Silicon particle size analysis

GKN is an international company with over 48,000 employees in more than 30 countries. Its main business sectors are automotive & aerospace component manufacturing and powder metallurgy. Sheepbridge Stokes is unique within GKN, because it manufactures cylinder liners for internal combustion engines, used in automotive, off-highway and marine applications. These are primarily manufactured from a variety of cast iron grades, but a small aluminium manufacturing facility has also been developed. The laboratory covers routine quality assurance testing of the liner materials (metallurgical structure and mechanical properties), and material developments (new, optimised cast iron grades and optimised aluminium-silicon alloys).

David Thompson and his colleagues are using the Domino Image Analysis system from Perceptive Instruments (now superceded by the Sorcerer Image Analysis system) to back up human estimations of the proportions of certain phases in the cast iron materials in marginal cases. Sheepbridge Stokes has developed a casting process for hypereutectic aluminium-silicon alloys to manufacture cylinder liners that can operate in an engine with no need for a bore coating. At silicon levels greater than 12%, excess silicon precipitates as "primary silicon" particles, which are very hard (approx. 900 Hv), and therefore wear resistant. It is important that these primary silicon particles are evenly dispersed and of the correct size and shape.

The particle-sizing module of Domino discriminates between the round/cubic primary silicon particles and the long, needle-like eutectic silicon structure. Size and shape filters have been developed to omit the eutectic silicon particles allowing only the primary silicon particles to be counted during the measurement process by virtue of their aspect ratio and circularity. Primary silicon particle size is then measured using the "X-feret" parameter. The region corresponding to the final cylinder liner's bore wearing surface is aligned in the X-axis. Therefore, the measurements accurately reflect the particle size that will be exposed to the piston ring's surface in the engine.

Samples are cut from a liner blank and ground flat using progressively finer silicon carbide grinding papers. Polishing is then completed with progressively finer diamond suspensions, with a final 1µm diamond polish to adequately reveal the microstructure. Samples can be examined and evaluated in their as-polished condition, due to the difference in colour between the aluminium matrix material and the two types of silicon. A x20 objective is used with plain, unpolarised, reflected light, produced by a Zeiss Axiovert inverted metallurgical microscope for illumination.

Once the data, (which is composed of 40 adjacent fields) has been collected, it is output from Domino into Microsoft Excel. The spreadsheet used has been set up to list the number of particles in each 10µm size "bin". Through knowledge of the frame area and the number of frames measured, the number of particles per cm² can be calculated for any range of particle sizes required by the customer.

The Domino image analysis system in use at Sheepbridge Stokes contains various modules. Its particle sizing and filtering capabilities are being used effectively in the case study example, but its linear measurement capabilities are also in use at the same site for coating thickness measurements and the measurement of diffusion layers formed during nitriding processes. The use of an automated system is far more efficient that manual measurement. To complete the particle counting without an image analysis system, each particle would have to be measured with a graticule on the microscope's ground glass screen. Considering that typically 1000 particles are measured in 40 adjacent fields, manual measurement would be impractical. An automated system, giving objective results, can be used to compare different production batches with confidence.

Useful links

GKN - www.gknplc.com
Sorcerer - www.perceptive.co.uk/imageanalysis

Contact

David Thompson - David.Thompson@sheepbridge.gknplc.com