Metodologia experimental para avaliação do desempenho de válvulas de segurança para ar comprimido e contrapressão atmosférica

Abstract

The safety of devices and equipment if the pressure is associated with the correct sizing of a safety valve. For this it is necessary to evaluate its performance. This work developed and tests methodologies for flow testing, seeking to follow the main existing standards, in addition to developing a simplified mathematical model to estimate the elevation and flow of safety valves that operate with compressed air and atmospheric backpressure. Four different methodologies were developed to control and carry out flow capacity testing: I) Directly powered and controlled by the compressor operation; II) Manually controlled from a control valve; III) Automatically controlled with pressure set point; and IV) Automatically controlled with flow setpoint. These methodologies were tested on two different valves. Additionally, a mathematical model was developed based on the theory of compressible flow in a convergent nozzle. Their results were compared with three different valves, in terms of disc lift and discharge flow. As for the methodologies, it was observed that Methodology I presents the greatest flow variability due to compressor cycles, as Methodologies II, III and IV have very close values to each other, with maximum deviation of 2.03% and 5.11% in relation to the average of the three methodologies, for the two tested valves. Of these, the shortest execution time of Methodology III stands out, with the IV reaching lower flow variability. To validate the experimental apparatus and Methodology III, a valve certified in accordance with the standards of the American Society of Mechanical Engineers (ASME) was tested and obtained a flow rate 0.95% above the flow stamped on the valve. The mathematical model, on the other hand, presents a difference between the estimated disc elevation and the experimental elevation of at most 3.94% and at least 0.60%. For the flow, with the mathematical model, a maximum difference of 7.36% and a minimum of 2.09% in relation to the experimental one was obtained.