There are direct and indirect methods. The direct method requires some very specialized measurement equipment called a "microflown" which measures air partical velocity directly. Then you need to moutn this on a some sort of scanning system. This was done by a guy name Behler from Auchen. He had some AES papers on this.
The indirect method requires that you have a known load on the divice most easily a plane wave tube. By measuring the response at severla location along and arround the tube one can calcultae the shape of the wavefront. But of course this later methiod involves a lot of math and careful measurements.
So the choices are spend a lot of money on equipment or a lot of time on the calculations. There is no simple method.
The indirect method requires that you have a known load on the divice most easily a plane wave tube. By measuring the response at severla location along and arround the tube one can calcultae the shape of the wavefront. But of course this later methiod involves a lot of math and careful measurements.
So the choices are spend a lot of money on equipment or a lot of time on the calculations. There is no simple method.
Dr. Geddes,
thank you for the offer. I would appreciate if you could provide the outline, so that I understand the coding requirements.
Would the coding require a use of a programming language or would a software package, i.e., Matlab, MatCAD, suffice?
If you feel that the discussion will become too involved, please use pm so that we do not unnecessarily clutter the forum.
Kindest regards,
M
thank you for the offer. I would appreciate if you could provide the outline, so that I understand the coding requirements.
Would the coding require a use of a programming language or would a software package, i.e., Matlab, MatCAD, suffice?
If you feel that the discussion will become too involved, please use pm so that we do not unnecessarily clutter the forum.
Kindest regards,
M
I did it in MathCAD, but data input was cumbersome.
Lets start with this:
The waves inside of any circular cross section tube can be defined by a series of Bessel functions and a wave number along the Z axis - the mode frequencies and the wavenumber along Z are not independent. This is shown in Theoretical Acoustics by Morse on page 509 (the following pages are essential reading for this technique).
If you write down the first (m,n)th equations (one for each mode m,n) for the value of the pressure at the boundary of the rigid walled tube then you will get a matrix equation where you have some square matrix of size m*n (which depends only on the wavenumber k) times the m*n coefficients of modal contributions that then equals the m*n measured pressure values along and arround the tube at the walls. Solve this equation and you have the modal contributions which can be used to reconstruct the wavefront anywhere inside the tube at some value of the wavenumber.
This equation will most likely be very near singular so you have to use SVD to decompose it into its most significant modes.
The rest, as they say, is just code.
Lets start with this:
The waves inside of any circular cross section tube can be defined by a series of Bessel functions and a wave number along the Z axis - the mode frequencies and the wavenumber along Z are not independent. This is shown in Theoretical Acoustics by Morse on page 509 (the following pages are essential reading for this technique).
If you write down the first (m,n)th equations (one for each mode m,n) for the value of the pressure at the boundary of the rigid walled tube then you will get a matrix equation where you have some square matrix of size m*n (which depends only on the wavenumber k) times the m*n coefficients of modal contributions that then equals the m*n measured pressure values along and arround the tube at the walls. Solve this equation and you have the modal contributions which can be used to reconstruct the wavefront anywhere inside the tube at some value of the wavenumber.
This equation will most likely be very near singular so you have to use SVD to decompose it into its most significant modes.
The rest, as they say, is just code.
Cornucopia
This [1] is just one of numerous articles (dissertations) on the subject of the acoustic behavior of compression drivers and the horns to which they are mounted. Shape of the wave front formed at the confluence of phase plug passages and beyond in the horn neck have a strong influence on horn performance and the signal audited in the far field.
Regards,
WHG
[1] ACOUSTICAL ANALYSIS OF HORN-LOADED COMPRESSION DRIVERS USING NUMERICAL TECHNIQUES Daniel R. Tengelsen
Acoustical Analysis of a Horn-Loaded Compression Drivers Using Numerical Analysis :: ETD
This [1] is just one of numerous articles (dissertations) on the subject of the acoustic behavior of compression drivers and the horns to which they are mounted. Shape of the wave front formed at the confluence of phase plug passages and beyond in the horn neck have a strong influence on horn performance and the signal audited in the far field.
Regards,
WHG
[1] ACOUSTICAL ANALYSIS OF HORN-LOADED COMPRESSION DRIVERS USING NUMERICAL TECHNIQUES Daniel R. Tengelsen
Acoustical Analysis of a Horn-Loaded Compression Drivers Using Numerical Analysis :: ETD
That is a good paper. Its a bit misguided since the horns that he is modeling are kind of obsolete these days, but he shows a real understanding of the math and its use. He misses some key aspects that more recent research has shown, such as the fact that nonlinear distortion in a horn compression driver system is not audibly significant, but that the audibility of the HOM are (he clearly shows how the HOM are dispersive and do not arrive in time or space the same as the fundamental sound.) It was interesting to see him show how HOM can be created from reflections.
Did anyone succesfully model a compression driver/horn combination beyond 5k using BEM/FEM? It seems that damping and damping precision have a lot of influence on result. Could you point me to a study where someone did?
@ dr. Geddes
Am I right when I assume that HOMs are excess phase? Or are they mixed phase?
@ dr. Geddes
Am I right when I assume that HOMs are excess phase? Or are they mixed phase?
Relevance?
Obsolesce
The referenced article [1] clearly demonstrates the consequences of applying different changes to the profile of an acoustic passage. The notion that a particular horn used in this investigation is obsolete or not, is irrelevant to the study mission.
Distortion
At some threshold (magnitude), ALL distortion becomes audible. Such thresholds are entirely dependent on the auditors hearing (listening) acuity. That is why there is dissenting views on what constitutes an audibly significant departure from an original signal.
[1] ACOUSTICAL ANALYSIS OF HORN-LOADED COMPRESSION DRIVERS USING NUMERICAL TECHNIQUES Daniel R. Tengelsen
Obsolesce
The referenced article [1] clearly demonstrates the consequences of applying different changes to the profile of an acoustic passage. The notion that a particular horn used in this investigation is obsolete or not, is irrelevant to the study mission.
Distortion
At some threshold (magnitude), ALL distortion becomes audible. Such thresholds are entirely dependent on the auditors hearing (listening) acuity. That is why there is dissenting views on what constitutes an audibly significant departure from an original signal.
[1] ACOUSTICAL ANALYSIS OF HORN-LOADED COMPRESSION DRIVERS USING NUMERICAL TECHNIQUES Daniel R. Tengelsen
12 kHz.
For higher read here:
http://digital.library.adelaide.edu.au/dspace/bitstream/2440/41350/2/03chapters5-8.pdf
Regards,
WHG
For higher read here:
http://digital.library.adelaide.edu.au/dspace/bitstream/2440/41350/2/03chapters5-8.pdf
Regards,
WHG
Seems like my recommendations from more than 20 years ago (without the use of numerical techniques I might add) have stood up well.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.