Filament / wall thickness / design principles

[Caleb Coburn]

Hi! I'm new (to 3D printing and 3D printed violins) and have some questions.

I'm mashing up the modular fiddle and hovalin. Partly for fun and partly for aesthetics.(its for my wife, and she likes the sound from modular fiddle but I want to make something that looks more traditional).  Once I'm happy with the body I plan to print it in carbon filled polycarbonate (just cause I have the material, don't want to deal with annealing, and think it would be a fun challenge since PC is hard to work with). I'm using a 0.6 mm tungsten carbide nozzle (apparently the cool new thing in printing high temp abrasives, with thermals like brass but way more wear-resistant then hardened steel).

The hovalin reddit group is giving me feedback that carbon fill isn't so great because it adds stiffness but not hardness and argue that a hard but flexible material is key. This group is all about as stiff as possible. Not sure which is right?

Also not sure about wall thickness. I read thinner top is louder, but wood is typically thicker than the 1.8 mm the fiddle is designed at. If I change materials do I need to optimize wall thickness? PC is much stronger than PLA but not as stiff (S3D puts it between PETG and PLA).

Also, so far all my mods to the mashup violin are free hand (bass bar, sound post, f-holes, etc.) I like to think it will sound nice but have no idea what it will do in reality. I've tried to follow guidelines for normal violins or take notes from the modular fiddle design, but if there are any good tips on how to tell if I'm making a big mistake for a plastic violin I would appreciate it. Or even if prototyping in normal PLA would sound similar to the same design done in CF-PC. 

I've also thought about salt annealing it (I know I said I didn't want to anneal, but freespirit1's youtube video (https://youtu.be/nRLJ4ylGTFc) is cool so I want to try that process on something. Maybe would be good for the CF-PLA being used in this group. No idea what that would do, but if stiffness really is the goal I guess it could be good. 

Appreciate any feedback!

[David Perry]

Hi Caleb,

Thanks for posting! And welcome to the wild world of 3D printing and 3D printed violins. It's pretty fun, and can also be rather frustrating. 

What a great mashup! I'll be very interested in what you make. 

Carbon filled PC!?!?!? You are brave. What printer are you using? It seems that you know what you're getting yourself into. I get sweaty just thinking about a long PC print like the body... :-)

OK let's talk details... First, materials in general. I have not done a thorough comparison of material properties, although I think that such a comparison would be illuminating. I also do not any formal training or experience in acoustic research, although I've read a lot of papers on the subject. 

My broad understanding is that what is most important is the ratio of stiffness to density. Tonewoods (the fancy wood used for instruments) are often described by their sound radiation coefficient, which is equal to the square root (stiffness/density cubed). More info here. Thus, a stiffer and less dense material is able to make a louder instrument. 

I could be wrong here, but I would say that strength, like fatigue stress or failure stress, is relatively unimportant. You need a 'strong enough' material, but strength does not correlate to resistance to creep. Creep resistance is more about type of plastic, additives, ambient temp relative to glass transition temp, and how stiff the material is relative to the given load. If you have a very stiff material, even if it breaks easily, it is less prone to creep than a less stiff but stronger material. 

Of course, high temp resistance is another matter, because the glass transition is a key indicator of how much creep you'll experience. If you have a material with a very high glass transition temp, it will creep less at room temp. PLA, not annealed, has a low Tg, thus it is highly susceptible to creep. However, it is also quite stiff, so that helps somewhat. What you need to do is allow a way for the bending stresses to be redirected (truss rod) so that the material can creep until the rods take the bulk of the load, and then ideally the parts will reach steady state. This does not happen for the body, though, so for that you must control creep via other means. 

The user on reddit stated (I'm paraphrasing from my memory of the post) that hardness is more important, as it is the hardness that will determine how efficiently the material transfers sound waves. That's a bit of a different way to think about it, and I think that it is less relevant to the behavior of violin plates (where the plate is the top and bottom thin sections of the body. These plates are not really transmitting vibrations so much as they are being excited by vibrations and vibrating in complex patterns from this excitation. The more you can maximize their excitation, and the diversity of vibration patterns in the plates (see Chladni patterns for violin plates), the more rich sound with good volume you will achieve. The skin of a drum does not need to be hard, it needs to be tight, which is analogous to stiffness more so than hardness. 

I think hardness may come into play in the acoustics for other instruments, or for electric instruments. If you have a pickup receiving a mechanical force, perhaps then hardness for the interfacing mechanical parts is really important. I'm not sure. I don't think that the Reddit user was wrong, but I think their analysis perhaps does not apply as much to violin plates. 

Now it gets complicated. It is not always desirable to have a very high stiffness to density ratio, and the propensity of the material to dampen sound is important, too. So truly, a good enough violin plate can be crafted out of a wide variety of woods and/or plastics, but those plates would be carved differently (thickness) for the different materials and material properties. 

Perhaps PC with carbon can make a great violin, but it might need dramatically different thickness in the plates than a PLA-CF body. I don't know. If you compare the young's modulus and density of the two materials, you should get a rough idea of how well the design will translate between them. 

Wooden violins have plates with variable thicknesses. The Modular Fiddle does not. The reason is because that introduces a lot of variables and I did not get around to it. Haha, not a good reason! I would like to get to that soon, I just need some time and enough money in the bank to fiddle with fiddles for a bit. 

I think the Modular Fiddle plates are modeled at 2.4mm thick. Wood varies I think from about 2.5 to 3.5mm, with some thicker sections getting up to 4.5 or 5mm. I'm not sure where that happens or how they are graduated, but I would love to learn more about this. 

Any thinner and the instrument will be riddled with wolf tones. Any thicker and it may sound more like you're playing a plastic box underwater. My advice is to start with something that works, then make very small changes to one thing at a time. Make it easy to swap out bodies. 

Salt annealing looks awesome! great results in that video, thanks for sharing! 

-David

[Caleb Coburn]

Thanks for being welcoming and enthusiastic! Appreciate your detailed response :)

I'm using a SeckitGo2. I have ~355 mm in z the way I set it up. I've built an insulated but not actively heated enclosure; I'm hoping it can handle a large PC print.

Interesting points on material properties. I agree strength doesn't matter (unless you really need a harmonic hammer). I haven't found much info on creep comparisons, not sure how CF-PLA and CF-PC compare. PC has temp in its favor but baseline stiffness and being amorphous are against it. Still, I spend some time in Phoenix and don't love the idea of PLA. Speaking of creep, with your PLA bodies do you find significant deformation after being strung up for a long time (assuming it doesn't get too hot)? Does it constantly need to be tuned?

HIPS might be interesting if stiffness/density^3 is thing to maximize. S3D material table lists HIPS as stiffer than PLA, good up to 100C, and 20% less dense. https://www.simplify3d.com/support/materials-guide/properties-table/?highlight=hipsre:" But its similar difficuklty to print as ABS, so heated bed and enclosure are musts

Your comments on stiffness vs hardness make sense to me.

re: "Now it gets complicated. It is not always desirable to have a very high stiffness to density ratio, and the propensity of the material to dampen sound is important, too. So truly, a good enough violin plate can be crafted out of a wide variety of woods and/or plastics, but those plates would be carved differently (thickness) for the different materials and material properties."  --- In the article you linked it focused on sound radiation coefficient and flexural rigidity. So I guess if the shape of the plate is fixed but the material changes, to mimic the sound the flexural rigidity as a function of position would need to be made the same (i.e. by varying thickness) but the loudness would be controlled by the material properties?

I guess the printing details and post processing (annealing, sanding, etc) will have a huge influence on the material's stiffness, stiffness anisotropy, and potentially weight per area (e.g. for an <100% infilled plate or if the layer lines are really bulging). That probably makes it really hard to know how to design without lots of iterating, even if I had the relevant information and was doing the math. I'll probably end up taking a shot in the dark and maybe iterate if I still have interest. I might try adding variable thickness plate as well, again as a shot in the dark, probably following an image like this: https://commons.wikimedia.org/wiki/File:Violin_bout_cross_section.png

You're right about modular fiddle plate thickness, I was taking the cross section on the side, which is thinner.

-Caleb

[David Perry]

You're welcome! I love diving into the technicalities of it all. Acoustics are pretty crazy, and even super experienced luthiers don't agree on a lot of things. 

355mm! Wow, that's cool. I would like to have a tall printer someday so I could print a full size violin body. SeckitGo looks interesting. I like their tagline "your second 3D printer kit." Clever.

I'm not sure how CF-PLA and CF-PC compare for creep, either. I agree with your assessment.

I have been surprised that I do not see more creep in my fiddle bodies. I have a five-string fiddle built up with a plain PLA body, and it was under tension for 6-8 months and the body shows no deformation. Now I leave it with loose strings just because I don't play it. I have fiddles with carbon fiber PLA bodies that have been under tension for a year plus and have no visible sagging in the top plate. I do have to tune them more than my traditional wooden violin. I have one Modular Fiddle set up with planetary geared pegs (they're pretty nice), a CF-PLA body, and octave strings. It stays in tune very well. I tuned it yesterday for the first time in at least 6 months, and all of the strings were a half step flat. So that is probably coming from creep either in bending in the neck or sagging in the top plate. Hard to say which. In the end, unless the fiddle gets hot they seem pretty stable, so I haven't done much research to find out what parts creep over time. 

It drives me crazy that the fiddles aren't safe in hot weather. Otherwise we have this perfect camping and outdoor fiddle!!! HIPS is an interesting idea, it does have some great inherent stiffness. I have tried to print it but without a heated enclosure it was a bit of a joke...

"...if the shape of the plate is fixed but the material changes, to mimic the sound the flexural rigidity as a function of position would need to be made the same (i.e. by varying thickness) but the loudness would be controlled by the material properties?"

Yes and no I think, but generally I think that is correct and it fits with my understanding. Loudness also has a lot to do with the overall size of the plates. The Modular Fiddle, as designed, can't possibly be as loud as a traditional violin due to the smaller body.

Yeah, print settings and material do have a big impact. Exactly as you say -- if you are overextruding and you end up with an extra 0.25-0.5mm on your plate thickness that makes a big difference in sound.

If you want to play with variable thickness plates, I would recommend doing some searching for some measurements and such. Luthiers have measured plates on all sorts of high performance violins and that data is generally available. David Golber did some scanning and CAD work and publishes in the VSA... I found his paper "Analysis and Synthesis of Violin Arches" interesting, but honestly I don't remember how applicable that is to the topic at hand...

Keep us posted! I am cranking through a few long-standing to-do items for fiddle stuff. Just finished the 3/4, am going to print an 'elite' neck tonight (with a false bout and more rounded neck profile at the request of a luthier). I think I'll make a tiny fiddle next, maybe 1/16 or 1/8...or some body iterations...or maybe parts to turn the fiddle into a uke...

[Caleb Coburn]

Got my first successful full violin print today! Obviously not CF-PC yet (this is just the cheapest PLA I could find). I used some wood glue to partially fill some print imperfections and I'll spray paint it after. carbon rod, strings etc should be arriving over the coming week so hopefully will get a chance to string it up and hear how it sounds soon. 

I'm not concerned about this one since its just practice, but I guess spray paint will affect sound, and not likely for the better (?) 

I'm curious about the idea of infilled plates. I sliced my first (failed) print wrong and it used some partial infill on the plate. Seems like an easy way to get very stiff and light plates. Again dunno how it would sound, maybe poor for sound propagation. 

Maybe I'll order some HIPS and give it a try. If my enclosure can take it, it would be fun. I guess if I reach my goal of printing PC well enough to do a violin I should be able to print HIPS, although I hear carbon fill makes warp-prone materials easier to print. Gotta manage what I make though, I'm not going to kit out each one with hardware and I don't have space for lots of violin bodies/violins.

I redesigned a bit to make it more modern looking. Still working on the mechanism for string tension in the neck without a pegbox, but I like the look. Kinda tie fighter. Thinking of calling it the VSOX (violin shaped object X) just to get ahead of violin aficionados when it inevitably doesn't sound like a Stradivarius 

[David Perry]

That is looking amazing! 

I think spray paint may have a small effect on sound, but probably not very much. 

Re: infilled plates. I have been assuming that a plate that is partially hollow, as in with infill, would have a dampening effect on sound. I don't know if this is correct, though, and you are right that if it could be thicker, but less dense, we could increase that stiffness to density ratio that we were thinking was so important! So I guess in the end I don't know and I think someone ought to try it. :-)

I'm digging the design! Thank you for sharing pics of the CAD. Funny, first I was like, 'oh man, those f-holes look too low.' But in the end I think it's because I'm used to looking at the shorter body of the Modular Fiddle. My sense of how a violin should look is totally out of whack. 

Wait wait, no pegbox??? So are you thinking you'll get the strings up to tension somehow and then use fine tuners for your tuning needs? I'm excited to see what you come up with for a tensioning system!

Haha, yeah, violins take up space quick. I've got a wall of them, and then a couple cardboard boxes filled with old and/or bad bodies. 

[Caleb Coburn]

build update: I made a violin :) super fun. Since its just a first version made in PLA I consider it more of a prototype rather than a final piece. But a fun milestone, and I think it looks cool.

Sound - IDK? I attached a simple clip of her playing it. I don't play strings or even have a musical ear or the ability to sing on key. My wife plays viola and thought it would be fun to have a violin, which was enough to get me working on this. Maybe I'll get her to teach me the basics. 

Tuning - so the tuning mechanism I ended up with in place of a pegbox is okay but not great. It only has 15 mm of travel and it turns out these strings are pretty stretchy. on top of that, the instrument itself isn't all that stiff, so tensioning one string makes the others flat. together, that makes it hard to tune. I ran out of travel on the a string, and even the extra 2 mm on the fine tuner wasn't enough. I ended up having to tension all the strings, and then one at a time back a single string all the way off, re-tie it on tighter to take up the slack induced by stretching of the string and bending of the instrument while the other strings were still tight, and then re-tension it. With a little re-design it could be nice. I think I will get rid of the tunnel routing and make a more gentle slope around the end of the neck piece so the bend radius isn't so tight, probably need to add ~5-10 mm of travel, and maybe something better for hiding the string ends. Maybe even just a cover that encloses it all to hide it from view. A big con is you need a hex wrench to tune, but it's similar sensitivity to a fine-tuner, so even I can easily tune it (I hadn't even handled a stringed instrument prior to making this one). The tunnels got a bit closed up during printing and I mangled them a bit trying to open them back up. PLA didn't cut well for me, it was just melting around my drill.

re: hollow plates

I did a bit of math / reading about this. From what I found, plastics have >2x density and ~1/4 stiffness of an average tone wood from the Wikipedia article. Since I'm using a 0.6 mm nozzle, if I do single perimeter plates with a carbon reinforced filament, from my calculations I need about 2.7 mm thick ~10-20 % infill between the plates to get a similar stiffness / density / sound radiation coefficient. I made a bunch of assumptions to calculate that (infill modulus proportional to infill percentage, effective modulus calculated from material_modulus*stiffness_actual / stiffness_solid). Also, from an article I lost track of about trying to make sound dampeners from honeycomb board (used extensively e.g. in airplanes where sound dampening is desirable) the engineered boards were described as "virtually transparent to sound", with significant absorption only occurring for open cells that are on the order of the wavelength of the sound being damped. So if the unit cell of the infill has dimensions ~3mm on a side, the equivalent frequency assuming speed of sound in air is like 100 kHz, and human hearing only goes out to ~20 kHz. Overall I'm optimistic about this direction for getting good sounding plates from plastic, but printing/slicing would be a trick. Printing you would want to have closed cell structures without holes do delamination of wall / infill. And slicing explicitly modeled single perimeters + honeycomb infill doesn't work well and ends up looking like a retraction torture test. I think people who sell 3d printed airplane files have techniques to get slicers to make really light stiff structures, but not something I have experience in. 

It has the additional benefit of using 30-40% less plastic for the same size body, but probably also makes it much less resistant to puncture damage.

stiffness of engineered plates:

https://www.youtube.com/watch?fbclid=IwAR3FBzBx2NBd2oPyI4k9fyRVSmklwVf_Da4fv4zB_KKXV8ph7gokV3GoCkw&v=4zpQwirFsbk&feature=youtu.be

material properties

https://www.3dxtech.com/blog/updated-test-data-on-carbon-fiber-filaments/

https://www.curbellplastics.com/Research-Solutions/Plastic-Properties

https://en.wikipedia.org/wiki/Tonewood

[Mark Millard]

Sooo. Working on the neck and body for my modular fiddle.  The skin on both the neck and body didn't really adhere to the rest of the print.  Any tips on minor adjustments to make to get it to adhere better?  Pretty new to 3d printing.  Having a lot of fun!

Mark

--
You received this message because you are subscribed to the Google Groups "OpenFab Violin Builders" group.
To unsubscribe from this group and stop receiving emails from it, send an email to OF-Fiddles+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/OF-Fiddles/3992d204-fd31-4a95-adad-8b9d9b40b145n%40googlegroups.com.

[Caleb Coburn]

Hi Mark,

I'm new to 3D printing too so probably others can help more. Would be good to get more info on exactly what your problem is though. If its layer adhesion between outer perimeters and infill then maybe increasing print temperature could help. But I'm not sure what you mean by "skin didn't adhere well"

Caleb

[David Perry]

Caleb -- looking good! I love that solution for the pegbox! That reminds me of something a luthier told me once that he did as an easy tuner solution. I have sketches of it somewhere, I'll see if I can find them. 

Ahhh, so with the honeycomb plates -- you are trying to turn the top and bottom plates into sheets with a thin outer solid layer and honeycomb interior with the honeycomb parallel to the plate. I see why that is a challenge for printing! While you can't quite get a honeycomb pattern, you could maybe achieve the same effect with some carefully calibrated infill. A bit of tinkering in Simplify3D shows me that if I do 20% triangular infill at 60 degrees, I get an infill pattern that to me looks like it might work. I set it to print with no perimeters so I could better visualize the pattern in 3D. 

Hi Mark! 

Yeah, upload some pictures if you can. Are your prints of other objects going ok? I print a lot of calibration cubes when I run into issues like you describe.

-David

[mill...@gmail.com]

David, sorry for the late reply (accidentally printed in a different feed).   Here are a couple pics.   e. I tried printing the neck twice using pla at 200 deg and again 215. In both cases the top layer on the fingerboard didn’t adhere to the rest of the print.  Also, one side of the neck printed really well, but the underside was very stringy.  See pics.   Would love your thoughts.

[David Perry]

Ha, now I'm sorry for the late reply. Did you get this figured out?? 

I would recommend printing some calibration cubes (20mm cubes). Print them with 2 perimeters, no infill, and no top. Then measure the wall thickness and make sure that your extrusion thickness matches what it is set to in your slicer. If not, adjust your flow rate to compensate. 

Once that is dialed, run a smaller print that's similar to the neck and see how it goes. If that prints well, then go for the neck. 

But I imagine by now you've either moved on or figured it out...

-David

[David Perry]

Hey Caleb -- how's your project going? Got any updates?

[Caleb Coburn]

I started a new cad model from scratch and followed the violin makes guide, more or less: http://www.makingtheviolin.com/uploads/makingtheviolinmanual.pdf

I printed the body with hollow walls, and at about a 50 degree tilt. That made it lighter and easier to print without internal structures for printability, fully rounded bottom and much better intersection with triangular infill. The body was printed in PETG as a test before I use the CF-PC. It sounds different than the PLA solid body, but the models are also completely different. Haven't had much time to work on it, so I don't have a working neck design. The tailpipe on my first one broke after a week or two of being strung, so I also redid the tailpiece design. Its super simple, uses no tailgut, but has fully integrated fine tuners. 

Dunno when I'll get it strung up, work getting busy hasn't let me put much time to this recently.

[David Perry]

Wow, that is looking great! I hope you find time to keep working on it, I'm excited to see your results.

I was reminded of your idea around hollow plates, or honeycomb plates, when I saw a very interesting cello on instagram yesterday. A maker in Belgium, Tim Duerinck, is experimenting with (amongst other things) the use of different kinds of fibers in composite violin family instruments. The cello in question uses glass fiber and an aramid honeycomb. The top plate is a thin layer of rigid fiberglass on either side of a 5mm aramid honeycomb. It's almost precisely what you were suggesting! He explains the instrument in this video. 

[Caleb Coburn]

Thanks for sharing the youtube David! That was very interesting and similar to what I've done here. 

I finished a working version of the neck and got it strung up. Maybe a bit of an improvement over the first attempt's sound, definitely lighter. 

Any advice on how to iterate for sound? i.e. if I'm able to determine a certain string is weak or the instrument sounds brassy/tinny overall, etc. how to go about tuning for better sound? Kind of a big if in there since I can't play it and tell myself if its any good, but still. You have experience on that part I gather :) 

Still need to do the body in CF PC, this is the PETG one. (neck is in cf-pc though). endpiece needs a redesign, I got the angle wrong to the string doesn't contact the end. Might benefit from using actual tailgut, and I think pegs are probably better than my screw version, although I do like how fine of control I have at the neck and once the instrument settles the fine tuners hopefully are enough for regular use, if the creep isn't too bad. Kept having the e-string cut through the plastic anchor on the screw though. Don't have a good solution for that but some tin foil bunched up at the cutting point is a working bandaid for now. It even went through the pc :( I guess there is a reason for the string wrapping, which had largely come off from all the retying I did on the last one. 

Bridge is also cf-pc. Its still heavier than the manual I followed says a wood one should be, but I don't think I can get it to be stiff enough without the extra mass. Not sure on that one. 

[David Perry]

Wow that cf-pc looks amazing. Gotta love cf finishes. 

Your build is coming together! So exciting. 

Iterating for sound is quite difficult. I have tried to learn more about violin acoustics, but it doesn’t appear that there is a simple way to say “oh my G string is weak; that means I need X.” Even the most devout researchers don’t seem to be able to draw clear conclusions about what does what to violin sound. 

I am in the process of iterating a V2 Modular Fiddle. It’s a bit larger body and has thicker plates, which then use infill to cut down weight. It works great!! I’ve been posting some updates to Instagram. Caleb, I’d like to give you a shout out for the infill idea; do you have an IG handle or website or anything I can plug? My account is @dsp39 for anyone that wants to check it out. 

So from my experience, I am usually trying to find the right balance of lively, flexible plates that are stiff enough to mitigate undesirable resonance but not overly dense such that the sound is damped. 

If the plates are too flexible, there is usually much variation in tone across the range of the instrument. You’ll probably have wolf tones, and there might be strong resonance at the A0 hemholz frequency due to the air resonance pushing the plates around (my own theory). 

If the plates are too massive, your violin will lose timbre and will sound like a plastic box. Like it’s a submarine underwater is always what it reminds me of. 

Somewhere between these two there is a balance. Your material will have a big impact. I would say switch to PLA to print test bodies (if the stiffness is closer to cf-pc than the petg). Then once you get close to good enough switch to your pc. It usually takes me six iterations to get to good enough, so plan to iterate. Also it helps to build up two fiddles so you can compare sound back to back. Try to change one thing at a time (within reason) and you’ll start to understand what effect different types of changes have.

Then once you find that balance there may be some fine tuning. 

Right now I’ve printed my body with one .6mm perimeter and 30% infill in CFPLA. The weight is a crazy 250g for the body. It is a bit flexible to the touch, but the sound is excellent overall. The G string is a bit muted, so I will try to fix that with plate graduation in the top plate to stiffen up the areas that feel soft to me. 

So there’s a bit of a brain dump. It’s very hard to know what to do to improve sound until you start making changes in the right direction. 

Oh last thought, if you would like to share your CAD model I’d be happy to review and offer suggestions!

-David 

To view this discussion on the web visit https://groups.google.com/d/msgid/OF-Fiddles/9aa777b2-765c-44aa-bfaf-c36feb3f2a1en%40googlegroups.com.

[Caleb Coburn]

Appreciate the thought re:credit for the infill idea. I'm not on IG / twitter and don't keep a personal website. I just have facebook/reddit and I guess linkedin/google scholar haha. Don't really use them anymore though, so not sure I have an internet presence that would benefit from the attribution.  You are welcome to just mention my name I guess.

I think sound on mine is pretty decent, but I can't really tell. I guess I'll have to find a feedback loop there. Are you in SF? I think I remember seeing that somewhere. Maybe not. I'm in the bay area so maybe at some point we could see the fiddle in person. (btw, is there a distinction between fiddle and violin?)

Thanks for the feedback on tuning for sound. Kind of my take too on the lack of consensus. I guess for traditional violin construction there isn't an easy way to have a one size fits all since materials are different for each piece.

250 g body is crazy light! I'm also running 0.6 walls on either side of infill. I don't remember the weight on this body, but all assembled with the screws, rod, 4 fine tuners, it comes to 449g ish, so also in the acceptable weight for a high quality fiddle range. I guess not bad for a full 4/4 plastic "violin shaped object" haha. 

Caleb

[Dobieag]

What about using a feature like in Prusa Slicer where you can change the infill settings in a local area. That way, you could beef up a specific location with more infill to make it stiffer. I know that would add a whole other dimension to testing but it popped into my mind so I thought I'd throw it out htere.

Mark

[David Perry]

Mark, I like the sound of that! I know nothing about Prusa Slicer so I think I'll have to look into it. I know that Cura also has some interesting infill options that allow a more structured infill in all three dimensions. I think they call it 3D infill, maybe? 

Caleb, I'm in Portland OR, but I travel to the bay area once a year or so, and will do so again once the pandemic passes. We will have to connect sometime!

Wow, 450 grams I think is pretty incredible! Nice work. What brand carbon filled PC are you using, and what are your print settings? I'm doing a deep dive into material options. 

For fiddle vs. violin -- a violin is a bowed instrument with specific dimensions. It is expected to have a proper string length, standard string spacing, standard fingerboard profile -- it's very much a standardized profile. A fiddle is less so, and could take a range of configurations. For example, a Hardanger fiddle has sympathetic understrings, short string lengths, and a flatter bridge. It is not a violin, but it does qualify as a fiddle. 

-David

[Caleb Coburn]

Mark - I've been thinking about different infill in different sections. I think its a great idea. Currently I'm more interested in Cura's ability (at least I think it has this feature) to treat an explicit model as infill. Agree with you that it would add dimension and IDK how to take advantage of it. 

David - for CF-PC I have been using 3D-best off of amazon (I think I liked it in a previous post, not sure). Thanks for the explanation on fiddle vs violin. I've been trying a new material recently - after you recommended trying PLA as a prototyping material I hopped online to order more, and got sidetracked on pursuing HIPS (I think we had mentioned it previously since the S3D material table makes it look like an ideal choice). As you indicated, it has been finicky and has taken me a while to figure out bed adhesion. It also struggles with overhangs, especially for thick walls and at high enough temps for decent layer bonding. I don't think I've got it dialed in perfectly, but I now have my 3rd full violin build done, all in HIPS. I sanded the lighter colored bits, which gives the color contrast:

Banana for scale. And scale for scale.

I made several changes, including a different bridge, thinner top plate, simpler tailpiece, and re-designing bits of the neck. It is very noticeably lighter and louder. The body came in at a whopping 220g. I think the sound is better, but hard to say. Still haven't gotten into the spectrum analysis or learned how to play one. But overall I am stoked at how this one has turned out. I may print the body in PLA to compare. I have a chinrest / shoulder rest printing out for it now. First time design so we'll see how it prints / works. I love the look of raw cuboid infill on the bridge. hopefully I can get my violinist friend to test drive it soon. Definitely some issues with the print, but nothing that kills functionality as far as I can tell. 

[David Perry]

Wow, I can't believe the weight on this instrument! It's crazy! 

GREAT thinking with the infill on the bridge. I'm totally going to try that now, too. 

Yeah I'm impressed you got HIPS to work this well. While you're at it, you might consider trying ASA, or maybe there's a carbon fiber filled ASA out there? I'd like to try some ASA for another project but I'm not confident I'll be able to print it very well on my machine. 

How is that tuning system working? I've made some progress on the one I'm designing that is similar. I'll post some screenshots of it in another thread. 

[Caleb Coburn]

Thanks! Yeah, HIPS is pretty cool filament, and while there aren't filled versions of it readily available, it made a pretty cool fiddle. I have a whole collection of bridges with different infill designs. My wife liked the one with only bottom and side perimeters (0 infill and top perims, flat topped. My goal was slicer estimated weight of ~2.5 g or less with enough stiffness to stand up to the strings. Honestly keeping the bridge weight reasonable with enough stiffness seems one of the hardest areas for 3d printing. However, HIPS is not as stiff as I hoped, definitely doesn't deserve the maxed bar that S3D material properties table gave it. its probably half PLA stiffness, not 5/4, and it fails easily. If it makes use of its high impact property even once it's stiffness is compromised. So a fun foray and a worthy instrument but not the end goal material.

I haven't been very interested in ASA, it just seems like expensive ABS to me. CF-ABS is readily available though and is a pretty great potential violin material for stiffness and weight. However, I have a new top material candidate that I'm excited about: CF-PP. This stuff: https://www.3dxtech.com/product/carbonx-pp-cf-polypropylene/

Per the TDS, it has a stiffness comparable to CF-PLA, but its ~13% less dense, heat deflection temp is 45 degrees above CF-PLA, and (maybe unintuitive why this is desirable) is much less still along the layers since PP is pretty flexible. Also, its not hygroscopic and doesn't need to be dried for printing or kept humidity regulated to have consistent mechanical properties. 

per 3Dxtech's website:

On the last point, my thought is that we really want to replicate spruce. But spruce is 2-3x less dense and 2-3x higher modulus along the grain. To match, as we have already discussed, we can make the plates hollow and thicker, with the goal of same mass-per-area as the wood and same overall plate stiffness (locally, if you want to grade it). So far so good, but spruce is extremely anisotropic. its modulus perpendicular to the grain is 10-20x lower than along it (https://www.fpl.fs.fed.us/documnts/fplgtr/fplgtr113/ch04.pdf). Luckily for this application, while normal 3d printed materials are pretty isotropic if later adhesion is good, carbon filament is like wood - its only stiff along the extrusions. Along the layers lines they are only as stiff as the base plastic. However, since fiber filled plastics for printing only have very short fibers, they are much less stiff than traditional composite materials, with stiffness anisotropy closer to 2-3 for PLA. But because PP is a flexible matrix, is has closer to 10x stiffer along extrusions, much closer to spruce. So I think if the infill pattern, plate thickness, printing orientation, etc. are chosen well, the CF-PP could be a prime material for a 3D printed violin that behaves similarly to a wood one, and is robust to chemicals, temperature, vibrations, etc. I've stated all of this very confidently, but we'll see if it pans out. I've not found any low priced CF-PP, so I might take a foray into at home filament extrusion since they sell pellets of it for cheaper. 

3Dxtech doesn't have anisotropy data for modulus, but you can get a good idea for it by comparing these:

https://cdn.shopify.com/s/files/1/2327/6017/files/TDS-Essentium-HTN-CF25_v1.0-Minus-3D.pdf?v=1609355683

https://cdn.shopify.com/s/files/1/2327/6017/files/TDS-Essentium-HTN_v1.0-Minus-3D.pdf?v=1612306657

after a bunch of math I think ~10% infill, 0.4 mm thick single perimeters, and total plate thickness of ~2.9 mm would be close to the properties of a 2.5 mm thick spruce board, bonus points if the infill pattern gives anisotropic stiffness. For all this though, I start to wonder what the CF rod in the designs is really for. Does it take all the force of the strings? If so does the along-grain stiffness need to match spruce? If the goal is a lightweight and flexible top-plate that is still strong enough to not fail under the tension is made easier by the rod taking most of the tension, can it be more flexible overall? Is the top plate being strained by the strings important to the sound? If I can achieve a top-plate that is as stiff as a spruce one, could/should I leave out the rod?

The tuning system works well. I haven't had any issues this time, and there is plenty of headroom for the strings to stretch and keep it in tune on this version without needing to loosen and retie the strings. But it requires a tool to tune, so is much less convenient for someone who has mastered tuning by pegs unless a full set of fine tuners is included.

Caleb 

[David Perry]

Very interesting. I had dismissed PP with CF because PP is such a flexible material, but you make some very interesting points here and it would be an awesome test. I may order some from 3DXTech, $75 for a kg is within reason for an interesting test... 

Regarding the truss rod -- the truss rod that runs from end pin and into the base of the neck is there to secure and reinforce the neck/body joint. It does not take any load away from the top plate, as the primary load there is the downward force from the strings. At least that is my intention and my understanding...

But one could use lighter-weight strings and have a more flexible top plate. That might be a good solution for the Modular Fiddle (less so a full bodied violin), as we have already accepted that it will be a quieter instrument. 

Glad to hear the tuning system is working! 

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/OF-Fiddles/70691399-2a5d-447a-9e11-f7d4428b8536n%40googlegroups.com.

[Caleb Coburn]

Would be very  interested in the results of a cf-pp fiddle if you make one!

"the truss rod that runs from end pin and into the base of the neck is there to secure and reinforce the neck/body joint. It does not take any load away from the top plate, as the primary load there is the downward force from the strings. At least that is my intention and my understanding"

I guess the top plate has two components to the stress - downward force ~20 lbs at the bridge and force along the strings ~50 lbs between the neck joint and the endpin. I've always assumed the 50lbs load would be a bigger issue, since the sound post and bass bar reinforce the top and 50 sounds bigger than 20 😅. It seems like the truss rod inside the cavity would share that load. Also, the luthier I took one of my pieces to mentioned that neck vibration can be important (not sure about that) and the rod might hinder those resonances. Maybe I should change my design to stop the truss at the joint instead of running it almost all the way up the neck (more like modular fiddle style rather than hovalin style)?. 

[Caleb Coburn]

I gave in and ordered a roll of the cf+pp. Interesting stuff, so far I'm very pleased with it! My full 4/4 violin body came out to 163 grams, which was very cool. the same model in PLA was 266g. 

I don't have a trained ear, but to my ear the new one sounds better than my previous best, see sound samples.

-Caleb

With the neck, etc pulled off an old one installed:

Body only (before hot air gun to kill the stringing): 

on the build-plate:

 

Sound files attached.

[Caleb Coburn]

One more set of pics: with my newer and much nicer looking chin/shoulder rest combo (in normal PETG):

[David Perry]

Awesome! Thanks for the update. 

Gosh it's hard to assess recordings. If I had to conclude anything...I think the PP sounds a bit quieter than the HIPS, but there's more timbre/resonance, and the open D has less of that too-loud resonant hollow ring. 

The weight you're getting on these is just crazy. I built a super minimal electric a week ago and got it to 330g. It BLOWS MY MIND that you're getting that weight with a full acoustic with fine tuners. 

Nice chin and shoulder rest! Do they integrate into the tailpiece or can you bolt them on to whatever? I'd love to try it out on my fiddles if you're up for sharing the files! I feel pretty lazy not 3D printing a chin rest...

-David

To view this discussion on the web visit https://groups.google.com/d/msgid/OF-Fiddles/467c2523-8cde-4136-a674-b01a7aad0006n%40googlegroups.com.