Formulations and numerical techniques for computation of acoustic pressure and its gradient by integral method

• Original expressions are proposed for the gradient of the radiated pressure • Efficient algorithms for acoustic calculations using integral methods are detailed • The integration techniques can handle grids in supersonic motion • The order of accuracy of the acoustic calculations is precisely quantified Starting from the Ffowcs Williams-Hawkings surface integral formulation for a moving medium, the article proposes rather simple expressions for the radiated pressure and its gradient in the time domain, that are valid for solid or porous, fixed or moving integration surfaces. Moreover, these original expressions allow calculations with integration surfaces in supersonic motion (such as rotating surfaces around propellers or rotors). Versions dedicated to fixed integration surfaces are also proposed. The usual locally compact and the fully non-compact integration techniques are recalled, with, for both, a detailed description of efficient calculation algorithms. Particular attention is devoted to the order of precision of the calculations. The time derivation techniques and integration schemes used in this study lead to a theoretical second order that can easily be increased for the locally compact integration method. The results of these expressions and integration methods are compared to the analytical solution for the case of a fixed monopole and for that of a rotating monopole. They clearly show the benefit of the direct gradient calculation over a calculation by finite differences of the pressure around the observation point, particularly for broadband signals.