| Pages:
1
..
6
7
8
9
10
..
13 |
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
I am not sure when 'Japanese Silk Gut' first appeared on the market as a commercial product but suspect that it was after 1945 - after the end of
World War 2.
The attached extract from an article in the 'Post Graduate Medical Journal', 1949 by A.M.C. Humphries, M.P.S. briefly describes Japanese Gut and its
manufacture - as applicable to medical sutures.
The material also appeared, post war, as 'Silk Fishing Gut' - a 'monofilament' leader that imitated true 'Spanish' silkworm gut. The attached images
are samples of Japanese gut fishing leaders that came on the market during the period that Japan was occupied by Allied forces post war (1945 to
1952). Old stock of this material is still being offered for sale on Ebay for anglers choosing to fish in the traditional way (akin to lute and oud
players interested in performing with 'authentic historical' strings!).
Note that the leaders are 'waterproof' yet must be soaked well in water before use!
Note also the long unjointed lengths of about 10 yards (3 metres).
I happen to have four samples of Japanese silk gut fishing leaders of unknown age but most likely postwar. The packages are not stamped 'made in
occupied Japan' so may date to some time after 1952 (until they became commercially uncompetitive with the newly introduced nylon leaders - about
1960?).
Next to examine and analyse these gut leader samples to see if they might be viable as instrument strings for the oud or lute.
[file]26531[/file] [file]26533[/file] [file]26535[/file] [file]26537[/file]
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
I have four samples of Japanese silk gut fishing leaders sent to me for evaluation as instrument strings some years ago by Alexander Rakov. They range
in diameter from 0.37 mm to 0.63 mm with advertised dry tensile strength from 10 lbs force to 25 lbs force (45 Newtons to 111 Newtons) and are about
90 cm in length.
The age of the samples is not known but is likely to be about 50 years old.
The leaders are semi transparent 'misty' grey in colour - stiff and springy - just like monofilament nylon in appearance. Unlike Spanish silkworm gut
the leaders can be tied dry into a tight knot without need for soaking in water and without breaking.
A small sample was boiled in water for several minutes to dissolve the binder and reveal the silk filament construction. The filaments were so
thoroughly saturated in the binder that this treatement did not remove all of the binder.
According to the Humphries article on Japanese gut previously posted, the binder used on Japanese Gut is a mixture of animal glue and an extract from
seaweed. The seaweed extract is most likely Agar-Agar a gelatin - like animal glue. It is extracted from red seaweed, solidifies at 37 °C, insoluble
in cold water but dissolves readily in boiling water. Absorbs water up to 20X its own weight. Sets to a firm gel at concentrations as low as 0.5%.
It is not eaten by bacteria so is used as a culture medium in laboratories (Petri dishes).
According to Humphries the critical proportions of binder to silk filament are 85% silk to 15% binder for optimum strength. The uniformity of the
diameter of the leaders suggests that during manufacture the leaders were passed through sizing dies.
The specific gravity of the samples under test was determined by measurement of leader length and diameter and by 'weighing' . The calculated S.G. is
1.36.
The relative transparency of the leaders indicates that they may have been made from degummed silk filament?
So - to all intents and purposes - Japanese gut may be nothing more or less than a low twist silk filament string made with a glue binder - akin to
the early strings described in Kanz at-Tuhaf or perhaps the late 19th C Acribelle silk violin strings?
On the other hand Humphries implies that raw silk was used. He states that the silk is reduced to a semi liquid state when boiled with the glue - but
more likely it is the sericin natural gum coating that becomes semi liquid not the silk filament (fibroin). Also the gut leader package states that
the gut is made from the finest virgin (not degummed?) silk - perhaps again implying that raw silk was used?
Next to test the 0.37 mm diameter leader as the top string on a lute.
[file]26561[/file] [file]26563[/file] [file]26565[/file] [file]26567[/file]
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
The gut sample has been fitted for testing as the first course on a lute of 60 cm vibrating string length - pitch f' (350 Hz) - equivalent to the
maximum pitch allowable for a plain gut string.
While tuning up the string to pitch the sound was quite promising but the string broke at around a tone below full pitch - equivalent to a string
tension of about 2.3 Kg.f.
The break occurred between the tuning peg and nut - the point of highest tension with the string being raised in pitch (the tension being somewhat
higher than that derived from the pitch).
The breaking load was measured on a test rig under direct loading conditions giving a value of 2.6 Kg.f the breaking load equivalent to an ultimate
tensile stress (UTS) of about 0.24 GPa.
The advertised breaking load of the gut leader is 10 lb.f or about 4.6 Kg.f. giving a UTS of about 0.42 GPa. If the string manufacturer provided a
reasonable safety factor of say 20%, the UTS of the string originally would have been about 0.5 GPa which is an average value for raw Bombyx mori
silk.
The test string therefore failed at a value of less than half the original tensile strength.
Why such a discrepancy? Did the manufacturer provide false information about the breaking load of the gut or is there some other explanation for the
failure.
Daylight and sunlight contain levels of ultraviolet light that are not only dangerous to unprotected exposed skin - more so these days than in the
past - but also destructive to both natural and man made fibres. Silk is one of the most sensitive materials susceptible to damage due to UV exposure
- the thinner the fibre the greater the damage - somewhat less so for raw silk where the sericin gum coating can provide some degree of protection
against degradation.
Checking out another technology for information.
Parachutes are made these days from nylon fabric but were once made from silk. The severity of degradation for nylon parachute fabric exposed to
summer sun is a loss in breaking strength of 52% after one week, 71% after two weeks and 94% after three weeks exposure. The degradation of silk
parachute fabric is even more severe.
For this reason parachutes used by the armed forces are given a service life or limited to a certain number of jumps. For the older type of silk
canopies service life was usually 7 years. Tests on 15 year old silk parachute canopies (non continuous exposure to daylight) showed that their
strength had fallen to 30% below original specification.
Sky diving anyone!
The age of the gut under test is about 50 years and it has been kept in transparent pakages so will have experienced at least some (but an unknown
level) exposure to daylight. The loss of strength, therefore, is most likely due to ultra violet degradation of the silk over time made worse by the
transparency of the gut.
So store all of your instrument strings - nylon, gut or silk - in dark envelopes in a drawer away from daylight exposure - particularly the thin
treble strings. Thicker bass strings are less liable to degradation than thin strings. Dyes - particularly the darker colours - may also afford some
protection against UV degradation.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
It is doubtful, if the once viable Chinese drawn Saturnid silkworm gut cottage industry survived the total war and post war situation in that country
(1937 to 1950) despite efforts to re-establish the industry in Formosa (now Taiwan).
The Japanese - like most other silk producing countries - did not appear to be successful in producing drawn silkworm gut that was able to compete
with the Spanish material. However they were able to offer their own version of silk gut for the international angling market - made from silk
filament with a glue binder.
We do not know the source of 'Acribelle' violin strings that were available around the end of the 19th C but I would guess that the Japanese gut silk
fishing leaders that appeared on the market post WW2 were those very 'Acribelle' strings - simply a convenient dual purpose adaptation.
This connection between fishing leaders and instrument strings is of course applicable today where PVF fishing leaders are successfully being employed
as instrument strings.
Interestingly there is a much earlier historical precedent for this practice.
The famous English diary writer Samuel Pepys made the following entry on March 18, 1667:
" This day Mr Caesar told me a pretty experiment of his, of angling with a 'minikin', a gut string varnished over which keeps it from swelling and is
beyond any hair for strength and smallness"
A 'minikin' was a small diameter treble lute string of the time. The usual practice then was to use horse hair for fishing leaders - hence Pepys's
interest. Presumably the lute string in question was made from sheep's gut waterproofed with a coat of varnish. It is not known from this description
if minikin strings were protected with varnish applied by the string makers or if Mr Caesar found it necessary to varnish the string for his angling
application.
The hair from the tail of a horse can range from about 0.1 mm to 0.28 mm in diameter. Samples of horse hair that I have to hand range in diameter from
about 0.1 mm to 0.26 mm.
It is hard to imagine that a gut top string of a lute could be as fine as even 0.28 mm in diameter let alone thinner. Perhaps the work horses in 17th
C England were a sturdier breed with much thicker tail hairs than horses today? Or was Mr Pepys just exaggerating?
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
The limited 'shelf life' of silk filament (due to deterioration when exposed to adverse environments) is something that probably should be taken a bit
more seriously for these experiments.
Weak acids are more damaging to the fibroin than weak alkalis so I should probably discontinue trials with acid based flexible hide glues as well as
chemical weighting - the latter being well known for its destructive effect on silk fibres. This will restrict future testing to untreated binders
(but include alkali based flexible binders) and loading with metal powders or thin wires incorporated into the inner core of bass strings.
Dyeing of the string binder may also provide a significant protective barrier against ultra violet penetration (which might have been one reason for
colouring the strings of an oud?). This would apply to strings made from either raw or degummed silk - both requiring a binder of some kind.
The treble strings made by Alexander Rakov that I have to hand are nicely made - smooth, uniform, and simply twisted from raw silk cooked up in the
Chinese way. They look like gut strings in colour.
The coil of 0.45 diameter silk string will be tested as the first course on one of my lutes with 60 cm vibrating string length at f' pitch (A440
standard) which should be about the upper practical limit for the string at about 35 Newton tension (3.6 kg.f) - a bit on the high side but we will
see how it goes.
The density of the string was measured and calculated to give a Specific Gravity of 1.26 - a bit on the low side for a string made from raw silk but
within the range expected, silk being a variable material.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
I have yet to prepare isinglas (fish swim bladder or sound) as a glue binder so am not sure how it will turn out (and there is really only one way to
find out). Before doing so there are a number of interesting early accounts on preparing the glue for various purposes.
In "The Art of Making Various Glues" by M. Duhamel du Monceau of the Royal Academy of Sciences, Paris, 1771 is the interesting statement that 'Fish
glue is still used to gloss fabrics of silk and especially ribbons. The Gauze-workers use it a great deal'. Fish glue in this context is isinglas.
This not only suggests that isinglas glue is flexible (when combined with silk fabrics and braided ribbons) - but it is tempting to make an
association with 'ganser' braided strings and use of isinglas as a binder.
A further entry - When used for sealing fabrics the glue is prepared in the usual fashion by heating and dissolving in water to which brandy
(eau-de-vie) is then added.
Another entry refers to isinglas dissolving more quickly in wine (than water) and even better in brandy - a glue that is quite different from
Strong-glue (hot hide glue?) which does not dissolve at all in alcohol.
Another reference to alcohol modified isinglas glue can be found in 'The Mechanic or Compendium of Practical Inventions - Volume 1' Liverpool ,19th C
under 'The Manufacture and Uses of Isinglass' page 333 - where it is stated that the isinglas is first broken into small pieces and then immersed in
common gin (ethyl alcohol/water), placed near a warm fire and then shaken frequently until the isinglas is fully dissolved. "This strong glue will
keep for many years and is a very convenient preparation for a number of purposes. By warming, it becomes fluid and fit for use"
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
The 'on loan' library copy of 'Epistles of the Bretheren of Purity, Epistle 5' by Owen Wright, O.U.P. 2010 has arrived and I have had a chance to
briefly read those parts concerning oud strings - so will take time out here to review and analyse the information. Previously, I only had part
translations by G.H. Farmer and Amnon Shiloah to refer to. The full translation includes some critical information omitted from the part translations
but otherwise is in agreement with the latter.
The Ikhwan al- Safa describe an oud as having four strings (double courses?) - no more no less. The thickness (diameter) of the lowest string (Bamm)
is 4/3 the diameter of the third string (Mathlath) which is 4/3 the diameter of the second string (mathna) which, in turn is 4/3 the diameter of the
highest string (Zir).
From this it is further stated that the lowest string should be made from 64 threads of silk, the third 48 threads, the second 36 threads and the
highest 27 threads. The string lengths are equal but their thicknesses (diameters) will differ according to the ratio 64:48:36:27.
The four strings are tuned a perfect fourth apart.
The highest string is first tightened as much as it can go without breaking before the other strings are tuned from it.
Strings that are identical in diameter, length and tension when plucked will sound identical. If they are identical in length (and tension?) but
different in diameter, the sound of the larger diameter strings will be lower than the thinner strings. If the strings are identical in length and
diameter but different in tension the sounds of the tenser strings will be higher than the the sounds of the slacker strings.
That's it - so let's see what might be deduced from this limited information concerning string design of a 10th C oud according to the Ikhwan
al-Safa.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
The first important bit of information is that the top string - Zir - is tuned to as high (a pitch) as it will go without breaking. This was also the
instructions for tuning the all gut strung European lute in the 16th C - the reason is to give maximum latitude for the thicker bass strings to sound
reasonably well.
The second is that the strings were made only from silk.
The third is that the four strings were tuned a fourth apart.
The fourth is that the string diameters increased in steps of 4/3 from treble to bass i.e. in the ratio 64:48:36:27
The fifth is that the strings were all made from silk filament and binder of the same density as there is no mention of density as it affects the
vibration of a string - only string diameter, string length and string tension.
Taking a practical example and assuming that the string length of an Ikhwan al-Safa oud is 63 cm. Silk is a variable material but similar to plain gut
in density - assumed here to be 1.3 grams/cm² - and other properties.
If an arbitrary pitch standard of a semi tone below modern A440 is assumed, the maximum pitch of the zir string would be f' (329 Hz) taking a silk
string as comparable to gut.
From these assumptions the tuning of the oud strings would be - Zir to Bamm - f'( 329 Hz), c' (247 Hz), g (185 Hz), d (139 Hz).
If the Zir string is assumed to be 0.45 mm in diameter then according to the ratios given by the Bretheren, the other string diameters would be 0.6
mm, 0.8 mm and 1.07 mm.
Given equal tension and string density - these diameters agree with the diameters calculated using the Mersenne-Taylor law for vibrating strings. As
the frequency of vibration of a string is inversely proportional to string diameter (length, tension and density being constant) then the tuned
frequencies of the strings are related by the 4/3 ratio (i.e. 139 X 4/3 = 185, 185 X 4/3 = 247 and 247 X 4/3 = 329).
All well and good but now comes a puzzle. The Bretheren go on to give the number of silk threads required to make each string as Zir 27, Mathlath 36,
Mathna 48 and Bamm 64 - 4/3 ratio again, 64:48:36:27. The problem is that the diameter of a bundle of silk threads depends upon the cross section area
of the bundle which is the number of threads X their c/s area.
So, taking the calculated c/s area of the 27 threads in the 0.45 mm Zir string as the c/s area of the silk threads to make the remaining strings -
their diameters would be 0.50 mm, 0.60 mm and 0.70 mm - far short of the required 0.6 mm, 0.8 mm and 1.07 mm mentioned above.
So did the Ihwan al-Safa get their geometry wrong in trying to demonstrate the universal application of the 4/3 ratio - or is there another
explanation for the apparent discrepancy?
More to follow
|
|
|
Brian Prunka
Oud Junkie
Posts: 2916
Registered: 1-30-2004
Location: Brooklyn, NY
Member Is Offline
Mood: Stringish
|
|
Quote: Originally posted by jdowning  |
So did the Ihwan al-Safa get their geometry wrong in trying to demonstrate the universal application of the 4/3 ratio - or is there another
explanation for the apparent discrepancy?
More to follow
|
I was curious about this and expecting this result as soon as I read your previous post. It seemed to me that applying 4/3 ratios to string diameters
was more likely to be idealistic philosophy than empirical observation of actual practice.
Curious to know what else you have up your sleeve!
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
Provided the string length, string tension and string material density remain constant for the strings tuned a fourth apart then the 4/3 relative
increase in string diameters given by the Ikhwan al-Safa is in agreement with the Mersenne-Taylor Law.
What does not seem to make sense is the relative proportions of the number of silk threads required to make the strings - these being in the same
ratios as the string diameters (64:48:36:27) according to the Bretheren.
Silk strings being made of a multitude of silk filaments must be twisted to form a uniform cylindrical string. Twisting the bundle of filaments
increases the string diameter - the greater the amount of twist the greater the diameter increase to a limit where the amount of twist causes filament
(and string) failure. Twisting a bundle of filaments reduces the tensile strength of a string (the outer filaments being stretched more than the inner
filaments) so that the top string should be made with as little filament twist as possible.
From limited trials reported earlier in this thread, the maximum diameter increase related to twisting the filaments is about 10% to 12%. Minimum
amount of twist to form a cylindrical string is about 2% to 5%. More testing will be required to confirm these figures.
So for our example twisting the thread bundles of the second, third and fourth strings measuring 0.5, 0.6 and 0.7 mm diameter (made from 36, 48 and 64
threads respectively) increases their diameters to a maximum of about 0.6, 0.7 and 0.8 mm respectively.
Time out - more to follow
|
|
|
Brian Prunka
Oud Junkie
Posts: 2916
Registered: 1-30-2004
Location: Brooklyn, NY
Member Is Offline
Mood: Stringish
|
|
Yes, but that assumes that equal string tension on all strings is a desirable result, which is not the case in my experience. There are many other
factors affecting perceived tension, as well as the loudness, projection, and tone quality of each string.
Of course, silk strings could possibly work well this way, but given there are many reasons why it might not it seemed unlikely to me. Add to that
the historical tendency of theorists to prescribe "ideal" practices that often do not correspond to the usual practices of musicians, and I become
suspicious when I come across an ideal mathematical correspondence such as this (in regards to resurrecting the actual historical practices of
musicians).
Regardless, fascinating stuff, thanks for taking the time to share this with us. Looking forward to reading about the rest of your discoveries.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
No, it has not been assumed that equal string tension is a desirable result - the assumed equal tension is just a starting point for this discussion -
which, as has now been demonstrated, clearly does not work by only increasing string diameter through twisting the string filaments (at least based
upon limited experimental data so far).
So what are the alternatives - according to the Mersenne-Taylor law?
1) Increase the linear density of the strings - by perhaps wrapping additional (but an unspecified mass of) silk material around a core made from the
number of filaments given by the Bretheren (like the Chinese wrapped strings)
2) Reducing string tension for the thicker strings.
Both of these options were explored earlier on this thread (page 4) concerning the Kanz al-Tuhaf silk strings where there is a similar difficulty with
thread count versus string diameter.
The Bretheren make no reference to string density as it affects string vibration (so perhaps option 1 can be discounted for lack of other evidence to
the contrary) but they do state (what appears to be obvious given equal density)) that if the strings "are identical in length but different in
tension, the sounds of the tenser ones will be higher and the sounds of the slacker ones lower"
So, looking at option 2) and assuming string tension has to be varied in order for a uniform density (1.3 gm/cc) monofilament silk string set of
maximum twisted diameters 0.45, 0.6, 0.7 and 0.8 mm to achieve pitches f' 329 Hz, c' 247 Hz, g 185 Hz and d 139 Hz, string tensions would have to
range from about 35 Newtons down to about 21 Newtons for the thickest Bamm string. Would this work satisfactorily? I don't know (yet) - it would
depend to some extent on the flexibility of the thicker strings, the acoustic response of the oud and the judgement of the listener.
The perception of a 10th C musician in judging the 'ideal' sound (i.e. one that works well) of an all silk strung oud is very likely to be far
different than that of a modern musician brought up on the sound of plastic/metal wound strings at higher than historically possible tensions (and
electronically amplified to boot)?
BTW the physics of strings as expressed by the Mersenne- Taylor law is not some kind of idealised historical mathematical concept.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
So - to recap for this example:
The silk thread bundles given by the Ikwan al-Safa are in number 27, 36, 48 and 64 yielding calculated untwisted diameters of 0.45 mm, 0.5 mm, 0.6 mm
and 0.7 mm. The required target diameters according to the Bretheren's specified 4/3 ratio diameter increases (and the Mersenne-Taylor law for equal
tension and 1.3 string material specific gravity) are 0.45mm, 0.6 mm, 0.8 mm and 1.07 mm for pitches (at A 415 standard) of f' 329 Hz, c' 247 Hz, g
185 Hz and d 139 Hz.
Twisting a bundle of filaments to form a cylindrical string shortens the original untwisted string length and increases the string diameter - the
greater the amount of twist the greater the increase in diameter. The volume of material - before and after - remains the same
The big question is can simply twisting the filament bundle result in sufficient increase in diameter to meet the target diameters (equal tension).
Based upon experimental data posted earlier on this thread the answer would appear to be negative for the second, third and fourth strings as these
will only increase in diameter (with simple twisting) to 0.56 mm, 0.69 mm, and 0.83 mm (diameter increases ranging from about 12% for the Mathna to
about 18% for the Bamm).
There is another alternative to increase the diameter further for the larger diameter strings and that is to use a multistrand roped construction -
Chinese style. The Chinese roped strings (for the qin at least) were made either from three or four strands. The early Chinese texts state that when
twisting their strings, the original untwisted length is reduced by 40% - this being equivalent to a calculated diameter increase of about 29% . If it
is assumed that the third string (Mathlath) is made up of three strands of 16 threads twisted together like a rope and the Bamm string is likewise
made from four strands of 16 threads ( 3 x 16 = 48 and 4 x 16 = 64) then the calculated twisted diameters would be 0.77 mm and 0.90 mm respectively.
Still not enough.
So, if the 64,48,36 and 27 thread count strings are to work string tension must be reduced from treble to bass (in accordance with the observations
made by the Brethren concerning string vibration and the Mersenne-Talor law).
This is how things work for all gut strung lute string tensions and so is valid for the oud.
So if the Ikhwan did not get their geometry wrong (in relating string thread number ratios to string diameter ratios) they may be telling us something
important about how the strings were twisted and constructed.
This being the case, the following string construction/tension is proposed for this example - all silk:
Zir string - simply twisted, minimal twist. Diameter 0.45 mm string tension 35 Newtons.
Mathna string - simply twisted, maximum twist. Diameter 0.56 mm, tension 31 Newtons.
Mathlath string - simply twisted, maximum twist . Diameter 0.69 mm, tension 27 Newtons.
Bamm string - roped, four strand. Diameter 0.90 mm, tension 25 Newtons.
Alternatively if the Mathlath string was made of a three strand roped construction the diameter would be 0.77 mm and tension 33 Newtons. This would
result in an uneven relative reduction in string tension so has been rejected as an alternative.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
Note that the proposed string design configuration for this example is based upon the Ikhwan al-Safa constraints on string thread count (64, 48, 36
and 27).
Otherwise, if more silk threads than the 4:3 ratio are used as string diameter increases it would be possible to make all of the strings of simply
twisted construction (i.e. no roped construction strings required) - increasing the degree of twist for the thicker strings in order to maintain
adequate linear flexibility.
Taking plain, simply twisted gut strings for comparison, the potential maximum tonal range (for this example, 63 cm string length) would be from a
high f' (low twist) string to a low F (high twist) string - or a potential maximum of two octaves (source N.R.I.).
Plain, simply twisted silk strings should have a similar performance - a tonal range f' to c - over four courses for this example - being well within
the potential two octave maximum range.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
Out of curiosity I have calculated the approximate Helmholtz resonance frequency for an Ikhwan al- Safa oud bowl using the previously posted example
with an assumed 63 cm string length (28 finger units of 2.25 cm) and sound hole of 4 units or 9 cm in diameter (or 4.5 cm radius)
The calculation procedure is covered in detail on page 10 of the 'Old Oud - New Project' topic on this forum.
http://www.mikeouds.com/messageboard/viewthread.php?tid=8488&pa...
The proposed geometry of the Ikhwan al- Safa oud is discussed here:
http://www.mikeouds.com/messageboard/viewthread.php?tid=11186&p...
The resonance frequency of the bowl cavity/sound hole relationship is proportional to the square root of the sound hole area divided by the volume of
the bowl cavity x the equivalent thickness of the sound hole (assuming that the Helmholtz formula - applicable to the resonance frequency of a sphere
with a short cylindrical neck - is approximately equivalent to the oud relationship).
To calculate the volume of the bowl precisely from the geometry requires a knowledge of integral calculus - a useful branch of mathematical science
that I once knew reasonably well but have long since forgotten. So, in order to calculate the volume of the bowl I must resort to what might be called
'macro integral calculus' - where the volume of the upper part of the bowl is determined from the calculated volume of individual slices added
together. As the section of the bowl is semi-circular, the volume of each slice is half the area of a circle of the radius of the slice multiplied by
the thickness of the slice - here taken to be a finger unit or 2.25 cm. The calculated volume does not include the neck block that is assumed to be 2
finger units thick (4.5 cm).
The volume of the lower part of the bowl is a quarter of a sphere so is easily determined from the radius of 8 units (or 18 cm) - i.e. 1.05 X radius
cubed or 1.05 x 18 x 18 x 18 = 6,107 cm³.
So to cut a long story short, the volume of the lower part of the bowl works out to be 6,107 c.c. and the upper part 10, 524 c.c. for a total volume
of 16,649 cubic centimetres.
The sound hole area is 3.142 x radius squared or 3.142 x 4.5 x 4.5 = 63.6 cm²
The 'equivalent length' of the sound hole is the sound board thickness + 1.7 X soundhole radius. Assuming a sound board thickness 2mm (or .02 cm)
gives an equivalent length of 0.02 + 1.7 x 4.5 = 7.75.
These figures give a calculated Helmholtz resonance frequency of the bowl cavity/soundhole diameter of about 122 Hz.
Assuming an arbitrary pitch standard of A415 (a semitone below modern concert pitch A440) the tuning of the individual courses a fourth apart is - as
previously posted - f' 329 Hz,
c' 247 Hz, g 185 Hz and d 139 Hz. However, if the pitch standard assumed is three semitones below A440 (i.e. A370) then the fourth course pitch is 122
Hz.
So what does this all mean?
This frequency happens to coincide with the calculated Helmholtz resonance frequency of the bowl and so will result in a volume amplification of the
sound of the thicker (less resonant) fourth course.
Interesting!
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
"Die Guitar und Ihr Bau" by Franz Jahnel was published in 1963 and covers the technology and design of the guitar, lute, mandoline and other fretted
instruments as well as string technology applicable to these instruments including gut and silk strings (Section F, part III).
Herr Jahnel mentions that in modern times (from the late 19th C) plain silk strings have been treated with chemicals to dissolve the outer silk
filaments of a string to make the string smooth (and more durable by protecting the inner layers against dirt and abrasion?).
(Silk protein (fibroin) will dissolve in strong acid or alkali solutions from which it can then be reconstituted - a method used today for spinning
fine filaments of silk for construction of bioengineered materials used for medical applications).
The C.A. Müller Company of Markneukirchen, Saxony patented a process in 1925 where the silk filaments of a strings would be soaked in a solution of
dissolved silk fibroin prior to being twisting into a string and dried at 115°C.
At the end of this brief article on plain silk string technology, the author mentions that during this period silk strings were being made smooth and
transparent by many simpler methods that nobody thought worthy of patenting. Silk strings prepared in these ways were made in Vogtlande and
Schönbach, Bohemia. Production of these strings ceased by the 1950's after plastic strings became available. "They appear from time to time as e' and
a' strings for violins but are not really recommended".
So here - quite likely - is the source of the dreaded silk 'Acribelle' violin strings mentioned earlier in this thread.
Unfortunately, the method of manufacture still remains unknown but is probably similar to the Japanese fishing leader product mentioned earlier.
The moral here is if you do not want people to copy your invention or secret process do not patent it!
The possibility of using a reconstituted silk protein solution rather than glue as a binder for plain silk strings might be worth investigating
further.
|
|
|
narciso
Oud Addict
Posts: 27
Registered: 11-15-2012
Member Is Offline
Mood: caleta
|
|
These are fascinating postings! Thanks!
Just wondering about the last comment above:
Reconstituted silk, after setting, presumably does not have the rubber-like elastic modulus of the traditional binding agents gum arabica or animal
glues. I wonder if that is an issue wrt playability concerns for a lute/guitar string?
Surface tension of the silk dope will also be a factor during spinning, affecting to what extent it can be drawn in air or in a liquid solvent as can
the traditional binding agents.
Anyway, after reading through your remarkable discussion I couldn't resist a bit of kitchen experimenting to see if I could get a basic
string-binding effect using supermarket gelatin, the most readily available protein glue these days!
I twisted together filaments of nylon sewing thread and drew them through a tub of gelatin kept warm in the liquid state (about 60C). I found I
could get a surprisingly uniform coverage, fixing the twist nicely as the gelatin cools.
I suppose the gelatin surface tension must be in some way optimal, along with the conveniently very short open time, (under 30s in my case) for this
to be at all manually feasible
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
Good for you narciso! I would encourage others to follow your example, experiment with making strings and report back their experience on this forum
for information.
Food grade gelatin (made from bones) can be used as a glue similar to (but weaker than) conventional woodworkers hide glue (also a gelatin) - so could
also be the basis of a binder (I have experimented with Knox brand gelatin as a binder ingredient). One advantage of pure food grade gelatin is that
it is crystal clear - although the coloured material (Jello brand) would also be OK for experimentation.
It is important that the binder completely saturates the filaments of a string rather than just the outer strands. This might best be achieved by
soaking a twisted string for some period of time in the hot binder solution and then squeezing the string through a sizing die of appropriate diameter
to produce a cylindrical string of uniform diameter. The die might be cooled to immediately chill the binder as it passes through. A non uniform
string would likely be false in intonation when stopped on a finger board.
I have no experience yet with reconstituted silk but I do not plan to spin the material into a thread (that requires technology beyond my means) only
to saturate the filaments with it as a binder in place of glue.
Hide glues (and gum arabic) are hard and brittle when cured but - as I have mentioned earlier in this thread, for hide glue at least - can be treated
to be 'rubbery' in consistency if required. Note that the silk filament itself (like gut) has rather poor elasticity (i.e. once stretched does not
return to its original state) so elasticity of the binder may not be a critical factor. Indeed the ancient Chinese silk qin strings were made by using
the natural gum coating (sericin) of the silk filament as the binder - and sericin is quite brittle.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
It might be worth mentioning that 100% pure silk fibroin is readily and widely available in a purified fine powder form (silk powder!) that is soluble
in water (and other solvents no doubt?). It is used as an additive for cosmetics and soap (makes the skin feel soft and 'silky' apparently) so should
be relatively safe to use.
It does not seem to be a very expensive product (and for string making only small quantities per string are required). This - as a convenient starting
point for making a binder - might save a lot of trouble and experimentation in avoiding the preparation of silk in solution prepared from raw silk.
From an historical perspective, if dissolved silk can be used as a binder for silk strings (yet to be demonstrated) then - as the chemicals required
to dissolve the silk were well known to the ancients (e.g. wood ash dissolved in water - known as lye) - it is within the bounds of possiblity that
they might also have made their silk strings with this kind of binder (although there is no historical record confirming that they did).
Reconstituted silk can be made from low value silk wastes - silk that otherwise cannot be used for woven textiles - so is commercially important as
the percentage of silk waste in the industry is significant.
For these trials I plan to make reconstituted silk from discarded, worn out silk clothing fabric as well as to test silk powder.
|
|
|
narciso
Oud Addict
Posts: 27
Registered: 11-15-2012
Member Is Offline
Mood: caleta
|
|
Looking forward to the results of your trials!
Curiosity piqued, I browsed around a little for accessible science literature to back up the idea that (at least in this context) reconstituted silk
can be manipulated like animal glue or a gum solution:
There certainly do seem to exist quite extensive molecular-level similarities between silk fibroins and collagen (i.e., the protein component of bone
glue).
Both have very long amino-acid sequences which assemble in solution into supramolecular fibrils.
The fibrils in both cases can apparently be readily coaxed into a similar gel form by controlling pH, salinity, temperature.
Moreover, a lot of the literature seems to be touting biomedical application of reconstituted silk specifically as a suture-forming material, so
very much a sort of ersatz collagen
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
Note that for bioengineering applications silk sutures and film must be free of any sericin gum as this will cause problems due to contamination of
organic tissue that it is in contact with. Not a problem for strings, of course, unless the sericin prevents proper adhesion of the reconstituted silk
- in which case the string filaments must first be degummed.
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
I have a quantity of worn out silk fabric (courtesy my wife) to experiment with and , hopefully, find good use for as reconstituted silk for string
making. The label attached to the 'donated' fabric declares that the material is 100% silk. My wife tells me that she thinks it is raw silk but I have
some doubts about that as the material has been dyed a buff colour. So could it be 'wild' silk ie not from the domesticated Bombyx Mori species? It
would seem from further preliminary research that the species of silk moth does matter when it comes to dissolving silk with chemicals.
The first step is to determine if scrap silk fabric is indeed real silk rather than a synthetic plastic 'look alike' material such as Rayon or
Nylon.
There are a couple of established tests to determine real silk.
The first is the 'burn test'. An open flame apllied to a silk thread will cause the thread to burn until the flame is withdrawn. The end of the thread
will form a brittle ball of ash - see attached image.
A chemical test is to use a so called 'silk reagent' in which a thread of Bombyx Mori silk will dissolve in a matter of minutes. Apparently this test
also distinguishes domesticated from wild silk - the wild silk presumably not dissolving (or at least not so quickly?)
The silk reagent is made from chemicals readily available 'off the shelf' from local pharmacists or hardware stores - chemicals that are,
nevertheless, not always non hazardous/non poisonous to use - so caution advised!
List of chemicals:
- Copper Sulphate crystals - from local pharmacy - poisonous.
- Sodium Hydroxide (caustic soda) pellets - from local hardware store for cleaning drains and general cleaning work - corrosive.
- Glycerin - from local pharmacy - non poisonous.
- distilled water - local harware store, for battery top up - non poisonous.
Dissolve 16 grams of Copper Sulphate in 150 cc water and add 8 grams Glycerin. Then add Sodium Hydroxide pellets slowly, bit by bit (this is a
strongly exothermic (heat generating) reaction so ALWAYS add the sodium hydroxide to the solution NEVER add the solution to the sodium hydroxide). A
light blue precipitate will first be formed. Continue adding the pellets until the precipitate clears to a dark blue solution.
Time out - more to follow.
[file]26956[/file] [file]26958[/file]
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
.... to continue.
My knowledge of pre-university chemistry is now a dim and distant memory but the above procedure (but I am not sure and so am open to correction)
should first result in a precipitate of copper hydroxide and sodium sulphate and then - with the addition of more sodium hydroxide - to an alkaline
rich solution of copper hydroxide and sodium sulphate.
It should be mentioned here that sodium hydroxide even at room temperature is a serious corrosive hazard, permanently destructive to the skin and eyes
- so use plastic gloves and eye protection when handling it - and work with small quantities at a time. Nothing to be frightened of - just be aware of
the dangers and act accordingly. Heating the stuff to a higher temperature increases the hazard - don't even try it!
Testing samples of my wife's silk fabric in the solution. The material did completely dissolve but not in a matter of minutes - more like about an
hour or so at room temperature.
Other known silk samples of Bombyx Mori silk - dyed and sericin coated - also completely dissolved in an hour or so rather than minutes.
So, it would seem that this 'silk reagent' will completely dissolve pretty well any silk - domesticated, wild, raw or degummed and dyed - at room
temperature in a relatively short period of time.
So - is my wife's silk fabric under test Bombyx Mori or some other 'wild' species of silk? It did not dissolve in a matter of 2 to 5 minutes so
perhaps it is 'wild' silk rather than 'raw' (Bombyx Mori cultivated) silk? On the other hand, perhaps the fact that the material has been dyed
increases the silk solubility time - the dye initially protecting the silk fibroin from attack by the reagent?
More to follow
|
|
|
jdowning
Oud Junkie
Posts: 3485
Registered: 8-2-2006
Location: Ontario, Canada
Member Is Offline
Mood: No Mood
|
|
The copper hydroxide reagent appears also to contain fine particles of metallic copper? (at least it cannot be gold unfortunately!) - see the
previously posted image where a couple of groups of particles can be seen floating on the surface (the remainder being at the bottom of the
container). So perhaps the reaction between the sodium hydroxide and copper sulphate is not quite as straightforward as first thought - and what is
the purpose of the glycerin?
Checking my antiquated set of 'Thorpe's Dictionary of Applied Chemistry' Vol. 5, page 93 provided some further information.
It is stated that "an interesting property of degummed silk (Bombyx Mori) is the ease with which it can be regenerated from solution in various
reagents without substantial degradation" and "Solutions of fibroin have also been recommended for the purpose of giving a silk like finish to cotton
goods"
Among the list of suitable solvents given that work at room temperature is an alkaline solution of Copper Hydroxide and also a concentrated solution
of Calcium Chloride (more on the latter later). The alkaline copper hydroxide solvent would appear to be rather old technology dating from the early
1930's in Japan.
As for glycerin (or glycerol) it is stated that the "addition of compounds such as glycerol or glucose results in a great increase in the rate of
solution of fibroin in alkaline copper hydroxide" (but doesn't explain why this should be the case).
Nevertheless, the alkaline copper hydroxide solution (if that is what I have made) seems to work in dissolving the scrap silk that I have to hand so
an attempt will next be made to produce a small quantity of reconstituted silk fibroin using this solvent - prepared as previously posted.
|
|
|
narciso
Oud Addict
Posts: 27
Registered: 11-15-2012
Member Is Offline
Mood: caleta
|
|
Conventional wisdom floating around the web has it that silk fabrics should never be washed with biological detergents (since silk is proteinaceous,
it is subject to degradation by proteases). Would that be viable as a gentler method I wonder? Although I suppose you'd be looking at a timescale of
weeks/months, rather than the matter of minutes you report...
Perhaps one could even 'train up' a crack strain of proteolytically active microbes specifically for the purpose !
|
|
|
| Pages:
1
..
6
7
8
9
10
..
13 |
![[*]](images/xmbprobrown/default_icon.gif){kind=icon}
