June 23, 2021 at 8:31 am

Erik Kostson

Ansys Employee

So as we said including viscosity should provide attenuation - we do not have an example, but I made a benchmark for 1D duct acoustic wave propagation in castor oil, which gives the correct values for transient acoustics, but we use the same matrices for harmonic so it should be the same there (and we have tried that also and it is working ok). So it should all work well when using MP and VISC so the dynamic viscosity in engineering data in workbench as shown below.

So looking at castor oil like properties (where the attenuation relation can be described by the Stoke relation), and a 2 cycle pulse excitation of 20 000 Hz, then using the Stoke relation for an attenuation (alpha) of 1 [m-1] assumed here, we should have a dynamic viscosity of 321 at this centre frequency (see below).

The pressure over 0.1 of a meter is for 100 Pa excitation then: 100 Pa/(exp(0.1)) ~ 90 Pa at a distance of 0.1 m from the end of the duct (excitation side). That is what we get also more or less (shown below), so it captures the attenuation fairly well.

So looking at castor oil like properties (where the attenuation relation can be described by the Stoke relation), and a 2 cycle pulse excitation of 20 000 Hz, then using the Stoke relation for an attenuation (alpha) of 1 [m-1] assumed here, we should have a dynamic viscosity of 321 at this centre frequency (see below).

The pressure over 0.1 of a meter is for 100 Pa excitation then: 100 Pa/(exp(0.1)) ~ 90 Pa at a distance of 0.1 m from the end of the duct (excitation side). That is what we get also more or less (shown below), so it captures the attenuation fairly well.