Sound excitation sources are fundamental to an acoustic analysis. You can introduce sound excitation sources via:
Specified pressure at nodes (D)
Normal velocity (harmonic analysis) or acceleration (transient analysis) on the exterior surface of the domain (SF)
Arbitrary velocity (harmonic analysis) or acceleration (transient analysis) on the exterior surface of the domain (BF)
Analytic incident wave sources (AWAVE)
Mass source (harmonic analysis) or mass source rate (transient analysis) at nodes, along lines, on surfaces, or in volumes (BF)
The following table shows all excitation sources available for acoustic analysis:
Table 8.1: Acoustic Excitation Sources
| Excitation Sources | FE Model Entities |
|---|---|
| Pressure | Nodes |
| Outward normal velocity (acceleration) | Nodes |
| Arbitrary nodal velocity (acceleration) | Nodes or elements |
| Analytic incident wave sources | Not applicable |
| Mass sources | Nodes |
| Diffuse sound field | Not applicable |
| Acoustic duct port | Nodes |
Excitation sources can be applied on the finite element model entities.
The following detailed descriptions of the available excitations are available:
For general information about applying loads, see Loading in the Basic Analysis Guide.
Pressure excitation (D,Node,PRES)
behaves as a Dirichlet pressure boundary (Pressure Boundary).
When applying pressure excitation, the pressure is enforced to a given value. Sound pressure reflected by other objects back to the excitation point cannot be taken into account.
Pressure excitation can be used only under conditions where the effect of reflected sound pressure is not required.
Outward normal velocity or acceleration can be applied to the exterior surface of
the model (SF,Nlist,SHLD).
Apply a minus sign of the outward normal velocity if an inward normal velocity is required.
For a harmonic analysis, a complex normal velocity to the surface is defined by the amplitude and phase angle. The program solves for the pressure on the normal velocity excitation surface.
Normal velocity excitation exists either on the structural surface or on the
transparent pressure wave port on which the incident wave propagates into the
acoustic domain and the reflected wave backs to the infinity. To absorb the
reflected wave on the transparent port, apply the impedance boundary to the port
surface (SF,Nlist,IMPD or
SF,Nlist,INF) along with the
velocity excitation. To distinguish the transparent wave port from the structural
surface, specify the transparent port surface
(SF,Nlist,PORT).
The following command applies outward normal velocity or acceleration to the nodes of the FE model:
SF,
Nlist,SHLD,Value,Value2
nsel,loc,z,0 ! select nodes at z = 0 sf,all,shld,vn,ang ! complex normal velocity sf,all,impd,z0 ! impedance boundary sf,all,port,1 ! transparent port
nsel,loc,z,0 ! select nodes at z = 0 sf,all,shld,vn,ang ! complex normal velocity sf,all,impd,z0 ! impedance boundary
Use tables (*DIM) in the SF command to define the frequency dependency of the normal velocity of acceleration, as shown:
*dim,vn,TABLE,2,1,1,FREQ ! normal velocity table *dim,ang,TABLE,2,1,1,FREQ ! angle table vn(1,0,1)=FreqB ! beginning frequency vn(2,0,1)=FreqE ! ending frequency vn(1,1,1)=vn1 ! normal velocity at FreqB vn(2,1,1)=vn2 ! normal velocity at FreqE ang(1,1,1)=ang1 ! phase angle at FreqB ang(2,1,1)=ang2 ! phase angle at FreqE nsel,s,loc,z,0 ! select nodes at z = 0 sf,all,shld,%vn%,%ang% ! tabular complex vn
An arbitrary velocity or acceleration can be applied to the nodes on the exterior
surface of the model (BF,Node,VELO).
The arbitrary velocities are projected to the outward normal direction on the excitation surface after interpolation during an acoustic solution. For a harmonic analysis, a complex velocity is defined by the amplitudes and phase angles of the components.
The arbitrary velocity or acceleration can be defined in a local Cartesian coordinate system (LOCAL), then assigned to the elements (VATT or ESYS prior to meshing or EMODIF after meshing). The program solves for pressure on the velocity excitation surface. If the reflected sound pressure waves that are passing through the velocity excitation surface are simulated, apply the impedance boundary condition to the excitation surface. The arbitrary velocity excitation exists either on the structural surface or on the transparent pressure wave port.
The following command applies arbitrary velocity to the nodes of the FE model:
BF,
Node,VELO,Vx,Vy,Vz,AngX,AngY,AngZ
et,1,220,,1 ! uncoupled acoustic element local,11 ! local coordinate esys,11 ! use local as element esys … nsel,loc,z,0 ! select nodes at z = 0 bf,all,velo,vx,vy,vz,angx,angy,angz ! complex arbitrary velocity sf,all,impd,z01 ! impedance boundary sf,all,port,1 ! transparent port
Use tables (*DIM) in the BF command to define the frequency dependency of the arbitrary velocity or acceleration, as shown:
*dim,vx,TABLE,2,1,1,FREQ ! vx table *dim,vy,TABLE,2,1,1,FREQ ! vy table *dim,vz,TABLE,2,1,1,FREQ ! vz table *dim,ax,TABLE,2,1,1,FREQ ! angle x table *dim,ay,TABLE,2,1,1,FREQ ! angle y table *dim,az,TABLE,2,1,1,FREQ ! angle z table vx(1,0,1)=FreqB ! beginning frequency vx(2,0,1)=FreqE ! ending frequency vx(1,1,1)=vx1 ! vx at FreqB vx(2,1,1)=vx2 ! vx at FreqE vy(1,1,1)=vy1 ! vy at FreqB vy(2,1,1)=vy2 ! vy at FreqE vz(1,1,1)=vz1 ! vz at FreqB vz(2,1,1)=vy2 ! vz at FreqE ax(1,1,1)=vx1 ! angle x at FreqB ax(2,1,1)=vx2 ! angle x at FreqE ay(1,1,1)=vy1 ! angle y at FreqB ay(2,1,1)=vy2 ! angle y at FreqE az(1,1,1)=vz1 ! angle z at FreqB az(2,1,1)=vy2 ! angle z at FreqE nsel,s,loc,z,0 ! select nodes at z = 0 bf,all,velo,%vx%,%vy%,%vz%,%ax%,%ay%,%az% ! complex velocity table
Acoustic analyses often use wave sources with analytic functions in harmonic analysis. The following table shows all available analytic wave sources for acoustic analyses:
Table 8.2: Acoustic Analytic Incident Wave Sources
| Wave Sources | Comments |
|---|---|
| Planar wave | Plane wave with the incident φ and θ in the global spherical coordinate system from infinity |
| Monopole or pulsating sphere | Spherical wave from the origin (xs,ys,zs) |
| Dipole | Consists of two monopoles or pulsating spheres with opposite signs and a small separation |
| Back enclosed loudspeaker | Loudspeaker with back enclosure acting as a monopole |
| Bare loudspeaker | Loudspeaker without back enclosure acting as a dipole |
To define various incident wave sources, issue the AWAVE command
Specify the integer number (WaveNum) or an acoustic
incident wave inside or outside of the model. Valid values are 1 through 20. One or
more wave types can be selected. The amplitude of the pressure or normal velocity is
used for the excitation.
A planar wave can be defined in terms of the amplitude and the spatial incident angles in the global spherical coordinate system, as shown in this figure:
Because the incident planar wave is approximated by the far-field wave front of a source far from the receiver, both the initial phase angle and the source original are ignored.
Specify the analytic incident wave sources and select either the total pressure field or the scattered field solver in an acoustic scattering analysis.
If the scattered parameter is required and the scattered pressure field is much smaller than the incident pressure field, use the scattered pressure field solver (ASOL,SCAT) to avoid numerical errors. Control the output result as needed for either the total or scattered nodal pressure in the model (ASCRES).
Specify incident wave sources as external sources (Opt2
= EXT on the AWAVE command) if a
scattering analysis is needed.
When analytic incident wave sources are used inside the model
(Opt2 = INT on the
AWAVE command), only the scattered pressure field solver is
activated, regardless of whether the ASOL,SCAT command is issued.
The source origin must be located inside the model. The plane wave incident source cannot be used inside the model.
The uniform normal velocity on the cross section can be used to launch the plane wave. When analytic incident wave sources are located inside the model, the nodal total pressure is always output, even though the scattered field solver is used.
block,0,xs,0,ys,0,zs ! geometry of model … awave,1,mono,velo,int,v0,ang,xs/2,ys/2,zs/2 ! incident wave inside of model
block,0,xs,0,ys,0,zs ! geometry of model … awave,1,dipole,pres,ext,p0,ang,-xs,-ys,-zs ! incident wave outside of model
It is not necessary to assign the internal analytic incident wave sources to the FE nodes. It is convenient to use the internal analytic incident wave sources rather than meshing the wave source structure, such as a pulsating sphere.
For more information, see Pure Scattered Pressure Formulation in the Mechanical APDL Theory Reference.
To excite sound waves in an acoustic model, use a mass source (harmonic analysis) or mass source rate (transient analysis).
The mass source is input by defining up to one scalar quantity
(Lab = MASS on the BF command) and
a phase angle. The mass source is specified at nodes (BF).
For a volume mass source, specify the mass source on the volumetric nodes.
For a surface mass source, specify the mass source on at least three nodes on an element face. The surface current source must coincide with the elements faces.
For a line mass source, specify the mass source at two nodes connected by an element edge. The line current source must coincide with the element edges.
A point mass source must be at the element nodes.
In general, a mass source launches the pressure wave in all directions. For a propagating or resonant system, a mass source can be used to excite the propagating modes or resonant modes of the structure.
Only proper modes can exist in the structure. To reduce the parasitic modes, choose the distribution of the mass source based on the pressure distribution of the excited mode.
When a mass source is applied to an exterior surface, the excited pressure is determined by p = qsc0. On an exterior or interior transparent port, the excited pressure is given by p = qsc0 / 2.
nsel,s,loc,z,0 ! select nodes at z = 0 bf,all,mass,masmag,masang ! complex mass source
Use tables (*DIM) in the BF command to define the frequency dependency of the arbitrary mass source or mass source rate, as shown:
*dim,masmag,TABLE,2,1,1,FREQ ! mass source amplitude table *dim,masang,TABLE,2,1,1,FREQ ! mass source angle table masmag (1,0,1)=FreqB ! beginning frequency masmag (2,0,1)=FreqE ! ending frequency masmag (1,1,1)=vx1 ! amplitude at FreqB masmag (2,1,1)=vx2 ! amplitude at FreqE masang (1,1,1)=vz1 ! angle at FreqB masang (2,1,1)=vy2 ! angle at FreqE nsel,s,loc,z,0 ! select nodes at z = 0 bf,all,mass,% masmag%,%masang% ! complex mass source table
For more information, see Mass Source in the Wave Equation in the Mechanical APDL Theory Reference.
The diffuse sound field is approached by the asymptotic model summing a high number of uncorrelated plane waves with random phases from all directions in free space. The DFSWAVE command defines the diffuse sound field.
The incident space of the diffuse sound field is mesh-free. A reference sphere
related to the structural panel defines the incident plane waves. The radius R of
the reference sphere should be at least 50 times the maximum dimension of the
structural panel. The energy of the diffuse sound field uniformly distributes on the
reference sphere surface in all directions. The sphere surface is equally divided
into N elementary surfaces.
The plane center of the structural panel should be located at the origin of the local Cartesian coordinate system (LOCAL) (defaults to the global Cartesian coordinate system). The +z axis of the Cartesian coordinate system must be consistent with the panel’s outward normal unit vector on the panel’s incident diffuse sound field side. The structural panel is meshed by solid or shell elements.
The receiving domain is meshed by acoustic elements and truncated by artificially matched layers (PML or IPML) or by absorbing elements, as shown in this figure:
If the effect of the acoustic fluid on the structural panel can be ignored, it is not necessary to create a receiving acoustic domain. The radiated sound far-field is calculated in the postprocessor (PRFAR,PLAT or PLFAR,PLAT) once the structural panel model is solved with the flagged equivalent source surface (SF,,MXWF).
In practice, the sphere surface is divided into M parallel rings along the z axis of a local Cartesian coordinate system, and the program generates the elementary surfaces, each having nearly the same area.
When defining the diffuse sound field, the DFSWAVE command specifies the local coordinate system number, the radius of the reference sphere, the reference power spectral density, mass density of the incident space, the sound speed in the incident space, the maximum incident angle of the plane waves, the number of the parallel rings, and the sampling options.
To excite the vibro-acoustics system, generate the SURF154 surface element on the surface of the structural elements.
The symmetry of a panel structure cannot be used to reduce the simulation size, as the incident plane waves have varying random phase angles.
To initiate multiple solutions (load steps) for random acoustics analysis with
multiple samplings, issue the MSOLVE command. The process is
controlled by the norm convergence tolerance
(VAL1 on
MSOLVE) or the number of multiple solutions if the
number of solution steps reaches the number specified
(NUMSLV
on MSOLVE). The program checks the norm
convergence by comparing two averaged sets of radiated sound powers with the
interval of the norm (VAL2 on
MSOLVE) over the frequency range.
To calculate the average transmission loss for multiple sampling phases at each frequency over the frequency range, issue the PRAS or PLAS command.
et,1,220,,0 ! coupled acoustic element et,2,220,,1,,1 ! acoustic PML element et,3,281 ! structure shell element et,4,154 ! surface element … cm,nod1,node ! group vibro-acoustics FSI interface nodes sf,all,fsi ! flag FSI interface … sectype,2,shell secdata,0.005,2 cmsel,s,nod1 ! select FSI interface nodes type,3 mat,2 secn,2 esurf ! generate shell element alls esel,s,type,,3 ! select shell element type,4 mat,2 esurf ! generate surface element … ! define diffuse sound field dfswave,0,15,1,1.225,340,90,20,all ! finish /solu antype,harmic harfrq,100,200 nsubst,100 msolve,5 ! five samples finish /post1 pras,dfstl,avg ! print transmission loss plas,dfstl,avg ! plot transmission loss finish
For more information, see Random Acoustics in the Mechanical APDL Theory Reference.
A single mode or multiple modes may exist in the acoustic duct. Launching a specified acoustic mode into a guided acoustic wave system with discontinuities excites multiple propagating or evanescent modes, depending on the working frequency and geometrical dimension of the duct.
If the lowest order mode is launched and the higher order modes decay as the parasitic evanescent modes near the discontinuities, the acoustic port can be used to terminate the inlet or outlet with the specified mode. While the multiple propagating modes are excited and propagate in the acoustic duct, PML or IPML should be used for the domain truncation.
Define the property of an analytic modal port via the following command:
APORT,
PortNum,Label,KCN,PRES,PHASE,--,VAL1,VAL2,VAL3,VAL4
The available analytic port types are planar wave, rectangular duct, circular duct, and coaxial duct.
et,1,220,, ! acoustic element et,2,220,,1,,1 ! acoustic PML element et,3,186 ! structural element ... nsel,s,loc,z,0 nsel,a,loc,z,1 sf,all,fsi ! fsi interface nsel,s,loc,x,-1 nsel,a,loc,x,1 nsel,r,loc,z,-5,5 cpcyc,all,,,2 ! coupled nodes for PBC ... nsel,s,loc,z,4 bf,all,port,1 ! flag interior port 1 aport,1,plan,0,p0,0,0,0,theta ! plane wave source port nsel,s,loc,z,-4) bf,all,port,2 ! flag interior port 2 aport,2,plan,0,0,0,0,0,theta ! plane wave output port nsel,s,loc,z,-5 nsel,a,loc,z,5 d,all,pres,0 ! zero pressure on PML exterior surface
Note that the transverse cross section of the acoustic port must be located on the x-y plane in the defined local coordinates system (LOCAL).
The low-order FLUID30 element does not support the higher modes in the coaxial duct.
For more information, see Analytic Port Modes in a Duct in the Mechanical APDL Theory Reference.