By Dr. Graham R. Rideal, Managing Director, Whitehouse Scientific Ltd., Whitehouse Scientific
SubscribeBackground - Sieving is one of the oldest methods of particle separation and is documented in ancient Egypt in the making of papyrus, the earliest form of paper.
More recently, sieve analysis has been described as the 'Cinderella' of particle metrology in that it does most of the work while getting little of the credit. The popularity of the method is reflected in the fact that there are literally millions of sieves currently being used around the world. Sieving remains unchallenged as the least expensive method of particle size analysis, however, with increasingly stringent quality assurance specifications being introduced, end users are demanding higher precision from the technique.
The weaving process employed in the manufacture of test sieves gives rise to fluctuations both in the nominal aperture size and the range of sizes in a given cloth. For example, according to the International Standards Organisation1, a nominal 63-micron test sieve can vary from 59.3 – 66.7 microns and still be within specification.
The maximum single aperture must not exceed 89 microns, which is not at all acceptable if the sieve is being used to screen out unwanted large contaminants. With such a variation in the tolerance, there could be serious consequences in industries such as pharmaceuticals where precision is paramount. It is therefore very important that the effective aperture size of a sieve is accurately known and, preferably traceable to an International standard such as the National Institute of Science and Technology, USA (NIST).
Microscopy is one method that has been used to calibrate sieves but it suffers a number of disadvantages including: expensive microscopy and image analysis equipment, two sets of measurements produced - warp and weft dimensions (see later), time consuming, the method examines only a fraction of a percentage of the sieve surface and the user usually has to send the sieves away for calibration.
An alternative method has been the use of calibration microspheres2. Here microscopy was used to characterise fine glass beads, which in turn were used to measure the sieve aperture size. The disadvantage here is that the accuracy of the method is dependant on the sphericity of the beads. Any deviation gives rise to inaccuracies because sieve diameters can be quite different from the equivalent spherical diameters of non-spherical particles as measured by microscopy. Furthermore, the wide particle size distribution of the standards results in poor resolution during interpolation.
The narrow particle size distributions of sieve calibration microspheres
The precision microsphere approach
To overcome the limitations of the broad particle size distribution microsphere standards, Whitehouse Scientific has produced a range of 30 narrow distribution glass microspheres to measure individual sieves from 20 – 3350 microns.
To avoid size definition problems which occur for non-spherical or out-of-shape particles, all the standards are certified by high precision electroformed sieves so only the particle breadth, rather than an equivalent spherical diameter is measured (this is the most appropriate dimension for calibrating wire woven sieves). Nevertheless, all the standards are shape sorted to maximize the sphericity of the microspheres so the results from the precision sieving are super-imposable with microscopic data.
Electroformed sieves are well defined and extremely accurate compared to wire woven sieves
Production and subdivision
For most sieve calibration standards down to 20 microns, a production Sonic Sifter (Gilson, USA3) was used to produce standards whose size peaked at the nominal sieve standard but only spanned one sieve either side of the ISO sieve series. Thus a 63-micron standard should have about 90% between 53 and 75 microns with approximately 50% retained on the 63-micron sieve.
The master batches were then was subdivided into appropriate weights for each test sieve. The weights were calculated to cover over 80% of the available apertures on any 200mm diameter test sieve. For example, a 200mm diameter, 63-micron test sieve has approximately 2.5 million apertures. A 1g bottle of the 63-micron calibration standard therefore contains an average of about 2 million microspheres. The single shot bottles in the series contain from 0.8g for the 20-micron sieve to 25g for the 3.35mm sieve.
The 100 stage spinning riffler4 specially designed by Whitehouse Scientific to subdivide the master batches was independently tested by the Community Bureau of References (BCR). The report concluded that the sample-to-sample variation was less than 1%.
Certifying the reference standards
Each electroformed sieve used in the analysis was calibrated microscopically using a Stage Reference graticules from both the National Physical Laboratory5 (UK) and NIST6 (the American equivalent). The apertures were identical to within 1 micron so there could be no ambiguity either from the certified sizes generated from the sieve analysis or from bottle-to-bottle variations in the single shot samples.
Calibration Graph - the high resolution achieved from the standards
Because the size distributions of the reference standards were so narrow, only three electroformed sieves were needed to cover the size range. A sieve stack was then assembled and 5 repeat analyses performed. The repeatability was quite exceptional with mean standard deviations always below 0.5%.
The higher resolution method of microscopy was then used to check the electroformed sieve data and, only when the two sets of data were super-imposable was the electroformed sieve data interpolated to construct a calibration graph supplied on each test certificate.
Calibrating a wire woven sieve
Because the standards are so spherical, the end point in the calibration process is very quickly achieved. Thus hand sieving, mechanical shaking and even Sonic sieving are all complete in about 1 minute. Hand sieving is the quickest method, as the sieve does not need to be clamped in a sieve shaker; however, some form of mechanical shaking is recommended to ensure operator independence and thus good repeatability from laboratory to laboratory.
Calibrating a sieve on a sieve shaker
To calibrate a sieve, a single shot bottle of the appropriate standard is pre-weighed before sprinkling on the sieve placed on a collecting pan. The assembly is then transferred to the sieve shaker and shaken for 1 minute. The percentage of the standard passing the sieve is calculated and the aperture size interpolated from the calibration graph on the test certificate. The analysis can be repeated up to five times as the standards come in sets of five but, in practice the method is so accurate that only one measurement is sufficient. The advantage of having a narrow distribution is clearly seen from the calibration graph where differences in the percent passing the sieve of 5% result in an aperture size difference of only 1 micron.
Comparison with the traditional microscope calibration
In the ISO and ASTM method of sieve calibration, individual aperture sizes are not measured, rather the single warp measurements (X dimension) and weft measurements (Y dimension) are accumulated – 80 to 300 counts in each dimension depending on the aperture size. Certification of a sieve therefore has the ambiguity of two dimensions. Furthermore, the recommended number of apertures to be counted can be as low as 0.01% of the total number of apertures on the sieve so does not give an accurate analysis of the complete sieve surface. By contrast, calibration by narrow size distribution microspheres analyses approximately 80% of the available apertures and give a single dimension, which is conceptually easy to understand – the diameter of a sphere that just passes.
Conclusion
This novel approach to sieve calibration fulfils all the requirements of the quality control laboratory in that it is simple, quick, accurate, repeatable, and above all, traceability to the International unit of length. Furthermore, the calibration can be performed in-house; the sieves do not need to be sent away.
Test certificate and single shot bottles of a standard
Although sieve analysis has been the ‘poor relations’ of particle size analysis and is increasingly under threat from more modern automatic measurement techniques such as Laser diffraction analysis, it is nevertheless an extremely important method of analysis for emerging businesses as it so cost effective (one sieve is about 200 times cheaper than a typical Laser Analyser). Now that the sieves can be easily and cheaply calibrated in-house and referenced to the top international standards, the traditional uncertainty of sieve analysis has finally been removed.
Bibliography
- International Standards - ISO 3310 -1:2000, see also ASTM E11
- F G Carpenter and V R Deitz, Glass Spheres for the Measurement of Effective Opening of Test Sieves, J. Res. NBS 47, 139 (1951) - now NIST
- Gilson Company, Lewis Center, OH 43035-0200, USA.
- T Alan, and A A Khan, Critical Evaluation of Powder Sampling Procedures, The Chemical Engineer, May 1970
- National Physical Laboratory, Teddington, Middlesex, TW 0LW, U.K
- NIST, 100 Bureau Drive, Gaithesburg, MD 20899-8520, USA
PDF Brochures