Strength of Glass and Glass Fibers Hong Li, Ph.D. Fiber Glass Science and Technology PPG Industries, Inc. Pittsburgh, Pennsylvania, USA
76th Conference on Glass Problems Columbus, Ohio, USA November 2 – 5, 2015
Fracture of Solids Inglis (1913), Griffith (1920), Orowan (1949)
sf = (Ego/4ro)1/2 (ztip/C)1/2 Structural
Mechanical
go[SiO2]: about 2.0 J/m2 (high vacuum); 0.5-0.7 J/m2 (surface dominated by silanol group); 0.4 -0.45 J/m2 (surface saturated with H2O); 0.3-0.4 J/m2 (surface in contact with water)
Weibull – weakest link & defect population (1951)
P(s) = 1 – exp[-(s/so)b]
Mechanical
ZrO2
Young's Modulus of Glass (GPa)
98 K2O
Al 2O3
Na2O TiO 2
96
MgO SiO 2
94 BCaO 2O3
Li2O B2O3 CaO SiO 2
Li2O
92 MgO
Na2O
TiO 2
90
K2O Al 2O3
88
ZrO2
-6
-5
-4
-3
-2
-1
0
1
2
3
Xi of given oxide in glass (mol%)
4
5
6
Young's Modulus of cyrstalline materials, E (GPa)
Part 1. Influence of Chemistry and Local Structure of Glass Conradt, 2nd Inter. Fiber Glass Tech. Symp. (1994) Volf, Chemical Approach to Glass 550
SiO2 (Stishovite)
500 450
Al2O3
400 BeO
350
MgO SnO2
300
TiO2
250
SiO2 (Coesite)
200
2:
CaO
1
SrO
150
ZnO
100
SiO2 (q)
50 0
SiO2 (c)
BaO
0
20
40
60
80
100
120
140
160
180
Oxide coefficient in glass model, E i (GPa/mol)
Li, et. al. IJAGS (2014)
sf = (Ego/4ro)1/2 (ztip/C)1/2 • Strength is proportional to glass Young’s modulus • Local structure of glass plays a key role
200
220
Local Structure and Packing Density Effect R. Conradt, 2nd Inter. Fiber Glass Tech. Symp. (1994) 4.3
250
Stishovite (~550 GPa) Perovskite
Crystalline Phase
challenge/opportunity
Young's modulus (GPa)
Glassy Phase
Coesite 200
2.9
150
Ilmenite
2.6
Enstatite
-quartz 100
2.3 2.2 50
cristobalite
Glass and Fiber Glass
SiO2
AlSi3O8
CaAl2Si2O8
Polymorphs
Albite
Anorthite
CaMgSi2O6
Diopside
CaSiO3
Wollastonite
MgSiO3
Na(Al,FeIII)Si2O6
Polymorphs
Jadeite
•
Coordination of T-O and material packing density are important
•
Challenge: creation of high-pressure like structure of T-O in glass under ambient conditions
Batch-to-Melt Conversion of Na2O-Al2O3-SiO2 Composition Batch treatment time at 1600oC: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 min Alumina input: C – corundum ( – Al2O3) A – alumina spinel (g – Al2O3) B – boehmite (g – AlO(OH)) G – gibbsite (g – Al(OH)3)
27Al
MAS NMR spectra of batch-to-melt
Christmann, Deubener, IJAGS (2016)
Challenge of keep Al in [AlO6] Tetrahedral coordinated Al, [AlO4] in melt is thermodynamically favorable over octahedral coordinated Al, [AlO6] in crystalline phase of raw material.
High Performance by Design 25%
38%
Average
10% new R-glass
E-glass
Pristine fiber tensile strength
UD Composite Tensile Modulus
Glass network structure can be modified for higher mechanical performance without significant process penalty (fiber forming temperature and crystallization) [1] H. Li, US 8,901,020 B2, PPG Industries Ohio, Inc. (2014) [2] H. Li, P. Westbrook, US 20150018194 A1, PPG Industries Ohio, Inc. (2015) [3] H. Li, et al. IJAGS (2014) 6
Part 2. Impact of Hydrolysis on Glass Surface Energy Proctor, et. al. Proc. Roy Soc. A297 (1967) p. 534
~2.5X
~1.5X
Liquid N2 Vacuum, RT
1X Ambient condition, RT
sf = (Ego/4ro)1/2 (ztip/C)1/2 • Surface hydrolysis is detrimental to silica fiber strength
Stress-assisted Hydrolysis Brow, et. al. JMS (2015)
Change in surface energy at crack tip f (LN2) / f (RT-50%RH) = 2.2 – 2.3
sf = (Ego/4ro)1/2 (ztip/C)1/2
• Under stress-free conditions: formation of “immobile” Si-OH groups • No significant impact on crack growth • Under applied stress conditions: formation of “mobile” Si-OH groups • Significant impact on crack growth • Newly created surface at the tip of critical flaws has much lower energy resulting from its reaction with mobile water species.
Stress-Assisted Hydrolysis Effect on Fiber Failure Fe2O3: 0.04 – 0.64 wt% Aging under 50oC-80% RH: 0 – 180 days 5.0
Brow, et al. (2015)
4.0 3.5 3.0 2.5 2.0 0.0
0.2
0.1
0.3
0.4
0.6
0.5
0.7
Fe2O3 (wt%) CeO2
4.5 4.0
SnO 2
5.0
3.5 3.0
MnO 2
Surface energy ratio, goLN /goAIr (LN/50%RH)
4.5
2.5 2.0 0.0
0.1
0.2
0.3
FeO (wt%)
0.4
0.5
In calculation of modulus ratio at fiber perspective failure strains in liquid nitrogen and air, respectively, glass secant modulus as a function of fiber failure strain was estimated based on work by Gupta and Kurkjian (JNCS, 351, 2005).
Glass Defect Formation through Reaction with Moisture Water • Hydroxyl groups formed on glass surface is controlled by water diffusion, which are immobile once formed under stress-free conditions decrease strength by lowering glass surface energy • Aging increases surface roughness decrease strength by introducing more surface defects • Aging results in possible formation of alkali and alkaline earth carbonates originated from ion exchange with moisture water decrease strength by introducing more surface defects • Aging may result in tip blunting of pre-existing surface flaws increase strength by increasing time of initiation of critical surface flaw
How to explain a “constant” ratio of failure strains between silicate glass fibers tested with and without presence of moisture water? Stress-assisted hydrolysis dominates glass fracture
Part 3. Impact of Defects
R.E. Mould, (1967)
1x101 Theoretical strength
1x100
Pristine Glass (as drawn)
3x10-1
Pristine Glass (heat treated)
1x10-1
Formed Glass
3x10-2
Used Glass
Damaged Glass
Strength, sm (x106 PSI)
3x100
1x10-2 3x10-3 1x10-3 3x10-4 10-8
Inherent Flaws
10 Å
Structural Flaws
100 Å
10-7
0.1 m
Fabrication
Microscopic Damage
1m
10-4 10-6 10-5 Surface flaw, C (inch)
sf = (Ego/4ro)1/2 (ztip/C)1/2
Visible Damag e
10-3
10-2
Surface Defect and Population Fiber Diameter 1000
100 0.001 2
0.01
0.1
1
Defect Geometry
Fiber diameter (mm)
1
lnln[1/(1-P(N)]
Fiber tensile strength (MPa)
10000
Weibull Analysis
0 -1
P(s) = 1 – exp[-(s/so)b]
-2 -3 -4
Griffith, Trans Roy. Soc. (1920) 69.2SiO2-11.8Al2O3-12K20-0.9Na2O-4.5CaO-0.9MnO
4
5
6
7
ln (strength, s(N), MPa)
8
9
Part 4: Damage, Healing, Environment Mould, J. Am. Ceram. (1960)
N2 – RT - 0.3%RH
• • •
Mould, J. Am. Ceram. (1961)
Aging of damaged glasses “recovers” its strength under low humidity Heat-treatment under low humidity further enhances surface defect “healing” Stress corrosion kinetics of the damaged glasses increases with level of water presence when the samples are under tension
Part 5. Chemical Tempering • De-alkalization Na2O (surface) + SO2 + 1/2O2 Na2SO4 (surface)
Na-rich matrix: high thermal expansion
Na-lean surface after de-alkalization: low thermal expansion
SiO2
CaO
Na2O
Sulfur-treated surface after bloom removal
Result: a compressive surface layer formed as the treated glass object cooled from dealkalization temperature to room temperature
Chemical Tempering Options Improvement on Glass Impact Resistance Wang and Tao, Glass Surface Chemical Treatment in Glass Surface Treatment Technology, Chemical Industry Press (Beijing, 2004)
De-alkalization Agent As-received De-alkalization Ion-Exchange* SO2 NH4Cl (NH4)2SO4 AlCl3 (NH4)2SO4+AlCl3 (10:1) NH4Cl+AlCl3 (10:1) NH4Cl+(NH4)2SO4 (1:1)
76 76 76 76 76 76 76
93 (22% ↑) 96 (26% ↑) 92 (21% ↑) 88 (16% ↑) 88 (16% ↑) 96 (26% ↑) 98 (29% ↑)
105.5 (39% ↑) 105.5 (39% ↑) 105.5 (39% ↑) 105.5 (39% ↑) 105.5 (39% ↑) 105.5 (39% ↑) 105.5 (39% ↑)
Combined# 144 (89% ↑) 152 (100% ↑) 135 (78% ↑) 120 (58% ↑) 130 (71% ↑) 129 (70% ↑) 132 (74%↑)
* Ion-exchange process: soak container in 70oC - 200 ml water solution containing 34g KNO3 - 69g KCl 8.5g K2SO4 and followed by heat-treatment at 500oC. #
De-alkalization first and followed by ion-exchange treatment
Ion Exchange Varshneya, JNCS 19 (1975)
Flexure Strength (MPa)
Varshneya, IJAGS 1 (2010)
Gy, Mater. Sci. Eng. B 149 (2008)
-250 MPa -100 MPa
Vickers indentation load (N)
(i) as-float (ii) thermal tempering (iii) chemical tempering
Karlsson, et. al. Glass Tech. Eur. J. Glass Sci. Tech. A (2010)
Source and Processing Temperature Range of Salts Ion
Source
Tmin (C)
Tmax (C)
Li+ Na + K+ Rb + Cs + Ag + Tl +
LiNO3 NaNO3 KNO3 RbNO3 CsNO3 AgNO3 TlNO3
261 307 334 310 414 212 206
600 380 410 370 510 444 430
ns
0.012 0.01 0.015-0.02 0.04 0.1 0.1-0.2
Rogoziński, Thesis (2012) http://dx.doi.org/10.5772/51427
• • •
Ion exchange temperature should be well below glass transition temperature Increase in glass transition temperature to improve ion exchange process efficiency at higher temperature Good surface quality of ion exchanged glass should be maintained
Summary A selective literature review on strength of glass and glass fiber was made covering effects of surface hydrolysis and surface flaw on useable strength of glass (USG). Applying Griffith-Inglis-Orowan theory on fracture of solids, specific examples are provided to elucidate importance of stress-assisted hydrolytic effect on USG, which highlights more pronounced detrimental impact of stress-assisted glass surface hydrolysis over the effect of stress-free hydrolysis. It is important to develop new glass chemistry to achieving greater pristine strength; in commercial applications development of new coating materials for bulk glass or new sizing for fiber glass is essential to significantly raise USG. The latter offers advantages of increase USG of existing glass or fiber glass products with minimum or without changing of existing processes, i.e., glass melting and product forming.
THANKS for You Attention!
Hong Li, Ph.D., Sr. Scientist Fiber Glass Science and Technology PPG Industries, Inc. Pittsburgh, PA, USA E-mail:
[email protected]