Tuning stability

[Rod Pearce]

How long would you expect a recently tuned instrument to stay in tune for?

 

I have tried to find some information by searching the forums but the number of hits from searching 'Tuning' is enormous. So I have opened a new thread on this subject.

 

I have tuned a number of instruments over the last 12 months, all reeds to within 1.5 cents after several passes. However, I have noticed that when checking the instruments after a few weeks / months a number of reeds have gone out of tune, or just not sounding correctly.

 

Is this to be expected ie an instrument'ss tuning should be expected to change periodically? If this were the case surely the majority of instruments in day to day use would be out of tune.

Is it that the reeds are getting past it and are no longer able to retain their tuning for long periods?

 

Any thoughts would be appreciated

[Lofty]

Colin Dipper tuned my Wheatstone Model 24 in 1994. It has been played a lot since then and its tuning is still excellent. I recently looked at the set of a few reeds which needed a bit more pressure than others to start sounding and that is much better now.

 

I think it’s probably not the reeds: more likely to be valves or reed shoes being a bit loose. Perhaps modern homes with low humidity are to blame.

 

Others who have more experience of maintenance and tuning may have different opinions.

 

Steve

[Rod Pearce]

Steve

 

Thanks for your feedback. Much appreciated.

 

As the concertinas in question have undergone a full refit of valves and pads, I would not expect the valves to be the problem. I am encouraged that your Wheatstone has remained stable, there is hope for me yet.

 

When tuning I did notice that a number of the reeds appeared to be quite thin as if they had been tuned a number of times previously. I will be fine tuning these instruments again soon so time will tell.

 

Regards

Rod

[ttonon]

Rod, I’m under the impression that the tuning of normally played steel tongues in concertinas would last an indefinite amount of playing, and I agree with Lofty that any changes are most likely due to changes in other materials than help define the pitch. 

 

I can substantiate my view with some technical data.  Some metals have a property called endurance limit, which means that, when subjected to periodic stress that completely reverses, as in the case of a vibrating reed tongue, and provided that the maximum stress experienced is limited to a certain level, the material can withstand an infinite number of cycles.  Of course no one has verified an unlimited number of cycles, so suffice it to say that the number of cycles is extremely high, even more than 100 million or so. 

 

Other materials do not have this property, and there is no maximum stress level for infinite endurance.  For these materials, as the periodically applied stress is applied, the number of cycles will always increase as the applied stress is lowered, but there's no lowest stress level that will allow the “infinite” number of cycles.  Such a state of affairs is represented on a “SN” or Stress-Number graph, available so easy now on the WWW.  Materials with an endurance limit have a curve that becomes horizontal at some lower level of periodically applied stress, while others have a curve that always drops as the number of stress reversals increases. 

 

Spring steel has an endurance limit, and thus, if the amplitudes of vibration of steel reed tongues remains below a certain maximum, they will theoretically never break, which probably means they will never undergo fatigue effects, always remaining perfectly elastic.  I do admit that my addition of this last “probably” phrase is my own guess, though it’s an educated guess that I believe many makers would support.  I remember Richard Morse making such a statement in this Forum.  The allowable stress limit for spring steel (AISI 1095) is about 43% of its ultimate yield strength, providing I believe a very convenient window for tongue design. 

 

In contrast, brass does not have an endurance limit, and we know brass reed tongues tend to break.  But there’s more to this story.  Several of us members have already discussed many of these issues here in a thread published in 2012,   I uploaded a supporting document for some of the relevant points I made in that thread, and for your convenience, I attach it here, which is a doc file showing two graphs.   The first graph is more relevant and I need to explain that the “endurance limit for brass” line in that graph is a stress level that would permit a very large number of cycles, and I don’t remember off hand what that number is, but at least 10^7.  Importantly, except for the longest reeds, steel tongues mostly have stresses under the maximum for infinite endurance, but many brass reeds experience stress levels above that (chosen) endurance limit. 

Best regards,

Tom

www.bluesbox.biz

Steel and Brass Stress.doc

[Rod Pearce]

Tom

 

Thanks for your very thorough and enlightening explanation. It will take me some time to assimilate it! I am unable to open the attachment, though.

 

I am taking from this and Lofty's reply that the reeds are probrbly OK and I should look at other factors.

 

Incidentally. I came across this thread after I posted my initial question. More food for thought.

 

Rod

 

 

 

[Dana Johnson]

I find that newly made reeds often go out of tune after being played awhile.  Removing material to create the reed profile changes the stresses in the steel and vibrating eventually equalizes things.  I find  reeds come off the grinder with a small amount of curvature, which I straighten out.  They are set in this condition, but the set can change with stress relief, changing pitch.  I have taken to running the reed pans on a vacuum ( my tuning rig ) for 3-5 hours either direction, after the initial tuning.  After tuning again, I find they stay in tune for a long time.  Lots of things can affect pitch, including set, pad or valve lift or stiffness, or even the variation of density of the reed pan with changes in humidity.  I don’t know if air density has a significant effect, but day to day changes in pitch are to be expected, and without knowing why a reed you just tuned is now off, you may be chasing a moving target to the detriment of the reed.

Dana

[Rod Pearce]

Dana

 

Thank you for replying. I am now feeling more comfortable with the idea that although the reeds were tuned pretty accurately, I should expect some settlement over time. I will of course be fine tuning them again and hopefully the long term results will be better.

 

Rod

[ttonon]

Hi Dana, good to hear from you.  To recapitulate, residual stresses are those stresses within a material even when there are no external forces on the material.  A good example is tempered glass, where the outside regions of the glass are in compression and the inside region is in tension, and such a state is intentionally produced by cooling outside regions faster than inside.  With all internal stresses, there are regions within the material that pull or push on other regions, achieving a static balance between tension and compression.  Thus, if some regions near the surface are in a push/pull with other regions, removing a particular outside region – as we do when tuning a reed - leaves no counterbalance to the forces produced by those other regions.  With unbalanced forces, the material can deform in some way, until a new state of internal stress establishes itself.  Whether we notice such deformation depends upon the internal stress pattern and how much material we remove.

 

I describe all that to support my contention that any tuning on a reed tongue surface that would cause a significant re-adjustment of internal stresses would most likely show up as a change in curvature along the length of the tongue.  I can see how such distortion could result in a noticeable change in pitch, particularly by way of a change in offset.  And such changes should be immediate. 

 

I’ve heard many claims that use of a reed relaxes internal residual stresses, but had my doubts, until I read your description of what happens to reed tongues after grinding.  By “grinding,” I’m assuming you mean that a region of the top or bottom surface of the tongue is remove in order to thin the reed in some area.  Such grinding is notorious for leaving residual tensile forces on the surface ground.  Thus, after grinding, the curve should be towards the side ground?  When you straighten it, you exceed the yield stress on that side by stretching it there.  Then, when the tongue springs back, it causes a residual state of compression on that side.  If, however, the opposite side experienced a compressive yield during your bending, upon spring back, it will wind up with a residual tensile stress on that side.  In general, residual compression near a surface is good, residual tension bad.  That’s because metals tend to fail in tension rather than compression and such residual tensile stresses will lower the fatigue resistance during cyclic (complete reversal of stress) loading, as in a vibrating tongue.  For other applications such as the leaf spring in a car, fabrication techniques are sometimes used to purposely impart a residual compressive stress on the outside side of the bending, even though residual tensile stresses remain on the inside of the bend curve.  There, the uni-directional bend cycling is favorable.  In the case of the tongue, I don’t think any residual tensile stresses that you leave on the one side when you straighten the strip should cause fatigue problems, simply because they are probably low magnitude.  And also because if it were, you would’ve noticed that straightened tongues tend to break. 

 

For the same reason, I think that the general statement, “usage relieves residual stress” can be misleading.  Firstly, during manufacture, all blue tempered spring steel 1095 is annealed at up around 1400 – 1500 F, which is a cherry glow, and that itself greatly relieves residual stresses.  Any residual stress in virgin material is very low.  In order for vibration to release residual stress, the vibration must raise stress in all portions of tongue beyond yield stress.  The tensile stress imparted by the vibration process alone, however, is below about 45% of the yield stress for all but the very longest tongues, otherwise you would notice frequent breakage.  Thus, for a release of residual tensile stress, the residual stress would have had to be about 75% of the yield stress, which is far above any residual stress after annealing. 

 

Incidentally, making tongues out of material that hasn’t been annealed, or from material that has an unknown history could be risky because for instance, if there are significant residual tensile forces such that the vibration process increases stresses much above about 45% ultimate yield stress, the endurance stress limit for that portion of the tongue will be exceeded, likely lowering the lifetime of the tongue.

 

I don’t know if air density has a significant effect, but day to day changes in pitch are to be expected, and without knowing why a reed you just tuned is now off, you may be chasing a moving target to the detriment of the reed.

 

 

It depends upon the accuracy and resolution of the instrument we measure with.  I can see that air density has some influence, but whether our ear/brain is sensitive enough, I can’t say.  Maybe when I finish my long-going analysis of tongue vibration, I can offer something interesting there.

 

Best regards

Tom

www.bluesbox.biz

 

 

[d.elliott]

After tuning, like a dwell period following the first couple of playings & sounding of the notes, some reeds change their set which can alter tone slightly, I tend to thing of this a 'playing in'. A practise commented on many times now. We bend the reeds past their elastic limit to change their curvature, the reed shape was stable before bending, the stresses balanced and the are stable after bending, static equilibrium. when we file a reed we alter the stress distribution through the reed,  it bends to a new position of static equilibrium. 

 

Vibration stress relief (VSR) has been used in industry for years, we used it on railway car (coach) bogies where we welded onto castings, I have seen it used on other fabrications and castings as well. I believe it is used on battle tank main gun barrels (Abrahams) . VSR  works best when the work piece, is not constrained ( like a reed tongue),  it fairly thin in section (like a reed tongue) and the vibrations achieve a resonant frequency (like a reed tongue). I have always ascribed the changes in set and very occasional pitch shortly after tuning to relaxation in the grain boundaries due to the phenomenon vibration stress relief. The conditions all fit, we know changes occur, and I have no other tenable explanation. I cannot prove it, regrettably.

[Dana Johnson]

Hi Tom, late reply, just saw this.

Grinding the top surface of the reed to create its overall profile creates a slight downward bend away from the ground face.  Which indicates to me that the balance in tension from whatever source changed in favor of the unground surface.  Straightening them moves the dislocations in the steel structure to allow the change to hold.  There is nothing guaranteeing that stresses created then will go away, only that they are not high enough to move the dislocations again.  However, adding the vibrational force may allow some of that stress to exceed the force needed to change the structure.  I find that hours of vibration are enough to allow these changes.

   Blue tempered 1095 steel may be annealed in some part of its creation, (in order to roll it to thickness, ) but it is hardened  after that and then reheated to around 600 deg.F to temper it and bring it down from file hard to about 55 on the Rockwell C scale.

   Different metals I am sure behave differently, but I find that in spring wire,  too small a coil size relative to wire gage creates compression failure on the inner diameter while the outer diameter shows no damage.  Not sure what that really means, since bending is not the same as pure tension or compression. 

Best,

Dana