STME58
866 posts
Jun 04, 2014
6:50 PM
|
I broke another draw 5 and was having a discussion with Gnarly about why this happens. He is not the only one who blames it on attempting to "bend past the floor", so there must be something to this idea. Being an engineer, I of course and interested in a more detailed mechanism for this failure. My best guess is that as you try to "bend past the floor" and get no response, you draw harder and fatigue the read. I have not proof of this, just a hypothesis.
In exploring it I pulled up a model I had made of a Seydel Session Steel D reed and ran a modal analysis on it. This finds the mode shapes (shape the reed bends to as it vibrates) and the resonant frequency of those shapes . I will present them here in case anyone is interested.
this is the first mode and likely the only one you would hear in normal playing. I must have gotten the measurements pretty close as the frequency of this reed is calculated at 1111 Hz, right around a C# and would only need a bit of filing to bring it up to D (1175 HZ in ET). The colors represent the stress in the material with blue being zero and red being the highest. Those of you who break reeds probably notice that they break right where the red is in this mode.The faint lines are the unbend shape of the reed.
This mode would be analogous to playing the first harmonic on a guitar. On a guitar the second mode is one octave above the first. Reed vibration is a bit more complex than string vibration. This reeds second mode frequency of 6714 is near G# two octaves above the D of the first mode. I don't know if this ever sounds in reality.
Mode 3 is a torsional mode. At 12750Hz some of us could not hear this. I wonder if this is the squeal you get sometimes along with an overblow.
Well, that was fun, but I still have no idea about "Bending to the floor"!
Last Edited by STME58 on Jun 04, 2014 7:02 PM
|
WinslowYerxa
607 posts
Jun 04, 2014
7:06 PM
|
I think you'e right.
My way of thinking about it is that if you give a reed energy it can't dissipate in normal vibration, you stress the reed in a way that can lead to early fatigue.
If you watch an isolated reed as you play it normally and then bend it down in pitch, its amplitude reduces the farther it gets from its plucked default pitch, and it moves closer and closer to the reedplate until it can no longer chop through the plate and re-emerge.
When you make the reed open, again its amplitude reduces as you bend it up, and the reed pushes farther and farther away from the plate until it can no longer descend to chop through the slot.
In a dual reed system, you have a tightly constrained bending range defined by the default closing pitch of one reed and the lowest opening pitch of the other. Especially in a system that bends less than a semitone, such as Draw 5, it's easy to push the reed farther than either reed in the dual reed system can handle.
The squeal also occurs when players form their embouchure in such a way that they tune their oral cavity to a note that the reed can't bend to - beginners often do this with the high notes. The reed has to dissipate the energy some way and I'm guessing that the squeal is a torsional vibration.
In the case of bending too far, again you're trying to get the reed to produce a pitch it can't match and the stress leads to early failure. Often it's the opening reed that breaks, such as Blow 4 when someone leans too harder bending on Draw 4.
----------
Winslow
Find out about the 2014 Spring Harmonica Collective!
SPONSORED BY
Lone Wolf Blues Company
Rockin’ Ron’s Music For Less
BlowsMeAway Productions
Slim’s Custom Cases
HarpGear
Seydel & Soehne
X-Reed Harmonicas
MasterHarp tuning tables
Last Edited by WinslowYerxa on Jun 04, 2014 9:07 PM
|
jnorem
251 posts
Jun 04, 2014
8:29 PM
|
Really interesting. Please, keep going.
----------
Call me J
|
STME58
867 posts
Jun 04, 2014
8:35 PM
|
Winslow, thank you for taking the time to respond. I appreciate you experience on the subject. What I find a bit puzzling is that, because of the laws that govern the force vs deflection of materials (Hooke's Law) The stress is highest when the deflection (strain in engineering terms) is highest. As you noted when the excitation moves away from the reeds natural resonance (the note it is tuned to) the deflection reduces. This means that necessarily the stress reduces.
As a possibly related example, ah singer can break a wine glass by hitting the resonant frequency of the glass, this is the frequency at which the vibrations are largest for the amount of energy input. If you can't get very near to the resonant frequency of the glass, it wont break.
Two reeds in a chamber with air flowing though it is much more complex than what I modeled.In hole 5 the resonant frequencies of he two reeds are closer than anywhere else on the harp. Do you think that as you bend the 5 draw down near the pitch of 5 blow, both reeds start swinging violently, leading to the large deflection and large stresses that causes failure, possibly in either reed? This would match the "bending to the floor" observation. IF you did this on draw 3, the blow reed might start swinging as the note approached its resonance but the draw reed is further away form its resonance and might not reinforce the motion as much.
I may be able to investigate this with a high speed camera when I get some time.
Last Edited by STME58 on Jun 04, 2014 8:39 PM
|
WinslowYerxa
609 posts
Jun 04, 2014
9:12 PM
|
While the stress of vibration amplitude reduces as you bend a note, there is a second stress of the reed being held at a flexed position that differs from its position when at rest, either below (for bending down in closing mode) or above (when opening).
Bending itself I don't think is any more destructive of reeds than normal playing. It's when you give the reed something it can't handle, like trying to bend it father than it can go, or simply trying to push it to extreme amplitudes, that it causes accelerated reed fatigue. I've killed far more reeds from blasting than from bending.
----------
Winslow
Find out about the 2014 Spring Harmonica Collective!
SPONSORED BY
Lone Wolf Blues Company
Rockin’ Ron’s Music For Less
BlowsMeAway Productions
Slim’s Custom Cases
HarpGear
Seydel & Soehne
X-Reed Harmonicas
MasterHarp tuning tables
Last Edited by WinslowYerxa on Jun 04, 2014 9:13 PM
|
Gnarly
1023 posts
Jun 04, 2014
10:26 PM
|
It was nice to talk to Steve today, and glad we can reunite online.
Whenever I am cautioning a player who is breaking draw 5, I usually reference Winslow's recording on the CD which accompanies his Dummies Book--you can clearly hear him drawing 5 down to almost unison with the 5 blow.
So that reed can be bent--but like sword swallowing, if you don't get it right, it can be a real pain.
|
Gipsy
70 posts
Jun 04, 2014
10:45 PM
|
A really interesting thread. I'm no engineer, but the suggestions advanced to explain the phenomenon seem to make sense. Just wondering whether there are any non audible frequencies are at play.
|
The Iceman
1697 posts
Jun 05, 2014
5:42 AM
|
One can develop a feel for where the "floor" is on any bend...
Imagine that you are in a small closet and have a helium balloon w/string attached.
When you let go of all muscle tension/impulse (fingers release the string), the balloon will rise to the ceiling naturally. Can't go above the ceiling.
In bending down, it's like reaching out, grabbing the string and pulling the balloon down to shoulder level.
Then, you place a hand above the balloon and let go of the string.
Now, this balloon will start to move left/right because it seeks release back to the ceiling.
Your hand must make slight adjustments to keep that balloon under it - this is reflective of how you must engage muscles to keep the pitch during bending.
As your hand approaches the floor with balloon under it, it will eventually touch the floor and you will start to feel a resistance to further downward motion.
You can continue the muscle tension and start to push the balloon into the floor. However, it won't go down any further and starts to lose its natural rounded shape. If you keep pushing, it will eventually burst, which is like a reed blowing out.
The trick is to sensitive yourself to that slight resistance at the floor and not exert any more force.
The interesting fact is that, at the "floor", the pitch you produce on a harmonica is about 24 cents flat of the correct pitch of the note.
The trick is to know where that "floor" is and aim for one foot above it. This way you will be at proper note pitch and also have "headroom" beneath it to incorporate a nice vibrato if you wish.
It's all about learning to sensitize your awareness.
----------
The Iceman
Last Edited by The Iceman on Jun 05, 2014 5:43 AM
|
Mahcks
6 posts
Jun 05, 2014
5:43 AM
|
Are single reed bends more stressfull on the reed than sympathetic bends?
I've noticed that boosting an overtone will give me more volume with the same amount of breath. Will this technique, assuming it causes stranger shapes in the reed, affect the reed's life span?
|
HarpNinja
3890 posts
Jun 05, 2014
6:38 AM
|
So when looking at reed efficientcy, especially as it pertains to air flow, what's the impact of embossing on the stress of the reed?
I've heard two different thoughts on embossing from two very well respected customizers - the first school is that tolerances should be tightest near the rivet end. The second is that those should be looser compared to the free end of the reed.
Does extreme embossing near the rivet end, where the reed is taking the most stress, help or hurt playability and longevity? What about the dampening effects of things like wax and polish?
IMHO, you can reach a point of diminishing returns either way.
----------
Mike
My Website
My Harmonica Effects Blog
Rock Harmonica Lessons
|
walterharp
1398 posts
Jun 05, 2014
6:47 AM
|
super cool simulations!
If you kill the 5 draw regularly, it also could have to do with the half step to the blow making it so the sympathetic blow reed cannot relieve the stress....as you said the interaction with the other reed is the key as that is what leads to bending... how ofter do people blow out 3 draws, the note with the biggest bend?
|
The Iceman
1700 posts
Jun 05, 2014
6:53 AM
|
you got it, walterharp.
Reeds that blow out are usually 5 inhale, 4 inhale, 6 inhale and the high end blow bends.
----------
The Iceman
|
Littoral
1087 posts
Jun 05, 2014
7:13 AM
|
This circle of people might know!
As OT as this may be I have never read anything definitive about how bending to get "all" notes was an intentional development in the design of the diatonic harp, much less the eventual chromatic "discovery".
How is it that this could happen AND work so well in proper pitched steps? I would expect bending to work more like a fretless bass. The closest answer I've gotten in the past was that bending was an accident. I haven't come close to accepting it was an accident but I haven't seen Jesus in my toast yet either.
Last Edited by Littoral on Jun 05, 2014 7:14 AM
|
The Iceman
1701 posts
Jun 05, 2014
7:19 AM
|
I used to give a seminar called "Harmonica Paradigm Shifts - the instrument touched by God".
Bending was never part of the equation when harmonicas were first developed.
----------
The Iceman
|
1847
1851 posts
Jun 05, 2014
8:05 AM
|
Bending was never part of the equation when harmonicas were first developed.
i hear that alot, it seems to be "collective wisdom"
but is there evidence of that?
is it possible in 1847 someone may have bent a note?
the person who came up with Richter tuning
obviously had a great deal of musical knowledge.
----------
i get a lot of request when i play my harmonica
"but i play it anyway"
|
The Iceman
1702 posts
Jun 05, 2014
8:11 AM
|
My "thesis" is that harmonica tuning was created to play the pop tunes of the day...These were German oom-pah type songs based around the 1 and 5 chord. The melodies were diatonic ones, so a major scale was included.
Harmonica was created to play in 1st position (although there was no concept of "positions" back then).
----------
The Iceman
|
1847
1852 posts
Jun 05, 2014
8:23 AM
|
Harmonica was created to play in 1st position (although there was no concept of "positions" back then).
all i am asking is.... where is the evidence of that?
i hear that all the time.
when richter tuning was introduced,
the "modes" in this case the other one's
had been in use for a thousand years
----------
----------
i get a lot of request when i play my harmonica
"but i play it anyway"
----------
i get a lot of request when i play my harmonica
"but i play it anyway"
|
The Iceman
1703 posts
Jun 05, 2014
8:27 AM
|
No hard evidence that I've found, but my thesis seems to make logical sense.
Winslow?
----------
The Iceman
|
Frank
4455 posts
Jun 05, 2014
8:34 AM
|
This is MBH...we don't need no stinking evidence here - that's why we live, breathe and thrive as a "harmonica community" here :)
What we need is more uncommon sense and educated speculation.
That my friends will get to the bottom of every mystery involving the harmonica and or climate change...
Last Edited by Frank on Jun 05, 2014 8:35 AM
|
WinslowYerxa
610 posts
Jun 05, 2014
9:29 AM
|
The earliest instructional materials I have are from 1870, when a few books were published in Chicago. The preponderance, if not all, of the tunes are for first position playing, and no mention is made of bending.
Interestingly, harmonicas are depicted with the name "Richter" on the covers, but the note layout differs in both the top and bottom octaves.
The first direct evidence we have in any form of note bending is in the sound recordings made in the early 1920s. A few years later, around 1929-1930, short "easy tricks" sorts of books were published that mention "choking" or "blueing" a note.
----------
Winslow
Find out about the 2014 Spring Harmonica Collective!
SPONSORED BY
Lone Wolf Blues Company
Rockin’ Ron’s Music For Less
BlowsMeAway Productions
Slim’s Custom Cases
HarpGear
Seydel & Soehne
X-Reed Harmonicas
MasterHarp tuning tables
|
Littoral
1089 posts
Jun 05, 2014
9:30 AM
|
The fact that the blow reeds produces the bent tones is insane. The functionality of the intervals works so well that it can't be an accident but there's no evidence it was by design.
That said, since there is nothing unique about specific frequencies our ability to articulate them to pitch is on us. Muscle memory and the ability to hear that well is magic.
|
HarpNinja
3891 posts
Jun 05, 2014
10:24 AM
|
I'd wager it is a happy accident. Those things do, in fact happen. Case in point, overbends. If someone really understood you could bend by design, the would have then, logically, attempted bending up in pitch too.
The notes are there when you bend, but it is imperfect in timber, pitch, and pattern.
----------
Mike
My Website
My Harmonica Effects Blog
Rock Harmonica Lessons
|
barbequebob
2582 posts
Jun 05, 2014
10:40 AM
|
The phenomenon of the blow reed doing the bending when you're trying to bend the draw note is absolutely dead on true and the first I ever read about it was when the very first Lee Oskar tool kits were issued around 1986-87, which made me scratch my head because it didn't make sense at all to me at the time.
However, what I did was to remove both cover plates and play 4 draw, the single most common note used for bending, especially by beginners, and so what I did (and this same phenomenon is stated in the LO tool kit was to place my finger on the 4 blow note as soon as the bend starts happening, and lo and behold, the bend completely stopped happening, and then I went on to do a blow note bend on 9 blow, and wen I put a finger on 9 draw, the bend completely stopped, and so it is the relationship of the two notes in the hole that makes it all work).
Let's face it, the instrument was never designed to do 90% of what harp players are presently doing at this moment in time and the idea of the instrument was to play nothing but very simple stuff and things likepositions, bending and overblowing were things the instrument were never designed to do. Even guitar strings were never really designed to bend notes for that matter either when one sits down and thinks aout it for a minute.
HarpNinja, I have to agree with you on the last sentence of your post, and a similar thing happens when bending guitar strings as well.
----------
Sincerely,
Barbeque Bob Maglinte
Boston, MA
http://www.barbequebob.com
CD available at http://www.cdbaby.com/cd/bbmaglinte
|
isaacullah
2791 posts
Jun 05, 2014
10:42 AM
|
Nice! I love this kind of simulation stuff (I do social simulation modeling for a living), and it makes it much clearer what's actually going on. I love that you are making this public too... I wonder if the "big boys" have done this sort of analysis too? If so, they sure didn't release the info! :)
----------
YouTube! Soundcloud!
|
Littoral
1090 posts
Jun 05, 2014
11:01 AM
|
STME58, these reed images would make a great logo (hat, avatar, shirt etc.). Any chance you could post an image of mode 2 without the parameters chart and the "w"?
|
STME58
868 posts
Jun 05, 2014
11:19 AM
|
It looks like BarbequeBob took Littoral's comment to mean he did not believe the blow read was involved in the draw bend. I took it more along the lines of a Tom Clancy quote "The difference between fiction and non-fiction is, fiction has to make sense!"
Either way, here is a link to a study on the subject.I think someone on this site pointed it out to me and I bookmarked it.
> Technical paper on bending
The got bends and overblows using a 60CC syringe to simulate the oral cavity. It might be interesting to use a similar rig to investigate reed life vs. depth of 5 draw bend.
Littoral, I can post am image as you described when I get a chance. The technical side of me wonders why you want mode 2 as that is not one that likely ever applies in a harmonica.
Last Edited by STME58 on Jun 05, 2014 11:20 AM
|
JustFuya
257 posts
Jun 05, 2014
11:19 AM
|
This is great stuff. I'd love to see a reed tweaked to the breaking point on a high speed camera. This video is interesting but doesn't satisfy my lust for destructive testing (on someone elses harp, of course.)
|
STME58
869 posts
Jun 05, 2014
11:36 AM
|
Here is one I have posted before of a Special 20 with opened covers. I need to figure out a system of mirrors that will let me see both reeds at the same time.
Slow Motion Reed
My company is toying with the idea of getting a second camera that could be synced with the first. If they do that I might be able to access them for an after hours G job. They actually encourage engineers playing around like this because what I learn on these G jobs frequently transfers to work related design and analysis.
Last Edited by STME58 on Jun 05, 2014 11:52 AM
|
Littoral
1091 posts
Jun 05, 2014
12:22 PM
|
The Slow Motion Reed video, the range of motion is amazing.
*Mode 2 image: yes, I noted that it wouldn't happen in reality but the bend in 2 is more visible than it is in 1 and 2 also shows the stress color better. 1 is cool too and maybe better because it is correct.
Last Edited by Littoral on Jun 05, 2014 12:26 PM
|
WinslowYerxa
612 posts
Jun 05, 2014
12:35 PM
|
The Bahnson/Antaki paper linked above cites an earlier groundbreaking study by Robert Johnston, published in in Acoustics Australia:
http://mis.ucd.ie/Members/RJohnst/oldharp.pdf
----------
Winslow
Find out about the 2014 Spring Harmonica Collective!
SPONSORED BY
Lone Wolf Blues Company
Rockin’ Ron’s Music For Less
BlowsMeAway Productions
Slim’s Custom Cases
HarpGear
Seydel & Soehne
X-Reed Harmonicas
MasterHarp tuning tables
|
barbequebob
2585 posts
Jun 05, 2014
12:39 PM
|
The two videos are nice to see and that technical paper on bending I remember seeing a while back. Good stuff!!
----------
Sincerely,
Barbeque Bob Maglinte
Boston, MA
http://www.barbequebob.com
CD available at http://www.cdbaby.com/cd/bbmaglinte
|
Philosofy
554 posts
Jun 05, 2014
6:47 PM
|
Just because I like to think out of the box, could the reason for early failure be sympathetic vibration from the opposing reed setting up some kind of undamped harmonic motion that stresses the reed to the breaking point? Perhaps the tone produced by the opposing reed is just the right frequency to push the breaking reed at the top of its vibration arc.
|
STME58
872 posts
Jun 05, 2014
10:57 PM
|
Here you go Littoral
Mode1
Mode2
Filter, this is not spam, please allow post!
|
nacoran
7797 posts
Jun 05, 2014
11:28 PM
|
Okay, my follow up question has to do with the theory I've proposed a couple times about the battle between steel and brass. Certain metals always fatigue every cycle, but a couple- steel and titanium, won't fatigue at all as long as they don't exceed a certain threshold. My theory has been that perhaps the people who are having luck with steel reeds are gentler players who are wrecking them, staying below that threshold. So, what is the deflection of a reed, (do you measure the deflection across the whole reed for something like that, or just at any given point?)
I still want to see titanium reeds, or maybe some of the liquid metals they use in golf clubs that are supposed to rebound so much more energy than regular metals.
----------
Nate
Facebook
Thread Organizer (A list of all sorts of useful threads)
First Post- May 8, 2009
|
STME58
873 posts
Jun 05, 2014
11:37 PM
|
Winslow wrote "Bending itself I don't think is any more destructive of reeds than normal playing. It's when you give the reed something it can't handle, like trying to bend it father than it can go, or simply trying to push it to extreme amplitudes, that it causes accelerated reed fatigue. I've killed far more reeds from blasting than from bending."
I agree with the first sentence of this. It gets a little tricky to parse because the same word, bending, can mean to reduce the pitch frequency or to change the shape of the reed. Changing the pitch frequency away from the resonant point (pitch bending) I would expect to result in less stress on the reed. The idea of bending the pitch "further than it can go" is a bit strange to me because as the forcing frequency moves further from the resonant frequency, the amplitude of the vibrations an the stress will be reduced. I have to admit I really don't understand exactly what the forcing frequency is. I suspect it might be a Helmholtz resonance generated by the volume of the oral cavity and the reed channel. Here are some calculations showing what oral cavity volume you need to resonate a various frequencies with 6mmx6mm reed slote 10, 15 an 20mm long Helmholtz Calcs
If you bend the reed further than it can go as in apply a load until it yields, that fits my understanding. Clearly the reed is getting more than it can handle as it fails, but just what is the mechanism?
My best guess at present is along the lines of what Philosofy posted. When the draw reed is bent down to the pitch of the blow reed, the blow reed is being excited at its resonance frequency. It seems reasonable to expect that this could cause large deflections in the reeds, especially on hole 5 where the draw reed is only a half step away from its resonant frequency at this point.
To prove or disprove this, all I need is high speed video of both reeds as a note is drawn down to the blow note. If this hypothesis is correct, the amplitude of the vibrations would be seen to increase.
Getting both reeds on camera is tricky. I might be able to use a couple of mirrors to show one of the reeds but I need to keep the light path from the lens to the reed similar in length ad the depth of field is quite small. In the Slow Motion Reed video above you can see how the rivets are out of focus and the tips of the reeds are in focus. TO get that image I had so much light on the harp it was hard to look at. I had to close my eyes to play it.
Last Edited by STME58 on Jun 05, 2014 11:47 PM
|
STME58
874 posts
Jun 06, 2014
12:25 AM
|
Nate, the deflection of the reed depends on how you load it. When you plink a reed (End load)it deflects to a slightly different shape than when you apply air pressure to one side (distributed load). The formulas for the shapes are in any good engineering handbook.
A reasonable first pass assumption for a harp read is an even pressure across the entire reed. In reality I am sure the pressure changes near the tip as the reed opens and closes.
Distributed load reed shape
The analysis in the above link shows that a brass A reed will last a little over 6 hours if played constantly at about .6PSI. It also shows the shape a flat reed will take under air pressure. I expect this curve is also the shape you would want your reed in at rest if you want it to enter the slot flat. Perhaps some of you customizers can tell me if that looks like a proper reed profile.
|
arzajac
1394 posts
Jun 06, 2014
3:06 AM
|
"I still want to see titanium reeds, or maybe some of the liquid metals they use in golf clubs that are supposed to rebound so much more energy than regular metals. "
I think that would likely squeal more than anything else! I think one of the reasons a Marine Band sounds so good is the balance between brightness and dampening.
"Just because I like to think out of the box, could the reason for early failure be sympathetic vibration from the opposing reed setting up some kind of undamped harmonic motion that stresses the reed to the breaking point? Perhaps the tone produced by the opposing reed is just the right frequency to push the breaking reed at the top of its vibration arc."
" It seems reasonable to expect that this could cause large deflections in the reeds, especially on hole 5 where the draw reed is only a half step away from its resonant frequency at this point. "
"To prove or disprove this, all I need is high speed video of both reeds as a note is drawn down to the blow note. If this hypothesis is correct, the amplitude of the vibrations would be seen to increase"
I was thinking I could do an objective comparison using Antaki's technique. I'd get a fair amount of brand new 5-hole reeds and set up an air pump to play the draw note at a particular flow until the reed fails. Reset with new reeds and then I would open up the resonance chamber to bend the note, say, 15 cents and repeat. Bend further and repeat. Maybe do this using four different reeds for each pitch to see if the time-to-failure is reproducible.
You wouldn't have to observe the reeds (as you say, that's technically challenging). We know what the reeds are doing by what pitch we are playing (no bend, bend, more bend, all the way to the floor). It would bve interesting to find what pitch causes the most stress and makes the reed fail fastest.
And this could be reproduced using the various brands reeds, too to compare. As Nate says, maybe steel reeds fracture under different stresses than brass. Anyway, having an objective measurement of time-to-failure between brands would be cool in of itself.
A big variable to control for would be reed shape. The reeds would all need to have the same shape both "profile" and curve.
Hohner reeds cost me $3 each. Anyone want to help crowdsource this project? (Lol!)
----------
Custom overblow harps. Harmonica service and repair.
Last Edited by arzajac on Jun 06, 2014 3:07 AM
|
MindTheGap
452 posts
Jun 06, 2014
4:26 AM
|
STME58 - Excellent! I do hope you continue with this and post the results. I'm not a metallurgist so I'd just ask how much you think that, in a real-word reed, the tuning scratches especially down at the rivet end might reduce the number of cycles before failure?
arzajac - I assume that what you propose is a thought experiment :-), but anyway...I used to work in an area that involved looking at component failure. In real world systems there is the concept of mean time to failure - i.e. a single number, but that usually comes from testing very large numbers of components and working out the probability curve. Filament light bulbs were the classic example. Some fail early, some late but there is a distribution. The manufacturers did the experiments so they could work out the economics of their business.
Unless you establish the width and shape of that curve, then drawing any conclusions around comparisons is dodgy. You could easily convince yourself of something from a few results and be wrong.
----------
mtg
|
MindTheGap
453 posts
Jun 06, 2014
4:30 AM
|
...Also, I like the conclusion that although it feels like bending a note is vexing the harp, it's actually stressing it less. Very good.
|
isaacullah
2792 posts
Jun 06, 2014
5:37 AM
|
About the hypothesis that the 5 hole (and higher holes) have reeds closer in pitch, and that the "bend" that is available is therefore more stressful, and hence why those reeds break. I've tuned a few harps now to "Easy Third" tuning, which replicates the notes of the middle octave in the lower octave. One of the results is that hole 2 is now the same tuning arrangement as hole 5. I haven't busted a reed in a long time, but I wonder if Hole 2 on these harps will suffer the same stresses as hole 5. On a "normal" harp, hole 2 almost NEVER breaks (I can't even think of an example I've heard of where hole 2 reed has broken). I wonder if it's more a function of the length to width to thickness profile of the reed than "bending stress". Or perhaps the two things are related? It makes sense that a longer reed has more area/volume to dissipate stresses, especially as the tensors change during reed movement. Having more area/volume would seem to increase the threshold to reach the critical shear stresses, right? But on the other hand, longer reeds swing more, and thus have larger stresses acting on them... hmm....
----------
YouTube! Soundcloud!
Last Edited by isaacullah on Jun 06, 2014 5:39 AM
|
arzajac
1395 posts
Jun 06, 2014
6:00 AM
|
"I assume that what you propose is a thought experiment"
I think it's doable. As you mention, the larger the number of samples (reeds), the more accurate the results. If I was to do this on my own, without help from others, I couldn't finance a very large number of reeds and so I would look for reproducibility and stick to looking for very simple answers. But if others wanted to pitch in, the sample size would go up and the results be much more accurate.
Also, there is so much more that could be measured. How does embossing affect reed life? Changing the reed shape? Comb material and chamber size? There are many more variables... Each of which requiring many new reeds to test.
Here's my keep-it-simple proposal:
Using a continuous flow to create a pitch in a hole of a harmonica, I would measure how long it took for reed failure to occur. I would record the pitch to a hard drive so that the characteristics of the changes in pitch at failure can be measured and compared with future samples.
Using an adjustable resonance chamber, I would alter the pitch (bend) to a series of samples to determine at what pitch reed failure occurs soonest. New reed(s) and a constant flow would be used for every sample so the only variable would be pitch.
The flowrate (not pressure) would probably be quite high since I reckon normal flow rates would not cause reed failure in a reasonable period of time. I suppose a flow rate that would cause reed failure in about 24 hours would be reasonable and allow the experiment to be carried out in a timely fashion. If that flowrate was very different than "normal" playing flow (like say, 10 to 100 times higher) then I would have to question the conclusions drawn and reevaluate the process of the experiment.
One other thing I am not sure about at this point is whether I would change both reeds after failure or just the failed one. Ideally I would change both, but that may impose financial constraints on the project. There's only so much one person can do...
----------
Custom overblow harps. Harmonica service and repair.
Last Edited by arzajac on Jun 06, 2014 6:01 AM
|
Sherwin
170 posts
Jun 06, 2014
6:16 AM
|
@arzajac......and the cost and trouble of conducting this test inside a soundproof chamber so that you don't drive your family crazy!
|
nacoran
7798 posts
Jun 06, 2014
11:40 AM
|
Again, as a thought experiment, since my math is awfully rusty- we've discussed reed thickness in other posts suggesting that a thinner reed (quality of sound out of the equation) may actually last a bit better- it's easier to bend a piece of aluminum foil and unbend it than is a 1/2 inch thick piece, having to do with, I think, the travel distance over a fulcrum point in the metal piece?
I can picture that in my head, but when you are changing the rate at which the reed swings you are, presumably changing where the reed is bending (or is more advanced aerodynamics involved). If the reed is bending in different place, is that, well, to use the example of the aluminum foil vs. the 1/2 thick piece again, if you want to bend the 1/2 thick piece X number of degrees it would be ridiculous to try to bend a 3 inch long piece, but you could bend a much longer piece without creasing it. If it's the opposite reed actually sounding, a bend is actually raising the pitch of that reed, shortening the vibrating section.
Would you then maybe want to have the lower reed in particular get thinner as it gets farther away from the rivet end, to take advantage of the aluminum foil aspect?
(Also, would a reed that dipped slightly in the middle along it's be less likely to experience torsional motion, like a bridge, where the edges are built up but the deck in the middle is thinner?)
Or I'm making things up with the little off brand CAD machine that is my brain. It's quite possible I'm imagining things.
----------
Nate
Facebook
Thread Organizer (A list of all sorts of useful threads)
First Post- May 8, 2009
|
STME58
875 posts
Jun 06, 2014
1:57 PM
|
Nate, nice description of the relationship between thickness and bending stress. It is counter-intuitive that sometimes making something thinner makes it less likely to break but you seem to have the principle down.
I tried modeling you idea of a reed thinned in the middle. I put a scoop on the top and the added material to the bottom until I got the same first mode resonance frequency of the original, 1110 Hz.
Scooped Reed Modal Analysis
I was surprised to see that the resonant frequency of the torsional mode went way down from the constant thickness reed. It is interesting to note that there is almost no stress in the thinned area of the reed. This might make a very responsive reed. It would be more likely to go into torsional vibration mode though. I wonder what would happen if I put the thicker section in the center? I'll try that later.
@arzajac, your study idea sounds good. Do you have the equipment space and time to pull that off? If you do, I would be willing to help out with the design of the study and the cost of a few reeds.
Last Edited by STME58 on Jun 06, 2014 2:19 PM
|
