Wednesday, 22 April 2015

Merewen's Pentahlon Documentation - Part 3 of 5 (Face Bead)

Reproduction of Facial Mosaic Cane and the Bead on which it is Found

Glass Work (Group IV) Intermediate

Merewen de Sweynesheie

Goal: To recreate a face bead made using my handmade mosaic cane.  I have never made murrini anywhere near this complex before, and wanted to challenge myself.  I had also never made murrini that wasn't round and had never made murrini that required multi-step piecing before.

1.      History and Making of Millefiori
2.      Examination of Extant Bead and Historical Materials
3.      Choice of Materials
4.      Creation of the Canes
5.      Application to the Bead
6.      Discussion of Results
7.      Troubleshooting/Understanding Possible Issues
8.      References

1. History and Making of Millefiori

Photo by Master Eirik Andersen.
Millefiori, or "thousand flowers" were likely invented in western Asia.  The technique has been used for thousands of years to create tiny designs in glass, starting from at least 300 B.C.

Millefiori (murrini, if not flower-shaped, or mosaic cane, if very detailed) are often made by stacking together rods of glass in patterns and melting these rods together.    Once the rods are melted together, the very thick resulting glass bar is heated and stretched out to form a much smaller rod of glass, with the pattern reduced in size in its cross-section.  The rod can then be sliced and pierced for use as a bead directly, or the slice can be applied to a core-formed bead in the flame to form a very complicated bead. This allows for very detailed designs in very small spaces.  For a smaller design, the resulting rod can end up very long, sometimes requiring two people to run in opposite directions to achieve the desired size, as the rod cools and hardens as it is pulled.  Since the glass stretches more when it it hot, pulling must start slowly and speed up as time progresses.  Smaller lengths of murrini can be created by "painting"/layering different colours of molten glass on the end of a ponti, or metal/glass rod.

2. Examination of Extant Bead and Historical Materials
Photo by THL Lassarfhina ingean Uilleag.
Origin - precise dating/location unknown; Phoenecian to Islamic period, 400 B.C. to 1400 A. D.

Location found - Page 60, The History of Beads from 100,000 B.C. to the Present, by Lois Sherr Dubin.  The beads on the page are said to be from Lois' private collection.

Design - Bright green bead with a wide, darker (black?) stripe around the center.  Within the stripe, mosaic cane designs alternate:  a woman's face in white, black, and red, and a flower in white, red, yellow, and black.  Facial mosaic cane appears to be at least partially made by "painting" molten glass onto a base, although stacking rods could have been used for eye and nose area.  Flower mosaic cane has been made using an optic mold to form the flower shape.  Some of the other beads in this collection have women with necklaces on them.  This one didn't.  The necklaces would have been created by layering a layer of black glass over top of a yellow glass rod, pulling to stringer thickness (1-3 mm thickness), cutting this stringer into sections, and then laying these sections side-by-side on the mosaic below the face during creation.

Size - Extant bead is 1.2 cm in diameter

Heat source - Originally a wood or charcoal-burning furnace would be used to melt the glass. (example to right)  This would require a second person to pump the bellows to keep the flames hot.

Shaping tools - Anything could have been used to shape the glass, from metal tools to stone and wet wood.  Whatever won't burn.

Bead release - A clay-based mixture would be used, depending on local availability of materials.  This would be applied to the mandrel to allow the bead to be removed from the mandrel after it cools.

Glass - A bead maker could make their own glass, reuse glass made into items by others, or buy imported glass.  All methods are equally period.  As for the chemical makeup of the glass, the formulas were very similar to today.  Soft glass (soda lime) would have been the preferred type of glass for beads because it is easier to work in low heat and under more conditions.

Annealing - Glass could be annealed in a tended fire, or was left unannealed.  Annealing is the process of bringing glass up to just below its melting temperature for a length of time dependent on the size of the piece, then slowly cooling past the stress point temperature.  This allows the stresses between glass molecules to relax and not re-form during cooling, creating a more durable piece.

Punties - punties could be made of anything that wouldn't burn and to which the glass would stick.  Punties are the items to which glass that you are working on can be stuck.  This keeps hands away from the flame, and can be especially helpful if the glass being used in the piece is very soft with small amounts of heat, like white is.

Mandrels - These could have been made of a metal that wouldn't melt at the temperatures used.  Mandrels are essentially the "hole" of the bead; they are what the glass bead is formed around.  Removing the mandrel after cooling creates the hole.

Moulds - For the type of flower murrini in this bead, an optic mould would have been used.  This can be one of two types - either  flat piece of some kind of material (modernly graphite or metal, but in period metal or wood could have been used) with a carved indentation the size and shape desired or a set of vertical rods designed to restrict the glass' spread in certain locations.  For both, the molten glass on the punty would be pushed into the mould to form the desired shape.  The glass would then be pulled out and more glass would be applied around the outside to fill out the design.

3. Choice of Materials

Photo by Master Eirik Andersen.
a)      Glass/Colours:  I use soft glass for beadmaking, which is consistent with period recipes.  I purchase this glass because of the dangers or the chemicals and heat, and specialist equipment needed.
    Red - a bright, opaque red was needed that would not discolour too easily with working.  I chose Effetre Pastel Red because of its bright colour.  It does have a tendency to bleed into other colours, but only tends to do so if worked heavily and hot while touching certain whites and other lighter colours.
    White - a solid, nonreactive white was required.  I used Effetre White because it would not react with the black that I chose.
    Black - a black that will stay looking black when stretched very thin was required to keep the detail in the murrini.  Creation is Messy Hades fit that bill.
    Yellow - many yellows on the market have a level of transparency or translucency, or will react with many other colours to form brown.  Effetre Butter Yellow had the right colour and opacity, and would not react heavily with the other colours I was using.
    Green - A bright green was needed.  Most bright greens in today's colour palette have a tendency to spread out over top of other colours, so I tried Effetre Grass Green, but also looked at other shades of green such as Effetre Cave Green and Creation is Messy Olive.

b) Heat source:  Nortel Mega Minor torch, with oxygen concentrator and propane.  I have tried the furnace method before, but working alone as I do, it was not feasible for this project.  In addition, the small homemade furnaces that people generally work with in the SCA are not likely to produce enough heat to melt the 1 inch by 1 inch chunk of glass that I would be working with, especially in the ambient temperatures in which I was working.  My torch with added oxygen barely managed to produce enough heat to successfully pull this cane.

c) Bead release:  Fusion bead release, chosen because it allows for flame drying or air drying.  Contains clay, graphite, and a mixture of other binders and other components.

d) Mandrels:  1/16" steel TIG welding rod, 308L.  Cheap and readily available.  A common type of mandrel used by bead makers.

e) Annealing: Glass Hive kiln.  I prefer to anneal my beads for durability, and, given the size and complexity of the face mosaic cane, I wanted to anneal that as well.

f) Shaping tools:  Brass Stump Shaper (created by Loren Stump, a modern master of detailed mosaic cane creation), brass stylus, hot fingers (when I dropped the glass when the punty broke - it holds a hot glass piece in "claws"), marble and graphite marvers (flat plates), steel chopsticks for pulling stringer.  The Stump Shaper is a slightly shaped rectangular brass paddle with a handle, that can be used to manipulate glass. 

g) Punties:  Lauscha Reformulated Clear glass rods in 8-10 mm size.  The bigger the punty for this size work, the easier it is to control and pull.  I use soft glass punties as opposed to metal or borosilicate punties because borosilicate must be sure to be completely removed, otherwise it can cause the glass to crack, and metal punties can detach upon cooling.  Soft glass punties do have two disadvantages - they can crack if cooled too much and reintroduced to the flame, and they can bend if they get too hot.  I chose clear glass because it is generally cheaper than coloured glass and has a higher heat tolerance before slumping than many other types of glass.  This did result in my unfinished murrini being dropped on the floor once.

4. Creation of the Canes

Photo by Master Eirik Andersen.
A.     Flower
I sat down to do the flowers with Effetre Red, Effetre White, Effetre Butter Yellow, and CiM Hades.  Here are the steps I followed:

a)      I started by making the white rod the center of the flower.
b)      I then coated the outside of this rod in red glass.
c)      To this I added alternating stripes of yellow and black using stringer around the outside.
d)      I melted these in a bit and shaped the core back closer to round, then added a second stripe of each on top of the first. 
e)      Then I melted back closer to round again.
f)       The petal shape would ideally be made in an optical mold, but I did not have one, nor the resources to buy one or make one, so I was trying to get the correct shape with basic tools.  To try to make the petals more petal-shaped, I tried indenting the black lines sharply with a brass stylus.  I then added another layer of black in the indents, and melted in slightly until I was close to round. 
g)      Next, I completed the layer of black by filling in any places where the yellow still showed.
h)      At this point, I melted more thoroughly and shaped the entire thing to round properly. 
i)        Finally, I added a layer of red around the outside, melted flat, and shaped to round one more time.
j)        I flattened the end and added a clear punty, then melted off the white rod, picked away the excess white, and added a second clear punty to the other side.  I did this because if I had left the white glass in the punty the murrini would have pulled unevenly because the white is too soft.

The resulting flower looked more like a poorly-made sun than a flower.  I decided to try again, and here were the changes I made to my procedure:

a)      Used a thick white stringer at the center.
b)      Coated the center white using red stringers.
c)      First layer was alternating yellow and black, second layer was two yellow to each black, third layer was two black to each yellow, and last layer was all black.

The result of this second attempt was a lot better than the first attempt in that it could conceivably look like a flower - just not quite the flower that was in the extant, which I would not be able to achieve accurately without an optical mold.

B.     Face
To make the face, I started by making eyes.  I went through three attempts before achieving eyes that I could use.  In the first eye cane, the pupil was not touching the black border, which was not accurate to the extant.  In the second cane, I used a reactive white (CiM Peace) to fill the eye, which turned silver in the finished cane and made the eye hard to see.  The third time, I used CiM Hades and Effetre White for the entire thing, and added the black lines on two sides of the square cane.  Steps for making the eyes (on third cane):

a)      Start with thick stringer of CiM Hades.  Add a line of Effetre White to either side using a stringer (180 degrees offset from one another).
b)      Use Hades stringer to coat the entire thing in a thin layer of black.
c)      Shape eye and flatten end.
d)      Add clear punty to flattened end and melt off black stringer.
e)      Pick off excess black stringer with needlenose pliers.
f)       Use white rod to add a thin layer of white all round, with two extra lines of white on the flatter sides.
g)      Shape to square.
h)      Use Hades stringer to add black to two sides of square, melt, and reshape to square.
i)        Add clear punty to other end.
j)        Carefully heat, reshaping back to square frequently.
k)      Pull.

I then worked on making the face.  My first attempt had a number of issues.  First, The punty broke and the chunk of glass fell on the floor.  I picked it up with hot fingers and got a new punty on.  The next issue was probably the biggest one - the largest punty I had was about 8 mm.  Since the chunk of glass was much larger than this, I was unable to get a big enough maria to pull evenly.  Despite the setbacks, the steps I took were as follows:

a)      Squeeze the end of a rod of Effetre White glass into a rectangle.
b)      Apply one eye to one side and one eye to the other.
c)      Apply a layer of black at the bottom of the nose.
d)      Apply a layer of white to the bottom of the whole piece.
e)      Apply a layer of red to the center of the bottom.
f)       Apply another layer of white, all the way around until the piece is round.
g)      Melt in and shape to round.
h)      Apply a few layers of white to the neck area.
i)        Shape, then coat with Hades.
j)        Apply another few layers of white for the décolletage and shape.
k)      Surround the whole thing with Hades and create the appropriate shape.
l)        Attach a punty to the second side.
m)   Thoroughly heat, then pull.

I tried this a second time to try to fix some of the design issues I had with my first time.  I did have a different issue this time, in that the head got twisted at one end before the application of the neck.  This meant that the image was mostly correct on one side, and the woman's head had an extreme sideways tilt at the other end.  These are the changes I made to my procedure:

a)      Applied one eye to the (clear) punty, then applied white for nose and second eye.
b)      Added two layers of white between the bottom of the nose and the lips.
c)      Did not apply white to the side of the eye area, only the top and bottom of the face.
d)      Tried to shape the décolletage more accurately.
e)      Did not add multiple layers of black at the bottom.
f)       Used two punties at each end to pull.

The resulting murrini pulled much more evenly.  Total time (per attempt) to complete one eye cane:  1-1.5 hours.  Total time (per attempt) to complete face mosaic cane using completed eye cane - 2 hours.

5. Application to the Bead

Photo by THL Lassarfhina ingean Uilleag.

Cutting the murrini - I had only one method available to me to cut the murrini, and that was with disc nippers.  Disc nippers, as their name implies, nip off a small chunk of glass.  Because you do not typically pre-score the glass with this tool, pieces can tend to chip off very unevenly.  Most professional murrini makers in our age will use a saw and flat lap machine to cut the murrini flat and smooth and even out the faces, but I do not have these tools.  The result of this was that some murrini were more uneven than others, some thicker than others, and some very thin.

Base bead colours - In making the bead, I first had to decide what glass to use in the base and the stripe.  I resolved to attempt three different greens, which were closest in colour to the picture of the extant:  Effetre 216 (grass green), Effetre Cave Green, and CiM Olive.  For the stripe, it looked black (I used Effetre black), but I wanted to try a slightly lighter colour so that the face wouldn't blend into its background, so I also tried CiM Adamantium, which is a darkish grey/brown.

Forming the stripe - I first tried making a smooth stripe by applying a large rod of black around a cave green base bead, but the stripe was very uneven and jagged, so I decided to apply multiple side-by-side lines of thick stringer around the bead instead.  This worked much better.  Alternately, I could have formed a base bead in the stripe colour and then added the "base bead colour" around the mandrel on the two sides of this.

Applying the murrini - When applying the murrini, I preheated the murrini by holding it on a brass tool beside the flame before using it.  I had no problem applying the final face murrini, but for some reason, I could not seem to apply the 1st attempt face murrini without the murrini cracking, even when the murrini had been preheated so hot that the edges were glowing.  I did have one problem with the (final) face murrini cracking in half on my final bead, but both pieces were still attached to the bead, almost touching each other, so I simply used my needle nosed pliers to push the halves back together.

The testing pattern was as follows:

a)      Grass green base with black stripe using full rod:  I first tried to make the black stripe using a full rod of glass.  Unfortunately, this ended up in very uneven application, so I gave up on the bead, added yellow dots, and decided to apply the stripe using thick stringers from then on.
b)      Cave green base with black stripe:  Made a donut-shaped bead and applied side-by-side stripes of black with a very thick stringer.  Did not melt stripe in all the way before adding murrini.  Not bad, but the black stripe was as wide as the visual area of the bead, blocking out the green ends.
c)      Cave green base with adamantium stripe:  Made a cylinder of cave green, applied side-by-side stripes of adamantium, melted in fully and reshaped to cylinder before applying murrini.  Murrini became very noticeably stretched out vertically from hole to hole and the adamantium and cave green did not appear to form a very defined line.  The adamantium actually expanded over the cave green.  These murrini had also been very thick, which may have contributed to the level of smearing of the design.  Either way, I decided not to use adamantium again, not to melt the stripe in fully before murrini application, and not to shape into a cylinder before applying the murrini.
d)      Grass green with black stripe:  This bead never made it as far as applying murrini.  Grass green has a known tendency to spread, and this bead was no exception.  The grass green "swallowed" the black stripe, and no amount of adding black made the stripe stay in sufficient quantity at the surface.  I added yellow dots and gave up.
e)      Olive bead with black stripe:  This time I made the base bead of olive into a cylinder and applied the black stripe.  I did not melt the stripe in fully, and did not reshape the bead into a cylinder after applying the stripe.  I applied the murrini and fully melted in the murrini.  There appeared to be a good division between the black and the olive green, but the face became slightly smeared.

6.  Discussion of Results

Photo by THL Lassarfhina ingean Uilleag.
Although I did not achieve a perfect replica, I believe my attempts to be good for the level of experience I have in this kind of work.  I ran out of time to repeat the work, explore further, and try to improve my results.

Murrini -
My second face murrini had the right eyes, and you could even see the pupil if you looked really really closely.  The shape of the décolletage was a little bit off, and the face was slightly tilted, but I felt it was a good attempt for this having been my first time doing something this complicated.  The size of this murrini and shape was much improved from the first.

The second flower murrini were much improved from the first, but still would have required the use of an optical mould to accurately copy the flower design of the extant.  From use, I discovered that the flower murrini must have had a black border to prevent spread of the red. Red does have a tendency to bleed into other colours, I just didn't realize that this would happen with the black - usually white is the biggest concern.

Beads - My beads tended to be a little bit larger than the extant - in the 15 mm range.  Fairly close, overall.  The ends of the beads were not perfectly puckered, but I was focusing on the rest of the design and letting that part suffer for achieving better results in everything else.

My green colour on the best beads ended up not being as bright as the extant, but I did not have an appropriate green colour to use that would not swallow the black stripe, so the olive and cave green were the best options I had.  I do not know what green would have been used to achieve the results in the extant, as most modern greens of that bright shade spread somewhat similarly to grass green.

The best beads seemed to be (b) and (e), for different reasons.  (e) had a better stripe and base bead definition, while (b) had much clearer face mosaic cane application.

7. Troubleshooting/Understanding Possible Issues

Photo by THL Lassarfhina ingean Uilleag.
Murrini creation is subject to a number of potential problems:

a)      Working with colours that react or bleed into one another - final design colours are not as intended
b)      Space between glass layers - air bubbles that can cause shattering of rod upon cooling
c)      Heat glass too hot while working - burns or boils, resulting in scum or air bubbles in the finished murrini
d)      Shaping colours incorrectly - once hot glass is adhered to other hot glass, the mistake can be difficult to correct
e)      Using colours too close in tones to one another - it will be difficult to discern the details of the design
f)       Using colours who have very similar working colours when hot - it can be hard to tell where the colours have been applied
g)      Making layers too thin - colour will pull thinner and not show up in final design
h)      Not heating enough before pulling the rod - murrini will not reach desired size before rod hardens too much
i)        Pulling too soon after removal from heat - murrini will be pulled to too small a size
j)        Heating or pulling unevenly - rod will vary in thickness over its length
k)      Not using a large enough punty to cover the surface fully - murrini will pull unevenly and there may be less or no useable material
l)        Murrini twists while building - details may be inaccurate at one end
m)   Murrini twists while pulling - resulting rod could be unusable for building into larger murrini
n)      Cut murrini too thinly - chips and cuts unevenly
o)      Cut murrini too thickly - difficult to apply to bead without shocking and breaking

Application of the murrini to the bead is also a delicate task.  Problems that can occur include:

a)      Heating the murrini too quickly, either before application or soon after applying to a bead - murrini pop off bead and/or shatter 
b)      Applying too little heat before application or soon after - murrini fall off of the bead
c)      Applying too much heat after the murrini is applied - smearing or mutation of the design
d)      Applying too little heat after application - murrini doesn't fully melt into bead
e)      Cooling the murrini face and pushing too hard - design expands and distorts
f)       Not cooling the murrini face while melting in - murrini design distorts

8. References

I do not have so many references at this point because I have learned the majority of the information I used in my work through experience, and learned a great deal of the history lampworking too long ago to remember the source.

Dubin, Lois Sherr, The History of Beads from 100,000 B.C. to the Present, 2009.  Abrams, New York, NY, USA.  Pages 60-61.  (Location of bead attempted)

Hornick, Sarah, 2013.  Bead making class in Chicago, Il.  (Use of silver and creation and raised application of murrini)

Liu, Robert K., Nubian Mosaic Face Beads, May 2014.  Ornament Magazine.  Accessed April 2015, from

No comments:

Post a Comment