High temp Modular Fiddle

[David Perry]

Hey Everyone,

I have been working for a year, off and on, on building up a high-temp Modular Fiddle. I got something that was OK, but not great, last summer. All parts were ProtoPasta HTPLA-CF. They warped and shrank minimally in the oven, and in the end I built up a playable but wonky instrument that can be left in a hot car, left in the direct sun -- you name it! You can read all about this build, and see a video of a Modular Fiddle failing in hot temps, on this ProtoPasta blog.

Now I have promised a high-temp fiddle to a local fiddler who is also one of my fiddle heroes. He doesn't want a wonky fiddle (he said this much more politely) so I have been trying to find a solution to my wonky high-temp fiddle. I'll tell you what I've tried, and I want to hear any ideas you have or experiments you've tried. 

I can print and anneal all parts in HTPLA-CF with the exception of the body. The body just warps too much in the oven. So I have a solution for all parts but the body. Here are three materials I've tried.

1. Annealed HTPLA-CF: this gives great stiffness and strength, but the body warps and the variation in dimensions makes assembly more difficult. I have tried two annealing methods: water bath (several tries, more deformation than oven) and oven (simple and fast). The body just won't stay in shape! Makes sense -- anything with thin walls like the body will want to move around once it softens. 
2. PET-CF: I specifically have tried Colorfabb XT-CF20. It's nice to print with, bridges well enough, and I got a usable body on my first print. It's much softer than CFPLA, though, and even is softer than some of my standard PLA. All this means that the body does not sound right. I could thicken up the walls of the instrument to improve it, but it still will not be as good as a PLA based instrument. Then I left it in the sun, and it warped!! It did not fail outright, which was interesting, but it did warp such that it went out of tune and the string heights got a little funky. 
3. Nylon based with CF. Lehvoss sent me some of their Luvocom 3F carbon fiber filled PA. I had a hell of a time printing with it, though. It seemed to have thin sections in it, and it over or underextruded constantly. In the end I was not able to print a functional body. My tests also showed that, despite a higher modulus than PLA, it just wasn't that stiff. 

Have any of you tried high performance or high temp materials?

I will probably try a body revision and then re-print with the XT-CF20. It held up for a while before warping, so at least it is high temp resistant, if not totally fail-proof. I just really don't want to sacrifice sound, and the CFPLA sounds SO GOOD. If you haven't printed a CFPLA body, do it now.

Perhaps I'll try printing a body with ProtoPasta's fiber filled PLA, and maybe that will anneal with less warp. Unlikely...

I could also try a carbon fiber ABS. 3DXTech has that and more.

Really, in the end, I think the solution is to anneal a HTPA-CF body, squish it back into shape quickly before it cools after the anneal, and then custom-fit a bridge to the instrument to accommodate the wonk. 

I appreciate any thoughts y'all have!

-David

[Matt Hova]

My friend had an idea that I have yet to try out, where you could assemble the entire body and neck together, fill up a pan with sand and then bake the violin in the sand. That would let the body keep its form and keep everything together. Worth a shot 🤷‍♂️

--
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/e7f50552-f397-4123-a413-d307341ba434o%40googlegroups.com.

[David Perry]

Nice! I like that idea, and I definitely think that a sand bath could help the warp. In talking with Alex from ProtoPasta, he has often suggested cutting the body in half (so that there is a separate top and bottom). Then print a mold/fixture that fits inside each half, and use that during annealing to hold the parts in place. 

I despise adding labor to the assembly, but perhaps the merits justify it.

I'm like 90% sure that nothing will be as stiff and resonant as ProtoPasta's CFPLA... so let's elaborate on ways to problem solve the annealing process. Alex has suggested printing slowly and with minimal fan cooling. Allowing the plastic to cool off more slowly reduces internal stress. 

Here are my ideas for how to stabilize the parts:

1. Hold parts in place while annealing
   1. Cut body as described and use molds.
   2. Cut a window in the back of the body, use a small mold that reaches through and fixtures the bridge area on the top plate. 
   3. Fill body with sand, and fully immerse in a tub of sand, anneal the whole thing.
2. Move plastic back after annealing (before cooling below Tg)
   1. Current approach is to use a butter knife to pry at f-holes and reform by eye.
   2. Add more holes to body to allow more precise reforming, use current technique.
   3. Make a mold for the exterior of the top plate, press through holes in bottom plate to push top plate against mold
3. Spot-anneal before full anneal: use a heat gun or other heat source to anneal specific areas of the body. In theory if small areas can be annealed independently without much warp, then those already annealed areas will not warp during the full anneal. I have sort of tried this and failed...it seems difficult to know if a small area has actually annealed, or perhaps it just doesn't work.
4. ? Gotta be more ways...

Another thought is to forego annealing altogether. Use plain old CFPLA or PLA with a heat trigger tailpiece. See below. The drawback, of course, is that the instrument is not truly heat proof, and the body may warp due to internal stress even if there is no string tension.

[OpenFab Violin Builders]

From Steven:

I keep thinking an oil bath could be differently painful as compared with a sand bath, not sure how to evaluate risk of accidentally reacting with the CFPLA, but there’s probably some non-reactive oil that could do the thing.  I also wonder about a water bath in a pressure vessel, your annealing temps aren’t that far off from boiling temps for water, there’s probably some nice math for what pressure you’d have to hold to keep the water in liquid phase, see if that sounds scary.

Couple questions: are we assuming deformation is from gravity?  Or from internal stresses selectively relaxing?  I assume you’ve tried annealing very very slowly (I.e. very slow temp ramps)?

More hairball: seal up your F holes and anneal with the body under internal positive pressure?  Feed a perfectly sized air bladder in to the F holes to reinforce from the inside.

Further questions: do you have pics you can share of the annealing deformation?  The blog post you referenced showed the thermal failure of unnanealed parts, I’m curious what it looks like when things go bad during annealing…

[David Perry]

I like what you're thinking -- to somehow hold pressure in the body of the instrument so that deformation can't happen. Unfortunately, the deformation is kind of all over the place -- it will cavitate in one section and blow out in another. 

To quickly respond to your questions, Steven:

• Yes I have tried heating very slowly. Absurdly slowly, actually! I was using a hot water bath and I ended up with more deformation than a fast heat in the water bath. In the end I abandoned the water bath because uneven pressure in the bath seemed to perhaps worsen the deformation. The part is just so soft and vulnerable while heating! 
• Oil is interesting, as is the pressure vessel. I suspect that any medium that can easily flow will simply be pushed around as the part deforms due to internal stresses. 

Photos, first, a nice body fresh of the printer. Quality isn't great, but notice the f holes and how they are nice and symmetric. 

Now this is also an R7 body but it has been annealed. This one is playable, but wonky. Note the difference between the F holes. The soundpost keeps the treble side of the top plate mostly in the right place, but then it dips substantially on the bass side. This is the most detrimental warping, but other areas exist as well.

The instrument, above, was annealed in an oven while adhered to the printer's build plate. 

From my experience, I believe that most of the deformation during annealing is due to internal stresses relaxing after reaching Tg and before annealing. Alex reinforced this as I was working on the process. Here are some quotes from Alex @ProtoPasta that help give background info:

"Regarding firmness, if you heat treat at 225F, for example, you can probe the part with your fingers at 225F as it bakes. That 225F part is going to be rigid when you put it in the oven, then get soft when warm, and then more firm (in between room temp and initial softness) in that same 225F environment without removal before any cool down or removal from the oven. This is how you can feel the change to crystalline from amorphous."

And more, first I propose to identify three phases of annealing:

Phase 1: Part softens as it warms.

Phase 2: Crystallization is achieved and part firms up while at higher temp

Phase 3: Part cools down and returns to rigid, with new crystalline state and higher temp resistance.

Alex replies:

"I agree with your 3 phases, but maybe there is also a Phase 0: Printing? The printing inputs of material, geometry, and parameters/conditions will also influence the result so holding those constant will give more confidence in understanding the heat treating influence though may not get you to the desired post-heat treating result. It's likely printing/cooling rate along with print temperature and geometry/orientation also has a strong influence on varying internal stresses throughout a part which would also affect deformation in heat treating.

If heating and cooling is even during heat treating, I believe warp is likely due to the internal stresses imparted during printing or a weight component (force of the part on itself) during the phases. I agree the part is most at risk to deform under its own weight during phase 1, but phase 2 and 3 could also put differential forces on the part based on geometry, support, or inconsistent thermals. If the part is firmly supported (fixture?), the time it is soft shouldn't matter, but to minimize that time, I'd think you want to the internal part temperature to be at peak crystallization, above 100 C as quickly as possible.

Does this make sense?

For example, I have worked with a series of 10x10x100 mm rectangular bars in Carbon Fiber HTPLA and a contant heat treating technique, only varying printing inputs. The result varied greatly from flat to 5 mm warp over 100 mm. I guess my point is that the printing process may matter just as much or more than the heat treating when seeking to minimize distortion. In this case, the slowest printing with least cooling fan was least distorted though the surface quality was also most poor, so there's a trade-off.

Sorry if I covered this topic previously, but I think it's important to consider a near-perfect heat treating technique may not be all that is required to get near-perfect form. If you are changing inputs prior to heat treating, that could affect the result when heat treating. Both pre-heat treating and heat treating conditions may need optimization in order to work together for a final best result.

Maybe water just doesn't get hot enough to be at that ideal crystallization temperature? Also, I wonder if it puts undesired forces on the part. I'm not sure, but I have abandoned the boiling technique as I haven't been pleased with the result nor the additional safety implications."

[Allen Irvan]

David, I found this water soluble product that is made for casting composite materials.  You pour it into the mold, let it set, do your annealing and such, and then wash it way...  I think you could pour it inside the fiddle and set it in a container and pour around the outside as well.  Or maybe do sand on the outside?

https://acmtucson.com/acm_products/aquapour/ 

Allen 

--
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/0fea562b-92b5-4797-a8bf-bfcfccc3a6f0n%40googlegroups.com.

--

Allen Irvan

[David Perry]

Huh, that is cool stuff! I'll look into it. Thanks, Allen!

[OpenFab Violin Builders]

If you'd like to receive fewer emails, you can adjust your settings here.

Here are more thoughts from Steven:

Yeah based on the pictures for sure this is localized internal stresses, I was targeting things that would fight against gravity (Oil and water to let buoyancy counter gravity).  Looking at the pictures I’m skeptical of the sand bath too, unless it’s like under compression somehow so the sand is jam-packed or whatever.

How aggressively have you chased down eliminating voids in the top btw?  I was surprised when watching mine print how many voids it was leaving with the approximately recommended print settings, and then also surprised at how good it sounded.  You’ve probably chased those to ground, but if you haven’t they could be exacerbating local stresses via a number mechanisms.  

Thinking through my spitball about holding the interior at pressure, that combined with a external tooled surface to be pressurized again that could be cool, almost like a blow mold process.  

Serious about annealing in a pressure cooker to allow higher temp with water, I think the body would fit in a large kitchen grade pressure cooker.  I guess that makes temp control extra hard (Inside a very sealed vessel).

Based on protopasta guy’s comments it seems like the target is to get to temp quickly, which makes sense, so it doesn’t have chance to deform before stresses get relaxed?  Makes me wonder about forcing convection through the body of the thing (to heat walls from both sides).  

This is a fun / hard problem, but I am concerned that if the resin manufacturer’s suggestions aren’t working then it’s a real hard problem.

[David Perry]

Good thoughts all around. I have not eliminated all voids, there are especially voids all along the bass bar and by the chin rest and near the neck joint. I will take a closer look at my annealed bodies and see if there seems to be greater deformation in these areas. 

Yes, get to temp quickly, but I don't know if it is possible to relax those internal stresses without deforming the part, unless of course you are holding the part rigid by other means.

Yes, it is a fun and hard problem. Unfortunately I think that it is a very hard problem. It makes me think hard about breaking the body into two parts. First, I'll try printing it nice and slow, annealing in the oven, and prying at it with a butter knife again. Maybe with a little practice I'll get it -- but it is expensive practice!!

-David