Foundry Manual


Foundry Manual, 1958, is an update to the 1944 Foundry Manual that was created primarily for use by foundry personnel aboard repair ships and tenders.

In this online version of the manual we have attempted to keep the flavor of the original layout while taking advantage of the Web’s universal accessibility. Different browsers and fonts will cause the text to move, but the text will remain roughly where it is in the original manual. In addition to errors we have attempted to preserve from the original this text was captured by optical character recognition. This process creates errors that are compounded while encoding for the Web.

Note page 169 is missing from the original that we scanned. Please contact us if you have access to another copy of this manual that has this page.

Please report any typos, or particularly annoying layout issues with the Mail Feedback Form for correction.

Richard Pekelney



NAVSHIPS 250-0334
Department of the Navy, Bureau of Ships
Bureau of Ships,
15 April 1958

NAVSHIPS 250-0334


Department of the Navy, Bureau of Ships
Department of the Navy, Bureau of Ships



Superintendent of Documents, U.S. Government Printing Office
Washington, D.C. – Price $3


Bureau of Ships,
15 April 1958

The Foundry Manual of 1944 has been revised to reflect the advancement in foundry technology and to indicate current foundry practice. The revised manual contains information for persons who operate or are employed in a foundry.

J. B. Duval, Jr.
Captain, USN
Assistant to the Assistant Chief of
Bureau for Shipbuilding and Fleet



This Manual is intended primarily for use by foundry personnel aboard repair ships and tenders. The recommended practices are based on procedures proved workable under Navy conditions and are supplemented by information from industrial sources.

The Manual is divided into two general sections. The first section, chapters 1 through 13, contains information of a general nature, such as “How Metals Solidify,” “Designing a Casting,” “Sands for Molds and Cores,” “Gates, Risers, and Chills,” and “Description and Operation of Melting Furnaces.” Subjects covered in these chapters are generally applicable to all of the metals that may be cast aboard ship.

The second section, chapters 14 through 21, contains information on specific types of alloys, such as “Copper-Base Alloys,” “Aluminum-Base Alloys,” “Cast Iron,” and “Steel.” Specific melting practices, suggestions for sand mixes, molding practices, gating, and risering are covered in these chapters.

This manual has been written with the “how-to-do-it” idea as the principal aim. Discussions as to the “why” of certain procedures have been kept to a minimum. This manual contains information that should result in the production of consistently better castings by repair ship personnel.


This page is blank.



Chapter I. How Metals Solidify1
The Start of Solidification1
Freezing Temperature of Metals3
Heat Transfer4
Gases in Metals5
Chapter II. Designing a Casting15
Strength Requirements15
Stress Concentrations15
Section Thickness16
Directional Solidification16
Wall Junctions17
Good Casting Design17
Chapter III. Patternmaking25
Functions of the Pattern25
Types of Patterns25
Pattern Materials25
Making the Pattern28
Finishing and Color Coding28
Maintenance, Care and Repair28
Calculation of Casting Weight29
Chapter IV. Sands for Molds and Cores39
Molding Sands39
Sand Properties40
All-Purpose Sand43
Properties of a 63 AFS Fineness Number Sand43
Molding Sand Mixtures44
Core Sand Mixes48
Methods for Testing Sand49
Chapter V. Making Molds61
Molding Tools and Accessories61
Types of Molds64
Molding Loose-Piece Patterns64
Molding Mounted Patterns66
False-Cope Molding and the Use of Broken Parts as Patterns66
Setting Cores, Chills, and Chaplets67
Closing Molds68


Chapter VI. Making Cores83
Coremaking Tools and Accessories83
Types of Cores83
Internal Support83
Facing, Ramming, Relief, and Venting of Cores83
Turning Out and Spraying84
Cleaning Assembly85
Storage of Cores85
Making a Pump-Housing Core85
Chapter VII. Gates, Risers, and Chills95
General Purpose95
Gating System95
Chapter VIII. Description and Operation of Melting Furnaces121
Oil-Fired Crucible Furnace121
Electric Indirect-Arc Furnace123
Electric Resistor Furnace127
Electric Induction Furnace128
Sintering the Monolithic Lining and Making the First Steel Heat131
Chapter IX. Pouring Castings139
Types of Ladles139
Ladle Linings139
Pouring the Mold140
Speed of Pouring141
Pouring Temperature141
Chapter X. Cleaning Castings147
Removing Gates and Risers147
Grinding and Finishing147
Chapter XI. Causes and Cures for Common Casting Defects149
Names of Defects149
Pattern Equipment150
Flask Equipment and Rigging151
Gating and Risering151
Molding Practice154
Pouring Practice155


Chapter XII. Heat Treatment of Castings179
Iron and Steel Castings179
Nonferrous Castings179
Brass and Bronze Treatment180
Stress-Relief Anne al180
Reasons for Heat Treatment179
Iron and Steel181
Chapter XIII. Composition of Castings185
Selection of Metal Mixtures185
Raw Materials and Calculation of Charges189
Chapter XIV. Copper-Base Alloys213
Selection of Alloy213
How Copper-Base Alloys Solidify213
Molding and Coremaking214
Causes and Cures for Common Casting Defects in Copper-Base Casting221
Welding and Brazing222
Chapter XV. Aluminum-Base Alloys227
Aluminum Silicon227
Aluminum Copper227
Aluminum Zinc227
Aluminum Magnesium227
How Aluminum Solidifies227
Molding and Coremaking228
Causes and Cures for Common Defects In Aluminum Castings231
Welding and Brazing232
Chapter XVI. Nickel-Base Alloys235
Modified S-Monel235
How Nickel-Base Alloys Solidify235
Molding and Coremaking235
Causes and Cures for Common Casting Defects in Nickel-Base Alloy Castings237
Welding and Brazing237


Chapter XVII. Cast Iron239
Selection of Alloys239
How Gray Cast Irons Solidify241
Molding and Coremaking242
Causes and Cures for Common Defects in Iron Castings247
Welding and Brazing248
Chapter XVIII. Steel251
Selection of Alloys251
How Steels Solidify252
Molding and Coremaking252
Welding and Brazing257
Chapter XIX. Copper259
Selection of Metal259
How Copper Solidifies259
Molding and Coremaking259
Causes and Cures for Common Casting Defects in Copper261
Chapter XX. Babbitting With Tin-Base Bearing Metal263
Selection of Alloys263
Preparation of Bearings263
Finishing of Bearing265
Bearing Failures265
Chapter XXI. Process Control269
Inspection and Test270



Chapter I
1Schematic Illustration of the Solidification of Metal in a Mold6
2Volume Change During the Cooling of a 0.35 Percent Carbon Steel7
3Types of Shrinkage7
(a) piping7
(b) gross shrinkage7
(c) centerline7
(d) microshrinkage7
4Cooling Curves of a Pure Metal, a Solid Solution Alloy, and an Eutectic Alloy8
5Melting Points of Metals and Alloys9
6Effect of Section Size on Size of Crystals10
7Schematic Representation of Crystal Growth11
8Dendrite Growth11
9High Magnification of Shrink Area in an Aluminum Casting, Showing Dendrites12
10Crystal Growth in a Gun Metal Casting Dumped Before Solidification was Complete12
11Preferred Orientation in Chill Zone Crystals12
12Dendritic Solidification and Dendritic-Equiaxed Solidification13
13Solubility of Hydrogen in Iron and Nickel at One Atmosphere Pressure13
14Mechanism of Pinhole Formation in Steel13
Chapter II
15Effect of Section Size on Physical Properties19
16Use of Fillets19
17Blending of Thin and Heavy Sections20
18Wheel Design20
19Recommended Wheel Designs21
20Transitions in Section Size21
21Simple Directional Solidification21
22Taper as an Aid to Directional Solidification21
23Hot Spot Location by the Method of Inscribed Circles22
24Reduction of Cross Section in L and V Junctions22
25Reduction of Cross Section in an X Junction22
26Various Treatments for a T Junction22
27Coring to Reduce Section in a Rib Junction22
28Removal of Heavy Section by Redesign22
29Hub Cross Section – Heavy Section23
30Hub Cross Section – Improved Design23
31Bracket Casting23
32Aluminum Yoke Casting23
Chapter III
33One Piece Pattern35
34Split Pattern35
35Core Print Construction36
36Chaplet Location With Pads36
37Mold Broken Due to a Lack of Taper36
38Clean Pattern Draw With Correct Taper36
39Pattern Draft36
40Distortion Allowance in a Simple Yoke Pattern37
41Plaster Patterns and Core Boxes37
42Making a Simple Plaster Pattern37
43Calculating Casting Weight37
44Calculating Casting Weight37


Chapter IV
45Permeability as Affected by the Grain Size of Sand53
46Permeability as Affected by Sand Fineness and Moisture53
47The Effect of Sand Grain Shape on Permeability53
48Permeability as Affected by the Amount of Binder53
49The Effect of Bentonite and Fireclay on Permeability53
50Green Strength as Affected by the Fineness of Sand53
51Green Strengths of Sands With Varying Fineness Numbers54
52Green Strength as Affected by the Shape of Sand Grains54
53Green Strength as Affected by Moisture and Varying Bentonite Contents54
54The Effect of Bentonite and Fireclay on Green Strength of Foundry Sand54
55The Effect of Bentonite on Sands With Various Moisture Contents54
56The Effect of Western and Southern Bentonite on Green Strength and Dry Strength54
57Green Strength as Affected by Mulling Time55
58Relationship Between Moisture Content, Bentonite Content, Green Compressive Strength, and Permeability for an All-Purpose Sand of 63 AFS Fineness Number55
59Relationship Between Moisture Content, Bentonite Content, Green Compressive Strength, and Dry Strength for an All-Purpose Sand of 63 AFS Fineness Number56
60General Green Compressive Strengths for Sands of Different Grain Class Numbers56
61Strength of Baked Cores as Affected by Baking Time and Baking Temperatures56
62Core Gas Generated by Two Different Core Binders57
63The Effect of Single Binders and Combined Binders on the Baked Strength of Cores 57
64Rammer Used for Test Specimen Preparation57
65Permeability Test Equipment58
66Strength Testing Equipment58
67Equipment for Drying Sand Specimens for Moisture Determination58
68Jar and Stirrer for Washing Sand58
69Sand Washing Equipment Assembled59
70The Difference in Sand Grain Distribution for Two Foundry Sands Having the Same Grain-Fineness Number59
Chapter V
71Molder’s Hand Tools69
72Additional Molder’s Tools70
73Double-Headed Chaplets71
74Stem Chaplets71
75Perforated Chaplets71
76Recommended Chaplet Design for Emergency Use71
77Anchoring Cores With Chaplets71
78Pattern Set in Drag With Gating System Parts71
79Hand Packing Riddled Sand Around the Pattern72
80Ramming a Deep Pocket72
81Striking Off the Drag73
82Drag Ready for the Cope73
83Cope With Pattern and Gating Pieces Set74
84Ramming the Partially Filled Cope74
85Venting the Cope75
86Start of the Pattern Draw75
87Pattern Completely Drawn76
88Setting the Core76
89Cope and Drag Ready for Closing77
90Clamped Mold With Weights and Pouring Basin78
91Pouring the Mold78
92Finished Pump Housing Casting79
93Propeller Set in the Drag79
94Propeller in the Drag With Parting Line Cut80
95Drawn Cope80
96Mold Ready for Closing81
97As-Cast Propeller82


Chapter VI
98Arbor for a Medium-Size Core87
99View of Inside of Core Showing Hollowing to Make the Core More Collapsible When Metal is Poured Around It87
100Section of Mold Showing Use of Lifting Eye for Supporting Heavy Core87
101Typical Lifting Hooks for Lifting Cores87
102Core Boxes for Pump Housing Core88
103Ramming Up the Core88
104Striking Off the Core89
105Placing the Reinforcing Rods89
106Cutting Vents90
107Drag Core Turned Out91
108Cope Core Turned Out92
109Applying Core Paste93
110Assembling the Two Core Halves93
Chapter VII
111Parts of a Simple Gating System106
112Illustration of Gating Ratio106
113Gating Nomenclature107
114Unfavorable Temperature Gradients in Bottom Gated Casting108
115Defect Due to Bottom Gating108
116Bottom Gate108
117Reverse Horn Gate108
118Reverse Horn Gate108
119Bottom Gating Through Side Risers109
120Bottom Gating Through Riser with Horn Gate109
121Sprue With Well at Base109
122Simple Top Gating109
123Pencil Gate109
124Typical Parting Gate109
125Parting Gate Through the Riser110
126Whirl Gate110
127Simple Step Gate. (Not Recommended.)111
128Thirty-Degree Mold Manipulation111
129Complete Mold Reversal111
130Pouring Cups111
131Pouring Basin111
132Solidification Time vs. A/V Ratio111
133Effectiveness of Square and Round Risers112
134Proper and Improper Riser Height112
135Poor Riser Size and Shape112
136Proper Riser Size and Shape112
137Riser Location at Heavy Sections112
138Cold Metal Riser (Not Recommended )112
139Hot Metal Riser113
140Feeding Through a Thin Section113
141Flanged Casting with Open Riser113
142Flanged Casting with Blind Riser113
143Inscribed Circle Method for Riser Contact113
144Effect of Keeping Top Risers Open113
145Effect of Keeping Blind Risers Open114
146Casting Defects Attributable to Shrinkage Voids and Atmospheric Pressure114
147Blind Riser Principle114
148Individual Zone Feeding for Multiple Risers114
149Padding to Avoid the Use of Chills or Risers115
150Padding to Prevent Centerline Shrinkage115


151Typical Padding of Sections115
152Shrinkage on the Thermal Centerlines of Unpadded Sections115
153Use of a Core to Make a Padded Section115
154Effect of Insulated Risers115
155Reduction in Riser Size Due to Insulation115
156Comparison of Ordinary Riser, Insulated Riser, and Exothermic Riser116
157Typical Internal Chills116
158Typical External Chills with Wires Welded-On or Cast-In to Hold Chill in Place116
159Use of External Chills in a Mold for an Aluminum Casting117
160Use of External Chills on a Bronze Casting117
161As-Cast Aluminum Casting Showing Location of External Chills118
162Gear Blank Mold Showing Location of External Chills118
163Principle of Tapering Edges of External Chill118
164Effect of Chill Mass and Area of Contact119
165Typical Application of External Chills to Unfed L,T,V,X, and W Junctions119
166Preferred Method of Applying External Chills by Staggering119
Chapter VIII
167Pit-Type Crucible Furnace133
168Crucible for Tilting Crucible Furnace133
169Tilting Crucible Furnace133
170Cross Section of a Stationary Crucible Furnace133
171Undercutting a Refractory Patch133
172Proper Burner Location134
173Proper Fit for Crucible Tongs134
174Electric Indirect-Arc Furnace134
175General Assembly View of Electric Indirect-Arc Furnace134
176Accessory Equipment for Electric Indirect-Arc Furnace135
177Properly Charged Electric Indirect-Arc Furnace135
178Electric Resistor Furnace135
179Electrode-Bracket Assembly for Electric Resistor Furnace136
180Cross Section of Electric Induction Furnace136
181Flow Lines in an Induction Furnace Melt136
182Essential Parts of an Induction Furnace137
183Typical Electric Induction Furnace137
184Induction-Furnace Control Panel137
185Method of Lining Induction Furnace Using a Steel Form137
Chapter IX
186Lip-Pouring Ladle143
187Teapot Ladle143
188Lip-Pouring Crucibles143
189Teapot Crucibles143
190Lining a Teapot Pouring Ladle143
191Proper Pouring Technique144
192Poor Pouring Technique144
193Use of Pouring Basin and Plug144
194Skim Core in Down Gate144
195Skim Core in Pouring Basin144
196Pyrometer Field When at Correct Temperature, Too High a Setting, and Too Low a Setting144
197Effect of Pouring Temperature on Grain Size145
Chapter XI
199Gross Shrink173
200Surface Shrink173


201Surface Shrink173
202Internal Shrink174
203Gating and Risering that Corrected Internal Shrink in Figure 202174
204Gross Shrink174
205Dross Inclusions174
207Expansion Scab174
208Erosion Scab and Inclusions175
209Metal Penetration and Veining175
210Hot Tear175
211Pin Holes175
215Cracked Casting176
217Blow and Expansion Scab177
Chapter XIII
222Example of Charge Calculation for Ounce Metal210
223Example of Charge Calculation for Gray Iron211
Chapter XIV
224Horizontal Molding of a Bushing223
225Vertical Molding of a Bushing223
226Gating a Manganese Bronze Casting223
227Gating a Number of Small Castings in Manganese Bronze or Red Brass223
228Gating for a Thin Nickel-Silver Casting223
229Poor Gating System for a Cupro-Nickel Check Valve223
230Improved Gating That Produced a Pressure-Tight Casting224
231Globe Valve – Poor Risering Practice224
232Globe Valve – Improved Risering Practice224
233High Pressure Elbow – Poor Risering Practice224
234High Pressure Elbow – Improved Risering Practice225
235Risers for a Cupro-Nickel Valve Body225
236Tapered Chills on a Flat G Metal Casting225
237Tapered Chills on a G Metal Bushing225
238Examples of Gassy and Gas-Free Metal225
Chapter XV
239Enlargement-Type Sprue Base233
240Well-Type Sprue Base233
241Coarse-Grained Structure (Caused by Iron Contamination )233
242Porosity. (Caused by Excessive Moisture in the Sand )233
Chapter XVI
243Poor Gating and Risering Practice for a Nickel-Base Alloy Casting238
244Improved Gating and Risering for Nickel-Base Alloy Casting238


Chapter XVII
245Knife Gate249
246Lap Gate249
247Riser for a Gray Iron Casting Molded in the Cope and Drag249
248Riser for a Gray Iron Casting Molded in the Drag249
249Riser for a Gray Iron Casting Molded in the Cope250
250Plan View of Runner, Riser, and Ingate250
251Operating Log for Cast Iron Heats250
Chapter XVIII
252Iron-Carbon Diagram258
253Tapered Chill258
254Steel Rods Used for Determining the Pouring Temperature of Steel258
Chapter XIX
255The Effect of Various Elements on the Electrical Conductivity of Copper262
256Properly Deoxidized Copper Sample262
257Partially Deoxidized Copper Sample262
258Gassy Copper Sample262
Chapter XX
259Jig for Babbitting Bearings267


Chapter I

Making a casting involves three basic steps: (1) heating metal until it melts, (2) pouring the liquid metal into a mold cavity, and (3) allowing the metal to cool and solidify in the shape of the mold cavity. Much of the art and science of making castings is concerned with control of the things that happen to metal as it solidifies. An understanding of how metals solidify, therefore, is necessary to the work of the foundry-man. The control of the solidification of metal to produce better castings is described in later chapters on casting design, gating, risering, and pouring.The change from hot molten metal to cool solid casting takes place in three main steps. The first step is the cooling of the metal from the pouring temperature to the solidification temperature. The difference between the pouring temperature and the solidification temperature is called the amount of superheat. The amount of superheat determines the amount of time the foundryman has available to work with the molten metal before it starts to solidify.

The second step is the cooling of the metal through the range of temperature at which it solidifies. During this step, the quality of the final casting is established. Shrink holes, blow holes, hot cracks, and many other defects form in a casting while it solidifies.

The third step is the cooling of the solid metal to room temperature. It is during this stage of cooling that warpage and casting stresses occur.


Solidification of a casting is brought about by the cooling effect of the mold. Within a few seconds after pouring, a thin layer of metal next to the mold wall is cool enough for solidification to begin. At this time, a thin skin or shell of solid metal forms. The shell gradually thickens as more and more metal is cooled, until all the metal has solidified. Solidification always starts at the surface and finishes in the center of a section. In other words, solidification follows the direction that the metal is cooled.

The way in which metal solidifies from mold walls is illustrated by the series of steel castings shown in figure 1. The metal that was still molten after various intervals of time was dumped out to show the progress of solidification. All metals behave in a similar manner. However, the time required to reach a given thickness of skin varies among the different metals.

The speed of solidification depends on how fast the necessary heat can be removed by the mold. The rate of heat removal depends on the relation between the volume and the surface area of the metal. Other things being equal, the thin sections will solidify before the thick ones. Outside corners of a casting solidify faster than other sections because more mold surface is available to conduct heat away from the casting. Inside corners are the slowest sections of the casting to solidify. The sand, in this case, is exposed to metal on two sides and becomes heated to high temperatures. Therefore, it cannot carry heat away so fast.Changes in design to control solidification rate sometimes can be made by the designer. If, however, a change in solidification rate is required for the production of a good casting, the foundryman is usually limited to methods that result in little or no change in the shape of the casting. The rate of solidification can be influenced in three other ways: (1) by changing the rate of heat removal from some parts of the mold with chills; (2) by proper gating and risering, mold manipulation, and control of pouring speed, and (3) by padding the section with extra metal that can be machined off later.


Metals, like most other materials, expand when they are heated. When cooled, they must contract or shrink. During the cooling of molten metal from its pouring temperature to room temperature, contraction occurs in three definite steps corresponding to the three steps of cooling. The first step, known as liquid contraction, takes place while the molten metal is cooling from its pouring temperature to its freezing temperature. The second, called solidification contraction, takes place when the metal solidifies. The third contraction takes place when the solidified casting cools from its freezing temperature to room temperature. This is called solid contraction. Of the three steps in contraction, the first liquid contraction causes least trouble to the foundryman because it is so small in amount.

Figure 2, which shows the change in volume of a steel alloy as it cools from the pouring temperature to room temperature, illustrates these contractions. In a similar way, most of the metals considered in this manual contract in volume when cooling and when solidifying. The amount of shrinkage in several metals and alloys is given in table 1. Notice that some compositions of gray cast iron expand slightly



NameCompositionDecrease in
Volume During
Total Decrease
in Volume,
Red brass85 Cu, 5 Zn, 5 Pb, 5 Sn6.310.6
Yellow brass70 Cu, 27 Zn, 2 Pb, 1 Sn6.412.4
Bearing bronze80 Cu, 10 Sn, 10 Pb7.311.2
Manganese bronze56-3/4 Cu, 40 Zn, 1-1/4 Fe, 1/2 Sn, 1 Al, 1/2 Mn4.611.5
Aluminum bronze90 Cu, 10 Al4.111.2
Nickel98 Ni, 1-1/2 Si, 0.1 C6.114.2
Monel67 Ni, 32 Cu6.313.9
Nickel silver20 Ni, 15 Zn, 65 Cu5.512.1
Carbon steel0.25 C, 0.2 Si, 0.6 Mn3.811.4
Nickel cast iron13 Ni, 7 Cu, 2 Cr, 3 C1.67.8
Gray cast iron2.18 C, 1.24 Si, 0.35 Mn
3.08 C, 1.68 Si, 0.44 Mn
3.69 C, 2.87 Si, 0.59 Mn
-1.65 (expands)


during solidification. This results from the formation of graphite, which is less dense than iron. The formation of graphite compensates for a part of the shrinkage of the iron.Reservoirs of molten metal, known as risers, are required to make up for the contraction that occurs during solidification. If risers are not provided at selected spots on the casting, shrinkage voids will occur in the casting. These voids can occur in different ways, depending on the shape of the casting and on the type of the metal. Piping, the type of shrinkage illustrated in figure 3a, occurs in pure metals and in alloys having narrow ranges of solidification temperature. Piping in a riser is usually a good indication that it is functioning properly. Gross shrinkage, illustrated in figure 3b, occurs at a heavy section of a casting which has been improperly fed. Centerline shrinkage, illustrated in figure 3c, occurs in the center of a section where the gradually thickening walls of solidified metal from two surfaces meet.Centerline shrinkage occurs most frequently in alloys having a short solidification range and low thermal conductivity. Microshrinkage, which is also known as microporosity, occurs as tiny voids scattered through an area of metal. It is caused by inability to feed metal into the spaces between the arms of the individual crystals or grains of metal. This type of shrinkage, which is illustrated in figure 3d, is most often found in metals having a long solidification temperature range. Microporosity may also be caused by gas being trapped between the arms of the crystals.After solidification, cast metal becomes more rigid as it cools to normal room temperature. This cooling is accompanied by contraction, which is allowed for by the patternmaker in making the pattern for the casting. Contraction in cast metals after solidification is resisted by the mold. Often, different cooling rates of thin and heavy sections result in uneven contraction. This uneven contraction can


severely stress the partially solidified, and still weak, heavier sections. Resistance to contraction of the casting results in severe “contraction stresses” which may tear the casting or which may remain in the casting until removed by suitable heat treatment. Sharp internal corners are natural points for these stresses.Some metals, such as steel, undergo other dimensional changes as they pass through certain temperature ranges in the solid state. In the case of castings with extreme variations in section thickness, it is possible for contraction to take place in some parts at the same time that expansion occurs in others. If the design of the junctions of these parts is not carefully considered, serious difficulties will occur in the foundry and in service.


Molten metal has the ability to dissolve many substances, just as water dissolves salt. The most important elements that are soluble in molten iron are other metals and five nonmetals–sulfur, phosphorus, carbon, nitrogen, and hydrogen. When substances are dissolved in a metal, they change many of its properties. For example, pure iron is relatively soft. A small amount of carbon dissolved in the iron makes it tough and hard. Iron containing a small amount of carbon is called steel. More carbon dissolved in the iron makes further changes in its properties. When enough carbon is dissolved in the molten iron, the excess carbon will form flakes of graphite during solidification. This metal is known as cast iron. The graphite flakes lower the effective cross section of the metal, lower the apparent hardness, and have a notch effect. These factors cause cast irons to have lower strengths and lower toughness than steels.

One of the most important changes in a metal as it dissolves other substances is a change in the freezing characteristics.

Pure metals and certain specific mixtures of metals, called eutectic mixtures, solidify without a change in temperature. It is necessary, however, to extract heat for solidification to occur. The solidification of pure metals and eutectic mixtures is very similar to the freezing of water. Water does not begin to freeze until the temperature is lowered to 32°F. The temperature of the ice and water does not change from 32°F. until all of the water is converted to ice. After this, the ice can be cooled to the temperature of its surroundings, whether they are zero or many degrees below zero. This type of temperature change during cooling, shown in figure 4a, is typical of pure metals, eutectic mixtures, and water. Actual

solidification temperatures are different for each material.Most of the metals used by foundrymen are impure and are not eutectic mixtures. These metals solidify over a range of temperature known as the solidification range. Mixtures of metals have many of the solidification characteristics of mixtures of salt and water. Just as the addition of salt to water changes the temperature at which water starts to freeze, so does the addition of one metal to another change the freezing point of the second metal. An example of such a mixture of metals is the copper-nickel system shown in figure 4b (right). A given mixture of copper and nickel will be liquid until it reaches the temperature that crosses the line marking the upper boundary of Area A + L. In the Area A + L, the mixture will be partly liquid, and in the Area A, it will be entirely solid. It will be noted that the addition of copper to nickel lowers the freezing temperature. On the other hand, the addition of nickel to copper raises the freezing temperature. A metal system which has the same general shape as the copper-nickel system is said to have complete solid solubility. Like the mixture of water and salt, metal mixtures of this type must be cooled well below the temperature at which freezing begins before they are completely solidified. In its simplest form, the cooling curve looks like that in figure 4b (left). The range of temperature between the upper and lower line is the solidification range.

Most of the metal mixtures used in the foundry do not have cooling curves as simple as those shown in figures 4a and 4b. As an example, the addition of tin to lead lowers the freezing temperature of the mixture (see figure 4c, right). The addition of lead to tin also lowers the freezing temperature of the mixture. However, there is one specific mixture which has a lower freezing temperature than either lead, tin, or any other mixture of the two. The mixture that has the lowest freezing temperature is the eutectic mixture. A typical set of alloys that has an eutectic mixture is that of the lead-tin system shown in figure 4c (right). A cooling curve for one lead-tin alloy is also shown in figure 4c (left). In such mixtures, the mechanism of solidification is quite complicated.

The melting temperatures of important metals are shown in figure 5. The melting temperatures of many metals are so high that they create real problems in selecting materials for handling the molten metal and for making the mold.


A casting is made up of many closely packed and joined grains or crystals of metal. Within

452605 0-58-2


any particular crystal, the atoms are arranged in regular orderly layers, like building blocks. On the other hand, there is no orderly arrangement of atoms in molten metal. Solidification, therefore, is the formation and growth of crystals, layer by layer, from the melt. The size of the crystals is controlled by the time required for the metal to solidify and by its cooling rate in the mold. Obviously, the heavy sections take more time to freeze than the light sections. As a result, the crystalline structure of a heavy section is usually coarser than that of the lighter members. This may be seen in figure 6.Although the physical properties of coarse-grained metals differ from those of fine-grained metals of the same chemical composition, this difference will not be considered in detail. As one example, coarse grains lower the strength of steel.

Metal crystals start to grow at the surface of the casting because this is where the molten metal first cools to its freezing temperature. Once a crystal starts to form, it grows progressively larger until its growth is stopped by other crystals around it or until there is no more molten metal to feed it. The growth of metal crystals is similar to the growth of frost crystals on a pane of glass.

A schematic drawing of the start and growth of metal crystals is shown in figure 7. The black square represents the original crystal center or nucleus which grows into a crystal or grain by the addition of layers of atoms from the melt. A three-dimensional sketch of crystal growth is shown in figure 8. Part (a) shows the crystal shortly after it has formed and has started to grow. In part (b), the crystal has become elongated and growth has started in two other directions. Still further growth is shown by part (c). The original body of the crystal has grown still longer and has become thicker in cross section. Two other sets of arms have started growing near the ends of the longest arms of the crystal. A still further stage of growth is shown in part (d). Crystals grow in this manner with continued branching and thickening of the arms. Because of its branching nature, the type of crystal shown in figure 8 is called a dendrite. When the metal is completely solidified, the arms will have grown and thickened until they have formed a continuous solid mass. A photograph of dendrites in a shrink area of an aluminum casting is shown in figure 9. The branching of the dendrite arms at right angles can be seen in this photograph. Close examination will also show where the growth of crystals was stopped by the growth of neighboring dendrites.

The first metal that solidifies at the mold surface will be composed of grains that are not

arranged in any particular pattern and that grow about the same length in each direction. Such grains are called randomly oriented, equiaxed grains. The crystals of zinc on the surface of galvanized steel are a familiar example. Another example of crystal structure is shown in figure 10. The faces of the individual crystals can be seen easily and growth would have continued if it had not been dumped to reveal the crystals.For a while after solidification begins at the surface of the casting, there will be a solid skin against the mold and the metal in the center will still be liquid. The growth of the metal crystals in the skin will take place by the building up of metal on some of the crystals of the surface layer which are favorably positioned for further growth. Figure 11 shows the small grains at the mold surface, with some of them positioned for further growth. The position for favorable growth is perpendicular to the mold wall and parallel to the direction of heat transfer from the casting. Properly oriented crystals will grow in toward the center because side growth will stop as soon as adjacent crystals meet. This type of crystal growth toward the center of the casting is known as columnar grain growth. Depending on the pouring temperature and the type of metal, growth of elongated grains may extend to the center of the casting. If the characteristics of the metal are such that it is impossible to feed properly the last parts of the dendrites, the casting defect known as centerline shrinkage is formed. This is shown in figure 12a. A point may be reached during solidification when the solidification temperature is reached by the entire remaining liquid metal. Nucleation and growth of crystals will then start throughout the melt and result in an equiaxed crystal structure in that part of the casting. Solidification which started as dendritic growth and finished as an equiaxed structure is shown in figure 12b.


The solidification of molten metal in the mold is a result of the extraction of heat from the metal by the sand that surrounds it. This process of heat extraction is called heat transfer.

The transfer of heat from the molten metal to the sand and its transfer away from the casting is most rapid at the time the mold cavity is first filled. As the casting cools and solidifies, the transfer of heat is carried on at a reduced rate. The rapid heat transfer in the early period of solidification is due to the ability of the sand to store a large amount of heat. As the maximum capacity of the sand to store heat is reached, the sand becomes saturated with heat,


and further transfer of heat from the casting to the mold is controlled by the ability of the sand to conduct the heat away. Because this is a much slower process than the absorption of heat by the sand, the transfer of heat away from the casting takes place at a lower rate. Many times, the rate of transfer is further slowed by an air gap which is formed when the solidified casting starts to contract and draw away from the mold. The presence of this air gap causes a further decrease in the rate of heat transfer. Chills produce an increased rate of solidification because of their increased heat-storage capacity, as compared to an equal volume of sand, and their ability to conduct heat at a rate much more rapid than that at which sand can conduct it.


Many defects in castings are caused by gases which dissolve in the metal and then are given off during solidification. These defects may range in size and form from microscopic porosity to large blow holes. Because of the large volume that a small weight of gas occupies, very little gas by weight can cause the foundryman a lot of trouble. As an example, at room temperature and atmospheric pressure, 0.001 percent by weight of hydrogen in a metal occupies a volume equal to that of the metal, and at 2,000°F., the same amount of hydrogen would occupy a volume equal to four times that of metal.

Gases may be absorbed by the metal during smelting, refining, melting, and casting. Here, we are primarily concerned with the gas absorption during melting. The gases in any melting process often come from water vapor in the air, or from water which is introduced into the melt by careless foundry practice.

A gas frequently absorbed by metals is the hydrogen produced from water vapor. The solubility of hydrogen in nickel and steel at various temperatures is shown in figure 13. Notice that it is possible to dissolve more hydrogen in molten metal than in solid metal. Therefore, gas that is absorbed during melting may escape when the molten metal cools and solidifies. If the gas cannot escape from the metal freely, bubbles are trapped in the casting causing defects. The treatment of metals to reduce their gas content before they are poured into the mold is discussed in later chapters dealing with the specific metals.

Gas defects in castings are not always caused by gas that is dissolved in the molten metal. In some cases, these defects are caused by gases driven into the metal from the mold.

The gases are trapped as the metal solidifies. In some cases, gas is generated by chemical reactions within the metal, such as may sometimes occur between carbon and oxygen in steel to form carbon monoxide.A good example of the formation of a casting defect due to gas in pinhole formation in steel. This takes place as shown in figure 14. When the molten steel comes in contact with moist sand in the mold, a thin skin of steel is formed almost immediately. At the same time, the water in the sand is changed to steam with an increase in volume of approximately 5,000 times. The steam is highly oxidizing to the steel and reacts with it. As a result, iron oxide and hydrogen are formed. The iron oxide produces the scale which is seen on steel castings when they are shaken out of the mold.

The hydrogen which is formed in this reaction passes through the thin layer of solid steel and enters the still molten steel. The hydrogen in the molten steel can then react with iron oxide, which is also dissolved in the steel. This reaction produces water vapor. As the steel cools, it must reject some of this water vapor and hydrogen, just as an ice cube must reject gas as it freezes. A bubble is formed and gradually grows as more steel solidifies. The bubbles become trapped between the rapidly growing crystals of steel and cause the familiar pinhole defect.


An understanding of the solidification or freezing of metals is important to the foundry-man who wants to know how to make good castings.

Solidification of a casting starts by the formation of solid grains next to the surface of the mold. These grains grow inwardly from the surface until they meet other grains growing from other surfaces. When these growing surfaces meet, the casting is solid.

Improper foundry practice will cause many defects which can be explained and avoided if proper attention is given to the way in which the metal solidifies. Casting defects which can occur if the freezing characteristics of metals are not taken into account are as follows: (1) microshrinkage, (2) centerline shrinkage, (3) shrink holes, (4) certain types of gas holes, (5) piping, and (6) hot tears. These defects can be minimized if proper attention is given to the practices described in later chapters, particularly Chapter 2, “Design,” and Chapter 7, “Gates, Risers, and Chills.”


Figure 1. Schematic illustration of the solidification of metal in a mold.
Figure 1. Schematic illustration of the solidification of metal in a mold.


Figure 2. Volume change during the cooling of a 0.35 percent carbon steel.
Figure 2. Volume change during the cooling of a 0.35 percent carbon steel.

Figure 3. Types of shrinkage
Figure 3. Types of shrinkage

(a) piping(b) gross shrinkage
(c) centerline(d) microshrinkage



Figure 4. Cooling curves of a pure metal, a solid solution alloy, and an eutectic alloy.
Figure 4. Cooling curves of a pure metal, a solid solution alloy, and an eutectic alloy.


Figure 5. Melting points of metals and alloys.
Figure 5. Melting points of metals and alloys.


Figure 6. Effect of section size on size of crystals.
Figure 6. Effect of section size on size of crystals.


Figure 7. Schematic representation of crystal growth.
Figure 7. Schematic representation of crystal growth.

Figure 8. Dendrite growth.
Figure 8. Dendrite growth.


Figure 9. High magnification of shrink area in an aluminum casting showing dendrites.
Figure 9. High magnification of shrink area in an aluminum casting showing dendrites.

Figure 10. Crystal growth in gun metal casting dumped before solidification was complete.
Figure 10. Crystal growth in gun metal casting dumped before solidification was complete.

Figure 11. Preferred orientation in chill zone crystals.
Figure 11. Preferred orientation in chill zone crystals.


Figure 12. Dendritic solidification and dendritic-equiaxed solidification.
Figure 12. Dendritic solidification and dendritic-equiaxed solidification.

Figure 13. Solubility of hydrogen in iron and nickel at one atmosphere pressure.
Figure 13. Solubility of hydrogen in iron and nickel at one atmosphere pressure.

Figure 14. Mechanism of pinhole formation in steel.
Figure 14. Mechanism of pinhole formation in steel.


This page is blank.


Chapter II

The design of a casting might seem to be something far removed from the field of interest of a Navy molder. He is usually called upon to make a casting from a loose pattern or from the broken parts of an existing casting. Very rarely is he consulted as to what is good casting design from the foundryman’s point of view. Nevertheless, an understanding of what constitutes good casting design will help the molder to make a consistently better product.Design influences the soundness, freedom from dirt, shrinkage, porosity, hot tears, and cracks found in a casting, and thus affects its serviceability. A capable foundryman may produce satisfactory castings that violate some of the principles of good design, but he will never produce them with any degree of consistency. Superior craftsmanship of the foundry-man should not be relied upon to overcome poor design.

Good casting design is based on two general considerations. The first thing to consider is the intended use of the casting, and the second is which alloy should be used. The intended use of the casting (that is, whether it is a supporting structure, moving part, pressure casting, or bearing) will be the major factor in determining the general shape of the casting. The amount of corrosion resistance, wear resistance, machinability, and strength that are needed will determine which alloy should be used. More often than not, a casting must meet a combination of requirements.

Many times, the same features of design which give trouble to the foundryman will also adversely affect the service life of the part. Therefore, the first step in the production of a casting should be a careful study of its design in the light of the information given in this chapter. This applies equally to a new design and to the replacement of a casting of an old design. In the replacement of a casting, the defective part should be thoroughly studied to determine if failure was in any way due to design faults; whether faulty design contributed to casting unsoundness, or whether it adversely affected the service strength of the solid part.


The amount of strength that is needed for a casting will be determined primarily by the part it plays in the structure or machine in which it is used. A casting should be designed so that the strength requirements are met with

the proper safety factor. Care should be taken not to overdesign a casting. Many times when a casting fails, certain regions in the vicinity of the failure will be made larger with the idea that additional strength will be gained with an increase in thickness. In reality, this overdesign frequently produces casting defects which offset the desired increase in strength.Sections that are heavier than necessary do not make use of all the strength that is available in the metal. As a general rule, a metal has lower strength per square inch of cross section when cast in thick sections than it does in thin sections. The effect of increasing section size on the strength and elongation of four different copper-base alloys is shown in figure 15. It is evident that the tin bronze and red brass are very sensitive to section thickness, while aluminum bronze and manganese bronze are less affected by section size. From this, it can be seen that the effect of section size on the properties of a casting must be considered if the casting is to make the best use of the metal poured into it.


One of the major factors that cause the untimely failure of castings is the concentration of stresses that results from improper design. Stresses, of course, are the forces and loads that cause a casting to crack, tear, or break.

Sharp corners and notches should be avoided in castings because they are points of high stress. The liberal use of fillets and rounded corners of proper size is the easiest way to reduce the concentration of stresses in corners. A sharp corner will also produce a plane of weakness in a casting where crystal growth from two sides meet. This is shown in figure 16a. The combination of high stresses and the plane of weakness result in early failure of the casting. The partial removal of this plane of weakness by rounding the corners is shown in figure 16b, and its complete elimination, in figure 16c.

The junction of thin and heavy sections is another point of stress concentration. The stresses in this case result from the rapid solidification and contraction of the thin section. This contraction will set up very high stresses at the junction with the hotter, weaker, heavy section and may produce hot tearing. Where sections of different thicknesses are necessary, they should be blended together to reduce the


stresses as much as possible. Recommended practices for the blending of junctions are shown in figure 17. Although shown for aluminum, the same practices should be followed for all metals.There are some castings in which the design must allow for the absorption of casting stresses in order to produce a good casting. A spoked wheel is an example. Correct and incorrect designs for wheels are shown in figure 18. The original design (with straight spokes) caused hot cracks at the junction of the spokes with the rim and hub. The modified design (with a curved spoke) produced a casting without hot tears. The modified design permits the spokes to stretch and distort slightly without tearing under the stresses set up by contraction. Two other patterns made to prevent tearing in a wheel casting are shown in figure 19.

Contraction stresses often cause warping of the casting. When distortion cannot be solved directly by design, as with the wheel casting, it must be allowed for by the patternmaker after consultation with the molder. Correction of this type of distortion is covered in Chapter 3, “Patternmaking.”


The minimum thickness that can be cast is determined by the ability of the metal to flow and fill thin sections without the use of an excessive pouring temperature. The normal minimum sections that can be cast from several metals are listed in table 2.


MaterialNormal Minimum Section Thickness, in.
Gray cast iron1/8
White cast iron1/8
Brass and bronze3/32

These minimum dimensions for thin sections may vary slightly with composition of the alloy, pouring temperature, and size or design of the casting. The use of adequate but not excessive section thickness in a casting cannot be stressed too strongly, because it is a major factor in good design.

An effort should be made at all times to increase gradually the section size toward the reservoir of liquid metal in the riser. This

will promote directional solidification described in the next paragraph. A sudden change in section thickness should be avoided wherever possible. Where a change in section thickness must be made, it should be gradual. A blending, or gradual change in section thickness reduces stresses at the junctions. Figure 20 shows various methods for changing from one section thickness to another.


Directional solidification means that solidification will start in one part of the mold and gradually move in a desired direction; it means that solidification will not start in some area where molten metal is needed to feed the casting. An effort is always made by the foundry-man to get solidification to progress toward the riser from the point furthermost from the riser. Casting design is a determining factor in the control of the direction of solidification, and every effort should be made to apply the principles of good design to reach this objective.

A slab casting of uniform dimension, shown in figure 21, demonstrates directional solidification. The metal is poured through the riser, and as it flows over the mold surface, it gives up some of its heat to the mold. Such a condition will mean that when the mold is filled, the metal at the right end will not be as hot as the metal near the riser. The first metal to solidify will then be the metal at the right, as shown in figure 21a. The mold to the left of the casting will also have been heated by the molten metal flowing over it and its ability to conduct heat away from the casting will be reduced so that the cooling of the casting in that area will be retarded. Figure 21b shows the casting with solidification in a more advanced stage. Because of controlled solidification, this will probably be a sound casting. However, the reduction of area at the corner is undesirable from the structural design standpoint.

In actual practice, conditions are usually such that directional solidification cannot be obtained as simply as described above because of the properties of the metal or the design of the casting. In such cases, the desired directional solidification of the casting must be obtained by other methods. In designing a casting to control directional solidification, tapering sections can be used. The sections are tapered with the larger dimensions toward the direction of feeding. When a flat casting is poured, solidification will begin at about the same rate from both sides and centerline shrinkage will be found because of the lack of directional solidification. Solidification of this type is known as progressive solidification and is shown in figure 22a. If this casting did not have


to have parallel sides, then a taper could be used to good advantage. Figure 22b shows the taper employed to obtain directional solidification. It will be noted that although solidification has taken place at the same rate from the opposing walls, the taper permits molten metal to feed the casting properly.If it is impossible to design a casting to make full use of directional solidification, then other aids must be used. The most effective and most easily used is the chill. Chills are used to start or speed up solidification in a desired section of a casting. Their application and use are covered in Chapter 7, “Gates, Risers, and Chills.” Another method of obtaining directional solidification in a casting is to taper the section intentionally and then remove the excess metal by machining. This method, called “padding,” is also described in chapter 7.


Junctions such as “L” and “T” sections must be given special consideration when designing a casting. Because a junction is normally heavier than any of the sections which it joins, it usually cools more slowly than adjacent sections. The method of inscribed circles, illustrated in figure 23, can be used to predict the location of hot spots, which are locations of final solidification and possible shrinkage. In the L section, the largest circle which can be drawn in the junction is larger than the largest circles that can be drawn in the walls. The same is true of the T section, where the circle at the junction is even larger than the one for the L section. The larger circles in both of the junctions predict the location of a hot spot, which will be unsound unless special precautions are taken. Figure 23b shows similar junctions with the progress of solidification indicated by the shaded areas. These sketches were made from actual laboratory studies of the solidification of the junctions. The location of the small white area in each case indicates the location of the hot spot. These small spots are within the large circles inscribed at the junctions as shown in figure 23a.

The joining of two walls may result in an L, V, X, or T-shaped junction. If small fillets and rounded corners are used in the L or V-type junction, a heavy section will be formed. Radii should be used so that the thickness in the junction will be the same as that in the adjoining walls. This is shown in figure 24. The area within the dashed line shows the amount of metal which should be eliminated to avoid hot spots. The wall thickness at the junction can be reduced further by using radii which will produce a junction thinner than the adjoining sections. Such junctions would be used only if they were

the part of the casting from which solidification was to start. The first method is most commonly used. Chills may also be used to produce a sound junction. They are described in Chapter 7, “Gates, Risers, and Chills.”An X section has a still greater tendency toward hot spots and unsoundness than do the L or V sections. The only way to reduce the wall section in this type of junction is to use a core, as shown in figure 25b, to produce a hole in the junction. A method which is preferred, especially when the junction is a result of ribbed construction, is that of staggering the sections so as to produce T junctions which can be more easily controlled with chills. Figure 25c shows the staggered design. Various treatments for a T section are shown in figure 26. A cored hole can be used, as in figure 26a; the section thickness can be used, as in figure 26b; the external chills can be used, as in figure 26c; or internal chills can be used, as in figure 26d. Internal chills should not be used without authorization from the foundry supervisor.

Many times, a large casting will require ribs to provide added strength at certain locations. The use of ribs produces a hot spot at the junction because it is thicker. The heavy section may also be reduced by using a core to make a hole at the junction of the rib with the casting section, as shown in figure 27.


Casting designs often cannot be ideal because the casting must be designed to do a certain job. Everything should be done, however, to give the casting a section having a gradual taper, so that the best possible conditions for solidification can be obtained. A detailed discussion of a good casting design cannot be given here, but a few examples are given of design features which can be of help to the molder and patternmaker in making a better casting.

A casting having a tubular section joining a flat base is shown in figure 28. As originally designed, the tubular section had a heavier wall than the plate. Redesigning eliminated the heavy section in the casting. A hub casting is shown in figure 29. The inscribed circle shows the heavy section which would be difficult to feed and would probably cause a shrinkage defect. A cross section of the same casting is shown in figure 30 as it was redesigned to eliminate the heavy section and make the casting more adaptable to directional solidification.

Many times, a casting can be designed to permit easier molding as well as to improve the feeding. The bracket shown in figure 31 is such a casting. The original design did not have


the shaded areas shown. This not only made the making of the mold difficult, but also resulted in heavy sections in the casting with the possibilities of shrinkage defects. By padding the area as shown by the shaded portions, the pattern was easier to draw and feeding of the lugs was simplified.Another example of good casting design is shown in figure 32. Note that the thin sections are connected to the heavy sections which are located so that they may be easily fed.


A few general rules can be made to assist the foundryman in producing a better casting. It must be remembered that in many cases, these rules cannot be followed to the letter. There also may be a conflict between rules. In such a case, a compromise must be made which will best suit the casting desired.

1. The casting thickness, weight, and size should be kept as small as possible, consistent with proper casting performance. (See “Strength Requirements,” page 19, and figure 15.)

2. All sections should be tapered so that they are thickest near the risers. Sections should never be tapered so that thick sections are far from the risers. If proper tapering is impossible, the section should have uniform thickness. (See “Section Thickness,” page 20, and figures 17 and 22.)

3. Abrupt changes in adjoining sections should never be allowed. (See figures 17 and 20.)4. Heavy sections should not be located so that feeding must take place through thin sections.

5. Use ribs to avoid warpage or to add stiffness. Ribbed construction can often be used to replace a heavier section.

6. Where junctions produce thick sections of metal (hot spots), use cores or other methods to eliminate the heavy section (figures 23, 24, 25, 26, and 27, “Wall Junctions,” page 22).

7. A casting should be made as simple as possible. The use of cores should be kept to a minimum. If a casting is complicated, consider the use of several simpler castings which can be welded together.

8. Avoid junctions of several walls or sections at one point.

9. Bosses, lugs, and pads should not be used unless absolutely necessary.

10. Allow for shrinkage and machine finish in dimensional tolerances. (Chapter 3, “Patternmaking,” Table 4, page 27, Table 5, page 28.)


Figure 15. Effect of section size on physical properties.
Figure 15. Effect of section size on physical properties.

Figure 16. Use of fillets.
Figure 16. Use of fillets.

452605 0-58-3


Figure 17. Blending of thin and heavy sections.
Figure 17. Blending of thin and heavy sections.

Figure 18. Wheel design.
Figure 18. Wheel design.


Figure 19. Recommended wheel designs.
Figure 19. Recommended wheel designs.

Figure 20. Transitions in section size.
Figure 20. Transitions in section size.

Figure 21. Simple Directional solidification.
Figure 21. Simple Directional solidification.

Figure 22. Taper as an aid to directional solidification.
Figure 22. Taper as an aid to directional solidification.


Figure 23. Hot spot location by the method of inscribed circles.
Figure 23. Hot spot location by the method of inscribed circles.

Figure 24. Reduction of cross section in L and V junctions.
Figure 24. Reduction of cross section in L and V junctions.

Figure 25. Reduction of cross section in an X junction.
Figure 25. Reduction of cross section in an X junction.

Figure 26. Various treatments for a T junction.
Figure 26. Various treatments for a T junction.

Figure 27. Coring to reduce section in a rib junction.
Figure 27. Coring to reduce section in a rib junction.

Figure 28. Removal of heavy section by redesign.
Figure 28. Removal of heavy section by redesign.


Figure 29. Hub cross section - heavy section.
Figure 29. Hub cross section – heavy section.

Figure 30. Hub cross section - improved design.
Figure 30. Hub cross section – improved design.

Figure 31. Bracket casting.
Figure 31. Bracket casting.

Figure 32. Aluminum yoke casting.
Figure 32. Aluminum yoke casting.


This page is blank.


Chapter III


A pattern is used to form the mold cavity into which molten metal is poured to produce a casting. As such, it is a tool in the hands of the foundryman. A great deal of success in producing a good casting depends on the quality and design of the pattern. For example, a pattern that does not have the proper draft is difficult to draw from the sand without breaking the mold.

The design of the casting itself, as well as that of the pattern, must be taken into consideration to make molding less difficult. The casting design should be as simple as possible, since it will determine the ease with which a pattern can be drawn from the mold, the number of loose pieces required in the pattern, and the number of cores needed.


There are three main types of patterns: loose patterns, mounted patterns, and core boxes.

Loose Patterns. The majority of molds made aboard repair ships are made with loose patterns, since castings required are usually few in number and not too often repeated. A loose pattern is the wood counterpart of the casting, with the proper allowance in dimensions for contraction and machining. A typical loose pattern is shown in figure 33. A loose pattern may be made in one piece or it may be split into the cope and drag pieces to make molding easier. A split pattern is shown in figure 34.

A loose pattern has the disadvantage of requiring a follow board or a false cope to make the parting line, or hand cutting the parting line. The different steps used to make molds from loose patterns are described in Chapter 5, “Making Molds.”

The original casting or the broken parts of a casting which have been put together may be used in an emergency as a loose pattern. In such a case, the part to be used as a pattern must be built up to allow for the contraction of the cast metal and prevent the new casting from being too small. A material known as “Celastic” (see allowance list), supplied in sheets, can be applied to the metal part. When Celastic dries, it will adhere firmly and form a hard surface which may be sandpapered or sawed like wood. For directions on the use of Celastic, see the

section on ‘Maintenance, Care, and Repair” in this chapter.Mounted Patterns. Patterns fastened permanently to a flat board, called a match plate, are known as mounted patterns.

The main advantage of the mounted pattern over the loose pattern is that it is easier to use and store. For these reasons, a mounted pattern is generally warranted when several of the castings (say, five or more) are to be made during one “run” or when the casting is made at frequent intervals.

Another advantage of the mounted pattern is that a pattern of the gating system also can be mounted on the match plate. This practice of molding the gating system eliminates the loose sand that often results when gates are hand cut. As a result, the castings produced usually are better than those produced with the loose patterns.

Core Boxes. Core boxes are actually negative patterns. When looking at a pattern, one sees the casting in its actual shape. A core box on the other hand shows the cavity which will be created by the core. Core boxes are used not only to make cores for holes in castings but also to make parts of a mold. In some cases, a pattern cannot be made so that it can be drawn. In such a case, the part of the casting which would hinder drawing is made as a core that can be placed in the mold after the pattern proper has been withdrawn. The making and proper use of cores is described in Chapter 6, “Making Cores.”


The most commonly used material for patterns is wood, because it is easy to work with and is readily available. Mahogany, white pine, and sugar pine are acceptable materials. Select kiln-dried white or sugar pine is most widely used because it is easily worked and is generally free of warping and cracking.

For pattern work, it is essential that the wood has a low moisture content, 5 to 6 percent if possible, in order to avoid warping and shrinking of the finished pattern.

Metal patterns are usually used as mounted patterns, with the gating included in the pattern. Their use is warranted only when a large number of castings must be made. Mounted metal


patterns are difficult to make and require special skills. The one distinct advantage of a metal pattern is that it does not warp on storage and when removed from storage, no preparation other than cleaning is necessary before use.A material which maybe used for an emergency pattern, when only a small number of castings are required and there is not sufficient time to make a wooden pattern, is plaster or gypsum cement. Gypsum cement is made from gypsum rock, finely ground and heated to high temperatures. When mixed with water, it forms a plastic mass which can be molded, shaped, or cast. Plaster patterns have the disadvantage of being very fragile and require careful handling, therefore, it is recommended for use only in an emergency.


The process of actually laying out a pattern comes under the work of the patternmaker. The various parts of proper pattern layout are discussed briefly here to provide the molder with information which may prove useful in determining any nonconformity between the casting and the original drawing.

Parting Line. The parting line divides the pattern into the parts that form the cavity in the cope (top) and drag (bottom) of the mold. Whenever possible, a casting is designed so a straight parting line can be used; that is, a single flat surface will divide the casting into cope and drag sections. Usually, a straight parting line is necessary if the pattern is to be mounted. When loose patterns are used, the mold may be made easier with a straight parting line than with a broken parting line.

Core Prints. A core print is a projection on the pattern designed to make an impression in the sand for locating and anchoring the core.

Although there are no fixed rules as to the length of core prints or how much taper they should have, practice requires that there should be sufficient bearing surface to support the weight of the core. The following table gives dimensions which have been found successful in practical application.


Size of CoreLength of Core Print
Up to 1 1/2-inch diameter2-inch core print
From 2-inch to 5-inch diameterAt least equal to the diameter of core
Above 5 inches in diameter6-inch core print (minimum)
In general, the length of a core print should equal or slightly exceed its diameter or width. When a core has prints in the cope, the cope prints should provide a “closing clearance” so as to avoid the possibility of crushing sand from the cope when closing the mold. This clearance, however, should not be excessive, as the core will shift under pressure from the molten metal. If it is possible for a core to be set “upside down” or “wrong end to,” locating or indexing lugs (tell-tales) should be provided to prevent this.A good practice for constructing core prints is shown in figure 35. It results in castings with fewer fins at the parting line. Fins tend to produce cracks, and require extra time to clean off the casting. Larger core prints provide better core location and support in the mold. In addition, they reduce the tendency for cracks to form in the cored openings from core fins.

The location, size, and type of vent holes, to allow gases to escape, should be indicated on the core prints and in the core box by means of strips or projections, or by some other appropriate means.

Chaplets. When the design of the core is such that additional support over and above that given by the core prints is needed, it is necessary to use chaplets. These chaplets are pieces of metal especially designed to support the core. Detailed description of chaplets and their use will be found in Chapter 5, “Making Molds.” Their use is to be avoided wherever possible, particularly on pressure castings. If chaplets are necessary, their location and size should be indicated on the pattern and core box by raised sections such as shown in figure 36. This additional metal in the mold cavity serves two main purposes; first, it accurately locates the best chaplet position and insures that the location will be consistently used; second, it provides an additional mass of metal to aid in the fusion of the chaplet, which is necessary to obtain pressure tightness.

Shrinkage Rules. The patternmaker uses rules which are somewhat longer than the numbers indicate. The size of such a rule allows for shrinkage of the casting. A 1/4-inch shrink rule, for instance, is 12 1/4-inches long, although the markings would indicate that it is only 12 inches long.

The shrinkage rule to be used in constructing a pattern must be selected for the metal which will be used in the casting. It must be remembered that the shrinkage rule will also vary with the casting design. For example, light and medium steel castings of simple design and no cores require a 1/4-inch rule,


whereas for pipes and valves where there is a considerable resistance offered to the contraction of the steel by the mold and cores, a 3/16- inch rule will be adequate. Shrinkage allowances for various metals and mold construction are listed in table 4.Machining Allowances. The machining allowance or finish is usually made on a pattern to provide for extra metal on the casting during heat treatment. This distortion is also a factor in determining machining allowance. In table 5 are listed some finish allowances which may be used as a guide.

Many times, the exact finish allowances can be found by referring to the original blueprints of the part to be cast.

Draft. Draft is the amount of taper given to the sides of projections, pockets, and the body of the pattern so that the pattern may be withdrawn from the mold without breaking the sand away. This also applies to core boxes. The breaking of sand due to a lack of taper is shown in figure 37. The same pattern with correct taper is shown in figure 38. When a straight piece, such as the face of a flange or a bushing, is made, the amount of draft is usually 1/8-inch per foot. In green sand molding, interior surfaces will require more draft than the exterior surfaces

that are to be machined. Some castings do not require finish since they are used in the rough state just as they come from the final cleaning operation. Most castings are finished only on certain surfaces, and no set rule can be given as to the amount of finish to be allowed. The finish is determined by the machine shop practice and by the size and shape of the casting. A casting may become distorted from stresses during the casting process or surfaces, because of the lower strength of isolated volumes of sand. The draft is dependent on the shape and size of the casting and should at all times be ample. The actual draft to be used is usually determined by consultation between the patternmaker and the molder. Proper and improper drafts are shown in figure 39.Distortion Allowance. Many times a casting is of a design which results in cooling stresses that cause distortion in the finished casting. The design may also be such that it cannot be corrected in the design. In such a case, the experience of the molder and pattern maker must be relied upon to produce a good casting. Distortion allowances must be made in a pattern and are usually determined by experience. Recorded information on castings of this type is very useful in determining distortion allowances on future work.

Casting AlloysPattern Dimension (inches)Type of ConstructionContraction
(inches per foot)
Gray Cast IronUp to 24Open construction1/8
From 25 to 48Open construction1/10
Over 48Open construction1/12
Up to 24Cored construction1/8
From 25 to 36Cored construction1/10
Over 36Cored construction1/12
Cast SteelUp to 24Open construction1/4
From 25 to 72Open construction3/16
Up to 18Cored construction1/4
From 19 to 48Cored construction3/16
From 49 to 66Cored construction5/32
AluminumUp to 48Open construction5/32
49 to 72Open construction9/64
Up to 24Cored construction5/32
Over 48Cored construction9/64 to 1/8
From 25 to 48Cored construction1/8 to 1/16
Bronze1/8 to 1/4



Casting AlloyPattern
Cast ironUp to 121/83/32
13 to 243/161/8
25 to 421/43/16
Cast steelUp to 123/161/8
13 to 241/43/16
25 to 425/165/16
Brass, bronze,
and aluminum
Up to 123/321/16
13 to 243/161/8
25 to 363/165/32

A typical casting which would require distortion allowances is a simple yoke casting shown in figure 40. Part (a) shows the casting as it was designed. The yoke made to this design is shown in part (b) with the arms widened out. The arrows indicate the direction of the cooling stresses which produced contraction in the cross member. Part (c) shows the pattern as made with distortion allowances, and part (d) shows the finished part. In part (d) the arrows again show the direction of the cooling stresses which were used to produce a straight yoke.


Skilled patternmakers are available aboard repair ships to make patterns. Construction of patterns, therefore, is not discussed in detail in this manual. Detailed information on patternmaking can be found in the patternmaker’ manuals aboard ship.

If broken parts are to be used as a pattern, extreme care must be taken to insure proper alignment of the parts when they are joined or placed for molding. The surfaces should be as smooth as possible and the size of the casting should be increased wherever necessary to compensate for contraction. The use of Celastic for this purpose is described in the section on “Maintenance, Care, and Repair.”

For applications where the quantity of castings required is small and the designs are quite simple, gypsum cement can be used as pattern material with success. See Pattern Materials.

One disadvantage in the use of this material is that it is fragile and is likely to be damaged in handling, molding, and storage. Internal support can be provided through the use of arbors, rods, wire frames, intermixed

hair, while external reinforcements can be provided by surface coating. Typical patterns produced in gypsum cement are shown in figure 41. Making of a simple pattern is shown in figure 42.


Shellac is usually used to fill the pores in wood patterns or to seal plaster patterns. The patterns are rubbed smooth to eliminate the possibility of sand adhering to the pattern because of a rough surface. Plaster patterns may be metal sprayed to produce a hard, smooth surface. After the surface of the patterns has been properly prepared, various parts are painted for identification.

The color code used for identifying different parts of a pattern is as follows:

1. Surfaces to be left UNFINISHED are painted BLACK.2. Surfaces to be MATCHED are painted RED.

3. Seats of, and for, LOOSE PIECES are marked by RED STRIPES on a YELLOW BACKGROUND.




The patterns normally made aboard repair ships are used for a few castings and then they must be stored. It is important that storage space be provided which is as free of moisture as possible. This precaution will maintain the patterns in good condition and prevent warping and cracking. The storage of patterns should be in properly constructed racks wherever possible. This will keep pattern damage down to a minimum.

A record should be kept of all patterns which are on hand. These records should contain a complete description of the pattern, pattern numbers, class of ship, size of part, and drawing and piece number. Such records are useful in locating a pattern for future use. They may also be used to provide a pattern for a similar casting which may be required. Time can be saved by slightly altering a pattern already on hand or using the pattern as designed and making alterations in the machining operations. Any permanent pattern changes, no matter how


small, should be noted on the pattern record, and if possible on the blueprint of the casting.Many times a core box has to be repaired or altered slightly. Sheet lead or sheet brass of varying thicknesses may be used. Celastic may also be used to repair a pattern or core box. Minor repairs to the pattern or core box may easily be made by a molder, but any repair of a major nature should be by a pattern-maker. After any repair is made, the pattern should be checked to make sure that it conforms to the drawing. A periodic check of patterns or core boxes and minor repair of them will go a long way toward keeping the patterns in good usable condition and prevent major repairs later.

Directions for Applying Celastic.1. Clean the surface where it is to be applied.

2. Cut pieces to the size required or a number of pieces to cover the required area.

3. Immerse the Celastic in the solvent (methyl ethyl ketone) until it becomes very pliable and sticky. In this state, it can be applied to the pattern and will shape very easily, even on irregular contours, by pressure from the fingers.

4. After the solvent has evaporated, the Celastic will adhere firmly to the pattern and the outer surface will be relatively hard. It may then be sanded and lacquered to a smooth surface.

WARNING: Celastic shrinks in thickness after dipping and drying, and proper allowance must be made. If a greater thickness is desired on any surface, one or more pieces may be applied to the original layer of the Celastic. Two small metal pans should be available for submerging the Celastic; any solvent left in the pan may be returned to the bottle.


The calculation of casting weights is important in the operation of any foundry. For that reason, some information on the methods and practices used is given.

It is obviously quite simple to calculate the weight required to pour a casting if the defective part is to be used as a pattern, or if it is on hand. Since risers and gates are usually round (and should be) in their cross section, it is easy to calculate their weight and add it to the weight of the casting.

Another simple method that can be used in cases where a small pattern of solid wood construction with no cores is to be used consists in weighing the patterns and multiplying this figure by the following:

For steel17.0
For cast iron16.5
For bronze18.5
For aluminum5.0

To this figure, the weights of heads and gates are added.

Caution must be used in following this practice; if the pattern is not of solid construction or if it is not made of white pine, an erroneous answer will be obtained. Sugar pine and mahogany have a greater density and a lower factor must be used to calculate the casting weights. Where neither of these methods is possible, it is necessary to break down the design into simple sections–such as rounds, squares, and plates–and calculate the weight of each section by determining its volume in cubic inches, multiplying this figure by the following weights per cubic inch, and then obtaining the total:

Pounds per
cubic inch
Cast steel0.284
Cast iron0.260
Compositions G and M0.317
Manganese bronze0.303

This method is demonstrated in the case of the designs shown in figures 43 and 44.

In table 6 are areas and volumes for calculating weights of castings. This table shows the various shapes and formulas which are useful in calculating casting weights.


Making a pattern is the job of a skilled patternmaker, but a knowledge of the factors involved in patternmaking is useful to the molder. Many times a defective casting can be traced to not enough draft, improper parting line, or insufficient core prints. A molder who is able to recognize a defect caused by improper pattern work or a pattern requiring repair can save himself a lot of time by having the pattern corrected.

The factors discussed in this chapter are not intended to supply all the answers relating to patternmaking. The molder should use this information to guide him in maintaining his patterns and recognizing when they are in need of attention.



Rectangle and ParallelogramRectangle and ParalleogramArea = ab

TriangleTriangleArea = 1/2 cd.
Area = SQRT(s(s-a)(s-b)(s-c)) when
s= 1/2(a + b + c)
Example: a = 3″, b = 4″, c = 5″
s = (3″ + 4″ + 5″)/2 = 6″
Area = SQRT(6 (6-3) (6-4) (6-5)) = 6 sq. in.

Regular Polygonspoligonn = Number of sides, s= Length of one side, r= Inside radius
Area = 1/2 nsr

of Sides
51.72047 s2 = 3.63273 r2
62.59809 s2 = 3.46408 r2
73.63395 s2 = 3.37099 r2
84.82847 s2 = 3.31368 r2
96.18181 s2 = 3.27574 r2
107.69416 s2 = 3.24922 r2
119.36570 s2 = 3.22987 r2
1211.19616 s2 = 3.21539 r2

TrapeziumTrapeziumArea = 1/2 [a (e + d) bd + ce]
Example: a = 10″, b = 3″, c = 5″, d = 6″, e = 8″
Area = 1/2 [10 (8 + 6) + (3 X 6) + (5 X 8)] = 99 sq. in.

SquareSquareThe diagonal of a square = A X 1.414
The side of a square inscribed in a given circle is: B X .707.


CircleCircleθ (the Greek letter Theta) = angle included between radii
π (pi) = 3.1416, D = Diameter, R = Radius, C = Chord.
h = Height of Arc, L = Length of Arc.
Circumference = πD = 2πR = 2 SQRT(π X Area)
Diameter = 2 R = Circumference / π = 2 SQRT(Area/π)
Radius = 1/2 D = Circumference / 2 π = SQRT(Area/π)
Radius = ((c/2)2 + h2)/2h
Area = 1/4 π D2 = 0.7854 D2 = π R2
Chord = 2 SQRT(h (D – h)) = 2R X sine 1/2θ
Height of Arc, h = R – SQRT(R2-(C/2)2)
Length of Arc, L = θ/360 x 2 π R = 0.0174533 Rθ
1/2 θ (in degrees) = 28.6479 L/R
Sine(1/2 θ) = (C/2) / R

Sector of a CircleSector of a CircleArea = 1/2 LR
Example: L = 10.472″, R = 5″
Area = 10.472/2 x 5 = 26.180 sq. in.
or Area = π R2 X θ/360 = 0.0087266 R2θ
Example: R = 5″, θ = 120°
Area = 3.1416 X 52 X 120/360 = 26.180 sq. in.

Segment of a CircleSegment of a CircleArea = πR2 X θ/360 – C(R – h)/2
Example: R = 5″, θ = 120°, C = 8.66″, h = 2.5″
Area = 3.1416 X 52 X 120/360 – (8.66(5 – 2.5))/2 = 15.355 sq. in.
Length of arc L = 0.0174533 R θ
Area = 1/2 [LR-C (R-h)]
Example: R = 5″, C = 8.66″, h = 2.5″, θ = 120°
L = 0.0174533 X 5 X 120 = 10.472″
Area = 1/2[(10.472 X 5) – 8.66(5 – 2.5)] = 15.355 sq. in.


Circular Ring
Circular RingArea = 0.7854 (D2-d2), or 0.7854 (D-d)(D+d)
Example: D = 10″, d = 3″
Area = 0.7854 (102 – 32) = 71.4714 sq. in.

SpandrelSpandrelArea = 0.2146 R2 = 0.1073 C2
Example: R = 3
Area = 0.2146 X 32 = 1.9314

Parabolic SegmentParabolic SegmentArea = 2/3 sh
Example: s = 3, h = 4
Area = 2/3 X 3 X 4 = 8

EllipseEllipseArea Tab = πab = 3.1416 ab
Example: a = 3, b = 4
Area = 3.1416 X 3 X 4 = 37.6992

Irregular FiguresIrregular FiguresArea may be found as follows:
Divide the figure into equal spaces as shown by the lines in the figure.
(1) Add lengths of dotted lines.
(2) Divide sum by number of spaces.
(3) Multiply result by “A.”


Ring of Circular Cross SectionRing of Circular Cross SectionArea of Surface = 4 π2Rr = 39.4784 Rr
Area of Surface = π2 Dd = 9.8696 Dd
Volume = 2 π2 Rr2 = 19.7392 Rr2
Volume = 1/4 π2 Dd2 = 2.4674 Dd2

SphereSphereSurface = 4 π r2 = 12.5664 r2 = π d2
Volume = 4/3 π r3 = 4.1888 r3
Volume = 1/6 π d3 = 0.5236 d3

Segment of a SphereSegment of a SphereSpherical Surface = 2 π rh = 1/4 π(c2 + 4h2) = 0.7854 (c2 + 4h2)
Total Surface = 1/4π (c2 + 8 rh) = 0.7854 (c2 + 8 rh)
Volume = 1/3 π h2 (3 r – h) = 1.0472 h2 (3 r – h)
Volume = 1/24 πh (3c2 + 4h2) = 0.1309 h (3c2 + 4h2)

Sector of a SphereSector of a SphereTotal Surface = 1/2πr (4 h + c) = 1.5708 r (4 h + c)
Volume = 2/3 πr2h = 2.0944 r2h

CylinderCylinderCylindrical Surface = π dh = 2 πrh = 6.2832 rh
Total Surface = 2 π r (r + h) = 6.2832 (r + h)
Volume = π r2h = 1/4 πd2h = 0 7854 d2h


PyramidPyramidA = area of base
P = perimeter of base
Lateral Area = 1/2 Ps
Volume = 1/3Ah

Frustum of a PyramidFrustum of a PyramidA = area of base
a = area of top
m = area of midsection
P = perimeter of base
p = perimeter of top
Lateral Area = 1/2s (P + p)
Volume = 1/3h (a + A + SQRT(aA)
Volume = 16h (A + a + 4m)

ConeConeConical Area = πrs = πr SQRT(r2 + h2)
Volume =1/3 π r2h = 1.0472 r2h = 0.2618 d2h

Frustum of a ConeFrustum of a ConeA = area of base
a = area of top
m = area of midsection
R = D / 2; r = d / 2
Area of Conical Surface = 1/2 πs (D + d) = 1.5708 s (D+d)
Volume = 1/3 h (R2 + Rr + r2) = 1.0472 h (R2 + Rr + r2)
Volume = 1/12 h (D2 + Dd + d2) = 0.2618 h (D2 + Dd + d2)
Volume = 1/3h (a + A + SQRT(aA)) = 1/6 h (a + A + 4m)


Figure 33. One piece pattern.
Figure 33. One piece pattern.

Figure 34. Split pattern.
Figure 34. Split pattern.

452605 0-58-4


Figure 35. Core print construction.
Figure 35. Core print construction.

Figure 36. Chaplet location with pads.
Figure 36. Chaplet location with pads.

Figure 37. Mold broken due to a lack of taper.
Figure 37. Mold broken due to a lack of taper.

Figure 38. Clean pattern draw with correct taper.
Figure 38. Clean pattern draw with correct taper.

Figure 39. Pattern draft.
Figure 39. Pattern draft.


Figure 40. Distortion allowance in a simple yoke pattern.
Figure 40. Distortion allowance in a simple yoke pattern.

Figure 41. Plaster patterns and core boxes.
Figure 41. Plaster patterns and core boxes.

Figure 42. Making a simple plaster pattern.
Figure 42. Making a simple plaster pattern.

Figure 43. Calculating casting weight.
Figure 43. Calculating casting weight.

Figure 44. Calculating casting weight.
Figure 44. Calculating casting weight.


This page is blank.


Chapter IV
The principal molding material used in foundries is silica sand. Silica sand is readily available, low in cost, and possesses properties that enable it to withstand the effects of molten metals.The primary function of any molding material is to maintain the shape of the casting cavity until the molten metal is poured and until the casting solidifies. The properties of silica sand that make it useful as a molding material are its refractoriness and its ability to be formed into complicated shapes easily. Its refractoriness enables it to withstand the intense heat from molten metals. Its ability to be formed into shapes is attained by the action of naturally occurring clay (clay that is quarried with the sand) or added clay, additional binders, and water. The binder maintains the sand in place until the casting is poured and solidified.

The three major parts of a molding sand are: (1) the sand grains, which provide the necessary refractory properties; (2) the bonding material, which may be a naturally occurring clay in the sand or an added material such as bentonite or cereal; and (3) water, which makes possible the bonding of the sand grains by the binder to make the sand a useful molding material.


Because storage space aboard repair ships is limited, it is to the molder’s advantage to stock only a few types of foundry sands. From this point of view, the use of an all-purpose sand is advantageous in that only one facility for new sand is required for all of the metals cast aboard ship. Many times, it may be impossible to obtain the all-purpose sand required, and a locally available sand will have to be used. In such instances, the various properties of the substitute sand will have to be determined before the sand is used in the foundry. All of the sand properties discussed in the section, “Sand Properties,” apply to natural sands as well as to synthetic and all-purpose sands.


Natural sands contain only the clay that is already associated with them when mined. Such a sand is often used as it is received, with only moisture added to obtain the desired properties. Albany sand is a typical example of a natural sand. A naturally bonded sand has the advantages of maintaining its moisture content for a

long period of time, having a wide working range for moisture, and permitting easier patching and finishing of molds. One disadvantage of natural sand is that its properties vary and are not so consistent as desired. Additions of bentonite are sometimes made to natural sands. Such a sand is called “semisynthetic.”SYNTHETIC SANDS

Sands that fall under the designation of “synthetic” sands are not actually synthesized from the various elements. They are made by mixing together the various individual materials that make up a molding sand. (See glossary, Synthetic Sand.) A more appropriate name would be “compounded” sands. However, the name synthetic has become established in the foundry industry, through usage, to designate a sand of this type.

Synthetic sands consist of a naturally occurring sand with a very low clay content, or a washed sand (all of the natural clay removed), and an added binder, such as bentonite. Synthetic sands have the following advantages over naturally bonded sands: (1) more uniform grain size, (2) higher refractoriness, (3) mold with less moisture, (4) require less binder, (5) the various properties are more easily controlled, and (6) less storage space is required, since the sand can be used for many different types of castings.


Sands that are used for a variety of casting sizes and types of metals are called “all-purpose” sands. In commercial practice, different sands are used to cast different metals and different sizes of castings of the same metal, but in a shipboard foundry, the limitation of storage space makes the practice of maintaining many special sands impossible. A synthetic sand used as a base for an all-purpose sand has the requirements for a molding sand for shipboard use. Naturally, some advantages will have to be sacrificed in using one sand for making all types of castings. The major factor that will be sacrificed in this respect is that of surface finish. However, the principal purpose of a shipboard foundry is to produce serviceable castings. Surface finish is often not a major requirement. As an example, a coarse-grained sand suitable for steel castings will produce rough surface finishes on lighter nonferrous castings made in the same sand. This is a minor disadvantage for an all-purpose sand when compared to its advantages for repair- ship use.



There are a great many properties of sand which are of interest to the production foundry-man. Among the most important are: (1) green permeability, (2) green strength, (3) dry strength, (4) moisture content, (5) clay content, and (6) grain fineness. These will be discussed in greater detail. The other properties include hot strength, sintering point, deformation, and collapsibility. The six properties selected as most important are those with which repair-ship molders should be most familiar. These are also the properties that can be determined by the use of the sand-test equipment aboard ship.


Green permeability is that property of a molding sand that permits the passage of air, gases, or steam through the sand. The openings between the sand grains in a mold give sand its permeability. There are four factors that control the permeability of foundry sand: (1) fineness of the sand grains, (2) shape of the sand grains, (3) the amount and type of binder, and (4) the moisture content. Permeability is expressed as a number that increases with an increasing openness of the sand.

Grain Fineness. Grain fineness is an indication of the grain size of the sands. It is expressed as a number that tells a molder if he has a fine sand, made up largely of very small sand grains, or a coarse sand, composed mainly of large sand grains. A detailed description of grain-fineness number is given under “Methods for Testing Sands,” in this chapter.

The general effect of grain size on permeability is shown by figure 45. Data for this curve were obtained by screening a given sand through a series of test screens and then making a permeability test on the sand retained on each screen. The permeability of the coarse sand is very high. As the sand grains become smaller, the permeability decreases rapidly. This decrease is due to the smaller voids or openings between the individual sand grains for the fine sand. Coarse sand grains have the same general size relation to fine sand grains as basketballs have to marbles.

The permeabilities of four typical foundry sands, ranging from coarse to fine, are shown in figure 46. The numbers shown on the graph are the grain-fineness numbers. The coarse sand, having a greater amount of large sand grains and large voids between the grains, has a high permeability. The other sands, having a greater amount of small sand grains and small voids, have lower permeabilities.

Shape of the Sand Grains. There are two primary shapes of sand grains, angular and rounded. There are many degrees of roundness or angularity between the two extremes. Angular grains can be compared to crushed stone. There are sharp edges and corners on the grains. The rounded sand grains have the appearance of beach pebbles that have been rounded by the action of the sea. Sharp angular sand grains cannot pack together as closely as rounded sand grains. As a result, sand with angular grains have a higher permeability than sands with rounded grains. The effect of grain shape on the permeability of molding sand is shown in figure 47. (The word “sharp,” incidentally, when applied to molding sands has nothing to do with grain shape. A sharp sand is simply a sand very low in Clay content.)Binder. The amount and type of binder also have an effect on the permeability of foundry sand. The effect of increasing amounts of bentonite on permeability is shown in figure 48. The permeabilities are shown for moisture contents of 2 and 4 percent. With 2 percent moisture, the sand shows a rapid decrease in permeability with increased bentonite content. Sands containing 4 percent moisture show a fairly constant permeability after 4 percent bentonite is reached. This type of information indicates that 4 percent of moisture in this particular sand would produce the best permeability over a range of bentonite contents. The type of binder also affects permeability, as shown in figure 49.

Moisture Content. The effect of moisture content on permeability was shown in figures 46 and 47. Low permeability at very low moisture content is caused by the dry clay particles filling the spaces between the sand grains. Figures 46 and 47 both show an increase in permeability to a maximum value, and then a decrease with further additions of water. The increase in permeability is produced when the moisture causes the clay particles to agglomerate or stick together. This action is similar to the addition of water to dust to form a firm piece of soil. When water is added in excess of the amount to produce this sticking together, the excess water begins to fill in the holes between the sand grains and as a result, the permeability goes down. This action is similar to the addition of water to a firm soil to produce mud.


Green strength is the strength of molding sand just after it has been tempered. (Refer to glossary, “temper.”) It is the strength which is required for the handling of the sand during the molding operation and, if a mold is poured soon after it is completed, it is the strength which must maintain the shape of the mold.


Green strength is expressed as the number of pounds per square inch required to crush a standard specimen. The same factors that control permeability also control the green strength of foundry sand. They are ( 1) grain fineness, (2) shape of the sand grains, (3) the amount and type of binder, and (4) the moisture content. Mulling practice or mixing practice also affect the green strength of sand. This is discussed in detail in the section on “Mixing.”Grain Fineness. The smaller the size of the sand grains in a given amount of molding sand, the greater will be the area of contact between the many grains. As a result, the green strength of the finer sand is high. A coarse sand, on the other hand, will have a much smaller contact area for the same amount of sand, and the green strength is lower. This is illustrated in figure 50. The green strength increases as the sand changes from a coarse sand to a fine sand. Figure 51 shows the variation in green strength for four different sands. The sand with the highest fineness number (108) is the finest sand and has the highest green strength for a given moisture content. The other sands become progressively weaker as they become coarser.

Grain Shape. The area of contact between the sand grains is also affected by the shape of the grains. Round grains pack together much more closely than sharp, angular grains and, as a result, have a stronger bond than the angular sand. A comparison of the green strengths of round and angular sands is made in figure 52.

Binder. Green strength is affected directly by the amount of binder which is added. The more binder used, the higher will be the green strength, as shown in figure 53. The type of binder used (clay, cereal, dextrine, or rosin) also affects the green strength of molding sand. The effect of bentonite and fireclay on the green strength is shown in figure 54.

Moisture Content. The effect of moisture on green strength is similar to its effect on permeability. The green strength increases with the first additions of water, reaches a maximum strength, and then starts to decrease. This is illustrated in figures 47 and 53.


The dry strength of sand mixtures is generally affected in the same way as green strength by grain fineness, grain shape, and moisture content. Different binders, however, can affect dry strength and green strength differently. For example, in comparison with western bentonite, southern bentonite produces a high green strength and a low dry strength. Southern bentonite is widely used for its low dry strength and the resulting easy shakeout of castings.

When cereal and dextrine are added to bentonite, the bonded mixtures give a higher dry strength. For more information on the behavior of different binders, see the next section.BINDERS

Binders are the materials added to molding sands to hold the individual sand grains together to provide a usable molding material.

Green strength, dry strength, and permeability are the properties of the sand which are directly affected by the amount and type of binder. The change in permeability with a change in bentonite content is shown in figure 48. Figure 49 shows the effect of bentonite and fire clay on permeability.

The change in green strength with a change in the bentonite content is shown in figures 53 and 55. In figure 55, it can be seen that for any given amount of bentonite, there is not a large change in green strength with a change in the moisture content. If the moisture content is maintained at a given value, the green strength can be changed over a considerable range by adjusting the amount of bentonite. Figure 54 shows the effect of fire clay and bentonite on green strength. This shows the advantage of bentonite over fire clay as a binder. The green strength due to fire clay decreases rapidly with increased moisture, while the green strength due to bentonite decreases much less for the same moisture contents.

The effects of blending western and southern bentonite on the green strength and dry strength of a sand with an AFS Fineness Number of 50 to 60 are shown in figure 56. There is a rather uniform decrease in dry strength with a changeover from western to southern bentonite. The green strength increases slightly from 100 percent western bentonite through the various mixtures and then increases rapidly as the 100 percent southern bentonite bond is used. This shows the difference in properties that result from the use of the two different bentonites, or mixtures of the two bentonites. The low dry strength of southern bentonite is especially advantageous when a sand mixture having good collapsibility is required, for instance, when casting alloys that are apt to hot tear easily.

Other binders (such as cereal, dextrine, and rosin) are often used as additives to augment or modify the clay binders. The cereal binders are wheat and corn flours. A cornflour binder slightly improves the green strength and makes a decided improvement in the dry strength. Wheat-flour, on the other hand, contributes very little to green strength, but improves the collapsibility of a sand. It is important to realize that the effects of all cereal


binders are not the same in influencing the properties of molding sands. Dextrine binders are a form of sugar and produce a much higher dry strength than do cereal binders. However, dextrines also cause a reduction in the green strength of the sand mixture. Molasses can be used as a substitute for dextrine, but its influence on sand properties is not so great as that of dextrine. Rosin binders are commercial byproducts that are used principally as core binders or in sand mixes for dry sand molds. A rosin-bonded sand has a very hard surface when baked, but has the disadvantage that it absorbs moisture on standing. Because of this characteristic, molds and cores made with rosin-bonded sands should be used as soon after baking as possible.GRAIN FINENESS

The effect of fineness of a foundry sand is discussed under the various other properties. Briefly, a fine sand will have a higher strength and lower permeability, for a given moisture and binder content, than will a coarse sand.


In a molding sand, hot strength and collapsibility are two properties which are important to the foundryman. Hot strength is the strength that a molding sand has when it is at the pouring temperature of the various molten metals. Hot strength is necessary in a sand mixture to retain the shape of the mold before solidification of the metal starts. Hot strength should not be confused with retained strength, which is the strength of molding sand after it has been heated and permitted to cool to room temperature. Collapsibility is the property that permits a sand mold or core to crumble when it is subjected to the forces exerted by a contracting casting. The determination of hot strength and collapsibility is impossible with the sand-testing equipment aboard ship, but general determinations of these properties can be made by observation. The two properties of hot strength and collapsibility go hand in hand, and one cannot be discussed without the other. The ideal foundry sand would have a high hot strength and good collapsibility, but this combination is difficult to attain, except through very close control of sand processing. The hot strength and collapsibility of the sand can be checked by observing the condition of the sand when shaking out a casting. If the sand is difficult to remove from deep pockets, then the sand lacks adequate collapsibility. A hot tear in a casting is an indication of too high a hot strength, and also a lack of collapsibility.

In this discussion of sand properties, it is obvious that all of the various factors affecting the properties of molding sand are dependent on

each other. This interdependence of properties must be kept in mind constantly, especially when trying to determine the cause of casting defects due to sand. The apparently obvious cause of a defect may not be the actual factor causing that defect, and in many cases it is a combination of sand properties that leads to a defect.REBUILDING OF SANDS

The binder in foundry sands is burned out by the heat of the molten metal. As a result, the green strength of the sand becomes lower and the permeability decreases as the sand is reused. The permeability decreases because of the increase in fines in the sand. The use of sand-testing equipment periodically to measure these properties of molding sands enables the molder to make appropriate additions to the sand before it has deteriorated to the point where it must be discarded. If a continuous check is made, corrections can be made by the addition of small amounts of binder, and more uniform day-to-day properties can be maintained.

Additions of new binder may be as little as one-third to one-half percent of the sand by weight if additions are made frequently and are made as shown necessary by test information. The actual amount of binder required will depend on the type of binder and on the manner in which it is added. The effect of fire clay and bentonite as binders is shown in figures 49 and 54. Note that the fire clay gives a much weaker bond than bentonite, and would require a larger addition to attain the same strength as a bentonite-bonded sand.


When rebonding sands, the use of a muller is necessary to obtain the maximum benefits. A much larger percentage of binder is required if the sand is mixed manually with a shovel.

Mulling Sand. To obtain the maximum properties from a molding sand, a muller should be used for the mixing of all foundry sands. It is especially important that core sands and facing sands be mixed in a muller, but the mixing of all sands in a muller provides a more uniform day-to-day operation. The use of a muller to mix and rebond sands is essential to good sand control, and shows up in the production of better castings.

The literature supplied with the mullers aboard repair ships should be consulted for proper operating instructions. The best results are obtained by mixing the dry sand and dry bond for at least one minute. This operation distributes the bond evenly throughout the sand. A part of the temper water is then added, the


sand mixed for a suitable period of time, the balance of the temper water added, and mixing completed. The total mixing time after the water additions should be approximately as shown in table 7.


Type of MixerSize of
cu ft
Time for
Time for
Sand, minutes
Simpson5 1/253

A mixing time longer than those listed in table 7 does not increase the green strength. This is shown in figure 57. It is good practice to make a series of tests for green strength after various mulling times to determine the time needed to attain the maximum green strength.

Mulling of sand distributes the clay and other binders over the individual sand grains by a kneading and smearing action. Such distribution of the binder is impossible to achieve by manual operation, no matter how thoroughly it is done. In addition to distributing the binder uniformly, the mulled sands require a smaller amount of binder than does a hand-mixed sand. The increased amount of binder required in a hand-mixed sand also results in a lower permeability than in a mulled sand of the same green strength.

Manual Mixing. Situations may arise when mulling of sand is impossible and manual mixing of the sand will have to be done. When such mixing is necessary, it should be done preferably the day before the sand is to be used.

The binder should be added to the sand heap in small amounts in the dry condition and mixed thoroughly after each addition. After the binder has been added and dry mixing completed, the temper water should be added a little at a time with a sprinkling can while the sand is being mixed. On completion of the mixing operation, the sand should be passed through a three or four-mesh riddle and permitted to stand (or temper) for at least a few hours. Preferably, a hand-mixed sand should be covered with wet burlap bags and permitted to stand overnight.


The all-purpose sand that is used in Navy foundries is a “compounded” or “synthetic” sand that has been developed by the Naval Research Laboratory. A wide range of properties can be attained in the molding sand with a minimum of bonding materials such as bentonite, cornstarch, and dextrine. Sand properties for an all-purpose sand having an AFS Fineness Number of 63 are discussed in the following section. Properties of sands having higher or lower AFS Fineness Numbers (finer or coarser sands) will generally vary as described in the section on Sand Properties. (See figures 45, 46, 50, and 51.)


The principal properties (green strength, permeability, and dry strength) of a 63 AFS Fineness sand are shown in figures 58 and 59. This graphical method of presenting the information is used so that the interrelation of the various properties can be easily seen.

The relationships between green compressive strength, moisture content, bentonite content, and permeability are shown in figure 58. The green strength of the sand increases with increased amounts of bentonite. Notice that for each bentonite content, there is a rapid increase in green strength with the first additions of moisture, and then a gradual decrease in green strength as the moisture content is increased. The broken lines in figure 58 show the various permeabilities that are obtained for the various bentonite and moisture contents.

The relationships between green compressive strength, moisture content, bentonite content, and dry strength are shown in figure 59. In figure 59, the broken lines show the dry strengths that are obtained with the various bentonite and moisture contents. Figures 58 and 59 provide information on the direction in which changes can be made to correct the sand properties, and also give information on the particular combinations of binder and moisture to use in a new mix to obtain certain desired properties.

As an example, assume that a sand was prepared with 4 percent bentonite and 4 percent moisture, and that it had a green compressive strength of 4.5 p.s.i. and a permeability of 95. Assume that this sand is found to be unsatisfactory because the green strength is too low, and it is desired to increase the green strength without changing the permeability. Reference to figure 58 shows that this change could be made by increasing the bentonite content to 5 percent


and reducing the moisture content to 3 percent. This new combination of bentonite and moisture contents would provide a sand that has a green strength of 7 p.s.i., with the permeability still at 95. From figure 59, it can be seen that this change in moisture and binder contents will probably cause a decrease in dry strength of only 10 p.s.i., reducing the dry strength from 110 to 100 p.s.i.As a second example, assume that a sand was prepared with 4 percent bentonite and 4.5 percent moisture. This sand would probably have a green compressive strength of 4.5 p. s. permeability of 90, and a dry strength of 120 p.s.i. Assume that this sand is found to cause difficulties in shake-out or to cause hot tearing in the casting. This would indicate that the dry strength might possibly be too high. References to figure 59 shows that by keeping the bentonite content at 4 percent, but decreasing the moisture content to 3 percent, the dry strength will be decreased to approximately 90 p.s.i. This change in moisture would produce only a small increase in the green strength from 4.5 to 5 p.s.i., and increase the permeability from 85 to 105.

When referring to these figures, it must be remembered that only bentonite was considered as a binder. Other materials can also be added as binders to improve green strength or dry strength. The effects of these other binders were discussed in the section on binders.

Another word of caution on figures 58 and 59. These figures should not be used as an indication of properties for all other sands that may have a similar fineness number and the same type of binder. Figures 58 and 59 were based on information obtained from a particular sand, and are used here mainly to show a method of presenting sand-property information in a condensed and usable form.

The green compressive strength of sands of the various grain class numbers that are to be used in shipboard foundries will vary generally as shown in figure 60. This figure should

not be taken to mean that there is a sharp separation between the properties of the different classes of sands. There will be some overlapping of the indicated areas because of differences in sand-grain distributions within sands having the same fineness numbers.It is recommended that a test series (such as that required to produce the information for figures 58 and 59) be made on each new shipment of sand before it is used in the foundry. Conducting such a series of tests and putting the information in graphical form would be a useful and informative way of conducting shipboard instruction periods. The information is developed by making a series of sand mixtures having different bentonite (or other binder) contents and different moisture contents. As an example, a series of 2 percent bentonite sand mixes with 1/2, 1, 2, 3, 4, 5, and 6 percent moisture can be tested for green strength, permeability, and dry strength. A second series of sand mixes containing 3 percent of bentonite and the same moisture contents can be tested to obtain the same properties. This procedure is then repeated for the remaining bentonite contents. The final information is then plotted to produce graphs similar to those shown in figures 58 and 59.

When a new shipment of sand is received aboard ship, a few spot tests can be made to determine how the new lot of sand compares with the previous lot. If the properties are reasonably close, the charts developed for the previous sand may be used for the new sand. However, if there is a significant difference in the physical properties, a complete series of tests should be conducted on the new lot of sand to develop a complete picture of the properties of the new sand.


Listed in the following tables are various examples of sand mixes that may be used as a starting point in preparing all-purpose sand for use aboard repair ships.



SandMaterials, percent by weightPropertiesCasting
Sea Coal
Skin dried470-10045.5
8. 070-8060 and



SandMaterials, percent by weightPropertiesCasting
7.5 – 9.0120to 500
Backing Sand
5.0 – 7.0120to 500
Skin dried550-7095. – 6.590-120100 and over


SandMaterials, percent by weightPropertiesCasting
5.0–5.55.0-10.050-100to 200


SandMaterials, percent by weightPropertiesCasting
Green470-10095.04. – 7.060-70to 2000
75.0 2000
470-10080.04.01. 015.0
The sand mixes given in the preceding tables are given only as a guide. The properties obtained with the all-purpose sands aboard ship will probably vary somewhat from those listed.


Cores used aboard repair ships are usually baked sand cores. Other types (such as green sand cores) have limited use and are not discussed here. Baked sand cores should have the following properties:

1. Hold their shape before and during the baking period.

2. Bake rapidly and thoroughly.3. Produce as little gas as possible when molten metal comes in contact with the core.

4. Be sufficiently permeable to permit the easy escape of gases formed during pouring.

5. Have hardness sufficient to resist the eroding action of flowing molten metal.

6. Have surface properties which will prevent metal penetration.


7. Be resistant to the heat contained in the metal at its pouring temperature.8. Have hot strength that is sufficient to withstand the weight of the molten metal at the pouring temperature and during the beginning stages of solidification.

9. Have good collapsibility so the core won’t cause hot tears or cracks in the casting.

10. Absorb a minimum amount of moisture if the mold is required to stand a considerable period of time before pouring. This also holds true if storage of cores is necessary.

11. Retain its strength properties during storage and withstand breakage during handling.


In addition to the special properties listed in the preceding section, the properties discussed for molding sands in the section “Sand Properties,” also apply to core sands.

There are three major factors which influence the properties of cores. They are (1) baking time and temperature, (2) type of core binder, and (3) collapsibility.

Baking Time and Temperature. The best combination of baking time and temperature varies with: (1) the type of binder used, (2) the ratio of oil to sand, and (3) the type of core ovens used. Figure 61 shows the dependence of the baked strength on baking time and temperature. It will be noted that the same strength was attained in one hour when baking at 450°F., as was attained in six hours at a baking temperature of 300°F. It is always good practice to make a series of tests on the effect of baking time and temperature on the baked strength of cores before using a new core mix. Such information will provide the shortest baking time to obtain a given strength for that mix. This type of investigation will also provide information on the baking characteristics of a core oven.

In the baking of oil sand cores, two things occur. First, the moisture is driven off. Following this, the temperature rises, causing drying and partial oxidation of the oil. In this way, the strength of the core is developed.

For proper baking of oil-sand cores, a uniform temperature is desired. This temperature should not be over 500°F. nor under 375°F. If linseed-oil cores are baked at a moderate temperature of 375°F. or 400°F., they will be quite strong. If the same cores are baked quickly at

500°F., they will be much weaker. Continuing the baking of the cores to the point where the bonding material decomposes must be avoided, as this causes the cores to lose strength.The size of the core must be considered in drying. The outer surface of a core will bake readily and will be the first part to develop maximum strength. If the temperature is maintained, the inside will continue to bake until it finally reaches maximum strength, but by that time, the outer surface of a large core may be overbaked and low in strength. The tendency for this to happen in large cores can be partly overcome by filling the center of the core with highly permeable material with a low moisture and bond content, by the use of well-perforated core plates, and by using low baking temperatures. It is not only a matter of heating the center of the core, but also of supplying it with oxygen. Thus, there is need for free circulation of air around and through the core while baking.

The most skillful and careful preparation of metal and mold can easily be canceled by careless technique, and the necessity for proper baking cannot be overemphasized. If cores are not properly baked, the following is likely to happen to the casting:

1. Excessive stress, possibly cracks, caused by the core continuing to bake from the heat of the metal, thus increasing in strength at the time the metal is freezing and contracting.2. Unsoundness caused by core gases not baked out.

3. Entrapped dirt due to eroded or spalled sand caused by low strength in the core.

When overbaked, the loss of strength of the core results in excessive breakage in handling or during casting, and cutting or eroding of the core surface.

To establish a full appreciation of the problems of drying cores, a series of 3-inch, 5-inch, and 8-inch cube cores should be made without rods and baked at temperatures of 400°F., 425°F., 450°F., 475°F., and 500°F. for varying predetermined times. After being taken out of the oven and cooled, they should be cut open with a saw to determine the extent to which they are baked. This simple test will aid in determining the proper times and temperatures to use for various cores in a given oven and under given atmospheric conditions.

Practice is necessary to determine accurately when a core is baked properly. A practical method is to observe the color of the


core. When it has turned a uniform nut brown, it is usually properly baked. A lighter color indicates insufficient baking, and a darker color indicates overbaking.Type of Core Binder. The type of core binder is important from the viewpoint of gas-generating properties as well as the strength the binder will develop. Figure 62 shows the volume of core gas generated from a linseed-oil compound and an oil-pitch mixture. The volume generated by both is the same for the first minute, but then the generation of gas from linseed oil decreases rapidly, while the generation of gas by the oil-pitch mixture decreases at a much lower rate. A core oil having gas-generating characteristics similar to those of linseed oil is preferred, since core gas is generated for a much shorter length of time and the possibility of defects due to core gas is lessened.

Combinations of several binders can be used to obtain a better overall combination of green strength, baked strength, and hot strength than can be obtained with the individual binders. For example, in a sand mixture containing core oil and cereal binder, the cereal binder contributes most of the green strength, whereas, the core oil contributes most of the baked strength. This is the reason for using combinations of cereal binder with oil binders. The strength attained from a cereal-oil combination is shown in figure 63. Notice that the strength obtained by the combination is higher than the total strengths of the individual binders.

Collapsibility. The sand-testing equipment used aboard ship does not permit the high-temperature testing of cores for collapsibility. A rule-of-thumb practice must be followed in determining this property. Close observation must be made in shaking out a casting to determine if the core mix had good collapsibility. A core that is still very hard during shakeout is said to lack collapsibility. If a crack should later be observed in the cored area of the casting, the core sand mixture definitely is too strong at high temperatures and the sand mix should be corrected. One remedy is to add about 2 percent of wood flour to the mixture.


Standard Materials. Core sand mixtures are made from clean, dry silica sands and various binders. The fineness of the sand is determined by the size of the core and the metal being poured. One important point in the mixing of core sand mixtures is to have the sand dry before any materials are added.

The materials used for binders are primarily corn flour, dextrine, raw linseed oil, and commercial core oils. Corn flour

and dextrine are cereal binders. Dextrine greatly increases the strength of baked cores and is used in small amounts with other binders. Dextrine-bonded cores have the disadvantage that they absorb moisture very easily and, therefore, should not be stored for any length of time before being used. Corn flour is used to give the core green strength and hold it together until it is baked. The cereal binders are used in combination with core oil to produce the desired strength. They are rarely used by themselves.Cereal binders have the following advantages that make them very useful binder materials: (1) good green strength, (2) good dry bond, (3) effective in angular sand, (4) core oil is not absorbed as in naturally bonded sands, (5) quick drying, and (6) fast and complete burn out. Core oils are used to provide a hard strong core after baking. They have the following advantages over other types of binders: (1) ability to coat the individual sand grains evenly with a reasonable amount of mixing, (2) generate a small amount of smoke and gas, (3) work clean in the core boxes, and (4) give cores good strength.

Substitute Materials. Aboard repair ships, the situation may arise where they standard core materials are unavailable. In such cases, substitute materials must be used. Substitute materials should be used only as an emergency measure. Molasses and pitch are two materials which can be obtained easily for use as core materials. Molasses should be mixed with water to form a thin solution known as “molasses water.” In this condition, it is added to the core mix as part of the temper water during the mulling operation. Pitch is seldom used alone. Used with dextrine, it imparts good strength to a core mix. Sea coal in small amounts is used with pitch to prevent the pitch from rehardening after it has cooled from the high temperatures caused by the molten metal.

If new washed silica sand is not available, reclaimed backing sand may be used for facing, if properly bonded. Some beach or dune sands, relatively free from crustaceous matter and feldspar, some fine building sands, and some natural sand deposits containing clay may be used. If bentonite is not available, portland cement, fireclay, or some natural clays may be used. The corn flour maybe replaced with ordinary wheat flour. Sugar or molasses will take the place of dextrine. Wherever substitutes must be used, the amount of organic materials and clay should be kept to a minimum and the amount of good clean sand grains to a maximum.

Other Core Materials. Silica flour and wood flour are added to core mixtures to get special properties. Silica flour is usually


added to prevent metal penetration and erosion of cores by molten metal. Silica flour must be used carefully and not used in excess. Excessive use may lead to hot tears because of too high a hot strength. Wood flour is not a binder but a filler material. Its use is that of softening or weakening a core so that it has better collapsibility.MIXING

Core sands should be mixed in a muller or some other type of mechanical mixer to obtain the maximum properties from the various binders. Many of the binders are added in very small amounts, and only a thorough mixing operation can distribute the binder uniformly throughout the sand. Manual mixing with a shovel requires the addition of much more binder to obtain the desired properties, and results are not consistent. Manual mixing of core sands is to be discouraged.

Mulling time has the same effect on core sand as on molding sand, as is shown in figure

57. The proper mulling time should be determined for each mix used. In the mixing of core sands, the additions are made in the following order with the mixer running: (1) sand, (2) dry ingredients, (3) run the mixture dry for a short time, (4) add liquids, and (5) continue to mix for the desired period of time. Laboratory tests have shown that if the core oil is added to the sand before the water and mixed for a short period of time, more consistent core properties will be obtained. If cereal binder is used, the batch should not be mixed too long before adding the liquids. Excessive mixing of the sand with cereal binders without the liquids will cause the batch to become sticky, and a longer length of time will be needed to bring the core mix to its proper condition.


The following tables suggest various representative core mixes. They are given primarily as a guide to obtain good core mixes for work aboard repair ships.



SandMaterials, percent by weightUse in


SandMaterials, percent by weightUse in
New550-7088. to
1000 lb


SandMaterials, percent by weightUse in



SandMaterials, percent by weightUse in
New550-7095. lbs
and over


SandMaterials, percent by weightUse in


CORE PASTE AND FILLERA very good core paste for use in joining core sections may be made from 3 percent bentonite, 6 percent dextrine, and 91 percent silica flour. The ingredients should be mixed dry, and water added to produce a mixture the consistency of soft putty.

A filler to seal the cracks between parts of the core may be made from 3 percent bentonite, 3 percent dextrine, and 94 percent silica flour. The ingredients are mixed dry, then water is added to make a mixture with the consistency of stiff putty. This material is pressed into the cracks between the core sections to prevent metal penetration.


Core and mold washes may be needed in some cases to prevent erosion of the sand and metal penetration into the sand. The following mix contains the same bonding material as the molding sands, with silica flour replacing the sand, and with sodium benzoate added to prevent the mixture from becoming sour.

Weight percent
Silica flour64.0
Sodium benzoate0.2

The dry material should be mixed thoroughly in a closed container. The water is then added, and the mixture stirred thoroughly. The mixture is sprayed onto the green core like paint and then baked, or it may be brushed on the dried core or mold. It must be allowed to dry thoroughly in air or be baked in an oven, and should be used only when absolutely necessary.

In most cases, the green or air-dried sand mixtures will produce excellent casting surfaces without use of the wash.In brass castings, where erosion and penetration are problems, a core wash made from a silica base is satisfactory. A plumbago wash is useful for bronze castings. A core wash for use with high-lead alloys and phosphor bronzes, may be made from a paste of plumbago and molasses water. Such a treatment should be followed by a thin coating of the regular core wash.


The testing of foundry sands should not be a series of tests for obtaining a great deal of meaningless information. Regular sand testing along with records of the results is the one way of establishing the cause of casting defects due to sand. Regular sand testing results in a day-to-day record of sand properties, and indicates to the molder how the sand properties behave. Proper interpretation of the results of sand tests permits the molder to make corrections to the sand before it is rammed up in molds, thereby not only saving time but also preventing casting losses.


A sample of sand, at least one quart in volume, should be taken from various sections of the sand heap and from a depth of at least six inches. The sand should be riddled through a 1/4 inch mesh riddle or the size of riddle used in the foundry. The same riddle size should be used for all sand tests.

Enough tempered sand is weighed out to make a rammed sample 2 inches high. The proper amount of sand can be determined by trial and error. The sand is then placed in the


specimen tube, which rests on the specimen-container pedestal. The tube with the pedestal is then placed under the rammer, as shown in figure 64. Care should be taken to keep the tube upright so as not to lose any of the sand.The rammer is lowered gently into the specimen tube until the rammer is supported by the sand. The rammer is raised slowly by the cam to the full 2-inch height, and permitted to fall. This is repeated until a total of three rams have been applied. The top of the rammer rod should be between the 1/32 inch tolerance marks for control work. If the end of the rammer rod is not within the tolerance, the specimen must be discarded and a new test specimen made. If the specimen is of the correct height, the rammer rod should be lifted carefully to clear the specimen tube, and the specimen tube removed from the pedestal. The type of rammer supplied for shipboard use is shown in figure 64.


The permeability of foundry sand is determined by measuring the rate of flow of air under a standard pressure through a standard specimen 2 inches high by 2 inches in diameter. The specimen is prepared as described in the previous section, “Test-Specimen Preparation.” The equipment for the permeability determination is shown in figure 65. The sand specimen, still in the tube, is placed in the mercury well with the sand specimen in the top position. The air chamber is raised to its proper position, released, and permitted to drop. When the water column in the manometer becomes steady, the permeability scale, which is on the curved part of the indicator, is rotated until the edge of the scale is opposite the top of the water column. The reading on the scale at this point is the permeability for control purposes. It is good practice to take permeability readings on three different specimens from the same lot of sand and to average the readings. (The test, as described, measures green permeability.)


Green compressive strength is the property most useful in foundry sand control in repair ship foundries. The specimen is prepared as described under “Test-Specimen Preparation,” and then stripped from the tube with the stripping post. The specimen used for permeability test is suitable if not damaged in previous test. The sand specimen is placed between the compression heads on the lower part of the test apparatus shown in figure 66. The face of the sand specimen which was the top face when the specimen was rammed should be placed against the right-hand compression head.

Care should be taken to seat the specimen carefully in the compression head. A small magnetic rider is placed on the scale against the compression head, and the arm is raised by the motor-driven mechanism or by hand. If hand operation is used, care must be taken to maintain a slow and uniform speed of operation, since the rate of motion of the arm affects the test results. When the specimen breaks, the motor automatically reverses itself and returns to its bottom position. With hand operation, the arm is returned manually when the specimen is seen to break. The magnetic rider will remain at the position attained by the arm when the break occurred. The green compressive strength is read from the back of the rider on the appropriate scale. The testing equipment must be maintained in good operating condition at all times, and sand from the broken specimens must be completely removed from the equipment after each test. Pay special attention to keeping grains of sand and dirt out of the bearings. Use only dry lubricants, such as graphite, on sand-testing equipment.DRY STRENGTH

For test specimen for determining dry strength is prepared as described in the section “Test-Specimen Preparation.” After the specimen is made and removed from the stripping post, it should be placed on a flat rigid plate and dried for at least two hours. Drying is done at a temperature between 220°F. and 230°F. The specimen is removed from the oven and cooled to room temperature in a container that will prevent moisture pickup by the dry specimen. The specimen is then tested in the same manner as described for obtaining green strength in the section “Green Strength.” The specimen should be tested as soon as it has cooled to room temperature, and should not be permitted to stand for any appreciable length of time before testing.


The moisture content of molding sands is determined with the apparatus shown in figure 67. A representative 50-gram sample of tempered sand is placed in the -special pan, which is then placed in the holder. The timer switch is set for 3 minutes. Setting the timer automatically starts the dryer, which runs for the set time interval. After drying is complete, the pan is removed and weighed. The loss in weight multiplied by two is the percent of moisture in the tempered sand.


A representative sample is obtained from the sand which is to be tested for clay content. The sand is then thoroughly dried and a 50-gram


sample taken. The sample is placed in the jar shown in figure 68 with 475 cc of distilled water and 25 cc of standard sodium hydroxide solution. The standard sodium hydroxide solution is made by dissolving 30 grams of sodium hydroxide in distilled water and diluting to 1000 cc. The jar containing the sand sample and solution is assembled with the stirrer and stirred for five minutes. (The assembled sand-washing equipment is shown in figure 69.) The stirrer is then removed from the jar and any adhering sand washed into the jar. The jar is then filled with distilled water to a depth of six inches from the bottom of the jar. The contents of the jar should be well stirred by hand and then allowed to settle for 10 minutes. The water is then siphoned off to a depth of 1 inch. Distilled water is then added again to a depth of 6 inches, the solution agitated and allowed to settle for 10 minutes a second time. The water is siphoned off a second time to a depth of 1 inch. Water is added a third time, the solution agitated and permitted to settle for a 5-minute period, after which the water is siphoned off again. Distilled water is added to a depth of 6 inches, the solution agitated, permitted to settle for 5 minutes, siphoned off to a depth of 1 inch, and the procedure repeated until the solution is clear after the 5-minute settling period. The glass cylinder is then removed from the base of the jar so as to leave the sand in the base. The sand is dried thoroughly in the base. The dry sand is weighed. The weight lost multiplied by two is the percent of AFS clay in the sand.If clay determinations are made on used sand, the result will not be a true clay content, since sea coal and other additives will be removed along with the clay. The determination would then give a false value.


Grain fineness is expressed as the grain fineness number and is used to represent the averagegrain size of a sand. The number is useful in comparing sands. Grain fineness numbers, however, do not tell the molder the distributionof grain sizes, and the distribution does affect the permeability and potential strength of the sands. Two sands may have the same grain fineness number and still differ widely in permeability, due to differences in their grain-size distribution. Clay content and the shape of sand grains also influence the sand properties, and may differ in sands having the same grain fineness number.

The sample for determining the grain fineness number should be washed of all clay as described under “Clay Content,” and thoroughly dried. A 50-gram sample of the sand is then screened through a series of standard sieves. The sand remaining on each screen should be

carefully weighed and recorded. The grain fineness number is then calculated as shown in table 18.


Size of Sample: 50 grams
Clay Content: 5.9 grams – 11.8 percent
Sand Grains: 44.1 grams – 88.2 percent

ScreenAmount Retained
on Screen
Grain Fineness Number =
Total Product / Total Percent of Retained Grain =
15243/88.2 = 173


A better method for comparing sands is to compare them by the actual amounts retained on each screen. A method for plotting this type of information is shown in figure 70. Two sands have been plotted for grain distribution. Notice that although both sands have the same grain fineness number, the size distribution of the grains is different.


The need for proper sand control through the use of sand-test equipment cannot be stressed too strongly. There is only one way to determine the properties of molding sands and core sands, and that is to make tests. Day-by-day testing of foundry sands provides the molder with information which enables him to keep the molding sand in proper condition. The recording of these test results, along with appropriate comments as to the type of castings made and any defects which may occur, can help the molder to determine the causes of casting defects, and point the way toward corrective measures.

As a summary to the chapter on foundry sand, the various factors affecting sand are tabulated below with the results produced in the sand.

452605 0-58-5


Molding Sand

Grain finenessSand too finePermeability reduced, green strength increased. Possible defects: blisters, pinholes, blowholes, misruns, and scabs.
Sand too coarsePermeability increased, green strength decreased. Possible defects: rough casting surface and metal penetration.
BinderToo much binderAccompanied by too little moisture, results in decreasing permeability, increasing green strength. Possible defects: hot cracks, tears, and scabs.
Too little binderLow green strength and high permeability. Possible defects: drops, cuts, washes, dirt, and stickers.
Moisture contentToo highPermeability and green strength decreased. Possible defects: blows, scabs, cuts, washes, pin holes, rat tails, and metal penetration.
Too lowPermeability and green strength too low. Possible defects: drops, cuts, washes, and dirty castings.


Figure 45. Permeability as affected by the grain size of sand.
Figure 45. Permeability as affected by the grain size of sand.

Figure 46. Permeability as affected by sand fineness and moisture.
Figure 46. Permeability as affected by sand fineness and moisture.

Figure 47. The effect of sand grain shape on permeability.
Figure 47. The effect of sand grain shape on permeability.

Figure 48. Permeability as affected by the amount of binder.
Figure 48. Permeability as affected by the amount of binder.

Figure 49. The effect of bentonite and fireclay on permeability.
Figure 49. The effect of bentonite and fireclay on permeability.

Figure 50. Green strength as affected by the fineness of sand.
Figure 50. Green strength as affected by the fineness of sand.


Figure 51. Green strengths of sands with varying fineness numbers.
Figure 51. Green strengths of sands with varying fineness numbers.

Figure 52. Green strength as affected by the shape of sand grains.
Figure 52. Green strength as affected by the shape of sand grains.

Figure 53. Green strength as affected by moisture and varying bentonite contents.
Figure 53. Green strength as affected by moisture and varying bentonite contents.

Figure 54. The effect of bentonite and fireclay on green strength of foundry sand.
Figure 54. The effect of bentonite and fireclay on green strength of foundry sand.

Figure 55. The effect of bentonite on sands with various moisture contents.
Figure 55. The effect of bentonite on sands with various moisture contents.

Figure 56. The effect of western and southern bentonite on green strength and dry strength.
Figure 56. The effect of western and southern bentonite on green strength and dry strength.


Figure 57. Green strength as affected by mulling time.
Figure 57. Green strength as affected by mulling time.

Figure 58. Relationship between moisture content bentonite content, green compressive strength, and permeability for an all-purpose sand of 63 AFS fineness number.
Figure 58. Relationship between moisture content bentonite content, green compressive strength, and permeability for an all-purpose sand of 63 AFS fineness number.


Figure 59. Relationship between moisture content, bentonite content, green compressive strength, and dry strength for an all-purpose sand of 63 AFS fineness number.
Figure 59. Relationship between moisture content, bentonite content, green compressive strength, and dry strength for an all-purpose sand of 63 AFS fineness number.

Figure 60. General green compressive strengths for sands of different grain class numbers.
Figure 60. General green compressive strengths for sands of different grain class numbers.

Figure 61. Strength of baked cores as affected by baking time and baking temperatures.
Figure 61. Strength of baked cores as affected by baking time and baking temperatures.


Figure 62. Core gas generated by two different core binders.
Figure 62. Core gas generated by two different core binders.

Figure 63. The effect of single binders and combined binders on the baked strength of cores.
Figure 63. The effect of single binders and combined binders on the baked strength of cores.

Figure 64. Rammer used for test specimen preparation.
Figure 64. Rammer used for test specimen preparation.


Figure 65. Permeability test equipment.
Figure 65. Permeability test equipment.

Figure 66. Strength testing equipment.
Figure 66. Strength testing equipment.

Figure 67. Equipment for drying sand specimens for moisture determination.
Figure 67. Equipment for drying sand specimens for moisture determination.

Figure 68. Jar and stirrer for washing sand.
Figure 68. Jar and stirrer for washing sand.


Figure 69. Sand washing equipment assembled.
Figure 69. Sand washing equipment assembled.

Figure 70. The difference in sand grain distribution for two foundry sands having the same grain-fineness number.
Figure 70. The difference in sand grain distribution for two foundry sands having the same grain-fineness number.