Maintaining sand quality requires frequent testing.

Successful Sand Testing System Design -- Part 2:

Maintaining Sand Quality Requires Frequent Testing

Foundry sands vary widely, requiring physical and chemical testing to preserve benefits.

Core Sand Influx

Core sand influx can cause wide variations in the composition and properties of molding sand. Caused by changes in product mix and varying amounts of core sand influx, they can be dealt with by anticipatory control of additives once the effects of the influx can be measured. Methylene blue, LOI and 25-micron clay tests can measure sand composition changes at shakeout. Frequent product mix changes make anticipatory controls essential.

The newest test to determine AFS clay content, the 25-micron clay test quickly performs clay washes, an important measure of shakeout sands. It uses a 25-micron sieve basket, like that shown in Fig. 5, in an ultrasonic scrubber to remove the AFS clay and retain the sand. An infrared dryer readies the sand for a screen analysis in less than half an hour, as compared to the old manual siphon and Autoclay methods that took hours.

Gas pressure testing of shakeout sand provides a quick alternative to LOI and volatiles testing as a means of measuring burnout. Using a microsplitter aids in providing a representative sand sample for gas pressure testing since the required specimen is small and segregation and sampling of shakeout sand can be difficult.

Core sand influx becomes more critical as the core sand/molding sand ratio per mold increases. The greater the ratio, the greater the fluctuation in the silica content of the molding sand at shakeout. This influx is also a function of the collapsibility of the cores, which depends on the binder system(s) used and the binder level.

Cone jolt toughness, friability and wet tensile tests are sensitive to the effects of core sand dilution. Heavy core sand influx can result in brittle molding sands.

Cone jolt toughness testing is dynamic in that failure is incremental with staged crack propagation. In the test, a standard specimen is jolted while bearing the full weight of a cone penetrator and the number of jolts until failure is counted. Simple and fast, it is ideally suited for production environments.

Failure in fewer than 30 jolts designates a brittle sand. Sands that have not been remulled or reused are classed as brittle. Laboratory mixes rate between 14 and 18 jolts. System sands develop toughness as they are reused and remulled. A heavy influx of core sand can cause the system sand composition and properties to resemble those of a brittle new sand or laboratory mix.

Broken molds and the inability to pull a deep pocket in a pattern often are the result of insufficient sand toughness, or brittleness. A cone jolt toughness tester measures the relative brittleness of the compacted sand, providing information quite different from compressive or tensile strength properties.

Though some molding sands may have adequate strength characteristics as measured in routine tests, they may be very brittle sands. There is an inverse relationship between cone jolt toughness and friability. Sands with low cone jolt toughness values usually exhibit high friabilities.

The cone jolt toughness test also can be used to provide a quantitative measure of when sand toughness may be too high. If it is too high, flowability problems can result, especially in restricted pattern areas. The maximum toughness allowing good flowability and compaction can be determined from the cone jolt toughness data by relating it to a record of molding problems. Excessive toughness can result in voids and soft areas in the mold.

Cone jolt toughness is alone among the usual mechanical tests that relates the lifting ability of the sand or the depth of sand that can be drawn from a deep pattern pocket.

Sand Mixtures

If the core sand influx is large and inconsistent, the sand properties and composition of the shakeout sand will vary widely. It is desirable to keep the influx as constant as possible. In systems in which the silica content is very inconsistent at shakeout, it is also variable when it eventually goes back into the mixer. The mixing helps, but with short mulling cycles, the sand coming out of the mixer is more likely to be nonuniform.

The variability in the core sand influx can be reduced in some cases by carrying as much of the core sand as possible out of the system in the casting. After it is shaken out of the casting, it gradually can be returned to the system in a controlled manner. It also could be sent through some type of reclamation process to minimize variations or effects of residual binder.

New sand additions often are based on a certain number of pounds per ton of metal poured, minus the core sand influx. Additional bond is added to coat the core and new sand, but the new sand addition is not reduced in relation to the core influx.

Though long considered an expensive and wasteful approach, many foundries still feel that better sand results are obtained. They claim core sand to be more difficult to coat than new sand, and the character of the green sand is deteriorated by high amounts of core sand influx. This approach minimizes the effects of residual binder, and the accumulation of other residuals (i.e., core binder solvents and ammoniacal nitrogen) is prevented.

Nitrogen and hydrogen in molding sands are acknowledged contributors to nitrogen defects in castings. Total nitrogen in molding sand does not correlate with the occurrence of nitrogen fissures because total nitrogen includes readily available and fixed forms of nitrogen. The fixed form occurs in a more stable molecular structure, like seacoal, and is not a problem. The kinetics of the pouring process are such that only the most readily available nitrogen, quickly picked up by the metal before it solidifies, contributes to nitrogen fissure defects.

Nitrogen in the form of ammonia originates from certain core binders, especially those containing urea, ammonium chloride, ammonium nitrate or hexamethylene tetramine, during pyrolysis. Ammoniacal nitrogen can be absorbed by the bentonite in the green sand. This nitrogen, which accumulates in the molding sand, is readily available at the metal-mold interface, and is known to contribute to some fissure defects in castings.

The ammoniacal nitrogen test is used to monitor molding sands where nitrogen fissure defects in castings are a problem (see Fig. 6).

Silica

The silica level of the sand has an important bearing on the sand quality and must be kept constant within a certain range. If the silica level is too high, the endothermic nature of the sand and its ability to dissipate heat is reduced, resulting in burn-on defects. The expansion tendency of the sand also is increased. If the silica level is too low, the sand loses its refractoriness and is subject to burn-in defects.

The AFS and 25-micron clay tests, and the specimen weight of the sand, provide information relating to the silica level of molding sand. The Silica Program Test Series requires more time, equipment and expertise, but provides a more accurate determination of the silica level, in addition to providing the M.B. clay, carbonaceous and fluxing material (which is further subdivided into inert fines and oolitics), and metallic contents of the molding sand. This test series can be used to indicate long-term trends occurring in sand composition.

Fine Removal, Lump Formation

The fineness and distribution of sand entering the system are not necessarily the fineness and distribution of the sand after it cycles through the system. Fines generation and removal and lump formation can have an important bearing on binder requirement, permeability, casting finish and the properties of the sand, and often are overlooked.

The sand should be checked regularly in the raw sand handling system and at various points in the green sand system for fineness and grain distribution. The presence of large balls and clusters in the sand can be extremely detrimental to casting quality.

A simple test for balls and lumps can be performed by weighing a sample of sand and passing it through a 3/8 in. sieve to extract its lumps. The lumps remaining on the sieve can then be weighed, riddled through a U.S.A. sieve No. 6 to extract the core lumps. The residue is again weighed and the results expressed in a percent as sand lumps and core lumps. The lumps can be examined further for LOI or moisture and M.B. clay to confirm their source as pieces of core, sand agglomerations, clay balls, shot, slag or tramp metal.

Sand containing 5% total lumps can result in serious mold quality problems, such as forming compaction voids and "bridging" problems in restricted pattern areas. Core lumps with residual binder and sand lumps containing high moisture can cause localized gas defects, while other foreign matter can cause obvious surface flaws.

Lumps can be formed at any point in the system, for example during mulling, sand cooling or at shakeout. Any rolling action of the sand can cause their formation, and is aggravated by high clay and compactability levels. Lump formation and removal should be monitored at several points in the system in order to determine their source, and steps taken to prevent their development.

Excessive hot strength due to moisture or Western bentonite can cause the formation of very large lumps at shakeout. They often are removed by screening and can be crushed using a lumpbreaker. The crushed material should be run through a finer screen before return to the system to prevent it from becoming a source for smaller balls and clusters.

Lumpbreakers can be installed after shakeout, followed by magnetic separators to remove metal particles. Lumpbreakers allow the use of a higher ratio of Western bentonite without the penalty of excessive lumping and the large loss of sand at shakeout.

The washed fineness and the fineness of the base sand do not represent the fineness of the molding sand used in contact with molten metal. The air-dried fineness test should be performed monthly to identify trends that may develop on sand. This test uses a 50 gram sample of the unwashed, prepared sand, which is air-dried to a compactability of less than 20%. After air-drying, a sieve analysis is performed using the usual procedure, with the exception of a 2 1/2 minute agitation period rather than the usual 15 minutes.

Fines removal, or the use of coarser core sand, changes the distribution of system sand. Since it is necessary to maintain a certain amount of fines in order to maintain casting surface quality, some foundries use between 10-12% fines on the 140 screen for burn-in control. If dust collector fines are added to the system sand, they should be tested for M.B. clay content, fineness and LOI so that the material returned can be classified and not be another unknown variable.

Sand Capacity

One of the best ways to sustain favorably constant sand properties in a system is to maintain a high sand-to-metal ratio and a large sand capacity.

Hot sand problems can be reduced by maintaining a large sand capacity in which sand is recycled more slowly through the system. Storage bins should be designed to discharge sand on a first in-first out basis. Sand sticking to hoppers can occur in sands having high clay levels, and results in only part of the sand being circulated. Where large portions of the sand cling to hoppers, it has been observed that the sand is recycled as much as four times as often as it should be, based on the capacity of the system, seriously compounding hot sand problems.

Cycle time and hot sand development can be monitored by sand temperature tests. Time required for the sand to become hot starting from a cool condition, after a weekend, for example, provides useful information reflecting system design.

Sand-to-Metal Ratio

Merely increasing the amount of sand in the system, or sand capacity, reduces hot sand problems because there is a greater volume of sand in comparison to the metal cast. But increasing the sand-to-metal ratio (casting weight per mold) provides other advantages. It builds a buffering effect into the system that minimizes variations.

The higher the sand-to-metal ratio, the more variation it takes in the process to upset the sand composition significantly. While the amount of burned out clay and carbons, core sand influx and fines generation, etc, per casting is the same, the effect of these is a reduced percentage of the total sand in the system.

In order to maximize sand-to-metal ratio, some consideration should be given to the amount of metal poured per mold and flask size. The economic contribution of "free riders" must be reviewed to assess their total effect on casting quality.

A prudent way to reduce the effects of clay and carbon burnout and new additive additions is to dilute shakeout sand with large amounts of spill sand and excess prepared sand that would be carried over to the shakeout on belt conveyors from the molding and mixing stations. The shakeout sand could then be blended with the spill and excess sand in the homogenizing drum and fed into the sand cooler.

After cooling, raw sand, clay and carbons could be added as dictated by a computer file based on pattern scheduling, and the sand further homogenized before being transferred to the return sand storage bunkers. The shakeout sand and new additions, instead of comprising 100% of the sand in the mixer, would then constitute a smaller percentage of the prepared sand consisting largely of the more homogeneous recycling sand in the system.

Figure 7 schematically depicts this type of system, together with some of the other ideal sand system features mentioned previously, such as an automatic compactability control at the mixer, weigh hoppers, control of the mixing cycle and bond development using a wattmeter (a scanner to detect unpoured molds at shakeout so that the bond and carbons additions can be adjusted), a lumpbreaker after the shakeout, followed by magnetic separators, a homogenizing drum and sand cooler.

All of these elements help to average the normal variations in the return sand. The improvement is evident through sand test data that shows reduced variation in the sand throughout the casting process.

Editor's note: The third and final installment next month will focus on sand quality and will cover additional testing procedures and data correlation. [Figure 7 Omitted]

PHOTO : Fig. 5. Equipment shown is used to determine the newer AFS 25-micron clay test that uses a

PHOTO : 25-micron sieve basket.

PHOTO : Fig. 6. Ammoniacal test device is used to monitor molding sands where nitrogen fissure

PHOTO : defects in castings are a problem.