Evaluation of a Modified Cone Jolt Test on Green Sand Properties

Transcription

1 Paper pdf, Page 1 of 8 Evaluation of a Modified Cone Jolt Test on Green Sand Properties S. N. Ramrattan, J. Rodriguez and A. Choudhury Western Michigan University, Kalamazoo, MI Copyright 2010 American Foundry Society ABSTRACT Foundry engineers have long known that certain AFS standard green sand properties tests provide limited information. This limitation is especially apparent when strict casting dimensional control is required. Furthermore, since sand conditioning involving moisture and/or clay directly affects mold quality, a more sensitive sand test is needed. Too much clay in the system is expensive, and reduced bond usage can diminish green strength essential for mold handling. Similarly, lower moisture levels equate to reduced moisture-related defects; however, insufficient moisture can cause friable molds. This paper addresses the benefits of a modified cone jolt green sand test. For this study, green sand specimens from working foundries were tested at various target compactability levels. Their major green properties were obtained and their inherent variability reported. The purpose of this paper was to identify if a foundry can monitor changes in sand properties caused by small changes in clay and moisture levels when using a modified cone jolt test. The modified cone jolt test was able to better differentiate among various compactability levels when compared to other green property tests. INTRODUCTION The majority of casting defects in sand casting facilities are sand related, but many metalcasters fail to recognize that molding sand quality is as important as metal quality. 1 The green sand molding process brings with it some potential drawbacks for casting parts of all shapes, sizes, and metals. Measurement and control of green sand properties is critical to the green sand foundry s success. The typical properties that are monitored are green strength, compactability, moisture, and methylene blue clay percentage. Many foundries incorporate different bond formulations to produce the green properties they feel they need to produce acceptable castings. Casting quality is directly related to mold quality and mold quality is directly related to sand control. To produce world-class castings competitively and to meet the ever-increasing demands for higher quality and dimensional reproducibility, sand control is essential. 2, 3 A good understanding of the ins and outs of a process is needed to deal with day-to-day problems. Over the years, several noted papers have been published on green sand system control. However, the common theme for success in green sand control is sticking to the basics. 2, 3 To effectively control green sand, the basic variables that the foundry engineers must monitor include the addition of water, bond, new sand, and carbonaceous material along with the subsequent mull time. The basic sand tests commonly used to control green sand systems include moisture, compactability, specimen weight, green strength, permeability, AFS/25-micron clay, AFS grain fineness, methylene blue clay, volatiles, and loss on ignition (LOI). 4,5 Moisture and compactability are the key tests for controlling water additions. Green strength, methylene blue and AFS or 25-micron clay provide important information relating to control of the bond. Density, specimen weight, permeability, grain fineness, and again AFS/25-micron clay relate to amounts of new sand additions. LOI and volatiles are used for controlling the carbonaceous material. Optional tests that also can be useful include green deformation, dry compressive strength, splitting strength, friability, cone jolt toughness, wet tensile strength, and volatiles. Most foundries carry out green sand testing as an essential part of process/quality control. Foundries depend on the AFS sand tests to help manage their new sands, green sand systems, chemically bonded cores and molds, sand additives, and reclaimed sands. The testing procedures are defined by the AFS. 6 Sand composition can be determined through laboratory testing, which provides an important means of recording and monitoring changes in the system. Green sand testing tells how well a foundry is controlling its system. Many of the tests are not timely enough to allow real-time control of the green sand system so automated controls is placed on the sand system. Still certain laboratory tests are essential for monitoring the sand system to achieve control. 7 Some test data such as LOI tends to change slowly, while other test data - green compression strength and compactability - may change more abruptly. Sand conditioning involving moisture and/or clay directly affects mold quality thus a more sensitive sand test is needed. The purpose of this paper was to identify if a foundry can monitor changes in sand properties caused by small changes in clay and moisture levels when using a

2 Paper pdf, Page 2 of 8 modified cone jolt test. The conventional cone jolt toughness test measures the bulk brittleness of the sand, and is related to difficulty in pulling deep pockets in a pattern and broken molds. 4 Though the conventional cone jolt toughness test has been around for many years and provides some general information on sand brittleness, it has found very limited usage in foundries as a sand control test. The main drawback in the test has been the high test-to-test variability caused by inadequate displacement control to end the test cycle. The modified cone jolt test overcomes this limitation, and allows additional test variables to be incorporated in the procedure. The modified cone jolt tester incorporates the following features: Specimen-holding base driven by a solenoid Ability to specify various drop height setting Ability to specify jolting frequency Ability to collect and chart data in real time Option to automatically stop test after specified displacement in specimen or at specimen break The working arrangement of these features is schematically represented in Figure 1. Figure 2. Modified cone jolt tester and data acquisition system. OBJECTIVE The objective of this study is to determine the applicability of the modified cone jolt test to identify significant difference in dynamic green sand properties for different sand. More specifically the following questions are raised: Is a modified cone jolt test sensitive to changes in active clay for a green sand system? Does AFS green properties test discern either clean or dirty sand systems? Figure 1. Schematic of modified cone jolt tester. For this study, samples will be collected from two working foundries, one from a grey iron foundry and the other from an aluminum foundry. Standard green property testing as well as dynamic tests will be conducted. The toughness test will be conducted using the modified cone jolt tester and data acquisition system (Figure 2). The experimental testing of both mechanical and dynamic green sand properties will be performed in a controlled laboratory setting. In this study, dirty sand refers to green sand riddled through a inch mesh screen (AFS standard) and clean sand refers to green sand riddled through a inch mesh screen. The test samples will be collected from working green sand foundries. All green sand data will be generated in the Metal Casting Laboratory at Western Michigan University. The determination of green properties will be accomplished by conducting a number of physical, mechanical, and dynamic test measurements on the green sand samples. These tests include moisture content, mold hardness, green & dry compressive strength, methylene blue clay percentage, friability, and modified cone jolt test. METHODOLOGY The laboratory procedure consisted of two major steps. First, green sand was tempered to the desired compactability level and second, a series of AFS tests for

3 Paper pdf, Page 3 of 8 green sand properties was performed. A minimum of sixty specimens were tested to determine an average and standard deviation for each green sand property. All specimens were prepared and tested at Western Michigan University, Metal Casting Laboratory. Ambient conditions were controlled, temperature at 71.6±1.8 F (22±1C), and relative humidity at 50±2%. PREPARATION OF GREEN SAND TO DESIRED COMPACTABILITY Two working green sand samples were shipped to Western Michigan University. The as-received green sand (15 pounds 66.7 N) sample was riddled through a inch mesh. To represent sand contaminants such as core butts and based on weight of sand (B.O.S) 1% polyethylene pellets (0.13-inch dia. 3.3 mm) were added to the green sand. The polyethylene pellets are allowed through a inch mesh screen during riddling (dirty sand) and prevented by a inch mesh screen during riddling (clean sand). Figure 3 shows typical clean and dirty specimens. This mixture was mulled to the desired compactability, which was continuously monitored. The water additions were then raised or lowered accordingly on that batch of sand to produce the target compactability. The sand was not discharged until the compactability was on target. Thus, the green sand systems used in this study were tempered to a desired compactability, property tested (as mulled), and re-tempered to start a new run of the experimental cycle. Figure 3. Cone jolt specimens, clean (left) and dirty (right). Green Sand Materials from Foundry A No cores were used at the iron foundry so that the only entering sand was sub-angular lake sand. The properties and grain size distribution of the green sand systems used in this study are shown in Table 1. The bond recipe contained an approximate five-to-one ratio of western bentonite to southern bentonite where total clay added was BOS (methylene blue clay also shown in Table 1), and water added to produce the desired compactability. Standard seacoal levels of 25% were also included in the pre-blend as well as 0.5% soda ash. Table 1. Properties of the Green Sand Systems Green Sand System Iron Foundry Aluminum Foundry USA Sieve No. Sample A Sample B Pan Screens 4 4 AFS-GFN Sand Type Silica Olivine Shape Sub-Angular Angular Roundness/ Sphericity (Krumbein) 0.5/ /0.3 ph LOI micron Clay, % 12 9 Methylene Blue Clay, % Dead Clay, % Green Sand Materials from Foundry B Silica shell cores were used at an aluminum olivine green sand foundry so the green sand system consisted of fine angular olivine sand and coarse (55 GFN) sub-angular lake sand. Table 1 shows te properties and grain size distribution of the green sand systems used in this study. The bond recipe contained a one-to-one ratio of Western Bentonite to Southern Bentonite where total clay added was BOS (methylene blue clay also shown in Table 1), and water added to produce the desired compactability. No additives were included in the pre-blend. Equipment Sand Muller (25 pound N); Digital Balance (0.01g sensitivity), Sand Squeezer; Stripping Post; AFS Standard Specimen Tube; Compactability Scale. Procedure 1. Follow standardized AFS procedure for compactability testing to fill specimen tube and strike off. 2. A sand squeezer (Figure 4), set at 140 psi (965.2kPa), was use to produce specimens within the specimen tube because a three ram green sand rammer adds to the cone jolt specimen-to-specimen variability.

4 Paper pdf, Page 4 of 8 3. A semi-automated stripping post (Figure 4) was used to strip the specimen from the specimen tube and directly measure percent compactability. The stripping equipment helps to reduce specimen-to-specimen variability. 4. Prepare green sand to desired compactability levels 44%, 38% and 32%, for Sample A, and 50%, 40% and 30%, for Sample B. Figure 4. Lab equipment to produce compactability and cone jolt specimens: Sand squeezer (left) and Stripping Post (right). AFS GREEN PROPERTY TESTS Selected AFS sand tests were used to measure properties of the green sand systems used in this study. The key variable on the sand systems was compactability. Compactability is a measure of sand temper that is inversely related to bulk density. It is a mechanical property related to the percentage decrease in height that the sand will compact at the molding machine. With automated molding systems, compactability must be held constant. 7 Moisture and compactability are the key tests for controlling water additions. The moisture content test and compactability test were run first. The moisture content test is simply a quantitative measure of the amount of water in the green sand. The higher the compactability, the wetter the sand, and moisture was adjusted to control compactability. The compactability test was used to control the temper of the sand. Green strength, methylene blue and 25-micron clay provided information relating to the bond. The green compressive strength tests are typically used to control the strength characteristics of the sand. The data is commonly used to relate to molding and handling properties of the sand. The methylene blue clay test allows measurement of live clay in molding sand. Live clay is capable of acting as bonding material. Molding sand also contains dead clay, which has been destroyed by temperature and therefore can no longer rehydrate and contribute to bonding. The 25-micron clay is a measure of the sum of the live and dead clay. The difference between the 25- micron clay and the methylene blue clay is considered the dead clay. Density, specimen weight, permeability and grain fineness relate to new sand additions. Before proceeding with the other tests that require a standard test specimen, the specimen weight must be determined. Compacted density can be determined simultaneously. It is important to record the specimen weight because the weight provides useful information regarding changes in sand composition. If the specimen weight increases, this indicates that the sand s silica content has increased, since silica is the heaviest component of the sand. If it decreases, either the additives have increased or there is an increase in the amount of dead clay and ash accumulating in the sand. In this way, it can be used as a guide for determining the need for new sand additions. Gases are produced in a mold from the heat of the molten metal. The water in the mold produces steam and the carbonaceous materials in the sand produce other gases. There must be a provision to vent these gases from the mold as they are produced or else gas defects will result. Permeability tests provide an important relative measure of the sand s venting characteristics. Additionally, the Mold Quality Indicator (MQI) test, which is inversely related to permeability, was also studied. The MQI number is a measurement of the resulting back pressure developed from resistance of airflow through a mold or core. 6 An MQI unit is typically deployed somewhere along the molding line to perform real time measurements on the molds waiting to receive the molten metal. With modifications to the original rubber contact head, this instrument was utilized with a Western Michigan University specimen holder as a second means of measuring and confirming the venting characteristics of the green sand specimens. The grain fineness and distribution of the base sand is another important factor. The fineness of molding sand can change as it used in the foundry. Finer sand equates to lower permeability because the voids between the sand grains are smaller. Low permeability produces a smoother casting surface finish because the voids between the sand grains are smaller. Low permeability, however, increases the likelihood of problems with blows, pinholes and other gas-related defects. Low-permeability sands also can produce expansion defects if the permeability is low as a result of high packing density of the sand grains. MQI also measures the venting characteristics of green sand but is inversely related to permeability. The ph of the foundry sand was determined. The LOI test was used to indicate the amount of carbonaceous material in the green sand systems. Dry compressive strength tests, which are run similarly to green strength

5 Paper pdf, Page 5 of 8 but on specimens dried at F ( C), are used to indicate the sand s shakeout characteristics when the sand is dried by the molten metal s heat. If dry compressive strength is high, the molds will be stronger and difficult to shake out. Mold hardness was conducted to give an indication of the molds resistance to indentation and penetration. Compressive strength tests have conventionally been used to control the strength characteristics of molding sands. However, tensile properties of molding sands are actually more critical because they are a weaker characteristic. Most mold failures, such as in pattern stripping, are actually tensile failures. Splitting strength is related to the tensile property of the green sand. Splitting strength tests relate to the degree of mulling and this test was conducted. The tests that have proven to be indicative of sand brittleness are friability and cone jolt toughness tests. Both tests were run. The cone jolt toughness test measures the sand s bulk brittleness, and is related to difficulty in pulling deep pockets in a pattern. Friability measures surface brittleness and abrasion resistance of the sand. Molding sands can become very friable if there is too high an influx of core sand or new sand and bond. New bond requires several passes through the mixer before its properties are developed. In general, friability is inversely related to compactability; lower compactability equals higher friability. A small drop in compactability, or a brief air-drying period, will produce a large increase in friability. The cone jolt toughness test is directly related to compactability and measures the green sand system s ability to absorb energy. The modified cone jolt toughness test uses the same testing approach except for a few modifications. The mass of the cone can be adjusted to represent a specified force on the sand (0.68 kg was use for testing). The drop height of the cone can be varied (0.06 in -1.5 mm - was use for testing). The frequency of drops can be varied via a solenoid controlled by a pulse generator (100 jolts/min was use for testing). The following explanation provides the procedures to operate the modified cone jolt toughness tester (Figure 2). A computer and data acquisition system is used for controlling, monitoring and plotting graphs of jolts versus displacement of a specimen. A standard AFS cone jolt specimen is placed between the base and the cone. When the test is initiated, a solenoid is cycled to automatically pick up and drop the specimen, simultaneously a linear voltage displacement transducer (LVDT) engages and measures longitudinal displacement of the specimen. The data acquisition system automatically logs and plots the jolts versus displacement curves. The test stops automatically when the specimen splits or displaces 0.05 in (1.25 mm) vertically. Table 2. AFS Test and Test Equipment AFS Test Test Equipment Used Percent Compactability Sand Rammer Bulk Density Digital Balance ( 0.01 g) Moisture Content Percent Digital Moisture Analyzer Specimen Weight (2x2in) Digital Balance ( 0.01 g) Permeability Digital Permmeter MQI Mold Quality Indicator Splitting Strength Sand Strength Machine Green Compressive Sand Strength Machine Strength Mold Hardness Mold Hardness Tester Percent Friability Friability Machine Cone Jolt Toughness WMU Cone Jolt Tester From the jolts versus displacement, curve specimen stiffness is identified from the slope of the curve. The distance a curve travels to the right indicates the degree of plastic deformation induced by the cone on the sand system. The area under the curve is an indication of the toughness of the sand system. Equipment Table 2 shows the AFS test and equipment for the routine tests conducted. The sand ph, LOI, grain fineness, methylene blue clay and 25-micron clay tests were conducted. Please see the AFS Mold and Core Test Handbook for this test equipment. Procedure All testing procedures are defined by the AFS 6. RESULTS AND DISCUSSION In this study green sand specimens from working foundries were tested at various target compactability and active clay levels; their major green properties were obtained and their inherent variability reported. Table 1 summarizes the information about the green sand systems used in this study. Tables 3 shows the green sand properties derived from various AFS tests that were conducted on the as-mulled green sand systems. One of the most fundamental tasks of sand control is monitoring sand temper. The moisture must be varied to produce the target compactability. The higher the clay, carbonaceous, and dead materials contained in the sand the higher the amount of moisture required to produce a given compactability. Thus, moisture cannot be used as a control because the moisture requirement varies. Moisture levels that are insufficient or in excess of what is required to activate the bond must be avoided. 6 The problem with moisture control arises from the fact that as the sand composition changes, the amount of moisture needed to produce the target compactability changes as well.

6 Paper pdf, Page 6 of 8 Table 3. Properties of Green Sand Samples at High, Medium and Low Compactability Levels AFS Test Sample A (Iron) Sample B (Aluminum) Compactability (%) 32 (0.167) 38 (0.833) 44 (0.667) 30 (0.167) 40 (0.167) 50 (0.167) Bulk Density (g/cm 3 ) 0.89 (0.007) 0.83 (0.003) 0.71 (0.012) 1.08 (0.008) 0.99 (0.002) 0.81 (0.016) Moisture content (%) 3.8 (0.017) 4.1 (0.083) 4.4 (0.067) 2.1 (0.090) 2.2 (0.050) 2.6 (0.097) Specimen weight (g) 147 (0.003) 147 (0.005) 145 (0.005) 168 (0.006) 169 (0.004) 174 (0.005) Permeability (#) 89 (0.833) 105 (0.500) 124 (1.167) 105 (1.500) 103 (1.830) 103 (0.667) MQI (#) 343 (1.667) 299 (5.000) 258 (3.667) 260 (7.160) 268 (6.500) 269 (1.000) Splitting Strength (psi) 4.2 (0.067) 4.7 (0.050) 4.5 (0.050) 4.5 (0.067) 4.4 (0.133) 4.9 (0.216) Compression Strength (psi) 27 (0.833) 26 (0.667) 25 (0.500) 26 (0.830) 26 (0.500) 26 (0.500) Mold Hardness (#) 97 (0.167) 97 (0.167) 95 (0.500) 98 (0.167) 98 (0.333) 97 (0.333) Friability (%) 14 (0.333) 8 (0.167) 4 (0.333) 18 (1.000) 15 (0.300) 2 (0.333) Cone Jolt (#) 23 (0.500) 57 (5.833) 121 (4.500) 13 (2.517) 20 (3.512) 27 (1.555) Note: italicized number in parenthesis indicates standard deviations The methylene blue clay and AFS clay provided information relating to the bond. The LOI indicated the initial amount of carbonaceous material in the sand system. Data for the percent methylene blue clay, 25- micron clay, and LOI are shown in Table 1. In Table 3, it is shown that Sample B (aluminum green sand) required less active clay and moisture to attain desired compactability; also, bulk density was higher for the green sand Sample B. In general, the buildup of dead clay can be troublesome, since it will require more water to achieve the targeted green properties. This excess water is related to the increase of casting defects and additional costs. Many factors affect sand permeability, but compaction, dead material, and grain fineness are the major variables. The higher the density to which the sand is compacted, the lower the permeability because the sand grains are forced tightly together, leaving smaller voids between the grains through which air can pass. Permeability increased with compactability for both samples, however Sample A (cast iron foundry) showed a better venting characteristics at the higher compactability levels. This could logically be contributed to the coarser sand used in Sample A. There was no difference between the mechanical properties for both systems; there was no difference in mold hardness, green compression and splitting strengths data for the various compactability levels and between Samples A and B (Table 3). Differences among the compactability levels and between the sand systems could not be detected by the basic green mechanical property tests for the sand systems studied. For the green sand samples tested, the lower the compactability the higher the friability. Dry sand defects such as inclusions, erosion, cuts and washes are avoided in conventional molding systems by keeping compactability levels high (35 50%). As would be expected, friability decreases with increase in the compactability level, but friability testing showed no significant difference in and between Samples A and B. As moisture increases, green and splitting strength reaches a maximum and then drops as the sand becomes over tempered. If green and splitting strength is low, the sand will have good flowability and the cost to maintain the system will be lower. If it is too low, broken molds and poor draws may become a problem. Low green and splitting strength indicates low clay content and/or poor mulling. If green and splitting strength is too high, the molds will be stronger, but difficult to shakeout, poor casting dimensions, poor flowability and high ramming resistance are likely problems. In addition, the cost to maintain the system will be higher due to use of excessive bond. 4,5 The mold hardness for all systems tested indicated there would be a good resistance to indentation.

7 Paper pdf, Page 7 of 8 problems and shift on mold handling systems used in foundries. The modified cone jolt test was able to better differentiate among the high, medium, and low compactability levels when compared to other green property tests (Figure 7). Only the friability test results for Sample B, the one with withstanding lower number of jolts, showed significant variation as well. Apart from sand type and grain size, these systems differed in their active clay levels, and the Jolts versus displacement plots obtained from the modified cone jolt test are able to differentiate the toughness of the respective sands clearly. The capacity of Sample A to sustain cyclic load is significantly greater than for Sample B, as shown in Figure 8. This kind of differentiation is what is required nowadays in industry, and shows the value of this modified cone jolt tester. Figure 5. Number of jolts versus compactability for clean and dirty sand from sample B, showing ± σ. Figure 6. Typical failure modes from modified cone jolt test. Cracked specimen for clean sand (left) and split specimen for dirty sand (right). The modified cone jolt test was used to compare sand cleanliness. Figure 5 shows that at any target compactability, the clean sand sustained greater cyclic loading when compared to the dirty sand. Since the clean sand theoretically has a higher density, compared to the dirty sand (contamination achieved with 1% by weight of polyethylene pellets), the cleaner sand showed superior dynamic properties. In addition, it can be observed that the dirty sand showed greater variability in green properties. Figure 6 illustrates typical failure modes for clean and dirty sand specimens, where the stress-rising effect of the pellets is producing a split failure, as opposed to a crack failure for clean specimens. Figure 7. Sensibility shown by various tests to changes in compactability for sample A. Both samples showed that with increased compactability there is greater toughness (Table 3). Sample A is a stiffer and tougher sand system and would resist cope down

8 Paper pdf, Page 8 of 8 The results obtained in this study are very encouraging, and presents the opportunity to develop a control tool for industry. More development and testing is required and already taking place, in order to improve on the testing procedure, and on the type of results that can be expected when other sand samples are tested. ACKNOWLEDGMENTS The authors would like to issue a posthumous recognition to Mary Beth Krysiak for her contributions and insight on the modified cone jolt tester used in this study. The authors gratefully acknowledge Glenn Hall, Peter Thannhauser, Abraham Poot, Alexander Hiday, Michael Horvath, and Ronald Davis from Western Michigan University, for their technical support. REFERENCES Figure 8. Number of jolts vs displacement at high, medium, and low compactabilities for sample A & B. LIMITATIONS In this study, only two high production foundry green sand systems were included. Other green silica sand systems, specialty sand, alternative molding media, and reclaimed sands were not considered in this study. CONCLUSIONS AND RECOMMENDATIONS It is possible for a foundry to better monitor variability in sand properties caused by small changes in clay and moisture levels using a modified cone jolt test. The modified cone jolt test was able to better differentiate among various compactability levels (moisture content) when compared to other green property tests. The modified cone jolt test was able to show clear differences between sand systems possessing different active clay levels when compared to other green property tests. 1. Aycardi, M., M.B. Krysiak, and D. Martin, Sand Lab to the Rescue, Modern Casting, (July 2009). 2. Ramrattan, S., A. Paudel, H. Makino, and M. Hirata, 2008, "Desirable Green Sand Properties via Aeration Sand Filling" No , Transactions, American Foundry Society. 3. Krysiak. M.B., Ramrattan, S.N., Cheah S.F., Thermal Distortion of Green Sand and Chemically Bonded Sand at Cast Iron Fill Temperatures, AFS Transactions, V 110 Paper No P , (2002). 4. Krysiak. M.B., Reducing Casting Defects: A Basic Green Sand Control Program, 2 parts, Modern Casting, (Apr., May.1994). 5. Krysiak, M.B., Pedicini, L.J., Sand Testing Design, 3 parts, Modern Casting, (Feb., Mar., Apr. 1990). 6. Mold and Core Test Handbook, 3rd. Edition, AFS (2001). 7. Rich, C., Krysiak, M.B., Monitoring Sand Temper Through True Compactability, Sand Control, Modern Casting, (Mar. 1993). At any target compactability, the clean sand sustained greater cyclic loading when compared to the dirty sand. Impurities, such as core butts (pellets), lower the bulk density and increase the compactability of a dirty sand system. Further, the dirty sand showed greater variability in green property data. Thus, a finer mesh screen (0.125 inch opening) might be a better option for measuring green sand properties.