JohnCockburn Posted April 1, 2015 Report Posted April 1, 2015 Mission Impossible? FEA.JPG No, a standard technique.
Peter K-G Posted April 1, 2015 Report Posted April 1, 2015 No, a standard technique. Maybe you missed the point in the image I posted? I know what FEA is, It might not be applicable on violin analysis
Carl Stross Posted April 1, 2015 Report Posted April 1, 2015 Maybe you missed the point in the image I posted? I know what FEA is, It might not be applicable on violin analysis Tuning works and FEA doesn't ??
Peter K-G Posted April 1, 2015 Report Posted April 1, 2015 Tuning works and FEA doesn't ?? Yes Carl, as it stands today tuning works (* better, surprised?* definition of works needs to be clearified
Carl Stross Posted April 1, 2015 Report Posted April 1, 2015 Yes Carl, as it stands today tuning works (* better, surprised? * definition of works needs to be clearified Can you prove it ?
Lars Silen Posted April 1, 2015 Report Posted April 1, 2015 Happy to report some results about my regraduation effort for adding meat to D and A. I have received suggestions from a few MNers. Assimilating them all led to lopping off 0.3mm from the top plate almost everywhere and lowering the bass bar by a couple of mm. I ended up getting significantly more meat on D and noticeably more meat on A. Overall, the balance is fine. Maybe a little bit of bridge tuning will suffice. It's now a much more powerful violin than before. M5 (with bar) is now 350 Hz. The only annoying thing is that the Warchal Amber strings have this metallic twang that doesn't seem to go away. So it looks like the top was just a little too stiff and too thick. A big "thank you" to those who offered suggestions. I have heard the metallic "twang" ... if I understand the description correctly for example on the D-string when the tap tone of the centre part of the g-side f-hole is clearly higher than the corresponding tap tone of the X-mode on the lower bout area. I would thin the area over the X-node between the tail piece and the widest part very slightly from the inside to get the tap tones closer. My question is then: Is the tap tone at the middle inner side of the f-hole clearly higher than the wide area beside the taipiece when you follow the fibers? If the tap tones are the other way around then I don't think my diagnosis is correct.
Don Noon Posted April 1, 2015 Author Report Posted April 1, 2015 I disagree. The free plate mode shapes and frequencies do transfer into the final assembled violin. But I do agree that plates M5 are linked to assembled violin they are called B1- for the top and B1+ for the back Transfer... Linked to... Yes, there IS definitely a modest correlation between the low plate modes and the lowest structure modes. You could probably also make a similar correlation between average thickness and B modes, too. So The heck What It's not a fabulous correlation, though, due to the aforementioned intervening little perturbations of ribs, blocks, neck, and soundpost. And the free plate modes do not really show up in the finished instrument... only some that sortof vaguely look like it, but are not really. And until anyone EVER shows that particular B mode frequencies are the key to wonderfulness, I will continue to concentrate on the other stuff. This is a red herring in my view. (this is not to say that if you make a 150g top plate and have B+ at 700 Hz it might be as good as anything; I'm talking about just being in the reasonable range, which is fairly large)
Peter K-G Posted April 1, 2015 Report Posted April 1, 2015 Yes Carl, as it stands today tuning works (* better, surprised? * definition of works needs to be clearified Can you prove it ? Sure I can, but first we need to agree on what to prove
DGV Posted April 1, 2015 Report Posted April 1, 2015 I have heard the metallic "twang" ... if I understand the description correctly for example on the D-string when the tap tone of the centre part of the g-side f-hole is clearly higher than the corresponding tap tone of the X-mode on the lower bout area. I would thin the area over the X-node between the tail piece and the widest part very slightly from the inside to get the tap tones closer. My question is then: Is the tap tone at the middle inner side of the f-hole clearly higher than the wide area beside the taipiece when you follow the fibers? If the tap tones are the other way around then I don't think my diagnosis is correct. I will try some tapping tonight or whenever I get a chance. But it is most probably the strings since I don't get this metallic twang with my old Dominant.
Carl Stross Posted April 1, 2015 Report Posted April 1, 2015 Sure I can, but first we need to agree on what to prove We already agreed : prove tuning works better than FEA.
Don Noon Posted April 1, 2015 Author Report Posted April 1, 2015 We already agreed : prove tuning works better than FEA. That's way too easy, judging by the FEA of the Titian. I could probably come closer to the signature modes building blindfolded with a sharpened spoon. (not that the signature modes matter much, but the important stuff is beyond calculation)
Carl Stross Posted April 1, 2015 Report Posted April 1, 2015 That's way too easy, judging by the FEA of the Titian. I could probably come closer to the signature modes building blindfolded with a sharpened spoon. (not that the signature modes matter much, but the important stuff is beyond calculation) Let's see how Peter does.
Don Noon Posted April 2, 2015 Author Report Posted April 2, 2015 I would rather see it proven that accurate taptones or signature mode frequencies matter. The Cremonese (or the folks that reworked their fiddles) obviously weren't very consistent about it. So, even if tap tuning is better than FEA at getting target signature modes, I still see it all as a useless distraction from figuring out how to make good violins.
Peter K-G Posted April 2, 2015 Report Posted April 2, 2015 Yes Carl, as it stands today tuning works (* better, surprised? * definition of works needs to be clearified Sure I can, but first we need to agree on what to prove We already agreed : prove tuning works better than FEA. Still missing: 1. Definition of works 2. What to prove Lets not continue this in Don's thread it could be a long one
JohnCockburn Posted April 2, 2015 Report Posted April 2, 2015 We already agreed : prove tuning works better than FEA. What Colin Gough does in the paper that Marty refers to is use computational modelling to investigate how the individual modes of the components couple to form the radiative modes responsible for sound output in the lower octaves. It is not an attempt to explain how to produce a "great" violin from first principles. It really has nothing to do with the claims of the tuners that "great" violins can be produced by accurate placement of the signature modes. So in this context, the challenge to "prove that tuning works better than FEA" is meaningless.
Carl Stross Posted April 2, 2015 Report Posted April 2, 2015 What Colin Gough does in the paper that Marty refers to is use computational modelling to investigate how the individual modes of the components couple to form the radiative modes responsible for sound output in the lower octaves. It is not an attempt to explain how to produce a "great" violin from first principles. It really has nothing to do with the claims of the tuners that "great" violins can be produced by accurate placement of the signature modes. So in this context, the challenge to "prove that tuning works better than FEA" is meaningless. Sure, but I am not talking ( or interested ) in that context or anyway related to what Colin Gough does. I am talking about using FEA to "design" the violin from first principles. Or rather to "optimize" given that we need to stick with an acceptable shape. I think the Peter Kreit method, which Peter K-G is using and drumming here, is simply not working at all. That's why we never got any tone samples. It's not that Peter is nasty , it's just that his violins do not sound good enough for his level of expectation. Of course, it is implied here that Peter has access to the capabilities needed to make a violin sound, put it to work - that's not at all trivial.
JohnCockburn Posted April 2, 2015 Report Posted April 2, 2015 Sure, but I am not talking ( or interested ) in that context or anyway related to what Colin Gough does. I am talking about using FEA to "design" the violin from first principles. A good illustration of one of the big problems for "violin science". The apparent lack of interest in the essential spade work that needs to be done to lead us incrementally towards full understanding (which is how real science usually advances).
uncle duke Posted April 2, 2015 Report Posted April 2, 2015 ...... Or rather to "optimize" given that we need to stick with an acceptable shape. It should be there in Gough's paper by the sounds of it. Carl, you say you don't care to play a violin anymore, you also say you don't care to try your hand at making one either. You say you are just interested in what makes a violin go. Seems to me Don could keep going with his "new plate" experiments and surpass what Gough has to say. I wonder if you would have the same opinion then as you do now? I doubt it. Lars is suggesting to DGV to try something with his recently adjusted violin. What's your opinion on what that may be about, Carl?
JohnCockburn Posted April 2, 2015 Report Posted April 2, 2015 It should be there in Gough's paper by the sounds of it. Carl, you say you don't care to play a violin anymore, you also say you don't care to try your hand at making one either. You say you are just interested in what makes a violin go. Seems to me Don could keep going with his "new plate" experiments and surpass what Gough has to say. I wonder if you would have the same opinion then as you do now? I doubt it. Lars is suggesting to DGV to try something with his recently adjusted violin. What's your opinion on what that may be about, Carl? So it seems to you that Don could surpass Gough's work, even though you haven't read Gough's paper? Good going.
Carl Stross Posted April 2, 2015 Report Posted April 2, 2015 A good illustration of one of the big problems for "violin science". The apparent lack of interest in the essential spade work that needs to be done to lead us incrementally towards full understanding (which is how real science usually advances). This is a bit unkind and implies you personally know how far or how deep "violin science" happens. You might not, simply because the most important chunk of today's violin science is done for profit. "Spade work" was done long ago. and certain lines of inquiry have been eliminated as lacking predictive capacity. For example, if someone takes a modal approach to violin "bettering" I could direct the respective researcher towards two result from Topology which give 100% assurance there are better ways to spend the time. There are quite a few quasi mathematical lines of reasoning ending with the same result. For me, the moment somebody gets busy with modes, it means he's either clueless or not interested as to what makes great violins tick. That's me, others may see things very differently and they're welcome. You should not understand from this that I somehow "put down" Colin's work. He's more than welcome to pursue the lines of research he choses and we're in no danger : he's not validating the models for the composites used in the Airbus 380 or simulating the flaps at various speeds, APs, temps, loads, rain, snow , fog, birds, etc etc etc or simulating one of the tires hitting a stone on the runway at various....... etc etc and estimating the amount of rubber the tire will lose. Meshing a surface, extracting normal modes and / or applying two forces is not quite where FEA happens. This is my beef with "violin science" : it uses overly simplified models and elementary tools. Sometimes, the researcher must simplify the models to match his tools and that's all fine being intended. Most of the times the researcher is simply unaware of the complexities behind the aspect he's concentrating on. Test yourself : take a look at various plate animations : what's missing from there which makes that picture downright useless ? Drop a PM when you figure it out - I'd be delighted if you got it right. ( Hint : one picture, any picture, is enough ).
Carl Stross Posted April 2, 2015 Report Posted April 2, 2015 It should be there in Gough's paper by the sounds of it. Carl, you say you don't care to play a violin anymore, you also say you don't care to try your hand at making one either. You say you are just interested in what makes a violin go. Seems to me Don could keep going with his "new plate" experiments and surpass what Gough has to say. I wonder if you would have the same opinion then as you do now? I doubt it. Lars is suggesting to DGV to try something with his recently adjusted violin. What's your opinion on what that may be about, Carl? So it seems to you that Don could surpass Gough's work, even though you haven't read Gough's paper? Good going. Actually, I think Uncle Duke got it right. From his previous paper, it seems Colin Gough is running straight down the road to nowhere. Don is on another road altogether. My bets are on Don. Don's practical results speak for themselves. Let's wait and see Colin's new paper : he might've changed course. We shouldn't anticipate beforehand.
Carl Stross Posted April 2, 2015 Report Posted April 2, 2015 It should be there in Gough's paper by the sounds of it. Carl, you say you don't care to play a violin anymore, you also say you don't care to try your hand at making one either. You say you are just interested in what makes a violin go. Seems to me Don could keep going with his "new plate" experiments and surpass what Gough has to say. I wonder if you would have the same opinion then as you do now? I doubt it. Lars is suggesting to DGV to try something with his recently adjusted violin. What's your opinion on what that may be about, Carl? (It's not that I don't care to play, I just can't. I suffered nerve damage in my left hand 37 years ago due ( probably ) to meningitis. Fingers 1, 3 and 4 were affected. 1 very little, 2 not at all. But I wasn't good at it, had no talent - no loss there whatsoever. I'm glad I didn't have to inflict my violin playing on innocent people. ) My opinion is to follow what Don says and does - I think Don is on his way to some fantastic results. As I said, my bets are on Don.
JohnCockburn Posted April 2, 2015 Report Posted April 2, 2015 This is a bit unkind and implies you personally know how far or how deep "violin science" happens. You might not, simply because the most important chunk of today's violin science is done for profit. by whom?
Carl Stross Posted April 2, 2015 Report Posted April 2, 2015 by whom? Private Companies involved in related fields.
David Burgess Posted April 2, 2015 Report Posted April 2, 2015 He's more than welcome to pursue the lines of research he choses and we're in no danger : he's not validating the models for the composites used in the Airbus 380 or simulating the flaps at various speeds, APs, temps, loads, rain, snow , fog, birds, etc etc etc or simulating one of the tires hitting a stone on the runway at various....... etc etc and estimating the amount of rubber the tire will lose. Meshing a surface, extracting normal modes and / or applying two forces is not quite where FEA happens. This is my beef with "violin science" : it uses overly simplified models and elementary tools. Sometimes, the researcher must simplify the models to match his tools and that's all fine being intended. Most of the times the researcher is simply unaware of the complexities behind the aspect he's concentrating on. Not necessarily. Some of the violin researchers are involved because the field is so complex, fascinating and challenging, compared to their professional fields of aircraft design and testing, or nuclear physics.
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now