| Abstract | In the process of development of a performance testing protocol for sugarcane harvesters, the need for an independent review of the literature pertaining to sugarcane loss measurement became apparent. This review was commissioned by SRDC to concentrate on the methodology of testing. The loss process is shown to be complex and the subject is one where the statistical aid to engineering investigation is so affected by this complexity that great care needs to be taken in trial procedures and analysis to ensure the validity of results. From the examination of recent detailed data, backed up by the historical information, suggestions for future development are included. The obvious facts from the literature are that: (a) the classical approach of taking measurements, in the field, of inputs and outputs of the desired clean cane is subject to the effect of variability such that the resources required for the establishment of adequate significance of differences makes the procedure almost impractical; (b) such testing in a test facility of machine processes can produce adequate results; (c) the direct method of collecting cane 'lost' does not cater for juice and minute fragment loss. This 'invisible' loss is shown to be a major consideration and its variability between treatments may invalidate comparisons using direct collection methods. A less obvious matter is the apparent reduction in the sucrose content of the desired clean cane as it passes through the harvest process. The latter consideration suggests that the measurement of the sucrose weight loss (a dry weight measure) would be of more value than the cane weight loss. Some recent trials have provided information on the sucrose weight of the input and output, and the results indicate that control of variability is superior to that obtained in the clean cane weight analysis. Future development approaches suggested are: (a) developing test facility techniques for machine process evaluation. Aim to establish the closure error; (b) examining the sucrose loss measurement. |
