| 4.3 Accuracy and long term stability |
|
5 Discussion
Several modifications and enhacements have been foreseen, some of which are discussed here. These further developments would be possible in the mechanism, microprocessor control system, and operating software.
Mechanism:
The mechanism periormed well with the exception of the spiral hair spring. Improving this component would undoubtedly result in considerably greater measurement accuracy. A properly machined spring would need to be custom manufactured for this purpose, as none are curtently commercially available.
Another problem that was experienced was that the chuck wobbled due to inperfections in the pin-joint between parts (F)and (G) in figure 3; this joint is necessary to enable the mechanism to be assembled. It is thought that a tapped hole in (F) and screw-like end to (G) would result in a screw joint that would substantially reduce this problem.
Mounting the entire mechanism in a rigid frame, and making provision for the test fluid beaker to be held securely beneath it, would eliminate the need for a clumsy retort stand. The experimental conditions would be easier to set up and reproduce accurately.
Microprocessor control system:
As mentioned temperature variations have a large effect on viscosity; therefore some means of temperature control would be advantageous. This function could be incorporated into the existing circuit, by using the digital to analogue converter and an additional comparator to perform analogue to digital conversion of the output voltage from a thermocouple. The processor would extract the temperature from this and compare it with an existing user-specified value. A heater immersed in the fluid and under control of the processor would maintain the liquid at a steady temperature. This modification would greatly enhance the viscometer's usefulness by making the measurements more precise; in addition, temperature-viscosity curves could be obtained automatically by the processor.
Another worthwile addition would be some form of analogue ouput, suitable for driving a chart recorder for example. Again the digital to analogue converter could be utilised, followed by a sample-and-hold amplifier so that ordinary motor speed control is not prevented.
A computer interface for PCs and Macintoshes would allow data logging and manipulation if desired. The tedium of copying by hand the 128 results of section 4.3, then typing them all into a calculator, illustrates the obvious advantages of using a computer to handle such mundane tasks. Such an interface could easily be added to the existing circuit, and need not prevent the viscometer's operation as a stand alone unit when desired.
Operating software:
Functional subroutines were developed in an extremely logical manner, which facilitates easy alterations to the operating software. In addition to the modifications required to allow the improvements discussed above, further programs could be incorporated to perform more functions, such as error analysis for example. Compensations for non-linearity or losses in the mechanism can be applied using a polynomial fit to the data, as opposed to the least squares straight line best fit that was used here. This could also be arranged to permit the testing of non-Newtonian fluids (see section 2.2).
Microprocessor control is extremely versatile and many program improvements
or application specific requirements may be implemented with ease.
| 4.3 Accuracy and long term stability |
|