Pattern Heating Although the use of a parting spray is effective, a better first step is to heat the patterns to to 11 °C (10 to 20 °F) above the temperature of the molding sand Cold patterns will cause the moisture in the molding sand to condense on the pattern face, which makes the sand stick to the pattern Many molding machines not have any provision for heating the pattern during the production run Whether the molding machine has the heating capability or not, the pattern should be preheated to the proper temperature prior to the start of the production run to assist in a rapid start-up References cited in this section High Pressure Molding, 1st ed., American Foundrymen's Society, 1973 D Boenisch, "Strength Problems in High Pressure Compacted Sand Molds," Paper presented at the Disamatic Convention, Disamatic Inc., 1971, p 69-84 D Boenisch and B Koehler, Sand Compaction and Grain Rupture in High Pressure Molding Machines, Giesserei, Vol 63 (No 17), Aug 1976, p 453-464 Mold Finishing After the mold has been compacted and the pattern removed, the mold is ready for the finishing operations These operations usually consist of blowing out any loose sand, looking for any molding defects, drilling the sprue cup (if applicable), and setting any necessary cores Once the finishing operations are complete, the cope can be accurately placed on the drag and the mold sent to the pouring station Mold blowoff is one of the areas that can cause surface finish problems if not property controlled Excessive amounts of air blown onto the mold can cause localized drying of the mold surface On the other hand, if the air contains large amounts of moisture, the mold face can become excessively wet, giving rise to rough finish and burn-in penetration, just as excessive amounts of water in the molding sand or excessive amounts of parting spray will Core setting is another part of the process that has undergone tremendous change in the last few years The increased demand for more accurate castings has affected cores and core setting just as it has molding With mold hardness in the 85+ range, it is no longer possible to press an oversize core into undersize core prints From the viewpoint of casting accuracy as well as cleaning room costs, it is equally unacceptable to place undersize cores in an oversize core print Modern core processes allow the possibility of making the same size core every time, just as the same size mold can be made every time Thus, it becomes apparent that every cavity in the corebox and every impression on the pattern must be as close as possible to the same dimensions This obviously places greater demands on the supplier of patterns and coreboxes as well as on the process itself Mold closing is the next step in the operation Some molding machines close the mold inside the machine, while others close the mold just outside the machine Still others utilize a separate piece of equipment to perform this function totally external to the molding machine The halves should not be allowed to remain separated any longer than absolutely necessary Separation of the mold halves for excessive periods of time will allow the mold faces to dry, and this can lead to cuts, washes, and a general degradation of the casting surface Transportation of the mold to the closing station is critical Jarring of the mold can cause green sand pockets to break away from the mold In some cases, the pocket may not break away until the mold is poured; thus, the mold, the cores, and the metal are wasted Rough transportation can easily cause heavy cores to shift off location, which can cause errors in casting dimensions, broken cores, and excessive metal around coreprints Mold closing is just as critical as mold transportation, and the same basic rules apply The mold must be treated as smoothly and gently as possible to avoid the same type of defects The mold guiding mechanism is as important at mold closing as it is during manufacture of the mold Too often the accuracy and smoothness with which the mold is closed is overlooked Again, the results can be drops and core movement as well as mold shift, crush, casting dimension problems, and so on After the mold is closed, mold transportation again becomes important The finished mold must be carefully transported to the pouring area, or problems such as those already mentioned will likely occur After the mold has been poured, the molten metal must be given time to solidify and coot to the proper temperature before it is removed from the mold During the solidifying process, the mold halves must be held solidly together Any movement will introduce the possibility of casting inaccuracies or increased demands for feed metal or both The loss of casting accuracy has obvious consequences The requirement for additional feed metal has the consequence that shrinkage cavities may form and will probably not be evident from the casting exterior Cooling time after solidification is critical for many casting/alloy combinations Insufficient cooling time can lead not only to dimensional problems due to lack of casting rigidity but also to hardness and internal stress problems, even to the point of cracking the casting Shakeout After the castings have cooled sufficiently, they can be shaken out, that is, separated from the sand mold Shakeout devices are available in a number of different configurations Many of the devices available are of the flat deck, vibratory type They range from normal intensity, frequency, and travel to high-intensity units that utilize a very short travel but high frequency Some shakeout units are rotary in nature and, depending on design, can also provide the added function of cooling the sand Another type is the vibratory barrel Deck-type shakeouts (Fig 16) are available in a number of different configurations for various applications The first is the stationary type Stationary refers to the casting and sprue, not the shakeout itself This type of shakeout is normally used by bringing the mold to the shakeout device; therefore, its primary application is for larger molds and low-tomedium production lines The deck-type shakeout is also available as a unit that provides the function of conveying the castings from one end of the unit to the other As mentioned earlier, either type is available in a variety of strokes, intensities, and frequencies Selection of the shakeout is a function of casting design Heavier castings can be quite successfully run using a longer-stroke shakeout, while thin-wall castings may require a short-stroke high-frequency unit to prevent breakage or damage to the casting Rotary-type shakeouts (Fig 17) are also available in different configurations The sand may exit at the same end that the sand and castings enter the unit, or it may exit at the opposite end This type of shakeout also provides the function of conveying the castings from one end of the unit to the other Rotational speed is adjustable on most units to allow flexibility in shakeout intensity In general, as rotational speed decreases, intensity decreases and castings are less likely to be damaged Light thin-section castings may not be suitable for this type of shakeout Although the castings themselves may not damage each other, the sprue is sometimes heavy enough that it can damage the castings Fig 16 Flat deck vibratory type shakeout device Fig 17 Rotary-type shakeout system Rotary plus cooling type shakeouts are also available in a configuration that not only holds the sand and castings together for an extended period but also affords the opportunity to cool the molding aggregate This type of device is designed such that the castings and sand are held together throughout the length of the drum (Fig 18) The castings and sprue aid in the breakdown of lumps Sand temperature samples are normally taken somewhere along the length of the drum to determine the amount of water necessary for cooling the sand The cool sand in turn cools the castings, often down to a temperature that can be comfortably handled at the exit of the drum Sand and castings are separated at the exit of the drum As with rotary sh akeouts, sprue can damage certain types of castings, especially as wall sections become thinner In-mold cooling time can become more critical when the castings and sand are kept together in the cooling device When castings are too hot, hardness problems can result In some cases, stresses can also be introduced into the castings because of the rapid quenching of the casting in the molding sand Fig 18 Rotary plus cooling type shakeout system in which the castings and water-cooled mold sand are separated at the drum exit Vibrating drum type shakeouts (Fig 19) combine the operating principles of rotating drum and vibrating deck units The vibrating section is round in cross section, but it does not rotate Instead, a rotating action is imparted to the sand and castings by the vibratory action As the drum vibrates, material is constantly agitated to produce particle migration in both axial and transverse directions The drum can be designed to provide a very rapid blending action or a gentle folding action, depending on process requirements Because air can be forcibly exhausted from the drum and because the surface of the sand within the drum is constantly changing, a limited amount of cooling is possible Additional information on shakeout is available in the article "Shakeout and Core Knockout" in this Volume Fig 19 Front (a) and side (b) views of a vibratory drum type shakeout system Sand/Casting Recovery What happens to the sand after shakeout is of great importance to the design and operation of the system (see the section "Sand Reclamation" in this article) Historically, the sand is returned from the shakeout to a storage bin, where it is kept until the next time it is mixed with additional clay, water, and carbonaceous materials Unfortunately, sand-to-metal ratios of 3:1 to 6:1 are quite common Sand-to-metal ratios in this range, combined with cooling times that allow the castings to become cool enough to separate from the sand, can easily create return sand temperatures of 120 °C (250 °F) and above (Ref 6) High sand temperatures cause innumerable problems not only with regard to molding and surface finish but also for the system itself Bentonite does not absorb water and become plastic to develop the necessary cohesive and adhesive strengths when sand is above 45 to 50 °C (115 to 120 °F) Therefore, the molding sand must be below these temperatures long enough for the muller to provide the necessary input of energy to coat the sand grains properly Hot sand, usually above 50 °C (120 °F), is difficult to temper and bond, and when above 70 °C (160 °F), hot sands are impossible to rebond (Ref 7) Unfortunately, sand is not easily cooled, especially in the quantity necessary to keep a molding line running Molding sand is a relatively good insulator and therefore tends to hold heat for long periods of time Storage quantity is therefore not the answer Not only does sand stored in a bin hold its heat for long periods of time but it also cools from the outside toward the center As it cools in this manner, moisture tends to migrate toward the cooler sand, which causes it to cake on the outside walls As time goes on, the caking on the outside wall becomes thicker until only a small portion of the sand is actually being circulated through the system Vibrators and bin poppers have been designed and can be of some help in combatting this rat holing tendency of return sand bins, but the ideal situation would be to cool the sand prior to storage Evaporative cooling is the only practical method of cooling the amount of sand needed in green sand systems Hot sands must therefore have water added in amounts that exceed those required for tempering if both cooling and tempering are to take place In addition, an ample supply of air must be present to carry away the heated water vapor The cooling of molding sand may be regarded as a two-stage process, although no sharp line separates the stages At temperatures in excess of approximately 70 °C (160 °F), added water causes a flash evaporation cooling effect (Ref 6) Temperature will continue to decrease fairly rapidly to about 60 °C (140 °F), but more slowly after that As sand temperature approaches ambient temperature, further cooling becomes more difficult and time consuming Conditions of high ambient temperature, especially when combined with high ambient humidity, can substantially reduce the effectiveness of cooling devices Therefore, ambient conditions should be considered carefully when sand systems are being designed or modified Some mullers have the capability of blowing air through the sand mixture and will cool the sand very effectively However, there are some disadvantages to this method It must be kept in mind that mulling (coating sand grains with bentonite) does not take place until the mixture is cool enough to be tempered and bonded (Ref 6) Cooling time must therefore be added to the mulling time Although western bentonite provides the mold stability needed by most foundries, it does require more time and energy to absorb water and develop the necessary properties (Ref 8) Thus, the job of the muller or mixer becomes even more difficult and time consuming The storage bin will still have the tendency to rat hole, thus returning sand more quickly and hotter to the muller and further aggravating the situation Control of solid additives and water becomes more difficult as the molding sand becomes hotter However, this is not an impossible situation; this method is used quite effectively in a number of foundries A few steps can be taken to provide some amount of cooling to the return sand in an existing system and to keep equipment costs as low as possible For example, water can be fogged on the return sand, preferably as early as possible Chains can then be dragged through the aggregate and/or plows can be used to turn the mixture over Additional air can be introduced by fans or other sources to enhance cooling Elevators can be vented to enhance air flow, but this provides little help because the sand is being conveyed in solid buckets The only assistance realized will be at the transfer points Although these and similar methods help to reduce return sand temperature, they are generally of only marginal value An effective job of cooling return sand normally requires the addition of water, along with forced air being blown or pulled through the aggregate by some type of auxiliary cooling device A number of auxiliary cooling devices are available that utilize forced air for evaporative cooling These units should always be placed as close as possible to the casting shakeout In fact, one type of unit, the shakeout-cooling drum, combines the functions of shakeout and sand cooling Cooling the sand at or near the shakeout enables tighter control, reduces the tendency toward rat holing in the return sand bin, and reduces the demand for cooling on the muller Because many muller designs make no provision for cooling, adequate external cooling is not only desirable but necessary Cooling the sand as early as possible reduces the total cycle time of the muller by reducing or eliminating the time necessary for cooling and provides a method for making mulling time more efficient Southern bentonite can be mulled in very quickly if the aggregate temperature is low enough As mentioned earlier, western bentonite is not mulled in very quickly, because it must swell by such a large amount (Ref 8) For this reason, it is advisable to keep the bentonite swelled and as active as possible Many of the auxiliary cooling devices can be controlled to the point where the level of return sand moisture will be such that the western bentonite will remain activated Normally, a retained moisture level of 1.8 to 2.0% will not only keep bentonites activated but will also reduce the amount of dusting at transfer points, thus reducing the load on dust collection equipment Cooling Devices As mentioned earlier, it is possible to realize some cooling by adding water to a return sand belt and then using some method of turning the sand over at various places along the length of the belt There are mechanized devices (Fig 20) that perform similar functions and provide air flow through the sand The effectiveness of these methods is often somewhat limited because of conveyor belt lengths; as belt lengths become shorter, the method becomes less effective Difficult sand temperature problems will require more serious measures Fig 20 Mechanized sand cooler used in high-production molding lines Drums used as cooling units are among the oldest of the effective devices (Fig 21) A cooling drum does not keep the sand and castings together; instead, this is a separate piece of equipment through which sand from the shakeout flows As with other cooling devices, water must be added to the molding sand to allow the air moving through the drum to provide the necessary cooling by evaporation Fig 21 Cutaway view of a sand cooling drum system Sequence of operations proceeds from right to left: 1, hot shakeout and spill sand enter, and helical flights convey sand forward to begin blending process; 2, cascading effect provides sand cooling as well as sand homogenization; 3, blended and cooled sand is discharged onto perforated cylinder, which screens off tramp metal and core butts while passing sand; 4, replaceable screen passes sand to discharge onto conveyor; 5, lumps that not pass final screen carry across to lifter paddles for discharge into overburden chamber The fluid bed cooler (Fig 22) is a vibratory type of conveyor through which the sand flows in a more or less continuous but controlled stream Air is pumped through the sand from underneath, causing the necessary evaporation and cooling Fig 22 Schematic of a fluid bed cooler Figure-Eight Cooler Similar to the continuous muller shown in Fig 13, the figure-eight cooler is designed so that air can be pumped through it and provide the necessary cooling This device has been used directly above the muller, but a more desirable location would be as close to the shakeout as possible for the reasons already mentioned Regardless of the equipment used, it is necessary to control the moisture additions so that sufficient moisture is available for cooling and bentonite activation without getting the return sand so wet that problems will be experienced with plugging up of the sand system The movement of air through the aggregate will almost certainly remove some of the finer material The higher the velocity of air movement, the better the cooling, but also the greater the loss of that fine material The loss of a certain amount of that material (such as dead, burnt clay and ash) can be beneficial Unfortunately, a number of beneficial materials can also be lost, such as the finer grains of sand, coal dust, and bentonite Any cooling device should be planned with a solids separator on the exhaust air so that these materials can be collected and fed back into the system at a controlled rate This will improve surface finish, and trapping and using the bentonites and coal dust will provide economic benefits Metal Separation and Screening The shakeout does the primary job of separating the sand from the sprue and castings Smaller pieces of metal can easily slip through the grating of the shakeout device and be processed along with the sand This will cause casting defects, and it may damage the equipment Therefore, it is advisable to remove as much of the tramp metal as possible When magnetic metals such as most irons and steels are being cast, the job is relatively easily accomplished with magnets The suggested practice is to install an over-belt magnet somewhere along the length of a conveyor belt and a pulley magnet at the discharge end of the same belt Placing both magnets on the same belt allows more complete separation of the magnetic particles Nonmagnetic alloys present a different problem Devices are available that separate the metallic particles based on density differences, but the most common method is to use screens Multiple screens are often used, and the mesh size from screen to screen becomes progressively finer Lumps are found in all sand systems and consist of system sand or core parts that have not been sufficiently heated to break down the binder For this reason, it is necessary to have a good screen in all systems The opening size in the screen should be as fine as is practical for the system involved Two basic types of screens are in use: flat deck and rotary The flat deck type is usually vibratory in nature and has the added function of providing further lump reduction as well as the screening function The rotary type of screen is normally a large barrel that continually rotates The exterior of the barrel has the desired size of holes in it to provide the screening action Because of the tumbling action within the screen, lump reduction similar to that obtained with the vibrating flat deck can be expected In both cases, the size of the screen should be as fine as is practical After the sand has been cooled, the tramp metal removed, and the core butts and lumps removed, the sand is ready to be returned to the storage hopper to be used again References cited in this section J.S Schumacher and R.W Heine, The Problem of Hot Molding Sands 1958 Revisited, Trans AFS, Vol 91, 1983, p 879-888 C.A Sanders, Foundry Sand Practice, American Colloid Company, 1973, p 441 J.S Schumacher, R.A Green, G.D Hanson, D.A Hentz, and H.J Galloway, Why Does Hot Sand Cause Problems?, Trans AFS, 1974, p 181-188 Computer-Aided Manufacture Recent years have seen a rapid advancement in the use of data processing units and data communication These advancements have made possible almost complete and instantaneous record keeping and, equally important, trend recognition The technology is advancing rapidly; there are systems currently in place that record on a continual basis the amounts of return sand, new sand, bentonite (or premix), and water that go into each batch of sand In many cases, mixing time and maximum current draw of the muller are also recorded With some systems, compactability can also be recorded In any case, output data, such as compactability and muller current draw, can be stored for a period of time, and a trend analysis can be done automatically Molding machines have also become more sophisticated With microcomputers and programmable controllers being used to control machine movements, it is possible to read the pattern number automatically when the pattern is installed Using information that had previously been stored in the memory of the computer or controller, the molding machine can optimize its molding parameters for the individual pattern A hypothetical case will illustrate the extent of the available information During a shift, a new pattern is installed on the molding machine The operator tells the machine that 1250 molds are needed Optimum molding parameters, poured weight, necessary cooling time, and so on, have already been determined during earlier runs and stored in the computer At any point during the run, the operator or someone operating a distant host computer can query the molding machine to find out which mold is going to reach shakeout next, how much cooling time it had, how much metal is required to complete the production run, how much time will be required to complete the production run based on existing molding rates, how many cores will be required to complete the run, how many molds have been made and/or poured, and so on These outputs can be used as control signals More water or less water can be added to the sand cooler when sand from the new molds reaches the cooling device Molding sand compresses more in the molding chamber/flask as sand becomes wetter (higher compactability), thus trend analysis can be done by recording mold compression during compaction, and the resulting information can be fed back to the sand preparation equipment The exact position required for an automatic pouring device can be set by the molding machine Daily production data reports can be printed out that will give information on each run; this information includes the number of castings, production rate, productivity, number of cored molds, and reasons for downtime (such as waiting for sand or metal) In the event of machine difficulty, the machine can help troubleshoot itself It is not only possible but practical to allow the molding machine to exchange data with a remote location (via telephone lines) if assistance in troubleshooting is needed The quantity of information that is available and transmittable depends on the mechanical and electronic design of the equipment Some units are designed to allow one-way communication (output), while others are designed to allow twoway communication (output and input) In the latter case, it is possible for a remote location to control some or all inputs to the production equipment These remote locations can consist of keyboard inputs from a host computer or even data output from other pieces of equipment The type of information available (either as inputs or outputs), the form the information is in, and the communication protocols may vary greatly among manufacturers It is therefore necessary to research the technical information available from each manufacturer to determine the best way for the various pieces of equipment to communicate and the best way to handle the information obtained Additional information on the role of computers in the manufacture of green sand molds is available in the Section "Computer Applications in Metal Casting" in this Volume Sand Reclamation Michael Zatkoff, Sandtechnik, Inc Reclamation is defined by the American Foundrymen's Society (AFS) Sand Reclamation and Reuse Committee 4-S as the physical, chemical, or thermal treatment of a refractory aggregate to allow its reuse without significantly lowering its original useful properties as required for the application involved To achieve this objective, one must evaluate the type of sand entering the reclamation system, the binder system used, and the area for its reuse This section will provide a brief review of sand reclamation systems for both chemically bonded (resin bonded) sands and clay-bonded sands (green sands) Detailed information on sand molding principles and processes can be found elsewhere in this Volume Reclamation of Chemically Bonded Sand The primary requirement of any reclamation system is to remove the resin coating around the sand grains This involves abrasion and attrition to break the bond, as well as classification to remove the fines that are generated The three basic reclamation systems are thermal, dry, and wet Selection of a system depends greatly on the type of organic binder to be removed from the sand grains More detailed information on organically bonded sand systems can be found in the article "Resin Binder Processes" in this Volume Wet Reclamation Systems Wet reclamation systems were used for clay-bonded system sands in the 1950s, but are now used for silicate binder systems only Silicate systems are very difficult to reclaim by dry processes and are impossible to reclaim in thermal systems This is because silicate is an inorganic system that melts rather than burns in the furnace The complete system includes lump-breaking and crushing equipment, an attrition unit, wet scrubber, dewatering system, and dryer The systems require about one pound of water per pound of sand reclaimed, and in some cases the water can be discharged directly into municipal sewer lines Most installations allow 100% reuse of the reclaimed sand, with makeup sand as the only new sand addition Dry Reclamation Systems Many factors determine the degree of cleanliness required in a reclaimed sand These factors include the type of resin system used for rebonding, the sand-to-metal ratio, the type of metal poured, the condition of the reclaimed sand, the type of new sand used, and the ratio of new sand to reclaimed sand Attrition reclaimers break down the sand lumps to a smaller grain size Some fines are removed, but the binder is not removed completely from the surfaces of the sand grains In most cases, these units produce a sand that requires a higher concentration of new sand when the attritor is coupled with a sand scrubber, as described below Additional scrubbing is sometimes required, and there are basically two types of scrubbers: mechanical and pneumatic Selection between the two types is primarily a question of wear, ease of maintenance, and energy consumption because the units provide comparable performance in terms of scrubbing action Pneumatic Scrubbing Figure 23 shows one cell of a pneumatic scrubber Sand is introduced by gravity at the top of the unit, and it flows down around the blast tube High-volume low-pressure air from a turbine blower flows through the nozzle and lifts the sand up through the blast tube to the target plate The sand grains undergo intense attrition in the tube by impacting on each other; further attrition occurs at the target as binder is removed from the sand grains These fines and resin husks are then removed from the system by a classification dust collection system Scrubbed sand falls from the target and is deflected to the next cell or is kept within the same cell for further scrubbing The degree of cleanliness attained is determined by the retention time in the cells (controlled by the deflector plate) and the number of cells Sand exiting the final cell should be screened to remove any foreign material that may be present in the refuse sand with dendrite arm spacing A solidification structure with dendrites parallel to the ingot axis yields optimal results However, this is not always possible A good ingot macrostructure requires a minimum energy input and, accordingly, a minimum melting rate Optimal melt rates and energy inputs depend on ingot diameter This means that the necessary melting rate for large-diameter ingots cannot be maintained for axis-parallel crystallization Figure 27 shows melting rates for various steels and alloys as a function of ingot diameter These are empirical values that were obtained from experience in operation These melting rates gave low microsegregation while achieving acceptable surface quality The importance of pool depth was also investigated by numerous researchers (Ref 28, 29, 30, 31, 32) Fig 27 Typical VAR melting rates for various steels and nickel- and cobalt-base superalloys References cited in this section 27 W.A Tiller and J.W Rutter, Can J Phys., Vol 311, 1956, p 96 28 W.H Sutton, in Proceedings of the Seventh International Vacuum Metallurgy Conference (Tokyo), The Iron and Steel Institute of Japan, 1982, p 904-915 29 J Preston, in Transactions of the Vacuum Metallurgy Conference, American Vacuum Society, 1965, p 366-379 30 A.S Ballentyne and A Mitchell, Iron-making Steelmaking, Vol 4, 1977, p 222-238 31 S Sawa et al., in Proceedings of the Fourth International Vacuum Metallurgy Conference (Tokyo), The Iron and Steel Institute of Japan, 1974, p 129-134 32 J.W Troutman, in Transactions of the Vacuum Metallurgy Conference, American Vacuum Society, 1968, p 599-613 Ingot Defects In spite of directional dendritic solidification, such defects as tree ring patterns, freckles, and white spots can occur in a remelted ingot This can lead to rejection of the ingot, particularly in the case of superalloys Tree ring patterns can be identified in a macroetched transverse section as light-etching rings They usually represent a negative crystal segregation Tree ring patterns seem to have little effect on superalloy material properties (Ref 33) Tree ring patterns are the result of a wide fluctuation in the remelting rate In modern remelting plants, however, the remelting rate is maintained at the desired value by means of an exact control of the melting rate during operation, so that the melting rate exhibits no significant fluctuations Freckles and white spots have a much greater effect on material properties than tree ring patterns, especially in the case of superalloys Both defects represent an important cause of the premature failure of turbine blades in aircraft engines Freckles are dark-etching circular or nearly circular spots that are generally rich in carbides or carbide-forming elements The formation of freckles is usually a result of a high pool depth or movement of the rotating liquid pool The liquid pool can be set into motion (rotation) by stray magnetic fields during remelting Freckles can be avoided by maintaining a low pool depth and by avoiding disturbing magnetic fields through the use of a coaxial current supply White spots are typical defects in VAR ingots They are recognizable as light-etched spots on a macroetched surface They are low in alloying elements, for example, titanium and niobium in Alloy 718 There are several mechanisms that could account for the formation of white spots (Ref 34, 35): • • • Relics of unmelted dendrites of the consumable electrode Pieces of the crown that fall into the pool and are not dissolved or melted Pieces of the shelf region transported into the solidifying interface All three of the above mechanisms, individually or combined, can be considered as sources of white spots This indicates that white spots cannot be completely avoided during vacuum arc remelting To minimize the frequency of occurrence of these defects, the following conditions should be observed during remelting: • • • Use the maximum acceptable metal rate permitted by the ingot macrostructure Use a short arc gap to minimize crown formation and to maximize arc stability Use a homogenous electrode free of cavities and cracks References cited in this section 33 R Schlatter, Giesserei, Vol 61, 1970, p 75-85 34 A Mitchell, in Proceedings of the Vacuum Metallurgy Conference, Pittsburgh, PA, 1986, p 55-61 35 F.J Wadier, in Proceedings of the Vacuum Metallurgy Conference, Pittsburgh, PA, 1984, p 119-128 Process Variables Atmosphere The heat generated by melting of the metal in vacuum arc remelting results from the electric arc between the consumable electrode and the liquid pool on the top of the ingot The pressure in the remelting vessel is usually of the order of 0.1 to Pa (10-3 to 10-2 mbar, or 7.5 × 10-4 to 0.0075 torr) In some exceptional cases, the melting is also carried out under inert gas with a pressure up to 10 kPa (100 mbar, or 75 torr) Evaporation losses of volatile alloying elements are minimized at this pressure Melt Rate As mentioned earlier, the melt rate is an important factor in the quality of the ingot macrostructure A modern VAR furnace is therefore equipped with a load cell system to measure the weight of the electrode at a particular interval of time The actual values of the melt rate are compared by computer with the desired set values Any difference between the measured melt rate and the desired value is eliminated by the proper accommodation of the power input Figure 28 shows the melt rate and the melting current at start-up, during steady-state melting, and during hot topping Startup and hot topping are usually controlled based on time The melting phase is controlled based on weight Fig 28 Process control parameters and set point functions in vacuum arc remelting Hot topping begins when a preselected residual electrode weight is reached A computer controls the melting parameters, which are stored in the form of up to 250 different recipes in the computer Figure 29 shows a simplified version of the computer furnace control With this, melting rate can be controlled with a precision of better than ±2% The computer also provides documentation in the form of tables and graphs for the relevant process parameters Fig 29 Schematic of automatic melt control system Remelting Variations Under Vacuum Apart from the remelting of a consumable electrode in a water-cooled copper crucible, there is a recent development of the vacuum arc remelting process, namely vacuum arc double electrode remelting (VADER) Figure 30 shows the basic design of the VADER process with a static crucible The arc is struck between the two horizontal electrodes that are to be melted Fig 30 Schematic of the VADER process As in vacuum arc remelting, the metal drops fall into a water-cooled copper mold Bath temperature, and therefore pool depth, can be very closely controlled Remelting can be done with minimal superheating; segregation is thus minimized The advantages of the VADER process over vacuum arc remelting are as follows (Ref 36): • • • • • Very low or no superheating of the pool and high rate of nucleation, producing a fine grain structure The lowest possible influence of magnetic fields on melting bath movement No condensation formation due to evaporation of liquid elements on the crucible walls Lower specific energy consumption Good ultrasonic testability due to the fine macrostructure of the ingot Reference cited in this section 36 J.W Pridgeon, F.M Darmava, J.S Huntington, and W.H Sutton, in Super-alloys Source Book, American Society for Metals, 1984 Vacuum Arc Skull Melting and Casting F Müller and E Weingärtner, Leybold AG, West Germany Titanium investment casting has recently gained the same importance as the precision casting of superalloys (see the article "Titanium and Titanium Alloys" in this Volume) Titanium skull melting originated at the Bureau of Mines in Albany, Oregon The first castings were made in 1953, although possibilities had been announced as early as 1948 and 1949 In the late 1950s, this technology was applied by research institutes, which were looking for a practical means of liquefying and pouring uranium into graphite molds, for example, to produce uranium carbide An early industrial vacuum arc skull melter was built in 1963 for the continuous production of uranium carbide This furnace had a crucible volume of approximately 0.01 m3 (0.35 ft3) and used a nonconsumable graphite electrode to liquefy the uranium pellets fed into the crucible The crucible tilting system was hydraulically driven The molds were stationary It was not until 1973 that one of the first skull melters for titanium went into operation; this furnace started production in 1974 in West Germany State-of-the-art titanium vacuum arc skull melting furnaces are often equipped with turntable systems for centrifugal casting (up to 350 rpm) Casting weights of more than 1000 kg (2200 lb) are possible Vacuum arc skull melting and casting is used for many titanium investment castings for aircraft, aerospace, medical, and chemical applications Electron beam skull melting is also used for titanium alloys (see the section "Electron Beam Melting and Casting" in this article) Furnaces Vacuum arc skull casting furnaces basically consist of a vacuum-tight chamber in which a titanium or titanium alloy electrode is driven down into a water-cooled copper crucible The dc power supply provides the fusing current needed to strike an electric arc between the consumable electrode and the crucible Because the crucible is water cooled, a solidified titanium skull forms at the crucible surface, thus avoiding direct contact between melt and crucible Once the predetermined amount of liquid titanium is contained in the crucible, the electrode is retracted, and the crucible is tilted to pour the melt into the investment casting mold positioned below For optimum mold filling, the mold can be preheated and/or rotated on a centrifugal turntable Figure 31 shows the operating principle of a modern 50 kg (110 lb) vacuum arc skull melting furnace At an operating pressure of approximately Pa (10-2 mbar, or 0.075 torr), the specific working current ranges from approximately kA/kg for small furnaces to about 0.2 kA/kg for large pouring weights This batch-type skull melting furnace allows for cycle times of approximately h for a full 50 kg (110 lb) pumping/melting/casting cycle, and in principle three consecutive pours can be obtained from one electrode This furnace basically consists of a vacuum chamber, an arc voltage-controlled electrode drive system, a skull crucible, a centrifugal casting system with stepless adjustable turntable speed, an automatically sequenced vacuum pump system, a power supply, and an electrical control system with control desk Fig 31 Schematic of a modern 50 kg (110 lb) vacuum arc skull melting and casting furnace 1, fast retraction system; 2, power cables; 3, electrode feeder ram; 4, power supplies; 5, consumable electrode; 6, skull crucible; 7, tundish shield; 8, mold arrangement; 9, centrifugal casting system; 10, chamber lid carriage The cylindrical vacuum chamber is equipped with two large dished doors that support the crucible with the tilting mechanism, the mold platform with the centrifugal casting system, and the casting tundish with its cover The crucible support system with an additional detachable device is also used for electrode loading The chamber is jacketed for water cooling in regions that are subject to heat radiation A top flange with a throat carries the electrode chamber and the electrode feeding system Viewing ports allow for video monitoring of the melting and pouring A vacuum pumping port is located in the cylindrical portion of the chamber Figure 32 shows a semicontinuously operated vacuum arc skull melter for charge weights of up to 1000 kg (2200 lb) The principal difference between this furnace and the smaller model (Fig 31) apart from capacity-related layout features is the rectangular vacuum chamber Again, the crucible and the tilting mechanism are carried by a dished door Fig 32 Schematic of a modern semicontinuously operating vacuum arc skull melter for charge weights of up to 1000 kg (2200 lb) 1, fast retraction system; 2, power cables; 3, power supplies; 4, electrode feeder ram; 5, consumable electrode; 6, skull crucible; 7, crucible carriage; 8, tundish shield; 9, mold arrangement; 10, vacuum pumping system; 11, centrifugal casting system In this furnace design, the centrifugal casting system is introduced from the bottom of the chamber to allow the horizontal connection of a separate cooling chamber, if desired Molds are loaded from the back side into the chamber and can be discharged through a front door, which also allows the use of a cooling and charging chamber with lock valves for continuous mold transport flow Modern vacuum arc skull melting furnaces are usually equipped with: • • • • • • • Coaxial power feed directly to the skull crucible to avoid electromagnetic fields that can disturb the melt bath Programmable control systems for crucible tilting to allow repeatable pouring profiles for consistent parameters Highly accurate electrode weighing system for precise determination of pouring weights XY adjustment system for coaxial positioning of the electrode in the skull crucible Compact air-cooled power-supply modules of high capacity for achieving high melt rates, thinner skulls, and correspondingly increased yields above 80% Proven vacuum pumping and measuring systems Forced argon cooling systems for faster mold cooling Electron Beam Melting and Casting W Dietrich and H Stephan, Leybold AG, West Germany Electron beam melting and casting technology is accepted worldwide for the production of niobium and tantalum ingots weighing up to 2500 kg (5500 lb) in furnaces with electron beams of 200 to 1500 kW Another application in East Germany and other Soviet bloc countries is the production of steel ingots weighing 3.3 to 18 Mg (3.6 to 20 tons) using electron beams of up to 1200 kW Furnaces of up to 2400 kW in electron beam power have been used since 1982 for recycling titanium scrap to produce 4.8 Mg (5.3 ton) slabs 1140 mm (45 in.) wide Furnaces of 200 to 1200 kW are used to refine nickel-base superalloys Other metals, such as vanadium and hafnium, are melted and refined in furnaces between 60 and 260 kW Approximately 150 furnaces with melting powers ranging from 20 to 300 kW are in operation in research facilities These furnaces are used in the development of new grades and purities of conventional and exotic metals and alloys, for example, uranium, copper, precious metals, rare-earth alloys, intermetallic materials, and ceramics The total power of installed electron beam melting and casting furnaces worldwide was approximately 25,000 kW at the end of 1987 Electron beam melting and casting includes melting, refining, and conversion processes for metals and alloys In electron beam melting, the feedstock is melted by impinging high-energy electrons Electron beam refining takes place in vacuum in the pool of a water-cooled copper crucible, ladle, trough, or hearth In electron beam refining, the material solidifies in a water-cooled continuous casting copper crucible or in an investment ceramic or graphite mold This technology can be used for all materials that not sublimate in vacuum Competing processes include sintering (for example, for refractory metals), vacuum arc melting and remelting (for reactive metals and superalloys), and electroslag melting and vacuum induction melting (for superalloys, specialty steels, and nonferrous metals) Some advantages and limitations of the competing vacuum processes are given in Table Additional information on some of these processes is available in the sections "Vacuum Arc Remelting (VAR)," "Electroslag Remelting (ESR)," and "Vacuum Induction Melting (VIM)" in this article Table Comparison of characteristics of electron beam melting and competing processes Metal Sintering Electron beam melting Vacuum arc melting Advantages Limitation Advantages Limitations Advantages Limitations Tungsten, molybdenum Small grain size most often used Refining limited; small batches; high energy consumption Moderate grain size; acceptable workability large ingots; low energy consumption Refining limited; costly electrode preparation; melting dangerous Highest possible purity; economical feedstock preparation; large ingots; low energy consumption Large grain size; brittle product; very rarely applied Tantalum, niobium Small grain size; good workability Same as above; rarely applied Alloying; moderate grain size; large ingots; low energy consumption Refining limited; expensive electrode; melting dangerous Same as above; most frequently used Alloying limited Hafnium, vanadium Same as above Alloying during remelting Almost no refining; costly electrode preparation; melting dangerous Good refining; economical feedstock preparation and ingot production; most often used High melting costs Zirconium, titanium Not used Very low contamination; wide range of alloying possible; large ingots low energy consumption; economical melting Limited refining; expensive feedstock preparation; only round ingots Economical feedstock preparation; refining of highdensity inclusions; melting of slabs, ingots, and rods; high production rate; low energy consumption Alloying limited; material losses from splatter; high furnace investment Electron Beam Melting and Casting Characteristics The characteristics of electron beam melting and casting technology are: • • • • The flexibility and controllability of the process temperature, speed, and reaction The use of a wide variety of feedstock materials in terms of material quality, size, and shape The different methods of material processing available Product quality, size, and quantity Contamination of the product is avoided by melting in a controlled vacuum and in water-cooled copper crucibles (Fig 33) Fig 33 Schematic of the electron beam melting process The energy efficiency of electron beam processing exceeds that of competing processes because of the control of the beam spot dwell time and distribution at the areas to be melted or maintained as liquid In addition, unnecessary heating of the ingot pool, as occurs in vacuum arc remelting, for example, is avoided Power losses of the electron beam inside the gun and between the gun nozzle and the target are very small, but approximately 20% of the beam power is lost because of beam reflection, radiation of the liquid metal, and heat conductivity of the water-cooled trough and crucible walls Electron Beam Melting and Casting Processes From the large variety of electron beam melting and casting processes shown in Fig 34 only the processes illustrated in (a), (c), (d), and (f) are related to processes used in foundry technology: • • • • Button melting processes for the quality control of steel and superalloy cast parts to control the content of lowdensity inclusions Drip melting process for the preparation of refractory and reactive metal feedstock material for electron beam and VAR skull melting and casting Continuous flow melting process for the feedstock refining of superalloys for VIM and electron beam casting Electron beam investment casting process Fig 34 Examples of electron beam melting and casting processes (a) Button melting with controlled solidification for quantitative determination of low-density inclusions (b) Consolidation of raw material, chips, and solid scrap to consumable electrodes for vacuum arc or electron beam remelting (c) Drip melting of horizontally or vertically fed feedstocks (d) Continuous flow refining/melting (e) Floating zone melting (f) Investment casting (g) Pelletizing (manufacture of pellets from scrap and other materials for scrap recycling) (h) Atomization and granulation of refractory and reactive metals Electron Beam Heat Source Specifications For all electron beam melting and casting processes, except for the crucible-free floating zone melting process, Piercetype electron beam guns with separately evacuated beam generating and prefocusing rooms are the key components of the furnaces used The essential features of these guns are: • • • • • Large power range of to 1200 kW Long free beam path of 250 to 1500 mm (10 to 60 in.) and the adjustable beam power distribution Beam deflection angle of ±45° and spot frequency up to 500 Hz Usable vacuum pressure range between and 0.0001 Pa (10-2 and 10-6 mbar, or 7.5 × 10-3 and 7.5 × 10-7 torr) Reliability of the gun and cathode system The power control and distribution system allows a very accurate distribution of beam power and energy for achieving the required heating for material melting, superheating, refining, and electrothermal effects Button Melting for Quality Control The button melting process (Fig 34a) serves to control the quality of feedstock materials for investment casting and to produce casting samples In contrast to the conventional electron beam melting process, this process is not used for refining, but only for flotation and concentration of low-density inclusions During the eight-step process, the sample is heated and drip melted Low-density inclusions are floated to the surface and concentrated in the center of the pool of molten metal during controlled solidification by computer-controlled reduction of beam power and circular electrothermal stirring The concentrated impurities can be identified and evaluated by conventional metallographic methods, but the size of the raft gives the first indication of the quantity of impurities in the metal Most button melting furnaces are completely automated and microprocessor controlled to guarantee process reproducibility Melting is usually carried out in the pressure range of to 0.001 Pa (10-2 to 10-5 mbar, or 7.6 × 10-3 to 7.6 × 10-6 torr) Drip Melting The drip melting processes (Fig 35c and d) are primarily used for the production of clean, mostly ductile ingots of refractory and reactive metals or of specialty steels The feedstock for the first melt (Fig 35c) can be compacted sponge, granular, powder, or scrap, which might be presintered in a vacuum heating furnace In some cases, loose raw materials can be consolidated in a water-cooled copper trough (Fig 35a) The consolidated ingot can then be fed horizontally for drip melting Raw material that is continuously consolidated in a water-cooled copper crucible with a retractable bottom plate (Fig 35b) can be fed horizontally or vertically for drip melting In both consolidation processes, only 20 to 80% of the material is melted Refining and losses of material by splattering are negligible Fig 35 Schematics of electron beam consolidation and drip melting processes (a) Consolidation of coarse and solid scrap (b) Continuous consolidation of raw material, chips, and solid scrap by direct feeding into a continuous casting crucible (c) Drip melting of horizontally fed compacts, sintered bars, or consolidates for initial melting of reactive and refractory metals (d) Drip melting of vertically fed vacuum induction melted or conventionally melted electrodes (e) Drip melting of horizontally and vertically fed materials for the production of alloys from feedstocks with very different melting points Drip melting of horizontally fed compacts is the most frequently used process for the production of ingots from refractory or reactive metals The resulting ingot is of sufficient purity, but has an area of inhomogeneity caused by the shadow of the horizontally fed bar Two or more electron guns are used in drip melting to make use of reflected electron beams and to reduce evaporation and splattering The end of a compact is welded to the front of the following one to avoid dropping semisolid material into the pool Table lists processing parameters that have been successfully used to electron beam melt various reactive and refractory metals and 4340 alloy steel Table Melting and refining data of refractory and reactive metals and alloy steels gained in laboratory and pilot production furnaces Metal Feedstock size, mm (in.) Ingot diameter, mm (in.) Ingot weight, kg (lb) Integral melting rate, kg/h (lb/h) Electron beam power of second melt, kW Operating vacuum pressure of last melt, Pa (torr) Total specific melting energy, kW · h/kg Material yield, % Hardness, HB Interstitial elements in feedstock and final ingot, ppm C Tungsten 40 (1.6) diam O N H 70 4100 30 10 60 (2.4) 37 (80) 10.5 (23) 119 × 10-3 (6 × 10-5) 10.3 93.1 10 200 200 5000 50 20 115 (4.5) 55 (2.2) diam, 100 (4) long 200 (440) 20 (44) 300 × 10-2 (1.5 × 10) 9.9 90 10 210 Tantalum 60 (2.4) diam 100 1200 140 10 80 (3.2) 65 (145) 16.7 (37) 130 × 10-3 (6 × 10-5) 6.0 92 10 75 30 65 650 13 10 160 (6.3) 523 (1150) 38.4 (85) 371 × 10-3 (2 × 10-5) 8.30 91.8 15 13 69 100 (4) diam 170 810 50 10 100 (4) diam 100 (4) 64 (140) 12.5 (27.5) 130 × 10-4 (6 × 10-6) 10.4 95 12 10 10 140 120 (4.7) square, 180 (7.1) long 200 750 60 10 60 (2.4) square, 160 (6.3) long Molybdenum (7.1) long 180 (7.1) 120 (4.7) square, 180 (7.1) long 290 10-3 (7.5 × 10-6) 5.2 96.8 10 12 11 140 160 5220 554 35 227 (500) 17.6 (39) 240 10-2 (7.5 × 10-5) 12.3 87 12 106 60 77 80 4500 330 40 180 (7.1) 80 (3.2) square, 150 (6) long 50.2 (111) 150 (6) Niobium 408 (900) 326 (720) 13.2 (29) 218 × 10-3 (3.8 × 10) 15.9 96.8 111 52 66 1870 95 × 10-3 (3.8 × 10) 38 93.1 170 25 160 Hafnium 60 (2.4) square 80 (3.2) diam 80 (3.2) 40 (90) 1.7 (3.7) 80 100 (4) diam 500 900 100 100 (4) diam 130 (5.1) 173 (380) 7.5 (16.5) 110 × 10-3 (3 × 10-5) 14.7 93.5 100 200 50 170 60 (2.4) square, 100 (4) long 950 95 30 100 (4) 19.5 (43) 14.3 (31.5) 80 × 10-3 (6 × 10-5) 4.6 88.2 545 30 120 950 95 30 150 (6) Zirconium 179 (395) 73 (161) 250 × 10-2 (1.5 × 10) 2.6 735 48 125 1045(a) 210(a) 16(a) 60 (2.4) square, 150 (6) long Vanadium 60 (2.4) diam, 80 (3.2) long long 80 (3.2) 7.7 (17) 2.8 (6.2) 80 × 10-3 (6 × 10-5) 3.1 91 235(a) 95(a) 13(a) 30-100 Titanium 100 (4) diam 2180 100 30 Ti-6Al-4V 100 (4) 28.2 (62) 45.2 (100) 87 0.8 (6 × 10-3) 1.91 99 1850 80 16 Ti-8Al-1Mo1V 100 (4) diam 890 70 100 (4) 28.2 (62) 22.5 (50) 60 0.4 (3 × 10-3) 2.66 98 730 50 3900 63 100 0.3 150 (6) 31 (70) 10 (22) 52 × 10-3 (3 × 10-5) 1.67 93 4300 2.4 26 0.08 150 (6) 31 (70) 80 (176) 80 2.0 (0.015) 1.0 99 3622 10 78 0.10 4340 steel 80 (3.2) diam (a) The reproducibility of the refining data could not be confirmed ... 4 1-4 7 47 60 % 3 1-3 7 57 70% 2 0-2 6 67 80% 11 -1 5 77 90% 8 9-9 1 Mullite 1 8-3 4 6 0-7 8 0.53.1 1. 0-3 .0 1850 3 360 Corundum 0. 2-1 .0 9 8-9 9.5 Trace 0. 3-1 .0 2000 363 0 Silica super duty 9 5-9 7 0 .1 5- 0.35... Super duty 4 9-5 6 4 0-4 4 1.52.5 2. 5-4 .0 17451 765 31753210 Medium duty 5 7-7 0 2 5-3 8 1.32.1 4. 0-7 .0 166 0 168 5 30203 065 Semi silica 7 2-8 0 1 8-2 6 1.01.5 1. 0-3 .0 164 0 168 5 29853 065 -5 2 2.02.8... 0.020.10 168 01700 3 060 3090 Conventional 9 4-9 7 0.4 5-1 .20 1.83.5 0.30.9 0.100.30 163 5 166 5 29753025 Chrome 3. 0 -6 .0 1 5- 34 2 8-3 3 1 4-1 9 1 1-1 7 1. 0-2 .0 12901425 2350 260 0 Magnesite 0. 7-1 .0 0. 3-1 .5