Upload
thomas-arnold
View
18
Download
0
Tags:
Embed Size (px)
Citation preview
OPTIMISING CORE-CLAD RATIOS IN GLASS EXTRUSIONS FOR OPTICAL FIBRE APPLICATIONSBy Thomas Arnold
4th Year MEng (Materials stream)
Individual project MM4MPR
• Introduction to optical fibres
• Problem statement and objectives
• Overview of previous projects
• Polymer extrusion
• Borosilicate extrusion
• Analysis of results
• Conclusions
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Outline
Cladding (low refractive index)
Core (high refractive index)
Light
Cladding
Core Light pulse
Input signalOutput signal
Monomode optical fibre
Light redirected into the core.
• Difference in refractive index of cladding and core
• Total internal reflection
Monomode optical fibre
• A core-clad ratio of 60% is optimum
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
How do optical fibres work?
Multi to single mode…
Ebendorff-Heidepriem, Heike, and Tanya M. Monro. "Analysis of glass flow during extrusion of optical fibre preforms." Optical Materials Express 2.3 (2012)
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
CoreClad
Die
Bobbin
Pu
nch
Flow of charges
Extrusion
Core charge
Clad charge
Extrusion of the preform(1) (2)
Stable section cut from the extrudate, where the change in cladding thickness is minimal.
Total cross sectional area
Clad cross sectional area
Core-clad ratio (%) = Core area/Total area
Extrudate (3) Preform
Above Tg, the glass charges become viscous liquids and the core charge is forced into the clad charge.
Bhowmick, K., Morvan, H. P., Furniss, D., Seddon, A. B., & Benson, T. M. “Co‐Extrusion of Multilayer Glass Fibre‐Optic Preforms: Prediction of Layer Dimensions in the Extrudate.” Journal of the American Ceramic Society (2013)
~Ø5 mm
Problem statement
• Monomode optical fibres researched at the University of Nottingham are manufactured via extrusion.• This provides a ‘preform’ which is drawn again
to give the final fibre length and diameter.
• Methods for optimising the core clad-ratio of the preform is not fully understood.• This is largely due to the difficulty associated
with analysis of chalcogenide preforms and the expense in carrying out these extrusions.
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Seddon, A.B. “Chalcogenide glasses: a review of their preparation, properties and applications.” Journal of Non-Crystalline Solids, 1995
Project objective• Find a suitable material to
replicate the extrusion process – clearly showing core and clad flow patterns.
• Perform two stack and six stack extrusions to understand the flow patterns of the core during extrusion.
• Post analysis of extrusions to determine a method for optimising the core-clad ratio (achieve 60%).
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Extruder set-up• Extrusion of the
preform is carried out in a controlled environment The
assembly
Billet Load cell
Furnace
Punch
Cooling coils
Bobbin
BarrelBarrel lining
T/C
Core charge
Extrusion of stack through a Ø4.76 mm die.
Clad charge
Core chargeClad chargeDie
(1) Barrel and die setup
(2) Extrusion flow
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Furniss, D., Glass Extruder - Operating manual, University of Nottingham
Previous work – chalcogenide glassesTwo layer extrusion
• Excellent light guiding characteristics.• Commonly used for monomode optical fibres.
But,• Very expensive – limiting experimental work
which can be carried out.
30 50 70 90 110 130 150 170 190 210 2300
20
40
60
80
100
W.H.C (94.03%)
S.D.S (89.44%)
Length along section/mm
Cor
e-cl
ad r
atio
/% A
rea
Three plots of “core-clad ratio vs. length along section” from previous work
Composition, Tg and viscosity effect the curve.
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Savage, S. D., Miller, C. A., Furniss, D., & Seddon, A. B. “Extrusion of chalcogenide glass preforms and drawing to multimode optical fibres.” Journal of Non-Crystalline Solids, (2008).Choong, W.C., Seddon, A.B., “Monomode Mid-Infrared Fibre Optics.” University of Nottingham MM4MPR Paper. 2012/13.
S.D Savage W.H Choong
Composition of core Ge17As18Se65 As40Se60
Composition of clad
Ge17As18Se62S3 Ge10As21.4Se66.6
Viscosity (core/clad) /Pa.s
107.2/Not known(320°C)
107.4/107.8 (230°C)
Difference in Tg/°C
35 6
Replicating the extrusion
• Due to high cost of chalcogenide glasses, an alternative low cost glass was to be used for this project.
• Initially, a polymer material was selected.
• Following problems with the polymer extrusion, an oxide glass was melted for the extrusions in this project.
Polymer Tg/°C Tm/°C CTE/10-6°C-
1
Optical transmissibility
PMMA 100 - 122 250 - 260 70 -150 Excellent
PS 92 - 100 240 - 260 10 – 150 Excellent
PVC 80 180 - 210 50 Good
Selecting the polymer…
Optical quality necessary for post extrusion analysis.
Non toxic and heat resistant
Optimising core-clad ratios, Thomas Arnold, 15/05/2014Tsao, Chia-Wen, and Don L. DeVoe. "Bonding of thermoplastic polymer microfluidics." Microfluidics and Nanofluidics (2009)
PMMA extrusion – two stack
• The extrusion was carried out at 137°C (~15°C above Tg).
• The extrudate showed significant die swell.
• We were unable to mitigate this problem, thought to be due to the thermal expansion and elastic characteristics of polymers.20.69 m
m
Die (Ø4.76 mm)
Bobbin
PMMAcharges
Barrel
Applied force from punch
Furnace
Heat
10 m
m
~Ø14 mm
Extrusion of stack (24 mm in height) through a Ø4.76 mm die.
2 charges with identical geometry
The samples – transparent and blue charges
Extrusion set-up
The resultant extrudate
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Den Doelder, C. F. J., and R. J. Koopmans. "The effect of molar mass distribution on extrudate swell of linear polymers." Journal of Non-Newtonian Fluid Mechanics (2008)
PMMA extrusion – six stack
20 mm
9 mm
Thickness/mm
1.97 2.10 1.90 2.11 1.93 1.99
100 105 110 115 120 125 130 135 140
Clear PMMA
Blue PMMA
Temperature/°C
Vis
cosi
ty/P
a.s/
m 4×107
4×106
4×105
The resultant extrudate• Die swell, contraction and
amalgamation of the PMMA layers was observed.
• Replication of fibre optic extrusion was not achieved.
Viscosity graph for the transparent and coloured PMMA samples
• Viscosity measurements showed clear differences between Tg values.
• Factor in the failure of the PMMA extrusion.
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Ebendorff-Heidepriem, H., Monro, T. M., van Eijkelenborg, M. A “Extruded high-NA microstructured polymer optical fibre.”(2007).
Oxide glass – selection and batching
• Borosilicate was selected because of its high water solubility – allowing simple cleaning of components.
• High borate content – water solubility.
• Silicate – strengthening of the glass.
• Sodium dioxide – network modifier, to encourage chemical bonding between the borate and silicate.
Molecular structure of borosilicate
Borate
molecules
Sodium dioxide atoms
Oxygen atomsSilicon atoms
Boron atom
Sil
icat
e m
olec
ule
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Glass samples melted
• 25g Batches of borosilicate were prepared with the composition: 5Na2O-75B2O3-20SiO2.
• This was divided into two melts, the second melt having 0.5wt% cobalt oxide added, giving a blue colour.
• Viscosity between glass melts was matched.
Manara, D., A. Grandjean, and D. R. Neuville. "Advances in understanding the structure of borosilicate glasses: A Raman spectroscopy study." American Mineralogist (2009)
Borosilicate extrusion – two stack
• The extrusion was carried out at a temperature of 535°C (above the Tg of borosilicate).
• The transparent borosilicate was the cladding charge and the cobalt oxide doped (blue) glass the core.
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
15 sections cut as shown for post processing
20 60 100 140 1800
20
40
60
80
100
W.H.C (94.03%)
S.D.S (89.44%)
Borosillicate
Length along section (mm)
Cor
e-cl
ad r
atio
/% A
rea
Two stack extrudate and cross sections
Core-clad ratio of borosilicate extrusion vs length along section, including data from previous studies for comparison.
Borosilicate extrusion – six stack
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
• Transparent and cobalt oxide doped glass showed clear core entry.
• Core entry for cores 1 –
5 was observed at 5.3, 14.3, 24.5, 39.8 and 58.5 mm respectively
Multi-stack extrudate
Cross section showing core entries
Multi-stack arrangement
Core 1
Clad
Core 3
Core 2
Core 5
Core 4
Alternating colours were used to allow cores to be identified in the preform.Each charge was ~3 mm in height.
Results
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
0 20 40 60 80 1000
20
40
60
80
100Core 1
Core 2
Core 3
Core 4
Core 5
2 layer stack
Length along section/mm
Cor
e-cl
ad r
atio
/%
Core-clad ratio of six stack borosilicate extrusion vs length along section, including two stack borosilicate extrusion.
CoreCore
charge height/mm
Clad charge
height/mm
Core height percentage /
%
Stable core-clad
ratio /%1 15 3 83% 952 12 6 67% 893 9 9 50% 804 6 12 33% 685 3 15 17% 44
1 2 3 4 5
Core 4 Core 3 Core 2 Core 1
Core 5
Clad
Core
Actual stack Equivalent stack
Core 4 Core 3 Core 2 Core 1
Core 5
Clad
Core
Actual stack Equivalent stack
Core 4 Core 3 Core 2 Core 1
Core 5
Clad
Core
Actual stack Equivalent stack
Core 4 Core 3 Core 2 Core 1
Core 5
Clad
Core 4
Actual stack Equivalent stack
This equates to a stable core/clad ratio of ~67%
~6 mm
~12 mm
Core<Clad Core>Clad
Equal core and clad charge height
10 20 30 40 50 60 70 80 9040
50
60
70
80
90
100
0
10
20
30
40
50
60Core-clad ratio (Max)Polynomial (Core-clad ratio (Max))
Core charge height% of total stack (17.88 mm)
Cor
e-cl
ad r
atio
/% A
rea
Cor
e en
try
pos
itio
n/m
m
Further analysis A polynomial relationship exists between peak core-clad ratio and core entry with the absolute stack height.
From this, an exact core-clad ratio of 60% can be expected from a clad charge height of 12.6 mm and a core charge height of 5.4 mm
A stable length of ~16 mm can be interpolated from the table for a core-clad ratio of 60%
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Peak core-clad ratio and entry of core along section of six stack borosilicate extrusion vs absolute stack height
CoreMax core/clad ratio/%
Position/mm
Stable length/mm
1 95.3 5.3 ~60.0
2 88.7 14.3 ~40.0
3 80.4 24.5 ~35.0
4 67.6 39.8 ~20.0
5 43.6 58.5 ~7.5
Core Core height percentage /%
Stable core-clad ratio /%
1 83% 95
2 67% 89
3 50% 80
4 33% 68
5 17% 44
Preform to fibre optic
If a fibre preform of Ø4.76 mm with a stable region of Ø16 mm is drawn to a fibre with a final of Ø1 mm …
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
Core
Clad16 mm
370 mm
Preform
Final fibre
• Extrusions to understand how to optimise core-clad ratios can be successfully carried out using borosilicate glasses.
• A polynomial relationship exists between peak core-clad ratio and core entry with the absolute stack height.o Further experiments are required with charges of
different geometries to validate the polynomial relationship established in this project.
• From this, an exact core-clad ratio of 60%:o A clad charge height 70% of the overall
stack height (12.6 mm). o A core charge height 30% of the overall
stack height (5.4 mm)• A resultant preform length of ~16 mm giving a final
fibre length of 370 mm.
Conclusions
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
↑Clad charge (thickness) Core charge ↓ (thickness)
Position along extrudate
Cor
e-cl
ad r
atio
by
area
Observations• Equal viscosities of core and clad
charges, leading to a steeper gradient, was validated.
• A core-clad ratio of ~67.6% with a stable length of ~20 mm is achievable, with a clad charge height of ~12 mm and a core charge height of ~6 mm.
• Reduction in core-clad ratio for cores 4 and 5 was observed following peak core-clad ratio.
Optimising core-clad ratios, Thomas Arnold, 15/05/2014
20 60 100 140 1800
20
40
60
80
100
W.H.C (94.03%)S.D.S (89.44%)J.BBorosillicate
Length along section (mm)
Cor
e-cl
ad r
atio
/% A
rea
Core-clad ratio of borosilicate extrusion vs length along section, including data from previous studies for comparison.
0 20 40 60 80 1000
20
40
60
80
100
Length along section/mm
Cor
e-cl
ad r
atio
/%
Core-clad ratio of six stack borosilicate extrusion vs length along section, including two stack borosilicate extrusion.
Core chargeClad chargeDie
Static material of clad charge