US3415307A - Process for casting ductile iron - Google Patents

Description

Dec. 10, 1968 sc uH ETAL PROCESS FOR CASTING DUCTILE IRON 5 Sheets-Sheet 1 Filed March 5, 1966 INVENTORS Arilmr E. Jehuh Martin A. Rico BY Andrlw b. Helix/'0 Attorney Dec. 10, 1968 sc u ET AL PROCESS FOR CASTING DUCTILE IRON 3 Sheets-Sheet 2 Filed March 5, 1966 Fig /0 Dec. 10, 1968 SCHUH ET AL 3,415,307

PROCESS FOR CASTING DUCTILE IRON Filed March 5. 1966 3 Sheets-Sheet 3 INVENTORS ArMur c. Schuln Martin A. Ri

A a/Few 5. Mnlizia Aiiarney United States Patent 3,415,307 PROCESS FOR CASTING DUCTILE IRON Arthur E. Schuh, Riverton, Martin A. Rice, Florence, and

Andrew B. Malizio, Delran, N.J., assignors to United States Pipe and Foundry Company, Birmingham, Ala.,

a corporation of New Jersey Filed Mar. 3, 1966, Ser. No. 533,133 7 Claims. (Cl. 164-414) ABSTRACT OF THE DISCLOSURE A method for casting ductile iron pipe wherein pipe are centrifugally cast in a metal mold provided with a dual coating consisting of a first layer of refractory coating material having a thickness of .01 to .06 inch and a second layer of powdered inoculating material and the metal is treated with inoculating material as it is poured into the mold, the metal in the pipe comprising 3.03.6% C, 2.3-3.75% Si and .02.07% Mg.

This invention relates to the production of cast iron. More particularly the invention relates to the production of substantially ferritic as-cast structures in spheroidal graphite iron. While the invention will be described in connection with the manufacture of ductile iron pipe for which the method is particularly useful, it will be immediately apparent that the method and principles described are equally applicable to other products.

When ductile iron pipe are cast in metal molds the ascast pipe have a carbidic structure. Wet spray refractory mold coatings on the order of .005" thick are used to provide a temporary thermal arrest and to assist in obtaining an even distribution of metal before solidification, but these coatings do not materially change the overall solidification rate of the pipe so that the resultant as-cast pipe is carbidic. On the other hand, it is known that if the solidification and cooling rate are retarded sufficiently, for example by using sand linings or a wet spray refractory coating of the order of .10" thick, a pearlitic structure containing minor amounts of carbide or ferrite can 'be obtained. In other words, at rapid freezing rates the pipe are substantially carbidic, but as the mold is provided with increasing coating thickness, the carbidic chilled structure is replaced by a pearlitic structure. In either case, it is necessary to anneal the pipe in order to obtain a ferritic structure which will meet the specifications as follows:

Tensile strength p.s.i 60,000 Yield strength p.s.i 45,000 Elongation "percent" 10 Charpy impact, 70 F. ft./lbs 7 Charpy impact, 40 F. ft./lbs 3 In this anneal the pipe are first heated to austenization temperature (eg 1720 F.) and then ferritized at a lower temperature (e.g. 1375 F.). While it is possible to ferritize the pearlitic pipe by use of only a secondary ferritizing anneal, the impact values are impaired by the large size of the ferrite grains, and possibly by sub-boundary structures which are associated with the slow freezing rates.

In ductile irons the number (and size) of the graphite spheroids that occur is a function of the freezing rate and the graphitization potential of the metal. The more rapid the freezing rate, the smaller the nodules become, until the freezing rate becomes so rapid that graphitization is essentially suppressed and large amounts of cementite form. Conversely, as the freezing rate becomes slower there is less tendency to form primary cementite, the

graphite nodules grow larger and pearlite is formed. Austenite which occurs at just below the solidification temperature can transform into pearlite plus graphite or into ferrite plus graphite. However, these are diffusion reactions and the transformation to ferrite plus graphite cannot occur to any appreciable extent during the cooling down of iron castings in normal casting procedures. Thus in order to convert the castings to ferrite it is necessary to subject them to an extended annealing cycle. On the other hand, the transformation to pearlite plus graphite can easily be obtained by using slow cooling rates and normal inoculation procedures.

The use of inoculation to induce graphitization in ductile iron is well known, and inoculation prior to pouring is the general practice. It is also known in the casting of pipe that the use of a coating of ferrosilicon or calciumsilicon on the metal mold, either bare or dressed with a conventional wet spray coating, will result in finer graphite nodules and reduced chill. These pipe are easier to anneal for carbide and pearlite removal because they have a fine grain structure, fine graphite nodules and reduced chill. Also, it is known that the placing of powdered inoculant in a sand-lined mold or the addition of inoculant to the metal as it is poured into the mold will result in improved graphite nodule structure, more spherical, less carbides, and small amounts of ferrite in the as-c ast structure. However, in every case it has been necesary to anneal all such pipe to obtain a satisfactory ferritic structure.

We have now discovered that by the proper control of composition, inoculation, freezing rate and cooling rate it is possible to obtain as-cast structures in ductile iron pipe which are chill free and contain large percentages of ferrite. While the previous uses of late inoculation during pouring of the casting were a move in the right direction, it has not been recognized previously that these procedures did not give optimum graphitization conditions. Furthermore, while previous methods provided for chill free casting by slowing the cooling rate of the metal, the resultant as-cast structures were substantially pearlitic. It was unknown that with a properly inoculated iron of proper composition there is a critical range of freezing and cooling rates between those commonly used in metal mold casting methods and those commonly used in sand casting methods wherein carbides will 'be suppressed and graphitization substantially completed so that the resultant matrix is substantially ferritic as cast.

The advantages of this invention will be apparent from the following description taken in conjunction with the drawings which are photomicrographs of etched specimens at X as follows:

FIGURE 1 is a typical as-cast microstructure of a 6" pipe cast on a thin wet spray coating in a De Lavaud pipe machine.

FIGURE 2 is a typical as-cast microstructure of a 6" pipe cast on a powdered inoculant coating applied to a bare mold in a De Lavaud casting machine.

FIGURE 3 is a typical as-cast microstructure of a 6" pipe cast on a thin wet spray coating overlaid with a coating of powdered inoculant in a De Lavaud pipe machine.

FIGURE 4 exemplifies as-cast, essentially chill free microstructures which are obtained with the use of resin sand or thick wet spray coatings for 6" pipe.

FIGURES 5-7 are as-cast microstructures of three different 24" pipe cast on three different thicknesses of Wet spray coating using conventional inoculation practices.

FIGURES 8-10 are as-cast microstructures of three pipe cast under similar conditions as the three above mentioned 24" pipe except that special inoculation practices were used.

FIGURES ll-l are as-cast microstructures of five 48 pipe cast using varying wet spray coating thickneses and inoculation procedures.

As can be seen by a study of FIGURES 1-3, presently used methods of casting ductile iron pipe in metal molds result in highly chilled, carbidic structures. On the other hand, pipe cast on resin bonded sand molds, a well known commercial practice, are substantially chill free but have a substantially pearlitic structure such as is shown in FIGURE 4. In contrast, the structures shown in FIG- URES 8-10 which were obtained by the present method are substantially carbide free and contain significant amounts of ferrite in the as-cast condition.

The process of the invention based on the above mentioned discovery broadly stated comprises: providing an iron of suitable composition, inoculating the molten metal to provide a high concentration of graphitization nuclei in the metal just before and during solidification (hereinafter referred to as ultra late inoculation), entrapping this transient nucleating effect at its maximum effectiveness by rapid freezing of the metal, and cooling the casting at a rate suitable to promote the transformation of austenite to ferrite plus graphite.

Iron compositions for use in the present process preferably are those which will result in a final pipe composition within the range of T.C. 3.13.4 Si 2.8-3.3 Mg .03.06

Within this composition range it is possible to obtain pipe with less than 5% pearlite and substantially no carbides, which pipe exhibit properties in the as-cast condition meeting the specifications stipulated for ductile iron pipe in this country. Of course, it will be apparent to those skilled in the art that the properties obtainable are dependent upon the total effect of the alloying elements; therefore, in order to get the desired properties a proper balance must be retained in the final composition; for example, the presence of higher amounts of carbide stabilizers such as magnesium, manganese, chromium, vanadium, etc., could not be tolerated in compositions having minimum graphitizing potential.

A still broader range of compositions which may advantageously be applied in plant production is as follows:

In this composition range the pearlite content may be held to less than 75%, the pipe will be substantially chill free, and the graphite nodules will be numerous and small in size. Generally, if a proper compositional balance is maintained in terms of the carbon to silicon relationship, pipe having compositions within this range will have satisfactory properties after a simple ferritizing anneal at about 1375 F.

Proper inoculation of the melt is essential if the desired results are to be obtained. Optimum inoculation of the melt is accomplished by introducing inoculant into the metal during pouring and solidification, and it is aimed at obtaining an extremely high population of graphitization nuclei in the melt just before and during solidification. When centrifugally casting pipe by the retractive pour method wherein the metal is distributed along the length of the mold by means of a long pouring trough, the inoculation procedure is preferably carried out by (1) post magnesium treatment inoculation in the ladle, (2) addition of granular inoculant to the metal as it flows down the chute which directs it into the pouring trough, (3) application of a coating of powdered inoculant to the mold surface upon which the metal is cast and (4) blowing inoculant powder into the stream of metal as it is discharged from the pouring trough in such a manner as to create a fog of inoculant powder inside the mold. When the proper size material is supplied at suitable rates which do not result in undissolved inoculant in the final product, there results a continuing formation of graphitizing nuclei in large numbers in the metal in the mold.

The underlying mechanism of the ultra late inoculation therefore appears to be one of providing a degree of dissolution of the inoculant such that on the one hand all discrete solid particles of inoculant become molten and an the other hand insufficient time is provided for complete homogeneous diffusion within the parent melt of the graphitizing constituents in the inoculant.

The post magnesium treatment inoculation procedure, usually minutes prior to pouring the casting is well known. In the present method, the procedure is the same, but the amount of inoculant added at this stage will depend upon the amount of inoculant to be added during the casting operation as well as upon the composition of the metal. Of course, other commonly used procedures such as add ing the inoculant with or as part of the magnesium treating agent can be used.

The ultra late inoculation procedures used during casting are as follows:

Chute inoculati0n.As molten metal is poured from the machine ladle at a uniform rate into the pouring chute, inoculant is added at a uniform rate throughout the period of pouring. The amount of inoculant added can vary over a relatively wide range, but a suitable working range has been found to be equal to .05.07 to 20% of the Weight of the pipe cast. An average addition of .12% is equivalent to .07% Si as calcium silicon or .10% Si as calcium bearing ferro-silicon. The inoculant is added in granular form of small size to permit dissolution before the metal solidifies but of large enough size to provide inoculant effect up to the time of solidification.

Spout inoculation-As the molten metal falls from the pouring spout of the casting trough to the mold surface inoculant powder is blown continuously into the stream of metal pouring from the trough. The amount added can vary over a wide range, but a working range of .075 to .25 by weight of the metal cast has been found to be satisfactory. An amount of .16% is equivalent to an addition of .10% silicon as calcium silicon and .14% silicon as 85% grade calcium containing ferro-silicon. Since the time between addition and solidification is extremely short, the powder should be sufficiently fine to permit it to go into solution, and it should not be added in amounts in excess of that which will go into solution.

Coating inoculation.As the molten metal falls from the spout of the pouring trough, it falls onto the mold which has been provided with a dual coating. The first layer of the coating is a refractory coating, and the second layer is a loose powder coating of an inoculant such as CaSi (hereinafter referred to as dry spray). From 5 to 15 grams per square foot of mold surface (1% to 5 pounds per ton of metal) comprises a good working range for the application of the dry spray coating. A nominal coating of 10.0 grams per sq. ft. amounts to an alloy addition of .12% for a 24" pipe having a wall thickness of .50" or is equivalent to a silicon addition of .07%. This contact with the inoculant from the outside together with the contact with inoculant from the inside resulting from the cloud of inoculant blown into the metal at the spout gives a thorough last moment inoculation of the metal before solidification.

The refractory coating referred to above is preferably a wet spray coating such as is well known in the centrifugal art. It is formed by spraying a slurry of refractory material such as bentonite and silica flour onto a hot mold so as to obtain a coating having the desired insulation value and a rough surface which assists in obtaining proper distribution and pick up of the molten metal in the rotating mold.

Calcium silicon (60% Si, 30% Ca) has been found to be one of the most potent inoculants for ultra late inoculation of ductile iron. It is commercially available in a form designated as 100 mesh which gives good results when used for chute, spout and mold coating additions. A representative size distribution is US. Standard Sieve: Percent retained This material has been found satisfactory for chute, spout and coating inoculation. Calcium-silicon of the size distribution given above has merely been set forth as one suitable material and it is pointed out that material of other size and distributon can be used as well as other known inoculants such as ferro-silicon. However, calcium silicon has been found particularly beneficial in obtaining good casting surfaces free of pinholes when used as the mold coating inoculant.

The use of all four inoculation procedures is preferred but it is possible to obtain sufficient inoculation by using less than four. For example, if the metal composition is too high in silicon to permit the addition of silicon inoculant during or after magnesium treatment, good results can be obtained with chute, spout and coating inoculation. Also, coating inoculation can be eliminated if good chute and spout inoculation is obtained, however, the use of the calcium-silicon coating is particularly desirable if good casting surfaces are diflicult to maintain or pinholing is encountered. The four inoculation methods have been set forth as examples of the various procedures which can be used to obtain the necessary number of inoculation nuclei in the metal at the time of solidification and are not set forth as limits on the scope of the invention.

The solidification rate and the cooling rate after solidification are as important as metal composition and ultralate inoculation. In the examples given later, solidification time is given as a function of interior mold coating thickness, etc. This time was measured from the beginning of pour to the end of eutectic arrest. The shorter the time the more rapid the solidification rate. The cooling rate referred to is the rate at which the casting cools through the transformation range; a temperature range below that of solidification. It is necessary to solidify the metal at a rate of speed sufficiently rapid to capture the nucleating effect of the ultra-late inoculation, but not too fast to result in the formation of carbides. After solidification it is essential that the casting cool at a rate which permits the desired ferritization to occur.

The desired solidification and cooling rates can be obtained by casting pipe in metal molds which are provided with a wet spray coating between .010" and .060" thick. The mold is preferably cooled externally with either air (normal convection) or water spray as needed. The use of a mold submerged in water, as used in the De Lavaud process, is not recommended because mold temperatures are inadequate to dry the coating slurry and the solidification and cooling rates are too fast, particularly for thinner castings.

The use of a refractory wet spray coating of between .010" and .060 thickness on the interior of a metal mold is given as a practical means of obtaining the desired solidification and cooling rates, and is not intended as a limitation on the invention. Obviously, there are many Ways of altering the solidification and cooling rates of a casting, and if equivalent rates can be obtained by other procedures then their use in the described process is contemplated.

When casting pipe which are less than .50 thick, difficulties are encountered because the solidification and/ or cooling rates are too rapid and either carbides are formed or considerable pearlite is retained. -If carbides are formed, thicker mold coatings or hotter molds may be used to slow down solidification. On the other hand, if pearlite is retained because the cooling rate is too rapid, the use of additional mold coating insulation may result in a solidification rate too slow to capture the full inoculation effect. Should this happen, the cooling rate can be slowed by an alternative method, namely by applying, usually by blowing in, a blanket of dry insulating material onto the inside of the freshly generated pipe, thereby not affecting the solidification rate.

In establishing the desired casting conditions, it is necessary to keep in mind that changes made to speed up the solidification rate will generally have an adverse effect on the rate of cooling the casting through the transformation temperature range, and slowing the cooling rate will adversely affect the solidification rate. In other words, attempts to speed up the solidification rate to capture the inoculation effect may at the same time result in increasing the cooling rate through the transformation range to the point where excess pearlite is retained, and attempts to slow the cooling rate to permit transformation to ferrite may result in loss of inoculation effect during solidification. If it is necessary to take measures to slow the solidification rate, there will generally be a corresponding favorable decrease in the cooling rate, but if it is necessary to slow the cooling rate, there will generally be an unfavorable decrease in the solidification rate unless the measures taken are carefully designed to affect only the cooling rate.

To illustrate the effects of modifying solidification and cooling rates, along with various melt chemistries and ultra late inoculation practices, on the microstructures and mechanical properties of different pipe, the following examples are offered:

Experimental 24" pipe having a length of 20' were cast in an air cooled metal mold rotated on a flat floor spinner. The mold had a wall thickness of 2" and the metal was retractively poured by means of a pouring trough and machine ladle mounted on a car which moved on tracks parallel to the axis of the mold. The metal was melted in an induction furnace and the target composition in the final pipe was the preferred range set forth above. The air cooled mold was internally coated by spraying with a slurry of diatomaceous silica and bentonite in water while the mold was hot so as to obtain rough insulating coatings of desired thickness. Using mold coating thicknesses of .015", .035" and .055" two series of castings were made. In the first series, the only inoculation given the metal was the conventional addition of 1% ferro-silicon Si, .5 Ca min.), added to the metal as it was reladled from the magnesium treating ladle to the transfer ladle. -In the second series of tests, 1% ferrosilicon was used to cover the Mg alloy during treatment, but during the casting operation, chute, spout and mold coating additions of CaSi were made. The results are shown in FIGURES 5-10 and Table I.

It can be seen that as the coating thickness increases, the graphite nodules become larger and fewer in number. The fact that there is less carbide in FIGURE 5 than in FIGURE 6 is due to the interaction of freezing rate and inoculation, i.e., the freezing rate was rapid enough to capture the waning effects of the reladle inoculation, but the graphitization power of this remaining effect was not sufificient to prevent the formation of carbides. In sharp contrast, FIGURES 8, 9 and 10 illustrate the effect that ultra late inoculation has in eliminating carbides and permitting the formation of large amounts of ferrite as the casting, while still in the mold, cooled through the ferritizing temperature range.

FIGURES 11, 12 and 13 illustrate the interaction of the variables in the process. These figures are photomicrographs taken from 48" diameter pipe cast in a metal mold having a thickness of approximately 3%". The mold was rotated in a flat floor spinner and was provided with a wet spray coating overlaid with a dry inoculant powder coating of CaSi. The metal was poured into the air cooled mold rapidly by means of a stationary horngate and was distributed in the mold by means of centrifugal force. An analysis of the microstructures together with the data set forth in Table II reveals the importance of both ultra-late inoculation and coating thickness. With the use of a dual coating comprising .03 wet spray and a thin dry inoculant powder coating, carbide is present (FIGURE 11). When the wet spray coating is increased to .09", the carbides are eliminated, but the matrix is primarily pearlitic (FIGURE 12). However, when the coating thickness is decreased to .050, the amount of dry spray coating is increased and inoculant is added to the metal in the horngate, carbide is eliminated and pearlite is decreased (FIGURE 13).

FIGURES 14 and 15 show the microstructures obtained in similar pipe retractivity cast by means of a pouring trough in the mold mentioned above. As the freezing rate is decreased over that used in conventional De Lavaud casting by using a mold coating of .03, freedom from carbide and reduced pearlite was obtained by using chute and coating inoculation, and finally a substantially ferritic structure was obtained when a wet spray coating thickness of .023" was used together with chute, spout and coating inoculation. See Table II which compares the casting conditions and properties of the pipe represented by FIGURES 11-15. y

In order to determine the effect of mold temperature at the start of casting, another series of castings was made in a 6" x 6' steel mold having a thickness of 1.125". The mold was mounted on a movable carriage on which means were provided for applying various time periods of water spray on the mold exterior. The metal was poured by means of a fixed elongated pouring trough. In addition to the use of ladle inoculation, ultra-late inoculation using CaSi was accomplished by adding 800 grams at the chute, by blowing in 1200 grams at the trough spout, and by applying 800 grams as a coating on top of the insulating wet spray coating. From the data in Table III it can be seen that the amount of retained pearlite was greatly reduced by the use of a hotter mold and reduced amounts of spray cooling. As would be expected, the thinner castings with their more rapid cooling rates tend to have retained carbides and pearlite, but these tests indicate that by proper control of casting conditions a substantially as-cast ferritic pipe can be obtained even with thickness as low as .35".

TABLE I Fig. Fig. 6 Fig. 7 Fig. 8 Fig. 9 Fig. 10

Composition of Pipe:

T.C 3. 34 3. 36 3. 32 3. 32 3.26 3. 3O 2. 91 3. 01 2. 92 3.07 3.05 2. 81 009 013 009 008 007 015 .23 .24 .23 .23 .20 .21 .03 .03 .03 .04 .04 .04 Mg .046 .058 .00 061 .049 035 Wall Thickness of Pipe, Inch .45 .46 49 .45 50 49 Coating Thickness, Inch... 015 035 055 015 035 055 solidification Time, Sec... 50 80 120 45 70 120 Inoculation:

Chute. No No No Yes Yes Yes Spout... No No No Yes Yes Yes Coating. No No No Yes Yes Yes Physical Prope T.S., Lbs. per sq in. 81, 000 90, 000 90, 000 Y.S., Lbs. per sq. 1 No Tests 57,000 ,000 60, 000 Elong., Percent 15. 5 12 Charpyrlmpact, F 11 70 t 8. 25 6. 25 -40 Tests 6.75 5 3 Microstructure, Perc Pearlite 70 50 65 5 Carbide 10 4O 15 0 0 0 TABLE II Fig. 11 Fig. 12 Fig. 13 Fig. 14 Fig. 15

Composition of Pipe:

T.C 3. 70 3. 43 3. 13 3. 32 3. Si.-- 3.08 2. 83 3. 21 3.14 3. 17 S 003 005 002 004 003 Mn 33 30 32 24 25 P". .19 .02 .02 .09 .02 Mg 031 068 049 049 036 Wall Thickness of Pipe, InclL- .82 72 65 72 72 Coating Thickness, Inch 03 09 05 03 023 solidification Time, Sec 150 300 210 120 100 Inoculation:

Ladle, Percent Si 75 1 1 Chute, Lbs. CaSi. No No 1% 1% Spout, Lbs. CaSL No No No 2 Coating, Lbs. CaSi .25 1 3 2 2 Physical Properties:

T.S., Lbs. per sq. in 71, 000 84, 000- 90,000 85, 000 77, 000 Y.S., Lbs. per sq. m 63, 000 56, 000 62, 000 60,000 55, 000 Elonm, Percent 10 10 10 17. Charp yBImpaet, Ft.-lbs.

70 3. 75 3. 25 4. 0 14. 5 Tests 2. 0 2. 0 2. 5 6.5

l 2 lbs. CaSi in horngate.

TABLE III Wet Spray Pipe Composition Mold Temp. Pipe Coating, Wall, Pipe Water Spray Percent At Time of No Inch Inch T C on Mold, See. Pearlite Casting, F.

. 012 36 3. 28 2. 91 (i0 230 012 .35 3. 28 3.01 20 25 450 012 36 3. 22 3. 06 None 7 650 .0 12 43 3. 50 2. 79 25 230 012 44 3. 50 2. 82 30 7 450 012 .42 3. 52 2.82 None 10 550 In a further series, pipe were cast in a 16" X 18' production type casting machine. The steel mold had a wall thickness of 1.42 and was mounted on a carriage which moved up inclined tracks as in the case in the De Lavaud method. Instead of the mold being submerged in a water box as is the usual De Lavaud practice, the mold was surrounded by a vented closure provided with an interruptable water spray system. A wet spray mold coating comprising diatomaceous silica (Celite 315) and bentonite binder was applied to the interior of the mold in a single pass. The coating had a fine toothy surface and a thickness of approximately .025". With a pouring temperature of 2350 F. and a pipe wall thickness of .50", the freezing time was found to be approximately 45-50 seconds. The mold temperature at the time of coating was 450 F. to 650 F. and at the time of pouring approximately 300 to 400 F. and the mold exterior was spray cooled for approximately two minutes per cycle. The metal was melted in a cold blast acid cupola, soda ash desulfurized and magnesium treated. In addition to ladle inoculation, chute, spout and mold coating inoculation was used as indicated in Table IV which lists representative pipe.

In annealing tests carried out on these 16 pipe it was found that after a simple ferritizing heat treatment all the pipe, including those containing as little as 2.34% silicon, contained 0-l0% pearlite and met the existing specifications. The heat treatment consisted of heating the pipe to 1375 F. and holding for minutes after which the pipe was air cooled. After such a heat treatment, pipe No. 7 had a Charpy value of 12 ft.-lbs. at F. and pipe No. 8 had a Charpy value of 9 ft.-lbs. at 70 F. Also, the ferritization heat treatment improved the impact value of pipe Nos. 9, 10 and 11.

Further series of pipe were cast in a 30"x 5.75 cast iron mold having a thickness of 4.25. The air cooled mold was mounted on a flat stationary floor spinner and the metal from an induction furnace was poured with a retractive trough. The first three pipe listed in Table V, Nos. 12-14, were cast, and again it is seen that a highly ferritized pipe resulted. As would be expected, when the silicon content was 2.49, the pearlite content increased. However, it was found that by applying an insulating layer of diatomaceous silica powder one and one-half inches thick on the inside of the pipe as soon as the metal TABLE IV Composition Inoculation, Grams/pipe Pearlite, Charpy Impact Percent 70 F., Ft.-lbs. Si Chute Spout Mold These tests indicate that if the silicon content is maintained in the approximate range of 2.8-3.3% that subwas cast, the pearlite content was reduced from 25% to 8% in a pipe containing approximately 2.50% silicon.

Pipe Nos. 15-17 were cast in a 36"x5.75 cast iron mold having a wall thickness of 5.6". The induction furnace metal was cast using a floor spinner and retractive trough, and the wet spray coating material was diatornaceous silica.

Using the same type machine and process as described above for 16" pipe, pipe 24''x 18' were cast in a mold having a thickness of 1.81". In addition to ladle inoculation CaSi was used at the chute, spout and mold in the amounts 800 grams, 1200 grams and 800 grams respectively. Representative pipe are listed in Table VI.

TABLE VI Wet Spray Pipe Composition Water Charpy, Ft.-ll0s.

Pipe Coating Wall, Spray, Pearlite,

No. Thickness, Inch 0 Sec. Percent +70 F. --40 F.

Inch

zation. However, it is truly amazmg that this p1pe was In addition, heavier wall p1pe were cast in 48" molds chill free and contained considerable ferrite as cast.

using induction furnace metal, a stationary floor spinner,

a retractive trough and diatomaceous silica wet spray coatings. In Table VII pipe Nos. 27-31 were cast in a cast iron mold 19 long having a wall thickness of 4 A, and pipe Nos. 32 and 33 were cast in a steel mold 6.5

2. Method according to claim 1 wherein the metal is poured into the mold by means of a horngate and inoculant material is added to the molten metal as it is poured into the horngate and as it pours out of the horngate into long with a wall thickness of 3.65". the mold.

TABLE VII Wet Spray Pipe Composition Inoculation, Lbs. Pear-lite, Charpy Pipe Coating, Wall, Percent Impact 70 F.

No. Inch Inch 0 Si Chute Spout Mold A study of the data listed in Tables III-VII reveals that when pipe having a composition within the preferred range and a wall thickness of approximately .50" or greater are cast using the present invention, they are consistently chill free, substantially ferritic and meet the specification for ductile iron pipe without annealing. On the other hand, when the silicon content falls outside the preferred range and/or the pipe thickness falls below .50", the results are not as uniform. The as-cast pearlite content varies over a broader percentage range, and the physical properties, while impressive, do not consistently meet the specification. Also, when the silicon content is higher than the preferred range, the Charpy impact values are impaired, as would be expected, and when this occurs in a pipe having 10% or more of pearlite, the impact values generally do not meet the specification. However, annealing tests have shown that these pipe, after a simple ferritizing anneal, have physical properties within the specification for ductile iron pipe.

From the above, it is apparent that the process of the present invention is particularly suited for making large diameter pipe which have greater wall thicknesses. On the other hand, the value of the process for casting smaller size pipe, even where the composition cannot be maintained in the preferred range, should not be overlooked, because substantial economies in annealing are effected, owing to the lower temperatures and faster annealing rates made possible.

We claim:

1. A method for casting ductile iron pipe wherein magnesium treated cast iron is cast in cylindrical centrifugal molds provided with a refractory coating and the metal is treated with an inoculating material to promote the formation of graphite, characterized in that, the mold is coated with a refractory coating between .01 and .06 inch thick, a coating of powdered inoculating material is ap plied over the refractory coating, the metal is treated with inoculating material as it is being poured into the mold, and the composition of the metal in the resulting pipe comprises 3.0-3.6% carbon, 2.30-3.75% silicon and .02.0'7% magnesiuum.

3. Method according to claim 1 wherein the metal is cast by retractive pouring by means of a long pouring trough and inoculating material is added to the molten metal at the trough chute and inoculant is blown into the stream of metal flowing from the trough spout onto the mold surface.

4. Method according to claim 3 wherein the mold is cooled by spraying water during part of the casting cycle.

5. Method according to claim 3 wherein from 5 to 15 grams of inoculant material are applied to each square foot of mold surface, from .05% to 20% of inoculant is added to the metal at the trough chute, and from .075 to .25% of inoculant is added to the metal at the trough spout.

6. A method according to claim 1 wherein the pipe are given a ferritizing anneal after they have been extracted from the casting mold.

7. A method according to claim 1 wherein the composition of the metal in the resulting pipe comprises 3.13.4% carbon, 28-33% silicon and .03-.06% magnesium and the pipe are substantially ferritic as cast.

References Cited UNITED STATES PATENTS 1,329,296 1/ 1920 Lavaud 1641 18 2,248,693 7/1941 Bartscherer 164-118 2,399,606 4/1946 Schuh et al 164-72 3,056,692 10/1962 Kitada 16472 X 3,321,304 5/1967 Snow 130 FOREIGN PATENTS 1,269,898 7/1961 France. 371,855 1962 Japan.

I. SPENCER OVERHOLSER, Primary Examiner.

E. MAR, Assistant Examiner.

U.S. Cl. X.R.

Images