Newest Developments in Sand Testing

(Click on Figures to see enlarged view)

In recent years there have been several advancements made in the area of sand testing and control. Some of these advancements involve incorporation of more modern technology to conventional sand tests, others involve development of new tests and equipment to address current challenges in sand control. 

Example of an advancements involving incorporation of more modern technology to conventional sand tests include the development of more modern permmeters and strength machines (Figures 1 and 2). 

Figure 1

Modern strength machines utilize load cells for measuring sand strength characteristics, and modern permmeters use new air flow technology for measuring permeability. In addition, these modern devices provide digital readouts of the data and allow exportation of the data to PC computers so that trends can be identified more quickly, leading to better control.

Figure 2

An example of the development of new tests to meet current challenges includes the introduction of the disc transverse test for testing of chemically bonded sands. First introduced to foundry sand control by Prof. D. Boenisch, this test offers a new alternate or supplement to tensile strength. 

Figure 3

Tensile strength has been the conventional test used to control the strength characteristics of the sand, but the tensile strength test is quite variable as far as repeatability, and is not sensitive enough to humidity which can cause deterioration of cores. Figure 3 shows the conventional tensile strength specimen (often referred to as a "dog bone" specimen. Figure 4

Figure 4

  compares it to the new disc transverse specimen which is 50 mm diameter x 8 mm thick. Figure 5 shows how the disc is supported on its edges and loaded across its diameter in the disc transverse 

Figure 5

strength test. The new disc transverse test offers several advantages over tensile strength including better repeatability due to a more consistent plane of failure, better simulation of stress that can cause core breakage, especially to many of today's thin and complex cores such as water jacket cores (Figure 6). 

Figure 6

The stress is better simulated because rarely are cores pulled apart as dog bones are in the tensile test. When thin, problematic cores break, the stress causing the core breakage is usually more similar to the stress applied in the disc transverse test. Another advantage to 

Figure7

this test arises from the fact that the disc transverse specimen is thin. As Boenisch pointed out, when a core is first prepared, when it leaves the core box, the strength gradient will resemble Figure 7. The exterior will have the highest strength where the core was in contact with the core box when blown, while the interior will have slightly less density and strength. After a core is stored, however, and its surface is subjected to the

Figure 8

 effects of humidity, the strength gradient can invert, producing a strength gradient as shown in Figure 8. If a thick core specimen, such as a dog bone is tested for this effect, the effect can be masked by the high strength interior, and the degradation, which is a surface effect, can go unrecognized. A final advantage of this test is its simple shape, a simple disc 50mm in diameter and 8 mm thick. Because this shape does not have critical contours like the dog bone, it can very easily be incorporated into production core boxes. Incorporating the test specimen into the core box allows several advantages. Differences in core machine or core box performance can be identified. Vent locations can be optimized to produce the highest density and strength. Disc specimens can also be left with the cores, if cores are stored, and if degradation due to humidity is a question, controls can be set up to specify a maximum shelf life or a minimum disc transverse strength. This type of control program would be much better than the conventional program of testing the sand mixture in the laboratory. While a sand mix can produce good laboratory strengths, the same sand mixture can produce bad cores in production if the core is not blown well, or if the cores are subjected to humidity during storage.

Another common challenge facing U.S. foundry men today is in dealing with the affects of core sand dilution. Today, the core binders used break down rapidly and thoroughly at the pouring temperatures resulting in a much higher influx of burned core sand entering the sand system at shakeout. This results in the formation of brittle sand. Core sand dilution is not the only cause of brittle sand, as hot sand, poor moisture clay relationships and even excessive new sand additions can also cause brittle sand. In dealing with brittle sand and the effects of core sand dilution, there are two types of effects; effects of condensates, and effects on rebonding.

Figure 9

 Effects of condensates are chemical effects that result from a waterproofing of the clay. As shown in Figure 9, water molecules, which are polar, orient themselves between the clay platelettes and they move in and out as the clay is hydrated and 

Figure 10

dehydrated. If organic condensates accumulate in the bentonite, however, these organic condensates can hinder the movement of the water molecules in and out from between the clay platelettes as illustrated in Figure 10. This produces a

Figure 11

 waterproofing effect that deteriorates the properties of the molding sand. The wet tensile test, Figure 11, is sensitive to this type of deterioration. In this test, as illustrated in Figure 12, a specimen is formed in the specimen tube and the specimen has a taper on the one end which fits into a lift off ring. During the test, the end of the specimen with the ring is 

Figure 12

subjected to a heat, and the heat drives the moisture in the sand back, creating a wet layer or condensation zone. This simulates what happens in the mold from the heat of the molten metal. A tensile stress is then applied to the lift off ring, and the wet tensile strength, or the strength of the wet layer is measured. Wet tensile strength is critical because is represents the weakest layer in the mold. Failure of the wet layer produces expansion defects such as scabs, buckles and rattails.

Figure 13

When condensates build up in the molding sand, wet tensile strength drops off. It must be understood, however, in applying the wet tensile test, that wet tensile strength is affected by the bond formulation if a blend of Western and Southern bentonite are used (Figure 13). The higher the amount of Western bentonite, the higher the wet tensile strength. The more Southern bentonite there is in the blend, the lower the wet tensile strength. Wet tensile strength lower than that predicted by the bond formulation then can signal effects of condensates, and can indicate that not enough new sand and clay are being added to the sand system to dilute this effect.

Figure 14

Another condensate from core binders is ammoniacal nitrogen (depending on the chemistry of the binder system employed). Some core binders give off ammonia as a decomposition product when they pyrolize. This ammonia can

Figure 15

 become absorbed in the clay in the molding sand and can be picked up by the molten metal and rejected during solidification to cause subsurface nitrogen fissure defects (Figure 14). Figure 15 shows a ammoniacal nitrogen tester that can be used to measure ammoniacal nitrogen in molding sand. Figure 16 shows several sand systems along with their ammoniacal nitrogen contents and 

whether or not nitrogen fissure defects were being experienced. With the exception of the first system which went through both good periods and bad periods, the remaining systems fell into two categories based on the presence or absence of the defect. The systems with ammoniacal nitrogen levels over 60 ppm were all reported to produce the defect, while those that ran below 60 ppm did not produce the defect. So as a general guideline, ammoniacal nitrogen levels over 60 ppm indicate the sand is prone to producing nitrogen fissures. The main source of ammoniacal nitrogen is from decomposition products of certain core binders, especially hot box binders cured with ammonium salt catalysts, shell binders, and phenolic urethane cold box binders as well as others.

Figure 17

  Figure 17 shows results of a case study from a foundry in Australia that was at one time running as high as 200 ppm ammoniacal nitrogen and which was experiencing severe problems with nitrogen fissures on a joint face of an engine block. The foundry was using a hot box binder at the time, and due to the problem, eventually switched to a warm box binder. Figure 17 shows the ammoniacal nitrogen level dropping along with the occurrence of the defect. Core binders are not the only source of ammoniacal nitrogen, however. Sometimes it appears from unexpected sources. In one foundry, it came from a woodflour that was from particle board. The resins and glues in the particle board were the source of the ammoniacal nitrogen and when the foundry switched to a woodflour containing only natural wood fibers, the problem disappeared. In another case, ammoniacal nitrogen came from well water that became contaminated from fertilizer.

Figure 18

Effects on rebonding are a physical effect of core sand dilution. Certain core binders leave a layer of lustrous carbon on the sand grain that can inhibit clay bonding. Figure 18 shows core sand grains that were exposed to heat with no lustrous carbon 

Figure 19

formation, whereas, Figure 19 shows core sand grains that were left with a layer of lustrous carbon after exposure to heat. Whether or not lustrous carbon forms on the sand grains is largely dependent on the binder chemistry (phenolic urethane binders have the highest tendency to produce lustrous carbon) and the amount of oxygen present when the binder pyrolyzes. If lustrous carbon is a residual left on core sand grains, when the core sand enters the greensand system at shakeout and eventually goes into the muller, it will be difficult to mull the clay onto the sand grain because lustrous carbon is almost like a lubricant. Negative effects on rebonding are difficult to detect with the conventional tests, such as green compression. The most sensitive tests to detect these effects include the friability test, and the cone jolt toughness test.

Figure 20

Figure 20 shows the friability test. In this test, two standard test specimens are rotated in a screen for one minute, and the weight percent of sand that abrades from the surface is measured. Molding sands that have a high friability are prone to producing erosion and inclusion type defects. Generally, under 10% friability is considered good. Molding sands that run over 10% friability are generally prone to erosion and inclusion type defects, because the molding sand cannot withstand the erosive flow of the molten metal. 

Figure 21

Figure 22

The cone jolt toughness test is shown in Figures 21 and 22. In this test, a standard specimen is formed with indentations in the top and bottom. The bottom indentation accommodates a centering pin on the test platform of the instrument, while the top indentation accommodates the rounded point of a cone shaped weight. In the test, a cam raises the test platform and then drops it repeatedly. The specimen is then subjected to a jolting action, while its top is under the weight of the cone. Eventually, the specimen fails and the number of jolts until failure is recorded. Molding sands that fail with less than 40 jolts are considered brittle and are prone to difficulties in pulling deep pockets in a pattern/broken molds. Molding sands that have a high influx of burned core sand often have very low cone jolt toughness. Cone jolt toughness, in contrast with friability which is a measure of surface brittleness, measures the bulk brittleness characteristics of the molding sand. Neither friability nor cone jolt toughness directly correlate with green compression. Some molding sands with high green strengths are actually very brittle and friable and are prone to producing molding difficulties in addition to erosion and inclusion type defects.

There are ways to reduce the effects on rebonding, including increasing the clay level prior to running heavily cored jobs, ensuring good mulling, mulling at a slightly higher compactability, adding or increasing Western bentonite, and adding a limited amount of moisture (less than 2.0% prior to the silo). Another tip in reducing effects of core dilution is to ensure that all lumps in the return sand are reduced to grain size, and avoidance of shakeout systems that produce lumps.

Figure 23

Another fairly recent advancement has been made in the area of controlling incoming bentonites and preblends. Figure 23 shows a soluble and leachable calcium and magnesium determinator that can be used for this purpose. This unit measures the soluble salts and exchangeable ions that affect the physical properties that the bentonite will develop in the sand mixture. Ensuring that the soluble salts and exchangeable ions are consistent provides assurance that the bentonite or preblend will perform consistently. 

Figure 24

Figure 24 shows typical values for Southern bentonites. Note that by observing leachable calcium, one can distinguish between Southern bentonite from Alabama versus Southern bentonite from Mississippi. Typical values for Western bentonites are shown in Figure 25. Note that there are high magnesium and low magnesium Western bentonites. 

Figure 25

Figure 26 shows the values obtained on an incoming preblend containing Southern bentonite and sea coal over various dates. It can be seen that the incoming preblend was consistent and consistent performance could therefore be expected. It also shows, from the leachable calcium figure, that the preblend contained the Alabama Southern bentonite as opposed to the Mississippi Southern bentonite. There are no other simple tests that provide this information.

Figure 26

Figure 27 shows an example of inconsistent shipments of Western bentonite supplied to a

Figure 27

 foundry in Canada. When laboratory mixes were prepared using the two different clays, the hot strength data, also shown in Figure 27, confirmed that the bentonites would produce different physical properties. If using the first bentonite, and then suddenly and unexpectedly supplied with the second bentonite, the molding sand properties would begin to change as the new bentonite was slowly introduced as new clay added at the mixer. This may necessitate changes in moisture, bond level, or other system parameters, in order to maintain sand within specifications, and may even contribute to molding difficulties and/or casting defects. 

Figure 28

Figure 28 shows application of the test to a preblend containing Western bentonite, Southern bentonite and carbons. A particular foundry was using preblend from two different suppliers, and while the formulation for the preblend was identical for the two suppliers, the performance was different. When using a preblend identified as Bond #1, the foundry experienced severe mold break up after pouring. When using Bond #9, there was no problem . The foundry suspected that Bond #1 contained more Southern bentonite than specified by the formulation and that the low hot strength of Southern bentonite was creating the problem. The top two rows of data show the base values obtained on the raw Western and Southern bentonite used in Bond #1. The next row shows the theoretical values that were calculated from the base values and the preblend formulation. The next row of data shows the actual data obtained when testing Bond #1. Since the theoretical values matched extremely well with the actual values, this proved that the preblend was properly prepared according to its formulation, and there was not more Southern bentonite in the preblend as suspected. Examining the values for Bond #9, shown in the last row, indicates that there was a difference between Bond #1 and Bond #9. Bond #9 apparently contained a low magnesium Western bentonite. In general, low magnesium Western bentonites have higher hot strengths. It is important to point out that this does not suggest that high magnesium Western bentonites are inferior. Many foundries use high magnesium Western bentonites quite successfully, and even prefer them due to better shakeout. The point here is that these tests can be used to check preblend formulations and raw bentonites for consistency as an incoming material test.

Figure 29	Figure 30	Figure 31
Figure 32	Figure 33	Figure 34

Figures 29 - 34 show various mold test tools that can be used to monitor mold and core quality. Figure 29 shows mold hardness testers, the B scale being the most common. This is a good mold test tool for low pressure greensand molds. Figure 30 shows mold strength testers for high pressure greensand molds. Figure 31 shows an electronic mold strength tester that is digital and has no delicate indicating hands that can become damaged or bent by misuse. Figure 2 shows a scratch hardness tester that can be used on chemically bonded sand. Figures 33 and 34 show the Mold Quality Indicator which can be applied to greensand or chemically bonded sand.

Figure 35

The Mold Quality Indicator is different than the other mold test tools in that it does not measure strength. The Mold Quality Indicator measures how open or porous the mold or core is. The higher the reading, the less porous, and more dense (grain to grain), the mold or core. This test is very similar to the old mold permeability test, except that the readings are the reverse (dense molds have lower permeabilities), and the MQI package is more portable and uses more modern technology including a digital readout. Figure 35 shows that for a given sand fineness, the higher the MQI reading, the higher the mold or core density. To compare apples to apples then, one must compare the same grade of sand. Comparing different grades could be misleading unless charted as in Figure 35. 

Figure 36	Figure 37	Figure 38
Figure 39	Figure 40	Figure 41

Figures 36 through 41 show how MQI, unlike any other sand test such as green compression, has a definite relationship to casting quality. In a laboratory test, furan no-bake specimens were prepared at different densities (number of rams) and different work times (Figure 36). Figure 37 shows a schematic of a Gertsman Penetration Test Casting that the core specimens were placed in. This test casting provides a very severe test for metal penetration. Figures 38, 39, 40, and 41 show the cored cavities in the test casting that were produced from the 10 ram, 8 ram, 6 ram and 4 ram specimens, respectively, with the MQI reading indicated in the lower right corner of each picture. It is very evident that as the ramming decreased and the sand aged, the MQI dropped and penetration increased.

A case study on application of the Mold Quality Indicator to greensand appeared in a paper entitled "Quality Improvements in Greensand Using the Mold Quality Indicator" by L. Hastings, Fahramet, Kubota Corp. In this particular case study, the foundry wanted to eliminate the need for a facing sand by obtaining higher compaction on their system sand. They monitored MQI on 53 molds over a 7 day period and rated them in terms of surface finish (Figure 42). 

Figure 42

They found that the best results were obtained when MQI was over 155 in their particular case. Next, they selected a pattern with several flat areas on both horizontal and vertical surfaces (Figure 43). They made a mold at their standard jolt and squeeze times of 7 and 8 seconds and the average MQI was 149.9

Figure 43

 (Figure 4). They increased both the jolt and squeeze times to 10 seconds and made another mold and the average MQI increased to 168.4. Figure 46 shows the casting results from these molds, which indicate an improvement on the vertical faces of the mold. They then ran at the new jolt and squeeze time for one month and reported a noticeable improvement in surface finish, especially on the vertical faces, and that roughness was rarely encountered. When roughness did occur at the new settings, it was felt to be more temperature related than compaction related. They consistently hold the MQI over 155 and reported that they find less need to "tweak" the system for clay content, moisture content, etc. They found that getting good compaction was half the battle. By using the MQI, they were able to optimize compaction and improve surface finish. The foundry then also decided to apply the instrument in their no-bake area as part of a continuing quality improvement program.

Figure 44

Figure 45

Figure 46

Figure 46 shows another area in which advancements are being made in the area of sand control. Sieving has been the conventional test for controlling sand fineness, but it is widely known that sieve analysis results are variable from sieve stack to sieve stack, from technician to technician, and from laboratory to laboratory. This presents many problems, especially when applied as an incoming material check between supplier and foundry. Today, new technology utilizing photo-optical particle size measurement is being introduced to foundry sand testing and control (Figure 46). This technology measures the particle diameters of sand grains as they fall in front of a camera that samples at very rapid speed. This provides more true particle sizing that is not subject to error from oversize openings that exist in sieve analysis. The photo-optical data can be correlated to sieve analysis, however, in order to provide a comparison and a common frame of reference to sieve analysis. The technology also significantly reduces testing time in addition to automating the particle size and distribution analysis.

In conclusion, sand quality is of fundamental importance to casting quality. Considering the cost of re-work and scrap, no foundry can afford to be operate without a good sand testing and control program. Advancements in the technology available today, along with the changing needs in sand control that occur with changes in sand systems, and the ever more stringent quality demands shape the development of new tests and equipment for monitoring and controlling sand and casting quality.