ASM Metals Handbook - Desk Edition (ASM_ 1998) WW part 11 ppt

Molten metal is then tapped from the melting furnace see the article "Melting Methods" into a ladle for pouring into the mold cavity, where it is allowed to solidify within the space def

Leaching Processes

Leaching is a separation process that uses aqueous solutions A suitable aqueous environment is selected which can decompose the mineral containing the valuable metal The objectives of leaching are:

• Production of a compound for further processing by pyrometallurgical techniques

• Production of a metal from impure metal or metal compounds that have been prepared by a pyrometallurgical process

• Direct production of a metal from an ore or concentrate

Selection of a particular objective depends on economic factors and the involved thermodynamic and kinetic conditions

of the system The theoretical possibility is limited by thermodynamic constraints, whereas kinetic constraints relate to the overall time required and affect the reactor size and design

The thermodynamics of leaching are concerned with the ability to decompose a particular compound so that it will selectively dissolve and become stable in the aqueous solution used for leaching Proper calculations can predict the maximum amount of mineral that can be leached until the system reaches equilibrium The rate of dissociation and dissolution is kinetically controlled Figures 9 and 10 are schematic diagrams of a mineral/water interface on a microscopic scale The concentration of the active chemical in the bulk of the solution is greater than the concentration near the surface of the mineral This occurs because the active chemical is removed from solution by reaction at the mineral surface Figure 9 shows that the rate at which the reaction or decomposition of the mineral will occur is determined by: (1) the diffusion of reactant, R, from the bulk of the solution to the surface of the mineral; (2) the reaction

of the reagent with the surface to form a soluble species; and finally, (3) the diffusion of the product metal species, M, away from the surface In Fig 10, diffusion of R through the porous product layer and diffusion of M through the product layer are also possible (Ref 15)

Fig 9 Schematic diagram of a mineral surface showing complete dissolution in water

Fig 10 Schematic diagram of a mineral surface showing decomposition in water and generation of a porous

layer of residue on the surface

The leaching of a compound in an aqueous environment can take several forms Simple dissolution reactions in water, acid, and alkali can be represented by Eq 6, 7, and 8 respectively:

CuSO4(s) + nH2O CuSO4·nH2O(aq) (Eq 6)

The leaching method used depends on the physical condition of the ore and the inherent mineralogy Simple leaching processes include in situ leaching, heap leaching, and agitation leaching In situ leaching refers to mineral dissolution with the ore "in place" underground In situ leaching is usually done in worked-out stops of high-grade mines, support pillars left behind after mining, or low-grade deposits This type of treatment requires the surrounding rock to be tight and impermeable to solution flow in order to contain the leaching solution This process is very time consuming but has low treatment costs, low equipment requirements, low capital costs, and the ability to treat low-grade ores

The heap-leaching process is similar to in situ leaching since it does not require extensive leaching equipment, such

as tanks, slurry pumps, and thickeners; hence, heap leaching requires low capital and maintenance costs Therefore, the principal feed materials for heap leaching are low-grade ores, ores not amenable to flotation, and discarded waste rock from previous processing with metal values below milling grade The leachant is sprayed or pumped over a heap of ore

As it percolates through the heap, it dissolves the desirable compounds from the ore An agglomeration step prior to leaching of the fine ore has shown significant improvement in precious metals recovery (Ref 17)

Agitation leaching is usually used with well-disseminated, fine-grained, high-grade ores This requires extensive crushing and grinding before leaching in order to expose the solution to the minerals where agitation improves the process kinetics Because of the extensive amount of equipment required for agitation leaching, recoveries of over 90% and short residence times are requirements of the process Agitation is accomplished by either bubble action using compressed air

or by mechanical agitation using impellers Pachuka tanks are commonly used in air-agitated leaching processes (Ref 18) When high-intensity agitation is required mechanical agitation with impellers is usually used Marine propellers produce

an axial flow, paddles cause a tangential flow, and turbine impellers produce a combined radial and axial flow pattern (Ref 19)

Pressure leaching is similar to agitation leaching except that the process is done at elevated pressures and temperatures Agitation leaching under normal pressure is limited to 100 °C (212 °F) As pressure on the solution is raised, the boiling point of water can be elevated Thus, higher leaching temperatures can be employed by increasing the pressure Therefore, the main objective of pressure leaching is to enhance the kinetics of metal dissolution by permitting higher operating temperatures and by increasing the solubility of gaseous species that may take part in the leaching reaction The oxidation of sulfides in aqueous solutions exemplifies the need for increased gas solubility At atmospheric pressure metal sulfides are insoluble even in strong acidic solutions Increasing the temperature and pressure of the system increases the solubility of oxygen in solution The increase in oxygen solubility and temperature causes rapid oxidation of some metal sulfides to sulfates making them soluble in acid solutions This ability to achieve oxidation in the leaching step by increasing the pressure can eliminate the need for pretreatment steps, such as oxidation roasting of sulfide ores

Pressure leaching is generally done in a stainless steel or titanium autoclave for high strength and corrosion resistance at higher temperature and pressure Linings of glass, lead, or refractory brick may be used under severe corrosion conditions Most autoclaves are equipped with agitators for mixing Both vertical and horizontal autoclaves are used in hydrometallurgy The Sherritt-Gordon process (Fig 11) uses pressure leaching during the production of nickel and cobalt metals (Ref 20)

Fig 11 Sherritt-Gordon process flow diagram for nickel and cobalt production

Solution Purification

Leaching with strong solutions usually produces an aqueous stream containing the desired metal values as well as some impurities due to the complexity of the mineral ore Processes for the purification of the leaching solution prior to metal recovery include precipitation, solvent extraction, and ion exchange

Precipitation involves the removal of ionic species from solution as compounds Precipitation is accomplished by making adjustments to the solution which cause formation of compounds that are no longer soluble in the solution The concept of the solubility product is used to predict, and to perform calculations concerning, precipitation from solution The solubility product for a given compound is defined as the product of the concentrations of cation and anion of the compound of interest, each raised to the power of its proportion in the compound For example, the solubility product for

a hypothetical compound MX would be

Changes in the pH of solutions can also be used to precipitate compounds (Ref 16) As the pH of a solution is increased and the solution becomes more basic, the hydroxide ion (OH-) concentration increases and solid hydroxides precipitate Iron, copper, cobalt, and nickel are precipitated selectively as hydroxides in solutions by raising the pH with milk of lime

to 2.5, 5.8, 8.3, and 9.4, respectively Precipitation can be used in a process to remove the impurities as well as to concentrate metal values in the form of a compound Desired metal values can also be removed from an impure solution and concentrated in a solid compound The recovery of sulfides of nickel, copper, lead, and zinc from leaching solutions

as precipitates requires further purification, but provides a low-cost treatment method with very low concentrations of metal values in the leach solution

Solvent extraction is a chemical process used to purify and concentrate a given species from aqueous solution This is accomplished by recycling an organic solution, which selectively exchanges the metal species of interest between an impure aqueous feed solution and a pure fresh aqueous solution The process relies on the immiscibilities of organic and aqueous solutions as well as the stabilities of the metal species in them Purification is then achieved by extracting a metal species from the impure aqueous solution to the organic solution and then stripping the metal species from the organic solution back to fresh aqueous solution A typical solvent-extraction flow sheet is shown in Fig 12 The organic must be selective to the species being purified, and the reaction must be reversible so that the metal species can be transferred from impure to fresh aqueous solutions via the organic phase Solvent-extraction chemistry can roughly be separated into two types of reactions: solvation and exchange reactions

Fig 12 Typical flow diagram for solvent extraction

Solvation involves transfer of neutral molecular species between aqueous and organic In the transfer from the aqueous to the organic phase, the neutral species simply dissolves in the organic solution The organic phases used for solvation can

be alcohols, ethers, esters, ketones or phosphorus-containing compounds such as trialkyl phosphates and trialkyl phosphine oxides Exchange reactions involve the formation of specific bonds between metal species and active compounds in the organic phase The metal species forms an organic salt with the active organic compounds, which then dissolves in the organic phase Exchange reactions can be cationic or anionic depending on the system involved (Ref 21)

The organic solutions typically used for exchange reactions are made up of a carrier, an extractant, and a modifier The carrier or diluent is the inert organic that makes up approximately 90% of the solution and acts as a vehicle for carrying the active extractant Common diluents are kerosine, and naphthene The extractant or active agent is the compound that contains the functional group capable of chemically reacting with the particular metal species in the aqueous phase The modifier, which is usually an alcohol, is added because it increases extracting power, increases selectivity, improves phase separation, and prevents formation of solid organic compounds Solvent extraction is typically performed in a combination of mixer-settler units to allow the countercurrent flow of organic and aqueous solutions from stage to stage

Ion exchange is accomplished by interchange of metal ions between aqueous solutions and a solid, insoluble resin The chemistry of the ion-exchange process is similar to the solvent extraction systems and is sometimes used as a substitute for solvent extraction to avoid problems of emulsion formation and solvent loss due to entrainment The ion-exchange process involves adsorption followed by elution Adsorption is the removal of metal ionic species from an aqueous solution when that solution is passed through a bed of ion-exchange resins Elution is the recovery of the metal ionic species in fresh solution by passing a suitable fresh solution through the previously loaded resins

Ion-exchange resins are classified as cationic resins and anionic resins Cationic resins exchange cationic species and are made of strong acid or weak acid groups and exchange H+ ions Anionic resins are strong bases, such as quaternary ammonium group bases, and weak bases, such as secondary or tertiary amine group bases In most cases, chloride ions exchange with anions in solution (Ref 22) The important properties of ion-exchange resins are capacity, selectivity, and mechanical properties The capacity of a particular resin is the amount of a specific inorganic group that the resin will hold per unit weight or volume The affinity of a resin for different ions in solution varies This selectivity of one ion over the other is described as a distribution coefficient, K, where K = (% equivalent of ions in resin)/(% equivalent of ions in solution)

For a particular resin, the selectivity coefficient varies with the species of interest making it possible to purify a particular metal ion from a complex solution Because ion-exchange resins are used over and over as the transfer media for purification, they require good mechanical properties Resins must be durable and resistant to breakage, must have low chemical degradation, and must be insoluble in aqueous solutions The ion-exchange equipment involves fairly high capital costs and a large plant area due to the large amounts of in-process material However, properly run ion-exchange facilities result in up to 99% efficiency during normal operation

Metal Deposition

Once an ore concentrate has been leached and purified, the metal of interest must be recovered from solution Three common techniques of metal reduction from aqueous solution are cementation, hydrogen reduction, and carbon adsorption

Cementation, or metallic replacement, is a classical process for recovering metals from aqueous solution Cementation

is essentially the precipitation or discharge of a noble or less-reactive metal in favor of a more-reactive metal The basic reaction between the two metals is electrochemical in nature and can be represented by the reaction in the following equation for the reduction of cadmium by zinc metal:

The above electrochemical reaction can be separated into two half-reactions:

+ 2e- Cd(s) (Eq 13b)

When the active zinc metal is added to a solution containing relatively noble cadmium ions, reduction takes place in microcells producing cadmium metal Since the two half-cell reactions require transfer of electrons it is essential that the solid is a conductor of electricity The tendency of one metal to displace or reduce another metal from solution is based on the electromotive series of metals (Ref 23) In general, when two metals are considered, the one that is more electropositive will tend to reduce a less-electropositive metal from solution Also, the greater the difference in potential between the two metals is, the greater is the driving force for the reaction However, even a small difference in potential results in an extensive degree of reduction Reduction of copper ions from solution by metallic iron using cementation is industrially practiced:

Copper sulfate solution is fed through open launders containing steel scrap where the displacement reaction occurs and produces a very pure copper that can be recovered in the bottom of the launders Cementation cones are commonly used now for the deposition of metals (Ref 24) The rate at which cementation reactions occur depends on initial concentrations, temperature, agitation, polarization characteristics of different metals, and addition agents (Ref 25)

Gaseous reduction of metals from aqueous solution can be done with reducing gases, such as hydrogen, carbon monoxide, and sulfur dioxide Hydrogen is the most widely used because it is relatively inexpensive The reaction products from carbon monoxide and sulfur dioxide have to be further treated after the reduction step The use of reducing gases also involves a replacement reaction:

The tendency of the above reaction to produce copper can be increased by increasing the pressure of hydrogen gas since the reaction moves to the right A similar reaction for nickel at room temperature and atmospheric pressure does not occur However, if the reaction is carried out in an autoclave that permits high hydrogen pressures and temperatures, nickel can be reduced from the solution Although the mechanism by which nickel is reduced from ammoniacal solutions

is more complex, nickel is reduced commercially at high temperature and pressure in the presence of ammonia (Ref 20)

Carbon adsorption, or reduction of metals from aqueous solution, is used almost exclusively to recover the noble precious metals, such as gold and silver This process is based on the principle that these metals can be reduced out of solution by solid carbon at low temperatures and deposited in metallic form on the carbon In a typical carbon-adsorption process, the metal leaching solution is fed to carbon columns and the metal is almost completely removed from the solution by adsorption on the solid carbon After the carbon is loaded, it is removed from the circuit and the metal is stripped away The carbon can also be added to the leaching liquor and agitated without the requirement of carbon columns, known as the "carbon in pulp" process Desorption or stripping is done by passing a hot, caustic stripping solution over the column Stripping is followed by electrowinning from solution to produce a very pure gold or silver product A typical flow sheet, including carbon adsorption, is shown in Fig 13 The loaded carbon can also be burned, which leaves a gold- or silver-rich ash

Fig 13 Heap-leaching charcoal-adsorption process for gold ores low in silver

References cited in this section

15 M.E Wadsworth and J.D Miller, Hydrometallurgical Processes, Rate Processes of Extractive Metallurgy,

H.Y Sohn and M.E Wadsworth, Ed., Plenum Press, 1979, p 133-241

16 T Rosenquist, Principles of Extractive Metallurgy, 2nd ed., McGraw-Hill, 1983, p 438-441

17 "Agglomeration-Heap Leaching Operations in the Precious Metals Industry: U.S Bureau of Mines Information Circular 8945," 1983

18 A.G.W Lamont, Air Agitation and Pachuca Tanks, Can J Chem Eng., Aug 1958, p 153

19 J.Y Oldshue, Fluid Mixing Technology, Chemical Engineering Division, McGraw-Hill, 1983

20 J.R Boldt, Jr and P Queneau, The Winning of Nickel, International Nickel, New York, 1967

21 P.J Bailes, C Hanson, and M Hughes, Liquid-Liquid Extraction-Metals, Chem Eng (UK), Aug 30, 1976

22 K Dorffner, Ion Exchangers, Properties, and Applications, Ann Arbor Science Publishing, 1973

23 D Jones, Principles and Prevention of Corrosion, 2nd ed., Prentice Hall, 1996, p 43

24 R.H Spedden, E.E Malouf, and J.D Prater, Cone-Type Precipitates for Improved Copper Recovery, J Met., Vol 18, 1966, p 1137-1141

25 J.D Miller, Cementation, Rate Processes of Extractive Metallurgy, H.Y Sohn and M.E Wadsworth, Ed.,

Plenum Press, 1979

Electrometallurgical Processes

Electrometallurgy, or electrolytic processing, deals with the production of metals from ions by application of electrical energy In contrast with electrochemical processing, where chemical reaction produces electricity as in batteries or corrosion, electrolytic processing uses electricity to perform chemical functions as in metal extraction Therefore, corrosion is also known as "extractive metallurgy in reverse." As has been discussed thus far, extraction usually implies

reduction of a compound, where as corrosion implies oxidation of a metal into a compound The term electrowinning is used to describe the recovery of metals from a solution or electrolyte on a negative electrode, or cathode, while the positive electrode, or anode, is inert to the ongoing reaction Electrorefining entails a purification process in which the anode is made of solid impure metal that actively dissolves in the electrolyte; deposition of pure metal occurs on the cathode These solutions could be generated during leaching and/or purification, as discussed in hydrometallurgy

Fused salt electrolysis is the production of metals which cannot be electrolyzed from aqueous solution due to their relative positions in the electromotive series and, therefore, are reduced from molten salts In other words, production of hydrogen from its oxide (water) is, energy-wise, more favorable than the production of metal from its compound Electrowinning and electrorefining concepts are equally applicable to molten salt electrolysis If the metal compound has high melting point and is difficult to melt, it may be dissolved in an inert carrier electrolyte (molten salt) which is more stable than the metal compound of interest The Hall-Héroult process for aluminum production essentially follows this scheme where alumina is dissolved in molten cryolite (sodium-aluminum fluoride)

Basic Concepts

The electromotive series is a listing of the standard half-cell potentials with respect to a reference electrode The reactions described in establishing an electromotive series are referred to as electrochemical reactions Electrochemical reactions involve oxidation (loss of electron) and reduction (gain of electron) and can be arranged in an electrochemical (galvanic) cell as shown in Fig 14 In general, the valence state of the metal is increased in oxidation and decreased in reduction The oxidation state of a pure metal atom is zero which gets positively ionized by losing electrons The electrons are picked up by another ionic species, which gets reduced in the solution If this other ionic species is another metal ion that

is reduced on the cathode, the process is known as cementation It should be memorized that the ions that move towards the anode are anions (negatively charged) and those that are attracted by the cathode are cations (positively charged) The ionic species in the solution (acidic) could be simply protons that can be reduced on the cathode as hydrogen gas

Fig 14 Electrochemical (galvanic) cell

The potentials on the electromotive series are the minimum theoretical volts required for depositing the metals from a molar solution saturated with its own ions at 25 °C (77 °F), and are measured against a standard reference electrode For the example of copper metal deposition using zinc metal (Eq 13), an electrochemical cell can be set up, as in Fig 14, by using a copper electrode as M in a solution of copper sulfate and a zinc electrode as M2 in a solution of zinc sulfate If the external circuit is short circuited, electrons will flow from the zinc electrode (anode) to the copper electrode (cathode) as zinc dissolves, which causes deposition of copper metal By placing a voltmeter between the cathode and the anode, the potential difference between the oxidation and reduction half-reactions can be measured Because measurement of potential requires measurement between the electrodes, an absolute potential of any half-reaction cannot be measured

The theoretical dissociation potential required for electrolysis of a compound is given by -cG/nF, where G is the Gibbs free energy change of the formation of the compound and is a function of temperature and activities of the ions, n is the number of electrons transferred in the oxidation-reduction reaction, and F is the Faraday Constant (96,486 C) The

theoretical current required for metal deposition from the solution is given by Faraday's law, which states that 1 equivalent weight of any metal can be deposited by passing 95,500 ampere-sec of charge (1 faraday) The product of the theoretical potential and current is the theoretical power required The current and energy efficiencies of an electrolytic cell may be measured as the percent ratios of these theoretical values and the actual amounts of potential applied and current consumed Efficient electrolytic cells are characterized by high current and energy efficiencies The theoretical

gram-dissociation potential must be exceeded to allow deposition of the metal of interest, but should be kept as low as possible for a higher energy efficiency Current, in general, is directly linked with the deposition rate, although higher current also implies higher ohmic losses and I2R heating

Electrowinning

The electrowinning arrangement usually recycles the acid between the leaching operation and the electrowinning operation for economic efficiency A simplified diagram of the leaching/electrowinning process is presented in Fig 15 For example, in the leaching step, the sulfuric acid is consumed as the metal dissolves, according to the reaction in the equation:

MO(s) + H2SO4(aq) H2O + MSO4(aq) (Eq 16)

Fig 15 Simplified block diagram showing the cyclical nature of the leaching/electrowinning process

The leaching solution, H2O + MSO4(aq), is transferred to the electrowinning step for deposition of the metal The leaching solution is fed into cells which consist of inert anodes (usually made of lead) and cathodes (which may be made of stainless steel, aluminum, or starter sheets of metal M), arranged as parallel alternating plates of cathodes and anodes In the electrowinning step, the power supply forces electrons to the cathode, and metal sulfate is reduced to produce the pure metal:

At the anode, oxygen is produced by decomposition of water and the regeneration of acid is accomplished by the reaction

in Eq 19, and recycled back to the leaching step (Eq 16):

H2O + O(g) + 2e- (Eq 18)

In this case, the potential required between cathode and anode in electrowinning is of the order of 1.25 to 1.75 V This potential is the combination of the decomposition potential defined earlier (the reversible potential between the cathode and the anode), the ohmic drops or IR losses due to the resistance of the electrolyte, connections and conductors, and polarization of the cell (Ref 23)

The decomposition potential is measured only when the cell reactions take place under conditions of infinitely low current

or net zero current In electrolysis, where current flows at a finite rate, additional voltage drops occur as a result of phenomena that take place near the electrode solution boundaries These extra voltages are called overvoltages and are caused by reactants not being supplied to electrodes as fast as products are removed (concentration overvoltage) or by

molecular phenomena (activation overvoltage) (Ref 23) The resistances of the electrical connections and the electrolyte cause generation of heat as current passes through the cell (Joule heating) The optimum solution temperature in most electrowinning plants is 30 to 45 °C (86 to 113 °F) Temperatures much higher than this can cause deterioration of the cathode deposit as well as increased corrosion of the anodes

Since the deposited metal must be removed to recycle the cathode, it is important that the metal be loosely bonded to the cathode For copper electrowinning, titanium, stainless steel, and copper have been used successfully Aluminum is used

to recover zinc and titanium, and stainless steel has been used for electrowinning manganese and cobalt Electrolytes for electrowinning may also be prepared by externally dissolving the metal in a solvent and pumping it into the cell

Electrorefining

The advantages of electrorefining over other refinement techniques are the high purity that can be obtained in the product and the recovery of secondary metals (which often are a series of precious metals) in a concentrated state The secondary metals are recovered from the concentrated form in a separate process Metals made by smelting of sulfides, such as copper, lead, and nickel, usually are refined by electrolysis The impurities that contaminate these metals are most often associated with sulfides themselves The precious metals gold, silver, platinum, and selenium, as well as iron, bismuth, arsenic, and antimony, are usually present The metal from the smelters, which is from 90 to 98.5% pure, is cast into anodes prior to electrorefining

The impure anodes are put into cells that contain an acid electrolyte of the anode metal ion Cathode sheets are placed between the anodes where the pure metal is deposited A dc power source is wired to the anodes and cathodes in a manner such that electrons are forced into the cathode sheets, causing reduction, and electrons are stripped from the anodes, causing oxidation At the anode, the oxidation reaction results in the dissolution of the metal by the reaction, M = M2+ +

2e-, and the electrons flow back through the power source to the cathode As the anode dissolves, the impurities either dissolve into the electrolyte with the metal or remain at the anode The metals that remain in the metallic state are more noble in the electromotive series than the metal that is being refined As the anode dissolves away, these noble metals fall

to the bottom of the cell and collect in a concentrated product that is commonly called "anode slime" or "anode mud." Anode mud is very valuable because it is usually made up of precious metals, such as gold, silver, platinum, and selenium In copper electrorefining, the slimes can be over 90% silver The metals that are less noble on the electromotive series of metals dissolve along with the metal that makes up the anode and, if not removed, do build up and contaminate the electrolyte

Approximately 85% of the anode is allowed to dissolve to retain its mechanical integrity before it is removed from the tank The unused portion is sent back to the smelter, where it is used as scrap and recast into more anodes The cycle time for anodes in a typical electrorefining operation is 14 to 28 days As the electrons are stripped from the anode and forced

to the cathode, metal-ion reduction and metal deposition occur on the cathode according to the reaction M2+ + 2e- = M

Because the less noble metals remain in the metallic state as the anode dissolves, the only metal that plates on the cathode

is the metal that made up the bulk of the anode Those impurities that are less noble and get dissolved in the electrolyte are not plated on the cathode and remain in solution as metallic ions Electrorefined product may also be collected either

on another metal sheet or a starter sheet of the metal being plated The cycle time for cathodes is somewhat longer than the anode cycle time Current densities on cathodes are kept low to ensure quantity deposits and often are approximately

215 A/m2 (20 A/ft2) The physical nature of the deposit is a function of the type of metal ion, current density, deposition temperature, interpolar distance, and deposition rates

The voltage required between anode and cathode during electrorefining is significantly lower than that required for electrowinning For example, electrorefining of copper requires only about 0.25 V since the reversible decomposition voltage between two copper electrodes is essentially zero The other voltage drops, such as the ohmic drops and polarization (overvoltage) exist in electrorefining however

The cells used for electrorefining are commonly rectangular vats that allow easy placement and removal of anodes and cathodes The bottoms of the tanks are sloped to allow the collection and flushing of the anode slimes Since most electrolytes are sulfuric acid base (with the exception of lead electrorefining), the lining material is PVC, rubber, or lead

Soluble impurities which dissolve with the anode and are not reduced at the cathode must be removed to avoid concentration in the electrolyte If not removed, these impurities lower the conductivity of the electrolyte, which in turn increases the voltage required and thus increases the energy required An example of this type of impurity is nickel in

copper electrorefining Nickel is an impurity in the anode that dissolves during electrolysis and collects in the electrolyte Most copper refiners keep the concentration of nickel below 10 to 20 g/L by continuously bleeding off a small portion of the electrolyte, removing nickel by chemical means, and returning the nickel-free electrolyte to the cells

Insoluble or more noble impurities that remain in the metallic state collect in the bottom of the electrorefining cells This anode mud collects until the anode-cathode cycles are complete At that time, the electrolyte is pumped from the cell, and the slimes are flushed out and collected by filtration Because the slimes are the final destination of the precious metals, they are further refined to recover gold, silver, platinum, and other associated precious metals

Molten Salt Electrolysis

As mentioned earlier, reactive metals that are more electropositive in the electromotive series than manganese cannot be electrowon or electrorefined from aqueous solutions because water will decompose at the cathode to form hydrogen gas before reduction of the metal occurs Therefore, electrolysis of very electropositive metals, such as aluminum, magnesium, lithium, and beryllium, is done from molten salt electrolytes

Present practice for aluminum electrolysis involves the use of the Hall-Héroult cell as shown in Fig 16 The cell is lined with carbon, which acts as the cathode; steel bars are embedded in the cathode lining as current collectors The consumable anodes are also of carbon and are gradually fed into the top of the cell For aluminum, the electrolyte used is cryolite (Na3AlF6) with 8 to 10% Al2O3 dissolved in it Other additives, such as CaF2 and AlF3, are added to obtain desirable physical properties, such as fluidity and electrical conductivity The melting point of the electrolyte is approximately 940 °C (1725 °F), and the Hall-Héroult cell operates at temperatures of approximately 960 to 1000 °C (1760 to 1830 °F) with a power rating of 10 to 12 kWh/kg aluminum

Fig 16 Hall-Héroult aluminum production cell with self-baking anode Source: Ref 5

Electrochemically, aluminum is reduced at the cathode from an ionic state to a metallic state by:

Al3+ + 3e- Al(l) (Eq 20)

This simplified reaction goes through a series of complex reactions that take place at the cathode (Ref 26) It does represent the overall production of molten aluminum, which forms a molten pool in the bottom of the cell Periodically, the molten pool of aluminum metal is drained or siphoned from the bottom of the cell and cast

At the anode, oxygen is oxidized from its ionic state to oxygen gas which, in turn, reacts with the carbon anode to form carbon dioxide gas, gradually consuming the anode material Two types of anodes are in use: prebaked and self-baking Prebaked anodes are individual carbon blocks that are replaced one after another as they are consumed Self-baking

anodes, as shown in Fig 16, are made up of a carbon paste that is fed into a steel frame above the cell As the anode descends in the cell it hardens, and new carbon paste is fed continually into the top of the steel frame

Impurities in the aluminum oxide raw material that are more noble than aluminum, such as iron and silicon, are reduced at the cathode along with the aluminum It is, therefore, important that raw materials be prepared in the Bayer's process as free of these metal oxides as possible By careful control of raw materials, aluminum with a purity of over 99% can be produced

Magnesium is produced by molten salt electrolysis Magnesium does not alloy with iron, and so the cells consist of iron pots as cathodes and carbon anodes The electrolyte used is a magnesium chloride-sodium chloride-calcium chloride mixture that allows the process to be run at approximately 750 °C (1382 °F) The raw material for magnesium may be either magnesium chloride or magnesium oxide Magnesium electrolysis differs from aluminum electrolysis in that the magnesium metal formed at the cathode is less dense than the electrolyte and rises to float on top of the electrolyte; this requires special precautions when the molten metal is recovered from the cell Production of calcium by molten salt electrolysis has been developed successfully from molten calcium oxide-calcium chloride salt mixtures using carbon anodes Presently, a great deal of research effort is concentrated on eliminating the consumable carbon anode from these fused salt electrolytic processes Inert ceramic anodes that conduct oxygen ions and are based on spinel and perovskite compositions have been developed Carbon electrodes impregnated with titanium diboride have been successfully used in the industry

The science of extractive metallurgy has more or less remained the same over several decades (Ref 27) However, significant progress has been made in the technology and application of this science to extract metals cheaply, efficiently, and in an environment-friendly manner

References cited in this section

5 T Rosenquist, Principles of Extractive Metallurgy, 2nd ed., McGraw-Hill, 1983

23 D Jones, Principles and Prevention of Corrosion, 2nd ed., Prentice Hall, 1996, p 43

26 K Grjotheim, C Krohn, M Malinovsky, K Matiasovsky, and J Thonstad, Aluminium Electrolysis: Fundamentals of the Hall-Heroult Process, Aluminium-Verlag GmbH, Dusseldorf, 1982

27 A.W Schlechten and C.A Natalie, Extractive Metallurgy, Metals Handbook: Desk Edition, ASM

International, 1985

General Introduction to Casting

Thomas S Piwonka, The University of Alabama

Introduction

METAL CASTING is the manufacturing method in which a metal or an alloy is melted, poured into a mold, and allowed

to solidify It is one of the oldest manufacturing methods known to humankind, and one of the most versatile Today's foundries make use of statistical process control and sophisticated solidification simulation software, and castings today are highly reliable, cost-effective components that are utilized in more than 90% of manufactured products Typical uses

of castings include municipal hardware, water distribution systems (pipes, pumps, and valves), automotive components (engine blocks, brakes, steering and suspension components, etc.), prosthetics, and gas turbine engine hardware

Casting Flow Chart

The steps in making a casting are shown in Fig 1 Before the casting can be made, the alloy must be melted, and the mold must be made The mold is usually made of a material with a higher melting point than that of the alloy, such as a refractory aggregate (e.g., silica sand) or a high-temperature alloy There are a wide variety of mold-making methods used

in casting, depending on the alloy to be poured, the number of castings to be made, the dimensional requirements of the casting, and the property requirements of the casting Figure 1 depicts the sand molding process, the most commonly used

molding (casting) method A more detailed account of this process can be found in the article "Molding Methods" in this Section

Fig 1 Simplified flow diagram of the basic operations for producing a steel casting Similar diagrams can be

applied to other ferrous and nonferrous alloys produced by sand molding

The right side of Fig 1 begins with the task of patternmaking A pattern is a specially made model of the component to be produced, used for producing molds Generally, sand is placed around the pattern, and, in the case of clay-bonded sand, rammed to the desired hardness In the case of chemical binders, the mold is chemically hardened after light manual or machine compaction Molds are usually produced in two halves so that the pattern can be easily removed When these two halves are reassembled, a cavity remains inside the mold in the shape of the pattern

Internal passageways within a casting are formed by the use of cores Cores are parts made of sand and binder that are sufficiently hard and strong to be inserted into a mold Thus, the cores shape the interior of the casting, which cannot be shaped by the pattern itself The patternmaker supplies core boxes for the production of precisely dimensioned cores These core boxes are filled with specially bonded core sand and compacted much like the mold itself Cores are placed in the drag, or bottom section, of the mold, and the mold is then closed by placing the cope, or top section, over the drag Mold closing completes the production of the mold, into which the molten metal is then poured

Casting production begins with melting of the metal (left side of Fig 1) Molten metal is then tapped from the melting furnace (see the article "Melting Methods") into a ladle for pouring into the mold cavity, where it is allowed to solidify within the space defined by the sand mold and cores After it has solidified, the casting is shaken out of the mold, and the risers and gates are removed Risers (also called "feeders") are shapes that are attached to the casting to provide a liquid-metal reservoir and control solidification Metal in the risers is needed to compensate for shrinkage that occurs during cooling and solidification Gates are the channels through which liquid metal flows into the mold cavity proper Cleaning and finishing, heat treatment, and inspection follow The article "Solidification of Metals and Alloys" provides a more detailed account of how castings solidify in the mold, shrinkage characteristics, as well as gating and risering practices

Casting Alloys

Alloys used for metal casting may have the same composition as alloys used in other forming methods However, many

of the alloys used in casting have been developed expressly for the casting process, and these alloys do not have analogs

for other forming processes They exist only in the cast form, as their performance depends on the unique metallurgical structure produced during solidification In addition, many of the alloys that exist in cast and wrought form often have slight variations in chemistry to facilitate the casting process It is important to know exactly which composition is meant when dealing with these alloys

Ferrous alloys include both cast irons and steels Cast irons generally refer to iron alloys containing 3 to 4% carbon, with silicon contents of 1.5 to 2.5% Cast irons, which are melted in induction furnaces, cupolas, and electric arc furnaces, are generally poured/cast into sand molds Steels include all alloys containing less than 2% carbon, with additions of small amounts of manganese and silicon, and other alloying elements as needed Most steel is arc melted and poured into sand molds

Nonferrous castings are used where ferrous castings would be too heavy, too expensive, or lack the properties required for the application These alloys are generally melted in crucible furnaces, reverberatory furnaces, or induction furnaces Molding techniques are similar to those used for ferrous alloys, but, because most nonferrous alloys melt and solidify at lower temperatures than ferrous alloys, metal molds are frequently used

Solidification of Metals and Alloys

Thomas S Piwonka, The University of Alabama

Introduction

SOLIDIFICATION is the transformation of liquid to solid During this process, atoms change their arrangement from randomized short-range order to regular positions on a crystallographic lattice In doing so, they give up energy in the form of heat, which must be removed by the mold The energy they give up is called the "latent heat of fusion."

Fig 1 Thermocouple trace for a pure metal solidifying in a mold

Solidification begins with a nucleation event Although it is possible under very closely controlled conditions to cool the metal far below its melting point ("undercooling"), in practice the undercooling experienced is very small This is because commercial melts always contain some sort of nucleating agent Most often this agent will be the wall of the mold itself

However, other agents may be deliberately added to the melt to control the nucleation event and the degree of undercooling For example, this is especially important in controlling the solidification of cast iron

Because the mold is cooler than the metal, nucleation will occur all over the surface of the mold Each nucleation event will produce an individual crystal, or grain, which then will attempt to grow These randomly oriented grains form a "chill zone" next to the mold wall The grains are oriented randomly with respect to the mold; that is, the major axis of each grain is randomly oriented As each metal grows most favorably in one principal crystallographic direction, only those grains favorably oriented with their growth direction most perpendicular to the mold wall will grow into the center of the casting (Fig 2) The grains in a pure metal will grow until they impinge on another grain

Fig 2 Solidification in conventional castings During the growth of the columnar zone, three regions can be

distinguished These are the liquid (L), the liquid plus solid (the so-called "mushy" zone), and the solid (S) regions

The fact that the grains most favorably oriented to the mold wall grow the fastest means that the final shape of the grains

in a pure metal casting is columnar As shown in Fig 2, the grains form parallel columns, growing progressively from the mold wall into the center of the casting

The density of the solid phase is different than that of the liquid phase In almost every metal, the solid is more dense than the liquid The solid will occupy less space than the liquid; in other words, it will shrink in volume Because this shrinkage occurs at the point of solidification, the volume deficit, which is called the shrinkage cavity, is found at the location of the last liquid to solidify Figure 3 shows the formation of a shrinkage cavity in a pure metal solidifying in a mold where all heat is removed through the mold If the final casting is to have the same volume as the liquid, a reservoir

of molten metal must be placed on the casting to feed liquid metal to the mold cavity This extra metal is called the feeder,

or the riser

Fig 3 Development of shrinkage void in a casting (a) Liquid (b) Liquid + solid (c) Liquid + solid (d) Solid

Alloys

The solidification of alloys is more complex Alloys are solutions of one or more metals or semimetals in another metal

As such, they solidify over a range of temperatures The temperature range over which solidification takes place under equilibrium conditions is found from the phase diagram The temperature at which solidification begins is called the liquidus temperature, and the temperature at which it ends in equilibrium solidification is called the solidus This range is often referred to as the "mushy zone," because the material in this temperature range is a mixture of liquid and solid (Fig 2) Figure 4 shows a simplified phase diagram for a binary alloy system Note that almost all commercial alloys have three or more components; as such, their solidification is more complex than for binary alloys

Fig 4 Simplified phase diagram, showing positions of liquidus, solidus, and eutectic temperatures, and the

mushy zone

Dendrite Formation. During alloy solidification, instabilities in the solid at the liquid/solid interface grow, leading to the formation of dendrites (Fig 5) Dendrites have a primary arm, secondary arms that branch from it, and tertiary arms that branch from the primary arm The spacing of the secondary arms is proportional to the rate at which heat is removed from the casting during solidification Each crystal consists of a single dendrite

Fig 5 Growing dendrite tip and dendrite root during columnar growth in a casting A dendritic form is usually

2

Segregation. During equilibrium alloy solidification, solute atoms are rejected into the liquid, according to the lever rule Thus, the composition of the unsolidified liquid changes during solidification, and the last liquid to solidify has a different composition than that of the first liquid to solidify The change in composition is referred to as segregation When large areas of the casting show differences in composition from the nominal composition of the alloy that was poured to make the casting, macrosegregation results However, segregation also occurs between the dendrite arms, where solidification is taking place

Segregation has a number of effects During solidification it causes the composition to change locally over the casting (Fig 6) This local change in composition may result in the formation of different phases or compounds, such as carbides

or intermetallic phases, which would not otherwise be stable at the nominal composition of the alloy These local phases may degrade casting properties Because the rejection of solute into the liquid changes not only the composition of the liquid but also its density, convection currents caused solely by segregation (solutal convection) may be set up in the casting The amount of shrinkage that occurs locally will also be a function of the local composition of the liquid when it solidifies A major effect of segregation is that heat treatment times may be lengthened; the time to homogenize a casting will depend on the dendrite arm spacing (the distance over which solute atoms must travel to go into solution) and the amount of segregation Because segregation causes the casting composition to vary locally, chemical etching can reveal the cast structure, which makes metallography possible

Fig 6 Isoconcentration profile lines in an Fe-25Cr-20Ni columnar dendrite showing that the chemical

concentration of elements is not constant within the dendrite

When metal is poured into a mold, the liquid has a velocity from the pouring operation In addition, as the metal cools at the mold wall, thermal convection currents are set up at the mold wall, with the cooling metal traveling downward along the mold wall These currents, combined with solutal convection and the residual liquid motion from pouring mean that the liquid in the unsolidified part of the casting is in motion This movement is usually beneficial, as the motion minimizes macrosegregation in small castings (although it magnifies it in large castings)

Alloys nucleate in the same manner as pure metals However, during solidification, the liquid motion may break off dendrites and carry them into the liquid Then the dendrites act as nuclei for more grains and multiply the number of grains that nucleate in the liquid In addition, solute rejection (segregation) may form a region of liquid where the melting point of the liquid rises above the temperature of the melt, meaning that the liquid in that region is now below its melting point Even though the actual temperature of the liquid has not changed, the change in composition resulting from segregation yields an alloy composition locally that has a melting temperature above that of the liquid This is called constitutional supercooling The result of these two mechanisms refines the final grain structure of the casting

Structural Zone Formation. Solidification begins, as in pure metals, with nucleation of solid on the walls of the mold, again forming a chill zone Then those grains that are most favorably oriented grow inward, forming a columnar zone However, because of segregation, there is constitutional supercooling and grain multiplication As a result, the last liquid to solidify does so as equiaxed grains Thus, the final structure of the casting has three zones (Fig 7) In the equiaxed zone, the grains do not grow progressively from the solid that has already solidified Instead, they grow simultaneously from their own nuclei, until they impinge on neighboring grains

Fig 7 Three structural zones forming in a casting (a) Early in the solidification process, solid nuclei appear at,

or close to, the mold wall For a short time, they increase in size and form the outer equiaxed zone Then, those

crystals (dendrites) that can grow parallel and opposite to the heat flow direction will advance most rapidly

Other orientations tend to be overgrown, due to mutual competition, leading to the formation of a columnar

zone (b) Beyond a certain stage in the development of the columnar dendrites, branches that become

detached from the latter can grow independently These tend to take up an equiaxed shape because their latent

heat is extracted rapidly through the undercooled melt The solidified region containing them is called the inner

equiaxed zone

The amount of the final cast structure that is columnar or equiaxed depends on the alloy composition and on the thermal gradient at the liquid/solid interface during solidification The thermal gradient is most easily controlled by controlling the rate of heat extraction from the casting, or the cooling rate Alloys that have a wide spread between the liquidus and the solidus temperature solidify with a mostly equiaxed grain structure at normal cooling rates, whereas alloys with small differences in solidus and liquidus temperatures solidify with a mostly columnar structure High cooling rates encourage columnar solidification because they establish high thermal gradients at the liquid/solid interface Low thermal gradients encourage equiaxed solidification

Shrinkage. Like pure metals, alloys shrink as they solidify However, feeding the shrinkage in the equiaxed zone is far more difficult than feeding shrinkage in the columnar zone In the equiaxed zone, the liquid must wind its way down tortuous interdendritic channels to feed the shrinkage occurring at the end of these channels As these channels narrow, feeding becomes finally so difficult that the metal is unable to reach the areas where solidification is occurring As a result, tiny micropores form between the equiaxed grains This condition is known as microporosity

Effects of Nonequilibrium Conditions. The discussion to this point has assumed that solidification takes place under equilibrium conditions However, in actual castings, this is not true Equilibrium solidification occurs at rates so slow that there is perfect diffusion in the liquid and in the solid, so that the composition of the alloy shown on the phase diagram is what is found everywhere in the casting However, real castings solidify much more quickly than the equilibrium rate, and many castings solidify at high rates The faster the solidification rate is, the greater the departure from equilibrium will be Departures from equilibrium exaggerate segregation They also depress the liquidus and solidus temperatures Indeed, it is common for the last metal to solidify as a eutectic because of departures from equilibrium during solidification It is important to remember in analyzing castings that there is this departure from equilibrium caused by solidification rate Because solidification rate varies with location in the casting (it is higher at the edges and corners of the casting, and lower in the interior of heavy sections), the departure from equilibrium varies with location in the casting

Eutectic Phases

When the liquid has the composition of the eutectic, it solidifies forming two distinct phases in intimate contact with each other (Fig 8) Eutectic, or near-eutectic, alloys are common in casting In fact, one of the most common casting alloys, cast iron, gets its distinctive microstructure from the controlled solidification of the eutectic phase

Fig 8 Irregular "Chinese script" eutectic consisting of faceted Mg2Sn phase (dark) in a magnesium matrix Etched with glycol 250×

Eutectics have the lowest melting point in the alloy and high fluidity Near-eutectic alloys flow easily into the mold When the temperature falls below the eutectic temperature, the casting is solid A more detailed account of eutectic phases can be found in the Section "Structure and Properties of Metals" in this Handbook

Imperfections

During solidification, atoms leave the liquid and arrange themselves as a solid During this process, many things can happen to cause imperfections (defects) in the solid

Porosity is the presence of pores in the casting These pores may be connected to the surface, where they can be detected

by dye penetrant techniques, or they may be wholly internal, where they require radiographic techniques to be discovered Macroporosity refers to pores that are large enough to see with the unaided eye on radiographic inspection, while microporosity refers to pores that are not visible without magnification

Both macroporosity and microporosity are caused by the combined action of metal shrinkage and gas evolution during solidification It has been shown that nucleation of pores is difficult in the absence of some sort of substrate, such as a nonmetallic inclusion, a grain refiner, or a second-phase particle Numerous investigations have shown that clean castings, those castings free from inclusions, have fewer pores than castings that contain inclusions Microporosity is found not only in castings but also in heavy section forgings, which have not been worked sufficiently to close it up

When the shrinkage and the gas combine to form macroporosity, properties are deleteriously affected Static properties are reduced at least by the portion of the cross-sectional area taken up with the pores Because there is no metal in the pores, there is no metal to support the load there, and the section acts as though its area was reduced Because the pores may also cause a stress concentration in the remaining material, static properties may be reduced by more than the percentage of cross-sectional area caused by the macroporosity

Dynamic properties are also affected A study of aluminum alloys showed that fatigue properties in some alloys were reduced 11% when specimens having x-ray quality equivalent to ASTM E 155 level 4 were tested, and that they were reduced 17% when specimens having quality of ASTM E 155 level 8 were tested

Static properties are mostly unaffected by microporosity Microporosity is found between dendrites, and, like macroporosity, it is caused by the inability of feed metal to reach the interdendritic areas of the casting where shrinkage is occurring and where gas is being evolved However, because this type of porosity occurs late in solidification, particularly

in long range freezing ("mushy freezing") alloys, it is particularly difficult to eliminate The most effective method is to increase the thermal gradient (often accomplished by increasing the solidification rate), which decreases the length of the mushy zone This technique may be limited by alloy and mold thermal properties, and by casting geometry (i.e., the design of the casting)

As long as the micropores are less than 0.2 mm in length, there is no effect on dynamic properties: fatigue properties of castings with pores that size or smaller are in the same range as those of castings where no micropores were found The shape of the micropore is as important as its size, with elongated pores having a greater effect than round pores Areas where microporosity is expected can be predicted by solidification modeling, similar to the prediction of macroporosity (see below) Microporosity can be healed by hot isostatic pressing (HIP) In one study comparing HIP and non-HIP samples, no difference was found in fatigue lives However, the HIP samples showed a lower crack growth rate than non-HIP samples In another study, HIP improved fatigue crack growth resistance only close to threshold levels As noted above, the design of the casting directly affects its tendency to solidify in a progressive manner, thereby affecting both the quality and the price of the cast component

Porosity and casting costs are minimized in casting designs that emphasize progressive solidification toward a gate or riser, tapered walls, and the avoidance of hot spots

Inclusions are nonmetallic particles found in the casting They may form during solidification, as some elements (notably manganese and sulfur in steel) precipitate from solution in the liquid during solidification More frequently, they form before solidification begins The former are sometimes called "indigenous" inclusions, and the latter are called

"exogenous" inclusions Inclusions are ceramic phases: they have little ductility A crack may form in the inclusion and propagate from the inclusion into the metal, or a crack may form at the interface between the metal and the inclusion In addition, because the inclusion and the metal have different coefficients of thermal expansion, thermally induced stresses may appear in the metal surrounding the inclusion during solidification As a result, the inclusion acts as a stress concentration point and reduces dynamic properties As in the case of microporosity, the size of the inclusion and its location determine its effect Small inclusions that are located well within the center of the cross section of the casting have little effect, whereas larger inclusions and those located near the surface of the casting may be particularly detrimental to properties Inclusions may also be a problem when machining surfaces, causing excessive tool wear and tool breakage

Exogenous inclusions are mostly oxides or mixtures of oxides and are primarily slag or dross particles, which are the oxides that result when the metal reacts with oxygen in the air during melting These inclusions are removed from the melt before pouring by filtration Most inclusions found in steel castings arise from the oxidation of metal during the pouring operation This is known as reoxidation, and takes place when the turbulent flow of the metal in the gating system causes the metal to break up into small droplets, which then react with the oxygen in the air in the gating system or casting cavity to form oxides Metalcasters use computer analysis of gating systems to indicate when reoxidation can be expected in a gating system and to eliminate them However, casting designs that require molten metal to jet through a section of the casting to fill other sections will recreate these inclusions and should be avoided

Oxide films are similar to inclusions and have also been found to reduce casting properties Oxide films form on the

surface of the molten metal as it fills the mold If this surface film is trapped within the casting instead of being carried into a riser, it is a linear discontinuity and an obvious site for crack initiation It has been shown that elimination of oxide films, in addition to substantially improving static properties, results in a five-fold improvement of fatigue life in axial-tension tension tests

Oxide films are particularly of concern in nonferrous castings, although they also must be controlled in steel and stainless steel castings Because of the high carbon content of cast iron, oxide films do not form on that particular metal If the film

is folded over on itself as a result of turbulent flow or "waterfalling" (when molten metal falls to a lower level in the casting during mold filling), the effects are particularly damaging Casting design influences how the metal fills the mold, and features of the design that require the metal to fall from one level to another while the mold is filling should be avoided so that waterfalls are eliminated Oxide films are avoided by filling the casting from the bottom in a controlled manner, by pumping the metal into the mold using pneumatic or electromagnetic pumps

Secondary phases that form during solidification may also nucleate cracks if they have the proper size and morphology An example is aluminum-silicon alloys, where the silicon is present in the eutectic phase as large platelets, which nucleate cracks, and along which cracks propagate The size of these platelets may be significantly reduced by modifying the alloy with additions of sodium or strontium However, such additions increase the size of micropores, and for this reason many foundrymen rely on accelerated solidification of the casting to refine the silicon As previously noted, solidification rates normally increase and the structure is thus refined in thin sections Heavy sections are to be avoided if a fine structure is desired Generally speaking, however, secondary phases in the structure of castings become important in limiting mechanical behavior of castings only in the absence of nonmetallic inclusions and microporosity

Hot tears form when casting sections are constrained by the mold from shrinking as they cool near the end of solidification These discontinuities are fairly large and most often weld-repaired If not repaired, their effect is not readily predictable; while generally they are detrimental to casting properties, under some circumstances they do not affect them Hot tears are caused by a combination of factors, including alloy type, metal cleanliness, and mold and core hardness However, poor casting design is the primary cause Castings should be designed so that solidifying sections are not subjected to tensile forces caused by shrinkage during solidification, as the solidifying alloy has little strength before it solidifies

Metal Penetration. Molten metal may penetrate the surface of the mold forming a rough surface, or in extreme cases,

it may actually become intimately mixed with the sand in the mold In iron castings, this is normally the result of the combination of metallostatic head (the pressure exerted on the molten iron at the bottom of the mold by the weight of the metal on top of it) and the surface tension relationships between the liquid iron and molding materials In cast iron, it is frequently also the result of the expansion of graphite at the end of solidification forcing liquid metal into the mold This penetration can occur if the casting is not properly designed with a tapered wall to promote directional solidification and avoid hot spots In steel castings, penetration may also occur as a result of formation of iron or manganese oxide on the surface of the molten metal These oxide phases react with the silica sand to form chemical penetration, which is difficult

to remove from the surface of the casting Use of mold coatings can protect the mold from this reaction

Gating and Risering

To make a casting, the metalcaster must fill the mold with metal and then control the solidification of the casting to prevent the formation of casting imperfections The techniques used to accomplish this are called gating and risering, or rigging

Gating

Gating System Components. The gating system is the plumbing system that fills the mold with molten metal Gating system nomenclature is shown in Fig 9 Metal is poured into the pouring basin, down a sprue, where it enters the runner, which delivers it to gates, through which it flows into the casting cavity in the mold Filters placed in the runners slow the metal and remove inclusions, and vents on the casting allow air to escape and relieve the back pressure that opposes mold filling

Fig 9 Basic components of a simple gating system for a horizontally parted mold

Recommended Practices. The gating system must fill the mold quickly, while minimizing turbulence Turbulence causes the molten metal to mix with the air in the sprue and runners The oxygen in the air reacts with the molten metal, forming oxide inclusions Unfortunately, it is difficult to minimize turbulence in gating systems in medium-size and large castings, as the velocity the metal reaches on falling down the sprue is usually so great that turbulence is unavoidable Ceramic filters slow the velocity of the metal in the runner, which fills the sprue with metal Good gating practice recommends that the system be unpressurized, that is, the cross-sectional area of the down sprue should be less than the total cross-sectional area of the runners, which, in turn, should be less than the total cross-sectional area of the gates

Runner cross sections are usually rectangular, so that metal does not swirl down the runners and entrap air Runners should be free from sharp edges, and gates should be filleted

The gating system should establish thermal gradients to promote a sound (porosity-free) casting The ideal situation is to have the casting freeze from thin sections (which freeze quickly) to thick sections, so that feed metal is available to feed shrinkage as it occurs As molten metal flows through the gating system, it loses heat to the runners and heats them up Thus, the first metal into the mold is coldest, and the last metal is hottest For this reason, gates should be placed into heavy sections of the casting, so that hot metal is available to feed shrinkage that occurs in the casting as the casting solidifies By having solidification take place progressively toward the gates and risers, shrinkage is avoided

Good gating systems should avoid reoxidation of the metal, avoid the formation of oxide folds, and, if possible, remove oxide and dross from the molten metal Gating systems should be designed to prevent the aspiration of air through the porous molding media The position of the casting in the mold should be given careful thought Normally the casting should have its longest dimension parallel to the parting plane Metal should not drop from one level in the casting cavity

to a lower level because the oxide film that forms on the top of the molten metal mixes into the metal in the casting, which causes inclusions

Filters strain out most of the slag or dross that may have been carried in from the melting operation, and pieces of mold refractory that may have come loose during molding Metal should not rain down from the top of the sprue or into the casting because it will react with the air in the runner system or the casting cavity Therefore, filters should not be placed

at the top of the sprue, and gates should be placed at the bottom of the mold, so that the casting cavity fills from the bottom The proper placement of filters in gating systems is shown in Fig 10 Filters are used in all alloys except titanium

Fig 10 Common methods of filter placement in horizontally parted molds (a) Parallel to parting line (b)

Between 0 and 90° to parting line (c) 90° to parting line Arrows indicate the direction of metal flow

Pouring metal down a sprue causes the metal to accelerate due to gravity The increased velocity caused by this acceleration leads to turbulent flow and the formation of inclusions One method of avoiding turbulent flow is to use some form of counter-gravity mold filling, as shown in Fig 11 Counter-gravity pouring systems allow the rate of fill to be controlled precisely; it can be speeded up or slowed down as necessary to compensate for changes in the cross-sectional area of the casting In filling the mold from the bottom, the rate of rise in the casting cavity should be no greater than 0.5 m/s (1.65 ft/s) for aluminum alloys and no greater than 0.3 m/s (1 ft/s) for copper-base alloys Filters are incorporated in the fill tubes used in counter-gravity gating systems Counter-gravity pouring systems operate at the same speed as conventional pouring lines; a number of techniques, such as rotating the mold around a horizontal axis at the bottom of the mold after pouring, are available to prevent molten metal from draining out of the casting cavity until solidification is complete

Fig 11 Schematic of the operations of the counter-gravity low-pressure casting process (a) Investment shell

mold in the casting chamber (b) Mold lowered to filling position (c) Mold containing solidified castings; most of the gating has flowed back into the melt

Gates must be easy to remove, and they must not distort the casting when they solidify If they are to be cut off or sawed off, room must be available for the cutting torch or saw blade When more than one gate is used to fill a casting, care should be taken that the solidification of the runner that connects the gates does not distort the casting when the runner, which must stay liquid long enough to fill the casting, freezes and contracts due to shrinkage

Gating design calculations may be done approximately by using Bernoulli's theorem, which is an energy and materials balance between points in the gating system More accurate gating designs may be made using a combination of expert systems and finite element or finite difference solutions of fluid flow equations These are part of commercial solidification software widely available to metalcasters

In a mold having more than one mold cavity of the same casting geometry, it is recommended that the gating of all of the castings be identical This match assures that all castings solidify similarly Differences in solidification sequence otherwise may cause castings to have properties that vary according to their position in the mold

Good pouring practice means filling the sprue rapidly and keeping it full while the casting cavity fills There are essentially three types of pouring ladles for gravity pouring (Fig 12) Keep in mind that slag and dross will form on the top of the metal in the ladle where the molten alloy is exposed to the atmosphere With the open lip ladle (Fig 12a), the

slag or dross must be held back during pouring The teapot ladle (Fig 12b) is most effective for keeping slag out of the mold Bottom pour ladles (Fig 12c) are effective only when the metal stream from the exit of the ladle to the sprue and the sprue itself are shrouded in an inert gas, because the high discharge velocities from the ladle cause stream spreading and lead to oxidation and inclusion formation in air

Fig 12 Types of pouring ladles (a) Open lip-pour ladle (b) Teapot ladle (c) Bottom-pour ladle

A variety of automatic pouring devices are available commercially that precisely and reproducibly fill the mold from ladles The advantages of automatic pouring include improved consistency of pouring, increased productivity, reductions

in pouring scrap, better process control, improved working conditions on the pouring line, and reduced environmental problems

Venting. An important aspect of mold filling, often overlooked, is allowing the air in the mold to escape as the mold is filled If air cannot easily leave the mold, it becomes compressed in the mold cavity and generates a back pressure against the flowing metal that holds up the metal, which may cause it to freeze or misrun before the mold cavity is filled Passages placed in the mold so that air can escape during pouring are called vents While it is obvious that impermeable molds, such as metal molds, need to be vented, it is less obvious that vents are needed in permeable molds, such as sand molds However, venting is every bit as important in sand, especially in high-pressure molding lines

Vents should be placed at the highest point in the casting cavity and in any part of the cavity that will be cut off from the rest of the cavity as the casting fills Vents should exit the mold on the cope surface Open risers make excellent vents; however, if blind risers are used, they should be vented Vents should also be placed at the end of runners Parting line vents should be used only in vertical parting line molds Large cores should be vented; if it is not possible to vent them directly to the atmosphere by cutting a hole in the mold, vent tubes (flexible plastic) should be run from the core to the top

of the cope

As the melting temperature of the alloy increases, the volume of air that expands on being heated by the incoming metal increases Thus the size and number of vents should increase as the melting point of the metal increases (e.g., vents in steel castings should be larger and more numerous than those in aluminum castings)

Risering

Riser design deals with the sizing and placement of reservoirs of feed metal to compensate for the shrinkage that occurs during solidification These reservoirs are called "risers" or "feeders." When liquid metal freezes, it undergoes three different volume changes: shrinkage of the liquid as it cools, shrinkage of the alloy as it transforms from liquid to solid, and shrinkage of the solid as it cools to room temperature While the shrinkage accompanying the liquid/solid transformation occurs in a pure metal at a single temperature (the freezing temperature), it occurs over the range of temperatures between the liquidus and the solidus or eutectic temperature in alloys Some of the volume change in alloy solidification is caused by segregation of alloying elements to the liquid during solidification In addition, the mold cavity expands as it heats up when molten metal enters the mold These effects mean that liquid metal must continue to be available during casting solidification to assure that no shrinkage voids form in the casting

The distribution of liquid and solid in a casting section during freezing depends on the freezing range of the alloy Pure metals, which freeze at a single temperature, freeze with a well-defined solidification front In that part of the casting in which the temperature is below the freezing point, the casting is solid, and in the rest of the casting, it is liquid However,

as the freezing ranges of alloys increase, the freezing front becomes diffuse (Fig 13) For a short freezing range alloy, such as steel, solidification begins with the formation of a skin of solid metal at the mold/metal interface, and there is a clear channel for metal to flow down to feed shrinkage In these alloys, the only place where shrinkage is expected is at the centerline of the casting, which is the last part of the casting to solidify In these alloys, tapering the casting cures the problem

Fig 13 Solidification (freezing) mode for pure metals and alloys (a) Freezing mode in pure metals, in which

the freezing range (liquidus-to-solidus interval) approaches zero Crystallization begins at the mold wall and

advances into the casting interior on a plane solidification front (b) Freezing mode in short freezing range alloys (<50 °C, or <90 °F) (c) Freezing mode in intermediate freezing range alloys (50 to 110 °C, or 90 to 200

°F) (d) Freezing mode in long freezing range alloys (>110 °C, or >200 °F)

For very long freezing range alloys (similar to many aluminum and copper-base alloys and superalloys), freezing begins

at the walls, but grains quickly nucleate and grow throughout the entire casting, making feeding difficult The result is that shrinkage is dispersed throughout the entire casting, usually in the form of microporosity This situation can be alleviated by increasing the rate at which freezing occurs, but this is limited by how efficiently heat can be absorbed by,

or transferred through, the mold As refractory molds are generally poor conductors of heat, increasing the thermal gradient is generally effective only in thin sections, especially in high thermal conductivity alloys, such as aluminum- or copper-base alloys

Gas solubility is a particular problem in some alloys Hydrogen is easily picked up from atmospheric humidity in aluminum melts, and nitrogen may be picked up in steel melts Unfortunately, the solubility of hydrogen in solid aluminum is about an order of magnitude less than it is in liquid aluminum The result is that hydrogen gas is rejected into the liquid from the solidifying alloy and forms gas bubbles in the liquid; if these bubbles cannot escape from the casting, they appear in the casting as pores Nitrogen gas porosity arising from a similar situation is sometimes found in ferrous castings, especially in the presence of certain resin binders that decompose and saturate the liquid iron with nitrogen

The goal of risering is to encourage the casting to feed progressively, that is, to feed from a remote point in the casting toward the riser; so that shrinkage is fed and gas that dissolves in the liquid can escape (Fig 14) Most castings, however, have complex shapes, which make it difficult to predict how solidification will proceed However, an approximation can

be made to help the metalcaster decide how to arrange and size risers This approximation depends on the fact that solidification rate is proportional to the square of the ratio of the casting section's surface area to its volume This

relationship is referred to as Chvorinov's law, which is given by the relationship: t = k · (V/A)2, where k is a constant

proportionality whose value is dependent on the thermal properties of the metal and the mold In other words, the freezing

time, t, is proportional to the square of the ratio of the section's volume to its surface area (V/A)2 Therefore, a thin plate will solidify faster than a sphere of equal volume In the same way, edges, corners, and fins on castings will solidify faster than straight sections, and concavities (internal corners and recesses in the casting) will solidify more slowly than straight sections Because feed metal cannot flow through portions of the casting that have already solidified, the first rule to follow in establishing progressive solidification is not to attempt to feed a thick section through a thin section Each thick (slow-to-solidify) section must be isolated from thin (fast-to-solidify) sections and provided with its own riser The riser for these thick sections may require a gate; however, in many cases it is not feasible or economical to place a gate into the riser In that case the riser is called a "blind" riser

Fig 14 Directional and progressive solidification in a casting equipped with a riser

For the riser to be effective, it must solidify after the casting section it feeds has solidified This is assured by making the surface area to volume ratio of the riser greater than that of the casting section it feeds There are a number of methods that can be used to estimate the size of risers that should be used for various alloys Sizing the riser is usually done without regard to the molding system or aggregate used, as the thermal conductivity of the mold affects both the casting and the riser; and it is only the ratio between them that is of interest in the estimation process

Considering only the amount of feed metal in a riser is not sufficient in riser design, the riser must also be able to deliver this metal to the areas of the casting where shrinkage occurs The feeding distance of risers depends on the alloy and on

the mold system used The fact that the edges of the casting freeze faster than the riser means that they provide an "end effect." The heat supplied by the mass of the solidifying riser keeps the area around it liquid longer than it would otherwise be, providing a "riser effect" to extend its feeding distance The combination of the end effect and the riser effect create progressive solidification from the edge of the casting to the riser However, as the freezing range increases, the amount of solid coexisting in the area of the casting between the liquidus and solidus temperatures also increases, which makes feeding difficult Feeding distance estimates are therefore particularly important in alloys with short freezing ranges, such as steels; there are a number of methods that have been developed to aid in this estimation However, in long freezing range alloys, these methods are ineffective because feed metal flow is so difficult in the mushy region

In long freezing range alloys, especially aluminum alloys, progressive solidification can be encouraged by using metal chills in combination with risers The chills are placed in the mold adjacent to the casting surface, so that they can pull heat out of the casting in that area (Fig 15) Risers are placed between the chilled areas of the casting to provide feed metal; these risers are often insulated so that their freezing rate is retarded even more The object is to increase the thermal gradient in the solidifying metal so that the actual part of the casting at temperatures between the liquidus and solidus is as short as possible

Fig 15 Use of chills to reduce the number of risers (T) on a steel flange casting (a) Side and top view of the

casting illustrating locations of the eight risers used when the workpiece is divided into feeding areas without

considering end effects (b) Top view of identical casting showing locations of five risers used when the workpiece is divided into feeding areas in which riser effect and end effect considerations are accounted for through the use of chills

One method of getting an accurate prediction of where shrinkage will form is to use commercial solidification simulation programs These programs do not predict the size of the riser that will feed shrinkage in a casting, but they do show exactly how solidification will proceed This means that the foundry must estimate the size of the risers by other means in designing the riser plus gating assembly Some commercial programs include expert system routines to help foundry engineers design gates and risers and then check them using fundamental heat flow equations that are solved using finite difference or finite element methods

The efficiency of risers can be significantly improved if they are surrounded by insulating or, in the case of copper-base and ferrous metals, exothermic materials Exothermic materials react chemically in the presence of heat to produce more heat The most common type of exothermic reaction is between iron oxide and aluminum powder: its reaction raises the temperature of the reactants to well above that of the melting point of steel Exothermic riser "sleeves" contain a controlled amount of reactants, which slow solidification and maintain the riser at a temperature above its melting point for an extended period of time so that it is able to continue feeding the casting In alloys with short freezing ranges, the risers solidify with a pronounced pipe in their center; care must be taken to be sure that the depth of the pipe does not enter the casting, causing scrap Another way to improve riser efficiency is to cover the top of the riser with insulating or exothermic materials This cover is especially important in ferrous metals, which lose substantial amounts of heat by radiation to the atmosphere When the risers are covered by exothermic or insulating materials, this heat loss is effectively stopped

Molding Methods

Thomas S Piwonka, The University of Alabama

Classification of Molding Methods

MOLDING METHODS and materials encompass a wide range of options Some molding (casting) processes are centuries old, while others did not exist two generations ago The selection of the correct one to use in a specific casting application depends on the alloy, the dimensional tolerances desired, the quality of the casting, and the price the customer

is willing to pay Most metals and alloys can be cast using any of the methods There are, however, limitations in some cases

Foundry processes can be classified based on the molding medium, such as sand molds, ceramic molds, and metallic molds This article uses such a classification system Molding methods may also be classified based on whether the molds are permanent or expendable Similarly, subclassifications can be developed from patterns, (i.e., whether or not the patterns are expendable) For permanent molding, processes can be classified by the type of mechanism used to fill the mold Table 1 provides one possible classification system for commonly employed molding and casting processes

Table 1 Classification system for foundry processes based on mold type

Expendable mold processes/Permanent patterns

Clay/water bonds (green sand molding)

Silica sand Olivine sand Chromite sand Zircon sand

Heat-cured resin binder processes

Shell process (Croning process) Furan hot box

Phenolic hot box Warm box (furfuryl/catalyst) Oven bake (core oil)

Cold box resin binder processes

Phenolic urethane Furan/SO 2 Free radical cure (acrylic/epoxy) Phenolic ester

No-bake resin binder processes Furan (acid catalyzed) Oil urethane Phenolic urethane Polyol urethane

Silicate and phosphate bonds

Sodium silicate/CO2 Shaw process (ceramic molding) Unicast process (ceramic molding) Alumina phosphate

Plaster bonds

Gypsum bond

No bond

Magnetic molding Vacuum molding

Expendable mold processes/Expendable patterns

Foamed patterns

Evaporative foam casting Replicast process

Wax patterns (investment casting)

Ethyl silicate bonded block molds Ethyl silicate bonded ceramic shell molds Colloidal silica bond

Plaster bond Counter-gravity low-pressure casting

Permanent mold processes

Die casting

High-pressure die casting Low-pressure die casting Gravity die casting (permanent mold)(a)

should be used

Of the expendable mold/permanent pattern processes listed in Table 1, green sand molding is the most prevalent Expendable mold/expendable pattern processes use wax patterns (investment casting) or foamed plastic patterns (lost foam casting) The investment casting process is one of the oldest casting processes known (it has been used for more than 6000 years), while the lost foam process has been used commercially for less than 20 years Permanent mold processes involve the use of metallic (usually ferrous) or solid graphite molds On a volume basis, die casting and permanent mold casting are the most important

Sand Molds

SILICA SAND, either bonded or unbonded, is the most popular molding media Sand has the advantage of being widely available, inexpensive, able to withstand the elevated temperatures involved in casting most alloys, and easily recycled In addition, it does not react with most metals and alloys, and it can be bonded with clay, resins, or silicate compounds Because sand molds are permeable, the air displaced by the metal when it enters the mold can escape through the mold

Bonded Sand Molds

In bonded sand molds, the sand particles are held together with chemical bonds The bonds that are developed must be easily broken when the casting is solid so that the casting can be removed from the mold

Green Sand Molds

Green sand molds are the most widely used of all sand molds They are made of sand, bonded by a mixture of clay and water Other materials may be added to the sand to control its strength and prevent metal penetration The sand is called

"green sand" because it is not baked or fired Its actual color is usually black

Patterns, Cores, and Molding Machines. The casting cavity is formed in the sand mold by compressing it around a pattern, which is a wooden, plastic, or metal form that has the shape of the casting, together with its gates, risers and core prints (Fig 1) In bonded sand processes, the pattern is removed from the mold and used repetitively to make molds The

dimensions of the pattern are slightly greater than that of the castings because the casting shrinks during solidification Castings distort as they cool, as a result of stresses produced by the resistance of the mold to casting shrinkage, so patterns must be designed to compensate for this distortion if accurate castings are to be produced As the pattern will be removed from the mold after it is made, draft is added to vertical surfaces so that the pattern can be drawn from the mold without scraping the sides of the mold cavity

Fig 1 Patterns for a sand casting and its gating and risering systems, for four different methods of mold

Cores are used to make the hollow parts of castings, such as the holes of a pipe fitting or the air path in an automotive intake manifold The cores are usually bonded with resins, which are stronger than the clay-water bond used in green sand The cores are supported in the mold by extensions at either end These extensions rest in core "prints" in the green sand The prints not only support the cores, they also position them accurately to maintain casting dimensions Cores are placed in the drag half of the mold after the pattern is drawn from the mold and before the cope is placed on the drag to close the mold

Match-plate patterns are split patterns in which the pattern that will make the cope is on one side of the plate, and the pattern that will make the drag is on the other side Match-plate patterns are used on mechanized molding machines for

intermediate runs of moderate sized castings More than one casting may be made in a mold, with the sprue, gates, and risers shared among the castings In match-plate machine molding, the sand is introduced into the molding box, and the pattern is placed on a machine table The machine then raises the table a few inches and then lets it fall back freely by gravity The sharp jolt experienced by the sand when the machine table stops falling compresses the sand in the box The procedure is repeated a number of times until the sand is properly compressed Usually the mold is then subjected to a squeeze pressure from the top (the part of the mold away from the pattern face) of the mold half to further compress the sand

For large castings and very high production runs, cope and drag patterns are used on highly automated molding machines The cope pattern is separate from the drag pattern, and separate molding stations are used to make each mold half Metal flasks hold the sand Cores may be placed in the molds by hand or by using automated fixtures In some machines, only squeezing is used to compact the sand; in others, both jolting and squeezing is used Figure 2 shows a jolt-squeeze machine Parting lines may be either horizontal or vertical (Fig 3)

Fig 2 Essential components of a high-pressure jolt-squeeze molding machine The compensating head

equalizes the pressure applied to each floating peen block as the sand filled flask is hydraulically raised against the peen blocks This develops a uniformly dense packing of molding sand against the entire surface of the pattern

Fig 3 Blow-fill pressure squeeze molding machine making vertically parted molds (a) Molding chamber filled

with sand (b) Sand compacted by squeeze pressure (c) Finished sand mold pushed out of molding chamber

Impact molding machines use rapidly released compressed air above the sand in the pattern box to send a shock wave through the sand to compress it (Fig 4) The flask is filled with sand by gravity and sealed against an air chamber before the compressed air is released Advantages of the impact molding process are highest mold density at the parting line and pattern surface, uniform mold strength across the mold, and less pattern wear than with other processes

Fig 4 Pressure wave molding machine that compacts sand by the rapid release of air pressure or an explosive

combustible gas mixture Part (a) shows the mold filled by gravity prior to being compacted by the pressure wave at (b)

Flaskless molding machines are popular, using either horizontal or vertical parting lines Only squeeze pressure is used to compact the sand Cope and drag halves may be made simultaneously, and the mold halves are made in the same attitude

as they will be used, so there is no need to rotate the molds

When molten metal is poured into a green sand mold, the moisture at the mold/metal interface flashes into steam This rapid expansion in volume (liquid transforming into vapor) flushes the air away from the interface The air and the expanding water vapor must escape from the mold In high-density molding processes, such as impact molding, the permeability of the mold is not sufficient, and vents must be cut into the mold to prevent the back pressure from the gases from blocking the entrance of the molten metal and causing the casting to only partially fill with metal

Sands. The aggregate usually used in green sand casting is silica sand, usually from former lake bottoms or river bank deposits Sand is usually 99% silica in the form of quartz, with the balance being oxides of iron, aluminum, titanium, magnesium, and calcium, depending on the source of the sand Quartz undergoes a series of crystallographic transformations as it is heated and expands in a nonlinear manner with temperature (Fig 5) These transformations can cause dimensional problems if not compensated for in the casting process Sand grains may be rounded, subangular, or angular in shape Rounded grains flow more easily and pack more densely; they also use less binder because their surface area to volume ratio is smaller than other shapes However, molds made from them have less permeability As sand is cycled through the system, it wears and eventually becomes too small to be used and must be discarded

Fig 5 Expansion characteristics of various foundry sands

Sand used for molding is usually held within strict size and size distribution limits The sand size is determined by passing

a sample of sand through a series of successively finer screens, and weighing the amount collected on each screen The

best combination of properties of green sand is usually found when almost all of the sand is caught on three or four adjacent screens; however, two screen sands have been used successfully for certain casting processes

Clays. The sand is bonded by a clay/water mixture Clays used in the United States are naturally occurring montmorillonites, sodium (Western) bentonite and calcium (Southern) bentonite Western bentonite is the most commonly used clay It develops a lower green strength than Southern bentonite (Fig 6) but has higher strength at casting temperatures (Fig 7) Southern bentonites, however, allow the sand mixture to flow more evenly over the pattern and thus aid the molding operation Usually, a foundry will prepare a mix of both Southern and Western bentonites to achieve a sand that is optimized for the castings it makes The sand, water, and clay are mixed in a muller for a period of one to two minutes; during this time the clay/water mixture coats the sand Water added to the clay may be driven out by the heat of the solidifying casting When this happens, the clay loses its bonding ability Water may be added to the clay to reconstitute it as long as the clay has not reached a temperature greater than 600 °C (1110 °F) Because the clay adjacent

to the molten metal has reached that temperature, and the clay in cooler portions of the mold has not, a portion of the clay must be replaced The amount of clay and water added in the muller must be sufficient to bond new sand that is added on each cycle (including recycled core sand, which was previously bonded using resins) and to compensate for the moisture and clay burn out in the previous casting cycle

Fig 6 Variation of mold sand properties with water content (a) Southern bentonite (b) Western bentonite

Fig 7 Effect of casting temperature on the durability (strength) of various foundry sands

Sand Additives. Other materials are added to the sand to improve its performance Chief among these materials for ferrous alloys is seacoal, which is powdered bituminous coal Seacoal cokes when it is heated, plastically expanding and filling the interstices between the sand grains at the casting surface, thus preventing penetration of molten metal into the sand In addition, it burns, assuring that the atmosphere at the mold/metal interface is reducing; so that fayalite, a low

melting point solution of iron oxide and silica, does not form on steel castings Cellulose may be added in amounts up to 0.2% to absorb excess water and control sand expansion Polymers may be added to control the surface energy of water

so that it coats the sand grains more effectively, and soda ash may be added in very small amounts to control the pH of the sand mixture

Sand Testing. Sand is tested to assure its quality Normal tests include mechanical properties, usually expressed as

"compactability," moisture content, active clay content, and volatile content, tested as "loss on ignition." Incoming sand is tested for grain fineness number, a measure of both size and size distribution, and it may be tested for acid demand value, especially when certain resin binders are used for cores

Sand Reclamation. Sand is normally reclaimed and reused When the casting is poured, the sand next to the molten metal is heated above 600 °C (1110 °F), the water evaporates, and the clay at the interface can no longer be used as a bonding agent However, most of the sand and clay in the mold is not significantly heated by the solidifying metal and can be recycled The sand is separated from the castings, cooled, and recycled through the muller, where new sand (to compensate for sand lost by attrition), water, and clay is added If pieces of core in which the resin binder has not decomposed are present, they are crushed Metal and other debris (such as core rods) are removed by magnets, and the sand is screened before being added to the muller It is important to cool the sand to a temperature below 50 °C (120 °F) for the clay-water bond to work properly The amount of makeup water and clay added depends on the sand to metal ratio

in the mold; obviously, larger and heavier castings will burn out more moisture and clay than lighter castings

The recycled sand includes cores bonded with resins, which may not be completely burned out If the resins are not completely removed from the sand grains, they interfere with the clay-water bond The only way to completely remove these resins is to thermally reclaim the sand, heating it to temperatures greater than 650 to 825 °C (1200 to 1500 °F) in air These temperatures, of course, also burn out the bentonites; for this reason, only a portion of sand may be thermally reclaimed on any cycle Note that the core sand is added to the system on each cycle Frequently, the core sand addition is all that is needed to make up for sand losses in processing

Sand Handling Systems. In a high production foundry, transporting the sand from mulling to molding to pouring to shakeout to recycling and back to mulling requires a large sand handling system comprising conveyor belts, storage areas, and the sand mixing and molding equipment Sand should not be blown into silos or bins because blowing causes the sand

to segregate by size, destroying its size distribution Dust from sand systems must be removed from the foundry air, and sand temperature must be controlled Fines in the sand must also be removed because their high surface area will absorb clay and moisture and drastically change sand properties Controlling the sand system may be difficult if there are wide swings in the size of castings poured during a shift The sand-to-metal ratio is constantly varying, making it difficult to predict just how much moisture and clay to add to the muller

Dry Sand Molds

Dry sand molding is sometimes used to make large castings, although resin-bonded molds have largely taken over for this process In dry sand molding, the water is removed from the mold prior to pouring The bond is created both by using bentonites, and also through the use of other organic binders, such as dextrin, molasses, glutrin, or pitch The molds are baked in ovens prior to pouring to drive out the moisture Baking temperature and time depend on the size of the mold and the binder used Dry sand molds are generally stronger than green sand, and they usually have better dimensional stability than green sand in larger castings Sand grain size is usually coarser in dry sand molding, to allow the increased permeability of the mold to aid in mold drying Therefore, mold washes and coatings are often applied to improve surface finish

As an alternative to drying the entire mold, only the surface of the casting cavity may be dried These "skin-dried molds" are made from the same sand mixtures as dry sand molds, but only the surface that will contact the molten metal is dried

to a depth of 6 to 12 mm ( to in.)

Resin-Bonded Sand Molds

Over the last thirty years, a number of resins have been developed as binders for sand, primarily for making cores, although they are also used as molds These binder systems generally fall into one of the following categories: no-bake, cold box, protein, hot/warm box, and shell (Croning) In these processes, sand and binder are mixed together and formed around a pattern or placed in a core box The resin may be cured by heat, or it may be cured by the setting of the resin To accelerate the curing of the resin, catalysts may be added, either in solid, liquid, or gaseous form

No-Bake Processes. The term "no-bake" generally refers to all of the binder systems in which a gas catalyzes the resin while the sand is in contact with the core box or pattern Originally it referred only to the phenolic-urethane-amine process, which is still the most popular core resin in use In addition to the phenolic urethane process, other no-bake binder systems include furan acid catalyzed, phenolic acid catalyzed, ester-cured, and alkyd-urethane binders Furan binders depend on the use of furfuryl alcohol, either by itself or modified with urea, phenol, formaldehyde, or other additions The binders cure at room temperature by the addition of an acid catalyst, and they have reaction products, such

as nitrogen gas and water The amount of binder used must usually be minimized, both for economic reasons and to limit the amount of nitrogen gas picked up in ferrous castings, where it can case pinholes, and water vapor picked up in nonferrous castings, where it causes microporosity The catalyst used depends on the rate of cure required and the type of metal cast Urea-containing binders react relatively quickly, and a popular catalyst for them is phosphoric acid Sulfonic acids may be used for low-nitrogen binders Sulfuric acid may also be used, although it is not recommended for use with ductile iron, where the generation of sulfur fumes during pouring causes the formation of flake graphite Binder addition

is generally 0.8 to 2.0% of the weight of the sand, and the catalyst addition is between 20 and 60% of the binder weight

Phenolic binder systems are based on resole phenolic resins, which contain phenol, formaldehyde, and modifying additions in water These are cured with sulfonic acid, added at a level between 20 and 45% by binder weight

Ester-cured phenolic binders are popular in steel foundries because they contain no furfuryl alcohol or nitrogen and do not use sulfur for curing The two-part binder consists of the water-soluble alkaline phenolic resin, plus 15 to 30% liquid ester coreactants A blend of esters is used to provide the required setting time but only partially cross-link the resin Thus, the resin retains a small degree of thermoplasticity, which permits the sand to absorb some expansion and reduces expansion defects in the castings

Alkyd urethane binders, also referred to as oil urethane binders, are three-part systems used for large molds that require longer setting times The three components are the following: part A, an alkyd oil resin; part B, a liquid amine/metallic catalyst; and part C, a polymeric isocyanate, such as polymeric methyl di-isocyanate (MDI) Part A is usually used between 1 and 2% by weight of sand, and part B is used between 2 and 10% of the weight of part A Part C is added between 18 and 20% of part A by weight Curing takes place in two stages: first the polymeric isocyanate reacts with the alkyd resin at a rate controlled by the amine catalyst This reaction produces a bond that is strong enough to allow the pattern to be stripped The second stage is the oxidation polymerization of the alkyd resin by atmospheric oxygen, accelerated by the metallic elements in the catalyst Because oxygen must diffuse into the sand, this latter cure is a function of sand section thickness The oxidation reaction can be accelerated by heating to between 150 and 200 °C (300

to 400 °F)

The phenolic urethane no-bake system is a three-part system consisting of the following: part I, a phenol formaldehyde resin dissolved in solvents; part II, a polymeric isocyanate dissolved in solvents; and part III, a liquid amine catalyst The resin and the isocyanate combine to produce a urethane bond However, unlike the alkyd urethane system, there is no second stage reaction and a deep through cure is obtained Binder levels of 0.7 to 2.0% of the sand weight are recommended, with a ratio of 55 to 45 or 60 to 40, part I to part II The amount of catalyst required is between 0.4 and 0.8% of the weight of the part I resin

Cold Box Processes. The term "cold box" generally refers to processes in which the sand, binder, and catalyst are mixed together before the sand mixture is introduced into the core box or pattern; in this case curing begins immediately, and working time is limited Cold box processes have become popular in high-production foundries The original process

is the phenolic urethane process It is similar to the phenolic urethane no-bake process described above, except that the catalyst is introduced as a vapor after the core is blown, producing a rapid cure, which enhances production rates Both di-methyl-ethylamine (DMEA) and triethylamine (TEA) catalysts are used In this process, shown schematically in Fig 8, a gassing system is used to meter out a predetermined quantity of liquid, change it to vapor, combine it with a carrier gas in the correct ratio, and deliver it to the core box A scrubber system is also used to prevent the discharge of the offensive odor of the catalyst into the work area Core box design must include consideration of the location of the catalyst vents to assure complete coverage of the core by the catalyst

Fig 8 The cold box coremaking process The wet sand mix, prepared by mixing sand with the two-component

liquid resin binder, is blown into the core box The core box is then situated between an upper gas input manifiold and a lower gas exhaust manifold The catalyst gas enters the core box through the blow ports or vents and passes through the core, causing almost instantaneous hardening of the resin-coated sand The core

is ready for ejection from the core box after purging with clean air for a few seconds After the catalyst gas passes through the core, it leaves the core box through vents into the exhaust manifold From the gas exhaust manifold, the catalyst gas is piped to an appropriate disposal unit

The sulfur dioxide (SO2) process uses a furfuryl alcohol/formaldehyde resin mixed with peroxide, such as methyl ethyl ketone peroxide (MEKP), and additives The sand is mixed with these substances, and after being compacted in the core box, SO2 gas is passed through the core The sand has a long bench life, but sulfur-containing binders are not recommended for ductile iron, as they cause flake graphite to form in the metal

The free radical cure (FRC) process uses three parts: a vinyl unsaturated urethane oligomer, an organic peroxide, and a vinyl silane adhesion promoter The mixture is cured with SO2 gas, which breaks down the peroxide, which in turn releases free radicals that cure the resin The binder contains no nitrogen or water, which decreases gas defects in castings, but it is susceptible to mold erosion in ferrous castings

The ester-cured phenolic process is similar to that of the no-bake ester-cured system Here, however, the catalyst is methyl formate, carried by air The cured strength is not quite as high as the urethane cold box system, but it is satisfactory for many applications, especially for making molds as a replacement for green sand

Protein-Based Binders. One drawback to the use of organic binders is that the binders must be completely burned out during cooling of the casting if the sand is to be fully recovered In very large cores, or, in the case of nonferrous alloys where the mold and core temperatures do not get sufficiently high so that the resins burn completely, core butts remain that must be crushed and the sand thermally reclaimed (burned) at high temperatures In addition, cores not used for production must be put through the same process to reclaim the sand; unfortunately, the value of the resin is lost A recent development to overcome this problem is the use of a protein-based binder

The binder is a proprietary mixture of biopolymers, which do not break down thermally to produce toxic gases Metal oxide catalysts are used to accelerate the curing reaction In addition, the cured binder is water soluble, which means that unused cores and core butts may be treated with water to recover both the binder and the sand, without using thermal treatments This binder was designed for use in nonferrous (aluminum and magnesium) applications, so it readily breaks down thermally at the lower solidification temperatures encountered in these alloys Cores made with it have mechanical properties comparable to traditional organic binders The process also yields castings with similar surface finishes

Temperature Control during Coremaking. Temperature control is very important during the non-heat-cured coremaking processes Curing is a chemical reaction, which doubles in rate roughly every 10 °C (18 °F) In other words, attempting to use these processes at too low ambient temperature (below 20 °C, or 70 °F) will slow the reaction rate and cause an incomplete cure and low strength Increasing the amount of resin to compensate for the low temperature merely

increases the amount of gas generated during pouring and leads to gas defects in the casting Too high a temperature (above 30 °C, or 90 °F) shortens bench life and reduces sand properties For consistent production during both hot and cold weather, sand heaters and coolers are essential A sand heater should regulate the sand temperature to within 1 °C (2

°F) under normal conditions Coremaking carried out under close temperature control minimizes the use of binder and catalyst, produces more uniform cores, and decreases foundry scrap

In the hot box process, the binder-sand mixture is wet A liquid thermosetting binder and a latent acid catalyst are mixed with dry sand and blown into a heated core box Upon heating, the catalyst releases acid, which induces rapid cure; therefore the core can be removed in 10 to 30 s After the cores are removed from the pattern, the cure is complete, as a result of the exothermic reaction and the heat absorbed by the core Large cores require post curing in an oven

Hot box resins are classified simply as furans or phenolics Furans contain furfuryl alcohol, and phenolics are based on phenol, while furan-modified phenolics have both All conventional hot box binders contain urea and formaldehyde The furan hot box resin has a fast cure compared to that of the phenolic-type system and can therefore be ejected faster from the core box Furan resin also provides superior shakeout and presents fewer disposal problems because of the lack of phenol Typical resin content is 1.5 to 2.0%

A simplified hot box reaction mechanism is liquid resin + catalyst + heat = solid resin + water + heat The heat is provided by using a heated pattern or core box Catalyst selection is based on the acid-demand value and other properties

of the sand Sand temperature changes of 10 °C (18 °F) or five units in acid-demand value of the sand require a catalyst adjustment to maintain optimum performance Both chloride and nitrate catalysts are used The chloride catalyst is the more reactive Hot box pattern temperatures are 230 to 290 °C (450 to 550 °F), and the patterns are designed for use between 220 and 245 °C (425 to 475 °F) Running the pattern temperatures higher than this range produces patterns or cores with weakened or friable surfaces, which produce castings with poor surface finish Hot box resins have a limited shelf life and increase in viscosity in storage They should be stored in temperature-controlled areas Hot box catalysts have indefinite shelf life

Warm box resins are minimum-water (%) furfuryl-alcohol-type resins formulated for a nitrogen content of less than 2.5% Because the resin/sand mix exhibits a high degree of rigid thermoset properties when fully cured, little or no post-strip distortion or sagging occurs High hot and cold tensile properties are characteristic of warm box sands and generally permit a binder level between 0.8 and 1.8%, or about 20% less than the conventional hot box resin content Warm box catalysts are copper salts based primarily on aromatic sulfonic acids in an aqueous methanol solution The catalysts are unreactive with the resin at room temperature but form strong acids when heated Pattern temperatures range from 150 to

230 °C (300 to 475 °F); the process was designed for use at 190 °C (375 °F)

The shell molding or "Croning process" uses a very fine sand coated with a thermosetting resin to produce a shell

mold only about 10 mm (0.4 in.) thick The very fine sand produces an excellent surface finish on the casting Use of sand

of such fineness would not be possible in a conventional mold because mold permeability would be too low; however, in

a thin shell mold, it is not a problem

The sand grains are coated with phenolic novolac resins and hexamethylenetetramine The dry, free-flowing sand is coated by being mulled at 150 to 280 °C (300 to 535 °F) for 10 to 30 s The coated sand is then placed in a dump box, and the box is rotated leaving the sand in contact with the hot pattern for a period of time Pattern temperatures are generally between 205 to 315 °C (400 to 600 °F) The thickness of the shell produced depends on the amount of time the sand is in contact with the pattern Dump boxes produce shells of variable thickness, and many shell mold foundries use the more expensive, but more accurate and controllable, contour patterns, in which the sand is blown between a heated pattern plate and a heated profiled back plate Pattern halves are glued together before pouring and may be poured in either the vertical

or horizontal position Because the molds are only a thin shell, they are often supported during pouring, either in fixtures

or in loose sand Patterns may also contain shell cores

In making shell cores, the sand is injected into a heated core box and then dumped out after a short time Again, the thickness of the shell is dependent on the time in contact with the heated core box Shell molds and cores generally produce castings with better surface finish and more accurate dimensions than other organic resin-binder processes

Quality Control of Resin-Binder Processes. In all organic resin processes, care must be used to prepare the pattern

or core box and to maintain it free from resin buildup during the production campaign It is especially important to inspect the vents on a regular basis to be certain that they remain clear of resin products that could clog the vents