Acoustics User Guide

SIEMENS
Acoustics User’s
Guide
Contents
Proprietary & Restricted Rights Notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
List of symbols . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
Overview of acoustic capabilities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
Assumptions and limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1-1
Theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Overview of acoustic theory . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1
Boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Imposed pressure (Dirichlet boundary condition) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Imposed normal velocity (Neumann boundary condition) . . . . . . . . . . . . . . . . . . . . . . . . 2-3
Imposed impedance or admittance (Robin boundary condition) . . . . . . . . . . . . . . . . . . . . 2-3
Open or radiating boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4
Finite element formulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-4
Frequency-dependent acoustic materials or absorbers . . . . . . . . . . . . . . . . . . . . . . . . . . 2-5
Coupling of fluid and structure (vibro-acoustics) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-7
Modal participation factors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-9
Output forms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
RMS pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
SPL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
Frequency weighted SPL . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-12
Defining the model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1

Defining the fluid volume . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
Defining porous or equivalent porous materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1
Defining spatially varying fluid properties in acoustic analysis . . . . . . . . . . . . . . . . . . . . . . . . 3-6
Defining boundary conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
Rigid wall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
Enforced pressure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
Defining the fluid-structure interface boundary condition . . . . . . . . . . . . . . . . . . . . . . . . . 3-8
Defining acoustic panel normal velocity and transfer admittance . . . . . . . . . . . . . . . . . . 3-11
Duct modes boundary condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-12
Coupled FEMAO . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-17
Impedance or admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-18
Defining infinite boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20
Defining acoustic loads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-20
Pre-computed enforced vibrations at fluid-structure interface . . . . . . . . . . . . . . . . . . . . . 3-20
Enforced acoustic pressure with complex data input . . . . . . . . . . . . . . . . . . . . . . . . . . 3-21
Fan noise boundary condition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-21
Acoustic source loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-23
Acoustics User’s Guide 3

Contents
Contents
Monopole, dipole, plane wave, and surface dipole acoustic sources . . . . . . . . . . . . . . . . 3-24
External loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-29
Defining microphone meshes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-30
Solution methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1

Overview of solution methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1
Input data requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1

Executive control commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Case control commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Acoustic load, boundary condition, and source control . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Import coupling or the area matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Fluid-structure interaction control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1
Random acoustic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4
Output requests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-5
Bulk entries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-13
Acoustic grids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-13
Acoustic elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14
Acoustic material definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14
Acoustic panel normal velocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14
Transfer admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-14
Impedance or admittance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15
Acoustic loads, boundary condition, and sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15
Modeling structure-acoustic interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-15
Acoustic and vibro-acoustic analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1

Overview of acoustic and vibro-acoustic analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1
Defining the fluid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1
Absorber elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Acoustic loading . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Single-point constraints . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Output requests (overview) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-2
Vibro-acoustic coupling punch output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Vibro-acoustic coupling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Superelement analysis for vibro-acoustic analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Running the job . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-3
Loudspeaker example (simplified version) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4
Acoustic analysis using automatically matched layer . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1

Exterior acoustics using automatically matched layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1
Assumptions and limitations of AML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2
Convex AML surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3
AML faces and infinite plane . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4
AML with non-homogeneous elements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6
AML and fluid damping coefficient . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6
Microphone mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-6
Acoustic computation process . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-10
Gearbox example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-11
4 Acoustics User’s Guide

Contents
FEM adaptive order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1

Finite Element Method Adaptive Order (FEMAO) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
Transfer vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1

Acoustic Transfer Vector (ATV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1
Vibro-Acoustic Transfer Vector (VATV) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-2
Other considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1

Superelements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1
Figures
2-1. Interior Fluid Domain and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1

3-1. Bounding box . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-10
3-2. Acoustic monopole source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-24
3-3. Acoustic dipole source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-26
3-4. Dipole sound field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-26
3-5. Acoustic plane wave source . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-27
5-1. Coupled and uncoupled structural faces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-3
5-2. Panel test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-13
6-1. Acoustic suspension loudspeaker . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-4
7-1. Convex mesh/surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1
7-2. Gearbox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-2
7-3. Convex AML surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-3
7-4. Non-convex AML surface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-4
7-5. AML faces (3) that contact the infinite plane are perpendicular . . . . . . . . . . . . . . . . . . . 7-5
7-6. AML faces (3) that contact the infinite plane are not perpendicular. . . . . . . . . . . . . . . . . 7-5
7-7. AML faces (3) that contact the infinite plane are not perpendicular and the entire reflectionless
artificial layer is not in contact with the infinite plane. . . . . . . . . . . . . . . . . . . . . . . . . . 7-5
7-8. Microphone mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-7
7-9. Acoustic radiation from a machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-8
7-12. Anechoic duct . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
7-13. Open duct at the right end . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
7-14. Open duct at the right end and ends in a baffle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-9
7-15. Open duct at the right end and top duct ends in a baffle . . . . . . . . . . . . . . . . . . . . . . 7-10
7-16. Acoustic radiation from a vibrating gearbox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-11
7-17. Gearbox . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-12
7-18. Microphone mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-13
8-1. f = 100Hz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
8-2. f = 1000Hz . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1
8-3. Shape function types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-4
8-4. Computation by FEMAO as a function of time and frequency . . . . . . . . . . . . . . . . . . . . 8-5
8-5. FEM mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6
8-6. FEMAO mesh . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6

Contents
Contents
8-7. FEMAO mesh with local refinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6

Proprietary & Restricted Rights Notice
© 2019 Siemens Product Lifecycle Management Software Inc. All Rights Reserved.
This software and related documentation are proprietary to Siemens Product Lifecycle Management
Software Inc. Siemens and the Siemens logo are registered trademarks of Siemens AG. Simcenter
3D is a trademark or registered trademark of Siemens Product Lifecycle Management Software Inc.
or its subsidiaries in the United States and in other countries.
NASTRAN is a registered trademark of the National Aeronautics and Space Administration.
Simcenter Nastran is an enhanced proprietary version developed and maintained by Siemens
Product Lifecycle Management Software Inc.
MSC is a registered trademark of MSC.Software Corporation. MSC.Nastran and MSC.Patran are
trademarks of MSC.Software Corporation.
All other trademarks are the property of their respective owners.
TAUCS Copyright and License
TAUCS Version 2.0, November 29, 2001. Copyright (c) 2001, 2002, 2003 by Sivan Toledo, Tel-Aviv
University, [email protected]. All Rights Reserved.
TAUCS License:
Your use or distribution of TAUCS or any derivative code implies that you agree to this License.
THIS MATERIAL IS PROVIDED AS IS, WITH ABSOLUTELY NO WARRANTY EXPRESSED OR
IMPLIED. ANY USE IS AT YOUR OWN RISK.
Permission is hereby granted to use or copy this program, provided that the Copyright, this License,
and the Availability of the original version is retained on all copies. User documentation of any code
that uses this code or any derivative code must cite the Copyright, this License, the Availability note,
and "Used by permission." If this code or any derivative code is accessible from within MATLAB, then
typing "help taucs" must cite the Copyright, and "type taucs" must also cite this License and the
Availability note. Permission to modify the code and to distribute modified code is granted, provided
the Copyright, this License, and the Availability note are retained, and a notice that the code was
modified is included. This software is provided to you free of charge.
Availability (TAUCS)
As of version 2.1, we distribute the code in 4 formats: zip and tarred-gzipped (tgz), with or without
binaries for external libraries. The bundled external libraries should allow you to build the test
programs on Linux, Windows, and MacOS X without installing additional software. We recommend
that you download the full distributions, and then perhaps replace the bundled libraries by higher
performance ones (e.g., with a BLAS library that is specifically optimized for your machine). If you
want to conserve bandwidth and you want to install the required libraries yourself, download the
lean distributions. The zip and tgz files are identical, except that on Linux, Unix, and MacOS,
unpacking the tgz file ensures that the configure script is marked as executable (unpack with tar
zxvpf), otherwise you will have to change its permissions manually.

Proprietary
Proprietary & Restricted
& Restricted Rights
Rights Notice
Notice
HDF5 (Hierarchical Data Format 5) Software Library and Utilities Copyright 2006-2016 by
The HDF Group
NCSA HDF5 (Hierarchical Data Format 5) Software Library and Utilities Copyright 1998-2006 by the
Board of Trustees of the University of Illinois. All rights reserved.
Redistribution and use in source and binary forms, with or without modification, are permitted for any
purpose (including commercial purposes) provided that the following conditions are met:
1. Redistributions of source code must retain the above copyright notice, this list of conditions,
and the following disclaimer.
2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions,
and the following disclaimer in the documentation and/or materials provided with the distribution.
3. In addition, redistributions of modified forms of the source or binary code must carry prominent
notices stating that the original code was changed and the date of the change.
4. All publications or advertising materials mentioning features or use of this software are asked,
but not required, to acknowledge that it was developed by The HDF Group and by the National
Center for Supercomputing Applications at the University of Illinois at Urbana-Champaign and
credit the contributors.
5. Neither the name of The HDF Group, the name of the University, nor the name of any Contributor
may be used to endorse or promote products derived from this software without specific prior
written permission from The HDF Group, the University, or the Contributor, respectively.
DISCLAIMER: THIS SOFTWARE IS PROVIDED BY THE HDF GROUP AND THE CONTRIBUTORS
"AS IS" WITH NO WARRANTY OF ANY KIND, EITHER EXPRESSED OR IMPLIED.
In no event shall The HDF Group or the Contributors be liable for any damages suffered by the users
arising out of the use of this software, even if advised of the possibility of such damage

List of symbols
Below is a list of the symbols used in this user's guide. Other symbols are defined where they appear.
Letter symbols Description

Specific admittance
Area or coupling matrix
Area or coupling matrix
Damping matrix
Bulk modulus of air
Damping or the impedance matrix of the fluid
Viscous damper
Damping matrix of the structure
Coefficient
Speed of sound
Speed of sound in ambient medium
Frequency
Mass density correction factor involving fluid viscous properties in the
Johnson-Champoux-Allard model
Bulk modulus correction factor involving fluid thermal properties in the
Johnson-Champoux-Allard model
Acoustic material damping
Equivalent frequency-dependent damping

Stiffness matrix
Stiffness matrix or inverse mass matrix of the fluid
Grounded Spring
Stiffness matrix of the structure
Structure factor or square of tortuosity

Wave number
Mass matrix

List
List of of symbols
symbols
Mass matrix or compressibility matrix of the fluid
Mass
Mass matrix of the structure
Shape functions for acoustic pressure
Shape functions for the structure
Outward normal to the fluid surface
Outward normal to the structural surface

Loading vector
Power
Prandtl number
Acoustic pressure
Acoustic volume velocity
Strength of acoustic source

Velocity potential
Static viscous permeability
Static thermal permeability

Unit radial vector
Gas constant of the acoustic fluid
Surface
Monopole amplitude
Displacement
Vector of fluid displacements
Acoustic particle velocity
Vector of fluid displacements normal to the surface with outward normal
Vector of structural displacements
Accelerations
Volume
Confidence level of the material

List of symbols
Frequency-dependent reactance
Source coordinates
Spatial location
Specific acoustic impedance
Greek letters Description
Tortuosity
Bulk modulus
Equivalent bulk modulus of the porous material
Real and imaginary parts of bulk modulus of the porous materials
Bulk modulus of ambient medium

Ratio of specific heats (isentropic expansion factor)
Kronecker’s delta
Dynamic viscosity
Viscous characteristic length
Thermal characteristic length
Kinematic viscosity
Thermal diffusivity
Acoustic fluid density
Equivalent density of the porous material
Real and imaginary parts of equivalent density of the porous materials
Frequency-dependent fluid density
Density of ambient medium
Speed of sound in air

Flow resistivity
Frequency-dependent flow resistivity
Vector of fluid modes
Vector of structural modes
Porosity
Mathematical
Description
operators

List
List of of symbols
symbols
Divergence operator
Gradient operator
Subscripts Description
State of the ambient medium at a given temperature (not to be confused with 0
degree temperature)
Equivalent
Fluid
Imaginary
Normal
Real
Structural

Chapter 1: Introduction
1.1 Overview of acoustic capabilities

Simcenter Nastran can be used for both interior and exterior acoustics. It allows you to perform a
pure acoustic analysis or a coupled vibro-acoustic analysis.
1.2 Assumptions and limitations

• Analysis is limited to linear acoustics.
• The current acoustic formulations in Simcenter Nastran assume that the medium is quiescent;
that is, no mean flow of the ambient medium exists.
Acoustics User’s Guide 1-1

Chapter 2: Theory
2.1 Overview of acoustic theory
Figure 2-1. Interior Fluid Domain and Boundary Conditions

In Simcenter Nastran, the acoustic equations are based on
• Small linear perturbations with negligible convective momentum terms.
• Linear pressure-density relationship.
Thus Euler's momentum equation is
Equation 2-1.
and the mass continuity equation is
Equation 2-2.
where:
• is the density of the ambient medium.
• is the acoustic particle velocity.

Chapter
Chapter 2: 2: Theory
Theory
• is the acoustic pressure.
• Qs is the volume velocity defined on a unit volume.
• δ is the Kronecker’s delta function defined such that the value of δ(x-xs) is zero everywhere
except at location defined by xs where the value is unity.
Under linear pressure variation,
Equation 2-3.
where is the speed of sound in the ambient medium. Substituting Equation 3 in Equation 2 gives
Equation 2-4.
with the compressibility or the bulk modulus β defined as
Equation 2-5.
Taking the time derivative of Equation 4 and rearranging some terms yields
Equation 2-6.
Similarly, using the Ñ operator on Equation 1 yields
Equation 2-7.
or
Equation 2-8.
2-2 Acoustics User’s Guide

Theory
2.2 Boundary conditions

You can specify the following boundary conditions for bounded and unbounded fluid domains.
2.2.1 Imposed pressure (Dirichlet boundary condition)
On surface S1, you can specify an enforced pressure
2.2.2 Imposed normal velocity (Neumann boundary condition)

On surface S2, you can define normal velocities as follows:
Equation 2-9.
where:
• is the outward normal.
• is fluid accelerations (positive along the normal direction).
On rigid boundaries, the velocity and hence acceleration is zero. Therefore, Equation 9 will reduce to
pressure gradient being zero.
2.2.3 Imposed impedance or admittance (Robin boundary condition)

On surface S3, you can define an impedance or admittance boundary condition as follows:
Equation 2-10.
where:
• is the specific acoustic impedance.
• is specific admittance (reciprocal of impedance).
Because of equation (see Overview of acoustic theory), Equation 10 becomes
Equation 2-11.

Chapter
Theory
2.2.4 Open or radiating boundary

An infinite boundary condition requires that the acoustic field vanishes at points farther away from
the source. The far field radiation boundary condition is specified using the Sommerfeld boundary
condition.
Equation 2-12.
An equivalent statement of Sommerfeld radiation condition is
Equation 2-13.
Thus, the boundary condition at infinity can be replaced by an impedance boundary condition (see
Impedance or admittance). This can be accomplished by modeling a fluid region that surrounds the
vibrating surface and by applying the impedance boundary condition on the free surface. The number
of fluid elements that need to be modeled to approximate infinity depends on the frequency range of
interest. At higher frequencies, more fluid exterior to the structure must be modeled. The impedance
boundary condition can then be applied using the CAABSF elements.
The Automatically Matched Layer (AML) boundary condition allows defining Equation 12 more
accurately.
2.3 Finite element formulation
Using Galerkin weighted residuals, expressing , and using as the weighting

function to minimize the residuals, gives
Equation 2-14.
Because of Gauss divergence theorem, Equation 14 can be expressed as follows:
Equation 2-15.
Equation 15 can be further simplified by substituting equation (see Imposed normal

velocity (Neumann boundary condition)) and equation

Theory
(see Imposed impedance or admittance (Robin boundary condition)) as
Equation 2-16.
Given that , Equation 16 reduces to
Equation 2-17.
where the Compressibility matrix is
Equation 2-18.
Inverse mass or mobility matrix
Area matrix
Impedance matrix
and Acoustic source vector
Note: The contribution to the damping matrix is due to the impedance boundary condition.
2.3.1 Frequency-dependent acoustic materials or absorbers

Acoustic absorbers can be characterized by a fluid that has frequency-dependent acoustic
characteristics. For such a case, the density and bulk modulus are expressed in terms of equivalent
density and equivalent bulk modulus which are complex quantities as follows:

Chapter
Theory
Equation 2-19.
where:
• is the equivalent frequency-dependent density.
• is the equivalent frequency-dependent bulk modulus.
Then,
Equation 2-20.
where defines the equivalent of structural damping constant for the fluid. The density and can
then be determined in terms of and as follows:
Equation 2-21.
Similarly,
Equation 2-22.
where
Equation 2-23.
Substituting Equation 20 and Equation 22 into equations
and

Theory
(see Finite element formulation) results in the following for the absorber
Equation 2-24.
Defining and such that
Equation 2-25.
Equation 24 can be re-written as
Equation 2-26.
Note: The imaginary part of results in where
Equation 2-27.
2.3.2 Coupling of fluid and structure (vibro-acoustics)

In a coupled vibro-acoustic analysis, the fluid pressure on the structure boundary causes surface
tractions on the structure
Equation 2-28.
where NS is a shape function for the structure,

Chapter
Theory
Equation 2-29.
and is the outward normal to the structure at the fluid boundary interface. Substituting Equation 29
in Equation 28 gives
Equation 2-30.
The structure equations of motion then becomes
Equation 2-31.
where
Equation 2-32.
and are structure and fluid shape functions, and =- .

At the fluid structure interface
Equation 2-33.
Equation (see Finite element formulation) then

becomes
Equation 2-34.
The combined fluid structure equations then become
Equation 2-35.
where:
• M is the mass matrix.

Theory
• B is the damping matrix.
• K is the stiffness matrix.
• P is the loading vector.
• Subscript s represents the partitions of the structure.
• Subscript f represents the fluid degrees of freedom.
• Degrees of freedom are displacements, u for the structure and pressure.
• Degrees of freedom are displacements, p for the fluid.
• [A] matrix is the coupling between the fluid and structure degrees of freedom at the wetted
interface.
Note, Equation 35 is unsymmetric. This equation is symmetrized by using a variable transformation

as follows:
Let a velocity potential q be defined such that
Equation 2-36.
Substituting Equation 36 in Equation 35, assuming harmonic time dependence, and then integrating
the fluid equation in time gives
Equation 2-37.
Also, because of equation (see Overview of acoustic theory), it follows that
Equation 2-38.
The pressure and the velocities can therefore be recovered from Equation 36 and Equation 38
respectively.
2.3.3 Modal participation factors

The modal solution for the displacement and the pressures can be written in the form

Chapter
Theory
Equation 2-39.
where:
• are the uncoupled, undamped structural mode shapes or eigenvectors.
• are the uncoupled, undamped rigid-walled acoustic modes.
• The vector is the modal amplitude or the modal basis function for the structure.
• The vector is the modal amplitude or the modal basis function for the fluid.
The eigenvectors or the mode shapes allow the transformation from physical space to the modal
space. Substituting Equation 39 in unsymmetrical equation
(see Finite element formulation) results in
Equation 2-40.
or
Equation 2-41.
where:
Equation 2-42.
Assuming a harmonic solution of the form , Equation 41 for the fluid degrees of freedom can be
reduced to

Theory
Equation 2-43.
which in turn can be further rewritten as
Equation 2-44.
where
Equation 2-45.
The fluid modal contribution at the grids is therefore,
Equation 2-46.
where is a diagonal matrix such that

The structure modal contribution for the response is given by
Equation 2-47.
where is a diagonal matrix such that

The fluid structure panel participation is then
Equation 2-48.
and the fluid structure grid participation factor is then
Equation 2-49.
where consists of columns of grids for panel .

Chapter
Theory
2.4 Output forms

Several forms of acoustic pressure are computed in an acoustic or vibro-acoustic solution.
2.4.1 RMS pressure

The mean square and the root mean square (RMS) values of acoustic pressure are defined by
Equation 2-50.
where is the pressure. For single frequency pressure, Equation 50 becomes
Equation 2-51.
where is the peak pressure at a given frequency.
2.4.2 SPL
Generally, sound pressure is expressed on a logarithmic scale as the Sound Pressure Level (SPL) in
decibels (dB). The SPL is defined by
Equation 2-52.
where is the reference pressure. For example, the reference pressure for air is
2.4.3 Frequency weighted SPL

Psychoacoustic studies have determined that the perception of loudness in the human ear is
frequency dependent. The SPL equation treats the pressure at all frequencies with the same
weighting. To compensate for the sensory perception of noise, a transformation function
Equation 2-53.

Theory
is used, where W is the weighting function. The most commonly used weighting function is the
A-weighting.

Chapter 3: Defining the model
3.1 Defining the fluid volume

The finite element representation of acoustic elements is similar to that in conventional structural
finite elements, but differs from the structural elements in the following ways:
• The fluid grid points referenced by acoustic elements must have a value of -1 for the CD field
(field 7) of the GRID card. These grids then constitute a set of fluid grids.
• The PSOLID physical property that is referenced by acoustic elements must specify PFLUID
for the FCTN field (field 8).
• The PSOLID physical property that is referenced by acoustic elements must in turn reference
a fluid material (MAT10, MAT10C, MATF10C, or MATPOR).
The following element types are supported:

• 3 D: CHEXA, CTETRA, CPENTA, and CPYRAM for 3-D elements. These elements must
reference either a PSOLID or a PMIC physical property.
• 2 D: CTRIA3 and CQUAD4. These elements must reference only a PMIC physical property.
• 1 D: CROD. These elements must reference only a PMIC physical property.
• 0 D: MICPNT.
3.2 Defining porous or equivalent porous materials

Porous materials are often used to dampen acoustic wave propagation. These materials typically
consist of a solid skeleton with pores or interstices through which acoustic waves can propagate. A
fully coupled vibro-acoustic solution is then needed to model the poro-elastic material. The limiting
case of poro-elastic materials is the case where the elastic skeleton is either completely rigid or limp
(almost zero rigidity). Such materials are characterized by various material models. Simcenter
Nastran supports the following porous material models:
• MAT10:
You can define acoustic absorber properties using the MAT10 material definition. The values
of ρS(ω), βS(ω), Ge(ω) and γ(ω) as described in Frequency-dependent acoustic materials or
absorbers are specified for RHO, BULK, GE, and GAMMA fields in the MAT10 (see MAT10 in
the Simcenter Nastran Quick Reference Guide ). The frequency-dependent tables are then
specified in the continuation line.
• MAT10C and MATF10C:

Chapter
Chapter 3: 3: Defining
Defining the model
the model
You can also define acoustic absorber properties using the MAT10C or MATF10C material
definition. These acoustic material bulk entries let you input complex density and speed of
sound for SOL 108 and SOL 111.
o MAT10C allows you to define constant or nominal properties.
o MATF10C allows you to define material properties in tabular format.
For example, in a SOL 111 solution, when a modal analysis is first performed, the solver uses the
nominal value to calculate the eigenvalue and eigenvector. In the frequency response, however,
the actual value is used if MATF10C is specified.
• MATPOR:
The MATPOR bulk entry allows you to specify the material models directly, and lets the solver
compute the equivalent density and speed of sound.
With the MATPOR bulk entry, you can select the Craggs, Delaney-Bazely/Miki, and
Johnson-Champoux-Allard empirical model and enter the relevant parameters.
o Craggs
The wave equation is given by
Equation 3-1.
where:
Equation 3-2.
where:
■ σ is the flow resistivity.
■ is the structure factor or square of tortuosity.
■ is the tortuosity.
■ is the porosity.
Then
Equation 3-3.
o Delaney-Bazely/Miki

Defining the model
The wave propagation is constant and the speed of sound is given by the following empirical
relationship.
Equation 3-4.
Equation 3-5.
Equation 3-6.
where:
■ k is the wavenumber.
■ is the speed of sound in air.
■ is the characteristic impedance.
■ is the density.
■ σ is the flow resistivity.
Table 3-1. Coefficients for Delaney-Bazely and Miki
Coefficient Delaney-Bazely and Miki
0.122
-0.618
0.180
-0.618
0.0785
-0.632

Chapter
Defining the model
the model
Table 3-1. Coefficients for Delaney-Bazely and Miki
Coefficient Delaney-Bazely and Miki
0.120
-0.632
The values of and can then be computed from equations
(see Frequency-dependent acoustic materials or absorbers)

and Equation 6. Similarly, the real and imaginary values of bulk modulus can be computed
using equations
(see Frequency-dependent acoustic materials or absorbers), Equation 4, and Equation 5.
o Johnson-Champoux-Allard (JCA)
The equivalent density and bulk modulus are as follows:
Equation 3-7.
Equation 3-8.
where:
Equation 3-9.

Defining the model
Equation 3-10.
Equation 3-11.
Equation 3-12.
where:
■ is the porosity.
■ is the flow resistivity.
■ is the tortuosity.
■ is the viscous characteristic length.
■ is the thermal characteristic length.
■ is the static viscous permeability.
■ is the dynamic viscosity.
■ is the speed of sound in air.
■ is the bulk modulus of air.
■ is the kinematic viscosity.
■ is the ratio of specific heats.
■ is the thermal diffusivity.
■ is the static thermal permeability.
■ is the Prandtl number.

Chapter
Defining the model
the model
■ is the mass density correction factor involving fluid viscous properties in the
JCA model.
■ is the bulk modulus correction factor involving fluid thermal properties in the
JCA model.
You can use MATPOR bulk entries in combination with CHEXA, CPENTA, CPYRAM, and
CTETRA elements that reference PSOLID property bulk entries.
For more information on MATPOR bulk entry, see MATPOR bulk entry in the Simcenter Nastran
Quick Reference Guide.
3.3 Defining spatially varying fluid properties in acoustic analysis

You can use results from a computational fluid dynamics (CFD) general notation system (CGNS) file
to define the fluid properties in acoustic radiation and propagation problems. By doing so, you can
account for spatial variation in the properties of the fluid across the mesh. This allows you to more
accurately model acoustic applications where significant thermal gradients exist. HVAC systems,
exhaust systems, and gas turbines are examples of such applications.
This capability is applicable to SOL 108 and 111 acoustic analysis with standard FE and
adaptive-order FEMAO meshes.
When you use the results from a CGNS file to define the fluid properties for an acoustic analysis, the
software does the following:
1. Reads the property data from a CGNS format file.

The CGNS format file is typically generated from a CFD mean flow simulation, and the results in
the CGNS file are given at either the grid point locations or element centroid locations of the CFD
mesh. The software can read the following results from the CGNS file:
• Temperature only
• Speed of sound only
• Temperature and mass density
• Temperature and pressure
• Speed of sound and mass density
• Speed of sound and pressure
When the software reads a single property from the CGNS file, such as temperature only, it uses
a constant value for the pressure that you must specify.
2. Using a maximum distance algorithm, maps the property data in the CGNS file to the grid points
of the acoustic mesh. The mapped property data is stored in a SCH5 file.

Defining the model
Note
You cannot modify the parameters that control the maximum distance algorithm from
within Simcenter Nastran.
3. If necessary, converts the property data at the grid point locations to speed of sound and mass
density values.
4. Using the shape functions for the elements in the acoustic mesh, interpolates the values for
speed of sound and mass density at the grid points to the Gauss points of the elements.
During the acoustic analysis solve, the software uses the mapped fluid properties. Thus, when the
software performs the integrations to obtain the element matrices during acoustic analysis solve, it
accounts for spatial variation in the fluid properties across each element.
Converting properties to speed of sound and mass density

The software uses speed of sound and mass density values when it calculates the element matrices.
Thus, if the software reads temperature or pressure data from the CGNS file, the data must be
converted to speed of sound and mass density data.
The software uses the following equation to convert temperature data to speed of sound data:
where c is the speed of sound, γ is the isentropic expansion factor, R is the gas constant, and T is
the (absolute) temperature.
The software uses the following equation to convert pressure data to mass density data:
where ρ is the mass density, γ is the isentropic expansion factor, p is the (absolute) pressure, and c is
the speed of sound.
Input file requirements

You use the data in a CGNS file to define the fluid properties for an acoustic analysis by adding the
new ACTEMP case control command and bulk entry to the SOL 108 or 111 input file. Because an
ACTEMP case control command and bulk entry are present, the software uses the material properties
it obtains from the ACTEMP bulk entry specification.
• Use the ACTEMP case control command to select the ACTEMP bulk entry from which the
software accesses the fluid property data. The ACTEMP case control command must be placed
above the subcases.
• Use the ACTEMP bulk entry to specify the following:

o The gas constant and isentropic expansion factor for the fluid, and, if applicable, the constant
pressure.
o The combination of properties to retrieve from the CGNS file, such as temperature and
pressure, speed of sound and mass density, and so on.

Chapter
Defining the model
the model
o The unit number of the SCH5 file.

The SCH5 file stores the CFD results that are mapped to the acoustic mesh.
o The IDs of the data tables in the CGNS file that you want to access.
• On the PSOLID bulk entry that the acoustic elements reference, in the FCTN field, specify
PFLUIDEX.
• In the File Management section, use an ASSIGN statement to associate the SCH5 file with
the unit number.
3.4 Defining boundary conditions
3.4.1 Rigid wall

For a rigid wall, no flow occurs across the boundary and hence from Euler’s equation the pressure
gradient is zero. The surface integral on the rigid boundary therefore vanishes by virtue of equation
(see Coupling of fluid and structure (vibro-acoustics)).

This is also called the natural boundary condition. This is analogous to a traction free boundary
condition, which is the natural boundary condition for the structural problem. No boundary condition
needs to be defined for a rigid wall boundary condition.
3.4.2 Enforced pressure
Single-point constraints of the fluid ( = 0) may be enforced on the fluid boundary using the SPC
entry with degree of freedom component set to 1. Non-zero enforced pressures can be applied
directly using SPCD and RLOADi entries with the TYPE entry for the RLOADi card set to DISP.
3.4.3 Defining the fluid-structure interface boundary condition

The vibro-acoustic coupling interface can be defined in one of two ways:
• Two-way (strong) coupling — That is, the vibration of the structure excites the fluid which in turn
causes pressure loading on the structure (see
in Coupling of fluid and structure (vibro-acoustics)).
• One-way (weak) coupling — Here, the effect of the fluid on the structure is assumed to be
negligible. However, the vibration of the structure on the fluid is assumed to be significant.

Defining the model
Two-way coupling
The coupled fluid-structure interaction is given by equation
For more information, see Coupling of fluid and structure (vibro-acoustics).

This can be specified using the ACMODL bulk entry. When no ACMODL is specified, Simcenter
Nastran uses default values to compute the wetted surface. As shown below, the fluid and the
structure are coupled by the matrix [A]. This matrix is called the coupling matrix or the area matrix and
is given by equation

Simcenter Nastran allows the structure and fluid mesh to be non-conformal or non-matching at
the fluid-structure interface.
For a matching mesh, computation of coupling or the area matrix is straight forward because
the shape function = . However, when structure and fluid meshes are non-conformal
(non-matching or dissimilar), this computation is quite complex.
The coupling algorithm uses the following multi-step approach to accurately compute the coupling or
the area matrix (assuming a non-matching fluid-structure interface).
1. Free face determination
Simcenter Nastran automatically determines the fluid and structural element free faces. If you
provide a set of elements, grids, or physical properties, Simcenter Nastran internally converts
these to valid element sets. Only those free faces that are associated with this limited set that you
provide are then retained for coupling computation.
2. Pairing
Starting with this step, the pairing algorithm works with one fluid face at a time. If you provide
a set of elements, grids, or physical properties, Simcenter Nastran internally converts these to
valid element sets. Only those free faces that are associated with this limited set that you provide
are then retained for coupling computation.
A bounding box is created around the fluid face as shown in the below Bounding box figure to find
one or more structural faces that it overlaps. The height of the bounding box is controlled with the
NORMAL field (outward normal), and the INTOL field (inward normal) on the ACMODL bulk entry.
The SRCHUNIT field on the ACMODL bulk entry changes the meaning of the NORMAL and
INTOL fields:
• If SRCHUNIT = REL, NORMAL, and INTOL are a ratio of the search box height to the
maximum edge length of the fluid free face (Default).
• If SRCHUNIT = ABS, NORMAL, and INTOL are a search distance in the model/absolute units.

Chapter
Defining the model
the model
Figure 3-1. Bounding box
3. Parallel check
The structural faces that are within the bounding box must be parallel to the fluid within a
specified tolerance. This is checked by ensuring that the normal of the fluid face and structure
faces form an angle less than the value of the OVLPANG field on the ACMODL bulk entry (default
= 60 degrees).
4. Subdivision
The grids from the structural faces from step 3 are projected onto the fluid face. A mesh is
generated on the fluid face from the set of fluid grids and the projected structural grids such that
the fluid and structural faces share a matching discretization. This virtual discretization is only
used for evaluating the coupling matrix and is not included in the results output.
5. Trimming
Subdivided structural face regions that are outside the fluid face are trimmed and eliminated from
coupling with the current fluid face. However, these eliminated structural regions are considered
for coupling when the above steps are repeated with other adjacent fluid faces.
6. Integration
The subdivided regions on the fluid face are associated to subdivided regions on the structural
faces for integration of matrix [A]. If multiple overlapping regions are paired with a single fluid
region, the structural region that is the closest to the fluid region is coupled to the fluid face.
Because the subdivided fluid and structural regions use a matching discretization, the resulting
surface integral, that is, the overlap area between a fluid face and the given overlapping structural
face, is exact and independent of structural/fluid mesh discretization.
One-way coupling
In the one-way (weak) coupling, the effect of fluid on the structure is neglected. That is, equation

Defining the model
(see Coupling of fluid and structure (vibro-acoustics)) becomes
Equation 3-13.
In this case, Equation 13 can be split into two sets of equations
Equation 3-14.
This allows the first equation to be solved independently of the second equation. The resulting
acceleration then becomes the load vector for the second set of equations. Because the size of
the equations to be solved is now considerably smaller, solve times are faster.
The one-way coupling option can be selected via the CTYPE parameter on the ACMODL bulk entry.
Note
Simcenter Nastran solves the two sets of equations sequentially.
3.4.4 Defining acoustic panel normal velocity and transfer admittance

In SOL 108 and SOL 111, the acoustic panel normal velocity and transfer admittance boundary
conditions are required for acoustic workflows such as muffler transmission loss. Transmission loss is
a measure of a product’s ability to reduce sound.
Acoustic panel normal velocity

Simcenter Nastran supports an acoustic panel normal velocity boundary that you can apply to 2D
acoustic element faces, for example, free faces of a 3D fluid element. This velocity boundary can be:
• Used to represent acoustically rigid yet vibrating panels. In this case, the applied panel velocity
corresponds with the particle velocity of the fluid in front of the panels.
• Combined with an acoustic impedance or admittance on the same panel. In this case, the
acoustic panel normal velocity represents the structural vibration of an acoustically treated and
therefore soft panel. Also, the particle velocity of the fluid in front of the panel is different from the
structural panel velocity, which is pre-defined.
You use the ACPNVEL bulk entry to specify the boundary. ACPNVEL supports the definition of:
• Magnitude and phase of acoustic velocity.

Chapter
Defining the model
the model
• Constant or frequency-dependent complex velocity.
Transfer admittance
The admittance can be measured, calculated from Mechel's formula, or derived from transfer
matrices. In its most general form, the transfer admittance is expressed as
Equation 3-15.
where:
• n1 is the normal velocity on the nodes of the face selection.
• n2 is the normal velocity on the nodes of the second face selection.
• 1 is the pressure on the nodes of the first face selection.
• 2 is the pressure on the nodes of the second face selection.
• 1, 2, 4, and 5 are complex admittance coefficients.
• 3 and 6 are complex source coefficients.
The six coefficients 1 through 6 are determined by the nature of the relation. Because these
coefficients have the dimension [velocity/pressure], they are called transfer admittance coefficients.
Note
The matrix element values depend on the structure between the nodes defined at two
sides of the surface. For example, in a physical problem, the two sides represent two sides
of a wall that contains perforations. Instead of modeling the perforations by using many
small elements, you model only the volumes on both sides of the wall. Then, you capture
the effect of the perforations, which causes the acoustic results between both sides of the
wall to be coupled, through the transfer admittance matrix.
In this case, the nature of the relation encompasses the number of holes and their porosity,
the viscosity of the fluid in the holes, and so on. These are the parameters in Mechel's
formula used to derive the transfer admittance values.
You use the ACTRAD bulk entry that references PACTRAD to specify the transfer admittance.
3.4.5 Duct modes boundary condition

You can apply a rectangular, cylindrical or annular duct modes boundary at the inlet of a duct modeled
with fluid elements. Duct modes are supported in an uncoupled acoustics solution running Simcenter
Nastran Direct Frequency Response (SOL108) with FEMAO.

Defining the model
A duct mode definition is similar to an acoustic load since it results in a propagating acoustic wave.
In duct acoustics, duct modes are important in applications such as HVAC and automotive exhaust
systems. Generally, applications in which exhaust noise reduction is important.
In long ducts, each acoustic mode can be expressed as the product of standing waves with a
propagating component. A specific duct mode is characterized by the mode orders and the amplitude.
The general solution for a duct mode progagating in a duct with an infinite length (z-direction)
is expressed as:
where,
describes the standing wave in x and y,
describes the propagating wave on the duct axis (z-direction),
is the amplitude,
are the mode orders relative to the two directions in the duct cross section,
are dimensions of the duct cross section, and
is the wave number.
Each duct mode has a characteristic cut-on frequency. As a result, specific duct modes propagate
only from a certain frequency onwards. This is demonstrated in the amplitude response plot below
where ω1, ω2, and ω3 correspond to the cut-on frequencies of three duct modes, each with their
own mode orders (m,n).

Chapter
Defining the model
the model
The software considers the boundaries you apply duct modes on as reflection-less. That is, wave
components that are reflected back to the duct mode boundary are completely absorbed.
There are two options to define duct modes on an inlet boundary:
• You can define a specific duct mode with given phase and amplitude.
The specific duct mode option is useful when you understand the amplitude and the frequency
range associated with the incoming noise. For example, you could know the amplitude of the
incident modes from previous experiments. You can define specific duct modes that have a
corresponding cut-on frequency within this range.
• You can request the distributed option.

The distributed duct mode option is useful when your noise source encompasses a broadband of
frequencies or you do not have experimental data for the incoming noise source.
With the distributed option, instead of defining a specific mode, you define the acoustic amplitude
which is applied to all modes. For each solution frequency, the software applies all possible
cut-on modes as incoherent sources with equal acoustic power. The software computes the
response of each mode separately, then sums them to compute the total response.
Multiple duct modes boundary conditions can be defined in the same model, but only one definition
can exist at a specific fluid boundary. For example, you could model a system of ducts which includes
multiple inlets. A unique duct mode boundary condition could be defined on each inlet.
Duct outlet boundary
To allow the acoustic waves to exit a duct system without reflections, you have two options.
• You can define the Anechoic End Duct (AED) boundary on the outlet. The AED is a reflectionless
boundary on a duct outlet. You can use this option when you are only interested in results within
the duct system, and not exterior acoustic radiation.
• You can define the Automatically Matched Layer (AML) at the outlet. When the AML is defined,
the acoustic energy at the outlet can radiate to exterior microphone locations. You can use this
option when you are interested in results within the duct system, and exterior acoustic radiation.
Inputs
• You model a system of ducts with a 3D fluid mesh. The duct inlet and outlet locations in which
you plan to define either a duct mode or an anechoic end duct should be modeled rectangular,
cylindrical or annular.
• The PACDUCT bulk entry defines the cross sectional properties for the duct mode and the
anechoic end duct (AED) boundaries.

Defining the model
o You enter the ID of a BSURFS bulk entry in the BID field. The BSURFS entry selects the fluid
element faces where the duct section boundary is applied.
o You specify that your cross section is a circular, annular, or rectangular with the GTYPE field,
and enter the dimension of your cross section with the DIMi fields.
o Relative to the coordinate system you select in the CID field, you define the cross
section origin with the XLOC,YLOC,ZDOC fields, and the duct axis direction with the
XVEC,YVEC,ZVEC fields.
o You can optionally offset the location of the axial position of the duct origin using the OFFSET
field.
• You define a duct mode with the ACDUCT bulk entry.

o You reference the ID of a PACDUCT entry with the PID field.
o You specify that your boundary is either a specific duct mode, or distributed duct modes
with the WTYPE field.
o You define the units of the duct mode amplitude as pressure, intensity, or power using the
MTYPE field.
o When you are defining a specific duct mode, you will define the mode number pair (m,n)
using the MODX1, MODY1 fields.
o For both the specific duct mode and the distributed duct modes, you will define the amplitude
as a complex data pair. You designate the form of this data pair, (real and imaginary) or
(magnitude and phase), with the FORM field.
• You reference the ID of your duct mode defined with the ACDUCT bulk entry from the case
control with the ALOAD case control command.
• You define your solution frequencies with the FREQi bulk entries. You select these solution
frequencies from the case control with the FREQUENCY case control command.
• You define an anechoic end duct (AED) boundary with the ACNDUCT bulk entry. You reference
the ID of a PACDUCT entry with the PID field.
• The DUCTFMAX parameter is available to define the maximum duct mode frequency using the
product (DUCTFMAX * maximum excitation frequency). This value determines the truncation
of the duct modes wavebasis. It is typically defined slightly over the maximum frequency of
interest. The DUCTFMAX default is 1.2.
Duct mode output

Two output options are available when a duct modes boundary condition is defined on a duct inlet,
and either an anechoic end duct boundary or an AML is defined on a duct outlet.
• The DMTRLOSS case control command is available to request the duct modes transmission
loss output.

Chapter
Defining the model
the model
Transmission loss is computed as the power introduced by the duct mode boundary conditions
divided by the sum of the modal transmitted power coefficients at the anechoic end duct boundary
or the power at the AML radiation surface. Mathematically it is written as:
• The DMTRCOEF case control command is available to request the duct modes transmission
coefficients.
Transmission Coefficients represent the amplitude of the output modes expressed in terms of
pressure, intensity, or power.
Input file example:

...
$Case Control
...
***Solution frequency selection
FREQUENCY = 100
***Duct tranmission loss output request

DMTRLOSS(PRINT) = YES
SUBCASE 1
***ALOAD command selects the duct mode with a set ID of 2.
ALOAD = 2
BEGIN BULK
***These parameters request FEMAO
ACADAPT STANDARD
ACORDER MINIMUM 1
ACORDER MAXIMUM 1
***Solution frequency definition

FREQ1 100100.00005.000000 180
***Anechoic end duct definition

ACNDUCT 1 2
***Anechoic end duct cross sectional properties

PACDUCT 2 2 CIRCULAR 28.5956 1 0+
+ 5171.33603.8923 1003.400.4762721.7123-40.879298
***Description referenced by the anechoic end duct

DESC 1Duct Outlet
***Fluid element face selection for anechoic end duct boundary

BSURFS 2 332342 91799 88669 91798+
+ 332418 88688 91799 91800 332464 91797 88671 91796+
+ 332623 91796 91795 91814 337443 91795 88671 88672+
+ 337444 91801 91821 91822 337797 88687 88688 91800
***Duct mode definition

ACDUCT 2 4SPECIFIC PRESS REAL +

Defining the model
+ 0 1 1000.00 0.0000
***Duct mode cross sectional properties

PACDUCT 4 4 0.0000CIRCULAR 25.3987 2 0+
+ 1823.95-167.270 1357.170.9493285.4486-50.314286
***Description referenced by the duct mode

DESC 2Duct Inlet
***Fluid element face selection for duct mode boundary

BSURFS 4 155804 98888 98889 98906+
+ 156115 103130 98882 98884 157499 103364 98885 98883+
+ 157502 102356 98884 98901 157503 102401 98884 102356+
+ 157661 98883 98902 98882 158279 98887 98888 98906+
+ 158293 98903 98883 98885 158894 98886 103409 98887+
+ 158897 98902 98883 98903 159463 98904 98906 98905
...
3.4.6 Coupled FEMAO

You can define coupling of a structure to an acoustic fluid in a vibro-acoustic FEMAO solution.
You can couple a structure to a fluid in a FEMAO (FEM Adaptive Order) solution. A FEMAO solution
is a higher-order polynomial method for acoustic and vibro-acoustic analyses. It provides more
accurate results and faster solution times by adapting the computational effort to the complexity of
the analysis. By using a FEMAO solution, the fluid elements can be relatively large, yet result in
accurate results at high frequencies.
In a vibro-acoustic simulation, the structure is expressed in modal DOFs and the acoustic fluid in
physical DOFs.
For FEMAO, SOL 108 Direct Frequency Response has been enhanced, and SOL 111 Modal
Frequency Response is a new solution. The enhanced functionality can be used in Simcenter
Nastran FEM vibro-acoustics solutions, such as:
• An engine in a vehicle engine compartment radiating sound both inside the engine bay and
outside the vehicle.
This type of solution could be used to model an automotive engine in a vehicle, surrounded by
an acoustic fluid which incorporates the geometric features of other structures, such as the
body-in-white and the passenger acoustic cavity. You can excite the engine, which radiates
sound into the fluid, and capture the sound response both inside the engine bay, and outside of
the vehicle where it contributes to pass-by noise.
• A panel transmission loss simulation, using modal coordinates for the structure and fluid meshes
with AML (AMLREG) in physical coordinates to represent reverberant and anechoic side on the
front and back of the panel.
Fluid-structure interaction
You can now couple structural FEMs to acoustic fluids as component FEMs in an assembly FEM,
and solve the solution using FEMAO.
Structural FEMs:

Chapter
Defining the model
the model
• Can be coupled to acoustic fluids as component FEMs in an assembly FEM and solved using
FEMAO.
• Can now be solved with fluids for acoustics and vibro-acoustics in both FEM and FEMAO
solutions SOL 108 and SOL 111.
SOL 111 FEMAO now supports:

• Two-way (strong) fluid-structure coupling.
• One-way (weak) fluid-structure coupling with only force excitation of the structure
You define the minimum and maximum order for the elements in the entire model.
Examples:
ACADAPT FINE
ACORDER MAXIMUM 10
ACORDER MINIMUM 2
You set the FEMAO options in the Bulk Data.
You define the adaption rule with the ACADAPT bulk entry.
You define the element order with the ACORDER bulk entry.
3.4.7 Impedance or admittance

The acoustic absorber element CAABSF defines frequency-dependent impedance or admittance
boundary conditions. It allows POINT, LINE, or AREA impedance to be specified on the free fluid
surface. The element topology is specified by CAABSF, and the frequency-dependent impedance
values are specified on the PAABSF entry.
Simcenter Nastran also supports constant real and imaginary impedance and admittance on a 2D
fluid free surface when CAABSF references the physical property PAABSF1.
If only one grid point is specified in CAABSF element, then the impedance
is the total impedance at the point. If two grids are specified, then the impedance is per unit length. If
three or four points are specified, then the impedance is per unit area or the specific impedance. The
PAABSF bulk entry allows you to define impedance as follows:

Defining the model
The resistance represents a damper quantity B that is frequency dependent. The reactance
represents a quantity of the type (ωM − K/ω). The impedance is defined by equation
For more information, see Imposed impedance or admittance (Robin boundary condition).
The scale factor S is used in computing element stiffness and damping terms as:
Equation 3-16.
The imaginary part of can then be construed as frequency-dependent stiffness matrix. Therefore,
Equation 3-17.
The absorption coefficient is then
Equation 3-18.
Note
• Parabolic element faces are also supported because only corner grids are used to
define the element faces.
• Point and line impedance are not supported in the acoustic method introduced in NX
Nastran 11, which includes AML, porous materials, and microphone elements.
• Point and line impedance are only supported through CAABSF that references
PAABSF if the system cell 617 (ACFORM) is set to 0. ACFORM = 0 reverts to the NX
Nastran10 acoustics behavior.

Chapter
Defining the model
the model
• Although PAABSF is not supported by the FEM Adaptive Order (FEMAO) method,
PAABSF1 is supported by both the standard FEM and the FEM Adaptive Order
(FEMAO) method.
3.4.8 Defining infinite boundary

See Open or radiating boundary for details on infinite boundary condition.
3.5 Defining acoustic loads

The following types of loading are available in Simcenter Nastran for fluid elements:
• Constant, frequency-dependent, or time-dependent enforced pressure at the grid points.
• An acoustic source characterized by a volumetric flow rate and corresponding to a power

spectral density function.
An acoustic source is assumed to be a pulsating sphere in infinite space and is defined on the
ACSRCE bulk entry. The ACSRCE entry is selected by the DLOAD Case Control command and
contains the material properties of the source and references a DAREA and TABLEDi entry. The
TABLEDi entry defines the power-versus-frequency curve characterizing the acoustic source. The
ACSRCE entry may also define a delay time and phase angle which is useful whenever multiple
sources are present.
In addition to the above acoustic loads, you can also apply any typical dynamic loads to the structural
portion of the model.
3.5.1 Pre-computed enforced vibrations at fluid-structure interface

When surface vibrations (displacements, velocities, or accelerations) on the structure are pre-defined,
the structural degrees of freedom are mainly used to compute the fluid loading using the coupling
matrix [A].
This reduces equation
(see One-way coupling) as follows:
Equation 3-19.
Equation 3-20.

Defining the model
The SPCSTR parameter can be used to instruct Simcenter Nastran to constrain the structural
degrees of freedom in the analysis set (after transferring the enforced motion to the fluid).
The parameter SPCSTR (see Parameters and Parameter Applicability Tables chapter in the Quick
Reference Guide) is used for fluid-structure analysis only.
3.5.2 Enforced acoustic pressure with complex data input

When acoustic pressures on certain grids are known (for example, the pressures are identified through
measurement or previous calculation), you can apply them using the enforced acoustic pressure card
ACPRESS. SOL 108 and SOL 111 acoustic and vibro-acoustic solutions support ACPRESS.
ACPRESS supports complex data and the definition of:
• Magnitude and phase of pressure on multiple grids.
• Constant and frequency dependency.
Note
Simcenter Nastran adds the degrees of freedom referenced by ACPRESS to the s-set.
For detailed information on the s-set, see the Understanding Sets and Matrix Operations
chapter in the User’s Guide.
3.5.3 Fan noise boundary condition

You can define a fan noise inside an acoustic domain. These aeroacoustic simulations for subsonic
fluid excitations, such as from a rotating fan or propeller, are solved using a FEM Adaptive Order
(FEMAO) acoustic solution. The method can be used to solve a variety of aeroacoustic simulations,
such as marine propeller sound propagation, cooling fan noise, or HVAC blower and duct acoustic
interaction.
A technique has been implemented to solve for the acoustic response using aeroacoustic excitations.
Previously solved CFD simulation surface pressures as a function of time are imported from a CGNS
(CFD General Notation System) file, converted to a blade force excitation, and are then applied as a
fan blade force using equivalent rotating acoustic forces.
You can read transient or frequency-dependent dynamic forces from spatially and time varying
computational fluid dynamics, acoustic, force loads, or stochastic pressure loads that were
transformed to Simcenter Nastran loads and stored in the external *.sc_h5 file by a pre/post software,
such as Simcenter 3D Pre/Post.
You create a SOL108 FEMAO Acoustics solution and then include single or multiple fan noise
sources in your simulation. The rotation speed (RPM) for all fans must be identical.
A fan noise can have a single blade, for which all blade forces will be assumed to be periodically
identical. If the fan noise has multiple blades, then blade forces can be unique for every blade. Fan
noises must be located on or inside the acoustic fluid boundary.
Fan excitation can be:
• Tonal Noise, which only includes blade passing frequencies and noise.

Chapter
Defining the model
the model
Note
If the imported CFD fan pressures are defined for only a single blade, you can only
apply Tonal Noise.
• If you have imported CFD fan pressures with varying forces on several blades, then you can
apply Tonal and Broadband noise, which includes blade passing frequencies and broadband
turbulent noise.
You use a fan noise load to excite the acoustic fluid around your structure.
In the solution Case Control, you specify the forcing frequencies as harmonics and sub-harmonics
to set the solution frequencies. Harmonics include only the fan. Subharmonics are available for
both the fan and shaft.
You create a fan noise load with the ACFAN bulk entry.
You define the fan noise properties with the PACFAN bulk entry.
You define fan noise harmonics with the FREQH bulk entry
Input file example:
...
$* FILE MANAGEMENT
ASSIGN SC_H5='D:\WorkDir\Project\nx13_LMS1306 Fan Noise\generated_fan_s,
egments.sc_h5' UNIT=201
ASSIGN SC_H5='D:\WorkDir\Project\nx13_LMS1306 Fan Noise\generated_fan_s,
egments.sc_h5' UNIT=202
SOL 108
CEND
$* CASE CONTROL
SUBCASE 1
LABEL = Subcase - Direct Frequency 1
ALOAD = 1
FREQUENCY = 101
SUBCASE 2
LABEL = Subcase - Direct Frequency 2
ALOAD = 2
FREQUENCY = 102
$* BULK DATA
BEGIN BULK
$REQH SID TYPE HTYPE PID START

FREQH 101 SINGLE SHAFT 101 1
$ACFAN PID TYPE RPM NBLADES
PACFAN 101 TONAL 2000.00 1
$REQH SID TYPE HTYPE PID START END STEP NSUB

FREQH 102 LSUB BLADE 102 10 50 4 2
$ACFAN PID TYPE RPM NBLADES
PACFAN 102 TONAL 2166.67 1
$* Load: Fan Noise(1)

$CFAN SID PID LUNF LDID LCID WINDOW

Defining the model
ACFAN 1 101 201 1 1 HANNING

$TI NAME LDID LDID DESCRIPT "ENDREC"
DTI DISTL 1 1 Fan1/Forces ENDREC
$* Load: Fan Noise(2)

$CFAN SID PID LUNF LDID LCID WINDOW
ACFAN 2 102 202 2 2 RECTANG
$TI NAME LDID LDID DESCRIPT "ENDREC"
DTI DISTL 2 2 Fan1/Forces ENDREC
3.5.4 Acoustic source loading
An acoustic source is assumed to be a sphere that oscillates in an infinite acoustic field. The
strength of the source is characterized by a frequency-dependent flux of volume velocity defined by
(surface area times velocity)
Equation 3-21.
where:
• is the strength of the acoustic source.
• r is the unit radial vector.
The power radiated from the source is related to its strength by:
Equation 3-22.
where:
• is the density of the fluid.
• is the speed of sound in the ambient medium at a given temperature.
• is the monopole amplitude.
Equation
(see Coupling of fluid and structure (vibro-acoustics)) can be re-arranged to yield

Chapter
Defining the model
the model
Equation 3-23.
The acoustic source strength can be specified explicitly as a dynamic nodal load (see in
Finite element formulation) or can be computed given the source radiated power (using equation
in Coupling of fluid and structure (vibro-acoustics)).
Acoustic source can be defined in terms of:
• Monopole amplitude.
• Source strength amplitude.
• Acoustic power.
If the source strength amplitude is known, this data can be specified directly as a dynamic load
using the RLOADi bulk entry. If the acoustic power is known then this data can be specified using
the ACSRCE bulk entry. If only the monopole amplitude is known, then either the acoustic power
or the source strength amplitude can be computed and in turn either the ACSRCE or RLOADi bulk
entry can then be used to specify the acoustic source information.
3.5.5 Monopole, dipole, plane wave, and surface dipole acoustic sources
Simcenter Nastran supports dedicated acoustic monopole, dipole, and plane wave sources in SOL
108 and SOL 111.
Acoustic monopole source
An acoustic monopole is a pulsating sound source that radiates equally in all directions.
Figure 3-2. Acoustic monopole source

Defining the model
Note
The plus sign means that the monopole source expands.
In Simcenter Nastran FEM acoustics, you can create acoustic monopoles:

• Inside the meshed fluid volume at a fluid grid point or inside an element.
• Outside the meshed fluid volume.
The monopole source generates an incident sound field at a location of a distance R due given by
Equation 3-24.
where:
• is the monopole amplitude.
• R is the distance from the source.
• k is the wave number.
The monopole source can also be specified by its volume velocity Qs. The relationship between the
monopole amplitude and volume velocity is given by
Equation 3-25.
Another method you can use to specify the monopole source is to use its acoustic power.
Equation 3-26.
where:
• is the fluid density.
• c is the speed of sound.
You use the bulk entry ACPOLE1 to specify a monopole source.

• The bulk entry supports power and monopole amplitude inputs that can be constant or frequency
dependent.

Chapter
Defining the model
the model
• In ACPOLE1, the source location is defined by the coordinate that can be inside or outside of
the meshed fluid volume. For sources outside the FEM mesh, you specify an AMLREG on all
or part of the free faces of the fluid model. This captures the effect of the monopole incident
field inside the fluid domain.
• ACPOLE1 inherits the density and speed of sound from the location of the monopole source. If
the source is outside the meshed fluid volume, the fluid properties are acquired by averaging
the properties used in those fluid elements that their free faces are used in the definition of
the AMLREG.
Acoustic dipole source

The dipole source can be visualized as an oscillating sphere with no deformation as shown in the
image below. The fluid near the source moves back and forth to produce sound. Unlike a monopole
source, the sound does not radiate equally in all directions.
Figure 3-3. Acoustic dipole source

Also, the dipole source can be visualized as two out-of-phase monopole sources separated by a
distance, where one monopole contracts (- sign) and the other one (+ sign) expands. In this case, the
dipole generates a sound field that can be written as
Equation 3-27.
(1) Evaluation point
Figure 3-4. Dipole sound field

Defining the model
where:
• d is the distance between two out of phase monopole sources.
• r is the line that connects the midpoint between the monopole sources and the evaluation point (1).
• is the angle between the line that connects those monopole sources and the line that connects
the evaluation point (1) and the midpoint between the monopole sources.
To specify the dipole source in your simulation, you define the dipole moment Sd
where:
• S is the source strength.
• d is the direction vector.
You use the ACPOLE2 bulk entry to specify a dipole source. The bulk entry supports:
• The definition of the dipole moment in a coordinate system.
• The source location inside or outside the meshed fluid volume. If the source is outside the
meshed fluid volume, you must define an AMLREG bulk entry to account correctly for the incident
field from the dipole.
Acoustic plane wave source

A plane wave source generates a plane wave on only one side of the space, in the positive direction
of the source vector.
(1) Incident field

(2) Vector
(3) Position
(4) Non-incident field
Figure 3-5. Acoustic plane wave source
The incident acoustic pressure pi due to a plane wave source is
Equation 3-28.
where:
• A is the amplitude of the plane wave.
• k is the wave number.

Chapter
Defining the model
the model
• d is the perpendicular distance from the plane source.
Note
• Simcenter Nastran uses surface integration to compute acoustic power. For a
monopole source, you can create a spherical microphone mesh that encloses the
source, and the software computes the acoustic power on this sphere. This also
means that you can define acoustic power for a monopole source. However, for plane
waves, you cannot define acoustic power because the plane waves are infinitely large
in terms of direction, and the surface integration would result in infinite numbers.
• You can use a preprocessor, such as Simcenter 3D Pre/Post, to create a collection of

discrete plane wave sources equally distributed in space that model an acoustic diffuse
field. This collection is usually used for a random vibro-acoustic analysis.
You use the bulk entry ACPLNW to specify a plane wave acoustic source and define location,
direction, and amplitude that can be constant or frequency dependent.
If the monopole, dipole, and plane wave sources are outside of the meshed fluid volume, you must
define an AMLREG bulk entry to correctly account for the incident field of these sources inside the
meshed fluid volume. Also, in this case, an incident-scattered formulation is used in Simcenter
Nastran FEM acoustics.
Surface dipole source

For aeroacoustic analyses, you can create an equivalent surface dipole acoustic source as a
boundary condition. A surface dipole is an acoustic source that you use to model the sound
generated by rigid surfaces located in a low-speed flow field application. In Simcenter Nastran,
surface dipoles are based on the Neumann formulation, which assumes that the mass density and
speed of sound around the rigid surfaces are uniform.
To model the sound generated by the interaction of a compressible or incompressible flow with a rigid
surface, define the surface dipole boundary condition with fluid pressures.
The aerodynamic or hydrodynamic fluid pressures are typically computed by computational fluid
dynamics (CFD) software, such as Simcenter Star-CCM+, and processed by a pre/post software for
later use by Simcenter Nastran. A pre/post software may read, map, Fourier transform, time-segment,
and write the dynamic fluid pressures to an sc_h5 file (*.sc_h5) in HDF5 format.
For example, you may want to examine sound generated by the interaction of low-speed flow with
surrounding surfaces, such as HVAC duct walls, or an automotive side mirror. In contrast to solving
the entire acoustics problem with volumetric flow information and generating an equivalent point
source for each CFD cell, you use a surface dipole source on, for example, a side mirror surface as a
boundary condition in your analysis. This reduces the data transfer from the CFD software to the
pre/post software, and consequently the computational effort by Simcenter Nastran.
This capability is supported in the following solutions and requires the FEM Adaptive Order (FEMAO)
method:
• SOL 108 for uncoupled acoustic and coupled vibro-acoustic analyses.

Defining the model
• SOL 111 for coupled vibro-acoustic analyses.
Note
The coupling can be weak or strong.
Surface dipole workflow:

1. Use the ASSIGN file management statement to select the .sc_h5 file that contains the processed
dynamic fluid pressures.
2. Use the ACSPO2 bulk entry to define a surface dipole for the frequency response analysis.
3. Use the PACSPO2 bulk entry to define the parameters of the surface dipole for the frequency
response analysis.
4. Use the ALOAD case control command to apply the dynamic loads to the response problem.
5. Use the ALOAD bulk entry to define a dynamic loading condition for the ACSPO2 bulk entry.
6. Use the DTI,DISTL bulk entry for the load description of the ACSPO2 bulk entry.
For more information, see the ACSPO2 and PACSPO2 bulk entries.
3.5.6 External loading

You can read frequency-dependent dynamic forces for a frequency response analysis or Power
Spectral Density (PSD) data for a random response analysis from an external file that is in the
open-source HDF5 format and commonly has a *.sc_h5 extension. This allows you to more efficiently
solve vibro-acoustic problems. Typically, these loads are spatially and time-varying computational
fluid dynamics, acoustic, force loads, or stochastic pressure loads that were transformed to Simcenter
Nastran loads and stored in the external file by a pre/post software, such as Simcenter 3D Pre/Post.
For example, you may want to transform deterministic aerodynamic forces from Simcenter Star-CCM+
that cause an automotive side window to vibrate and use these loads in a vibro-acoustic solution to
examine the acoustic response at a driver's ear to the structural excitation.
This capability is supported by SOL 108 and SOL 111 solutions with standard (fixed low-order) FEM
or FEM Adaptive Order (FEMAO) method, which is a higher-order polynomial method.
In a random response run:
1. Use the ASSIGN file management statement to select the .sc_h5 file that contains the Simcenter
Nastran loads.
2. Use the RLOADEX bulk entry to define frequency-dependent dynamic forces obtained from a
file of the form:
3. Use the RANDPEX bulk entry to reference a PSD specification that is a simple PSD across all
degrees of freedom for use in a random analysis.

Chapter
Defining the model
the model
4. Use the DLOAD case control command to apply the dynamic forces to the response problem.
5. Use the DLOAD bulk entry to define a dynamic loading condition for the RLOADEX bulk entry.
6. Use the DTI,DISTL bulk entry for the load description of the RLOADEX and RANDPEX bulk
entries
7. Use the RANDOM case control command to select the identification number of a RANDPEX
bulk entry.
8. Depending on the organization of your subcases for a random analysis, you may need to include
ANALSIS = RANDOM in the subcases that contain the RANDOM case control command.
For more information, see the ASSIGN case control command, the RANDPEX, RLOADEX, and
DTI,DISTL bulk entries.
For more information on ANALYSIS = RANDOM, see the RANDOM case control command.
3.6 Defining microphone meshes
You can request acoustic results at arbitrary locations exterior or interior to the fluid. These locations
are defined with the 0D, 1D, 2D, or 3D microphone mesh. A microphone mesh is also known as a
field point mesh.
When a microphone mesh is interior or on the Automatically Matched Layer (AML) boundary,
Simcenter Nastran interpolates the results from the fluid grids to the microphone location. However,
when the microphone mesh is exterior to the AML boundary, Simcenter Nastran uses the acoustic
results at the boundary of the FE domain and a boundary integral to obtain the acoustic response.
For more information on AML and microphone meshes exterior to the AML boundary, see Acoustic
analysis using automatically matched layer.
You can request pressure output with the PRESSURE case control command, acoustic intensity
with the ACINTENSITY case control command, and acoustic velocity with the ACVELOCITY case
control command for the fluid grids referenced by a 0D, 1D, 2D, and 3D microphone mesh. You
can also request acoustic power for the fluid grids referenced by a 2D microphone mesh with the
ACPOWER case control command.
Note
When you define a 2D microphone mesh for power output over multiple acoustic domains
and you also use acoustic incident or scattered sources, Simcenter Nastran can only
approximate the power for the 2D microphone faces crossing from one domain to another.
The following summarizes the 0D, 1D, 2D, and 3D microphone mesh definitions on fluid grid points:
• You can specify the 0D microphone mesh using the MICPNT bulk entry.
• You can specify the 1D microphone mesh using the CROD entry, which references the PMIC
bulk entry for the physical property.

Defining the model
• You can specify the 2D microphone mesh using the CTRIA3 and CQUAD4 entries, which
reference the PMIC bulk entry for the physical property.
• You can specify the 3D microphone mesh using the CTETRA, CHEXA, CPENTA, and CPYRAM
entries, which reference the PMIC bulk entry for the physical property.

Chapter 4: Solution methods
4.1 Overview of solution methods

Direct and modal
Acoustic and vibro-acoustic analyses can be performed using both direct and modal methods.
• Direct dynamic analysis can be performed using solution sequences 107 through 109 and
solution 200 with ANALYSIS = DFREQ.
Note
In a direct frequency response solution (SOL 108):
o You can request the global iterative solver. The solve can result in significant
computational savings through preconditioner reuse. To activate the global
iterative solver in SOL 108, set either NASTRAN ITER = YES in the Nastran input file
or add iter = yes to the Nastran command line.
o You can request the Padé via Lanczos (PvL) approach for vibro-acoustic models
in which the structure and fluid are strongly coupled. The PvL approach utilizes
factorization of the matrix at selected frequencies for all responses and therefore
significantly reduces the computational time for problems with large number of
frequencies.
To request the PvL approach, set either the new keyword krylov or the new
system cell 679 (krylov) to YES.
NASTRAN KRYLOV=YES
The iterative solver support for frequency-dependent solutions and the Pade via
Lanczos (PvL) approach are mutually exclusive. If you request both PvL and
global iterative solver solution, the PvL approach will take precedence over the
frequency-dependent iterative solver approach.
• Modal analysis can be performed using solution sequences 103, 110, 111, 112, and 200 with
ANALYSIS = MFREQ and ANALYSIS = MTRAN.

Chapter
Chapter 4: 4: Solution
Solution methods
methods
Note
In modal solutions (103, 110, 111, 112, and 200), the normal modes for the structure
and the fluid are computed independently. The uncoupled modes of the fluid and
structure are then used to compute modal frequency response. Thus, the separately
evaluated fluid and structural modes from, for example, a SOL 103 run, can be used
to perform a restart into a modal response calculation.
Modal
When you create a vibro-acoustic analysis with SOL 111, you can optionally use modal coordinates
for the structure and physical coordinates for the fluid.
When you specify METHOD(STRUCTURE) instead of METHOD(FLUID) in the case control section,
the solver:
1. Computes the normal modes for the structure.
2. Does not compute the modes for the fluid. The fluid is retained in the physical degrees of freedom.
You use this mixed-coordinates approach for frequency-dependent fluid. For example, the fluid
contains porous material or has an Automatically Matched Layer (AML) boundary condition.
For information on AML, see Acoustic analysis using automatically matched layer.
Note
You can also use modal coordinates for the fluid and physical coordinates for the structure.
However, you rarely use this approach.

Chapter 5: Input data requirements
5.1 Executive control commands

No extra executive control commands are necessary to perform an acoustic or a vibro-acoustic
solution. Both modal and direct solutions are available for selection.
5.2 Case control commands

The standard case control commands for the modal and direct solutions can be used for both
acoustics as well as vibro-acoustics. Single and multipoint constraints (SPC and MPC) can be
imposed on the model and the dynamics loads referenced via the DLOAD entry.
5.2.1 Acoustic load, boundary condition, and source control

Use the case control command ALOAD to specify the enforced acoustic pressure load with:
• Complex data input (ACPRESS).
• Panel normal velocity boundary condition (ACPNVEL).
• Monopole source (ACPOLE1).
• Dipole source (ACPOLE2).
• Plane wave sources (ACPLNW).
ALOAD can optionally reference a corresponding bulk entry ALOAD.

For more information on the ALOAD bulk entry, see Acoustic loads, boundary condition, and sources.
5.2.2 Import coupling or the area matrix

The case control command A2GG can be used to import the coupling or the area matrix in the
DMIG format.
5.2.3 Fluid-structure interaction control

The FLSTCNT case control command (see FLSTCNT in the Simcenter Nastran Quick Reference
Guide) can be used to specify the following:
• Formulation to use symmetric or asymmetric, that is, equation
or equation

Chapter
Chapter 5: 5: InputInput
data data requirements
requirements

Setting ACSYM = YES requests a symmetric formulation and ACSYM = NO requests an
asymmetric formulation.
• The type of pressure output, peak or RMS. RMS output is available in Modal and Direct
Frequency Response analysis only (SOLs 111, 108, and 200).
• The reference pressure to use for computing pressure in decibels. A peak reference pressure
must be specified for PREFDB. Pressure is expressed in decibels as follows dB = 20
log(P/PREFDB). For example, the reference pressure for air is PREFDB = 2.0E−05 Pascals.
• Whether a coupled solution of structure and fluid needs to be performed. The default is ASCOUP
= YES. Specifying ASCOUP = NO disables the coupled solution.
• If the coupling or the area matrix needs to be punched out. This is controlled by AGGPCH on the
FLSTCNT case control command. If AGGPCH = YES, Simcenter Nastran writes the area or
coupling matrix in the DMIG format to the punch file.
You can then import and optionally scale this AGG matrix in a consecutive solution that includes
the same model definition at the coupled interface. Use the A2GG case control command and
the ASCOUP describer on the FLSTCNT case control command to request this import.
See the FLSTCNT case control command in the Simcenter Nastran Quick Reference Guide.
• The output of the coupling information file is controlled by the SKINOUT parameter. Depending
on the option used, a Simcenter Nastran punch file *.pch, coupling information output file *.dat,
and/or OACCQ output data block is created. The name of the files is the base name of the deck
appended with _coupling_info.pch and _coupling_info.dat. For example, if the original input file is
test.dat, the debug files are test_coupling_info.pch and test_coupling_info.dat.
The OACCQ output data block contains distance information for the coupled and uncoupled fluid
and structure faces. The output is in addition to the coupling information written to the .dat
and .pch files.
You can use this information to display the coupling data as follows.

Input data requirements
(1) Uncoupled
Structural Faces
(red)
(2) Coupled
Structural Faces
(blue)
Figure 5-1. Coupled and uncoupled structural faces
For detailed information on the OACCQ data block, see the DMAP Programmer's Guide.
The SKINOUT input options are:
o SKINOUT = PUNCH:
A punch file, a coupling information file and an OACCQ output data block are output.
The punch file (*.pch) contains a list of the original structural and fluid element IDs that
participated in the coupling. The debug data deck (*.dat) contains:
■ Dummy shell elements that represent the coupled structural faces, and are assigned
to a dummy pshell with ID = 1.
■ Dummy shell elements that represent the coupled fluid faces, and are assigned to a
dummy pshell with ID = 2.
o SKINOUT = FREEFACE:
A coupling information file and an OACCQ output data block are created (no punch file). The
coupling information file contains:
■ Dummy shell elements that represent the coupled structural faces, and are assigned
to a dummy pshell property with ID = 1.
■ Dummy shell elements that represent the coupled fluid faces, and are assigned to a
dummy pshell property with ID = 2.
■ Dummy shell elements that represent the uncoupled structural free faces, and are
assigned to a dummy pshell property with ID = 3.
■ Dummy shell elements that represent the uncoupled fluid free faces, and are assigned
to a dummy pshell property with ID = 4.
o SKINOUT = STOP:

Chapter
requirements
Works like SKINOUT = PUNCH, except the solution stops immediately after the debug files
are created. It is recommended that you use SKINOUT,STOP in the first run, then review the
coupling interface in a pre/post system before continuing with the remainder of the solution.
o SKINOUT = NONE (default):

No coupling information files are created.
• To request if the SFE Akusmod fort.70 acoustic coupling matrix needs to be read into Simcenter
Nastran. When SFEF70 = YES, Simcenter Nastran does not compute the coupling, and instead
uses the coupling definition from the external file. When SFEF70 = NO (default), the external
coupling file is not imported. Simcenter Nastran expects the AKUSMOD file in the same directory
where the job is being run, and expects the file name to be fort.70. An ASSIGN statement which
uses UNIT = 70 must be defined in the file management section of your input file if the coupling
file is not named fort.70 or if it is in a location other than where the job is run.
Example
ASSIGN OUTPUT2 = '/directory_path/user_file_name.70' UNIT = 70
5.2.4 Random acoustic analysis

For distributed acoustic plane wave problems, you can perform power spectral density (PSD) random
acoustics analysis. This capability is supported for both direct and modal solutions.
To request that the software calculate the random response, in your input file include the RANDOM
case control command. Then, organize the subcases as follows:
1. Create a frequency response subcase for each acoustic loading. Make sure that all of these
subcases reference the same set of frequencies.
The acoustic loads can include acoustic monopole, dipoles, or plane waves. A typical acoustic
loading, is that of an acoustic diffuse field. You can represent an acoustic diffuse field with a
number of subcases that each contain an acoustic plane wave source. Orient each plane wave
source differently with respect to the object that sees the acoustic diffuse field. For example, the
object might be a panel or a spacecraft. To define a diffuse acoustic field, you can define these
sources as uncorrelated sources in a RANDOM subcase by specifying the diagonal (auto-PSD)
factors in the RANDPS bulk entries only.
2. Below the frequency response subcases, include a random analysis subcase for each PSD
function that you want to evaluate. Include ANALYSIS = RANDOM in these subcases.
The software uses the frequency response functions that are calculated in the immediately
preceding frequency response subcases as the inputs to the random calculations.
To specify cross-power spectral density and cross-correlation functions, use the RCROSS or
RCROSSC bulk entries. When you do so, set the RTYPE to PRESS to obtain acoustic pressure
output.

Example
Suppose that the fluid in a vibro-acoustic problem is excited by three plane wave loadings, and
you want to evaluate the acoustic random response for two PSD functions. You can organize the
subcase structure of the input file as follows:
SUBCASE 1
$
$ Subcase 1 calculates the frequency response function for the
$ loading specified by ALOAD 111 at the frequencies specified by
$ FREQUENCY set 13
$
FREQUENCY=13
ALOAD=111
$
SUBCASE 2
$
$ FREQUENCY set 13
$
FREQUENCY=13
ALOAD=211
$
SUBCASE 3
$
$ FREQUENCY set 13
$
FREQUENCY=13
ALOAD=311
$
SUBCASE 4
$
$ Subcase 4 uses the frequency responses from Subcases 1-3 to
$ calculate the random response of the structure for the PSD function
$ specified by RANDOM 100
$
ANALYSIS=RANDOM
RANDOM=100
$
SUBCASE 5
$
$ Subcase 5 uses the frequency responses from Subcases 1-3 to
$ calculate the random response of the structure for the PSD function
$ specified by RANDOM 200
$
ANALYSIS=RANDOM
RANDOM=200
5.2.5 Output requests

You can use the following case control commands to request results output from acoustic and
vibro-acoustic types of frequency response analysis:

Chapter
requirements
Table 5-1. Acoustic and vibro-acoustic analysis output requests

Direct or Acoustic or
Case control
Output description modal vibro-acoustic
command
solutions analysis
Acoustic intensity at fluid grids referenced by
ACINTENSITY Both Both
microphone elements
Acoustic power for 2D microphone surfaces and
ACPOWER Both Both
AML regions
Acoustic velocity at fluid grids referenced by
ACVELOCITY Both Both
microphone elements
Requests the duct modes transmission loss
DMTRLOSS Direct Acoustic
output.
Requests the duct modes transmission
DMTRCOEF Direct Acoustic
coefficients.
Contributions of grids in structural panels to
GRDCON Both Vibro-acoustic
acoustic response
Incident acoustic power for 2D microphone
INPOWER Both Both
surfaces and acoustic free faces
Contributions of structural modes to acoustic
MODCON Modal Both
response
Contributions of structural panels to acoustic
PANCON Both Vibro-acoustic
response
PEAKOUT Limits the output to the peak acoustic responses Both Both
Acoustic pressure at fluid grids or fluid grids
PRESSURE Both Both
referenced by microphone points
TRLOSS Acoustic transmission loss Both Both
Transmitted acoustic power for 2D microphone
TRPOWER Both Both
surfaces and AML regions
To minimize noise radiated from a structure without performing an acoustic or vibro-acoustic analysis,
you can use the ERP case control command to help identify the areas of a structure that are the
primary sources of the noise due to structural vibration.
For more information on the ERP case control command, see the Equivalent Radiated Power Output
section of the Simcenter Nastran Basic Dynamic Analysis User's Guide.
Modal, panel, and grid contribution output
You can use the MODCON, PANCON, and GRDCON case control commands to produce output that
relates the acoustic response at specified locations at each excitation frequency to structural or fluid
modes, structural panels, or grids in structural panels.
• Use MODCON to examine how each structural mode contributes to the acoustic response, how
each fluid mode contributes to the structural response, or how with the PANELMC describer,
each structural mode contributes to the acoustic response on a structural panel-by-structural
panel basis.
For more information on the MODCON case control command including applications to structural
only frequency and transient response analysis, see the Frequency Response Solution Control
and Output section of the Simcenter Nastran Basic Dynamic Analysis User's Guide.

• Use PANCON to examine how the motion of structural panels contribute to the acoustic response.
• Use GRDCON to examine how the motion of grids in structural panels contribute to the acoustic
response.
Note
PANCON and GRDCON results are only valid when the excitation is from structural loads
and not acoustic loads.
PANCON case control command
The PANCON case control command output quantifies how the motion of structural panels contributes
to the acoustic response at a specific location in the fluid.
Structural panels are surfaces at the interface of the structure and fluid. To define structural panels,
use the PANEL bulk entry.
When you use the PANCON case control command, you can specify whether the output is in SORT1
or SORT2 format and whether the output is normalized.
For SORT1 output, at each excitation frequency (as specified by FREQ) and response (as specified
by SETMC grid and DOF entries) combination, the software ranks the contribution that each panel
makes to the response based on the results of the following equation:
where:
MAGi(j) = magnitude of contribution at excitation frequency j from panel i (real)
Xi(j) = panel contribution to response at excitation frequency j from panel i (real or complex)
X(j) = total response at excitation frequency j (real or complex)
You can use the TOPP describer to specify the number of top contributing panels that the software
outputs for each excitation frequency and response combination. In the output, the panels are listed
from the highest contributor to the lowest contributor.
For example, if you specify TOPP = 10, for each excitation frequency and response combination, the
software outputs the 10 panels that contribute the most. If you specify TOPP = ALL, the software
outputs the contribution of all the panels listed from the highest contributor to the lowest contributor. If
you do not specify the TOPP describer, by default, the software outputs the 5 panels that contribute
the most.
You can use the ABS, NORM, or BOTH describers to specify if the output is normalized.
For example, specify NORM to output normalized panel contributions and specify ABS to output
non-normalized panel contributions. Specify BOTH to output both normalized and non-normalized
panel contributions.
For SORT1 output, the software calculates the normalized panel contributions as follows:

Chapter
requirements
where NMCi(j) is the normalized panel contribution to the response at excitation frequency j from
panel i.
For SORT2 output, at each panel and response (as specified by SETMC grid and DOF entries)
combination, the software ranks the contribution that the panel makes at each excitation frequency
(as specified by FREQ) to the response based on the results of the following equation:
where:
MAGi(j) = magnitude of contribution at excitation frequency j from panel i (real)
Xi(j) = panel contribution to response at excitation frequency j from panel i (real or complex)
Xi*(j) = complex conjugate of Xi(j)
X*(j) = complex conjugate of X(j)
N = total number of excitation frequency increments
You can use the TOPP describer to specify the number of excitation frequencies where the panel
contributes most to the response for the software to output for each panel and response combination.
In the output, the excitation frequencies are listed from the highest contributor to the lowest contributor.
For example, if you specify TOPP = 10, for each panel and response combination, the software
outputs the contribution of the panel at the 10 excitation frequencies where the contribution of the
panel is greatest. If you specify TOPP = ALL, the software outputs the contribution of the panel at
all frequencies listed from the highest contributor to the lowest contributor. If you do not specify the
TOPP describer, by default, the software outputs the contribution of the panel at the 5 excitation
frequencies where the contribution of the panel is greatest.
For example, specify NORM to output normalized panel contributions and specify ABS to output
non-normalized panel contributions. Specify BOTH to output both normalized and non-normalized
panel contributions.
For SORT2 output, the software calculates the normalized panel contributions as follows:

where NMCi(j) is the normalized panel contribution to the response at excitation frequency j from
panel i.
Example SORT1 input:
...
$ CASE CONTROL
SETMC 99 = PRES/198(T1)
SET 500 = TOP, BOTTOM
PANCON(SORT1,PRINT,ABS,TOPP=3,PANEL=500) = 99
FREQ = 1
...
BEGIN BULK
FREQ1,1,86.0,27.0,2
...
ENDDATA
Example SORT1 output:

...
COMPLEX ABSOLUTE PANEL CONTRIBUTIONS
FREQUENCY = 8.600000E+01 PRESSURE, GRID = 198
CONTRIBUTIONS FROM STRUCTURAL MODES
PANEL REAL IMAGINARY
TOTAL -1.649896E-09 -1.889964E-08

BOTTOM 8.855658E-10 -5.356264E-09
COMPLEX ABSOLUTE PANEL CONTRIBUTIONS

PANEL REAL IMAGINARY
TOTAL -9.694239E-09 -1.137355E-08

BOTTOM -1.736440E-09 -3.504814E-09
TOP -3.024307E-09 -2.259977E-09
...
GRDCON case control command
The GRDCON case control command output quantifies how the motion of grid points in structural
panels contributes to the acoustic response at a specific location in the fluid.
When you use the GRDCON case control command, you can specify whether the output is in SORT1
or SORT2 format and whether the output is normalized.
For SORT1 output, at each excitation frequency (as specified by FREQ) and response (as specified
by SETMC grid and DOF entries) combination, the software ranks the contribution that each grid in a
structural panel makes to the response based on the results of the following equation:
where:
MAGi(j) = magnitude of contribution at excitation frequency j from grid i (real)

Chapter
requirements
Xi(j) = grid contribution to response at excitation frequency j from grid i (real or complex)
You can use the TOPG describer to specify the number of top contributing grids that the software
outputs for each excitation frequency and response combination. In the output, the grids are listed
from the highest contributor to the lowest contributor.
For example, if you specify TOPG = 10, for each excitation frequency and response combination,
the software outputs the 10 grids that contribute the most. If you specify TOPG = ALL, the software
outputs the contribution of all the grids listed from the highest contributor to the lowest contributor. If
you do not specify the TOPG describer, by default, the software outputs the 5 grids that contribute
the most.
For example, specify NORM to output normalized grid contributions and specify ABS to output
non-normalized grid contributions. Specify BOTH to output both normalized and non-normalized
grid contributions.
For SORT1 output, the software calculates the normalized grid contributions as follows:
where NMCi(j) is the normalized grid contribution to the response at excitation frequency j from grid i.
For SORT2 output, at each grid and response (as specified by SETMC grid and DOF entries)
combination, the software ranks the contribution that the grid makes at each excitation frequency (as
specified by FREQ) to the response based on the results of the following equation:
where:
MAGi(j) = magnitude of contribution at excitation frequency j from grid i (real)
Xi(j) = grid contribution to response at excitation frequency j from grid i (real or complex)
Xi*(j) = complex conjugate of Xi(j)
X*(j) = complex conjugate of X(j)
N = total number of excitation frequency increments
You can use the TOPG describer to specify the number of excitation frequencies where the grid
contributes most to the response for the software to output for each grid and response combination. In
the output, the excitation frequencies are listed from the highest contributor to the lowest contributor.

For example, if you specify TOPG = 10, for each grid and response combination, the software
outputs the contribution of the grid at the 10 excitation frequencies where the contribution of the
grid is greatest. If you specify TOPG = ALL, the software outputs the contribution of the grid at all
frequencies listed from the highest contributor to the lowest contributor. If you do not specify the
TOPG describer, by default, the software outputs the contribution of the grid at the 5 excitation
frequencies where the contribution of the grid is greatest.
For example, specify NORM to output normalized grid contributions and specify ABS to output
non-normalized grid contributions. Specify BOTH to output both normalized and non-normalized
grid contributions.
For SORT2 output, the software calculates the normalized grid contributions as follows:
where NMCi(j) is the normalized grid contribution to the response at excitation frequency j from grid i.
Example SORT1 input:
...
$ CASE CONTROL
SETMC 99 = PRES/198(T1)
SET 700 = 10544 10305 10110 10777
GRDCON(SORT1,PRINT,ABS,TOPG=2,GRID=700) = 99
FREQ = 1
...
BEGIN BULK
FREQ1,1,86.0,27.0,2
...
ENDDATA
Example SORT1 output:

...
COMPLEX ABSOLUTE GRID CONTRIBUTIONS
GRID REAL IMAGINARY
TOTAL -5.846166E-10 -1.356268E-09

10305 3.565532E-15 7.451980E-15
10777 -5.422780E-17 -7.540371E-17
COMPLEX ABSOLUTE GRID CONTRIBUTIONS

GRID REAL IMAGINARY
TOTAL -7.430111E-10 -4.562482E-10

10305 2.862127E-15 -7.918542E-15
10777 -1.468421E-17 1.060137E-17
...

Chapter
requirements
Incident and transmitted acoustic power and acoustic power transmission loss output
For direct and modal acoustic frequency response analysis, you can request results output for
acoustic power, incident acoustic power, transmitted acoustic power, and acoustic power transmission
loss in deterministic and random acoustic analysis.
To output acoustic power for 2D microphone surfaces and AML regions, use the ACPOWER case
control command. Because acoustic power results include both incident and scattered (reflected)
components, you can use the INPOWER, TRPOWER, and TRLOSS case control command to
examine acoustic results in more detail and compute acoustic transmission loss.
• Use INPOWER to isolate the acoustic power that is incident on a surface like a 2D microphone
mesh or a set of acoustic free faces (BSURFS) and that is attributable to acoustic sources like
monopoles or acoustic plane waves, but not reflections.
• Use TRPOWER to request the acoustic power that is transmitted through an automatically
matched layer (AML) region or a 2D microphone mesh that includes both incident and scattered
components.
• Use TRLOSS to request the acoustic power transmission loss. The software calculates the
acoustic power transmission loss from the INPOWER and TRPOWER results as follows:
Calculating acoustic power transmission loss through a structural panel
As an example, you can use TRLOSS to calculate the acoustic power transmission loss through a
structural panel.
In a laboratory setting, you can measure the acoustic power transmission loss through a structural
panel by mounting the panel in an aperture of a wall between two rooms. One room is reverberant
and contains a diffuse acoustic plane wave source. The other room is anechoic and has no pure
acoustic sources. The anechoic room also contains the instrumentation for measuring the acoustic
power. Because the room is anechoic, you measure the acoustic power that radiates from the panel,
and the measurement is not influenced by reflections.
A simulation for such a laboratory setup is shown below:

(1) AMLREG1 (reverberant

side)
(2) Structural panel
(3) Acoustic free faces
(reverberant side)
(4) AMLREG2 (anechoic
side)
Figure 5-2. Panel test
In the simulation, both rooms are represented by fluid meshes using AML regions. AMLREG1
represents the reverberant room where the fluid domain opens to the half space in front of the panel.
The acoustic power source within the reverberant room can be modeled by a set of random plane
waves typically outside the fluid mesh. AMLREG2 represents the anechoic room where the fluid
domain opens to the half space behind the panel.
1. Use INPOWER to calculate the incident acoustic power on the free acoustic faces on the
reverberant room side that couple with the panel.
2. Use TRPOWER to calculate the acoustic power flowing through AMLREG2.
3. Use TRLOSS to calculate the acoustic power transmission loss through the panel.
Random results
If you request random results with the INPOWER, TRPOWER, and TRLOSS case control commands,
only real output is supported.
Limiting acoustic output to peak responses

In a SOL 108 or SOL 111 acoustic or vibro-acoustic frequency response analysis, you can limit the
output to frequencies where peak responses occur. This capability is called peakout.
For more information on peakout, see the Limiting Output to Peak Responses section of the
Simcenter Nastran Basic Dynamic Analysis User's Guide.
5.3 Bulk entries

The following describes the bulk entries used for acoustic solutions.
5.3.1 Acoustic grids

GRID bulk data entries (see GRID in the Simcenter Nastran Quick Reference Guide) are used to
specify the geometric coordinates and coordinate system of the fluid degrees of freedom. Acoustic
fluid grids are defined with CD = -1 and acoustic elements reference only fluid grids. At the boundary

Chapter
requirements
between the structure and the fluid, the grids can be coincident. However, structural elements cannot
connect to fluid grids and vice versa.
5.3.2 Acoustic elements

Acoustic elements are 3-D and are modeled using the CHEXA, CTETRA, CPENTA, and CPYRAM
elements. These elements reference a PSOLID entry ( see also PSOLID in the Simcenter Nastran
Quick Reference Guide) to define the physical property.
The FCTN field must be set to “PFLUID” for acoustic elements. In addition the MID field must
reference a MAT10, MAT10C, MATF10C, or MATPOR material.
The 2-D elements CTRIA3 and CQUAD4 are used to define microphone elements (see Microphone
mesh). In addition, you can define CROD and MICPNT to define a 1-D and 0-D element respectively.
The CROD, CTRIA3, and CQUAD4 elements in this case must reference PMIC physical property
5.3.3 Acoustic material definition

Material properties for acoustic elements are defined using MAT10, MAT10C, MATF10C, or MATPOR
bulk entries.
In the MAT10 entry, the density is specified in the RHO field and the speed of sound in the C field.
Porous materials with equivalent properties can be specified using GE and GAMMA fields, and the
frequency-dependent tables via the continuation line entries. Equations
and
(see Frequency-dependent acoustic materials or absorbers) that define ρs, βs, Ge, and γ are then
specified as entries in the MAT10.
For detailed information on MAT10, MAT10C, MATF10C, or MATPOR, see the Simcenter Nastran
Quick Reference Guide.
5.3.4 Acoustic panel normal velocity

You use the ACPNVEL bulk entry to specify the acoustic panel normal velocity boundary. ACPNVEL
supports the definition of:
• Magnitude and phase of acoustic velocity.
• Constant or frequency-dependent complex velocity.
5.3.5 Transfer admittance

You use the ACTRAD bulk entry that references PACTRAD to specify the transfer admittance
between two sets of acoustic free faces.

5.3.6 Impedance or admittance

Frequency-dependent absorbers are specified using CAABSF elements (see CAABSF in the
Simcenter Nastran Quick Reference Guide). CAABSF allows POINT, LINE, or AREA impedance
to be specified on the free fluid surface. The grids G1 through G4 can reference only fluid grids.
The CAABSF elements define the element topology and in turn reference the physical property
PAABSF or PAABSF1.
The PAABSF physical property table is used to specify constant complex impedance or
frequency-dependent impedance, or a combination of both.
The PAABSF1 physical property table is used to specify constant real and imaginary impedance and
admittance on a 2D fluid free surface.
5.3.7 Acoustic loads, boundary condition, and sources

• Use the ACSRCE bulk entry (see ACSRCE in the Simcenter Nastran Quick Reference Guide)
to define the acoustic source loading where you define the source by acoustic power as a
function of frequency. Alternatively, you can directly define acoustic volume acceleration using
the RLOAD1 or RLOAD2 entries.
• Use the ALOAD bulk entry to specify a combination of the following:

o Monopole source using the ACPOLE1 bulk entry.
o Dipole source using the ACPOLE2 bulk entry.
o Plane wave source using the ACPLNW bulk entry.
o Enforced acoustic pressure load with complex data input using the ACPRESS bulk entry.
o Panel normal velocity boundary condition using the ACPNVEL bulk entry.
Note
The ALOAD bulk entry does not support scaling of loads.
5.3.8 Modeling structure-acoustic interface

The ACMODL bulk entry (see ACMODL in the Simcenter Nastran Quick Reference Guide) can be
used to model the structure-acoustic interface. The search parameters for coupling are specified via
NORMAL and OVLPANG fields. The units for the search parameters can be in absolute or relative
units (specified using SRCHUNIT field). Structural and fluid elements that need to be coupled can be
specified using FSET and SSET fields respectively. The input can be a set of grids, elements, or
physical properties. Simcenter Nastran converts the input into relevant elements.
Note
Because a FSET or SSET field can select a SET3 bulk entry, you can also use 2D PLOTEL
elements in the ACMODL bulk entry. PLOTEL elements are elements that you can add to
your model for plotting and visualization in a preprocessor, such as Simcenter 3D.

Chapter 6: Acoustic and vibro-acoustic analyses
6.1 Overview of acoustic and vibro-acoustic analyses

• The 3-D elements CHEXA, CPENTA, CPYRAM, or CTETRA can be used to model the fluid.
• The 2-D elements CTRIA3 and CQUAD4 are used to model microphone elements.
• The rod element CROD is used to model the 1-D microphone element.
• The MICPNT element can be used for modeling a 0-D microphone element or microphone point.
• 3-D frequency-dependent absorbers, in which the absorptive properties are included on the
MAT10 or MAT10C bulk entry (fluid material).
• 2-D/1-D frequency-dependent absorbers (CAABSF).
• In a coupled vibro-acoustic problem, the interface between the fluid and structure can be modeled
with coincident or non-coincident unconnected grids. The fluid grids and structural grids can be
coincident but unconnected.
Note
For more information on various coupling options, see Defining the fluid-structure
interface boundary condition.
• Coupled vibro-acoustic analysis is available in the direct dynamic solution sequences 107 through
109 and 200 with ANALYSIS = DFREQ, and the modal dynamic solution sequences 103, 110,
111, 112, and 200 with ANALYSIS = MFREQ and ANALYSIS = MTRAN. Note that in solutions
103, 110, 111, 112, and 200, the normal modes are computed separately for the fluid and
structural parts of the model. In other words, the uncoupled modes of the fluid and structure are
used in the modal formulation of the stiffness, mass, and damping. In SOLs 110 through 112,
the SDAMPING Case Control command and the parameters G and W3 are applied only to the
structural portion of the model. Design Sensitivities may be computed in SOL 200.
6.1.1 Defining the fluid

You define fluid grid points by specifying a value of -1 for CD in field 7 on the GRID bulk entry.
Fluid elements are defined using the CHEXA, CPENTA, CPYRAM, and CTETRA bulk entries.
However, in the PSOLID entry, you must specify the character value PFLUID for FCTN in field 8 and
MID, field 3, must reference a MAT10, MAT10C, MATF10C, or MATPOR material entry.
The MAT10 entry defines the bulk modulus and the mass density properties of the fluid, along with
optional absorber properties.

Chapter
Chapter 6: 6: Acoustic
Acoustic and vibro-acoustic
and vibro-acoustic analyses
analyses
Microphone elements, which are also fluid elements, are defined using the CHEXA, CPENTA,
CPYRAM, CTETRA, CTRIA3, CQUAD4, CROD, and MICPNT bulk entries. Except for MICPNT,
these bulk entries must reference the PMIC bulk entry for the physical property.
Additionally, you can model different fluid domains and connect them using a surface-to-surface
glue definition.
Note
Surface-to-surface gluing can be used between acoustic mesh faces. Similar to structural
glue conditions, the meshes between the acoustic glued faces can be dissimilar. The
acoustic-to-acoustic glue inputs are consistent with the structural-to-structural capability,
except that on the BGPARM bulk entry, GLUETYPE, PENTYPE, and PENT are ignored.
PENN is used to calculate the acoustics penalty matrix [K].
Refer to the Gluing Elements chapter in the Simcenter Nastran User's Guide for more
information.
6.1.2 Absorber elements

The acoustic absorber elements are special elements needed to model soundproofing materials on
the structural surface.
See Impedance or admittance for more information.
6.1.3 Acoustic loading

Loads on fluid elements are, in most cases, analogous to enforced displacements on structural
elements. The types of loading available for fluid elements are detailed in Defining acoustic loads.
6.1.4 Single-point constraints

Single-point constraints of the fluid (P = 0.0) may be enforced on the fluid boundary using the
SPC entry or the PS field on the GRID entry. The fluid pressure degree-of-freedom is defined as
component 1 on the SPC entry.
6.1.5 Output requests (overview)

The following output can be requested on the fluid:
• Pressure and peak sound pressure levels at the fluid points, sound pressure level in dB and dBA.
• RMS sound pressure level in some Nastran solutions.
• Separate output for the fluid and structural portion of the model.
• Particle velocity at element centroid.
• Modal and panel contribution output.
• Equivalent radiated power for some shell elements and output equivalent radiated power for
panels that contain shell elements.

Acoustic and vibro-acoustic analyses
• Coupling information files with the FLSTCNT case control command.
See the Output requests and Fluid-structure interaction control sections in the Input Data
Requirements (Summary) chapter for more information.
6.1.6 Vibro-acoustic coupling punch output

You can request the output of the vibro-acoustic coupling matrix AGG to the punch file with the
AGGPCH describer on the FLSTCNT case control command.
For more information, see Fluid-structure interaction control.
6.1.7 Vibro-acoustic coupling

The model for fluid/structural systems has an explicit FE mesh for both the structural components and
the interior fluid cavity. The equation of motion for the coupled system is shown by equation
(see Coupling of fluid and structure (vibro-acoustics)).

The coupling algorithm uses a multi-step approach, which is detailed in Two-way coupling.
6.1.8 Superelement analysis for vibro-acoustic analysis

Superelements may be used with the restrictions listed in the Superelements section.
6.1.9 Running the job

Diagnostic messages
Diagnostic messages are generated automatically by the software. The following messages indicate
the error factors in area for free body motions. Nonzero numbers indicate a hole in the model as
indicated by the T2 value.
^^^ DMAP INFORMATION MESSAGE 9055 (SEMG) - THE FL./STR. INTERFACE CHECK IS FORCES
AND MOMENTS RESULTING FROM A UNIT INCREASE IN PRESSURE, OR CHANGES IN THE FLUID PRESSURE
RESULTING FROM RIGID BODY MOTIONS OF THE STRUCTURE. THESE VALUES ARE DIRECTLY PROPORTIONAL
TO THE OPEN SURFACE OF THE FLUID.
FL./STR. INTERFACE CHECK
T1 T2 T3 R1 R2 R3
1 2.3554080E-18 -8.4260993E-02 -1.4653091E-28 6.0271138E-10 2.6100844E-19 1.8405548E-10
Diagnosing problems
Some recommended techniques are as follows:
• Carefully check the special diagnostic messages. Holes in the boundaries cause nonzero
resultant area factors.
• You can include the SKINOUT describer on the FLSTCNT case control command to request
coupling information files. See Requesting coupling information files below.
• Run tests without the boundary coupling to estimate the frequencies.

Chapter
analyses
• Temporarily switch to an alternate eigenvalue method and/or a smaller range to ensure modes
are correct.
• If a model is large, try subdividing it into smaller sections that can be checked more conveniently.
Requesting coupling information files

You can include the SKINOUT describer on the FLSTCNT case control command to request debug
files. Depending on the option used, a punch file *.pch and/or debug data deck *.dat is created. The
resulting files have _acdbg appended to their name. For example, if the original input file is test.dat,
the debug files are test_acdbg.dat and test _acdbg.pch.
The SKINOUT input options are described in detail in the Fluid-structure interaction control section in
the Input Data Requirements chapter.
6.2 Loudspeaker example (simplified version)
Figure 6-1. Acoustic suspension loudspeaker

Test problem description
A simple test problem illustrates the fundamentals of acoustic analysis for enclosed containers
with flexible walls. The physical problem represents an initial attempt at design and analysis of
an enclosure for an acoustic suspension loudspeaker system. The objective was to calculate the
resonant frequencies and responses of the system without the mass and impedance of the speaker
coil and magnet components.

Physical description
The structure consists of a simple rectangular box with wood walls as shown by the plate model in
Figure 6-1. A single cutout is provided for the speaker and a thin polyethylene cone was modeled
with triangular shells. The physical properties of the model are listed below. Note the use of the MKS
system of units, which illustrates the flexibility of Simcenter Nastran.
Width: 0.5 M Depth: 0.4 M
Height: 0.6 M Hole Diam. 0.345 M
Box E Modulus: 11.61E9 Box Density: 562 Kg/M**3
Wall thickness: 0.015 M
Air Wave speed: 344 M/Sec Air Density: 1.11KG/M**3
Cone E Modulus: 3.4E9 Cone Density: 450.0
Cone thickness: 0.1E-3 M Cone Depth: 0.04 M
Executive and case control

The configuration shown below is a direct complex modal solution using the Lanczos method to
extract eigenvalues.
Bulk data
The key data of note in the Bulk Data section are as follows:
• ACMODL bulk entry is not explicitly listed in the input deck. However, this means that Simcenter
Nastran uses internally ACMODL entry with defaults.
• The speaker cone consists of a light, thin plastic material, and the box is made from wood.
• The cone was meshed with CTRIA3s and the box with CQUAD4s, while CHEXAs represent
the air.
Results from Simcenter Nastran

Three runs (out of many) are described below. They illustrate a recommended sequence for the
analysis process.
Run 1: Real eigenvalue analysis
Before the coupled structure is analyzed, it is important to understand the behavior of the structural
and fluid models separately. Fortunately they can be included in the same data file and the Real
Eigenvalue Solution Sequence (SOL 103) may by used. The results for the uncoupled speaker box
and the acoustic modes are shown below. Note that the natural frequencies for both systems occur
in the same range, which indicates that the enclosure modes will interact with the acoustic cavity
resonances. Results for the acoustics were checked by one-dimensional wave solutions. Results for
the box were checked against calculated natural frequencies of simply-supported plates.
Mode no. Frequency, Hz Type
1 0.0 Air- Constant Pressure (fictitious)
2 292 Air- 1st z (up/down)
3 297 Box- u(y) Back panel
4 351 Air- 1st x (Left/Right)
5 359 Box- u(x) Sides in-Phase

Chapter
analyses
Mode no. Frequency, Hz Type

6 379 Box- u(x) Sides out-of-phase
7 450 Box- u(z) Top/bottom
8 453 Air- 1st y (Fore/Aft)
Run 2: Complex eigenvalue analysis, coupled modes

The coupled natural frequencies must be obtained from an unsymmetrical matrix equation which
requires a Complex Eigenvalue method even in the undamped case. The CLAN method is
recommended for most large-order non-superelement jobs. The only changes from the Real Modes
job was the addition of an ACMODL input and the changes in eigenvalue method.
Mode no. Frequency Type of dominant motion

1 0.0 Air- Constant Pressure
2 291 Air- u(z)
3 295 Coupled- u(y)- back panel
4 336 Coupled- u(x)- sides
5 371 Coupled- u(y)- cone
6 373 Coupled- u(x)- sides
7 445 Coupled- u(z)- top/bottom
8 453 Coupled- u(y)- cone
Results above are interpreted by examining the frequency shifts from the uncoupled system. The
modes with u(y) fore-aft motion are most likely to be excited by the speaker. Modes 1, 2, 3, 4, 6, and
7 are close to their uncoupled equivalent. Modes 5 and 8 are new combinations of higher modes.
The results were confirmed by several methods:
1. The job was rerun using the INV method on the EIGC Bulk Data and produced nearly identical
results.
2. A printout of the interface area matrix was obtained by using a DMAP alter and verified by hand
calculations.
3. The natural symmetry of the geometry produced symmetric and antisymmetric results relative to
the natural structural planes of symmetry (except for roots with close frequencies).
Excerpt of an input file demonstrating the 2nd run:

ACOUSTIC TEST PROBLEM DATA FILES
ID SPEAK39F,DNH
SOL 107
DIAG 8,12 $ PRINTS MATRIX TRAILERS AND ROOT-TRACKING MESSAGES.
CEND
TITLE=SPEAKER BOX -WITH CONE, SIMPLE CORNER SUPPORTS
SUBTITLE = COUPLED BOUNDARY, NON-MATCHING ELEMENTS
ECHO= UNSORT
SEALL=ALL
SPC=20
$ USES MKS SYSTEM
CMETHOD = 7 $LANCZOS
SET 20= 3,13,23,43,82,83,84,91,93,95,103,113,123,

131,135,153,163,171,173,175,183,
1013,1023,1043,1082,1083,1084,1091,1095,1113,
1131,1135,1163,1173,1183,1193
DISP= 20 $ FOR MINIMUM PRINTOUT
$ DISP(PLOT)=ALL
$
BEGIN BULK
$
PARAM,POST,0
PARAM,COUPMASS,1
$
$ BOX PROPERTIES - WOOD
MAT1,11,11.61+9,,0.3,562.0
PSHELL,1000,11,.015,11,,11
$
$ SPEAKER CONE
MAT1,3,3.4+9,,0.3,450.0
PSHELL,10,3,0.1-3,3,,,0.223
$
$ PROPERTIES OF AIR
MAT10,100,131.94+3,1.115
PSOLID,100,100,,2,,1,PFLUID
$
$ EIGEN METHODS
EIGR,20,MGIV,1.0,600.0
EIGC,7, CLAN,,,,,,,+CLAN
+CLAN,0.0,10.0,0.0,1600.0,100.0,,20
$ USE INVERSE POWER TO CHECK LOW ROOTS
EIGC,107,INV,MAX,,,,,,+EC1
+EC1,0.0,100.0,0.0,1800.0,100.0,12,9
$
$ FIX BOX AT BOTTOM CORNERS
SPC1,20,123,1001,1005,1031,1035
$ COORDINATE SYSTEM AT CENTER OF HOLE
CORD1C,83,283,113,85
$
$
$FLUID GRID POINTS
$ NOTE VALUE OF -1 IN FIELD 7 INDICATES 1 DOF.
GRID,1,, -.25, -.2, -.3,-1
GRID,2,,-.125, -.2, -.3,-1
GRID,3,, 0.0, -.2, -.3,-1
GRID,4,, .125, -.2, -.3,-1
GRID,5,, .25, -.2, -.3,-1
GRID,11,, -.25, -.0667, -.3,-1
GRID,12,,-.125, -.0667, -.3,-1
GRID,13,, 0.0, -.0667, -.3,-1
GRID,14,, .125, -.0667, -.3,-1
GRID,15,, .25, -.0667, -.3,-1
GRID,21,, -.25, 0.0667, -.3,-1
GRID,22,,-.125, 0.0667, -.3,-1
GRID,23,, 0.0, 0.0667, -.3,-1
GRID,24,, .125, 0.0667, -.3,-1
.25, -.2, 0.,-1
-
ETC.

Chapter
analyses
-
$ STRUCTURE GRIDS
GRID,1001,, -.25, -.2, -.3
GRID,1002,,-.125, -.2, -.3
GRID,1003,, 0.0, -.2, -.3
GRID,1004,, .125, -.2, -.3
GRID,1005,, .25, -.2, -.3
GRID,1011,, -.25, -.0667, -.3
GRID,1012,,-.125, -.0667, -.3,,6
GRID,1013,, 0.0, -.0667, -.3,,6
-
ETC.
-
$ OPTIONAL ASET DATA TO USE WITH HESS METHOD
$ASET1,1,1,THRU,195
$ASET1,123,1003,1012,1014,1022,1024,1033
$ASET1,123,1163,1172,1174,1182,1184,1193
$ASET1,123,1042,1043,1044,1081,1082,1083,1084
$ASET1,123,1065,1122,1123,1124
$ASET1,123,1072,1074,1112,1113,1114,1152,1154
$ASET1,123,1051,1061,1131,1141
$ASET1,123,1055,1065,1135,1145
$
$
$SOLID ELEMENTS FOR AIR
CHEXA 1 100 1 2 12 11 41 42 +001
+001 52 51
CHEXA 2 100 2 3 13 12 42 43 +011
+011 53 52
CHEXA 3 100 3 4 14 13 43 44 +021
+021 54 53
CHEXA 4 100 4 5 15 14 44 45 +031
+031 55 54
CHEXA 11 100 11 12 22 21 51 52 +041
+041 62 61
CHEXA 12 100 12 13 23 22 52 53 +051
+051 63 62
CHEXA 13 100 13 14 24 23 53 54 +061
+061 64 63
CHEXA 14 100 14 15 25 24 54 55 +071
+071 65 64
CHEXA 21 100 21 22 32 31 61 62
+081
+081 72 71
-
ETC
-
$STRUCTURAL ELEMENTS
CQUAD4 1001 1000 1001 1002 1012 1011
CQUAD4 1002 1000 1002 1003 1013 1012
CQUAD4 1003 1000 1003 1004 1014 1013
CQUAD4 1004 1000 1004 1005 1015 1014
CQUAD4 1011 1000 1011 1012 1022 1021
CQUAD4 1012 1000 1012 1013 1023 1022
CQUAD4 1013 1000 1013 1014 1024 1023

CQUAD4 1014 1000 1014 1015 1025 1024

-
ETC
-
$
$SIMPLE CONE
$
CTRIA3, 5042,10,1042,1043,1083
CTRIA3, 5043,10,1043,1044,1083
CTRIA3, 5044,10,1044,1084,1083
CTRIA3, 5084,10,1084,1124,1083
CTRIA3, 5124,10,1124,1123,1083
CTRIA3, 5123,10,1123,1122,1083
CTRIA3, 5122,10,1122,1082,1083
CTRIA3, 5082,10,1082,1042,1083
ENDDATA
Run 3: Complex eigenvalue analysis, modal formulation

A more efficient method for solving a coupled matrix problem is to reduce the size of the matrices
by using modal coordinates. In this case, both the structural displacements and fluid pressures are
replaced in the matrix solution by generalized coordinates representing the uncoupled real modes.
For the modal cases, the real modes (25 total) below 800 Hz. were used. In the coupled cases, 13
modes were obtained below 500 Hz. All results compared to within 1%.
On a Sun SPARCstation 1+ the CPU time comparisons are:
Type solution Solution size CPU time (seconds)

Uncoupled Modes 538 110.2
Direct, No Modes 538 586.3
Structural Modes Only 114 143.2
All Modes 25 106.9
The conclusion is that the modal method (SOL 110) reduces the costs with very little effect on
accuracy. It is recommended that you use a liberal number of real modes (twice the number of
coupled modes) to represent the system.
The complete input file spkall.dat for the modal method is available in
install_directory\nxnr\nast\misc\doc\advdynamics\.

Chapter 7: Acoustic analysis using automatically matched layer
7.1 Exterior acoustics using automatically matched layer

Modeling exterior acoustics using the finite element method (FEM) has some challenges. In
particular, a non-reflecting (anechoic) boundary condition is required at the free fluid boundary.
In previous releases, you defined acoustic absorbers on the free fluid mesh boundary with a
characteristic impedance (density x speed of sound). To prevent reflections with this method, this
boundary had to be several wavelengths from the vibrating source. As a result, these models tended
to be large, and perfect absorption was still difficult for a range of frequencies.
Another recommended method of representing the non-reflective acoustic boundary condition is the
Automatically Matched Layer (AML) method. The AML method uses a reflectionless artificial layer
that absorbs outgoing waves regardless of their frequency and angle of incidence. As depicted in the
following figure, the AML is defined on a convex shape boundary, which uses the AMLREG bulk entry.
Figure 7-1. Convex mesh/surface
An AML region can be modeled close to the vibrating structure or acoustic source with good accuracy
resulting in much smaller FE models.
The following example demonstrates how an AML can be specified to represent the radiation from
a vibrating gearbox.

Chapter
Acoustic analysis
analysis usingusing automatically
automatically matched
matched layerlayer
(1) Structural Mesh

(2) Fluid Mesh
(3) AML
Figure 7-2. Gearbox
The AML produces accurate results for the FE domain, yet the FE domain represents a small part of
the fluid, which in reality is infinite. You can request acoustic results at arbitrary locations exterior
or interior to the fluid. For output requests exterior to the meshed volume, Simcenter Nastran
uses the acoustic results at the boundary of the FE domain and a boundary integral to obtain the
acoustic response. These exterior and interior locations are defined with the microphone mesh.
See Microphone mesh.
The AML can be specified in a direct frequency response solution (SOL 108) or in a modal frequency
response solution (SOL 111). When an AML is defined in a modal frequency response solution, the
structure is reduced to modal coordinates, although the acoustic fluid remains in physical coordinates.
7.2 Assumptions and limitations of AML

• The AML boundary condition can be specified only in a direct frequency response SOL 108
or in a modal frequency response SOL 111.

Acoustic analysis using automatically matched layer
Note
Because the AML boundary condition creates frequency-dependent element matrices,
the structure is modeled using modal coordinates in SOL 111. The fluid, however,
is still retained in physical coordinates.
• The AML surface must be defined on fluid faces only.
• The AML surface must be convex.
• The AML region must not protrude through an infinite plane that represents a reflection boundary.
• The face of an AML surface must not overlap or intersect with an infinite plane.
• AML faces that contact the infinite plane must be perpendicular to it.
• The infinite plane must be parallel to the plane of two global axes within a tolerance.
For more information about infinite planes, see Infinite planes..
7.3 Convex AML surface

The AML surface must be convex so that modal behavior between parts of a structure can be
accounted for.
The following examples illustrate convex and non-convex AML surfaces.
Legend for all examples:
(1) Structural Mesh (grey)
(2) Fluid Mesh (blue)
(3) AML (green)
The waves radiating from the structure are

not absorbed, as no non-convex regions exist
between the two vibrating surfaces.
Figure 7-3. Convex AML surface

Chapter
Acoustic analysis
matched layerlayer
The waves radiating from the structure are

absorbed in the non-convex region between the
two vibrating surfaces.
Figure 7-4. Non-convex AML surface
7.4 AML faces and infinite plane
AML faces that contact the infinite plane must be perpendicular to it so that the solver computes
accurate results. The following examples illustrate perpendicular and non-perpendicular AML faces.
(3) AML (green)
(4) Microphone point (yellow)
(5) Infinite Plane (purple)
(6) Local surface normal (black)
(7) Outer boundary of reflectionless artificial layer (green dashes) formed by the solver when it
extrudes the AML surface.

This model produces an accurate solution.

• The modeled fluid domain matches a
half-space where both the FE mesh and
the reflectionless artificial layer respect the
infinite plane.
Note
In half-space, the sound radiation
is only half of the radiation in a
free field because of the acoustic
reflection boundary.
Figure 7-5. AML faces (3) that contact the
infinite plane are perpendicular • The solver accounts for all microphone points
outside the FE mesh including points close to
the infinite plane.
This model may not provide an accurate solution

for exterior microphone points because the
reflectionless artificial layer intersects the infinite
plane. This may also adversely affect the
accuracy of interior microphone point results.
Figure 7-6. AML faces (3) that contact the

infinite plane are not perpendicular.
This model may not provide an accurate solution

because of the following issues:
• The solver may not follow the domain
boundaries.
• The solver may not account for microphone

points that are outside the FE mesh and
close to the infinite plane.
Figure 7-7. AML faces (3) that contact the • For interior microphone points, the entire
infinite plane are not perpendicular and the reflectionless artificial layer is not in contact
entire reflectionless artificial layer is not in with the infinite plane.
contact with the infinite plane.

Chapter
Acoustic analysis
matched layerlayer
7.5 AML with non-homogeneous elements

The elements that define the AML surface do not have to be homogeneous.
Although the acoustic results derived in the interior will be still accurate, the results on the exterior will
not be. This is because Simcenter Nastran computes the acoustic radiation on the exterior using a
free space Green’s function that assumes homogeneity.
7.6 AML and fluid damping coefficient

The software supports the definition of a user-specified, uniform fluid damping coefficient with the
GFL parameter, even when an automatically matched layer (AML) is present.
In a coupled vibro-acoustic analysis, you apply GFL to the fluid portion of a model.
For more information on the GFL parameter, see G - Parameters.
7.7 Microphone mesh

Microphone mesh
Microphone locations can be:

• Exterior to the convex AML boundary.
• Interior to the convex AML boundary.
• On the convex AML boundary.
When a microphone mesh is:

• Interior or on the AML boundary, Simcenter Nastran interpolates the results from the fluid grids
to the microphone location.
• Exterior to the AML boundary, Simcenter Nastran uses the pressure and velocity on the AML
(or on the physical boundary if requested), and a boundary integral, to compute the results at
the exterior microphone locations.
For information on output requests and microphone mesh definitions, see Defining microphone
meshes.
The following example demonstrates a 2D microphone mesh.

(1) 2D microphone mesh
Figure 7-8. Microphone mesh
Infinite planes
Acoustic results for microphone meshes exterior to the AML region are derived using boundary
integrals. You can define additional boundaries called infinite planes exterior to AML region to
influence the derived acoustic results.
The planes are created with the IPLANE bulk entry, which includes the TYPE field to designate them
as a zero velocity or a zero pressure acoustic reflection boundary. You then select up to three infinite
planes with the AMLREG entry. If you select multiple infinite planes, they must be perpendicular
to one another.
• An infinite plane with TYPE = 0 defines a rigid, reflective boundary in which the velocity is zero.
This is also known as a symmetric acoustic boundary. For example, air-to-ground is a zero
velocity reflective boundary.
• An infinite plane with TYPE = 1 defines a pressure release reflective boundary in which the
pressure is zero. This is also known as an anti-symmetric acoustic boundary. For example,
air-to-water is a zero pressure reflective boundary.
You should only define a microphone mesh on the side of an infinite plane in which the fluid elements
are defined. The software will not compute results on the non-fluid side.

Chapter
Acoustic analysis
matched layerlayer
Location of pressure and pressure gradients
You can use the RADSURF field on the AMLREG entry to select the location of pressure and pressure
gradients for computing the acoustic results on microphone locations exterior to the AML region.
• When RADSURF = AML, the pressure and pressure gradients at the AML are used to compute
the results exterior to the AML region.
• When RADSURF = PHYB, the pressure and pressure gradients on the physical boundary are
used to compute the results exterior to the AML region. The physical boundary is defined as all
fluid free faces excluding those on the AML and those on an infinite plane.
The following examples illustrate the use of AML, RADSURF (field of the AMLREG bulk entry),
and infinite plane.
(3) AML (green)
(4) RADSURF (red)
(5) Infinite Plane (purple)
• RADSURF = AML surface
• No infinite plane
Figure 7-9. Acoustic radiation from a machine
• RADSURF (red) = AML
• FE fluid domain
modeled to the infinite
plane implies AML
surface is not closed
Figure 7-10. Acoustic radiation from a
machine

• RADSURF (red) = PHYB

(should yield same results
as with RADSURF = AML)
Figure 7-11. Acoustic radiation from a

machine
• Duct ends at the right end

and continues indefinitely
• RADSURF is not needed as

interest is on what happens
Figure 7-12. Anechoic duct inside the duct
• RADSURF (red) = AML
• RADSURF is needed,
if the results are required
outside FE domain
Figure 7-13. Open duct at the right end
• Duct ends in a baffle

(ground, wall, and so on)
acting as an infinite plane
Figure 7-14. Open duct at the right end and
ends in a baffle

Chapter
Acoustic analysis
matched layerlayer
• No RADSURF at the open

duct on the right end
• RADSURF is only required

if sound is outside the
conventional fluid domain
Figure 7-15. Open duct at the right end and

top duct ends in a baffle
7.8 Acoustic computation process

The computation for acoustic results in the conventional FE domain and at microphone points
involves two steps:
1. The results are computed for only the conventional FE domain.
2. In a post-processing step in Simcenter Nastran, the results at microphone points are computed
as follows:
• Points inside or on the conventional FE domain:
Nodal results are used to interpolate acoustic pressure results at fluid grid points referenced
by microphone elements.
• Points outside the conventional FE domain:

Kirchhoff-Helmholtz boundary integral is used in conjunction with the acoustic pressure and
the gradients from either the AML surface or from the physical boundary.
The choice of using the AML boundary or physical boundary is user selectable.

7.9 Gearbox example
Figure 7-16. Acoustic radiation from a vibrating gearbox

Test problem description
A test problem illustrates the fundamentals of an exterior acoustic analysis using an automatically
matched layer. The physical problem represents an initial attempt at design and analysis of an
acoustic radiation from a vibrating gearbox. The objective was to carry out a weakly coupled
vibro-acoustic response due to the vibrations of the gearbox.
Frequency-dependent forces and moment loads on the structural mesh were applied at the
connection between the gearbox shafts and the housing parts, and nodes on the gearbox flange were
constrained to simulate a connection between the gearbox and engine. However, only the forces
coming from the engine shaft were considered.
Physical description
The ready-to-be-solved structure consisted of a simplified gearbox made from steel without internal
shafts and gears as shown by the model in the following figure.

Chapter
Acoustic analysis
matched layerlayer
(1) Structural Mesh

(2) Fluid Mesh
(3) AML
Figure 7-17. Gearbox
The microphone mesh as shown in the following figure was used to place microphones in space
(exterior to the convex AML boundary), where acoustic pressure results were requested.

(1) 2D microphone mesh
Figure 7-18. Microphone mesh
Some physical properties of the model are listed below.
Material Properties
Air Density: 1.225 kg/m^3 Speed of Sound: 340 m/sec
Steel Density: 7.829e-006 kg/mm^3 E Modulus: 2.0694e+008 mN/mm^2
Executive and case control

The SKINOUT describer on the FLSCNT case control command was used to request coupling
information. Several subcase runs with different RPMs using the same model were made to
investigate the effects of increased RPM. The configuration shown below was a direct frequency
response SOL 108 solution.
Bulk data
The key data of note in the Bulk Data section are as follows:
• The FREQ1 card was used to apply several forcing frequencies ranging from 1000 to 2000Hz to
the model and all subcases.
• The structure and the fluid were meshed with linear tetrahedral solids, and the spherical
microphone mesh was modeled with linear quadrilaterals.
• The PMIC property was used as dummy property for the microphone elements.

Chapter
Acoustic analysis
matched layerlayer
• The structure was made from steel and the fluid material was air with a constant mass density
and speed of sound.
• The CTYPE describer on the ACMODL card, where you can set fluid-structure interface modeling
parameters, was used to specify the fluid-structure coupling type.
• The AML was defined on a convex shape boundary using the AMLREG bulk entry. The SID field
of the AMLREG bulk entry referenced a BSURFS bulk entry. The element faces on the BSURFS
bulk entry were all on fluid elements.
Results from Simcenter Nastran

Acoustic Power [microW] radiated through the AML at 3000 RPM
Frequency, Hz
Real Imaginary
1000 7.894135E+00 3.093594E+00
1100 5.114406E+00 9.429158E-01
1200 6.318355E+03 8.360485E+02
1300 4.341093E+01 5.550558E+00
1400 1.160105E+01 4.358886E-01
1500 1.253860E+02 -8.275199E-01
1600 3.911594E+00 -1.572538E-02
1700 3.019994E+02 1.949584E+00
1800 6.132312E+00 4.249183E-01
1900 3.751575E+01 1.073381E+00
2000 4.130656E+01 -1.386058E+00
Excerpt of the input file:

$*
$* Simcenter Nastran VERSION 11.0
$*
$* ANALYSIS TYPE: Vibro-Acoustic
$* SOLUTION TYPE: SOL 108 Direct Frequency Response
$* SOLVER INPUT FILE: gearbox.dat
$*
$*
$* UNITS: mm (milli-newton)
$* ... LENGTH : mm
$* ... TIME : sec
$* ... MASS : kilogram (kg)
$* ... TEMPERATURE : deg Celsius
$* ... FORCE : milli-newton
$* ... THERMAL ENERGY : mN-mm (micro-joule)
$*
NASTRAN FREQVM=1
ID,NASTRAN,gearbox
SOL 108
CEND
$*
$*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$*
$* CASE CONTROL

$*
$*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$*
ECHO = NONE
FLSTCNT ACSYM=YES ACOUT=PEAK ASCOUP=YES PREFDB=2.0-8 SKINOUT=FREEFACE ,
AGGPCH=NO SFEF70=NO
SPC = 181
OUTPUT
ACPOWER(AMLREG=ALL,PRINT) = YES
DISPLACEMENT(PLOT,REAL) = ALL
$* Step: RPM=3000 - Data Source 1
SUBCASE 1
LABEL = RPM=3000 - Data Source 1
DLOAD = 401
FREQUENCY = 201
SUBCASE 2
DLOAD = 402
FREQUENCY = 201
SUBCASE 3
DLOAD = 403
FREQUENCY = 201
$*
$*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$*
$* BULK DATA
$*
$*$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$$
$*
BEGIN BULK
$*
$* SOLUTION CARDS
$*
$* Modeling Object: Forcing Frequencies - Direct1
FREQ1 201 1000.00100.0000 10
$*
$* PARAM CARDS
$*
PARAM G 0.0000
PARAM GFL 0.0000
PARAM K6ROT100.0000
PARAM OIBULK YES
PARAM OMACHPR YES
PARAM POST -2
PARAM POSTEXT YES
PARAM SPCSTR NO
PARAM UNITSYS MN-MM
$*
$* GRID CARDS
$*
...

Chapter
Acoustic analysis
matched layerlayer
$*
$* ELEMENT CARDS
$*
...
$* Mesh Collector: Solid(1)

$* Mesh: 3d_mesh(1)
CTETRA 178102 1 53999 61516 66858 61413
CTETRA 178103 1 53999 61516 61413 54954
...
$* Mesh Collector: AcousticalMesh_ForTutorial_finished_Microphone Surface(1)

$* Mesh: AcousticalMesh_ForTutorial_finished_Sphere Primitive Mesh (1)
CQUAD4 469922 19 111668 111672 111673 111669
CQUAD4 469923 19 111668 111669 111685 111684
...
CQUAD4 469974 19 111706 111722 111723 111707

CQUAD4 469975 19 111703 111707 111723 111719
$*
$* PROPERTY CARDS
$*
$* Property: PSOLID1
PSOLID 1 1 0 SMECH
$* Property: AcousticalMesh_ForTutorial_finished_PSOLID - Acoustic Fluid1
PSOLID 18 5 PFLUID
$* Property: AcousticalMesh_ForTutorial_finished_PMIC1
PMIC 19
$*
$* MATERIAL CARDS
$*
$* Material: Steel
MAT1 12.0694+8 0.2880007.8290-61.1280-5
MATT1 1 1 2 3
TABLEM1 1 1 +
+ 20.00002.0694+8 21.11002.0694+8 ENDT
TABLEM1 2 1 +
+ 20.00000.288000 21.11000.288000 23.89000.288000 37.78000.288000+
+ 51.67000.289000 65.56000.289000 79.44000.290000 93.33000.290000+
+ 107.22000.291000121.11000.291000135.00000.291000148.89000.292000+
+ 162.78000.292000176.67000.293000190.56000.293000204.44000.293000+
+ 218.33000.294000232.22000.294000246.11000.294000260.00000.295000+
+ 273.89000.295000287.78000.296000301.67000.296000315.56000.296000+
+ 329.44000.297000343.33000.297000357.22000.298000371.11000.298000+
+ 385.00000.298000398.89000.299000412.78000.299000426.67000.300000+
+ 440.56000.300000454.44000.301000468.33000.301000482.22000.302000+
+ 496.11000.302000510.00000.303000523.89000.304000537.78000.304000+
+ 551.67000.305000565.56000.306000579.44000.307000593.33000.308000+
+ 607.22000.309000621.11000.310000635.00000.311000648.89000.312000+
+ 662.78000.313000676.67000.314000690.56000.316000704.44000.317000+
+ ENDT
TABLEM1 3 1 +
+ 20.00001.1280-5 93.33001.1790-5107.22001.1880-5121.11001.1970-5+
+ 135.00001.2060-5148.89001.2132-5162.78001.2222-5176.67001.2312-5+

+ 190.56001.2402-5204.44001.2492-5218.33001.2582-5232.22001.2672-5+
+ 246.11001.2762-5260.00001.2852-5273.89001.2942-5287.78001.3032-5+
+ 301.67001.3122-5315.56001.3212-5329.44001.3302-5343.33001.3392-5+
+ 357.22001.3464-5371.11001.3554-5385.00001.3644-5398.89001.3734-5+
+ 412.78001.3806-5426.67001.3896-5440.56001.3986-5454.44001.4058-5+
+ 468.33001.4148-5482.22001.4238-5496.11001.4310-5510.00001.4400-5+
+ 523.89001.4472-5537.78001.4544-5551.67001.4616-5565.56001.4688-5+
+ 579.44001.4742-5593.33001.4796-5607.22001.4832-5621.11001.4886-5+
+ 635.00001.4904-5648.89001.4922-5662.78001.4922-5676.67001.4922-5+
+ 690.56001.4904-5704.44001.4886-5718.33001.4850-5732.22001.4796-5+
+ 746.11001.4724-5760.00001.4634-5773.89001.4472-5787.78001.4256-5+
+ 801.67001.4004-5815.56001.3680-5 ENDT
$* Material: Air_acoustic
MAT10 5 1.2250-93.4000+5
$*
$* FLUID-STRUCTURE INTERFACE
$*
$* Modeling Object: Fluid-Structure Interface Modeling Parameters1
ACMODL 1.000000 60.0000 REL+
+ 0.500000 0 WEAK
$*
$* LOAD AND CONSTRAINT CARDS
$*
$* Constraint: Fixed(1)
SPC 181 68337 123456 0.0000
SPC 181 68338 123456 0.0000
SPC 181 68339 123456 0.0000
SPC 181 68340 123456 0.0000
SPC 181 68341 123456 0.0000
SPC 181 68342 123456 0.0000
SPC 181 68343 123456 0.0000
SPC 181 68344 123456 0.0000
SPC 181 68348 123456 0.0000
SPC 181 68349 123456 0.0000
SPC 181 68350 123456 0.0000
SPC 181 68351 123456 0.0000
$*
$* Simulation Object: Automatically Matched Layer(1)
AMLREG 182 1Automatically Matched Layer(1) +
+ 5 AML 0 0 0
$* Region: AmlRegion1
BSURFS 1 254119 79946 81121 79948+
+ 254222 79416 79418 79419 254283 78598 78594 78596+
+ 254307 78655 78656 78657 254337 82861 80250 80251+
...
$* Solver Set: Force - Node on Node 68306:DOF1(2)

RLOAD1 2 1 0.0000 0.0000 1 2 0
TABLED1 1 1 +
+ 0.0000 -1722.36.243900 1699.62 12.4878 1059.11 18.7317 -92.390+
+ 24.9756 1370.47 31.2195448.2240 37.4634 1004.60 43.7073 1321.89+
+ 49.9512549.0030 56.1951216.8290 62.4390397.1410 68.6829877.1630+
+ 74.9268 1173.46 81.1707 1028.60 87.4146993.3700 93.6585 1653.23+
+ 99.9024 1945.41106.1460 1133.06112.3900502.3700118.6340 42.9523+
+ 124.8780105.0700131.1220167.1880137.3660187.0870143.6100 -70.311+

Chapter
Acoustic analysis
matched layerlayer
...
$* Load: Force - Node on Node 68347:DOF6(36)

$* Solver Set: Force - Node on Node 68347:DOF6(179)
MOMENT 179 68347 01.000000 0.0000 0.00001.0000+6
DLOAD 4031.0000001.000000 1221.000000 1241.000000 126+
+ 1.000000 1281.000000 1301.000000 1321.000000 134+
+ 1.000000 1361.000000 1381.000000 1401.000000 142+
+ 1.000000 1441.000000 1461.000000 1481.000000 150+
+ 1.000000 1521.000000 1541.000000 1561.000000 158+
+ 1.000000 1601.000000 1621.000000 1641.000000 166+
+ 1.000000 1681.000000 1701.000000 1721.000000 174+
+ 1.000000 1761.000000 1781.000000 180
ENDDATA
A comparable, complete input file and a detailed workflow is available from the Simcenter 3D AML
acoustic activity in the Self-Paced Catalog on the Learning Advantage website. However, the input
file in that activity may produce slightly different results.

Chapter 8: FEM adaptive order
8.1 Finite Element Method Adaptive Order (FEMAO)
Finite Element Method Adaptive Order (FEMAO) is a higher-order polynomial method for acoustic
and vibro-acoustic analyses. It provides more accurate results and faster solve times by adapting the
computational effort to the complexity of the analysis. You can use FEMAO with SOL 108 or SOL 111.
Working principles of the FEMAO method

The FEMAO method adapts each element’s basis of shape functions at each frequency to provide an
accurate representation of the acoustic pressure inside the element.
The FEMAO method uses:
• Higher-order acoustic shape functions for high frequencies, large elements, or a combination of
both.
• Lower-order acoustic shape functions for low frequencies, small elements, or a combination of
both.
The order of shape functions in an element can be as high as polynomial order 10. At order 1, an
element that uses linear shape functions can span only 1/8 to 1/6 of a wavelength. With the standard
(fixed low-order) FEM method, you need 6 to 8 elements per wavelength. The maximum admissible
frequency would be x Hz. However, at order 10 with FEMAO, you need only about a 6/10 element per
wavelength. The maximum admissible frequency would be 10x Hz. Thus, if you use FEMAO, you
can use the same mesh to compute frequencies more than 10 times higher than with standard FEM.
The following example shows the same 2D mesh with different element orders for two frequencies:
Figure 8-1. f = 100Hz Figure 8-2. f = 1000Hz
When you use FEMAO, you specify the level of accuracy you want in a Simcenter Nastran adaptation
rule. To ensure the chosen accuracy, the FEMAO method chooses the optimal polynomial order of
shape functions per element. The order adaptation is based on the following parameters:

Chapter
Chapter 8: 8: FEM FEM adaptive
adaptive orderorder
Parameter Modeling considerations Solver behavior

You can use larger elements for
voids and local mesh refinement
near geometric boundaries to
FEMAO automatically chooses
accurately capture the geometric
Element size the order per element by taking
of the acoustic domain (for
the element size into account.
example, an engine surface) or
where local variations of the fluid
properties exist.
FEMAO uses lower-order
shape functions (lower number
DOFs per element) at low
You can use larger elements frequencies, and higher-order
even if you need to capture a shape functions (higher number
Frequency DOFs per element) at high
high maximum frequency of
interest in the mesh. frequencies.
The solver can return the
maximum admissible frequency
as a quality result.
You can define the speed of
FEMAO incorporates the
sound using the Simcenter
specified speed of sound
Local speed of sound Nastran material bulk entries
per element to determine the
MAT10, MAT10C, MATF10C,
element order (shape functions).
and MATPOR.
The FINE adaptation rule uses
You can use the Simcenter more higher-order elements
Nastran bulk entry ACADAPT at lower frequencies, and for
Adaptation rule to define the adaptation rule, the same mesh, the COARSE
and set the rule to COARSE, adaptation rule uses more
STANDARD, or FINE. lower-order elements at lower
frequencies.
FEMAO automatically chooses
2 for the lowest and 10 for the
highest polynomial order unless
limits are specified.
The default value for the
minimum order number is 2
You can use the Simcenter instead of 1 for the following
Nastran bulk entry ACORDER reasons:
Polynomial order to control the lowest and highest
• Some acoustic analyses
allowed polynomial order for the
require a pressure gradient
entire model.
(spatial derivative of the
pressure) next to pressure
computation. On the surface
of a first-order element
mesh, where the mesh
represents a set of combined
elements, only the computed

FEM adaptive order
Parameter Modeling considerations Solver behavior
pressure field is continuous.

The pressure gradient field,
which is associated with
acoustic particle velocity,
is piecewise constant with
abrupt changes between the
element faces. However,
for a surface with free
second-order element faces,
the pressure gradient is
continuous. Thus, the use
of second-order element
faces results in increased
computational accuracy.
• Incident acoustic pressure

from, for example, acoustic
monopole sources outside
of the FEM domain, is also
more accurately captured
with second-order elements.
Shape functions
A large number of polynomial shape functions is required to represent the pressure field within each
element. In first-order, linear elements, the number of shape functions and DOFs is the same as the
number of physical nodes in the element. In higher-order elements, the number of shape functions
and DOFs is much higher. Because FEMAO adapts the order on a per element and per frequency
basis, it allows a more efficient representation of the pressure field.
Note
FEMAO uses the hierarchical Lobatto shape functions, which are more flexible and efficient
than the Lagrangian shape functions used in standard (low-order) FEM. (For detailed
information on shape functions, see Reference.)
The following table outlines typical numbers of shape functions defined on a tetrahedral element as a
function of the order.
Internal (vertex +
Order External (bubble) Total shape functions
edge + face)
1 4 0 4
2 10 0 10
3 20 0 20
4 34 1 35
5 52 4 56
6 74 10 84

Chapter
adaptive orderorder
Internal (vertex +
Order External (bubble) Total shape functions
edge + face)
7 100 20 120
8 130 35 165
9 164 56 220
10 202 84 286
Four types of shape functions are shown below on a tetrahedral element:
Figure 8-3. Shape function types

FEM adaptive order
FEMAO benefits
• Accuracy through element order

Standard FEM does not adjust the element order (shape functions). In standard FEM, you
typically use a single mesh for the full frequency range of the analysis, which results in an
over-discretized model at lower frequencies. At higher frequencies, the standard FEM model
becomes under-discretized and less accurate. It may even miss the targeted accuracy because
no automatic correction is in place.
FEMAO, however, adjusts the element order automatically, and the model is represented each
time with the right number of DOFs for each frequency to reach the desired accuracy.
• Performance through adaptivity

An over-discretized model results in longer solve times.
In FEMAO, however, the order adaptation for each frequency guarantees the optimal model size
and consequently the optimal solve time for each frequency. This yields much faster computation
times compared to the fixed number of DOFs approach that standard FEM uses.
Figure 8-4. Computation by FEMAO as a function of time and frequency
• Performance through shape function efficiency

The higher-order shape function basis is also more efficient in capturing the acoustic pressure
field within each element compared to a standard first-order or second-order elements-based
FEM approach.
For example, if you fill a volume with large elements using higher-order shape functions, which
results in more DOFs and shape functions in a single element, and then fill the same volume with
small elements using first-order or second-order shape functions as in standard FEM, the total

Chapter
adaptive orderorder
number of DOFs to reach the same accuracy is higher for the standard FEM than for the FEMAO
method. Thus, FEMAO becomes more efficient as the frequency increases.
• Pre-processing
FEMAO allows you to use large elements in the acoustic domain. This results in a lean FEMAO
model that contains fewer elements than an equivalent standard FEM model. This also means
that you can mesh the model faster in any pre-processor, and a coarser-meshed model improves
graphics performance.
The below images show the differences between a typical FEM mesh, a FEMAO mesh, and a
FEMAO mesh with local refinement. For a 1 meter box, all are designed for a frequency of
3000 Hz.
Figure 8-5. FEM mesh Figure 8-6. FEMAO mesh Figure 8-7. FEMAO mesh
with local refinement
A coarse mesh may not represent the boundary accurately or provide an accurate solution, so a
standard FEM model must use many small elements for an accurate high-frequency solution.
In FEMAO, you should ensure that the spatial variations of the geometry, fluid properties (sound
velocity and density), and boundary conditions (velocity and admittance) are well represented by
the geometrically linear or the quadratic mesh. For this, you can use local refinement.
Example
Consider a velocity boundary condition, which is applied as acoustic panel velocity on
the fluid mesh, with spatial variations of the order of a centimeter required on two faces
of the unit box. The FEM mesh in Figure 8-6 typically yields a poor representation.
Therefore, you should use mesh local refinements as shown in Figure 8-7.
However, if the vibrations originate from a meshed structural panel next to the free fluid
faces, the FEMAO mesh in Figure 8-6 with coarse elements at the fluid structure
interface is supported. In this case, the coupling matrix also includes the higher-order
DOFs of the coarse fluid elements. The structural mesh must be discretized finely
enough to capture the spatial variations in the velocity boundary condition it imposes
on the fluid.

FEM adaptive order
Maximum frequency
• The maximum frequency of a FEMAO mesh is reached when the order of any element Pef is
greater than 10. If you assume 8 elements per wavelength for linear elements, the edge size to
use is
h < (1/8) * (c0 / fmax)
For higher-order elements this becomes
h < (Pef / 8) * (c0 / fmax)
This means that the maximum frequency can be estimated by
Equation 8-1.
where:
o Pef is the order of the element.
o c0 is the speed of sound in the ambient medium.
o h is the element dimension.
However, the Lobatto shape function basis of FEMAO is more efficient compared to standard
FEM. This means that at higher orders, fewer DOFs can be used per wavelength to represent
the acoustic field accurately.
When the ACADAPT adaptation rule is:
o Coarse, 2.2 times fewer DOFs can be used per wavelength.
o Standard, 2 times fewer DOFs can be used per wavelength.
o Fine, 1.6 times fewer DOFs can be used per wavelength.
The maximum frequency criterion therefore becomes
Equation 8-2.
Bulk entries
When you use FEMAO, make sure that your model uses the correct bulk entries:
• For a simple acoustic source, use the acoustic monopole source ACPOLE1. Do not use
ACSRCE.
• For a damping and stiffness absorber, use CAABSF with the bulk entry PAABSF1. Do not use
the CAABSF/PAABSF combination.

Chapter
adaptive orderorder
• For gluing features, use the ACTRAD bulk entry to enforce acoustic continuity across two
surfaces of acoustic meshes. Do not use BGSET and BGADD.
Note
The software ignores unsupported bulk entries.
Fluid to structure interaction

The FEMAO method supports the coupling of a structure to a fluid. In an analysis where you request
the computation of vibro-acoustic transfer vectors (VATVs), FEMAO only supports strong coupling.
Couplings in an analysis without VATV computation may be weak or strong.
For more information on coupling, see Defining the fluid-structure interface boundary condition.
For more information on VATVs, see Vibro-Acoustic Transfer Vector (VATV).
Fluid-structure output requests

Available outputs with fluid-structure coupling and FEMAO are:
• Pressure, velocity, intensity, acoustic power, and transmission loss.
• Panel pressure, modal pressure, and grid pressure contribution as listed in the table.
FEM fluid output requests computed by FEMAO
Panel Modal Grid
contribution contribution contribution
VATV Fluid-structure
Solution Pressure Pressure Pressure
computation interaction
SOL 108 No Strong No No No
SOL 108 No Weak Yes No No
SOL 108 Yes Strong Yes No Yes
SOL 111 No Strong Yes Yes No
SOL 111 No Weak Yes Yes No
SOL 111 Yes Strong Yes No Yes
Note
o Panel pressure contribution with the exception of the number of structural panels
(TOPP describer).
o Modal pressure contribution with the exception of the number of structural modes
(TOPS describer). The contributions are computed on all panels. You cannot
specify a subset of panels.
Performance checking
The solver writes performance indicators, such as computation time (in seconds) and memory (in
GB) per frequency per subcase to the .log file and OUTPUT2 (.op2) file for post-processing in a
pre/post software, such as Simcenter 3D.

FEM adaptive order
Quality checking
For quality assessment, element order per element results from an acoustic analysis will be written
to the .op2 file.
Workflow
1. Create a SOL 108 acoustic or vibro-acoustic solution.
2. Set the following bulk entries:

a. ACADAPT to invoke the FEMAO solution and to define the adaptation (refinement) rule.
Note
The RULE parameter ensures that the numerical error stays within an acceptable
range. This allows you to perform and compare multiple solutions with the same
mesh, but different RULE values.
b. ACORDER to specify the lowest and highest allowed polynomial order for FEMAO.
Reference
Efficient implementation of high-order finite elements for Helmholtz problems, Hadrien Bériot, Albert
Prinn, and Gwénaël Gabard, International Journal for Numerical Methods in Engineering, 2015.

Chapter 9: Transfer vectors
9.1 Acoustic Transfer Vector (ATV)

An Acoustic transfer vector (ATV) is a frequency-dependent acoustical representation of fluid volume
that you can use in SOL 108 and 111 solutions to solve external acoustics problems. ATVs relate
the response at microphone points and 2D microphone elements to excitation at the fluid-structure
(coupling) interface. The primary benefit of ATVs is computational efficiency.
ATVs are similar to output transformation matrices (OTMs). They both use a matrix to relate the
response at specific spatial locations or over specific regions to the excitation at specific locations or
from specific regions.
For more information on OTMs, see the Output Transformation Matrices section of the Simcenter
Nastran Basic Dynamic Analysis User's Guide.
For example, suppose you want to examine how the acoustic response at specific locations to
structural excitation varies with respect to structural materials, loads, and so on. Rather than solving
the entire acoustics problem repeatedly, you can solve the acoustics problem once to create an ATV,
and then use the ATV repeatedly to examine how the response varies.
To use the ATV capability, two Simcenter Nastran runs are required.
• The first run is the ATV computation run. During this run, the software creates and writes the
matrix representation of the ATV to an OP2 file. The ATV computation run must be SOL 108.
• The second run is the ATV response run. During this run, the software retrieves and uses the
matrix representation of the ATV to calculate the acoustic response at microphone points and
elements to the excitation of the structural portion of the model. The ATV response run can be
either SOL 108 or 111 and can use one ATV only.
Note
Only a SOL 108 ATV response run supports an empty A-set. For example, the A-set is
empty when you constrain the structure in all degrees of freedom.
In the ATV computation run:
• Use the ATVOUT case control command to trigger the creation of the ATV, reference the ATVFS
bulk entry that specifies the coupling interface, and reference the SET case control command
that specifies the microphone elements that comprise 2D microphone meshes.
• Use the ATVFS bulk entry to specify the BSURFS bulk entries that define the coupling interface.
• Use a FREQi bulk entry to specify the frequencies at which the software calculates the matrix
representation of the ATV.

Chapter
Chapter 9: 9: Transfer
Transfer vectors
vectors
Note
During the ATV response run, the software interpolates ATV data over the frequency
range defined by the FREQi bulk entry. Because the software does not extrapolate
ATV data, a fatal error occurs if the frequency is out of the range defined by the
FREQi bulk entry.
In the ATV response run:

• Use the ATVBULK bulk entry to select the OP2 file that contains the matrix representation of
the ATV and to offset the element, grid point, and property IDs in the ATV to ensure that they
are unique from those in the structural portion of the model.
• Use an ACMODL bulk entry to specify the coupling interface parameters.
• Use the PRESSURE case control command to request acoustic pressure results at microphone
points.
• Use the PANCON case control command to request the acoustic pressure at microphone points
that are attributable to the vibration of selected structural panels.
o Use PANEL bulk entries to define the structural panels.
o Use SETMC case control commands to specify the microphone points.

For microphone points, specify RTYPE = PRES.
• Use the GRDCON case control command to request the response at microphone points to
vibration at grids in structural panels.
• Use the MODCON case control command to request the contribution of modes to the acoustic
pressure at microphone points. This is valid for a SOL 111 ATV response run only.
o Use SETMC case control commands to specify the microphone points.
For microphone points, specify RTYPE = PRES.
When you use an ATV, acoustic particle velocity and acoustic intensity results cannot be computed,
and the ACINTENSITY and ACVELOCITY case control commands are ignored if they are specified.
9.2 Vibro-Acoustic Transfer Vector (VATV)

The Vibro-acoustic transfer vectors (VATVs) let you efficiently compute vibro-acoustic pressure due
to aerodynamic or acoustic loads on flexible structures in the context of two-way (strong) coupled
vibro-acoustic problems. VATVs contain matrices that link the structural grids of a flexible structure
with microphone points or 2D microphone elements.
The VATVs are independent of the loading and you can reuse them for various loads in multiple
vibro-acoustic problems as long the structure, acoustic fluid, and the location of microphones remain
the same.

Transfer vectors
For example, suppose you want to examine how the acoustic response at specific locations to
structural excitation varies with respect to loads. Rather than solving the entire acoustics problem
repeatedly, you can solve the vibro-acoustics problem once to create a VATV, and then use the VATV
repeatedly to examine how the response varies.
Using a VATV in your vibro-acoustic problems involves two runs:
1. The solver computes the VATV using SOL 108 or SOL 111. During this run, Simcenter Nastran
writes the VATV matrices, which stores the results in pressure format for fluid grid points
referenced by microphone elements and normal nodal force format for structural grids, to a
name_vatv.op2 file.
2. You use the computed VATV in a frequency response or random analysis. During this run,
the solver retrieves and uses the matrix representation of the VATV to calculate the acoustic
response at microphone points and elements to the excitation of the structural portion of the
model. The VATV response analysis can be only SOL 108 and can use one VATV only.
In the VATV computation run:

• Use the VATVOUT case control command to trigger the creation of the VATV. The command
references a VATVFS bulk entry and specifies the Fortran unit number for the .op2 file to which
the VATV results are written.
• Use the VATVFS bulk entry to specify the BSURF or BSURFS bulk entries that define the
structural element faces of a pressure boundary. The structural element faces describe the
free surface region.
• Use an ACMODL bulk entry to specify the coupling interface parameters.
• Use a FREQi bulk entry to specify a frequency range at which the software calculates the matrix
representation of the VATV.
Note
During the VATV response run, the software derives the exact frequencies from the
VATV matrix computed frequencies in the specified frequency range.
In the VATV response run:

• Use the VATVBK bulk entry to select the .op2 file that contains the matrix representation of
the VATV.
• Use the FREQV bulk entry to specify a frequency range of interest.
• Use the PRESSURE case control command to request acoustic pressure at microphone points.
• Use the ACPOWER case control command to request acoustic power at 2D microphone
elements.
• Use the PANCON case control command to request the acoustic pressure or power at
microphone points that are attributable to the vibration of selected structural panels.
o Use PANEL bulk entries to define the structural panels.

Chapter
Chapter 9: 9: Transfer
Transfer vectors
vectors
o Use SETMC case control commands to specify the microphone points and response type.
Specify RTYPE = PRES for acoustic pressure.
• Use the GRDCON case control command to request the response at microphone points to
vibration at grids in structural panels.
When you use a VATV, acoustic particle velocity and acoustic intensity results cannot be computed,
and the ACINTENSITY and ACVELOCITY case control commands are ignored if they are specified.
For more information, see the VATVOUT case control command, and the VATVFS, FREQV, and
VATVBK bulk entries.

Chapter 10: Other considerations
10.1 Superelements
Superelements may be used with the following restrictions:
1. A superelement may contain either fluid or structural points, but not both. The residual structure
may contain both.
2. The grid points at the fluid-structure interface may be assigned to the residual structure only. This
requires the specification of q-set points using the SEQSETi and EIGR or EIGRL Bulk Data
entries and the METHOD case control command.
3. Superelements must be linear. That is, any non-linear or frequency-dependent effects must be in
the residual. Consequently, the presence of the Exterior acoustics using automatically matched
layer or elements that reference porous materials must be defined in the residual. This applies to
the AML boundary condition, CAABSF absorber element (frequency-dependent), and elements
that reference MAT10 material definition pointing to frequency-dependent tables.

Siemens Industry Software
Headquarters
Europe
Granite Park One
Stephenson House
5800 Granite Parkway
Sir William Siemens Square
Suite 600
Frimley, Camberley
Plano, TX 75024
Surrey, GU16 8QD
USA
+44 (0) 1276 413200
+1 972 987 3000
Asia-Pacific
Americas
Suites 4301-4302, 43/F
Granite Park One
AIA Kowloon Tower, Landmark East
5800 Granite Parkway
100 How Ming Street
Suite 600
Kwun Tong, Kowloon
Plano, TX 75024
Hong Kong
USA
+852 2230 3308
+1 314 264 8499
About Siemens PLM Software
© 2018 Siemens Product Lifecycle Management

Siemens PLM Software, a business unit of the Siemens
Software Inc. Siemens and the Siemens logo are
Industry Automation Division, is a leading global provider
registered trademarks of Siemens AG. D-Cubed,
of product lifecycle management (PLM) software and
Femap, Geolus, GO PLM, I-deas, Insight, JT, NX,
services with 7 million licensed seats and 71,000 customers
Parasolid, Solid Edge, Teamcenter, Tecnomatix and
worldwide. Headquartered in Plano, Texas, Siemens
Velocity Series are trademarks or registered trademarks
PLM Software works collaboratively with companies
of Siemens Product Lifecycle Management Software
to deliver open solutions that help them turn more
Inc. or its subsidiaries in the United States and in other
ideas into successful products. For more information
countries. All other trademarks, registered trademarks
on Siemens PLM Software products and services, visit
or service marks belong to their respective holders.
www.siemens.com/plm.

Acoustics User Guide

Uploaded by

Copyright:

Available Formats

Acoustics User Guide

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Acoustics User Guide

Uploaded by

Copyright:

Available Formats

SIEMENS

Proprietary & Restricted Rights Notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

Defining the model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1

Acoustics User’s Guide 3

Solution methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4-1

Input data requirements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-1

Acoustic and vibro-acoustic analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6-1

Acoustic analysis using automatically matched layer . . . . . . . . . . . . . . . . . . . . . . . . . . . 7-1

4 Acoustics User’s Guide

FEM adaptive order . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-1

Transfer vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9-1

Other considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10-1

2-1. Interior Fluid Domain and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-1

Acoustics User’s Guide 5

8-7. FEMAO mesh with local refinement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8-6

6 Acoustics User’s Guide

Acoustics User’s Guide 7

8 Acoustics User’s Guide

Letter symbols Description

Area or coupling matrix

Damping or the impedance matrix of the fluid

Damping matrix of the structure

Equivalent frequency-dependent damping

Stiffness matrix or inverse mass matrix of the fluid

Stiffness matrix of the structure

Structure factor or square of tortuosity

Acoustics User’s Guide 9

Mass matrix or compressibility matrix of the fluid

Mass matrix of the structure

Shape functions for acoustic pressure

Shape functions for the structure

Outward normal to the fluid surface

Outward normal to the structural surface

Acoustic volume velocity

Strength of acoustic source

Static thermal permeability

Acoustic particle velocity

Vector of fluid displacements normal to the surface with outward normal

Vector of structural displacements

10 Acoustics User’s Guide

Equivalent bulk modulus of the porous material

Real and imaginary parts of bulk modulus of the porous materials

Bulk modulus of ambient medium

Real and imaginary parts of equivalent density of the porous materials

Frequency-dependent fluid density

Density of ambient medium

Speed of sound in air

Vector of fluid modes

Vector of structural modes

Acoustics User’s Guide 11

12 Acoustics User’s Guide

1.1 Overview of acoustic capabilities

1.2 Assumptions and limitations

Acoustics User’s Guide 1-1

2.1 Overview of acoustic theory

Figure 2-1. Interior Fluid Domain and Boundary Conditions

• Linear pressure-density relationship.

Thus Euler's momentum equation is

• is the density of the ambient medium.

• is the acoustic particle velocity.