تبدیل هیلبرت به عنوان یک ابزار کنترل کیفیت در فرستنده فشار خازنی

The capacitive pressure transmitters consist of an isolating diaphragm connected to a sensing line through some inner tubes which contain silicone oil. The accuracy of the sensor is determined by the efficiency of the filling process with silicone oil in the transmitter inner chambers, so that the fact of not filling completely any of the transmitter chambers can affect the dynamic response of it. In fact, the oil loss syndrome is one of the most important breakdowns, since the sensor dynamic behaviour is no longer linear. In this work the Hilbert transform is applied to detect the non-linearity in the sensor response to a sinusoidal pressure wave. Such sensor suffers the oil loss syndrome at an incipient phase. The ill sensor is compared with a linear one and the efficiency of the analysis is proved for the detection of a non-linear behaviour as opposed to the traditional methods based on Fourier analysis. At last, a non-linear model taken from literature is validated with empirical data.

The capacitive pressure transmitters, and specially the Rosemount type, are widely used in Nuclear Plants nowadays [1]. They are compound by an isolating diaphragm connected to a sensing line through some inner tubes which contain silicone oil [2]. The accuracy of the sensor is determined by the efficiency of the filling process with such silicone in the sensor inner chambers [3], so that the fact of not filling completely any of the sensor chambers can affect the dynamic response of it. So, the oil loss syndrome is one of the most noticeable sensor breakdowns and consequently, the dynamic response of it is no longer linear. Several organizations, like EPRI and NUREG have studied the syndrome to understand it and detect it [4] and [5]. In this work the sensor response to a sinusoidal pressure wave is analyzed to detect a possible non-linear behaviour. To achieve that, the instantaneous amplitudes and frequencies diagram from a linear sensor will be compared with a non-linear one. Besides, a mathematical model will be established to justify such behaviour. In this way, the analysis through the Hilbert transform is proposed as a quality control test for sensors to evaluate the silicone oil filling process.

Applying the Hilbert transform to the empiric response of a pressure sensor excited by a harmonic pressure wave, a non-linear behaviour has been detected associated to an incipient oil loss in the inner chamber of such sensor. Besides it has been proved that such incipient detection cannot be carried out with the traditional techniques via Fourier analysis. It is proposed to apply the Hilbert transform technique as a proof to verify the adequate filling process of the sensor chambers just after manufacturing, from the sensitivity of that technique to detect small oil losses. A theoretical model has been taken from literature and it explains the non-linearity associated to the oil loss and it has been used to validate the methodology based on Hilbert transform.