US2734242A - Ingot molds - Google Patents

Description

E. MARBURG 2,734,242

INGOT MOLDS Feb. 14, 1956 Filed June 30, 1954 v4 SheetS-Sheet l Feb. 14, 1956 E. MARBURG 2,734,242

INGOT Moms Filed June 30, 1954 4 Sheets-Sheet 2 77M HFT/5e Pou/e M//Varf 0075/05 60e/9064 Wan/5 /A/5/5 @Bevagna/V5 /N/r/.qL I

WEG

l I l I I I I I I I I I l I I I I l Feb. 14, 1956 Filed June 30, 1954 EL MARBURG INGOT MOLDS F155. H5111@ 'TL/f H., el L [mem:

wm@ i Feb. 14, 1956 E. MARBURG 2,734,242

INGOT MOLDS Filed June 30, 1954 4 Sheets-Shaml 4 United States Patent O INGoT MoLDs dgar Marburg, Pittsburgh, Pa., assignor to United States Steel Corporation, a corporation of New .iersey Applicaties .time 30, 1954, serial N0. 440,353

4 Claims. (Cl. 22-139) This invention relates to improvements in ingot molds.

The present application is a continuation-impart of my earlier depending, application Serial No. 299,635, led July 18,. 1952,v now abandoned.

Both the uniformity of composition and the internal soundness of a steel ingot. are dependent upon its mode of solidilication. In changing from the liquid to the solid state, steel contracts about 4 percent; in cooling from the solid'us temperature (at which solidication is just cornplete) to atmospheric temperature, steel contracts an additionalfl' percent. As4 a result oi?l contraction during solidicati'on and cooling, a central cavity or pipe, which may extend below midhegh`t, develops in ingots cast without top protection. The continuitj of the pipe may be broken by bridging; that is, complete trans-'solidication, at one or several places below the top of the ingot. Pipe cavity above the uppermost bridge is called open or primary pipe; that below this bridge is referred to as' secondary pipe. To prevent pipe from developing in ingots, collars of refractory insulating material (generally tire' clay) are set on` mold tops; they extend a'few inches down inside the mold. The hot top is poured after the mold is filled. Since ,the metal inthe hot top, or sinkhead, remainsliquid longer than that in the ingot body, it may drain in'to the latter and compensate for ingot shrinkage. For the' purpose of describing and claiming the present invention, I define the term ingot body as refer-ring to the portion of an ingot below the hot-top junction (plane of lower edge. of hot top) in a hottopped ingot, and the entire ingot below the height of pour in a non-hot-topped ingot.

Initial' soliditi'cation of ingots occurs approximately according to Feilds Equation d=k\/ (d=depth in inches; t=tirne in minutes). If a' be plotted against t in this equation, a parabolic curve results. According to this curve, rates of solidiication are extremely fast. near the surface of an ingot body; they decrease to slow rates in the interior.

Steel solidies as dendrites, or tree-like crystals. Since purer metal solidifies at a higher temperature than that at which less pure metal solidies, the solidifying crystals are always purer than'. the adjacent liquid metal. In particular the layer of' liquid metalA immediately adjacent tol the face of solidieation has a higher content of segregating' efe'ments, principally carbon, phosphorus and sulphur, than does either the solid metal or the remaining liquid metal. Thel degree of segregation. that occurs is a function of the rate of heat extraction. Nearthe surface of an, ingot body, where heat extraction and solidification rates are extremely fast, little or no segregation occurs; as solidiication rates decrease inwardly, the intensity of segregation in the segregated layer increases.

The segregated liquid layer may be entrapped by the following mechanism: elongated or columnar crystals `grow' inwardly approximately perpendicular to the surface of an ingot body. A columnar dendrite may grow slightly in advance of its neighbors,` and penetrate the segregated liquid layer. Since the segregated layer is slightly in- 'ice sulating, penetrating columnar crystals provide a preferred course for heat flow. Thus solidiiication may proceed inwardly and upwardly beyond the segregated layer, and entrap the latter. Because transverse solidication proceeds inwardly at the same time, the above mechanism results in layers of segregation that slope inwardly toward the top of the ingot body, known as inverted-V segregation. ln large ingots the above mechanism is repeated, and a series of approximately parallel lines of inverted-V segregation isv developed.

Another type of segregation in ingots is V segregation, which occurs along the vertical axis. V segregation results from the entrapment of segregated liquid layers adjacent to the face of vertical solidication by the soliditication at higher temperature of purer metal above the segregated layers. V segregation is most concentrated in the upper portions of ingot bodies and decreases in intensity downwardly. In big-end-up ingots V segregation is generally confined to the upper 15 percent of the ingot body; it may extend below mid-height in big-enddown ingots. v

if liquid metal be completely entrapped by solid metal, pores or voids result from contraction of theA entrapped metal in solidifying. Thus porosity of aY minor degree is associated with both and. inverted-V segregation. Major porosity develops along the vertical axes of ingots of certain: steels asV a result of the entrapment of a vertical column ot' incompletely solidified metal by transso]iditicationy justy below. the hot-top junction. This defeet, referred to as axial porosity, is somewhat-,similar ink appearance to V segregation. But, whereas the latter is primarily segregationwith associated porosity, the former is V porosity with associated segregation. n Y

The axial defects described above, V segregation and V porosity, are the most serious internal defects in ingot bodies, and are objectionable for many product applications. Because ot high breakage in forming sheet and strip products from the upper portions of ingot bodies, the top cuts of ingot bodies of deep-drawing steels are often diverted to other use. Forging may cause internal cracks to develop in ingot bodies with unsound centers; axial defects may also cause weakness in forged products. or adversely atect their surface appearance. In prodnets that areV center-drilled or bored, center segregation may adversely affect tool life, and/or maycause defects in surface of bore. Laminations in sheet or strip .product, blisters in galvanized sheet or strip product, and pinholes in tin plate are other product defects caused by unsound ingot centers.

An object of the present invention is to provide irnprov'ed ingot molds which minimize axial segregation and porosity in ingot bodies cast therein.

As hereinafter explained', it is desirable that ingot bodies solidify' as rapidly as possible in a vertical direction upwardly from the bottom, and as slowly as possible transversely across the top; a further object is to provide ingot molds which promote vertical solidiiication, and retard transverse solidiiication across the top.

A further object is to provide ingot molds in which the foregoing. objects are attained by simple proportioning of the thickness of various parts of the mold walls.

In accomplishing. these and. other objectsy of the invention, i have providedv improved details of structure, preferred forms of which are shown in the accompanying drawings, inwhich: v

Figure l is a schematic view which shows a typical soliditication pattern of a big-end-up ingot;

Figure 2 is a schematic View which shows a typical soliditication pattern of a big-end-down ingot;

Figure 3 is a top plan view of a big-end-up ingot mold which embodies featuresv ofthe present invention;

Figure 4 is a side elevational view of the mold shown in Figure 3, the mold being equipped with a hot top;

Figure 5 is a horizontal sectional view taken on line V-V of Figure 4;

Figure 6 is a horizontal sectional view taken on line VI-VI of Figure 4;

Figure 7 is a horizontal sectional view taken on line VII-VII of Figure 4;

Figure 8 is a vertical sectional View of a big-end-down mold constructed in accordance with the invention, the mold being equipped with a hot top;

Figure 9 is a view similar to Figure 8, except that the mold lacks a hot top;

Figure l0 is a vertical sectional view of a modified form of a big-end-down mold in accordance with my invention, the mold being equipped with a hot top; and

Figure ll is a graph which compares mold wall temperatures of a mold constructed as shown in Figure 8 with those of a conventional mold.

The progress of soliditication of an ingot body may be depicted by isochrones representing the face of solidilication at various intervals after pour. Figure l shows the solidiication pattern of a 32 by 32 inch big-end-up ingot reconstructed from measurements of shell thicknesses of six ingots dumped at various intervals after pour. Figure 2 shows a solidication pattern for the 29-inch-wide section of a-29 by 66 inch big-end-down ingot. Although the patterns for the big-end-down and the big-end-up ingot are generally similar, they differ in certain important respects, to be discussed below. The patterns may be considered from three viewpoints: (a) solidiiication transversely from sides to middle; (b) solidication vertically from base to top; and (c) accelerated solidiication resulting from overlapping of transverse and vertical components of heat extraction, originating at points of tangency Feilds Equation d=k\/t. I have found that inner lines of inverted-V segregation develop as a result of acceleration of transverse solidification; hence they coincide with loci af of this acceleration. Accelerated `transverse solidiication is first completed to the middle of the ingot at point c, which represents the tip of the base cone abaca of vertical soliditication. When trans-soliditication is completed at c, there remains a relatively large column of liquid metal above point c, as bounded in Figure l approximately by isochrone 100. At point c transverse components of heat extraction from all four sides of the ingot are added to vertical components effecting vertical soliditication. As a result, vertical solidification rates are greaty ly accelerated at point c, as indicated by the wide spacing of isochrones along the vertical axis above this point. Because of the extremely fast rates of final accelerated vertical solidification above point c, vertical solidication reaches the top of the ingot body before transverse solidification is completed across the top. Dashed lines ece trace the cores of vertical soliditication thus developed in ingot bodies of both types.

I have found that V segregation is developed within and extends across the vertical cores of ingot bodies. Since segregates are lighter than liquid steel, they have a tendency to float upwardly and escape into the sinkhead. Their ability to do so, however, depends upon the width of the vertical core and that of the adjacent zones of accelerated transverse'soliditication at the hot-top junction, that is, upon widths Vt and Vu. As indicated by the s0- lidication patterns, Vt and Va are wider in big-end-up ingots than in big-end-down ingots of comparable sizes. For this reason, the former ingot bodies develop less axial segregation and porosity than do the latter.

Widths Vt and Va are principally dependent upon two factors;` (l) the width to height (w/h) ratio of an ingot body, and (2) the ratio of the rates of transverse solidification across the top of the ingot body to those of vertical soliditication. Wide vertical cores and sound ingot bodies are favored by high w/h ratios. Practical limitations of rolling mill equipment and economic considerations, however, dictate the upper limit of w/h ratios of ingot bodies. It would almost certainly prove uneconomical to increase w/h ratios of ingot bodies to the extent necessary to eliminate all axial defects.

Wide vertical cores (and sound ingot bodies) are also favored by slow rates of transverse soliditication across the top of an ingot relative to vertical solidication. Since experiments have demonstrated that transverse solidification rates vary with mold-wall thickness (at least in mold walls over 2 inches thick)it would naturally seem desirable to design mold walls thin at the top, and thick at the base (since rates of accelerated vertical solidilication in the vertical core are caused principally by transverse components of heat extraction to the lower portion of the mold wall). Indeed, mold walls are commonly so designed, with the base 1 to 4 inches thicker than the top, and with a uniform taper between the two ends. Unfortunately, however, tapered mold walls are ineffective in producing differential solidication in ingots. Numerous dumped-ingot experiments have revealed that, except for accelerated soliditication from the base, solidilication proceeds at uniform rates throughout the height of ingots cast in molds with tapered walls. It is extremely diicult to analyze exactly heat losses from mold walls. Heat will ow through the walls, vertically as well as transversely, from hotter to colder portions. In addition to conducted heat, heat will be radiated from the outer surface in amounts varying with the surface temperature. Regardless of the exact analysis of heat losses from mold walls, however, it is well-known that molds become hotter in their middle portions than at their end portions (this can be verified by observation; an oval-shaped portion in the middle of each wall becomes heated to a bright red to orange color, while the end portions remain dull red to black). Since heat-extraction rates will vary inversely with mold temperatures, however, it is undesirable t0 have a higher temperature at the middle than at the top of the mold wall.

The purpose of the present mold design, as will be clear from the following description, is to control temperatures in mold walls so as to cause greater rates of transverse heat-extraction from the base and middle than from the top portion of a mold, hence to cause solidification to proceed in a favorable manner.

Figures 3 to 7 show a big-end-upV mold constructed in accordance with my invention. This mold is equipped with a hot top I-I which is of any standard or desired construction and extends a short distance into the top of the mold cavity. The upper portions 10 of the mold walls are thin and the lower portions 12 thick. Immediately below the bottom of the thin upper portion 10 is a relatively short, thickened section 13. Between this thickened section and the top of the thick bottom portion 12 is a relatively short thin section 14. The trace of the inside surface of the Walls in any vertical plane above the base is a straight line.

When molten steel is introduced to the mold just dcscribed, the thin portions 10 and 14 become hotter than the thick portions 12 and 13. Consequently heat ows from the top portion 10 and from the thin section 14 into the thick section 13, as indicated by arrows. Hence a horizontal plane may be passed through section 13 across which no vertical heat tiow occurs. Similarly heat flows upwardly from thin section 14 into thick section 13 and downwardly from thin section 14ito thick portion 12. Hence a horizontal plane across which no vertical heat ow occurs may also be passed through thin section 14.

Because of the middle thick-and-thin arrangement, heat wenn# from the upper thinportion@ 111A tlows' downwardly into the comparatively small volume represented by the up* per half of the thick section: 13. Similarly the lower thick. portion: 12.re'ceivesheat only from thezlowcrhalf of -thin section 14. Thus the. normally faster rates of solidiiication of thick base portion 12 and normally. slower rates: of thin upper portion are dist-urbedtof at minimum) extent. Asa result, the thin` top and thick basepon tions ofzrnolds` are enabled toeffect differential soliditication yin ingotsin the manner desired.,

Figure 8 shows a big-endfdowrfr.I mold constructed in accordance with my invention. This moldisl eq'tripped;A with a hot top H similar to that shown in Figure' 4. The mold walls havea relatively thick lower portion 17 and a relatively thin' upper portion 181 At the' bottom of the latter is a relatively narrow thick section I9. Between theA thick section 19 and the top` of the thick lower portion 1"]"is` a relatively narrow thin section` 20. The trac of' the inside surface in' any vertical plane again is a straight line. Thick,section' 19 andj thin section 20 act tov disrupt vertical ow of heat in the same manner as the corresponding thick section 13` andi thin section 14 the big-ed'up mold' shown in' Figures 3 to 7:.

Figure 9 shows a big-enddown mold which is similar to Figure 8, except that it lacks a hot top. Semi-killed, or gas-evolving steels, such as represented in Figure 2, are commonly cast without hot tops. Molten metal is poured into the mold to a height a few inches short of the mold top. The height of ingot body cast therein is comparable to the portion of a hot-topped ingot below the hot-top junction. The solidication pattern of this ingot, Figure 2, has already been discussed.

Figure 10 shows an alternative mold construction which can be used to facilitate manufacture. Although Figure 10 shows a big-end-down mold equipped with a hot top, it is apparent that the same mold can be used without a hot top and that the same principles can be applied to a big-end-up mold. In large foundries, a sand slinger frequently is used to make the outer sand mold for casting the ingot mold. A sand slinger has revolving blades which throw sand with tremendous force into the peripheral space between an outer housing and the pattern (similar to the ingot mold exterior). It is apparent that sand so placed would not pack tightly in the dead space below the thick ring 13 of Figure 4 or 19 of Figures 8 and 9. Consequently the mold shown in Figure 10 has a thickened portion 22 and a thinner portion 23 separated by a re-entrant angle modified to enable sand to be packed tightly with a slinger. The action of the thick and thin portions in the mold walls is the same as already described for the other embodiments.

Height d in Figures 4, 8, 9 and l0 represents the vertical distance between the top of the ingot body and the beginning of the uppermost thickened section and may be evaluated as about 10 to 70 percent of the height of the ingot body. While these are extreme limits of effectiveness, the preferred range is about 35 to 50 percent of the height of the body. The actual mold extends a few inches above this height to allow for insertion of a hot top or to leave some clearance above the height of pour in a uonhottopped ingot. The minimum spacing c can be as low as it is practicable to cast corrugations. The upper limit of spacing may be specified as about 24 inches. While these are extreme limitations, the preferred spacing of corrugations is about l2 inches.

Figure l1 shows a graph which demonstrates the electiveness of my invention. In this graph the solid line curves represent mold wall temperatures at various vertical positions and at various intervals after pour with a mold constructed as shown in Figure 8, of inside dimensions 71/2 by 71/2 inches at the top, 81/2 by 81/2 inches at the base. The broken line curves represent corresponding temperatures in a conventional mold (i. e. a mold whose walls taper uniformly), of the same internal dimen sions, and of the same wall thickness at the top (l1/zvao increQLand at the base E .i1/zI inches) as the mold shown' in lrgnre it; In aan instance the. temperatures were' determined by inserting al thermocoupl in oneot'v the m'oldwallsl at thelreig'ht indicatedand tol a depth 1,411 inch fromy the insidefface, a second thern'iocouple in another wall at" thel same height' to the midfthickness depth,- and a third thermocouple in a different wall at this height to a depth" lidi inch' from the outside face, and` awe'ragingy the three readings. 'rneb'rok-en-lne curves show that the conventional mold'v is hot-test a-tj the middle portion; also it is hotter at the baserA portion than vat theA top1 (at least for the aboutA 7 minutes, required for most of this' ingot to'sol'idify). In arr-exact opposite manner, the. experiment-ali molo is least not at the middle thick rirrgfr the duration of solidiiie'ation-v it isr alsol less hot at the base than atfthe top ponan. Thus the'experimentat mold. extracts heat faster front the'- base and middlepornon, than. it does from the top of' an ingot (in the manner desired',v for favorable solidiiication).

Another; test was conclut-:ted in which i'ng'ots of 0.25'- perc'enti carbonhseei cast in tl'lef above' two inol'ds 'were back=poured with. steel. containing 2 percent copper about 61 minnresan 30: secendsafter pour. Snce steel containing copper etches darker than steel without copper, the metal,v that had been liquid at back-pour was revealed as a dark column in the macroetch of a longitudinal section of each ingot. Measurements of these macroetches reveal that, at the time of back-pour, transverse soliditcation had proceeded 2%6 inches across the top of the ingot cast in the regular mold, but only 2 inches across the top of the ingot cast in the mold of special design, Figure 8, Vertical solidilication had progressed to a height only about 51/2 inches above the base of the former mold; it had attained a height of 13 inches in the latter mold. These macroetches provide evidence that vertical solidification had progressed almost 21/2 times as far in the mold of special design as in the regular mold; but transverse solidi` ication had only proceeded 86 percent as far across the top of the former ingot.

Ingots of various grades,.plaincarbon, low-alloy, and

f sulfur steels, have been cast in big-end-up and big-enddown molds of the design shown in Figures 4 and 8. Comparison ingots of the same steels have been cast in conventional molds of the same internal dimensions, but of uniform wall taper. The ingots have been split longitudinally along the center plane, machined and ground to a smooth surface, and macroetched. All of the ingots cast in molds o-r conventional design contained secondary pipe, V segregation and axial porosity to a greater or less extent. In contrast, all the ingots cast in the special molds with one exception (the big-end-down low-alloy steel ingot had some axial porosity) were completely solid and sound.

From the foregoing description it is seen that the present invention affords an effective means for disrupting the vertical llow of heat in ingot mold walls by simple proportioning of the wall thickness. The invention effectively eliminates the most serious axial defects in ingots, namely secondary pipe, V segregation and axial porosity.

While several embodiments of my invention have been shown and described, it will be apparent that other adaptations and modifications may be made without departing from the scope of the following claims.

I claim:

l. An ingot mold comprising solid side walls which define a cavity adapted to receive molten metal for casting an ingot body and which extend a relatively short distance above the normal height of the ingot body, said walls below this height being of greater thickness in their lower portion than in their upper portion to accelerate solidilication of the lower portion of an ingot body both transversely andvertically and to retard transverse solidtication of theupper portion, the trace of the inside surfaces of said walls in any vertical plane away from the lower end being a straight line, and means on the exterior of said walls between their thick and thin portions spaced below the normal height of an ingot body by a distance which is 10 to 70 percent of the height of the ingot body for disrupting vertical heat ow in these walls, said means including a thickened section and a thinner section im mediately therebelow, both of said sections` being of relatively short vertical dimensions. i

2. An ingot mold comprising solid side walls which define a cavity adapted to receive molten metal for casting an ingot body and which extend a relatively short distance above the normal height of the ingot body, said walls below this height being of greater thickness in their lower portion than in their upper portion to accelerate solidiiication of the lower portion of an ingot body both transversely and vertically and to retard transverse solidification of the upper portion, the trace of the inside surfaces of said walls in any vertical plane away from the lower end being a straight line, a relatively short thick section in said walls at the bottom of their upper thin portion spaced below the normal height of an ingot body ,by a disseltlions cooperating to disrupt vertical heat flow in said w s.

Y 3. A mold as defined in claim 2 which is ofthe big-endup type and in which the portion of the walls above the normal height of the ingot body is adapted to receive a hot top.

4. A mold as defined in claim 2 which is of the bigend down type and in which the portion of the walls above the normal height of the ingot body is adapted to receive a hot top when a hot-topped ingot is cast therein and to provide clearance above the height of pour when a nonhot-topped ingot is cast therein.

References Cited in the le of this patent UNITED STATES PATENTS

Claims (1)

1. AN INGOT MOLD COMPRISING SOLID SIDE WALLS WHICH DEFINE A CAVITY ADAPTED TO RECEIVE MOLTEN METAL FOR CASING

Images