The measurement of active clay is critical to the control of foundry green sand. The amount of active clay present in green sand directly affects the strength, moisture, compactability, and mechanical properties of the mold. By improving the characterization of clay levels, the moisture and compactability could be more accurately controlled.

Inadequate compactability control can cause green sand casting defects, and the associated costs of scrap, rework, labor, and energy. A previous study reported that a foundry required an MB clay control within +/-0.35% to minimize scrap gray iron castings. These costs warrant investigations into alternative methods of control. 

The foundry industry would benefit from a faster, simple, accurate, and low-cost alternative to measure active clay in green sand. The AFS 2210-17-S and AFS 2211-17-S  MB Clay Tests have been used to determine active clay levels in foundry green sands since 1967. These tests are dependent on an operator making a visual determination, which leads to inherent inaccuracies based upon studies and input from numerous foundries that are AFS corporate members. Due to this potential bias and other factors, the test reproducibility within and between foundries is currently limited in maintaining ongoing accuracy.

Casting defects are consistently attributed to variations in green sand systems and limitations of the clay control methods for green sand. The standard MB test is dependent on a cation exchange reaction where active binding anionic sites on clay are bound by methylene blue dye, resulting in the sequestration of these dye molecules from solution. The amount of the bound dye molecules is determined by adding the remaining dye solution dropwise to filter paper and reading the resulting “halo.” The reading of this halo requires operators with specialized training and experience. There is often a need to monitor technicians to ensure that there is no drift in calling the halo point. Because of the discretion in calling the halo point, some foundries have reported operator bias factors can affect MB reported test results when the operator calls the halo point to try to be within the control ranges that they are tasked with staying within. This is often evidenced by histogram distributions truncated at the control limit values. 

An alternative method to the standard MB test was published in 2013 and 2020. This non-standardized dye absorption and titration methodology offered an accurate and digital measurement of active clay. Moreover, the alternative method provides an avenue to automation much easier than the current AFS standard for active clay measurement.

In this research, the previously developed method from 2013 and 2020 was modified in various procedure aspects to minimize variation, reduce costs, and simplify the method for lower sample processing times and less user error. The proposed modifications included replacing the centrifuge with filtration, as foundries typically do not have a centrifuge with sufficient force for this method. Furthermore, preliminary data suggested that using filtration leads to more reproducible results than using centrifugation in the testing method.

The new spectrophotometric method uses Cu(II) triethylenetetramine, a dye that binds stronger to the active clay binding sites, thereby allowing for lower dye concentrations in the test solution and a dye that can be measured spectrophotometrically. Methylene blue dye needs to be used at a much higher concentration, with a dye that is opaque and cannot be measured spectrophotometrically unless dramatically diluted. 

The goal of these studies was to validate a new standard method for determining active clay content of green sands, with lower variability and a shorter test time than current standard tests. This would also allow for a quicker reaction by the foundry to changing green sand characteristics. In addition, the consumables will be lower cost, safe for use in an industrial environment, and result in easier clean-up compared to the current AFS standard MB Clay test.

This method was first optimized in a university laboratory and tested as a standard method for use in foundries. Then, the method was further optimized and tested in three working foundries. Then a Gage Repeatability & Reproducibility (GR&R) study was conducted, comparing the new optimized spectrophotometric test to the standard MB test.

This analysis suggests that the new spectrophotometric test has variability that is comparable to the standard AFS MB test. This research has refined the spectrophotometric method, making it user-friendly with minimal training, accelerating the testing process.

This test has been submitted to be approved as an AFS standard test for active clay in green sands and included in the “AFS Mold & Core Test Handbook.”

Experimental Procedure and Materials

The new spectrophotometric procedure is comprised of four main steps: (1) weighing and mixing of the green sand with the copper triethylenetetramine solution, (2) equilibrium settling, 3) filtration of the sand from the dye, and 4) spectrophotometric reading of the dye absorbance at a specific wavelength of 578 nm. Much like the standard MB test, the active clay in this procedure effectively sequesters the dye from the solution, resulting in a less concentrated dye solution that can be measured. This dye more specifically binds active clay sites, and thus a lower concentration of dye can be used. Therefore, it is more translucent to light and thus can be measured spectrophotometrically, which allows for a digital quantitative reading. 

The equipment/consumables, specifications and the specific models and item numbers that are used in this project are listed in Table 1. Instruments and supplies that are used are shown in Figure 1.

Dye Solution Preparation

A dye solution of 3.33 mmol/L of Cu(II)-triethylenetetramine (Cu-T)  solution was prepared. 1.005 g of liquid triethylenetetramine was weighed and dissolved with 50 mL deionized (DI) water in a beaker. 66.67 mL of 0.1 mol/L of copper sulfate solution was measured using a measuring cylinder and an adjustable 1-10 mL pipette. Both solutions were transferred to a 2L volumetric flask, and the flask was then filled to 2L with DI water. A magnetic stir bar was used to gently stir the solution overnight while covered to allow for sufficient formation of dye complexes. The Cu-T solution was then transferred to a large amber container and stored in a cool and dark place. Due to the hygroscopic nature of triethylenetetramine, rapid handling and proper storage of the solution is required.

Sample Weight and Dosing

To optimize the dye usage, a study was completed on multiple foundry laboratory scales to determine the level of scale accuracy. Various tests were performed to determine the lowest dye and sand sample sizes that could be used while maintaining an acceptable variance. A green sand sample size of 0.5 g (± 0.1 g) along with 10 mL (± 0.1 mL) of Cu-T dye was used in the optimized procedure. A 15 mL centrifuge tube was weighed, recorded, and tared, and 10 mL (± 0.1 mL) was pipetted using a bottle top dispenser with recirculation valve into the centrifuge tube. The dye was then weighed, recorded, and tared again. Using a glass funnel, and a 1/16 or 1/8 teaspoon measure, green sand samples were scooped and transferred into the centrifuge tube. Density and properties of the foundry sands may necessitate the use of the different measuring spoon capacity. The green sand sample was weighed and recorded. With a 1/16 or a pinch scoop, the green sand can be easily measured to 0.5 g (± 0.1 g) and need not be the exact amount as the weight of the sand sample and dye can be adjusted in the CEC calculation.

Mixing/ Agitation

The prepared samples were first mixed by vigorously shaking by hand for 5 seconds and then mixed by vortex mixer at speed setting of approximately 2000 rpm for 10 seconds. The samples are then allowed to sit for at least 2 minutes to allow sufficient settling and equilibration of the green sand. Then the samples are vortexed and allowed to settle a second time for 2 minutes.

Separation by Syringe Filtration

Using a 3 mL disposable syringe 3 mL of the supernatant dye was pipetted and filtered through a 25mm 0.45 µm PTFE or 13mm 0.45 µm polyethersulfone (PES) syringe filter into a 4.5 mL clear polystyrene disposable cuvette. The 13mm 0.45 µm syringe filter was preferred by the operators as it was easier to expel the liquid into the full cuvette volume. Square bottom cuvettes were necessary to minimize variance in our tests and trials, as those that have a tapered bottom proved to be problematic in the spectrophotometer.

UV-VIS Spectrophotometry

The UV-VIS spectrophotometer is first tared with a cuvette that is filled with deionized (DI) water. The outer walls of the sample cuvette as prepared above were wiped with disposable wipes and inserted into the spectrophotometer. Absorbance was measured at the peak wavelength of the Cu-T solution of 578 nm.

The linearity of the exchange solution, the absorbance of the Cu-T dye at various concentration levels was confirmed. Based on testing, it was shown that the dye remained stable for at least 6 months and this linearity was maintained during this period. For the test study, the spectrophotometers used at WMU and supplied to industry partners were all calibrated using standard filters.

Details regarding dye mixing/ verification instructions and recommendations for maintaining calibration consistency of absorbance-measuring equipment will be available in the AFS procedure that is forthcoming.

CEC (Cation Exchange Capacity) Determination

CEC was determined by the difference in the concentration of Cu-T of the sample to a referenced standard, considering the volume of exchange solution and valance of the copper cation, and adjusting for the weight and the moisture content of the sample. Since there is an equivalent exchange of sample and solution, the valance of the Cu-T is the charge of the dye, which is 2. The sample concentration of Cu-T was calculated by dividing the sample absorbance obtained from the spectrophotometer by the absorbance of the straight Cu-T dye batch sample and this ratio is multiplied by the calculated batch concentration of the straight Cu-T dye. This straight Cu-T dye concentration should typically be very close to the 3.33 mmol/L that is prepared as described above. Moisture content was determined by weighing sand sample before and after drying in an oven at 107°C (224.6°F). Moisture content was adjusted in the CEC calculation only and not during other methodology. CEC was calculated using the equation below.

CEC = cation exchange capacity, expressed 
in cmolc / kg
Xo = concentration of Cu(II) [mmol / l]
Ao = UV Vis Spectrophotometer Absorbance of straight dye
As = UV Vis Spectrophotometer Absorbance Value of Sand Sample
D = dye volume [mL]
z =  valency of index cation
M sample  = sample weight [mg]
W 107° = sample moisture measured at 
107C (224.6F)” 
 

Results

The objective of this research was to optimize the new spectrophotometric method for integration into foundries as an alternative method for measuring the active clay in green sands. Trials were conducted as described at three foundries, EJ, John Deere (JD) and Metal Technologies Inc (MTI). The active clay content was measured for several green sand samples in foundry testing.

During this testing, the three foundries conducted tests on several individual samples that were taken during their own operations in their respective facilities. These samples are labeled as MTI, EJ, or JD samples to indicate their origin in the data tables. The three foundries each used Cu-T dye from the same batch as prepared by Western Michigan University (WMU).

Testing in Foundries

The optimized procedure was disseminated to be used in all foundries, and testing was then conducted at each foundry. Results for the in-foundry testing of green sand are shown in Table 2. The average SD value for each foundry were reported to be 0.21 (MTI), 0.12 (EJ) and 0.22 (JD).

Conversion Between Cationic Exchange Capacity (CEC) and Standard Methylene Blue (Mb) Active Clay Measurements

Active clay has traditionally been measured as a percentage, with this percentage measured as compared to active bentonite. This bentonite is assumed as “100%” active.

However, this standard bentonite can vary in activity, dependent on the source of the standard, with most standard being 80%–90% active. Studies have shown the samples are consistent based on type of bentonite (sodium or calcium). The CEC measurement is a direct measure of cation exchange activity and allows for direct comparison between foundries regardless of clay blend used.

CEC can be converted over to active clay to compare the new spectrophotometric active clay test to the traditional methylene blue active clay measurement. The method to convert to a MB equivalent is similar to that used in the current standard MB tests. A known standard of clay at a known concentration is tested using the new test technique and an MB conversion factor is calculated and applied to the subsequent CEC results from each sand test.
The active clay measurements from testing that was conducted on daily foundry sand samples are shown in Figure 2; these measurements are converted to traditional active clay percentage (in orange/dashed lines) and compared to traditional MB test results (in gray/solid lines). As shown here, the measurements generally have good agreement for the JD and EJ foundries, while there is an offset between the measurements for the MTI foundry. 

After some additional analysis, it was determined that this is likely due to a normalization issue with a change in active clay standard and should be able to be corrected with appropriate normalization to the standard bentonite clay. These types of issues can be avoided using the CEC measurement, which is a direct measurement of cationic activity of clay and not dependent on bentonite standards.

Comparison of the New Digital Active Clay Measurement in Green Sands with the Traditional Mb Active Clay Measurement

As previously noted, Figure 2 shows a comparison from the results of the new method testing from testing and the traditional MB test for all three foundries, MTI, EJ, and JD all presented in percent active clay. The results for JD and EJ foundries have consistent active clay content using either test. MTI results show the noted discrepancy, but the overall data show a consistent trend in active clay. 

A GR&R study was performed to determine if the optimized spectrophotometric test had similar or better variability when different operators performed the test as compared to variability of the methylene blue test. After the GR&R tests were performed, it was concluded that the new test had similar or better variability when compared to the traditional methylene blue historical study results conducted independently by three foundries. (Table 3).

A concern with the traditional MB test is the influence that the operator can have biasing results to meet internal specifications. The qualitative aspect in calling the halo point allows wide latitude to the operator. Figure 3 shows the MB test data for a foundry during a four-month period. This foundry has an internal specification range with a lower limit of 7.2% and an upper limit of 7.8% active clay. As shown, the data is skewed to the upper specification limit, suggesting that the operators are biasing the results to meet the upper specification limits. 

Conclusion

In this study we have demonstrated that a new spectrophotometric method for measuring active clay in green sands works well with low variability over the entire range of clay levels used in most foundries. This method has been simplified by using filtration for separation of sand from the dye as opposed to centrifugation that was used in previous work. Not only does this simplify the process to work well in foundries and lower sample measurement time, but this also eliminates the need to use a high-speed centrifuge, which is a costly piece of equipment with required periodic maintenance. 

Measurements from the new test are taken as a CEC measurement, which can be correlated to the standard MB test measurement for comparison. We recommend eventual conversion by foundries to the new CEC measurement instead of the standard MB measurement that is dependent on purchased bentonite standards that can vary by supplier. However, the new CEC measurements compared well with converted measurements to the standard MB test. We suggest that during the transition to the new spectrophotometric test, foundries take standard MB measurements and the new measurements side-by-side for comparison until they are comfortable with a full transition to the new test.

The standard MB test method is prone to subjectivity and takes a great deal of training and experience for success. Additionally, it was observed that several operators run the test, keeping a known target in mind, potentially biasing readings. We tested the alternative active clay test, comparing it with the standard MB test in a GR&R study. In this study we demonstrate that the alternative test has similar variance as the standard MB test when operators measured a series of random test samples without knowledge of the estimated active clay percentage going into the test (as in a GR&R study). 

Overall, we demonstrate in this research study that the spectrophotometric measurement active clay test offers an enhanced alternative option to the standard MB test. This new test is relatively simple to perform, without extensive training or experience required, as is necessary for the standard MB test. The digital test also has low startup and operation costs, and the test can be conducted in 10 minutes or less time. This test is also conducive to automated operation, which may be explored in subsequent research studies.