jdowning - 9-30-2012 at 12:18 PM
Plain nylon strings have a lower density than PVF or gut so - for strings of the same diameter - are relatively less 'bright' sounding.
In larger diameters the lower pitch range of a nylon string is limited by 'inharmonicity' - making a string sound unacceptably 'dull'. The lower limit
is a subjective judgement, however, dependent on a number of factors including the tolerance of the listener. The situation may be a bit a bit worse
for a lute player (strings plucked with soft fingertips) than it is for an oud player.
For lute strings, 'Pyramid' limit maximum diameter of their plain nylon strings to about 1 mm. Range limit depends upon string length so, for example,
a plain nylon string of 61 cm length has a lowest acceptable pitch of d (147 Hz ) - according to the 'Pyramid' lute string calculator.
In the early 1980's some lutenists claimed that if a plain nylon string was twisted (like a gut string) then treble strings sounded relatively
'brighter' and the lower pitch limit of the larger diameter strings could be extended. Some applied the necessary twist with the string mounted on a
lute while others pre-twisted a string and set the twist by heat before mounting the string on a lute - all apparently with a satisfactory relatively
improved string performance.
The attached article by Eph. Segerman and Djilda Abbott of 'Northern Renaissance Instruments' published in FoMRHI in 1978 (Comm. 126) - summarises the
situation and provides a theoretical design solution (for the 'rocket scientists').
I am not sure if this technique is still used by lutenists and have never tried it (being a bit dubious about its claimed effectiveness) so thought
that it might be of interest (to both oudists and lutists) to run some trials for information and to try to evaluate the effectiveness of the
technique.
More to follow.
ameer - 9-30-2012 at 05:02 PM
As a fan of nylon (or more generally unwound) third courses I'm interested to see what results you come up with.
jdowning - 10-1-2012 at 12:05 PM
The scope of this investigation is limited to the thickest gauges of plain nylon in order to determine if twisting results in any significant increase
in elasticity resulting in extending the lower tone range.
High twist strings are not suitable for thin trebles as the twist reduces the breaking strength of a string and the problem with 'inharmonicity' is of
greater significance in the larger diameter, stiffer strings.
For this reason 'Pyramid', for example, limit the maximum diameter of plain nylon lute strings to 1 mm after which wound strings are required in order
to sound well at the lower pitches. For a string length of 61 cm 'Pyramid' give the lower tonal limit of just below d pitch (147 Hz or third course
Arabic tuning at A440). For a denser plain gut string this limit drops to just below B (123 Hz).
I happen to have two third course plain nylon guitar strings available for testing - the first of unknown manufacturer measuring 1.05 mm average
diameter the other (rectified nylon?) from a set of Augustine Red measuring 1.04 mm diameter. Tests to measure string tension will be undertaken on my
string test rig set to a string length of 61 cm with reference d pitch (i.e. third course Arabic tuning).
The first question to be addressed is how is it possible to twist a solid nylon monofilament string? A plain silk string made from a multitude of silk
fibres is easy to envisage as the filaments themselves are visible (under magnification). One explanation is that monofilament nylon being extruded
through a die has its molecules arranged longitudinally (which gives the string its tensile strength). So on a molecular level the monofilament nylon
string might be considered to be a fibrous material just like silk or gut. So lets run a test to find out.
To apply twist to the test strings I happen to have a string twisting device (used for experiments with and manufacture of plain silk strings) It is a
simple device consisting of a geared winder at one end (to save time in twisting a string) and a sliding dead weight at the other end ( to apply
tension to a string during winding to keep everything straight and kink free) - see attached image.
A much simpler alternative would be to hang the string from an overhead hook (in a doorway would be a convenient location) and hang the tensioning
weight at the other end. The weight is then spun around by hand and the revolutions counted - slow but sure!
For these tests the tension load applied is 1.84 Kg - just a guess at this stage (about half anticipated string tension) - for a length of string to
be twisted of about 70 cm.
jdowning - 10-2-2012 at 12:16 PM
The approach here is to twist the string on the twisting rig and then to set the twist with a hot air gun - as reported in the previously posted
FoMRHI article. No details were provided in the article about the twisting operation - so these trials must start from 'square one'.
The hot air gun in this case - used to soften the nylon and 'set' the twist - is my wife's hair dryer. Maximum hot air temperature from the dryer -
measured using an oven thermometer - is 75°C too hot to touch but which seems to be about the right temperature to set the twist with minimum
unwinding (about 15 turns) once released from the twisting rig.
Heat was applied by moving the dryer quickly along the length of the strings for several passes so as not to concentrate heat at any one place along
the string.
After twisting the string was then tested on the string test rig to measure tension at d pitch.
Here are the first preliminary results:
#1 string, plain nylon, 1.05 mm mean diameter, make unknown.
Test string length 61 cm, no twist.
pitch d (147 Hz) - measured tension 3.044 Kg - sustain about 6 seconds.
Twist applied over 73 cm length - 105 turns. Heated to 75°C to set the twist. Reduction in length about 3% due to twist. Increase to 1.08 mm mean
diameter.
Measured tension d pitch - 3.172 Kg
Additional twist applied to 175 turns total. Mean diameter increase to 1.05 mm.
Angle of twist about 35°
Measured tension d pitch (147 Hz) - 3.314 Kg - sustain about 8 seconds.
Tension at c pitch (132 Hz) - 2.613 Kg - sustain about 6 seconds.
Tension at B pitch (124 Hz) - 2.325 Kg - sustain about 5 seconds.
Pitch when tested at the original tension for the untwisted string of 3.044 Kg was about d - 80 cents - almost a semi tone drop in pitch.
Attempts to apply further 70 twists to the string failed when the string was observed to begin to wind upon itself and kink before suddenly breaking
before heat could be applied to set the twist.
So the preliminary results confirm that a monofilament nylon string can be twisted and the degree of twist measured - just like a gut or silk
monofilament string - and that the twisting does result in increased string elasticity with a consequent lowering of the pitch range.
jdowning - 10-2-2012 at 02:25 PM
Some further observations on the test results.
After twisting the string felt smooth but slightly 'bumpy'. This unevenness is barely visible however (see the macro images of the string) - human
touch being sensitive enough to detect the minutest of surface irregularities. The twist can be clearly seen in the macro images.
The measured tensions closely matched tensions calculated using the Arto Wikla calculator for the measured diameters - increased diameter due to
twisting equating to increased string tension.
The measured degree of string twist of about 35° is considered to be a medium twist for a plain gut or silk string. This degree of twist may be about
the maximum judging by the failed attempt to increase the amount of twist of the test string.
From a subjective assessment of string sustain measured on the string test rig it would seem that increased twist - as might be expected - results in
increased string elasticity and hence a reduction in the lower acceptable tone range limit, in this case, by perhaps about 3 semitones?
ameer - 10-5-2012 at 02:20 PM
I wonder if there is any merrit to twisting PVF or nylgut? What about twisting 3 or 4 smaller strings? Would that work or would they break?
jdowning - 10-5-2012 at 03:34 PM
I don't know but am only interested here in testing plain (and low cost) nylon strings to improve performance.
Monofilament PVF and 'Nylgut' strings already have a performance edge over plain nylon as they are denser materials so can be made relatively smaller
in diameter. The same applies to gut and silk strings.
Why don't you have a go at twisting PVF and 'nylgut' monofilament strings to find out if there might be any further benefit to be gained for oud
strings?
By twisting 3 or 4 smaller nylon strings do you mean making nylon strings of roped construction - like the modern roped gut or so called 'catline'?
That has already been tried for lute basses as far as I can remember. Not sure about the result but they do not appear to be commercially available so
were likely not a success.
Alternatively just twisting 3 or 4 smaller diameter nylon strings together to form a simply twisted larger diameter string would probably work - if
the nylon strands once twisted could be persuaded to stick tightly together as an homogeneous mass - and if the 'bumpy' end result is not considered
to be a problem. Again, I suspect that this approach has likely already been tried and tested (for lute strings) without resounding success.
jdowning - 10-6-2012 at 04:44 AM
When test string #1 broke while attempting to apply more twist, the released energy caused the string to tangle up on itself as the twists unwound
explosively.
After carefully unwinding the mess there was still sufficient length available to run further tests on the test rig. The original twist of 175 turns
seemed still to have been retained (judging from the twist angle) so a further 35 turns were applied followed by heating.
It was observed that the original length (knot to knot) prior to twisting reduced to a length by about 3 cm as the turns were applied. The measured
diameter after twisting a total of nominal 210 turns averaged 1.05 mm or about the same diameter as previously measured after 175 turns except that
this time around the tension for d pitch had fallen to 3.280 Kg on test (compared to 3.314 Kg previously).
Interestingly the sustain had increased from about 6 - 8 seconds to 9 -10 seconds. As this appears to be the limit of twist that could be applied
without the string distorting (or breaking under torsion) this may be the highest level of sustain achievable for a plain nylon string of this
diameter. This would have to be verified by a further series of tests.
By way of comparison the sustain of a Pyramid #650 orange oud set 3rd course wound string at d pitch, 61 cm vibrating string length, is about 15 -16
seconds as measured on the test rig.
All of the tensions recorded on the rig at the various diameters correlated closely with calculated values using the Arto Wikla calculator.
Tests on string #2 - Augustine Red label guitar plain nylon.
Vibrating string length - 61 cm
Original diameter untwisted - 1.04 mm
pitch at d - tension 3.012 Kg (cf A.W. 2.890 Kg)
pitch at c - tension 2.378 Kg (cf A.W. 2.382 Kg)
pitch at B - tension 2.098 Kg (cf A.W. 2.122 Kg)
Maximum sustain about 5 secs.
Applying 175 turns followed by heat treatment:
diameter - 1.06 mm
at d pitch - tension 3.118 Kg (cf A.W. 3.117 Kg)
at c pitch - tension 2.487 Kg (cf A.W. 2.474 Kg)
at B pitch - tension 2.206 Kg (cf A.W. 2.200 Kg)
Retwisting to add 35 turns to a nominal maximum 215 turns. At this point the string was observed to start buckling close to the winding hook - an
indication that maximum turns had been reached (see attached image)
adding twist shortened the string by 1.8 cm from 66 cm knot to knot. Heating relaxed and extended the string length by 0.6 cm to give a total
shrinkage of 1.2 cm.
Diameter 1.07 mm
pitch at d - tension 3.122 Kg
Sustain 10 seconds on test rig.
Note that the calculated Arto Wikla values for tension versus diameter above assume a string density for nylon of 1.08 gm/cc rather than default 1.04
gm/cc. Both of the test strings are about 30 years vintage so the assumed value is consistent with the density of nylon in use then (according to NRI
data).
[file]24329[/file]
jdowning - 10-6-2012 at 06:45 AM
To summarise the results of these preliminary trials:
1) Plain monofilament nylon strings (measuring about 1.05 mm diameter in this case) can be twisted up to a maximum limit of about 3 turns per
centimeter untwisted length.
2) Twisting increases diameter and reduces overall string length.
3) The increased diameter equates to increased tension for a given pitch (consistent with Mersenne-Taylor law).
4) At maximum twist angle of about 35° increased string elasticity results in about a doubling of string sustain.
5 Heating a twisted string to 75°C will set the twist and prevent unwinding (apart from a few turns when the string is released from the twisting
rig).
Control of heat application might be more precisely controlled for future trials by measuring string length contraction during winding and then
measuring string relaxation when heated. The greater the length of time a twisted string is heated the greater the increase in length of the finished
string (and reduction in diameter from the fully twisted state).
No tests have been undertaken so far on an instrument to verify performance (subjectively).
These limited trials suggest that twisting of thicker monofilament nylon strings might be a straightforward way to significantly improve performance
of cheaper strings.
jdowning - 10-7-2012 at 11:33 AM
Having gained an appreciation of the possibilities for improvement associated with twisting monofilament nylon, for the next round of trials an
attempt will be made to obtain more precise data for comparison. Hopefully some low cost, guitar 3rd, nylon strings for testing will be available from
one of the local music shops without having to go too far afield.
In assessing string sustain it was found best to measure this with a stop watch - listening for the duration of tone decay at the sound box sound hole
of the test rig. Recording an audio signal using a Zoom H2 digital recorder and analysing the result using 'Audacity' software was tried but found to
be a less sensitive solution.
For the second series, the elasticity of the strings will also be determined by loading a string sample on the test rig and measuring the consequent
longitudinal extension. The Modulus of Elasticity (Young's Modulus) of the string material is then defined as Stress/Strain where:
Stress = Load/String C.S. Area and
Strain = Extension of the String/ Original length.
jdowning - 7-10-2013 at 10:52 AM
In order to extend testing to larger diameter monofilament nylon strings than those generally available for guitar/lute/oud third courses (about 1 mm
diameter limit) it has been necessary to look for other sources.
One source of larger diameter nylon monofilament strings are those made for harp that are available in diameters up to about 1.5 mm.
Another potential source might be tennis racket strings that are made in a variety of materials (nylons, copolymer polyesters etc) - not only as
monofilament but also multifilament and braided construction and in diameters up to about 1.4 mm. Prices are also reasonable - starting from about 60
cents a metre (or less if purchased in bulk coils).
The question is can these strings be made to perform as instrument strings after twisting to increase flexibility? Only one way to find out!
For preliminary trials I have obtained two coils of tennis racket strings - nominal 1.25 mm diameter (#17 gauge) - from a company in Calgary ('Golden
Set International').
One coil measuring 12.2 metres in length is a monofilament copolymer polyester - "Power Cord 17" - costing just over $6 per coil.
The other coil measures 13.6 metres is of multifilament/braided nylon (?) construction - "Anabolic 17" - costing about $6 per coil (currently on
sale).
The monofilament string is a smooth extrusion quite stiff compared to nylon. Measuring volume and mass of a metre length, the specific gravity works
out to about 1.4, significantly greater than nylon (or silk and gut) but less than PVF carbon.
The braided string is much more flexible. Measured specific gravity is about 1.04 - the same as nylon.
The monofilament supplied is dark grey in colour - an error as I ordered Amber for colour. However, the colour will make no difference to the test
results so dark grey it is.
Testing will begin next with the monofilament string sample.
[file]27153[/file] [file]27155[/file] [file]27157[/file] [file]27159[/file]
jdowning - 7-18-2013 at 12:17 PM
For the twisting of monofilament co-polymer polyester or nylon to succeed in making the string more flexible, the degree of twist must be 'set' with
heat after twisting. These plastic materials are thermoplastics - that is they can be softened (or melted) by heat but will return to their 'hard'
state on cooling.
The temperature at which a hard material like plastic starts to soften is Tg or 'glass transition temperature' which is below the melting temperature.
So, for example Tg for Nylon 6 is said to be 47°C and for polyester PET (PolyEthylene Terephthalate) is 70°C.
As the co-polymer of the tennis racket string under test is unknown is has been necessary to run a few (albeit rough) tests to find the approximate Tg
for the string. Immersing the string in boiling water (i.e. known temperature 100°C) did not soften the string so a heat gun capable of producing air
temperatures up to around 300°C at the gun outlet was used. Measuring air temperatures at varying distances from the gun with an oven thermometer it
was found that the approximate Tg for the material under test is about 120°C with melting point around 220°C.
The co-polymer tennis racket string is quite hard and stiff and not suitable for an instrument string without further processing. On the twisting rig
it was possible - on a preliminary trial - to apply about 170 turns to an initial string length of 84 cm length. The twist was set with the heat gun
used 'freehand' until the string was observed to 'relax' slightly from its fully twisted state (i.e the string length was measured and observed to
start to lengthen slightly). In the fully twisted, heat set, state string length had reduced to about 78 cm with an increase in diameter from 1.23 to
1.25 mm diameter to 1.28 to 1.30 diameter. The constant load applied to the string during twisting was 3.6 Kg.
Next to see how the modified string performs under varying string tensions on the string test rig.
jdowning - 7-20-2013 at 08:54 AM
Earlier in this thread forum member ameer wondered if it might be possible to make 'roped' construction bass strings from twisted nylon monofilament.
I thought that this had been tried before but without much success (commercially).
It turns out that Northern Renaissance Instruments has been making 'Processed Polyester' strings for a few years now - although their most up to date
price list is dated 2009 so I am not sure if there is any demand for them. It does show, however, that it can be done.
http://www.nrinstruments.demon.co.uk/ProcPolySt.html
One supplier of monofilament nylon harp strings tells me that their strings are made by 'DuPont' - as filament for manufacturing of brushes. Checking
the DuPont website, the most likely composition is their 'Tynex' nylon monofilament (apparently also used for guitar strings). The DuPont technical
paper 'Filament Performance in Brushes' provides very useful information about the properties of 'Tynex' (and other filament grades).
http://www2.dupont.com/Filaments/en_US/assets/downloads/Paintbrush/...
The diameters of monofilament 'Tynex' available in the larger diameters are 0.914mm, 1.016mm, 1.143mm, 1.270mm, 1.394mm and 1.524mm - all matching the
harp string diameters offered for sale.
http://www2.dupont.com/Filaments/en_US/assets/downloads/Toothbrush/...
jdowning - 7-20-2013 at 12:13 PM
The first attempt at twisting the co-polymer tennis racket string has now been tested on my string test rig at vibrating string lengths of 58.5 cm, 61
cm, and 67.5 cm (Modern Turkish, Arabic and 16th C Oud/Lute) with string tensions of 2.0, 2.5, 3.0, 3.5 Kg force.
Even after twisting and heat setting, the string still feels excessively stiff. String sustain ranged from about 5 seconds to about 7 seconds -
greater for the longer string lengths with higher tensions - but quite a bit less than modern wound strings having a sustain of 15 seconds or so.
The approximate tonal range (A440 standard) for the 58.5 cm string ranged from Fsharp to B - 50cents (2kg to 3.5kg tension respectively). For the 61
cm string length from F to A sharp and for the 67.5cm string length Dsharp + 40cents to Gsharp +20cents.
So for an Arabic oud of 61 cm string length tuned say - C F A d g c' - this string could be used for the fourth course A (110 Hz) at a tension of 3.26
Kg force (or about 32 Newtons tension).
Tested on a lute with string length of 67.5 cm, the string tuned to G pitch (about 3.25 Kg tension) had a 'woody' timbre totally different from the
'brassy' sound of a wound string.
Next to test if the string can be processed to take more twists to reduce stiffness and improve the current marginal performance.
jdowning - 7-23-2013 at 12:03 PM
Starting with a new length of co-polymer 'Power Cord 17' on the twisting rig - this time tensioned with a 4.4 Kg load (increased from 3.6Kg for first
trial) for an initial length (knot to knot) of 107.5 cm - the sample was twisted in steps of 17.5 twists measuring the reduction in length at each
step.
For the first 35 twists there was no measurable reduction in length - presumably as any slackness was 'taken up' (due to preliminary tensioning of the
mounting loops and their knots at each end of the string).
At 175 twists of the geared winding mechanism (an 'old fashioned' hand cranked drill with 3.5 : 1 gear ratio), the tension in the string - twisted
like a spring - was sufficient to rotate the winding mechanism backwards when released. At this point the string had shortened to a length of 99.7 cm
due to the twisting.
Twisting was continued until at 230 twists the string suddenly failed at the lower knot. String length at failure was about 93.3 cm
The length of the broken string was about 97.0 cm with a uniform diameter of 1.30/1.31 mm (increased from 1.23 mm for the untwisted string) - so after
breaking the string had untwisted somewhat (to about the length at 205 turns). Interestingly, the broken string after 24 hours has remained stable -
retaining its broken length and diameter without need for any heat application. Presumably the stress level in the string at maximum twist before
failure was sufficient to induce permanent displacement (after slippage) of the molecule chains in the string?
The surface of the twisted string felt smooth to the touch and it was not possible to detect the degree of twist in the string (due to the dark
colour) except at the point of failure.
The attached macro image of the string at the point of failure shows that the degree of twist is about 35° after failure and relaxation of the string
material.
This degree of twist is within the range of 'High Twist' gut strings judging from the images of HT gut on the 'Aquila' website (range 35° to about
40°).
Due to the catastrophic failure of the test string, the string is not straight but buckled in places so may next require correction with heat stress
relief under load. It is also quite stiff despite the twisting so perhaps will also be 'softened' when stress relieved?
More to follow.
[file]27211[/file]
jdowning - 7-24-2013 at 12:11 PM
The length of broken test string was straightened by again mounting it on the twisting rig under a tension load of 4.4 Kg. Some additional twists were
applied (about 32) until resistance was felt in the winding gear. The string was then stress relieved with a hot air gun until the string softened
enough to just start to increase in length under load.
The load was then removed and the string left overnight to relax before being retested on the string test rig. By this time the string diameter had
reduced from 1.30/1.31 mm to 1.27/1.28 mm. On test there was no significant change in the measured tonal range compared to the twisted string first
tested except that the sustain of the string had improved slightly - about a 15% increase for all string lengths (58.5 cm, 61 cm, and 67.5 cm) at 3 Kg
force tension.
The string length has again been heat treated under a load of 4.4 Kg with the hot air gun until the string length had increased by a constant 13 mm
and diameter reduced to 1.26/1.27 mm.
The string will next be again tested on the string test rig to determine if there is any significant improvement in performance.
jdowning - 7-25-2013 at 12:10 PM
No significant improvement can be reported for the heat treated string - maximum sustain measured (by ear with a stop watch) on the string test rig is
still around 8 seconds. So I think that this is about the limit for a string of this material and diameter. The co-polymer is quite hard and stiff
even after applying maximum twist.
The braided 'Anabolic 17' racket string was next tested but although quite elastic the sustain did not exceed 5 seconds at best on the string test
rig. Twisting the string did not significantly improve performance (max 6 seconds sustain) and resulted in a slightly uneven string - similar to that
of a high twist gut string.
I do not know how these results compare to high twist gut strings of similar diameter - not much different perhaps?
Previous tests with plain nylon guitar 3rd strings - where sustain after twisting and heat treatment was measured at 10 seconds on the string test rig
- suggest that perhaps this the way to go, with twisted monofilament nylon harp strings in the larger diameters (Dupont 'Tynex').
More on this later once some sample harp strings have been aquired for testing.