Con is used to model a conical flare horn segment. See Attachment 1 for examples of conical horns.
Exp is used to model an exponential flare horn segment. See Attachment 2 for examples of exponential horns.
Par (parabolic flare) is used to model a rectangular cross-sectional area horn segment having two parallel straight sides and two sloping straight sides.
Other available horn flares are:
Bes - Bessel
Hyp - Hyperbolic-exponential
Lec - Le Cléac'h
Obl - Oblate spheroidal
Rad - Radius
Sph - Spherical wave
Tra - Tractrix
Attachments
The results will be a little different because the simulation uses horn segments to specify the flared port, meaning that the internal end correction is not taken into account.
Do we get a physical description of the flange? Width? Angle? Depth? I have read obout flanges and corrections for 40 years. Never seen numbers. I generally do a large round over or thumbnail profile. I guess a chamfer could also work.
As far as i know it's just a flat (no roundover) flange. E.g. a tube flush mounted to enclosure in contrast to a tube protruding into (out of) the box.
You didn't post a PAR enclosure because 99% of builds have 2 parallel straight sides. If anything, HR should default to PAR and the modeler would have to switch to the other horn profiles if they are going to build a non-PAR enclosure.
Wow, not every day you get a picture ref by David of something you've created🙂
I of course used Hornresp👍 to model them and get the curve data etc.
The generally-accepted standard end correction formulas (as used in Hornresp) assume that a flanged tube terminates in an infinite baffle, and that the end of an unflanged tube is free.
I didn't post a Par enclosure example because you queried the need for the Con and Exp options, not for the remaining Par option. I assumed that you already understood the need for the Par flare because you had stated that "99% of BUILT models are for rectangular enclosures".
There are not many cases in Hornresp where the segment flare type is automatically specified by the program, so the concept of having a default flare does not generally apply. One example of the flare being set by Hornresp however is when a paraflex horn template is created using the Input Wizard and conical segments are specified. Conical segments are used simply because given the options of Con, Exp or Par, Con is the first alphabetically and there is no certainty that the built system will use parabolic segments anyway. It is left up to the user to alter the flare in the simulation model as necessary. Making Par the default flare profile anywhere in the program, is not going to happen.
Hornresp Update 5540-240429
Hi Everyone,
BUG FIX 1
The system volume for offset driver BP6S and BP6P loudspeakers was not always shown to three decimal places on the schematic diagram. This has now been fixed.
BUG FIX 2
When the L12 value for a stepped segment PH or TH loudspeaker was changed in the loudspeaker wizard the S2 value was copied to S2S and the S4 value was copied to S4S. This problem along with a number of other similar related issues has now been resolved.
My thanks to Bluewoldemort for reporting the bug in Post #14,622.
Kind regards,
David
Hi Everyone,
BUG FIX 1
The system volume for offset driver BP6S and BP6P loudspeakers was not always shown to three decimal places on the schematic diagram. This has now been fixed.
BUG FIX 2
When the L12 value for a stepped segment PH or TH loudspeaker was changed in the loudspeaker wizard the S2 value was copied to S2S and the S4 value was copied to S4S. This problem along with a number of other similar related issues has now been resolved.
My thanks to Bluewoldemort for reporting the bug in Post #14,622.
Kind regards,
David
Having two parallel straight sides does not automatically qualify as being a PAR segment. If the two other sides (given a rectangular cross section) are also parallel, it doesn't matter what kind of segment is specified, because they all behave like straight tubes in that case. If the two non-parallel sides are straight and expanding, the segment will be PAR, but if they expand in a more complex way (i.e. not straight lines), the shape will be CON, BES etc.
No built paraflex enclosure has a horn section that look like the CONical picture below.
I was not asking you to change the default horn profile. However, it would help new HR users learn the difference between what is modeled and what is built. It also prevents folks from accusing HR for their built enclosure not measuring right.
Good thing Filter Wizard doesn't default to Bessel! Bessel is before Butterworth alphabetically.
No CONical horn profile has parallel walls. ALL square and rectangular enclosures are parabolic due to the 2 parallel sides.
B and C are the only 2 horn profiles that can be folded into a square or rectangle.
B & C = PARabolic.
Just to clarify, your CONical picture shows two exponential horns, not conical ones.
The paraflex horn designs that I am aware of would be modelled assuming cylindrical segments, so either Con, Exp or Par could be used.
See typical example below:
Then what did you mean by the following statement in Post #14,645?
"If anything, HR should default to PAR"
I agree, which is why it doesn't.
The image below shows an exponential horn with two parallel sides in a rectangular enclosure. It could just have easily been a conical horn.
(Horn flare is defined by the rate of area expansion).
If we're talking about the geometrical axisymmetric conical horn, I agree, it does not have parallel walls. If we're talking about a horn having the area expansion of a conical horn, it can have parallel walls. As can any other expansion law.
A rectangular enclosure is only parabolic if it has two parallel sides and and the other two sides are expanding linearly, i.e. the area expands linearly with distance. In the low frequency domain, the only thing that matters is the area expansion. It so happens that for a geometry like your figures B and C, the outer part of that geometry has an area that expands linearly with distance, and the appropriate model for such a segment is parabolic. The inner part has no expansion, and the type of segment does not matter, because all types will behave the same when there is no area expansion. They basically default to straight ducts.
This is not about defining types of enclosures, but about using the mathematical model that best describes the geometry you have. I.e. simulating what you actually build. For instance, the horn in the image below has exponential area expansion, but two parallel sides. It would not be accurately modeled by a PAR segment. You could either use an EXP segment, or a series of PAR segments. (But in reality, this example expands so quickly that you need to take the curvature of the wave fronts into account when modeling it.)
BTW, did you ever wonder why it is called a parabolic segment? Because if you make an axisymmetric version of it, the walls have the shape of a parabola. See image below.
You can see that in the Hornresp schematic too, which displays the axisymmetric version of the design. So the name indicates that PAR segments are not limited to rectangular enclosures with two parallel sides.
Are you saying exponential because the spread between stepped segments could be
S1-S2 = 2" width,
S2-S3 = 4",
S3-S4 = 16",
S4-S5 = 256"
in that enclosure?
Four stepped cylindrical segments as shown in Attachment 1 can be simulated using either Con, Exp or Par. Because the segment throat and mouth areas are the same, the flare rate for each segment is zero and Hornresp doesn't care which profile is specified. My personal preference is to simply use Con in such cases.
The four stepped cylindrical segments could be roughly approximated using a single exponential segment as shown in Attachment 2, if required.
Attachments
I haven't come across one, but there is no reason why it couldn't be done.
For the example shown previously if the driver was reversed in direction and moved to the opposite side on top of the existing enclosure, a low pass filter throat chamber could be positioned behind the driver and feed down into the throat of the current horn.
The problem is 99% of people are not building cylinder shaped horn subwoofers.
Conical (straight) and exponential (curved) segments mean the top, bottom, and both sides are expanding. 99% of the people are not building those enclosures either.
The only horn profile that 99% of people are building is parabolic, whether the horn is negative, straight, or positive flare or the enclosure is direct radiator or bandpass...I E square or rectangular enclosures.
A circle can be converted to a square.
4 or more expanding sides cannot be converted to 2 expanding sides PHYSICALLY.