The reduction and melting of iron and steel in blast furnaces, basic oxygen converters, and
electric furnaces were discussed in Chapter 4.
Most steel is cast from furnaces of the types already described, but over half of gray iron and malleable iron base for castings is melted in the cupola. Electric furnaces are being used more for melting iron. They cost more to oper- ate than cupolas, but do not have the pollution control problems and expenses.
The Cupola
The cupola is simple and economical for melt- ing pig and scrap iron. It is essentially a vertical steel-shrouded and refractory-lined furnace.
The construction of a typical cupola is shown in Figure 9-27. Cupolas are made in many sizes, commonly about 1–2 m (≈4–7 ft) in outside diameter and 9 to over 12 m (≈30 to over 40 ft) high. These sizes may turn out 5–10 metric tons (≈5–10 tons) of melted metal per hour.
A cupola must be prepared and heated care- fully to avoid damage. The refractory lining is repaired or replaced as needed, and the bottom doors are propped shut. A layer of sand sloping toward the tap hole is rammed over the bottom.
Excelsior, rags, and wood are placed on the sand to protect it from the initial charge of fuel, which may be a meter or more (several feet) thick, and to ignite the fuel. Materials are charged through a door 4–8 m (≈15–25 ft) above the bottom.
After the initial charge has become hot, alter- nate layers of fuel and metal with flux are added.
The fuel may be a good grade of low sulfur coke, anthracite coal, or carbon briquettes. The flux to help form slag to remove impurities and retard oxidation of metal is usually limestone, but sometimes soda ash, fluorspar, and proprietary substances are added. The proportion of metal to fuel by weight ranges in practice from about 6:1 to 12:1. A ratio of 10:1, which is common, means 100 kg of coke are needed to melt 1 metric ton (≈200 lb of coke for a ton) of iron. Fuel cost is lower with a lower proportion of coke; the melt- ing and output rates are higher with more coke.
Once the charge and cupola have had a chance to become heated uniformly in an hour or two, the forced draft is started. Air is blown through the wind box and tuyeres around the hearth into the furnace. About 850 m3 of air are required to melt 1 metric ton of iron (about 30,000 ft3/ton). Then, the metal begins to melt.
The tap hole is plugged with a bott of clay. The molten metal seeps down through the coke and collects at the bottom. Tapping is done by break- ing the clay plug in the tap hole. Flow may be permitted at intervals or may be continuous during the melt. For the latter case, the melting rate must be in balance with the capacity of the tap hole. The slag floats on top of the metal. If the molten metal is allowed to accumulate in the hearth, the slag flows off through an opening higher than the tap hole, called the cinder notch, in the back of the cupola. Otherwise, the slag may be skimmed off the metal as it flows out of the cupola. When sufficient metal has been melted, the bottom of the cupola is dropped to spill the remaining contents onto the ground to cool.
Calculations
The cupola does little to refine the metal, and the composition of its product depends largely upon what is put into it. The proportions of the metals charged into the cupola must be calcu- lated carefully to assure a uniform and pre- dictable product. These calculations are based upon knowing the amounts of carbon, silicon, manganese, phosphorus, and sulfur in the pig and scrap iron, and the nature of the reactions that take place in the cupola.
The following example shows how the calcu- lations are made for 1,500 kg and alternately for 3,000 lb of iron fed into the cupola. The typical raw materials available in the storage yard of the foundry are listed in Table 9-4 with their analyses. In other foundries, still other materials, such as machinery steel scrap and iron of various analyses, also may be stocked depending upon the sources of supply and the desired product. In this example, on the basis of current practices and results in the foundry, it is decided to compose the charge of 10% No.
1 pig iron, 20% No. 2 pig iron, 30% new scrap iron, and 40% returns from previous melts. The amount of each element that may be expected in the product can now be ascertained on the basis of the reactions in the cupola.
1. The amount of carbon in the iron remains substantially unchanged during the process.
Some carbon is oxidized, but about the same amount is picked up from the fuel. The
Figure 9-27. External and internal views of a classic cupola.
amount contributed by each ingredient can be found by the following calculations.
For 1,500 kg of iron:
No. 1 pig iron
1,500 × 0.10 × 0.035 = 5.25 kg No. 2 pig iron
1,500 × 0.20 × 0.035 = 10.5 kg New scrap iron
1,500 × 0.30 × 0.034 = 15.3 kg Returns
1,500 × 0.40 × 0.033 = 19.8 kg
Total 50.85 kg
For 3,000 lb of iron:
No. 1 pig iron
3,000 × .10 × .035 = 10.5 lb No. 2 pig iron
3,000 × .20 × .035 = 21.0 lb New scrap iron
3,000 × .30 × .034 = 30.6 lb Returns
3,000 × .40 × .033 = 39.6 lb
Total 101.7 lb
So, the final percentage of carbon equals:
(Eq. 9-2) 2. The silicon content can be expected to be re- duced by 10% from oxidation. The amount of silicon in the charge can be found by making the following calculations.
For 1,500 kg of iron:
No. 1 pig iron
1,500 × 0.10 × 0.025 = 3.75 kg No. 2 pig iron
1,500 × 0.20 × 0.030 = 9.0 kg New scrap iron
1,500 × 0.30 × 0.023 = 10.35 kg Returns
1,500 × 0.40 × 0.025 = 15.0 kg
Total 38.1 kg
Table 9-4. Compositions of some typical metals for cupola melting (%)
Carbon Silicon Manganese Phosphorus Sulfur
No. 1 pig iron 3.5 2.50 0.72 0.18 0.016
No. 2 pig iron 3.5 3.00 0.63 0.12 0.018
Cast-iron scrap 3.4 2.30 0.50 0.20 0.030
Returns (risers, defective
castings, etc.) 3.3 2.50 0.65 0.17 0.035
50 85
1 500 100 3 39 101 7 3 000 100 .
, . % .
× = = , ×
For 3,000 lb of iron:
No. 1 pig iron
3,000 × .10 × .025 = 7.5 lb No. 2 pig iron
3,000 × .20 × .030 = 18.0 lb New scrap iron
3,000 × .30 × .023 = 20.7 lb Returns
3,000 × .40 × .025 = 30.0 lb
Total 76.2 lb
The final percentage of silicon equals:
(Eq. 9-3) 3. The manganese content is expected to be
reduced by 20% from oxidation. In the same way as for the other elements, the manga- nese content is estimated to be 9.12 kg for a 1,500-kg charge and 18.24 lb for a 3,000-lb charge, and the final percentage of manga- nese equals:
(Eq. 9-4) 4. Phosphorus losses in the cupola are neg- ligible. In the same way as for the other elements, the amount of phosphorus in the charge is calculated to be 2.55 kg for 1,500 kg of iron and 5.10 lb for 3,000 lb of iron. From this, the final percentage of phosphorus is equal to:
(Eq. 9-5) 5. The iron loses almost no sulfur in melting,
but picks up about 4% of the sulfur in the coke. The quantity of sulfur in the metal charge is calculated to be 0.423 kg for 1,500
38 1 0 9
1 500 100 2 29 76 2 0 9 3 000 100 . .
, . % . .
,
× × = = × ×
9 12 0 8
1 500 100 0 49 18 24 0 8 3 000 100
. .
, . % . .
,
× × = = × ×
2 55
1 500 100 0 17 5 10 3 000 100 .
, . % .
× = = , ×
kg of iron and 0.846 lb for 3,000 lb of iron.
The iron to coke ratio is to be 8:1, and the sulfur content of the coke is known to be 0.50%. Thus, the quantity of sulfur in the coke is (1,500/8) 0.005 = 0.9375 kg and, of that, 4% or 0.0375 kg is added to the iron for a 1,500-kg charge. Alternatively, the sulfur in the coke is (3,000/8) 0.005 = 1.875 lb, and .075 lb is added to a 3,000-lb charge. The final sulfur content is estimated to be:
or (Eq. 9-6)
In summary, the composition of the iron from the cupola is estimated to be 3.39% carbon, 2.29% silicon, 0.49% manganese, 0.17% phos- phorus, and 0.03% sulfur.
If more or less of any of the elements is want- ed, the results can be changed by specifying raw materials in different amounts and of different kinds and by adding ferroalloys of elements to the metal in the ladle after melting. Ferroalloys are less expensive than the pure elements. For instance, if in the example 2.50% silicon is de- sired, about 2 kg (≈4 lb) of 50% ferrosilicon can be added to each 500 kg (1,102 lb) of molten iron.
Sometimes elements are added by bubbling gas carriers through the metal in the ladle.
Melting Nonferrous Metals
Some nonferrous metal melting is done in al- most all types of furnaces. More is done in induc- tion furnaces for reasons of convenience, ease of operation, and fewer environmental problems, but oil- and gas-fired crucible furnaces have advan- tages. Although the fuel cost is about the same for both types of furnaces, an electric furnace may cost twice as much initially as an oil- or gas-fired one. Modern electric furnaces have complicated controls and are costly to service, whereas all a fossil-fuel furnace needs is relining from time to time. On the other hand, oil- and gas-fired furnaces create heat, fumes, and noise problems. A stack is needed, and a bag-house may be required depending upon local environ- mental regulations. It is also difficult to find
( . . )
, . %
0 423 0 0375 100
1 500 0 03
+ × =
=(. +. )× =
, . %
846 075 100 3 000 0 03
people to work around hot fumes, particularly in hot weather.
Two types of crucible furnaces are stationary and tilting furnaces. The stationary type of crucible furnace requires that the crucible be lifted in and out for pouring. A typical furnace of this type is shown in Figure 9-28. When a stationary furnace is sunk into the floor or deck of a foundry, it is known as a pit-type furnace.
A tilting type of crucible furnace, shown in Figure 9-29, requires a crucible with a suitable lip for pouring metal when the furnace is tilted.
Most nonferrous metals and alloys oxidize, absorb gases and other substances, and form dross readily when heated. Various practices are followed for each kind of metal to preserve purity and obtain good castings. Space does not permit descriptions of all practices. Full information may be found in treatises on the subject.
Aluminum
Aluminum and its alloys have a marked ten- dency to absorb hydrogen when heated. This gas is released on cooling and causes detrimental pinholes and porosity in castings. Exposure to hydrogen-forming media, such as water vapor, must be avoided. Clean and dry melting stock and crucible are important; a slight excess of air in the furnace atmosphere is desirable.
Figure 9-28. Stationary, high-speed, gas-fired, crucible-type furnace for melting nonferrous metals. (Courtesy McEnglevan Industrial Furnace Company, Inc.)
Figure 9-29. Hydraulic, nose-pour type, crucible furnace for melting nonferrous metals. (Courtesy McEnglevan Industrial Furnace Company, Inc.)
Molten aluminum reacts readily with oxygen to form a film on the surface. Fortunately, if not broken, the dross serves as a good shield against hydrogen and further oxidation. How- ever, excessive dross may become trapped in the metal, particularly if the metal is agitated, and appears as defects in the final casting. Both the amount of oxidation and the tendency to absorb hydrogen increase with temperature and time.
Excessive temperature also causes coarse grains in castings. These conditions dictate the melting procedure for best results. Aluminum should not be heated more than about 56° C (100° F) above the necessary pouring temperature. Tem- perature should be checked with an immersion pyrometer. The melting time should be short with as little agitation as possible.
Aluminum does not ordinarily require as much flux for protection as some other metals
because of its oxide film. However, at times the gas or oxide in the metal must be reduced.
Chlorine or nitrogen may be bubbled through the molten metal. Solid fluxes, containing alu- minum or zinc chloride, may be added. Various proprietary fluxes are commonly used to help dry the surface dross and facilitate skimming it from the metal.
Vacuum Melting
It has become necessary to develop ways of melting and pouring some metals and alloys in the absence of air to make and keep them pure and clean. This is done in a number of ways. A typical installation is depicted in Figure 9-30, which represents true vacuum melting and pouring. Operating temperatures with this equipment can range up to 1,649° C (3,000° F).
Some such furnaces operate at less than one millionth of normal atmospheric pressure, and most under 10 microns (.133 microbars of Hg at 0° C). The metal in the crucible is melted by induction and the induced field stirs the liquid constantly and aids in the release of gases. This is called vacuum induction melting (VIM). En- ergy for melting is supplied by electron beams in some installations. The metal is poured by tilting the crucible; in some cases, the whole furnace tilts.
Arrangements such as those already de- scribed eliminate the two contaminating media, air and slag, from the casting process. The third source of impurities, the crucible, is neutral- ized by the consumable electrode method. This utilizes an arc in a vacuum, between an initial charge and an electrode of the metal melted, inside a water-cooled copper crucible. As metal is melted off the electrode by the arc, it is quickly solidified, especially where in contact with the cooled crucible. Diffusion between crucible and melt is minimized, and the crystalline structure is improved. This process is called vacuum arc remelting (VAR) and is commonly a second step after VIM for high-purity alloys.
Vacuum arc direct electrode remelting (VA- DER) is a process that remelts like VAR but is faster and produces smaller grains in the metal.
Two ingots to be remelted are held horizontally in line in a vacuum over a mold. An arc is struck between the ingots and drops of molten metal
fall into the mold. All the energy is used for melting; none is required to maintain a molten pool. Melting rates at least three times greater than what is required for VAR are reported.
A process called degassing entails pouring only in a vacuum. The metal is melted by con- ventional means in air. The vacuum need not be extremely high to draw off substantial propor- tions of hydrogen and other gases. Four common methods are depicted in Figure 9-31.
Stream degassing is done with a ladle of mol- ten steel positioned over an evacuated chamber containing another ladle or ingot mold. When the nozzle is opened, steel pours into the cham- ber, giving out gas as it falls. Only one ladle is needed for ladle degassing; it is filled with molten steel and placed in the vacuum chamber, which is then evacuated. During the operation, continuous stirring, by induction or helium in- jected at the bottom, constantly brings untreated Figure 9-30. Schematic drawing of a vacuum melting furnace.
steel to the surface for degassing. In the D-H and R-H processes, increments of molten steel are degassed in small chambers. The D-H chamber (or the ladle) surges up and down. On the first stroke, steel enters the chamber and is degassed and, on the reverse, it sinks back into the ladle.
In contrast, the R-H process is continuous. Mol- ten steel, aided by the injection of gas in the up leg, flows up into the chamber, is degassed, and then drops back down the down leg.
Uranium, titanium, and alloys with reactive elements can be effectively cast only in a vacu- um. Vacuum melting improves the properties of metals that can be melted by ordinary means, such as steel and aluminum. Raw materials of good grade are, of course, necessary for a supe- rior product. Metals melted and cast in a vacuum can be kept pure because they are not exposed to new contamination and can be further purified because gases are drawn from them. Oxidizers,
Figure 9-31. Techniques of vacuum degassing. (Courtesy Metal Progress, ASM International)
such as manganese and silicon in steel, are not needed, and no slag is formed. In vacuum melting, precise amounts of pure carbon or hydrogen may be added to reduce oxides. As gases are evolved, they are extracted by the pumps, and the reactions are driven to comple- tion. The addition of highly reactive elements like zirconium, titanium, and aluminum to an alloy may be delayed until after the gas content of the melt has been depleted.
Vacuum-melted steel may contain 3–20 parts of oxygen per million parts of steel. This is about one-twentieth of the proportion in commercial air-melted steel; other impurities are reduced to about the same extent. Pure and clean metals are substantially stronger, more ductile, and more resistant to fatigue and corrosion. For instance, the Charpy impact strength at room temperature has been reported to be 298 J (220 ft-lbf) for vacuum-melted 430 stainless steel
compared to 11 J (8 ft-lbf) for air melted. The benefits obtained have been found particularly desirable for high-temperature alloys, such as nickel and cobalt alloys. Thus vacuum melting is especially in demand for metals to make prod- ucts such as turbine blades, buckets, bearings, castings, after-burner parts, highest-quality tool steels, and all parts that must serve at high temperatures with high strength.
Vacuum melting is inherently more costly than air melting. Although liquid melts up to 14 metric tons (15 tons) are reported for VIM, most installations are of much smaller capacity because of high equipment costs. Melts of tens of metric tons (tons) are produced by VAR, and as much as several hundred metric tons (tons) by degassing. An 1,100-kg (≈2,500-lb) capacity production unit of the type illustrated in Figure 9-30 may cost several million dollars with all equipment and accessories.