The typical system
sand contains varying percentages of sand, bond, additives, and
water. The sand system, the muller, and the molding machine
process this mixture into a finished mold. To form the casting,
molten metal is poured into the mold after which the heat of the
metal causes damage to the mold and its materials. The damage
that occurs will vary directly with the distance from the metal-
mold interface. With successive cycles, if there is no cooling
period, the heat content in the sand builds up to excessive levels.
How Does Hot Sand Affect the Foundry?
Hot sand will affect every aspect of a green sand molding
operation. Depending on the foundry's ability to control the sand
temperature, the results can range from higher than normal scrap
rate to complete loss of system control.
Among the casting defects caused by hot sand are: sand
inclusions, washes or erosion scabs, surface roughness, pin holes,
crushes and broken molds. Sand inclusions, whether from erosion
Simpson Report Page 2
type of defects or expansion type of defects, and their
reduction with the use of cooled sand, are the major
consideration in this paper. Erosion defects are surface
defects that consist of lump-like protrusions of metal above
the intended surface of the casting. The lump replicates the
depression caused by the erosion of the mold. Loose sand
may or may not become embedded in the surface of the
casting. The excess metal must be removed. Even more
important than the removal of the scab is the location of the
loose sand. This embedded sand may or may not scrap the
casting, but it will cause a detrimental effect on any required
machining.
The mechanism of erosion defects is the dislodging of sand
grains or sand masses from the mold surface by metal
impingement, metal turbulence, or gad agitation. Increased
moisture in the molding sand enhances the erosion tendency.
The increased moisture gives increased steam potential, which
gives rise to more turbulence in the metal flow.
Scabbing defects are caused by the mold surface rapidly
heating and converting the moisture into steam. The steam
moves further in to the mold and condenses and adds to the
normal moisture content of the sand. This combination forms
a condensation layer of high moisture content. If the
expansion forces developed by the expanding sand grains in
the dry layer exceed the strength of the wet layer and the
constraints of the mold, the mold surface begins to deform and
expansion defects result. Rattails buckles, and scabs occur
when the mold surface is unable t accommodate the expansion
of the sand on heating by the metal poured. Rattails and
buckles are the same type of defects in an insipient form,
which did not have time to develop into scabs before the mold
was filled and the casting solidified. Rattails occur primarily
on the drag surface, while buckles and scabs usually occur on
the cope surface or on vertical surfaces.
TOTAL SAND SYSTEM CONSIDERATIONS
In the late 1970's an Ad Hoc Committee of the AFS Molding
Methods and Materials Division was formed to determine
what needed to be done to foundry system sand to prepare it
for green sand molding. In the 1980 AFS Transactions, Lee,
principal author, reported the committee's findings. The
conclusions reached were the following:
1) There is a wide and unfavorable variability in the quality
of molding sand produced in mechanized foundries.
2) Standard sand testing does not generate enough
significant data to design or control sand systems.
3) There is not sufficient time or control in the muller to
correct for imbalances created in use, shakeout, and
handling of sand.
4) Moisture control by automatic final-state correction is
inadequate.
5) It is a conceptual error to concentrate controls in the muller, a
sand system includes the mold pouring and cooling zone,
shakeout, return belts and elevators, screens, bins and
hoppers, and prepared sand delivery-as well as the molding
machine.
6) Control must be exercised in each part of a sand system in
order to effect control of prepared sand.
7) Used sand must be monitored and makeup additions adjusted
accordingly.
Because of the variability that the total system contributes, the
muller cannot be relied upon to maintain control and reduce the
variability. Completeness of mulling is a very important factor in
developing the maximum strength of the bond available and plays
a very important part in resistance to erosion and expansion
defects. The main objective in the preparation of molding sand is
the uniform and complete dispersion of the clay binders and the
water, in the proper amounts, in the sand aggregate. For the
successful production of castings, it is essential that the supply of
molding sand be consistent in properties and quality. Because
system shakeout sand varies widely in moisture content, residual
binder, and temperature, some means of compensation for these
variations must be developed. The use of a sand cooler to control
the variable temperature of the return sand is one way to
compensate for the shortcomings of the system.
Appendix A gives more detail on the principles of physics that ate
applied in cooling sand.
FOUNDRY SAND AND CASTING SCRAP RESULTS
The foundry to be considered is a gray iron foundry located in the
Midwest of the United States. The foundry decided that because
of the very high temperature of the prepared sand and the high
variability of the sand, which in turn produced a high scrap, they
would install coolers in both of their production molding lines.
The shakeout sand temperature before the cooler installation are
listed in Table 1. The data show that the foundry's assessment of
the sand temperature, but it was also extremely variable within a
day, or from day to day.
The type of cooler installed is the double-pan continuous cooler
shown in Figure 1. The unique design of this cooler, which
utilizes the evaporative cooling of air and water with controlled
retention tie, meets the requirements for consistency in properties
and quality of return sand. The cooler capacity/volume is
relatively large in relation to the feed rate and, therefore, the
retained mass absorbs the variations in the entering sand. The two
pans each contain a counter-rotating head, to which are attached
four mixing plows. The purpose of the plows is to move the sand
forward, from charge point to discharge point, and to keep the
sand in a state of agitation. This allows the cooling water and the
Simpson Report Page 3
air to be intimately mixed with the sand causes evaporative as
well as convective cooling. Because the incoming sand is
highly variable, the double-pan design and counter-rotating
plows will constantly back-blend the sand. The plows are set
at different vertical heights from the cooler bottom to get
overlap in mixing from top to bottom.
The cooling air, supplied from a separate motor-driven fan, is
introduced not the plenum chamber behind the cooler's inner
wall. The air purges the sand of the water vapor produced by
controlled moisture additions to the hot return sand. A self-
adjusting door keeps the discharge rate equal to the incoming
rate. The door is actuated by an electro-hydraulic device that
controls the desired volume and retention time of the and in
the cooler.
The amount of moisture added will depend on the sand geed
rate and the temperature of the incoming sand. The moisture
content is maintained at a level that will assure enough water
for cooling with a targeted residual water in the sand
discharge.
Table 1 shows the shakeout temperature data from line #1
prior to the cooler installation. All the shakeout sand
temperatures exceed 230F and have a large range in
temperatures. The shakeout sand went into one bucket
elevator and two transfer belts before entering the sand storage
bin. Additional ambient cooling would take place at all of
these points. The temperature data from line 2 (not shown) is
similar in nature. Both lines operate with the same type of
molding equipment, and patterns are run interchangeably on
both lines. The foundry has kept records on prepared sand
temperature and still does record the temperature of each
tested sand sample.
Table 2 gives the data on the average, range, and standard
deviation of prepared sand temperatures for both lines. These
sand sample temperatures are compared for February, a cold
ambient temperature period, and July and August, a warm
ambient temperature period, and for both before cooler
installation and after cooler installation. It must be noted that
these are temperature of sands prepared on two separate but
similar lines. Variations come from slight metal mix changes,
mulling equipment, and handling equipment. The decrease in
temperature, in actual degrees change and also in the percent
change in the range and standard deviation, indicates that in all
cases the sand temperature is lowered and the variability has
been lowered on the "after sand cooler installation" samples.
The more consistent the sand is the more consistent the molds
will be and, in turn, the more consistent the castings will be.
Figure 2 presents the shakeout sand temperature before and
after the cooler installation. This data, which is based on a
single complete day's comparison, shows the effect of the
cooler in the
Table 1
Line #1 Shakeout Sand Temperature Prior to Cooler Installation
Temperature C Temperature F
12:00 PM 220 428
12:30 220 428
1:00 120 248
1:30 200 392
2:00 190 372
2:30 220 428
3:00 220 428
4:00 220 428
5:30 160 320
6:00 330 626
6:30 260 500
7:30 150 302
8:00 150 302
8:30 150 302
9:00 160 320
9:30 220 428
10:00 110 230
10:30 130 266
___________________________________________________________
Average Temp 190 375
system. Before the cooler was installed the average shakeout
temperature was 375F, and after the cooler was installed the
shakeout temperature was 258F, with the cooler in the system the
heat buildup in the sand system was controlled to a lower average
shakeout sand temperature.
Figure 3 shows the shakeout temperature and the out-of-cooler
temperature for the same day on line #1. The shakeout sand
temperature, even with a cooler in the line, has a great deal of
variability, while the out-of-cooler sand runs in a very narrow
band on temperature. The sand out of the cooler has retained
moisture, therefore, the amount of moisture required for temper at
the muller is less. This increased moisture makes the mulling of
the sand less difficult.
Figure 4 shows the difference in prepared sand temperature for
July and August, a high ambient temperature period for line #1,
while Figure 5 shows the difference n prepared sand temperature
for February, a low ambient temperature for sand, before the
cooler was installed, was higher for the cold month of February.
This was most likely due to the sand-to-metal ratio and line
utilization at that time. The average prepared sand temperature,
after cooler installation, in both cases was below 110F. The
higher temperature of the ambient air used in the cooler may
explain why the cooling, percentage-wise, in the warm months
was not as great as in the cold.
Simpson Report Page 4
Table 2.
Prepared Sand Temperature for Both Lines
Before and After Cooler Installation
LINE #1
Before After
Cooler Cooler Actual Percent
Installation Installation Change Change
February February
Average Temp (F) 141.3 100.2 -41.1 -29.1
Range 30.1 13.0 -18.1 -60.1
Standard Deviation 12.9 5.6 -7.3 -56.6
July & July &
August August
Average Temp (F) 135.0 107.2 -27.2 -20.1
Range 18.3 13.0 -6.3 -34.4
Standard Deviation 7.9 5.6 -2.2 -40.5
LINE #2
Before After
Cooler Cooler Actual Percent
Installation Installation Change Change
February February
Average Temp (F) 139.8 94.8 -45.0 -32.2
Range 19.2 15.1 -4.1 -21.3
Standard Deviation 8.3 6.5 -2.8 -33.7
July & July &
August August
Average Temp (F) 144.8 100.9 -43.9 -30.9
Range 24.9 9.7 -15.2 -61.0
Standard Deviation 10.8 4.1 -6.7 -62.0
Figure 6 shows the difference in temperature for line #2 for
July and August, the warm period, while Figure 7 show the
temperature for February, the cold period. On these two
graphs the temperature of the prepared sand was higher for the
warm months than the cold month sand, as would be expected.
In both cases the temperature of the prepared sand was around
100F or less.
Table 3 lists the sand properties before and after the cooler
installation on line #1 and Table 4 lists the same properties
before and after cooler installation on line #2. The sand
properties have not changed significantly. The standard
deviation has gone down in many cases and stayed constant in
the others. The molds made are more consistent and the scrap
rate is lower after cooler installation. The amount of bond
added, per batch, has not changed. The MB clay is slightly
higher than on the before-cooler sand. The cooler sand allows
the water to combine with the bond in a shorter time and, in
turn, this appears to tie up the bond so not as much is drawn out
by the dust collection system on the mullers.
Table 3.
Line #1 Prepared Sand Properties
Before and After Cooler Installation
February July & August
(Cold Ambient Air) (Warm Ambient Air)
Before
Cooler
Installation
After
Cooler
Installation
Before
Cooler
Installation
After
Cooler
Installation
Moisture
Average 3.1 3.7 3.8 3.6
Std. Dev. 0.1 0.2 0.2 0.2
Compactability
Average 45.7 47.4 50.0 44.0
Std. Dev. 2.7 2.0 3.7 2.5
Green Strength
Average 17.8 19.3 16.6 18.0
Std. Dev. 1.2 1.1 1.3 1.2
M.B. Clay
Average 7.9 8.5 8.1 8.2
Std. Dev. 0.4 0.4 0.3 0.3
Temperature
Average (F) 141.3 100.2 135.0 107.2
Std. Dev. 12.9 5.6 7.9 5.6
Table 4
Line #2 Prepared Sand Properties
Before and After Cooler Installation
February July & August
(Cold Ambient Air) (Warm Ambient Air)
Before
Cooler
Installation
After
Cooler
Installation
Before
Cooler
Installation
After
Cooler
Installation
Moisture
Average 3.1 3.7 3.7 3.8
Std. Dev. 0.3 0.3 0.3 0.3
Compactability
Average 43.2 44.4 50.0 43.2
Std. Dev. 3.8 4.8 4.9 4.3
Green Strength
Average 18.2 16.9 14.4 17.0
Std. Dev. 1.2 1.1 1.7 1.4
M.B. Clay
Average 8.3 8.3 8.0 8.3
Std. Dev. 0.5 0.3 0.6 0.3
Temperature
Average (F) 139.8 94.8 144.8 100.9
Std. Dev. 8.3 6.5 10.9 4.1
Simpson Report Page 5
Because of lack of data on the sand temperatures entering and
exiting the cooler, line #1 was monitored completely for the
two shifts run one day. The results on checks on line #2 were
not significantly different, so only the line #1 data is
presented. Table 5 lists all temperatures and properties for one
complete day. The data gathered was the shakeout
temperature, shakeout plus spill sand temperature, sand
temperature out of the cooler, electronic probe temperature of
sand entering the cooler, laboratory sand sample temperature,
sand to metal ratio, moisture on the laboratory sample,
compactibility, MB Clay, green compression, and
permeability. The shakeout into the cooler and the out-of-
cooler sand was sampled in a gallon paint can and a
thermometer was probed into the sand in four to five places.
This probing gives the range shown in Table 5. There is no
explanation for the lower green compression strength for the
first five to six hours of operation. The afternoon and evening
strengths are back up to the expected 22-24 psi range. The molds
were of comparable quality and the next day's scrap sheet did not
show any significant changes.
One of the measures of the consistency of the sand is the scrap
rate reduction. The principal sand defect was sand inclusions.
Figure 8 shows the scrap rate before and after the cooler
installation. This scrap reduction was 34.5% for the nine-month
period following installation versus the nine-month period before
installation. Not only was the amount of scrap reduced, but the
variation in the scrap rate was also reduced. The metal mix for
the two nine-month periods was very similar, and the various
castings were made on both lines during both periods. The
foundry stated that the reduction in scrap, due to reduced sand
inclusions, was great enough to return the investment in the
coolers in less than 15 months.
Table 5.
Temperatures and Properties on Line #1
(Typical results after sand cooler installation)
T
i
m
e
S
h
a
k
e
o
u
t
T
e
m
p

(
F
)
O
u
t

o
f

C
o
o
l
e
r
T
e
m
p

(
F
)
S
a
n
d

L
a
b
T
e
m
p

(
F
)
M
o
i
s
t
u
r
e
C
o
m
p
a
c
t

(
%
)
M
B

C
l
a
y

(
%
)
G
r
e
e
n
C
o
m
p
r
e
s
s
i
o
n
P
e
r
m
e
a
b
i
l
i
t
y
6:30 90 3.7 49 8.4 19.4 130
7:30 85 4.0 44 8.7 19.8 125
9:00
190-
215
174 104
9:15
9:30
200-
235
167 90 3.8 47 8.6 18.2 134
9:45 178
10:00 88
10:15
230-
240
171
10:30
200-
235
210-
215
187 102
10:45
11:00
170-
190
175 107
11:15 84
11:30
200-
215
190-
210
11:45
190-
210
185-
195
180 96
12:00
12:15
12:30
180-
195
175 93 90 4.0 48 8.5 17.4 140
12:45
1:00
210-
225
180-
195
T
i
m
e
S
h
a
k
e
o
u
t
T
e
m
p

(
F
)
O
u
t

o
f

C
o
o
l
e
r
T
e
m
p

(
F
)
S
a
n
d

L
a
b
T
e
m
p

(
F
)
M
o
i
s
t
u
r
e
C
o
m
p
a
c
t

(
%
)
M
B

C
l
a
y

(
%
)
G
r
e
e
n
C
o
m
p
r
e
s
s
i
o
n
P
e
r
m
e
a
b
i
l
i
t
y
1:15 98
1:30
190-
210
171 94
1:45
195-
230
2:00
240-
265
210-
240
105
2:15
350-
500
300-
325
100
2:30
275-
305
300-
340
90
2:45
230-
250
200-
240
106 92 3.8 45 8.8 24.0 145
3:00
250-
275
180 100
3:15 102
3:30
250-
290
96
3:45 92 3.7 44 8.6 23.0 135
4:00
275-
300
102
4:15
260-
305
200-
230
190
4:30
260-
290
200-
230
100
4:45
220-
250
104
5:00
6:35
240-
270
200-
230
100 96 3.7 41 8.8 24.2 130
8:40 92 3.7 38 8.8 23.6 120
Simpson Report Page 6
SUMMARY
The high and variable temperature of molding sand makes control
difficult and increases the scrap produced on any line in the foundry.
Installation of sand coolers decreased the scrap produced and decreased
the variability of all sand properties. The installation of the sand
coolers reduced the scrap rate enough to return the investment on the
coolers in less than 15 months. Not only was the scrap reduced, but the
general overall casting appearance was substantially improved.
ACKNOWLEDGEMENTS
The authors wish to thank Fred Schuster, National Engineering Co., for
his assistance in gathering the data. Our thanks also go to Carol Vaseff,
National Engineering Co., for her assistance in preparing the graphics.
REFERENCES
1. J. S. Schumacher, E. H. King, and R. W. Heine; "The Problem of
Hot Molding Sand," AFS Transactions, vol 66, pp 261-267 (1958)
2. R. S. Lee; AFS Committee 80 A-2 Report, "Sand System
Functions, Requirements, and Specifications," AFS Transactions,
vol 88, pp 209-216 (1980)
BIBLIOGRAPHY
AFS Plant and Plant Engineering Committee; "The Cold Facts of
About Cooling Foundry Sand," AFS Transactions, vol 72, pp 872-881
(1964).
Brendt, H. "Sand Processing," Giesserei, vol 67, Issue 21, pp 667-675
(1980).
Graham, A.L., H. W. Dietert, T. H. Hanna, and R. T. Dakswicz;
"Moldability Testing and Control,: AFS Transactions, vol 71, pp 394-
403 (1963)
Heine, H. J. ; "A New Look at Molding Sand, Part 1 and 2, Foundry,
vol 109, no. 10, pp 28-32, pp 102-107 (1981)
Schumacher, J. S; "The Problem of Hot Molding Sand: 1958
Revisited," AFS Transactions, vol 91, pp 879-888 (1983).
Schumacher, J.S., R. A. Green, G. D. Hanson, D. A. Hentz, and H. J.
Galloway; "Why Does Hot Sand Cause Problems: Parts 1, 2 and 3,"
AFS Transactions, vol 82, pp 181-188 (1974); vol 83, pp 441-446
(1975; vol 84, pp 385-388 (1976).
Scott, J.; "Hot Sands: Properties, Problems, and Remedies Quantified,"
AFS Transactions, vol 81, pp 314-319 (1972).
Teumissan, W.; "Hot Sand: a New Approach," Foundry Trade Journal,
vol 132, no 2887, pp 463-467 (Apr 6, 1972).
Zimnawoda, H.W.; "How Continuous Mulling kReconditions Molding
Sand," Foundry, vol 96, no 7, pp 70-75 (July 1960)
APPENDIX A
Sand Cooling and the Laws of Physics
The total amount of the to be removed from foundry return sand
is a function of: 1) the initial temperature of the sand; 2) the
desired temperature of the cooled sand, and weight of sand to be
cooled; 3) the ambient condition of the cooling air and to a
lesser extent, 4) the temperature of the cooling water.
The cooling of foundry sand by evaporative methods follows the
laws of physics. The use of water evaporation is the most
economical method of cooling sand. For every pound of water
evaporated approximately 1000 Btu of heat are removed from
the sand. To calculate the total heat that must be removed to
give the desired cooling, the following formula, ignoring
residual moisture content of the sand, is used:
Q=(T
1
-T
2
) (c) (m)
Where
Q = Heat Expressed in British thermal units (Btu)
T
1
= Initial temperature (F)
T
2
= Final sand temperature (F)
c = Thermal capacity (Btu/lb/F), app. 0.19 for silica sand
m = Sand mass in pounds
For example, to cool 1 ton of sand per minute from 300F to
100F requires the removal from the sand of
(300-100) (0.19) (2000) = 76,000 Btu per minute
or roughly the evaporation of (ignoring the sensible heat effects
of the water and air)
76,000 Btu/1,000 Btu/lb = 76.0 lb of water per minute
the rate of heat flow is primarily limited by the thermal
conductivity of the sand through which the heat must flow and
the temperature differential existing between the sand grain
center and its surface. No one should make the mistake of
believing that just because sand grains are small heat removal
can be accomplished almost instantaneously, even if the method
of heat extraction is extremely efficient.
To reduce the temperature, the heat must be transferred from the
sand grains to another media, which must be disposable without
creating atmospheric pollution and which must be readily
available at low cost. Air and water vapor fulfill these
requirements, provided both are free of solids when discarded.
Both the air and water must come into contact with as many
sand grain surfaces as is practical. The intimately mixed air,
sand, and water must be in contact a sufficient time for the heat
to flow from the center of the sand grain to its surface, where
the heat exchange can take place.
Whenever a liquid (water) is evaporated, heat must be supplied
from some source, which, in this case, is the sand grain heat
content. The higher the temperature, the higher the vapor
pressure of the water and, the higher the water vapor pressure,
the higher the rate of evaporation. If the sand surface
temperature is over 212F (100C) the vapor pressure then
exceeds the atmospheric pressure of the air around it. Below
212F the rate of evaporation is somewhat reduced, but
continued evaporation at a slower rate does proceed.
The water vapor initially formed must be removed from the
sand surface so that further vapor can form. By mechanically
fluidizing the sand mass and mixing it with a continuous supply
for fresh, cool air, the water vapor formed is flushed away from
the sand surface by the fluidized air. The lower the temperature
and relative humidity of the incoming air, the more effectively it
will cool the sand in a given exposure or treatment time.