Acousto-chemical analysis in multi-transducer sonochemical reactors for biodiesel production

Abstract

Biodiesel is a powerful alternative fuel that is less polluting and problematic to produce and implement. The production process of biodiesel also gives us the byproduct glycerol, which is a useful feedstock to produce hydrogen and syngas as fuels. With such high value as a fuel we are in need of better production technologies for biodiesel, which is currently being pursued through sonochemical reactors. The development of continuous sonochemical reactors for biodiesel production is a crucial requirement for the biofuel industry. Sonochemical reactors make use of ultrasound and acoustic cavitation to produce biodiesel from waste cooking oils (WCO). In this work we carried out both numerical simulation and experimental analysis of sonochemical reactors with multiple transducers. Through simulation, the effect of double vs a single transducer has been tested for a continuous flow sonochemical reactor. In experimental work three different cases with different ultrasound systems (bath, probe and bath+probe) have been tested. In both the studies, acoustic pressure and biodiesel conversion are analyzed. Results for the simulation show that in shorter reactors, the high cavitation from two transducers dampens the acoustic pressures leading to low conversion. However, at taller heights the effect of combined cavitation is less severe and the acoustic pressure and biodiesel yield are very similar between the designs having single and double transducers. From experiments it was found that the biodiesel conversion depends on several acoustic conditions mainly cavitation. A meticulous and insightful analysis was made to understand the difference in bath type and probe type ultrasound systems on acoustic pressure and biodiesel conversion.

Multi-frequency sonoreactor characterisation in the frequency domain using a semi-empirical bubbly liquid model

“…By linearising the system of differential equations involved for the Caflisch model, Commander and Prosperetti [23] derived an analytical expression for a complex wavenumber to describe the behaviour of bubbly liquids in the frequency domain. This model had regained popularity in the past decade to describe the attenuation in ultrasonic systems [13] , [48] . The complex wavenumber squared for the monodisperse Commander and Prosperetti model can be expressed as: …”

“…The simpler approach assumes a constant value of for the entire domain [49] , [53] while the other involves specifying as a spatial function of pressure [18] . In this study, the latter was adapted with reference to past works [13] , [18] , [48] as it was believed to be more comprehensive. The method assumes a linear relationship between the bubble fraction and the local absolute pressure as: …”

“…Frequency domain modelling is commonly used due its ease of implementation, lower resource requirements and being much easier to solve then their time-dependant counterparts. This is evident in the large number of recent studies where frequency domain simulations were used to design and optimise ultrasonic equipment [10] , [11] , [12] , [13] . The specification of wave attenuation is also simpler in frequency domain models.…”

“…Current simulations can be categorised into those that consider the propagation medium as a pure liquid [12] , [14] , [15] , [16] , [17] , and those that consider the effects of bubbles in the medium [13] , [18] , [19] , [20] , [21] , [22] . Bubbly-liquid models gained popularity in recent times as it is believed to provide a better representation of the fluid property in ultrasonic systems.…”

Numerical modeling and verification of a sonobioreactor and its application on two model microorganisms

“…The 3D-mode shapes of the whole setup provided a clear picture of the pressure distribution inside the simulated setup. To be useful, mode shapes should have the maximum pressure along the center of the reactor, whereas the walls of the reactor have lower pressure [37,38]. Mode shapes were evaluated for each peak in the specific frequency range to decide which one was useful.…”