Method of improving the removal of investment casting shells

Abstract

The removal of an investment casting shell surrounding a metallic part is improved by adding a salt of alkali or alkaline earth metal to at least one of the layers of the shell.

Description

FIELD OF THE INVENTION

[0001] This invention relates generally to investment casting and, more particularly, to a method of improving the removal of investment casting shells.

BACKGROUND OF THE INVENTION

[0002] Investment casting, which has also been called lost wax, lost pattern and precision casting, is used to produce high quality metal articles that meet relatively close dimensional tolerances. Typically, an investment casting is made by first constructing a thin-walled ceramic mold, known as an investment casting shell, into which a molten metal can be introduced.

[0003] Shells are usually constructed by first making a facsimile or pattern from a meltable substrate of the metal object to be made by investment casting. Suitable meltable substrates may include, for example, wax, polystyrene or plastic.

[0004] Next, a ceramic shell is formed around the pattern. This may be accomplished by dipping the pattern into a slurry containing a mixture of liquid refractory binders such as colloidal silica or ethyl silicate, plus a refractory powder such as quartz, fused silica, zircon, alumina or aluminosilicate and then sieving dry refractory grains onto the freshly dipped pattern. The most commonly used dry refractory grains include quartz, fused silica, zircon, alumina and aluminosilicate.

[0005] The steps of dipping the pattern into a refractory slurry and then sieving onto the freshly dipped pattern dry refractory grains may be repeated until the desired thickness of the shell is obtained. However, it is preferable if each coat of slurry and refractory grains is air-dried before subsequent coats are applied.

[0006] The shells are built up to a thickness in the range of about 1/8 to about 1/2 of an inch (from about 0.31 to about 1.27 cm). After the final dipping and sieving, the shell is thoroughly air-dried. The shells made by this procedure have been called "stuccoed" shells because of the texture of the shell's surface.

[0007] The shell is then heated to at least the melting point of the meltable substrate. In this step, the pattern is melted away leaving only the shell and any residual meltable substrate. The shell is then heated to a temperature high enough to vaporize any residual meltable substrate from the shell. Usually before the shell has cooled from this high temperature heating, the shell is filled with molten metal. Various methods have been used to introduce molten metal into shells including gravity, pressure, vacuum and centrifugal methods. When the molten metal in the casting mold has solidified and cooled sufficiently, the casting may be removed from the shell.

[0008] Investment casting molds must withstand significant mechanical and drying stresses during their manufacture. Ceramic shells are designed having high green (air dried) strength to prevent damage during the shell building process. Once the desired mold thickness is achieved, it is dewaxed and preheated to approximately 1800.degree. F. At this point, it is removed from the high temperature furnace and immediately filled with liquid (molten) metal. If the mold deforms while the metal is solidifying (or in a plastic state), the casting dimensions will likely be out of specification. To prevent high temperature deformation, molds are designed to have substantial hot strength. Once the casting is solidified and cooled, low fired strength is desired to facilitate the knock out or removal of the ceramic mold from the metal casting.

[0009] Most investment casting molds contain significant quantities of silica. The silica usually starts as an amorphous (vitreous) material. Fused silicas and aluminosilicates are the most common mold materials. When exposed to temperatures above approximately 1800.degree. F., amorphous silica devitrifies (crystallizes) forming beta cristobalite. Cristobalite has low (alpha) and high (beta) temperature forms. The beta form has a specific gravity very close to that of amorphous silica so mold dimensions remain constant and stresses associated with the phase transformation are minimal. Upon cooling, beta cristobalite transforms to the alpha form. This phase transformation is accompanied by an approximate 4% volume change that creates numerous cracks in the shell, thereby facilitating mold removal. Cristobalite reduces the fired strength of silica containing investment casting molds.

[0010] Although investment casting has been known and used for thousands of years, the investment casting market continues to grow as the demand for more intricate and complicated parts increase. Because of the great demand for high-quality, precision castings, there continuously remains a need to develop new ways to make investment casting shells more efficiently, cost-effective and defect-free. For instance, if shell strength was maintained to the point of metal solidification, followed by a reduction in strength as the shell cools, improvements in productivity could be realized through improved knock out (shell removal). This is particularly desirable for non-ferrous alloys, e.g. alloys of aluminum, copper and magnesium, because their melting and pouring temperatures are insufficient to promote cristobalite formation and easy knock out.

[0011] The knock out is especially difficult when the part presents a blind hole or a small cavity in which the ceramic is under compression. The compression occurs during the cooling of the metal parts, which in general have a higher coefficient of thermal expansion (CTE) than the ceramic shells. This effect is especially accentuated in non-ferrous castings because of the high CTE of this metal (>18 10.sup.-6 m/m).

[0012] Non-ferrous castings produced by investment casters are rather fragile, so they are cleaned by water or sand blasting, compared with the aggressive shot blast and vibratory cleaning for steel and high temperature alloy castings. Residual ceramic on steel castings is dissolved away using concentrated acids and bases or molten salt baths. Chemical incompatibility excludes their use on aluminum and magnesium. If a binder was developed having low fired strength and associated easy knock out properties upon exposure to temperatures at or below 1800.degree. F., aluminum casting cleanup could be greatly improved.

[0013] Accordingly, it would be desirable to provide an improved method of removing an investment casting shell surrounding a metallic part.

SUMMARY OF THE INVENTION

[0014] The method of the invention calls for adding a salt of alkali or alkaline earth metal to at least one of the layers of an investment casting shell. The addition of a salt of alkali or alkaline earth metal effectively improves the removal of the investment casting shell surrounding a metallic part.

DETAILED DESCRIPTION OF THE INVENTION

[0015] The present invention is directed to a method of improving the removal of an investment casting shell surrounding a metallic part. In accordance with the invention, a salt of alkali or alkaline earth metal is added to at least one layer of the investment casting shell.

[0016] The salt of alkali or alkaline earth metals which may be used in the practice of the invention include calcium carbonate, calcium sulfate, calcium magnesium carbonate, magnesium carbonate, magnesium sulfate, strontium carbonate and mixtures thereof The preferred salt of alkali or alkaline earth metal for use in improving the removal of an investment casting shell from a metallic part is calcium carbonate.

[0017] The salt of alkali or alkaline earth metal can be added to at least one of the layers of the investment casting shell by any conventional method generally known to those skilled in the art. In a preferred embodiment, the salt of alkali or alkaline earth metal is added to at least one layer of the refractory stucco. However, in the practice of the present invention, the salt of alkali or alkaline earth metal may alternatively be added to at least one layer of the refractory slurry or to at least one layer of both the refractory slurry and the refractory stucco.

[0018] The salt of alkali or alkaline earth metal is used at a concentration that will effectively improve the removal of an investment casting shell surrounding a metallic part. It is preferred that the amount of salt of alkali or alkaline earth metal which is added to at least one layer of the shell be in the range of about 1 to about 30% by weight of the shell. More preferably, the amount of salt of alkali or alkaline earth metal is from about 5 to about 25%, with about 8 to about 20% being most preferred.

[0019] The present inventor has discovered that adding a salt of alkali or alkaline earth metal to at least one layer of an investment casting shell effectively improves the removal of the shell surrounding the metallic part.

EXAMPLES

[0020] The following examples are intended to be illustrative of the present invention and to teach one of ordinary skill how to make and use the invention. These examples are not intended to limit the invention or its protection in any way.

Example 1

CaCO.sub.3 Mixed in Refractory Slurry

[0021] Slurries were prepared using the following formulas: TABLE-US-00001 TABLE 1 Slurry Ingredients Concentrations (ratios) Colloidal silica.sup.1 1920 g Deionized water 386 g Latrix .RTM. 6305 polymer.sup.2 138 g Nalcast .RTM. P1 (-200 mesh) fused silica.sup.3 1200 g Nalcast .RTM. P2 (-120 mesh) fused silica.sup.4 3600 g Nalco .RTM. 8815 anionic wetting agent.sup.5 1.0 g Dow Corning .RTM. Y-30 antifoam.sup.6 4.0 g CaCO.sub.3 fine Marblemite .RTM..sup.7 0.0 g (0%), 288 g (6%), 576 (12%), 864 g (18%), 1200 g (25%) CaCO.sub.3 coarse 40-200.sup.8 480 g (10%), 960 (20%), 1680 g (30%) .sup.1Nalcoag .RTM. 1130 (8 nanometer, sodium stabilized) (available from Ondeo Nalco Company) .sup.2Styrene butadiene latex at 10% based on diluted colloidal silica (available from Ondeo Nalco Company) .sup.3Available from Ondeo Nalco Company .sup.4Available from Ondeo Nalco Company .sup.570% sodium dioctyl sulfosuccinate (available from Ondeo Nalco Company) .sup.630% silicone emulsion (available from Dow Corning Corporation of Midland, Michigan) .sup.7Cultured marble calcium carbonate (available from Imerys Corporation of Roswell, Georgia) .sup.8Screened grade calcium carbonate (available from Imerys Corporation of Roswell, Georgia)

[0022] After seventy-two hours of mixing, the viscosities of the slurries were measured and adjusted using a number four Zahn cup. The viscosities ranged from 14-18 seconds. Minor binder additions (colloidal silica+water+polymer) were made to obtain the desired rheology. Once adjusted, the slurries were ready for dipping.

[0023] Wax patterns were cleaned and etched using Nalco.RTM. 6270 pattern cleaner (available from Ondeo Nalco Company) followed by a water rinse. Wax bars were dipped into each slurry followed by Nalcast.RTM. S1 and S2 fused silica stucco (available from Ondeo Nalco Company) (applied by the rainfall method). Dry times started at 1.5 hours and progressed up to 3.5 hours as coats were added. The final shells had two coats with Nalcast.RTM. S1 stucco, three coats with Nalcast.RTM. S2 stucco plus one seal coat (no stucco). All coats were dried at 73-75.degree. F., 35-45% relative humidity and air flows of 200-300 feet per minute. After a twenty-four hour final dry, the shells were placed into a desiccator for an additional twenty-four hours prior to testing.

[0024] Several shell properties were evaluated using modulus of rupture (MoR) bars prepared from the experimental slurries. The bars were broken with a three point bending fixture on an ATS universal test machine (available from Applied Test Systems, Inc. of Butler, Pa.). The analog output (voltage) was fed into a personal computer containing an analog-to-digital conversion board and data acquisition software. The data was stored as a load versus time, or load versus displacement plot. Calculations and analyses were performed using data acquisition software or spreadsheet programs. The following physical properties were determined for the MoR specimens:

Fracture Load

[0025] The fracture load is the maximum load that the test specimen is capable of supporting. The higher the load, the stronger the test specimen. It is affected by the shell thickness, slurry and shell composition. This property is important for predicting shell cracking and related casting defects. The fracture load is measured and recorded for test specimens in the green (air dried), fired (held at 1800.degree. F. for one hour and cooled to room temperature) and hot (held at 1800.degree. F. for one hour and broken at temperature) condition. Results are normalized and expressed as an Adjusted Fracture Load (AFL). The AFL is simply the fracture load divided by the specimen width for a two inch test span.

Shell Thickness

[0026] Shell thickness is influenced by slurry and shell composition, combined with the shell building process. Thickness fluctuations are indicative of process instability. Non-uniform shell thickness creates stresses within the shell during drying, dewaxing, preheating and pouring. Severe cases lead to mold failure. The mold surrounds and insulates the cooling metal. Changes in thickness can affect casting microstructure, shrinkage, fill and solidification rates.

Modulus of Rupture

[0027] A flat ceramic plate is prepared using a rectangular wax bar as the pattern. Typical dimensions are 1.times.8.times.1/4 inches. The bar is invested using the desired shell system. After drying, the edges are removed with a belt sander. The two remaining plates are separated from the wax, yielding two test specimens. The specimens are broken using a three point loading apparatus on an ATS universal test machine. MoRs are calculated for bars in the green, fired and hot conditions. MoR = 3 .times. PL 2 .times. bh 2 ##EQU1## where P=Fracture load in pounds [0028] L=Specimen length in inches (distance between supports) [0029] b=Specimen width at point of failure in inches [0030] h=Specimen thickness at point of failure in inches

[0031] The MoR is a fracture stress. It is influenced by fracture load and specimen dimensions. Shell thickness is of particular importance since the stress is inversely proportional to this value squared. The uneven nature of the shell surface makes this dimension difficult to accurately measure, resulting in large standard deviations. This deficiency is overcome by breaking and measuring a sufficient number of test specimens.

Bending or Deflection

[0032] The test specimen bends as the load is applied. The maximum deflection is recorded as the specimen breaks. Bending increases with flexibility and polymer concentration. A flexible shell is capable of withstanding the expansion and contraction of a wax pattern during the shell building process. Bending is measured for bars in the green condition.

Fracture Index

[0033] The fracture index is a measure of the work or energy required to break a shell in the green condition. It is indicative of shell "toughness", i.e., the higher the index, the tougher the material. For example, a polypropylene bottle is "tougher" than a glass bottle and therefore has a higher fracture index. The index is an indicator of crack resistance. High index shells require more energy to break them than low index systems.

[0034] The fracture index is influenced by slurry and shell composition. Polymer additives increase the index. Soft polymers produce higher index shells than stiff ones. The index is proportional to shell flexibility. A shell that is capable of yielding absorbs more energy than a rigid, brittle one.

[0035] The fracture index is determined by integrating the area beneath the load/displacement curve for a MoR test specimen. The index measures (force).times.(distance) when monitoring displacement or (force).times.(time) when monitoring load time. To convert from (force).times.(time) to (force).times.(distance), the loading rate is used. Test results are normalized by simply dividing the index value by the specimen width for a two inch test span.

Modulus of Elasticity

[0036] The modulus of elasticity, also known as Young modulus, MoE or elastic modulus is a proportionality constant obtained from the stress-strain curve. It can be calculated from MoR (bending) analysis. The modulus of elasticity is a measure of stiffness of the material. A stiff, rigid material will display a strain/stress curve with a steep slope and high MoE. A soft, flexible material will exhibit a strain/stress curve with a flatter slope and low MoE. The modulus of elasticity is independent of sample dimension. It provides an accurate mean for comparing different shell systems.

[0037] As shown below in Table 2, the fired strength decreased with the amount of fine or coarse CaCO.sub.3 in the shell. The best system was achieved with the addition of 10-20% coarse CaCO.sub.3, which showed approximately a 30% decrease in fired MoR with a minimal effect on the green properties of the shell. TABLE-US-00002 TABLE 2 Summary green data MoR % CaCO.sub.3 (psi) AFL (lbs) MoE (kpsi) F Index Bending (mils) Fine 0% 556.73 9.02 381 26.43 4.77 6% 448.06 8.20 231 28.72 5.46 12% 406.47 7.96 228 28.11 5.03 18% 353.20 7.51 185 26.25 5.29 30% 261.82 6.26 115 23.59 5.75 Coarse 10% 588.83 8.29 346 31.26 5.44 20% 482.67 9.00 253 31.57 5.37 30% 385.94 9.46 186 34.70 5.33 Summary fired data MoR (psi) AFL (lbs) % CaCO.sub.3 Fine 0% 1016.82 15.33 6% 804.23 14.82 12% 651.27 12.59 18% 504.93 11.54 25% 397.56 9.05 % CaCO.sub.3 Coarse 10% 752.87 13.91 20% 738.12 15.11 30% 530.53 11.87

Example 2

CaCO.sub.3 Added as Discrete Dry Refractory (Stucco) Coats

[0038] Slurries were prepared using the following formulas: TABLE-US-00003 TABLE 3 Slurry Ingredients Concentrations (ratios) Colloidal silica 1920 g Deionized water 401 g TX-11280 polymer 138 g Nalcast .RTM. P1 (-200 mesh) fused silica 1200 g Nalcast .RTM. P2 (-120 mesh) fused silica 3600 g Nalco .RTM. 8815 anionic wetting agent 2.2 g Dow Corning .RTM. Y-30 antifoam 4.0 g Alternative layers of CaCO.sub.3 and SiO.sub.2 stucco were used during the shell preparation sequence as follows: Stucco Stucco Stucco Coat Coat Coat Stucco Coat Stucco Coat Formulation # #1 #2 #3 #4 #5 1 S1 S1 S2 S2 S2 2 S1 S1 CaCO.sub.3.sup.9 S2 S2 3 S1 S1 CaCO.sub.3 50/50 S2 + S2 CaCO.sub.3 4 S1 S1 CaCO.sub.3 CaCO.sub.3 S2 5 S1 S1 CaCO.sub.3 CaCO.sub.3 CaCO.sub.3 The shell test methods were also the same. .sup.9Screened grade calcium carbonate (available from Imerys Corporation of Roswell, Georgia)

[0039] As shown below in Table 4, the addition of the CaCO.sub.3 in the stucco layer decreased fired shell strength without compromising green strength. The best results were obtained with 1 CaCO.sub.3 stucco layer showing a 35% fired strength decrease with a minimal effect on hot strength. The green, hot and fired MoR results for shells with and without CaCO.sub.3 stucco coat were as follows: TABLE-US-00004 TABLE 4 Summary green data MoR Bending Formulation # (psi) AFL (lbs) MoE (kpsi) F Index (mils) 1 434.51 9.38 175 39.17 6.42 2 470.52 7.11 238 28.64 6.30 3 504.09 5.95 294 23.10 6.53 4 473.57 5.30 284 22.07 6.34 5 455.91 4.40 292 18.92 6.18 Summary hot and fired data Hot Fired Formulation # MoR (psi) MoR (psi) Hot AFL (lbs) Fired AFL (lbs) 1 1204.13 707.82 19.80 14.36 2 1023.13 465.12 13.14 7.40 3 619.81 377.00 9.05 4.96 4 700.56 254.34 7.57 3.09 5 518.13 131.40 4.65 1.38

Example 3

CaCO.sub.3 Mixed with Dry Refractory Stucco

[0040] Slurries were prepared using the following formulas: TABLE-US-00005 TABLE 5 Slurry Ingredients Concentrations (ratios) Colloidal silica 1920 g Deionized water 401 g TX-11280 polymer 138 g Nalcast .RTM. P1 (-200 mesh) fused silica 1200 g Nalcast .RTM. P2 (-120 mesh) fused silica 3600 g Nalco .RTM. 8815 anionic wetting agent 2.2 g Dow Corning .RTM. Y-30 antifoam 4.0 g Blends of SiO.sub.2 and CaCO.sub.3 were used as stucco coats during the shell preparation sequence as follows: Stucco Stucco Stucco Formu- Coat Coat Coat Stucco Coat Stucco Coat lation # #1 #2 #3 #4 #5 1 S1 S1 S2 S2 S2 2 S1 S1 S2 + 5% S2 + 5% CaCO.sub.3 S2 + 5% CaCO.sub.3.sup.10 CaCO.sub.3 3 S1 S1 S2 + 10% S2 + 10% CaCO.sub.3 S2 + 10% CaCO.sub.3 CaCO.sub.3 4 S1 S1 S2 + 20% S2 + 20% CaCO.sub.3 S2 + 20% CaCO.sub.3 CaCO.sub.3 5 S1 S1 S2 + 30% S2 + 30% CaCO.sub.3 S2 + 30% CaCO.sub.3 CaCO.sub.3 .sup.10Screened grade calcium carbonate (available from Imerys Corporation of Roswell, Georgia)

[0041] An erosion test was added to fully characterize the benefit of the CaCO.sub.3 addition on the knock out properties of the shells. The new technique simply consisted of a simulation of the sand blasting process used in the industry. Shell coupons of dimensions: L, W=0.75'' to 1.5'' and 0.4'' thick max were placed in a fixture located on one of the walls of the blasting cabinet. A spring blade held the coupon against a stainless steel mask with a 0.5 x 0.5'' window. This mask ensured that identical surface area was exposed to the blast from one sample to the next and a homogenous blasting media flowed across the surface. The thickness of the coupon was measured prior to the test. The coupon was then exposed to the blasting media until it was perforated. The time to perforation was measured and an erosion speed was calculated for each sample. The test was repeated with a representative number of coupons to allow an accurate determination of the erosion speed.

[0042] As shown below in Table 6,,the addition of the CaCO.sub.3 blended with SiO.sub.2 in the stucco layer decreased fired shell strength without compromising green strength. The best results were obtained with 10-20% CaCO.sub.3 addition to the stucco layer showing a 20 to 40% fired strength decrease with a minimal effect on hot strength. The erosion speed data show that a five fold increase in erosion speed was achieved after 24 and 48 hours cooling shells containing 8 and 10% CaCO.sub.3. The green, hot, fired MOR and erosion test results for shells with and without CaCO.sub.3 stucco coat were as follows: TABLE-US-00006 TABLE 6 Summary Green Data F % CaCO.sub.3 MoR (psi) AFL (lbs) MoE (kpsi) Index Bending (mils) 0 564.40 9.44 268 38.38 6.32 5 566.48 9.11 306 33.16 5.78 10 580.84 8.64 305 32.67 6.02 20 577.19 8.64 305 32.67 6.02 30 559.01 6.71 355 24.90 5.62 Summary hot and fired data Fired % CaCO.sub.3 Hot MoR (psi) MoR (psi) Hot AFL (lbs) Fired AFL (lbs) 0 1326.26 932.15 19.28 16.22 5 1127.35 809.54 15.44 11.94 10 923.31 729.07 15.44 11.14 20 723.13 558.85 9.75 6.54 30 793.39 496.62 7.81 5.59 Summary erosion data 2 Hr cooling 24 Hr cooling 48 Hr cooling % CaCO.sub.3 (mils/sec) (mils/sec) (mils/sec) 0 4.89 4.40 4.79 2 4.51 5.13 6.38 4 4.37 10.74 9.64 6 4.46 10.71 14.71 8 4.34 14.70 25.10 10 4.23 10.58 24.28

[0043] While the present invention is described above in connection with preferred or illustrative embodiments, these embodiments are not intended to be exhaustive or limiting of the invention. Rather, the invention is intended to cover all alternatives, modifications and equivalents included within its spirit and scope, as defined by the appended claims.

Claims

1-3. (canceled)

4. A method of forming an investment casting shell in a shell preparation sequence and improving the removal of the investment casting shell surrounding a metallic part, wherein the shell preparation sequence comprises deposition of alternating layers of a refractory slurry and a dry refractory stucco onto a pattern, wherein an effective amount of the salt of an alkali or alkaline earth metal is added to at least one layer of the dry refractory stucco during the shell preparation sequence, and wherein the refractory slurry layers are free of added alkali or alkaline earth metal salt during the shell preparation sequence.

5-11. (canceled)

12 The method of claim 4, wherein the salt of the alkali or alkaline earth metal is selected from the group consisting of: calcium carbonate, calcium sulfate, calcium magnesium carbonate, magnesium carbonate, magnesium sulfate, strontium carbonate, and mixtures thereof.

13. The method of claim 4, including about 1 percent to about 30 percent of the salt of the alkali or alkaline earth metal, based on weight of the shell.