Submitted in partial fulfillment of the academic requirements for the Masters degree in Technology, Chemistry department, Durban University of Technology, Durban, South Africa, 2015. / The focus of most lubricating grease testing has been based on performance and appearance rather than determining the concentration of chemical components. The primary reason being lubricating grease is a difficult matrix to work with from the perspective of the analytical laboratory. The purpose of this study was to develop a simple, reproducible method for elemental determination in lubricating grease and to apply the developed method to work out a flushing procedure for the filling lines in a grease manufacturing plant.
The first part of the experimental work focused on developing a suitable and efficient sample preparation technique. Three techniques were explored: direct dilution, microwave assisted acid digestion and emulsification. Direct dilution involved shear mixing the lubricating grease with metal free base oil and diluting it with an organic solvent. Use of these solvents caused plasma destabilization or even plasma extinction and their use posed health risks for laboratory personnel. Microwave digestion involved mineralising the lubricating grease using an optimised microwave assisted acid digestion procedure. In the third sample preparation technique, microemulsions were formed by mixing the lubricating grease with a relatively small volume of a strong mineral acid mixture followed by the surfactant (triton X-100) at room temperature and pressure (RTP). This approach does not require the destruction of the organic matter or the use of large amounts of organic solvents. For all three techniques the sample was subsequently analysed for Al, B, Ba, Ca, Li, Na, S, Mo and Zn.
All analysis was performed using an Optima ICP-OES with previously optimised parameters. The direct dilution method gave the most inconsistent results with relative standard deviation (RSD) as high as 56% for calcium, 79% for boron and 66% for lithium. Microwave digestion and emulsification gave comparable results, with the calibration curves of oil emulsions not differing significantly from aqueous ones. For microwave assisted acid digestion the limits of detection ranged from 0.028 mg/L for sodium to 0.255 mg/L for boron. Correlation coefficient values (r2) of all the elements were greater than 0.99. Likewise the limits of detection for emulsification ranged from 0.03 mg/L for aluminium to 0.37 mg/L for sulphur. The correlation coefficients for all the elements were greater than 0.99 and this indicates that the calibration curves were sufficient for analysing the digested grease samples.
Five quality assurance samples were analysed using both methods and in addition a t-test performed at the 99.9% confidence level and 4 degrees of freedom showed that the two sample preparation techniques gave similar results. Emulsification has several advantages over microwave digestion technique and superior analytical performance over direct sample dilution using solvents, and hence was chosen as the method of choice for routine analysis of lubricating grease.
The second part of the experimental work focused on developing a flushing procedure for filling lines in a lubricating grease plant. A flushing procedure is essential in order to minimise waste, which will in turn minimise production costs and avoids incurring disposal costs. In this series of experiments, the microwave digestion sample preparation technique was utilised for the subsequent determination of the flushing mass required in between product changes. The flushing procedure was implemented and monitored by the use of statistical quality control tools for a specified period of time, and as indicated by Shewart Control charts, the process was in statistical control.
Identifer | oai:union.ndltd.org:netd.ac.za/oai:union.ndltd.org:dut/oai:localhost:10321/1283 |
Date | January 2015 |
Creators | Marume, Cathrine |
Contributors | Govender, S., Redhi, Gyanasivan Govindsamy |
Source Sets | South African National ETD Portal |
Language | English |
Detected Language | English |
Type | Thesis |
Format | 128 p |
Page generated in 0.0017 seconds