Brass Instruments Materials

[Enthalpy]

Hello everyone !

Could liquid crystal polymer make mouthpieces for brass wind instruments?

Mouthpieces are almost always made of metal presently. Wood is too sensitive to saliva, usual polymers sound badly supposedly because they aren't stiff enough. But metal is cold when playing outside or in a church.

With high mechanical damping and 10GPa Young's modulus, more if loaded with graphite choppers for instance, LCP should perform better than usual polymers. If the sound is decent but differs from metal, it may fit some scores.

At least Vectra A950 is authorized for food contact by the FDA. LCP can be processed by injection to make cheaper mouthpieces.

Marc Schaefer, aka Enthalpy

[Phi for All]

18 hours ago, Enthalpy said:

usual polymers sound badly supposedly because they aren't stiff enough.

The stiffness of the rim relates more to the seal around the embouchure and the player's comfort, doesn't it? Using polymers instead of steel or brass seems less expensive, and might allow you to play with different sizes of cup to get a pleasing sound. You can also change the rim's flatness/roundness more easily, I would think. The throats and backbores are fairly standard, but if polymers make it easier to change the rims and cups I'd say it's worth looking into. 

[hypervalent_iodine]

Plastic mouthpieces are very common, but not preferred for normal playing. I have only ever used brass mouthpieces, but my guess is that the plastic would change the tonal qualities of the instrument, which would be why they aren’t preferred. It wouldn’t be to do with stiffness so much as how much the plastic diffuses the sound. You can get metal sleeves to go around the mouthpiece, called brass tone modifiers, and this is meant to improve the sound quality quite a bit, though I have not tried them. Personally, I don’t see that plastic would be too much different to metal once you get going. They warm up fairly quickly. 

[Enthalpy]

Thanks for your interest!

I have no personal opinion. I only play the contrabass tuba among the brass, and very badly.

I'll search for the video where Allison Balsom tries a trumpet of injected thermoplastic. She reverts very quickly to a metal mouthpiece, and then she finds something not too unpleasant to tell about the instrument - but you just hear the plastic. It's one trial that needs no lengthy wondering and interpretation.

The effect is very similar to what I observed with a clarinet of thin injected thermoplastic. The difference is huge, and not in the right direction.

Maybe the explanations I proposed
Oval resonances
Bending resonances
bring some enlightenment, as the figures could fit. Or maybe not. But It's a reasonable assumption that the polymer's lower E-modulus is the reason, so I suggest to try a stiffer polymer, optionally loaded with graphite choppers.

 

==========

Here's the video with the p-trumpet:
NLAHSgZaMU0 music 0:51
it sounds absolutely dull, so the plastic mouthpiece (at 3:10) doesn't make much hearing difference to me. And easy blowing, yes - just like the thin injected thermoplastic clarinet I tried. Both aspects reveal huge losses at the walls, which should logically result from the lack of stiffness.

==========

A more direct opinion by an other trumpet player, supposedly not paid by the manufacturer, there
s7Uv5Ld0sJU
absolutely dull sound (he doesn't tell it like that). He finds a difference between a plastic and metal mouthpiece.

Both instruments intonate terribly, especially on low notes. I suppose it's a problem of shape more than material.

[hypervalent_iodine]

1 hour ago, Enthalpy said:

Thanks for your interest!

I have no personal opinion. I only play the contrabass tuba among the brass, and very badly.

I'll search for the video where Allison Balsom tries a trumpet of injected thermoplastic. She reverts very quickly to a metal mouthpiece, and then she finds something not too unpleasant to tell about the instrument - but you just hear the plastic. It's one trial that needs no lengthy wondering and interpretation.

The effect is very similar to what I observed with a clarinet of thin injected thermoplastic. The difference is huge, and not in the right direction.

Maybe the explanations I proposed
Oval resonances
Bending resonances
bring some enlightenment, as the figures could fit. Or maybe not. But It's a reasonable assumption that the polymer's lower E-modulus is the reason, so I suggest to try a stiffer polymer, optionally loaded with graphite choppers.

 

==========

Here's the video with the p-trumpet:
NLAHSgZaMU0 music 0:51
it sounds absolutely dull, so the plastic mouthpiece (at 3:10) doesn't make much hearing difference to me. And easy blowing, yes - just like the thin injected thermoplastic clarinet I tried. Both aspects reveal huge losses at the walls, which should logically result from the lack of stiffness.

==========

A more direct opinion by an other trumpet player, supposedly not paid by the manufacturer, there
s7Uv5Ld0sJU
absolutely dull sound (he doesn't tell it like that). He finds a difference between a plastic and metal mouthpiece.

Both instruments intonate terribly, especially on low notes. I suppose it's a problem of shape more than material.

You can definitely hear the difference if you listen to them side by side. Plastic doesn’t give the same smooth, clear tone IMO. It’s not entirely surprising. Some people also say that you can tell a tonal difference between different finishes on trumpets as well, though I don’t know how much of that is actually true.

[Enthalpy]

I heard it about saxophones too. Silver shall give a more brilliant tone than varnish. But how much do the eyes mislead the ears? And while a saxophone body can vibrate, how to explain the alleged effect at a bassoon bocal?

For flutes, I had the golden opportunity to compare materials at the head
111316-woodwind-materials
I could convince myself there was a repeatable difference, but more in the response than in the sound, and it would be imperceptible through computer loudspeakers. The flute body should matter more than the head. But if we hear the plastic at a trumpet record, in real life it must be striking, just like it was at the clarinet.

I just love the way professional musicians tell "Absolutely amaaazing" about the piece of junk in their hands to earn their two bucks.

Oboes exist of PMMA and, even over PC loudspeakers, they sound just like plastic.
111316-woodwind-materials
NrJy8tNlBuQ 1:55, same musician and reed, different materials
8AJnQk3ECYE 0:38
So would polyketone or LCP sound better?

Acoustics had difficulties to explain the effect up to now, which let some people deny it. That would be a mistake to my opinion, as the acoustics of musical instruments is difficult, and we know so little about sound perception. I hope to have found explanation candidates
111316-woodwind-materials
111316-woodwind-materials
and plan to search credible explanations for brass too.

 

[Enthalpy]

And a (not favourable) opinion about a plastic trombone, with sound
SITNi-5uVR0
it sounds terrible even with the metal mouthpiece. The musician gives more reasons against the instrument.

[Enthalpy]

Being both stiff and lossy, could Liquid Crystal Polymers (LCP) make varnishes for brass instruments?

Silver, or varnish over brass, resonate less strongly, and are said to leave a more brilliant sound in the air column. The varnish must be stiff to dampen the metal, so metal deformations induce forces in it. I expect most present varnishes to be too soft, lossy polyketones too. The varnish has also to be lossy. What matters is approximately the product of both: E", the loss modulus at relevant frequency.

From the LCP doc already cited for woodwind materials, on page 22
hipolymers.com.ar
the A950 resin formulation offers E=10GPa and 6% internal loss, wow.

How to make a varnish from the resin isn't quite clear. The few solvents aren't so friendly to skin nor metal. But at least, brass instruments accept a process at higher temperature and solvents. Or could the resin be just molten?

A stretched varnish film may improve. At fibres for ropes, orienting the molecules raises E to >200GPa, and losses use to decrease but less quickly. Something like a motorised brush, similar to the polishing wheels used by luthiers, might act on a drying or solidifying film. Crossed layers may improve.

In case it works for trumpets, don't forget flutes, including altos, and saxophones.

Marc Schaefer, aka Enthalpy

Edited August 8, 2019 by Enthalpy

[Enthalpy]

Despite being a labrophone (a brass instrument), the cornetto is made of wood, its mouthpiece too
wikipedia

To resemble historical instruments, the wood is usually European and imperfectly water-resistent. Condensation would swell the wood at the bore and split the body. Common practice at ancient instruments is to impregnate the bore with linseed oil or an other siccative oil that reacts with air to make a solid water-repellent layer.

Alas, siccative oils get toxic as they react with air to harden. It may have sickened the romantic painters, together with absinth. I worry for cornetto players who might impregnate the mouthpiece itself with linseed oil. An impregnated cloth is typically pulled from the narrow end of the bore, here the mouthpiece.

This is especially acute for the mute cornetto, whose mouthpiece is integral to the body made of light-colour wood looking like beech, box, pear or cherry. Other cornettos have a mouthpiece of darker wood, possibly some Dalbergia that resists water.

Alternative treatments:

• Use a non-siccative oil. Repeat the treatment often and forever then. Palm oil is solid.
• Squalane maybe. I mistrust (?) paraffin due to residual polyaromatics.
• Use a non-toxic varnish if this exists. Wax?
• Inject a polymer as a barrier layer, as is done for oboes. Needs tooling.

Some materials for the mouthpiece, or for some mouthpiece insert concealed in the mute cornetto, or for the whole body:

• Ebonite. Turned like wood, well-known from single reeds. Dark.
• Exotic wood. Years drying, dark.
• Titanium. Stainless steel (nickel-free). Bronze. Or metal plated with Au, Pd, Rh...
• POM, PETP...? Light colours too.
• LCP is stiffer than most wood, some is allowed for food contact. Light colours too.
  scienceforums

Marc Schaefer, aka Enthalpy

[Enthalpy]

Air pressure oscillations create strong forces in the tube of a wind instrument at the bends, I hope to detail that some day. I haven't seen much consideration about it in academic research, but luthiers try to stiffen the instruments against these forces, more so at brass instruments.

The sketch illustrate the stiffeners at a U-turn. More locations use them.

The most usual design, first left, provides but more than a stable spacing between the tube ends. The narrow connection to the tube doesn't help the thin wall.

The known design, second left, is meant to provide stiffness against the varied orientation of the forces at the bend. It injects force over a longer line at the tube, but still perpendicularly to the thin tube that offers little stiffness.

I propose instead to inject the force tangentially to the tube's wall so the flexibility of the thin wall doesn't spoil the stiffness. I give three examples at the right of the sketch. I believe that the parts depicted in grey can be cut and bent from a flat sheet even in the rightmost example.

This applies to woodwind too.

Marc Schaefer, aka Enthalpy

[Enthalpy]

At bends, a wave in an air column creates net forces on the wall even if the section is uniform. This applies to woodwinds too.

Imagine the small angle element is physically closed at its entrance and exit. The acoustic pressure inside won't let the closed volume element accelerate, so the sum of the forces on the entrance and exit sections compensate the net force on the wall, hence the two initial formulas.

They tell already that where and when the acoustic pressure is positive, a bend is pushed outwards the curve, and where and when negative, inwards. For instance the ends of a bend are pushed away from an other by an acoustic pressure positive in the whole bend. Or a full loop of tube gets a net force if the acoustic pressure is positive at one side and negative at the opposite side of the loop. Drawn examples could come, in the future hence maybe.

With similar sections, the forces resemble the ones I described at woodwinds' tone holes. The many bends, the narrow long tubes give brass instruments many reasons to vibrate.

On the picture, I give a few net forces integrated over a bend, for the cos and sine components of a sinusoidal pressure distribution. Extreme cases give the expected value. The example for the vertical (on the drawing) force takes a pressure negative at the lower end of the bend and positive at the upper end, that is, the air speed is maximum around zero angle. The air's inertia is consistent with a vertical force.

The forces could have been computed equivalently from air speed, as is done for a rocket's thrust. The tube's vibrations created by the acoustic pressure act also on the air column's inertia. As I described about the oval and bending deformations of woodwinds' walls, resonances at the tube amplify the deformations and change their phase so the effect is to weaken the resonances of the air column. I should explain that some day, hopefully.

I'm pretty sure luthiers have known that for over a century, because they add stiffeners to brass instruments at corresponding locations and orientations. But when I read many research papers three decades ago, I didn't stumble upon the equivalent of my explanation.

Marc Schaefer, aka Enthalpy

[Enthalpy]

This illustrates how the distribution of air pressure pushes on the bends of a tube of constant section. Less pedantic, more visual than the former message.

The standing wave has, at some instant, the locations of its maximum pressure (antinode) indicated in red, of its pressure nodes and maximum speed in orange, and of its minimum pressure (antinode) in blue.

In the left case, the pressure distribution component P_c*cos(k*R*phi) matters. The higher pressure in the right-hand bend pushes it outwards because the outer curve has a bigger area than the inner curve (the previous message's computation is less direct but easier). The lower pressure in the left-hand bends pushes them inwards. The forces add to drive the tube vibration horizontally. Its response depends on its proper modes.

In the right case, P_s*sin(k*R*phi) matters. The overpressure at the bend's top pushes it up, the underpressure at the bottom too. The shape may well create tube resonances with a vertical movement component there, and these resonances are excited.

Shorter air wavelengths may fit mechanical resonance frequencies better. Several half-waves in a bend cancel out their effect partly, not fully. Local forces can also create more local deformation, for instance to open a bend: luthiers put stiffeners there. Local forces can even add up if the wave numbers match between the air and a tube's bending mode.

Marc Schaefer, aka Enthalpy

Edited July 3, 2021 by Enthalpy

[Enthalpy]

Nearly all brass instruments have bodies of thin tube, where the dispersion relation of bending waves simplifies.

mu*w² = EI*k⁴ where mu is the mass per length unit, w=2pi*F and k is the wave number: k*lambda=2pi.

For a thin tube, mu ~ 2pi*R*e*rho and I ~ pi*R³*e where R is the radius, e the thickness and rho the density, so
EI/mu ~ R²*0.5*E/rho

w ~ sqrt(0.5*E/rho)*R*k² with:

Sqrt(0.5*E/rho) ~
============================================
 860 Polypropylene
 940 Ebonite
1000 ABS
2300 Vectra LCP, isotropic
2020 Sterling Ag
2596 CuZn30
2620 CuZn20
2753 CuNi18Zn27
2640 Ni (Co?), electroformed
3400 Ni (Co?), laminated
6200 Graphite filament winding
============================================
Sqrt(m/s)

Thickness affects resonant frequencies little, so metals are faster than polymers. Graphite composites and anisotropic LCP are even faster. E/rho is the squared speed of a compressive wave, it's about the same for laminated alloys of Fe, Ni, Co, Ti, Al, Mg and for Ni electroformed slowly at +40°C. Cu alloys are a bit slower. Faster alloys are impractical.

==========

A wave can propagate as quickly in air as if it bends a tube. Several half-waves fitting in a bend cumulate their effect then.

w = kc with c ~ 348m/s in exhaled air, so velocities coincide if sqrt(0.5*E/rho)*R*k = c.

* For CuZn30 and 2R = 12mm, k = 22/m, F ~ 1.2kHz and lambda/2 ~ 0.14m.
* For CuZn30 and 2R = 20mm, k = 13/m, F ~ 0.7kHz and lambda/2 ~ 0.23m.

These frequencies are well audible, they fit fundamentals or harmonics, and some bends are longer than the corresponding half-wave, so I expect this coincidence happens. The effect must matter more at a chromatic horn whose curves contain many half-waves.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Figures about the effect of a bend at a trumpet. The trumpet is (the photo of) a YTR3335, the bend is the last before the flare.

From counting pixels, the outer diameter at mid-bend is already D_o=13.2mm. I take 0.4mm thickness (matters little) hence D_i=12.4mm. These are uniform in my model. CuZn30 (8530kg/m³ and E=115GPa) gives the section, in SI units, EI=37.9, µ=0.137kg/m and sqrt(EI/µ)=16.6.

The bend has mean D=74mm R=37mm, is 2*58mm long, and the straight sections to the pistons are 93mm long.

I model the resonance modes by zero movement at the pistons and no rotation at the far "end", due the continuity of the lower and upper pipes. The far end includes a half of the bend, as it contributes mass and compliance, so X=58mm+93mm=151mm.

The proper modes follow then nearly kX~(N+1/2)pi/2 with N>=1. Considering for instance N=1: k~15.6rad/m and this mode resonates around 643Hz, unfortunately the range where the air column of a trumpet has the smallest losses.

As computed here on Jun 28, 2021, scalable P_s=1Pa and kR=0.43 (now k in air with 348m/s) act on the bend as
F_s = 0.82*P_s*S = 99µN
The static deformation, without any resonance of two beams each 76+76mm long is 0.38nm.
The hands dampen little there, so the brass' resonance brings easily Q=30 and the deformation is 11nm with 90° phase shift.

The movement of the bend provides an increasing volume where the instant pressure is higher, that is a loss. I take the same 0.82*S, that is 99mm². A better computation would be similar to Jun 28, 2021 here, and thermodynamics might even prove both figures must be equal whatever the shape.

11nm deformation, 99mm² and 643Hz provide 4.4mm³/s lossy conductance for the initial 1Pa. That contributes to 227MOhm, very close to the 150MOhm input impedance reported in
"Interactions between wind instruments and their players" by Wolfe, J., Fletcher, N.H. and Smith, J.

• The pressure antinode may not reside at mid-bend. At random within 180°, it has 50% chances to reside within +-45°. Then, P_s*S and the force are sqrt(1/2) as big, the effect of the movement on the loss too, so the conductance is half as big. 454MOhm is still a big contribution to the 150MOhm.
• Notes don't match exactly mechanical resonances. But semitones are F*1.059 apart so some note is F*1.029 close to a resonance and excites Q=30 efficiently.
• This bend resonates also at estimated 1.8kHz, 3.5kHz... The trumpet's first bend after the leadpipe begins at lower frequencies. All tubes, especially the longer slides, resonate as well. This cumulates dozens of resonances within the hearing and playing ranges. Notes have many overtones too. Chances are that most notes have several components damped by movements at the bends.

The reported effect of the wall material too fits qualitatively my model:

• Thin plastic vibrates more, the sound is dull and the instrument consumes much breath.
• Thicker walls vibrate less, the sound is more brilliant and the instrument consumes less breath.
• Silver dampens the vibrations (it sounds "poc poc" instead of "ting ting" when tapping) so the walls absorb less power and the sound is brighter.

Marc Schaefer, aka Enthalpy

[Enthalpy]

How to reduce the losses at the bends?

Difficult and unsure: determine the mechanical resonant frequencies, check where the air pressure nodes are at these frequencies, locate the bends where the air column excites the mechanical resonances less. Though, the instrument has many mechanical resonances and it emits many notes containing many overtones.

Choose bend radii that couple little with the air column at the mechanical resonant frequencies. Could this explain the quest for very big radii at the slide trombone and others? The difference of path length doesn't explain it well, the air flow neither. Just an unsubstantiated hypothesis from me.

Stiffen the instrument. Some luthier try harder already. I proposed improvements here on Jun 13, 2021
scienceforums
Fewer mechanical resonances remain in the sensitive frequency range, and at higher frequencies, where the air column is lossy anyway, they hurt less. Better: the stiffer instrument moves less, that's equivalent to thicker walls without much added weight. Two bores in one wooden part excel for that, as at the boot of the bassoon and old baritone oboes, and I proposed to build contrabass clarinets the same way.

Dampen the mechanical resonances. My estimate from last message suggests a big effect with Q=30, but with Q=3 it would be small.

==========

Its material can dampen the resonance of the tube. I consider this is the effect of plain silver tube. Whether thin silver plated over 0.3mm brass acts at all, and by this effect, I doubt it.

Miyazawa's PCM alloy improved a flute headjoint more than sterling silver did in my trial, so it could be tried at bocals, at brass instruments... Rumours, nothing more, want it to comprise 65% Ag, Cu, Au and Pd. Sadly, I didn't tap the PCM headjoint to compare its poc poc with sterling silver.

Cheaper materials dampen vibrations too, at least for strong amplitudes. NiCo resists corrosion, it can be brazed and plated, maybe it dampens if electroformed too. CuMn is know too, but it needs plating against corrosion.

==========

If adding stiffeners to the tube, they could concentrate a bit of damping material where it acts best, especially for costly sterling silver and PCM. The preferred location is where the material deforms most for a given tube deformation, and the section must let the damping material contribute most to the tube deformation. Examples:

Stiffeners on long tube are better irregular so no uniform half-wave fits in between. For a slide trombone, the stiffeners should leave elasticity, so the moving part can dampen the standing one too. If stiffeners converge, say at a saxophone or bassoon bocal, the lossy material can be at the node instead, if the node provides most compliance.

==========

Stiffeners can use materials unsuited to tubes. Cast Zn maybe, useful at thick parts, while corrosion seems to exclude Mg alloys. Some elastomers excel too: perfluorosilicones, silicones, polyketones, many elastomers - with proper shape and coupling with the stiffer metal. Baroque trumpets wrap heavy rope around the tubes: for that purpose? Modern materials could save this weight.

Marc Schaefer, aka Enthalpy

[Enthalpy]

Some brass luthiers allege the flare's material matters especially. Here are some physics arguments.

The acoustic pressure acts axially on the wall where the section changes. This creates mechanical movements that produce a volume oscillation where the section changes. Mechanical resonances amplify the movements and shift its phase, so  the volume oscillation acts as a loss and dampens the resonance of the air column, as already explained for other wall effects.

The force at the flare can also produce a volume oscillation at a bend and reciprocally, as they are mechanically coupled. Not considered here.

When a section varies slowly as compared with a quarter wavelength, the pressure of a propagating wave varies as the reciprocal of the diameter. For a standing wave in a conical tube too. This law shall serve for the explanation. As the area varies with the diameter squared, a wave pushes more strongly at the flare than at a bend for instance, proportionally to the diameter. The effect of axial wall movements on the air column's losses too increases as the diameter. That's why flares and their material matter.

At the top of a trumpet's conventional range, the fundamental has a pressure antinode where the flare is about 3* as wide as the leadpipe, more at the overtones. The effect is roughly 9* as big as at bends, which I already estimated important. The air column's input impedance has only dropped by roughly 3 at such frequencies, so I agree that flares and their materials can matter a lot to brass instruments and woodwinds too. If a resonance with Q=30 fell in this range, some notes would be unplayable.

==========

One can stiffen the body to raise the resonances. The flare itself is very stiff axially. An unhelped bend, loaded by the flare, would resonate very slowly, but brass instruments bind the tube sections to an other. At the trumpet, the binding elements are skewed, almost parallel to the tubes. This suggests that luthiers have known my story here for a good century.

The binding elements could hold at the tubes through tangential sheets for stiffness, as I propose in the previous messages. They could also be trusses or sheets rather than skewed bars.

==========

One can also dampen the mechanical resonance. For the trumpet, the musician's lips do it at the leadpipe which is connected to the flare.

I suggest to use damping materials at the binding elements. Sterling Ag, PCM, NiCo, CuMn... as mentioned in other messages. This would hopefuly be cheaper and more efficient than flares of damping material.

Marc Schaefer, aka Enthalpy

[Enthalpy]

To reduce losses at woodwinds and brass instruments, I have suggested walls and stiffeners of damping materials. It's not that clear. Depending in the frequency mismatch between a mechanical resonance and the fundamental or overtone to be preserved, material losses can increase or decrease the power picked from the air column.

Here's a graph of the conductance G = Re(Y) of a series RLC circuit representing the absorption by a wall, around a mechanical resonance and for different Q. The black dips indicate semitone spacing, the mechanical resonance falls randomly among them, I displayed them half-way between aligned and anti-aligned.

With side holes, oval deformation of a metal tube needs an amplification factor >30 to begin to act, the bending of the body too. Against these losses, a damping material is good.

At bends, I estimated an amplification =30 and even less acts strongly. Damping can be useful, but not far from the resonance, where it worsens the loss.

At a flare, even an amplification <10 has a big effect, hence also far from the resonance. There, Q=50-200 can outperform Q=20, but not near the resonance, and Q=10 is still best. Nothing simple hence.

Stiffness looks desirable in all cases.

==========

According to Schilke's trials, beryllium bronze outperforms sterling silver at a trumpet flare. Silver dampens much, beryllium bronze very little. Beryllium bronze is stiffer (E~135GPa) than the usual luthier copper alloys. This might fit my thoughts here.

Beryllium bronze is expensive, uneasy to work, and bears the doubtful reputation of being unhealthy. Cu-Cr alloys are stiff too: 130GPa for Cu, already CuCr1 offers 140GPa while CuZn30 drops to 115GPa. But CuCr1 isn't as light as CuBe2.

==========

Series-produced instruments are known to vary. Maybe the tiny differences in hole positions, undercut... explain it, especially through the alignment of the overtones.

Or maybe the body's mechanical resonances, aligned or not with fundamentals and overtones, play a role? The dispersion of wood's stiffness, of worked metal sheet thickness, shift the resonances easily by a quartertone.

This suggests to fine-tune the mechanical resonances so they fall between the important sound components at design pitch. Near the end of the fabrication, measure the significant mechanical resonances, decide how much to shift each, run software on the matrix of resonance sensitivity to mass, add or remove the right mass at the right locations.

==========

My ramblings here are just paperwork. Experiments decide.
Marc Schaefer, aka Enthalpy