Acoustic Characterization of Catalyst

Abstract: The WHO has set a sound level threshold of 55 dB, beyond which noise is considered harmful to humans when exposed for an extended period of time. It is, however, estimated by the European Environmental Agency that over 110 million people in the EU are regularly exposed to levels of traffic noise above 55 dB. This work will investigate one of the constituent parts of the devices that deal with reducing noise emissions from road vehicles, specifically exhaust noise. Exhaust after-treatment systems deals, not only with filtering exhaust particulates but also with sound filtration. Their acoustic transmission properties are evermore of interest as noise pollution standards become more stringent. Catalytic converters are found in these after-treatment devices to filter NOx, CO, and HC and is used to treat exhaust gas from essentially all internal combustion engines. As exhaust particle emission standards also become more stringent, manufacturers may find themselves using more catalyst substrates in their after-treatment devices, meaning that their acoustic performance is put even further into the forefront. This project aims to model the acoustic two-ports of catalytic converters and perform model validation through transmission loss measurements. Since the catalytic converter channels are of a size where the boundary layer occupy a large proportion of the channel area, viscous effects will start to significantly affect the wave propagation through these channels. To account for this, the Kirchhoff solution for capillaries is used as the basis of the model and its validity for non-circular pipes also become of interest. Since the catalytic converters being investigated have square channels as well as channels formed from corrugation, a model of their equivalent hydraulic radius is put forward. The proposed sinusoidal approximation of corrugation is especially noteworthy since this catalyst geometry has seldomely been treated in previous works. Measurement data for model validation has been acquired with a transmission loss rig using an over-determined plane wave decomposition from three microphone measurements. Source-switching has been used to retrieve linearly independent measurements in order to solve the two-port scattering matrix. It is found that the Kirchhoff model and the proposed hydraulic radius model yield transmission loss predictions that are accurate to measurements. For catalysts with square channels, this result validates previous works that have attempted to model transmission loss through catalytic converters. Additionally, it can be concluded that the newly proposed sinusoidal model for corrugation may be used to accurately model catalysts whose channels are formed with waves between concentric liners. Finally, this result should, in the future, also be validated with the addition of a mean flow through the catalyst in order to, more closely, replicate its operating conditions.