3 In a batch manufacturing process the viscosity of the compound increases during the reaction cycle and determines the end-point of the reaction. Samples of the compound are taken throughout the whole period of the reaction and sent to the laboratory for viscosity assessment.

The laboratory tests cannot be completed in less than three hours.

The delay during testing is a major source of under-utilization of both equipment and operators. Records have been kept of the laboratory measurements of viscosity and the power taken by the stirrer in the reactor during several operating cycles. When plotted as two separate moving mean and moving range charts this reveals the following data:

Date and time Moving mean Moving mean viscosity stirrer power 07/04 07.30 1020 21 09.30 2250 27 11.30 3240 28 13.30 4810 35 Batch completed and discharged 18.00 1230 22 21.00 2680 22 08/04 00.00 3710 28 03.00 3980 33 06.00 5980 36 Batch completed and discharged 13.00 2240 22 16.00 3320 30 19.00 3800 35 22.00 5040 31 Batch completed and discharged 09/04 04.00 1510 25 07.00 2680 27 10.00 3240 28 13.00 4220 30 16.00 5410 37 Batch completed and discharged 23.00 1880 19 10/04 02.00 3410 24 05.00 4190 26 08.00 4990 32 Batch completed and discharged Standard error of the means – viscosity – 490 Standard error of the means – stirrer power – 90 Is there a significant correlation between these two measured parameters?
If the specification for viscosity is 4500 to 6000, could the measure of stirrer power be used for effective control of the process?