ENGINEERING RESEARCH INSTITUTE UNIVERSITY OF MICHIGAN ANN ARBOR PROGRESS REPORT RISERING OF DUCTILE CAST IRON By R. A. FLINN Associate Professor of Chemical and Metallurgical Engineering and Production Engineering W. A., SPINDLER. Assistant Professor of Production Engineering Project M862 THE INTERNATIONAL NICKEL COMPANY, INC. NEW YORK, NEW YORK June, 1953

'/^^ )J c~~~~~~~~~~~~~~~Lrir~~~~~~~~~~~~~~~~~~~~~~~ — - - -r — n... e"71~ UMae. Ivn~

TABLE OF CONTENTS Page LIST OF TABLES iii LIST OF FIGURES iv I. SUMMARY II. PROCEDURE 3 Castings Investigated 3 Riser Design 5 Neck Design 5 Gating 8 Mold Conditions 8 Clamping vs. Weighting of Molds 8 Melting Conditions, Compositions Explored 8 Inspection 11 III. DISCUSSION OF DATA 11 Section A. Risering Curve for Ductile Iron 14 X-Axis 14 Y-Axis 14 Section B. Variables Affecting Riser Performance 16 1. Neck Design 16 2. Mold Restraint 18 3. Pouring Temperature 18 4. Composition 18 IV. RECOMMENDATIONS FOR FURTHER WORK 20 APPENDIX 21 ii

LIST OF TABLES Page I. EXPERIMENTAL CASTINGS USED IN THIS INVESTIGATION 3 II. SUMMARY OF ANALYSES OF ALL HEATS 12 III. EFFECT OF MOLD RESTRAINT ON SHRINKAGE AND CASTING DIMENSIONS 18 IV. EFFECT OF POURING TEMPERATURE ON SHRINKAGE 18 V. EFFECT OF METAL COMPOSITION ON RISERING 20 iii

LIST OF FIGURES Figure Page 1. Tentative Positions of Risering Curves for Ductile Iron. 2 2. Seven Castings Used in This Investigation, Shown Sectioned through Risers. 4 3Ao Dimensions for Conventional Necks for Side Blind Risers on Ductile Iron Castings. 6 3B. Dimensions for Knock-off Necks for Side Blind Risers on Ductile Iron Castings. 7 4. Gate Design Used in This Investigation 9 5. Effect of Mold Restraint. 10 6. Four Types of Defects Encountered in This Investigation.13 7. Effect of Height of Neck. 17 8. Effect of Pouring Temperature. 19 iv

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN PROGRESS REPORT RISERING OF DUCTILE CAST IRON I. SUMMARY The objective of this research is to determine formulae for the most economical risering system to eliminate shrinkage cavities in ductile iron castings in the range 3.4 - 3.9% TC, 1.8 - 2.8% Si. Three types of shrinkage defects are encountered: 1) Continuous centerline shrinkage from the riser into the casting; 2) Isolated centerline shrinkage in the casting, usually near the center in the plan view; and 3) Surface shrinkage as shown by a depression in the casting skin either near the riser or near the center of the cope surface. To feed any of these types of shrinkage it is axiomatic that: (1) The riser must freeze more slowly than the casting and (2) an open feeding channel must exist between the riser and the casting. To fulfill condition (1) the following calculations are made. First the surface area/volume ratio of the casting is determined. Since this ratio governs the cooling rate, the riser dimensions must be selected to provide a smaller A/V value (slower cooling rate). In practice the ratio of A/V for the casting to A/V for the riser should be greater than 1.3; the exact values are shown in the graph, Fig. 1. The most efficient, practical side blind riser is cylindrical with height equal to diameter. A satisfactory riser of this geometry may be calculated from the simple formula: Diameter of riser = 7.8 Volume of casting Surface area of casting

1.5 - -38 TENTATIVE POSITION OF RISERING 1.4 CURVE FOR DUCTILE IRON; 40 C.E. s4.21; SIDE BLIND RISERS; CLAMPED MOLDS; GREEN SAND; 1.3 ~ ~ ~' * - GATED INTO RISERS 43 2 IN. DIA.X3-3/4 1.2 4 N.CUBE 41n BE 45 -I 1 ~~.~~~ I 48 g! Z COMPLETELY SOUND o 1 __o ~1 1 1 150 -u- ^/ I {I 2X2X6 IN, 0 0.9 ~ 53' 0.8 _ 1 56. o > \ IX3X71N. o.6 " II- 63 0.5 67 0 /~7 ~ 3/4 X4X 7IN. CASTING a. D SHRINKAGE | \: 0; 0.4 I1 Q..\ |1/2X4X71N. CASTING 0.3 ~ 3/8 X 4X7 IN. GASTING 0.2 - I I __I 83 0.1 ~ I r- ~ ~ ~ 91 0 i n__ _ _ _ _ 100 0.7 08 0.9 1.0 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8 1.9 Vc^ (COOLING RATE OF CASTING) FREEZING RATIO- SA (COOLING RATE OF RISER) Vr FIG. I 2

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN A more precise value, or a riser of different proportions, may be selected from Fig. 1. To fulfill condition (2) the effective feeding distance of a given riser must not be exceeded. For example, it is not possible to feed a sand cast 21-in. bar, 1 in. in diameter, with a single riser at one end. The effective feeding distance of ductile iron appears to be similar to that of steel, but this requires further study. Additional work is also needed regarding the effects of top open and blind risers, exothermic compounds, carbon equivalent, knockoff and neck design, larger castings, use of a common riser for several castings, and cupola melted metal. II. Procedure Since the earliest experiments with magnesium-treated iron, it has been evident that castings of this material are prone to greater shrinkage than soft gray iron castings with flake graphite. This shrinkage problem has assumed particular importance for ductile iron because the major use has been in highly stressed parts in which shrinkage defects can result in serious failure in service. Castings Investigated It seemed important, therefore, to develop general risering formulae which would insure sound castings without expensive experimentation by the producer with each new casting design. A series of standard castings of various sections and shapes within the scope of the melting equipment was therefore selected to determine if such formulae could be developed. These experimental shapes are listed in Table I below and illustrated in Fig. 2. TABLE I EXPERIMENTAL CASTINGS USED IN THE INVESTIGATION Casting dimensions, Area of Surface, Volume A/V, Weight,......... in.' in'i. i. n.in,. lbs 1/4 x 4 x 7 61.5 7.0 8.78 1.9 3/8 x 4 x 7 64.3 10.5 6.12 2.8 1/2 x 4 x 7 67.0 14.0 4.78 3.8 3/4 x 4 x 7 72.5 21.0 3.45 5.7 1 x 3 x 7 62.0 21,0 2.95 5.7 2 (Dia.) x 3-3/4 29.9 11.8 2.54 3.2 2 x 2x6 56.0 24~0 2.33 6.5 1-3/4 x 4 x 9 117.5 6300 1.87 17.0 4 x 4 x 4 96.0 64.o 1.50 17.3 _\~ - - ~3~

DA x - 3/4 O I~~~~-o 139- 1 Fig. 2. Seven Castings Used in This Investigation., Shown Sectioned through Ries

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Riser Design Since the cooling rate of a casting increases as the ratio of surface area to volume grows larger, it is obvious that a riser with a smaller A/V ratio than the casting is needed if the riser is to feed the casting. This elementary principle arises from the fact that the amount of heat in either the riser or the casting is a function of the volume, while the rate of heat dissipation is a function of surface area, The cooling rate decreases with large volume and increases with large surface area. A few simple calculations will show that a given weight of metal will exhibit the slowest cooling rate in a sphere and increasing rates in a bar and a plate respectivelyo The most efficient riser is therefore spherical but because of practical molding consideration, cylinders with heights equal to the diameters were employed. In the later stages of the work the upper portion of the cylinder was rounded to practically hemispherical shape as illustrated in the photograph, Fig. 2. Most of the work has been concentrated on side blind risers. Analgous work in steel by Caine has shown only minor differences between side blind risers and open side and top risers treated with exothermic insulating compounds. The data of this investigation can probably be applied generally, although spot checks will be made in the near future. Neck Design This problem has received much less general attention in the past than it has deserved. A well designed neck has two very important functions: (1) to provide the proper thermal gradient between the riser and the casting and (2) to allow easy knockoff without breaking into the casting but with a minimum of grinding. Considerable experimentation was required to arrive at the neck design shown in Fig. 3A. The risering formulae which have been developed apply only if this design is used. The design of 3B provides for easier knockoff. As an example of improper neck construction, the earlier phases of the work were conducted with the neck height equal to the casting section thickness. Although it was believed that this procedure would provide a maximum channel for feed metal, it had the unforeseen effect of including the neck-to-casting junction in the zone of metal which froze last. This resulted, of course, in a spongy zone at the breakoff point of the neck. The best design used to date employs a neck height equal to 1/3 the riser diameter. Atmospheric riser cores were used consistently to develop the risering curve, Fig. 1. Although they are not essential in every case, it was decided to maintain this standard practice. In general, in the lighter sections the metal seems to freeze around the riser core and render it ineffective,.~ - -~5

2 D D ~ ~D D 1 * _ II CASTING.+.DIMENSIONS FOR CONVENTIONAL NECKS FOR SIDE BLIND RISERS ON DUCTILE IRON CASTINGS FIG. 3A 6

18 CASTING F-I. D D 6 DIMENSIONS FOR KNOCK-OFF NECKS FOR SIDE BLIND RISERS ON DUCTILE IRON CASTINGS. FIG. 3B 7

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Gat ing A standard gating arrangement delivering hot metal to the riser was used as shown in Fig. 4. The same tapered sprue and runner areas0,.31 in.2 and 0.375 in.2 respectively, were used throughout the work. Mold Conditions Since the most common sand mixture is of the greensand-seacoal type, the following standard mix has been used throughout the investigation. New sand was used each time, so that the reconditioning variable was not encountered. Mix (lbs ) Properties 200 Michigan City Sand A.F.S. 55 Moisture 2.9 - 303 % 9 Western Bentonite Permeability 130 - 150 10 Seacoal Green Comp. Str. 7 - 10 psi 6 Water Green Deformation.017 -.021 psi Dietert Flowability 75 - 80 Hand ramming to a mold hardness of approximately 77-85 has been followed. Patterns have been mounted for accurate control of neck dimensions. Clamping vs. Weighting of Molds In the early stages, two 40-lb weights were used with 9 x 9-in. flasks, but castings were consistently oversize with a crown in the center (Fig. 5). It is believed that this effect was caused by the graphitization reaction (expansion). Clamping has minimized the growth of castings, restricting the graphitization expansion to the original mold volume, and has provided sound castings with minimum riser size. The risering curve is therefore based on clamped molds. Melting Conditions, Compositions Explored Thirty heats were made in the 150-lb high-frequency induction furnace with a basic magnesite lining. The charge was composed of Lyle pig, steel punchings, and standard ferroalloys. The melt was heated to 2770~F and tapped into a preheated ladle with the addition of 2% of NMA No. 1 alloy. Instead of using another ladle, the metal was poured back in the hot furnace while adding.55% Si as 85% ferrosilicon and then immediately returned to the ladle. No power was applied after the initial tap. After the initial survey of pouring-temperature effect, the pouring temperature was maintained in the range 2540 - 2620 ~F. 8

Fig. 4. Gate Design Used in This Investigation.

Fig. 5. Effect of Mold Restraint. 1 x 3 x 7-in. Castings Sectioned through Risers. Casting No. 76-3 Made in 13 x 13-in. Flask and Weighted; No. 76-4 in 13 x 13-in. Flask and Clamped; No. 76-5 in 9 x 9-in. Flask and Clamped.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN Three analysis ranges were surveyed, but the principal attention was given to the 3.6% C and 1.8% Si type. TC Mn P S 1 3.6.2.02 1.8 2 3.56.02 2.5 3 3.9.2.02 2.5 A summary of the analysis is given in Table II. Inspect ion In a research problem of this type, it is difficult to select an inspection tool for shrinkage that will be universally acceptable. As a compromise between the overexacting microscopic inspection for microshrinkage and the crude fracture test, a sectioning and sandblasting technique was developed. A cut was made through the center of the riser and the centerline of the casting with a soft abrasive cut-off wheel, using a voluminous supply of coolanto The cut surfaces were then carefully sandblasted to remove any flowed metal and to expose shrinkage. This technique was more delicate than hot acid etching, which was also tried. Photographs of typical castings are shown in Fig. 6. In addition to the chemical analysis, a metallographic inspection was made of each heat. Completely spheroidal graphite was obtained consistently. Some carbides were obtained in the lighter sections and further work is needed on the effects of these free carbides. III. DISCUSSION OF DATA The most useful summary of data is contained in risering curve, Fig. 1, and this will be discussed first (Section A, below). It is important to understand that this curve holds true only for clamped molds; with side blind risers and of the neck design of Figo 3A, The curve is probably a good first approximation, ie., to 1/4-in. riser diameter, for open insulated risers or top risers, but these points require further study. The observations on which Fig. 1 are based are contained in Appendix Io Following the discussion of the risering curve the effects of the following variables will be considered in Section B: (1) neck design, (2) mold restraint, (3) pouring temperature, and (4) metal compositiono 11

TABLE II SUMMARY OF ANALYSES OF ALL HEATS Heat No. T.C. Si Mn P Date 33 3.68 1.88.28.02 June 25, 1952 34 3.53 2.07.25.02 July 3, 1952 36 3.60 1.62.23.02 July 8, 1952 39 3.68 1.75.21.02 July 15, 1952 41 July 16, 1952 51 3.63 1.86.26.03 August 4, 1952 53 3.48 2.70.26.03 August 7, 1952 54 3.59 2.55.23.03 August 8, 1952 61 3.45 2.34.24.02 August 13, 1952 62 3.61 2.55.31.02 August 13, 1952 64 3.56 2.48.24.02 August 14, 1952 65 3.54 2.51.25.02 August 14, 1952 66 3.61 2.50.23.02 August 19, 1952 67 3~59 2.65.27.03 August 29, 1952 68 3.58 2.45.26.03 76 3.57 1.80.26.03 October 3, 1952 88 3.59 1.74.26.03 October 24, 1952 99 3.58 1.89.25.03 November 7, 1952 111 3.56 1.88.26.02 November 22, 1952 113 3.54 1.87.25.02 November 25, 1952 132 3.61 1.83.28.01 January 5, 1953 135 3.65 1.76.26 01 January 10, 1953 136 3.70 1.79.24.03 January 20, 1953 138 3.66 1.93.28.03 January 26, 1953 139 3.58 1.72.23.03 February 2, 1953 142 3.60 1.76.28.03 149 3.98 2.38.20.03 February 23, 1953 190 3.76 2.49.29.03 March 31, 1953 198 3.80 1.93.20.03 April 15, 1953 220 3.75 1.92,27.02 May 2, 1953 12

Fig. 6. Four Types of Defects Encountered in This Investigation. Casting No. 220-5 Has Riser Shrinkage; No. 138-5 Has Centerline Shrinkage; No. 135-3 Has Surface Shrinkage in Addition to Centerline; No. 65-3 Has Bulging (Thicker at Center than at Edges ).

- ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN Section A, Risering Curve for Ductile Iron The most familiar approach to riser dimensioning has been developed by Caine for steel and discussed in many issues of the A. Fe S. Transactions. It will be shown that this is perhaps unnecessarily complicated for ductile iron because of the difference in solidification behavior compared with steel, First, the axes of Fig. 1 should be explained. In the case of steel, shrinkage cavities can arise from two sources' lo The riser freezes before the casting and therefore, instead of supplying liquid metal to compensate for liquid-to-solid shrinkage, may actually draw liquid from the casting and add to the casting shrinkage cavity. 2. The riser freezes after the casting but the quantity of liquid irequiredby the casting cannot be provided completely by the riser. These two important variables are represented by the x and y axes respectively as follows: X-Axis: Ratio of Cooling Rates of Casting and Risering. As discussed earlier, the cooling rate is proportional to the surface area/volume ratio, since the area determines rate of heat dissipation and the volume determines quantity of heat. The higher the A/V ratio, the faster the cooling rate. Therefore, if the casting is to solidify before the riser, the A/V for the casting must be greater than the A/V for the riser. In other words the ratio of A/V for the casting to A/V for the riser must be greater than 1. A quick reference to Fig. 1 will show that this is true for all combinations in the sound region of the graph. Y-Axis. Now the y axis may be considered briefly. The relative amount of feed metal in the riser is given simply by the ratio volume of riser/ volume of casting. This is an important ratio in the case of steel castings of large surface and small volume such as thin plates. Here it is necessary but not sufficient for the riser to cool slower than the casting. It is easy to design a small chunky riser that would freeze slower than a large plate. The large plate also requires a large amount of liquid metal to compensate for the liquid-to-solid shrinkage of its relatively large volume, In this case the volume ratio rather than the A/V ratio become a dominating factor. In steel, the volume of the riser can never be less than 6 per cent of the volume of the casting and the curve is therefore asymptotic to this level, In ductile iron there is an advantage in the expansion react ion which occurs on solidification~ liquid -> austenite + graphite The demand for feed metal is therefore less than in the case of steel, To a first 14

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN approximation then, the only consideration is to have a riser that will freeze more slowly than the casting by a factor of 1,3, Very light plate sections show a slight tendency to follow the feed metal demand of steel and require more metal than is provided by the riser with the 1.3 ratio. This may be due to the lesser amount of graphite formed on solidification, as shown by the presence of primary carbides in these light sections. Under conditions more favorable to graphitization, such as verylow sulfur cupola metal treatedwith a minimum of magnesium, these sections may follow the behavior of the heavier castings. In this case the only consideration is the relative freezing rate of casting and riser. The most efficient side blind riser has height equal to the diameter and if this ratio is maintained a very simple formula may be derived. The surface area of the riser is tdh + 2 d2 or if h = d: td2 + 2 d2 6 d2 4 4 4 The volume is td2 d3 h or if h = d ~ 6td2 A = 4 6 V d ^ d 4 The expression A/V (casting) A/V (riser) becomes: A/V (casting) _ 1 3 6/d d riser = 7.8 A/V (casting) Volume 7.8 of casting Area As an example, the riser for a 1 x 4 x 4-in, plate may be calculated: A = 4(1 x 4) +2(4 x 4) = 48 V = 4x4 x x 1 16 = 15

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN d riser = 7-8 = 2.6 in. 3 It should be re-emphasized that the effective feeding distance of the riser must not be exceeded, For the present the formulae developed by Pellini for steel may be used in the absence of any data for ductile iron, These are as follows: Maximum Feeding Distance Shape of Casting from Edge of Riser Plates having width-tothickness ratio of 1 to 4-1/2 times casting thickness 3 or greater Bars having width equal to thickness (square cross 6J-, where T is the casting sections) thickness Therefore it would be better to consider an 8 x 8 x 1-in. plate as composed of four 4 x 4 x 1-in. sections, each with a riser of the dimensions just calculated, than to attempt to use one side riser. Of course a single, centrally located riser might be used if satisfactory cleaning techniques were available. Section B. Variables Affecting Riser Performance As mentioned earlier, the risering curve is based on standard neck design, mold restraint, pouring temperature, and metal composition. The relative importance of these variables will now be discussed. 1. Neck Design. An offhand analysis of neck design indicates that a large area of neck contact would tend to keep a feeding channel open from the riser to the casting. It soon developed, however, that a large neck prevented a satisfactory thermal gradient from the casting to the center of the riser, The result was that the riser, the riser neck, and the zone of the casting adjoining the riser formed a "hot spot" and contained shrinkage. When the riser neck was made smaller, the proper gradient was developed and the casting, neck, and riser froze in that order. These effects are demonstrated in Fig. 7. Notice the shrinkage in the casting with the full neck compared with the sound castings with neck height = 1/3 D. As a result of these experiments a standard neck dimensioning diagram has been developed (Fig. 3). Unless this diagram is followed the risering formulae just discussed do not apply. 16

Fig. 7. Effect of Height of Neck. 2 x 2 x 6-in. Castings Sectioned through Riser. Casting No. 113-13 Has 2-in. Neck Height; No. 53-5 Has 1-in. Height.

ENGINEERING RESEARCH INSTITUTE * UNIVERSITY OF MICHIGAN 2. Mold Restraint. At the beginning of the investigation the standard practice was to use two 40-lb weights on the 9 x 9-in. or 13 x 13 in flasks. It soon developed that the castings were 1/32 to 1/16 in. oversize and crowned in the center. Clamping of the molds reduced this effect as shown by Table III below and Fig. 5. Shrinkage was lessened at the same time. Apparently the graphitiza tion reaction at solidification tends to bulge the castings and if restrained the remaining liquid is forced into zones of possible shrinkage. TABLE III EFFECT OF MOLD RESTRAINT ON SHRINKAGE AND CASTING DIMENSIONS All Castings 1 x 3 x 7-in. with 2-1/2-in.-dia., 2-1/2-in.-high Riser Heat No. 76: 3.57% C, c26% Mn,.03% P, 1.80% Si Casting Mold Flask Size Shrinkage- Casting No. Restraint in. Riser Centerline Thickness 76-3 - Weighted 13 x 13 13/16 in, into 1 in. 1-1/32 in. casting 76-4 Clamped 13 x 13 3/16 in.out of 3/16 in. 1-1/64 in. casting 76-5 Clamped 9 x 9 7/16 in. out of none 1 in, casting ~. Pouring Temperature. Pouring-temperature effects were explored over a wide range (2780-2400F) in several heats. No pronounced effect was noted in the composition investigated^ as shown by Table IV below and Fig. 8, TABLE IV EFFECT OF POURING TEMPERATURE ON SHRINKAGE All Castings 1 x 3 x 7-in. with 2-1/2-ino-dia., 2-1/2-in,-high Riser Heat No. 61: 3.45% C, o24% Mn,.02% P, 2o34% Si Casting No. Pouring Temperature, ~ F. Riser Shrinkage, in, - 61-1 2780 1-1/8 into casting 61-2 2730 1-1/16 into casting 61-3 2600 1-1/16 into casting 61-4 2470 1 into casting 61-5 2400 1-1/16 into casting 4. Composition. Two heats (149 and 198) were made at higher carbon equivalents, 4.77 and 4.44 respectively. The higher-CE, irons showed sometimes more and sometimes less shrinkage (Table V). It is believed that these effects are within + 1/4-in, of the riser diameter in the casting range shown. They require further investigation, however, along with the metal history (cupola vs. induction melting), 18

Fig. 8. Effect of Pouring Temperature. 1 x 3 x 7-in. Castings Sectioned through Risers. Casting No. 6i-i Poured at 27800F, No. 61-2 at 27300F, No. 61-3 at 2600F, No. 61-4 at 2470"F, No. 61-5 at 24000F.

ENGINEERING RESEARCH INSTITUTE ~ UNIVERSITY OF MICHIGAN TABLE V EFFECT OF METAL COMPOSITION ON RISERING Riser, Heat Riser Casting in. No. C Si C.oE Shrinkage, in. 1 x 3 x 7 2-3/8 clamped 149 3.98 2.38 4.77 1-1/4 into casting 1 x 3 x 7 2-3/8 clamped 142 3.60 1.76 4.19 1-1/8 into casting 1 x 3 x 7 2-3/4 weighted 39 3.68 1.75 4.26 1-7/8 into casting 1 x 3 x 7 2-3/4 weighted 62 3.61 2.55 4,46 1-5/16 into casting 2 X 2 X 6 3 clamped 149 3,98 2.38 4.77 3/4 out of casting 2 x 2 x 6 3 clamped 139 3.58 1o72 4.15 1/2 into casting 3/4 x 4 x 7 2-1/8 clamped 198 3580 1.93 4,44 1-3/16 into casting 3/4 x 4 x 7 2-1/8 clamped 111 3.56 1,88 4.19 5/8 out of casting 2(Dia.) x 3-3/4 2-1/2 clamped 198 3.80 1.93 4.44 1-1/2 into casting 2(Dia.) x 3-3/4 2-1/2 clamped 132 3.61 1.83 4.22 5/8 out of casting IV. RECOMMENDATIONS FOR FURTHER WORK This progress report has been compiled at this time to provide the Risering Committee with a resume of the work at the University gf Michigan. Throughout the report the need for further work has been indicated, particularly on the following topics: Risering of larger castings Effective feeding distance Risering curves for open risers with and without exothermic compound s Top blind risers versus side blind risers Effects of composition and previous history of metal, e. g,, cupola metal Effect of common riser for several castings In addition to these research items, it is highly important that the risering calculations be tested for commercial castings. It is hoped that the committee members will do this and criticize the utility of the risering curve freely. We shall be glad to cooperate in any calculations or analysis of data. 20

APPENDIX

APPENDIX SUMMARY OF OBSERVATIONS OF CASTINGS USED FOR FIG. I, TEE RISERING CURVE Riser Neck Heat and Size, Dimensions, Casting Carbon in. in. Restraint Number Equivalent Soundness 2-in.-Diameter x 3-3/4-in,.-Long Casting 2-1/2 1-15/16 half circle x 7/16 Clamped 132-6 Low Completely sound 2-5/8 1-7/8 dia. x 7/16 Weighted 67-7 Int.* Surface shrink 2-5/8 1-7/8 dia. x 7/16 Weighted 54-6 Int. Riser froze over 2-3/4 1-7/8 dia. x 7/16 Weighted 53-6 Int. Completely sound 2-7/8 1-7/8 dia. x 7/16 Weighted 51-7 Low Completely sound 3 1-7/8 dia. x 7/16 Weighted 39-8 Low Completely sound 2-1/2 1-7/8 dia. x 7/16 Weighted 62-5 Int. Surface shrink 3 2-1/2 1-15/16 half circle x 1/2 Clamped 136-6 Low Riser shrink 1-3/4 in. Large atmospheric pressure core 2-3/8 13/16 x 2-3/8 to 13/16 x 3/8 x 1/8 Clamped 190-6 High Riser shrink 1-5/8 in. 2-1/2 7/8 x 2-1/2 to 7/8 x 7/16 x 9/6k Clamped 198-6 High Riser shrink 1-1/2 in. Preliminary work has been done on top blind risers. *Intermediate 4-in.Cube Casting 4-1/4 1-7/16 x 3 x 3/4 Clamped 88-6 Low Riser shrink B/8 in, 4-1/8 1-1/2 x 2-3/4 x 7/8 Clamped 99-6 Low Riser shrink 3/k in 4 1-3/8 x 3 x 13/16 Clamped 111-6 Low Riser shrink 1-5/16 in. 4-3/8 1-3/8 x 3-1/2 x 3/8 Weighted 76-6 Low Riser shrink 1/16 in. 4-1/2 2 x 3-1/2 x 1/2 Weighted 64-5 Int. Completely sound Preliminary work has been done on top blind risers.

APPENDIX (cont.) Riser Neck Heat and Size Dimensions Casting Carbon in. in. Restraint Number Equivalent Soundness 2 x 2 x 6-in. Casting 3 15/16 x 2-5/8 x 7/16 Weighted 53-5 Int. Riser shrink 1/8 in. 3 1 x 2-1/4 x 9/16 Clamped 132-4 Low Riser shrink 13/16 in. 3 1 x 2-1/4 x 9/16 Clamped 135-2 Low Riser shrink 1 in. 3 1 x 2-1/4 x 9/16 Clamped 136-5 Low Riser shrink 3/4 in. Large atmospheric pressure cored 3 1 x 1 x 9/16 Clamped 138-6 Low Centerline 3/8 in. 3 1 x 1-1/2 x 1/2 Clamped 139-5 Low Riser shrink 15/16 in. On centerline of casting 3 1 x 1-1/2 x 1/2 Clamped 139-6 Low Riser shrink 1/2 in. 3 1 x 3 x 7/16 Clamped 142-5 Low Riser shrink 11/16 in. 3 1 x 3 to 2-1/4 x 7/16 Clamped 142-6 Low Riser shrink 5/16 in. 3 1 x 3 to 2-1/4 x 7/16 Clamped 149-5 High Completely sound 3 1 x 3 to 1 x 7/16 Clamped 149-6 High Completely sound 2-7/8 1 x 2-7/8 to 1 x 7/16 Clamped 190-5 High Riser shrink 7/8 in, 2-3/4 1-15/16 x 1-3/4 x 3/8 Weighted 62-4 Int. Riser shrink 3/16 in. 2-7/8 1-7/8 x 1-3/4 x 1/2 Weighted 54-5 Int. Centerline 3/4 in. 3-1/4 1-15/16 x 1-3/4 x 1/2 Weighted 36-4 Low Centerline 1 in. 3 2 x 2-1/4 x 9/16 Clamped 132-5 Low Riser shrink 1/2 in. 3 2 x 2-5/8 x 7/16 Weighted 113-13 Low Riser shrink 3/4 in. 3-1/4 2 x 1-1/2 x 7/16 Weighted 39-3 Low Riser shrink 7/16 in. 3-1/4 2 x 2-1/4 x 3/4 Weighted 34-6 Low Riser shrink 1/2 in. Preliminary work has been done on top blind risers.

APPENDIX (cont.) Riser Neck Heat and Size, Dimensions, Casting Carbon in. in. Restraint Number Equivalent Soundness 1-3/4 x 4 x 9-in. Casting 3-3/4 1-1/4 x 3-3/4 to 1-7/8 x 9/16 Clamped 149-7 High Riser shrink 1 in. 3-7/8 1-15/16x3-7/8 to 1-15/16 x5/8x/4 Clamped 190-7 High Riser shrink 15/16 in. Improperly rammed) 1-7/8 in thick. 3-7/8 1-5/16 x 3-7/8 to 1-5/16x/8x7/32 Cramped 198-7 High Riser shrink 1-9/16 in. I x 3 x 7-in. Casting 2-1/2 7/8 x 1-3/4 x 7/16 9 x 9-in, moldClamped 76-5 Low Completely sound Flat 2-1/2 7/8 x 1-3/4 x 7/16 13 x 13-in. mold,Clamped 76-4 Low Centerline 3/16 in. 2-1/2 7/8 x 1-3/4 x 7/16 13 x 13-in. mold,Weighted 76-3 Low Riser shrink 13/16 in. 2-1/2 7/8 x 1-3/4 x 9/16 9 x 9-in, mold.,Clamped 111-4 Low Riser shrink 1 in. 2-1/2 3/4 x 2-1/4 x 1/2 9 x 9-in, moldWeighted 65-3 Int. Centerline 9/16 in 2-1/2 I x 2-1/4 x 1/2 9 x 9-in. mold,Weighted 64-1 Int. Riser shrink 1-3/8 in. 2-1/2 1 x 2-1/4 x 1/2 13 x 13-in. moldWeighted 61-1 to 61-5 Int. Riser shrink 1-1/16 in, 2-1/2 1 x x 3/8 13 x 13-in. mold, Weighted 68-1 Int. Riser shrink 7/16 in. 2-/4 1 x 2-1/2 x 736 1 x 13-in. mold,Weighted Many castings Low Riser shrink 3/4 in. 2-5/8 1 x 2-1/4 x 7/16 13 x 13-in, mold, Weighted 53-3 Into Centerline 1/8 in. 2-1/4 1 x 2 x 1/2 9 x 9-in, moldWeighted 62-2'Int. Surface shrink 3/32 in 2-3/8 1 x 2 x 1/2 -9x 9-in, mold Weighted 54-2 nt. Riser shrink 1-1/2 in. 2-3/8 3/4 x 2-1/4 x 7/16 9 x 9-in, mold,Weighted 68-4 Int, Riser shrink 1-1/4 in. 2-3/8 3/4 x 2-1/4 x 1/2 9 x 9-in, mold,Clamped 88-3 Low Riser froze over Surface shrink 2-3/8 3/4 x 2-1/4 x 9/16 9 x 9-in, mold, Clamped. 111-3 Low Centerline 5/16 in. 2-1/4 3/4 x 1-11/16 x 7/16 9 x 9-in. mold)Clamped 132-3 Low Riser shrink 2 in. Surface shrink

APPENDIX (cont.) Riser Neck Heat and Size, Dimensions, Casting Carbon in. in, Restraint Number Equivalent Soundness 1 x 3 x 7-in. Casting (cont.) 2-3/8 13/16 x 13/16 x 1/2 9 x 9-in. mold, Clamped 138-5 Low Centerline 15/16 in. 2-3/8 13/16 x 2-3/8 x 11/32 9 x 9-in. mold, Clamped 142-2 Low Riser shrink 1-1/4 in. 2-3/8 13/16 x 2-3/8 to 1-5/16 9 x 9-in. mold, Clamped 142-3 Low Riser shrink I in, x 11/32 2-3/8 13/16 x 2-3/8 to 1-7/8 9 x 9-in. moldClamped 142-4 Low Riser shrink 1-1/8 in. x 11/32 2-3/8 13/16 x 2-3/8 to 1-3/16 9 x 9-in. mold. Clamped 149-4 High Riser shrink 1-1/4 in. x 11/32 2-1/2 7/8 2-1/2 to 7/8 9 x 9-in. mold)Clamped 198-4 High Riser shrink 1-3/16 in. v x 7/16 x 9/64 Some work has been done on gating not into riser Some work has been done on end blind risers Preliminary work has been done on top blind risers 3A/4 x 4 x 7-in. Casting 2-1/8 3/4 x 1-1/2 x 9/16 Clamped 111-1 Low Centerline shrink 1/4 in. 2 11/16 x 1-9/16 x 7/16 Clamped 132-2 Low Centerline shrink 1 in. 2-1/8 3/4 x 1-1/2 x 1/2 Clamped 88-1 Low Riser froze over - surface shrink 2-1/8 3/4 x 1-1/2 x 1/2 Clamped 99-1 Low Mold crushed in clamping 2 3/4 x 1-7/8 x 7/16 Weighted 62-1 Int,* Riser shrink 1-/8 in. *Intermediate

APPENDIX (cont.) Riser Neck Heat and Size, Dimensions, Casting Carbon in. in. Restraint Number Equivalent Soundness 3/4 x 4 x 7-in. Casting (cont.) 2-1/8 3/4 x 2 x 7/16 Weighted 54-7 Int. Riser shrink 1-1/8 in. 2-1/4 3/4 x 2 x 1/2 Weighted 54-3 Int. Centerline 1-15/16 in. 2-3/8 3/4 x 2-1/8 x 7/16 Weighted 51-2 Low Centerline 7/8 in. 2 11/16 x 11/16 x 7/16 Clamped 138-4 Low Centerline 1-5/8 in. 2 11/16 x 2 to 11/16 x 5/16 Clamped 190-3 High Riser shrink 2-5/16 in. 2-1/8 3/4 x 2-1/8 to 3/4 x 3/8 x 1/8 Clamped 198-3 High Riser shrink 1-3/16 ino Preliminary work has also been done on top blind riser 1/2 x 4 x 7-in. Casting 1-1/2 1/2 x 1-1/8 x 5/16 Clamped 135-1 Low Riser shrink 1-7/16 in, 1-1/2 1/2 x 1-1/8 x 5/16 Clamped 132-1 Low Riser froze over surface shrink 3/64 in. 1-1/2 1/2 x 1-3/8 x 3/8 Weighted 62-7 Int,* Riser shrink 2-1/2 in. 1-5/8 1/2 x 1-1/2 x 3/8 Weighted 53-4 Into Centerline 15/16 ino 1-3/4 1/2 x 1-5/8 x 7/16 Weighted 51-1 Low Centerline 5/8 ino 1-1/2, 1/2 x 1/2 x 5/16 Clamped 138-3 Low Centerline 1-1/8 ino 1-1/2 1/2 x 3/4 x 1/4 Clamped 139-2 Low Riser shrink 3/16 in. 1-1/2 1/2 x 1-1/8 x 5/16 Clamped 149-1 High Riser froze over 1-1/2 1/2 x 1-1/2 to 1/2 x 3/16 Clamped 149-2 High Riser froze over 1-1/2 1/2 x 1-1/2 to 1/2 x 3/16 Clamped 190-2 High Riser froze over 15/8 1/2 x 1-5/8 to 1/2 x 1/4 x5/64 Clamped 198-2 High Surface shrink 1/64 in. 1/4 gate, atmospheric pressure core did not work 2-1/4 1/2 x 1-7/8 x 1/2 Weighted 33-5 Low Riser shrink 2 ino Gated into casting Intermed iate

APPEWDIX (cont.) Riser Neck Heat and Size, Dimensions, Casting Carbon in. in. Restraint Number Equivalent Soundness 1/2 x 4 x 7-in.Casting (cont.) no riser Weighted 533-6 Low Surface shrink 1/16 in. Centerline 1-11/16 in, no riser Clamped 136-2 Low Surface shrink 1/16 in. Centerline 1-1/8 in, Preliminary work has also been done on top blind risers 3/8 x 4 x 7-in. Casting RJ -1 1-1/8 3/8 x 9/16 x 3/16 - 1/4 gate Clamped 139-1 Low Centerline 5/8 in. ll/8 3/8 x 3/8 x 5/16 Clamped 138-2 Low Centerline 1 in. 1-1/4 3/8 x i-1/4 x 3/8 Weighted 64-2 Int.* Centerline 2-1/8 in. no riser Weighted 62-6 nt. Surface shrink 1/32 in. Centerline 2-3/4 in. no riser Clamped 136-1 Low Surface shrink 1/32 in. Centerline 2-1/4 in. no riser Clamped 190-1 High Surface shrink 1/32 in. Centerline 2-1/2 in. l-1/4 3/8 x 1-1/4 to 7/16 x 3/16 Clamped 198-1 High Wavy surface shrink x 1/16 5 3/16 gate lA x 4 x 7-in. Casting no riser Clamped 138-1 Low Centerline 1-3/4 in, Intermediate

UNIVERSITY OF MICHIGAN 3 9015 02826 3153