Just a quick teaser - one of the new possibilities how to define an enclosure. Time to learn Bezier curves 
The corresponding script snippet:
box = {
point P0 100 0 5
point P1 140 20 8
point P2 100 250 10
point PB 0 250 10
bezier P0 P1 P2 PB
}
Mesh.Enclosure = {
Plan = box
Spacing = 0,20
}
The corresponding script snippet:
box = {
point P0 100 0 5
point P1 140 20 8
point P2 100 250 10
point PB 0 250 10
bezier P0 P1 P2 PB
}
Mesh.Enclosure = {
Plan = box
Spacing = 0,20
}
Two more samples -
box1 = {
point P0 100 0 8
point P1 125 25 8
point P2 50 200 16
point PB 0 200 20
line P0 P1
line P1 P2
line P2 PB
}
arcs = {
point P0 100 0 8
cpoint P1 100 50
point P2 150 50 10
cpoint P3 150 150
point P4 50 150 20
point PB 0 150 20
arc P0 P1 P2
arc P2 P3 P4
line P4 PB
}
box1 = {
point P0 100 0 8
point P1 125 25 8
point P2 50 200 16
point PB 0 200 20
line P0 P1
line P1 P2
line P2 PB
}
arcs = {
point P0 100 0 8
cpoint P1 100 50
point P2 150 50 10
cpoint P3 150 150
point P4 50 150 20
point PB 0 150 20
arc P0 P1 P2
arc P2 P3 P4
line P4 PB
}
From the looks the enclosure is ”meshed around“ a waveguide definition, meaning it is already finished, ready to run in abec/akabak. For initial development, quarter symmetry will be good, after which one would want to simulate the full enclosure for corrections. mabat has said before it will be possible to localize it on the baffle as needed, i. e. on the upper half.
I don’t know how inserting a second source (in case of a two way) would work and determining their combined polar pattern, but I guess we are about to find out soon.
This becomes a real practical problem and the scripting does not overcome that. A waveguide transitioning into a baffle needs a lot of resolution to avoid problems, using half symmetry makes that twice as difficult. It is not just about the solve time but the fact that beyond a certain number of elements ABEC can end up thrashing away forever and never come up with a solution. 10,000 elements is a reasonable ceiling and by that point the sim time is heading towards days not hours. Needing to adjust a setting that you missed is quite disheartening. If the mesh steps on the waveguide are not selected properly there can end up being lots of long thin triangles created which cause real problems for accurate simulation. To keep the smoothness of the curve the solution is usually smaller element size which just begins the loop.
From what mabat has described that does not sound like what he intends to do. There would be the possibility to combine two simulations together into one but that would require some background knowledge of ABEC which most people waiting for the scripting approach don't have (yet).
Yes, especially with a more economic termination, it will get interesting.
Concerning your second point, if the new meshing script can produce this https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-6973452 and we can also define another source with source definition (can we define two sources in one ath.cfg?), I assumed this was within the possibilities of the script. But I could be wrong.
Concerning your second point, if the new meshing script can produce this https://www.diyaudio.com/community/...-design-the-easy-way-ath4.338806/post-6973452 and we can also define another source with source definition (can we define two sources in one ath.cfg?), I assumed this was within the possibilities of the script. But I could be wrong.
I'm sure it is possible but I didn't get the impression that mabat intended to do it that way.
High frequencies and a dense large mesh = really long solve times potentially tending towards unsolvable.
Simulating the waveguide and the woofer at the same time makes things quite tricky especially if we are talking 15" sized devices.
The waveguide can be simulated in quarter symmetry in a finite cabinet quite successfully and by keeping the cabinet more minimal but still representative of the main important dimensions pretty accurate directivity can be simulated.
A woofer at low frequency can be simulated in half symmetry with much less fine elements and still get accurate directivity results. Two simulations to tell you if the important parameters of a design work as intended. There is no real need to have them together and it is considerably more difficult to do so.
I thought the same way before that having it all together would be great, reality has shown me the error in my thinking
I understand your working premises, and I think you are right the working method will be quarter symmetry. I have never thought about half symmetry for the woofer section before. It is however possible to run a full enclosure, and for the verticals, it will make a difference due to the extended baffle below or above the sources. We will soon find out about the limits of calculation power and the meshing script.
I am especially interested to see what happens when the termination of a waveguide is defined in a more economic, space efficient way.
I am especially interested to see what happens when the termination of a waveguide is defined in a more economic, space efficient way.
Just a quick teaser - one of the new possibilities how to define an enclosure. Time to learn Bezier curves
View attachment 1037637
The corresponding script snippet:
box = {
point P0 100 0 5
point P1 140 20 8
point P2 100 250 10
point PB 0 250 10
bezier P0 P1 P2 PB
}
Mesh.Enclosure = {
Plan = box
Spacing = 0,20
}
TAKE MY MONEY PLEASE
mabat,
I would be interested to know how the meshing script copes with a shorter termination. fluid had some trouble, when he created an enclosure for the waveguide from post #9045
This is the full cfg, is the meshing of an enclosure successful in this case?
I would be interested to know how the meshing script copes with a shorter termination. fluid had some trouble, when he created an enclosure for the waveguide from post #9045
This is the full cfg, is the meshing of an enclosure successful in this case?
Code:
; ____ Parameters ____
Throat.Angle = 10.08
Throat.Diameter = 25.4
Throat.Profile = 1
Coverage.Angle = 45 -9*sin(p)^2
Length = 79
Term.s = 1.3 -0.55*sin(p)^2
Term.n = 3.2 +0.55*sin(p)^2
OS.k = 0.42 +0.02*sin(p)^2
Term.q = 0.94 +0.042*sin(p)^2
; ____ Source Mode ____
;Source.Shape = 2
Source.Shape = 1
Source.Curv = 0
Source.Radius = -1
Source.Velocity = 1
; ____ 3D Mesh Settings ____
Mesh.InterfaceResolution = 8.0
Mesh.LengthSegments = 30
Mesh.AngularSegments = 64
Mesh.ThroatResolution = 2
Mesh.CornerSegments = 5
Mesh.InterfaceOffset = 3
; ____ ABEC Settings ____
ABEC.MeshFrequency = 1000
ABEC.NumFrequencies = 30
ABEC.Abscissa = 2
ABEC.SimType = 1
;ABEC.SimProfile = 1
ABEC.f1 = 1000
ABEC.f2 = 20000
; ____ VCAS Output ____
ABEC.Polars:SPL = {
MapAngleRange = 0,90,19
Distance = 2
Offset = 200
}
ABEC.Polars:SPL_Sono = {
MapAngleRange = -90,90,19
Distance = 2
NormAngle = 10
Offset = 200
}
ABEC.Polars:SPL_Sono = {
MapAngleRange = -90,90,19
Distance = 2
Offset = 200
}
ABEC.Polars:SPL_H_norm = {
MapAngleRange = 0,90,19
Distance = 2
NormAngle = 10
Offset = 200
}
ABEC.Polars:SPL_V_norm = {
MapAngleRange = 0,90,19
Distance = 2
NormAngle = 10
Offset = 200
Inclination = 90
}
ABEC.Polars:SPL_D_norm = {
MapAngleRange = 0,90,19
Distance = 2
NormAngle = 10
Offset = 200
Inclination = 42
}
; ____ Ath Report ____
Report = {
Title = "asy 265 mm 50 deg waveguide"
Width = 1200
Height = 800
NormAngle = 10
}
; ____ File Output ____
Output.ABECProject = 1
Output.STL = 0I don't know what you mean by a "shorter termination" but it's no problem -
To answer the question how difficult it will be to add an anclosure to an existing project - a few more lines of code.
Only one acoustic source is possible at the moment (which I don't see as a limitation).
One has always to be careful about mesh quality but that's not difficult.
To answer the question how difficult it will be to add an anclosure to an existing project - a few more lines of code.
Only one acoustic source is possible at the moment (which I don't see as a limitation).
One has always to be careful about mesh quality but that's not difficult.
Last edited: