## General Mechanical

#### Harmonic Acoustics_ Setting the Boundary Conditions at Water-Air Interface

• Saiful
Subscriber

Hello Folks,

I am a new learner of Harmonic Acoustics. Recently I am modelling a project for acoustic propagation in an aquarium of a rectangular shape whose top is opened to the atmospheric air. I have a sound source at exactly the middle of the tank. I can model the source as a monopole acoustic pressure source. The source produces 76 dB (dB re 20 uPa) at air (calibrated). I want to simulate the sound propagation inside the aquarium. My understanding so far is that for the five glass surfaces I can treat them as hard boundary (default in ANSYS as well) since they are smooth and good reflector of sound. Can anyone help me how can I select the boundary condition at the open surface? The open surface is the water/air interface and they have two dissimilar impedances. How can I model such boundaries?

• Erik Kostson
Ansys Employee
Hi

There is a big specific impedance mismatch between air and water (since it is impedance~rho-c_speedofsound), so the acoustic waves from/in the water will be reflected there at the interface (watre/air) since the reflection coefficient is close to 1, and the same if the waves come from the air into water.
So the water is like a waveguide and the acoustic energy stays inside there .

Now the boundary condition (BC). So i think a zero acoustic pressure (p=0), so sound soft BC, should be OK there at the air/water interface (otherwise we can model the air and the water as acoustic media).
All the best

Erik
• Saiful
Subscriber
Hello Erik Thank you for your answer. Yes, there is a huge impedance mismatch. However, can you tell me how can I define a soft boundary at the top? I was thinking to define the top as an impedance boundary with resistance (343*1.21 = 415 Pa.s/m) and zero reactance. or do I need to use the impedance of water (1481*1000 Pa.s/m)?
Any thoughts on it?

• Erik Kostson
Ansys Employee
HI just use the pressure BC in Harmonic acoustic and set it to 0 so p=0 at the interface.

Erik