Abstract
The paper is a case history of the solution of a control problem in mechanical engineering. The system concerned, namely a pneumatic control system which regulated the leakage to the atmosphere of carbon dioxide through a rotating face seal, had exhibited the familiar symptoms of drift and hunting and had defied ad hoc methods to cure it. The difficulties stemmed from the fact that the gain of the element being controlled could assume indefinitely large values, whilst in addition, the loop contained one highly non-linear element.
The solution of the problem did not require highly sophisticated techniques. In contrast, a combination of frequency-response testing and the application of quite elementary servo theory enabled the causes of the malfunctioning of the system to be pin-pointed and indicated the modifications required.
Details are presented of the four phases of the investigation which comprised the initial frequency-response testing of the malfunctioning system, the theoretical analysis, laboratory tests on a simulator model of the system and final proving tests on the modified system at site.
Get full access to this article
View all access options for this article.