Characterizing Wetting Saturation Using Centrifugal Force
The steady-state pressure drop and collection efficiency of the media depend on the local airflow pattern inside the media, which is dependent upon the local fluid saturation in the media. On the other hand, unfortunately, the local fluid saturation depends on the very same local airflow pattern in the media. This saturation-airflow interdependence makes it extremely difficult to mathematically describe the fluid mechanics of a coalescence filter. It is therefore not surprising that despite decades of research in this field, it is still impossible to predict the performance of a coalescing filter in terms of filter properties (e.g., fiber diameter, material contact angle, filter porosity, flow velocity...).
To break the saturation-airflow interdependence, we proposed to replace airflow with a more controllable and predictable centrifugal force. Centrifugal desaturation experiment was conducted using an Eppendorf 5810 centrifuge consisting of a S-4-104 rotor with a radius of 18.9 cm and a rotational speed ranging from 200 to 3900 rpm with an increment of 10 rpm. The rotor drove four free-swing buckets each having a capacity of 750 ml. As the centrifuge lid had to be closed during the centrifuge operation, we developed an image/video capturing system (Arducam 8MP IMX219 Autofocus USB Camera Module with a small light having a brightness of 15 lumens) to monitor the drainage of the fluid from the sample in real-time. The camera module had a minimum focusing distance of 40 mm which made it suitable for capturing fluid drainage in a tight enclosure (Bucket-1). Bucket-2 contained a Raspberry Pi 4 Model B powered by a 5V/3A power bank having a capacity of 15,000 mAh. Raspberry Pi was used as a small single-board computer to operate the camera module. As the Raspberry Pi could not be hardwired to an external PC when the centrifuge was in use, a remote wireless connection was established through SSH terminal to operate it and to transfer data. This enabled us to collect image/video data in real-time through the microSD card embedded in the Raspberry Pi.
The centrifugal desaturation was also compared with pressure-driven desaturation with multiphase VOF simulations in a 2D random fibrous geometry.
This allowed us to propose force-saturation curve for the first time instead of the traditional capillary pressure-saturation curve. We believe the centrifugal method can better represent the fabric’s resistance to fluid drainage as its results are not affected by the localized airflow through the media, and thereby it allows one to more directly characterize the capillarity of the media.
The centrifugal desaturation was shown to provide fluid saturation information as a function of filter microstructure only (besides the centrifugal force). This unique attribute allows one to design the material (e.g., polymer) and microstructure of a filter such that it maintains the desired level of fluid saturation when operating in different conditions (e.g., different fluids) or when used for different applications.