You've got the job, now make it!: four approaches to rigging a ductile iron casting.
In the heat of the battle for an order, foundries promise the best possible technology to meet the customer's requirements - at a competitive price. Once the job is in the foundry's hands, however, it must temper its enthusiasm with the realities of producing the part at the quoted price - with its existing equipment, technology and production schedules.
Four members of the AFS Iron Gating and Risering Committee (5-M) faced these realities in a 1996 Casting Congress panel presentation that produced four distinct approaches to rigging a hypothetical ductile iron hub casting [ILLUSTRATION FOR FIGURE 1 OMITTED].
The casting was designed for production on commonly available equipment, with or without consumable products to enhance yield. A standard Grade 65-45-12 ductile iron was chosen and a 1000-plus production run was specified.
A goal was set to maximize casting yield and produce sound, competitively priced castings. The modulus risering method was generally used as a starting point to reach this target.
Test pouring or computer simulation was used to evaluate rigging effectiveness.
Equipment Available: A Herman horizontal (18 x 18 in.) and two Disamatics (2013 and 2070) were considered, with attention given to pour weight, details/dimensions, and operating and material costs.
The mold machine should operate near maximum productivity and produce consistent demand from the melt department. Scheduling with other production would be needed to assure adequate molten iron without costly melt delays, nor jeopardizing other jobs.
Pour weight is also critical because automatic pouring requires each mold to be poured in a specific amount of time. Based only on pour weight, 2-4 hubs would fit on the Herman, 1-2 on the 2103 and 3-5 on the 2070.
Both Disamatics can produce the center hole in green sand, while the Herman would require a core. Considering casting size and spacing requirements for a hard mold, the options became: Herman (two-on), 2013 (two-on) and 2070 (three-on).
Core cost makes the Herman the least efficient. Only three castings will fit on the 2070, resulting in low pour weight and poor mold utilization. Due to offering the best efficiency, the 2013 (two-on) was chosen.
Modulus Calculation: The AFS Weight/Order program was used. The casting was divided as shown in Fig. 2. Compensation was made for the noncooling surfaces of each section and heat saturation in the corners. Also, the center hole was too small to contribute to the cooling.
Calculations showed #4 freezing off before #1, producing an isolated heavysection. Therefore, #4 was thickened with a feed pad. Thickness was calculated by assuming that #4 was an infinite plate and using the formula: Modulus = Area/Perimeter.
Riser Calculation: The size of the feeding system was also determined using modulus relationships. As a rule of thumb, the contact modulus should be 35% of the casting modulus and the riser modulus should be 120%. This assumes an accurate casting modulus; otherwise, the entire system may be over- or under-risered.
A rectangular riser was chosen to provide a flat knock-off surface. To simplify calculations, the width was made equal to the thickness and the length, 1.5 times the width. The formula, Modulus = Volume/Surface Area, was used to solve for the riser width.
To put the contact along the riser thermal centerline for maximum performance, the thickness was measured from the top of the contact, rather than the plate line. This added about 0.50 in. thickness to the swing riser.
Gating Layout: The gating system was designed for bottom-filling to reduce turbulence as the cavity fills and to provide "hot" risers for effective piping [ILLUSTRATION FOR FIGURE 3 OMITTED]. The gating was sized to meet automatic pour-rate objectives (4-6 sec).
A pressurized system (choke at ingates) was chosen. This permits small ingates for easier knock-off and grinding. While this suggests a potential for flow-type defects, the ingates are entering a heavy section and the effects of the high pressure are minimized. Also, by the time pressure builds up enough in the down runner to force iron into the small choke plates, the cavities are nearly full, minimizing the chance of flow defects when iron enters a casting from two directions.
Gating Calculations: AFS-recommended gating ratios weren't used. Grede's experience is that these ratios produce large runners that reduce yield without improving quality.
To achieve the desired pour rate, the gating was sized to increase the flow by 5% for each junction, moving away from the choke (at the ingates). A thin section from the runners to the risers allows metal to also enter the casting from above, compensating for the reduction in cavity-fill rate experienced with bottom-filling as back-pressure builds. It also assures that the risers fill with hot iron. The thin section freezes off fast to isolate the risers from the gating. The risers include cavity gas vents.
Another thin, wide cross-section below the pour cup dissipates the energy of the iron stream entering the cup while allowing adequate flow. The down runner was also tapered to prevent aspiration.
Gating size calculations were based on the law of continuity (Flow Rate = Area x Velocity). The velocity in a given part of the gating is determined by friction and the head potential (measured from the top of the plate down to the gating section being sized). Using a friction factor (determined through experimentation), the calculation becomes:
Flow Rate = Friction Factor x Area x Distance from Top of Plate.
The known flow rate (estimated pour weight/desired pour time) is used in the above formula to solve for each cross-sectional area.
Testing/Test Pouring: Two samples were poured, one under normal foundry conditions and the other, under adverse conditions (high magnesium levels, soft molding, excessively hot iron).
Both castings were clean and free of surface defects. The casting made under normal conditions was sound. The casting made under the adverse conditions contained a small area of Level 1 shrink in the hub, 180 [degrees] from the riser.
Results/Conclusion/Yield: With a total pour weight of 62 lb and a casting weight of 33.4 lb, the yield was 53.8%, about standard for the foundry. Increasing the lap size would accommodate potential process variations, but must be balanced against added cost.
Lufkin Industries, Inc.
Equipment Available: The majority of Lufkin's castings range from 300 - 10,000 lb, limiting the available equipment for this small hub to impact green sand molding or flaskless nobake molding. Since the small bore couldn't be consistently produced on the impact line without a core and was also a poor flask fit, flaskless nobake molding was chosen.
Modulus Calculation: While this casting lends itself to simpler two-dimensional methods, solid computer modeling and the mass properties feature of the software were selected. The moduli calculated by the software are shown in Fig. 4.
A correction factor for reduced cooling effect in the small hub bore was calculated using Karsay's formula. This increased the modulus for #1.
While computer simulation disclosed that the sand adjacent to the web would superheat, it was suspected that the web would cool fast enough to isolate the rim from the bore area, preventing feeding between the two sections.
Using the simulation information and modulus calculations, it was decided that, for this foundry and molding method, an exothermic mini riser placed just above the web area would be the best feeding choice. Placing exothermic material here eliminates any need of padding to connect the two sections.
Riser Calculation: In addition to selecting a riser with a modulus equal to or greater than the casting modulus, the riser was checked to see that it contained at least 3% of the casting weight (foundry's procedure for ductile iron).
A thin sand core between the riser and the casting prevents the exothermic material from burning into the casting surface and simplifies knock-off.
Gating Layout: The flaskless nobake system eliminates layout restrictions of other systems. This layout includes a square pouring basin for slag floatation. A thin, tapered, rectangular sprue eliminates vortexing and allows the gating to quickly charge. A sprue well cushions the vertical drop and reduces turbulence [ILLUSTRATION FOR FIGURE 5 OMITTED].
A horizontal silica-cloth filter was used, with a choke just ahead of the filter to minimize turbulence (reoxidation) in the cleaned iron. The choke keeps the runner system full while the cavity fills. The runner above the filter was kept thin to assure that the filter is wetted on both sides.
Gating Calculations: Initial gating calculations were made to select a filter and to size gating components, using Novacast software. The final gating combines computer calculations, manufacturers' recommendations and adjustments based on rigging experience.
Flow/Temperature Simulation/Analysis: Computer flow analysis was used to design a hydraulic "lock" just ahead of the ingate, assuring that the iron would enter the rim portion of the casting first, rather than squirt prematurely into the hub area. (This is more important in larger castings where coldshut defects are likely.)
Powercast computer temperature simulation showed that at the end of the fill, the iron was hottest at the ingate and riser as intended. No shrinkage was predicted.
Testing/Test Pouring: Actual castings were made, one with and one without the riser. Sectioning disclosed a shrink in the calculated heavy section of the riserless casting. None was found in the risered casting.
Results/Conclusion/Yield: Final results showed a casting weight of 17 lb and a pour weight of 25.2 lb. The casting, when poured using a loose laminated object manufacturing (LOM) pattern and the designed gating system, was sound and the yield was 67%.
Equipment Available: As a foundry consumable supplier (not a foundry), no limitations on equipment were set. Thus, the effect of using an extremely high production process and vertical casting on a Disamatic molding machine was examined.
Modulus Calculation: The full range of design technology was available to develop a riser and gating system - from simple use of known rules and a hand calculator to computerized thermal and fluid-flow calculations.
For this exercise, hand calculation was chosen, along with a roles-based computer riser program, to demonstrate that castings produced on high-production equipment can be readily rigged using traditional design techniques.
To evaluate the rigging design, a computer-based finite element program was used.
While pure modulus calculations using Chvorinov's Rule can be directly applied to ductile iron, this may lead to larger risers and unnecessary padding because such calculations don't consider benefits of graphite expansion during solidification.
The hub was split into sections and the moduli were calculated using standard volume and surface area equations [ILLUSTRATION FOR FIGURE 6 OMITTED].
Widely accepted practice for feeding well-controlled ductile iron through one section to another section suggests that, for sound sections, the modulus of the section providing the metal should be at least 85% of the modulus of the section receiving the metal.
The initial calculation for the central neck (#6) was adjusted for loss of cooling effect through the small central cylinder of sand. A pad was added to the web in #5 and the modulus recalculated when it became clear that the original web would freeze off before #6 obtained its feed metal.
Riser Calculation: FeederCalc for Iron was used to calculate the riser. This rules-based computer program incorporates chemistry, pouring temperature, casting weight, inscribed sphere, mold hardness, riser location and riser type.
The program determined that a 7/10 size Kalpur Direct Pouring Unit would feed two castings and calculated the area of the riser contacts. The direct pouring unit is an insulating riser sleeve with in integral foam filter. It eliminates a separate pouring cup and an extensive runner system.
Gating Layout: The gating consists of two direct pouring units, two riser basins, four riser contacts (calculated by the riser program) as ingates and a small connection runner to fill four casting cavities [ILLUSTRATION FOR FIGURE 7 OMITTED]. The riser basins provide a channel for metal feeding and a smooth transition of the filtered iron from the riser into the casting.
Gating Calculations: The small runner between the two direct pour units was calculated to balance the system. The runner entrance at the bottom of the top riser basin has the same area as the two casting contacts.
The runner exit at the top of the bottom riser was calculated using the law of continuity and the height of the mold cavity from the top of the mold. This compensates for differences in head height between the two levels of castings and allows smooth, even filling of both levels.
Flow/Temperature Simulation/Analysis: A finite element program, MagmaSoft, was used to examine the results of this design. Flow simulations disclosed even, consistent filling of the four cavities. The velocity at the ingate was minimal. While some turbulence was observed inside the cavities, it wasn't considered severe enough to cause oxidation defects.
Solidification: At complete filling, the highest temperatures were found in the riser, riser basin and neck. Simulations at intermediate solidification levels showed that the neck was open long enough to feed the liquid contraction of the iron, providing the necessary feed metal as the casting proceeded through solidification and expansion of the graphite.
At 90% solidification, the casting was sound and iron contraction was shown at the top of the top riser and in the runner above the lower riser. Ultimately, all four castings were sound and presumed clean because of low turbulence and use of filters.
Results/Conclusion/Yield: The total casting weight (four castings) is 66.8 lb and the total pour weight is 82.19lb - a yield of 81.3%. The riser contact/ingate should break off easily, with little to no grinding. Remelt losses will be small and the high yield should assure profitability.
[TABULAR DATA FOR TABLE 1 OMITTED]
Brillion Iron Works
Equipment Available: A Herman and a Brillion-built Moldblower were considered. Four-on would be a good fit on the Herman, however, a core would be needed. Two-on would be satisfactory for the Moldblower. The Moldblower was chosen because no core was required and it can produce this casting at a high molding rate.
Modulus Calculation: The AFS Weight/Order program, which incorporates the Area/Perimeter calculations, was used [ILLUSTRATION FOR FIGURE 8 OMITTED]. The noncooling faces and the center perimeter (bold lines) were subtracted from the section perimeters. The initial calculations indicated that the center section wouldn't feed from the outside and a 0.25 in. feed pad was added to the right-hand section.
Riser Calculation: The geometric method was used. The feed metal volume was determined as a percentage of the total casting cavity weight, and the riser height and pressure section height were based on the mold size with a safety factor added. Thermal walls around the feed metal and below the neck were built to equal two times the casting modulus. A 90 [degrees] pie wedge was removed from the riser top to initiate piping.
Neck-size calculations are normally made using 40-60% of the casting modulus, depending on the length of the neck. With this short neck, 40% was used. Various neck proportions were considered; the final choice was based on despruing area.
Gating Layout: Figure 9 shows the gating layout. The desired pour time, pouring ratios and available space determined the gating dimensions. Features include:
* a standard pour cup flattened on four sides to prevent swirling;
* an inoculation pellet print in the downsprue base;
* a cope runner before the filter, with a 45 [degrees] enlargement for slag flotation;
* a filter to eliminate extensive runners
for slag flotation, improving yield;
* a tapered drag runner to smooth flow, improve yield and maintain a full system;
* a thin choke for flow control and knock-off. Freezes off to isolate the spree from the riser and promote piping;
* a hot riser for progressive solidification;
* a single riser feeding two casting cavities for improved yield.
Pour Time/Choke Size: As a guideline for determining pour time, the square root of the combined weight of the castings and riser was taken and adjusted for a system charge time of 1 sec (5.69 sec adjusted pour time).
Dividing the combined weight by the adjusted pour time, a pour rate of 7.87 lb/sec was obtained. To pour 1 lb per sec on this equipment, 0.1 sq in. of area are required. Thus, 0.787 sq in of choke area will achieve the calculated pour time.
Gating Ratios and Areas: The gating ratios and areas were chosen to keep positive pressure on the downstream choke and allow for slag flotation to maximize filter life. A 10% enlargement of the runner past the choke accommodates friction and prevents oxidation defects from squirting a stream into the riser.
Testing/Test Pouring: The gating system was verified by molding a LOM prototype pattern on a short-run floor. Thermocouples placed in the mold checked thermal distribution as solidification proceeded.
Based on many samples poured in this foundry and plotted in terms of computed modulus vs. time-to-solidification in minutes, a "best-fit" curve has been developed to confirm calculated modulus values.
Results/Conclusion/Yield: Readings taken at all thermocouple locations confirmed a directional solidification pattern. The result was a sound, shrink-free casting - as predicted by the original calculations.
Based on a casting weight for two castings of 33.4 lb and a pour weight of 55.7 lb, a mold yield of 60% was obtained.
Table 1 summarizes the results, showing there are many ways to produce ductile iron castings that satisfy customers' requirements for quality and price, as well as foundries' requirements for high yield and profitability using current technology.
|Printer friendly Cite/link Email Feedback|
|Date:||Jun 1, 1997|
|Previous Article:||Students examine metalcasting opportunities.|
|Next Article:||GRAC puts aluminum filtration to the test and into practice.|