Method of Glass Property Measurement
|
|
Density |
The density of suitably selected piece of glass is measured by weighing it in air and dividing this weight by the buoyancy when the sample is suspended in water (Archimedes'method). Occasionally, the glass may be affected by water, in which case a suitable inert liquid such as kerosene may be selected. This method of measurement is accurate to 0.001 g/ml. When bulk glass specimen includes bubbles, the pycnometer method is employed using a pulverized specimen. Automatic measurements of density can be carried out using pycnometers that utilize He gas as the displacement fluid. A powder or granular sample, on the order of a few milligrams, is all that is required to perform the measurement. The accuracy of such instruments is approximately 0.002 g/ml.
The density of high temperature glass melt is measured by the Archimedes'method. Two platinum balls of different diameters are individually dipped in melt and weighed. The density of the melt is calculated from the weight difference between the two balls dipped in melt and the volume difference between the two balls. |
Elastic Moduli |
The techniques commonly used for measuring the elastic moduli of glasses
(i.e. Young’s modulus, shear modulus, and Poisson's ratio) can be devided
into three categories: (1) Those based on the measurement of the stress-strain
curve, (2) those based on the propagation of ultrasonic waves, and (3)
those based on the estimation of natural vibration frequencies of a bar
(ASTM C 623-92).
|
Modulus of Rupture |
The strength of glass depends on the surface state of the test specimen. It is closely related to the testing method such as the three-point bending test, the four-point bending test (ASTM 158-95) and the biaxial ring test. The dimensions of the test piece, loading speed, and atmosphere must be specified.
Abraded strength -
In order to compare different glasses, the abraded strength is determined. Many procedures are employed for normalizing or controlling the surface defect condition of glass specimens or articles used before the strength tests. A typical technique is as follows: Dry SiC grits of specific mesh size are dropped from a fixed height on the specimen surface to be tested and exposed away from the edge using a mask. (ASTM C158-84)
|
Fracture Toughness |
Fracture toughness, KIC, is defined as the value of the stress intensity
factor at which the crack propagation becomes fast. There are two methods
for determination of fracture toughness. One method uses notched specimens
and the other uses specimens with indented microcrack. Specimens from materials
exhibiting static fatigue are usually tested in an inert environment in
order to prevent subcritical crack growth during loading..
Notched specimen |
The measurement consists of loading the notched specimens by
increasing the load untill fracture occurs. There are various methods that can
be used depending on the shape of notch and type of loading. The single-edge
precracked beam (SEPB) method is a typical method used. A rectangular test beam
with a single-edge precrack is tested by three point bending. KIC is
calculated from the breaking load, the length of the precrack, the dimension of
the test beam and the span length. (JIS R1607)
|
Specimen with an indented microcrack |
The advantage of this method lies in the fact that no special specimens are necessary, so that it can also be used for fast inspection. There are two kinds of methods for measuring fracture toughness using indentation-produced cracks. |
(a) |
IF(indentation fracture)method |
A Vickers indenter produces imprint diagonal and radial cracks in indentation diagonals under sufficient loads. The KIC is calculated using the dimension of diagonal depression and crack length. (JIS R 1607). Care must be taken not to apply to some glasses such as silica glass and borosilicate glass for which significant densification occurs via indentation. |
(b) |
IS(indentation strength-in-bending) method |
After creating a central crack using the Vickers indenter, the bending strength of the specimen is measured. The imprint must be on the side loaded by tension, with one of the cracks perpendicular to the tensile stress. The KIC is calculated using the Young’s modulus, the Vickers hardness, the bending strength and the indenter load by means of a semiempirical equation. |
|
Thermal Shock Resistance |
Test samples of glassware are heated in an oven or a hot bath, then they
are rapidly transferred from the hot bath to a cold water bath. The number
of glassware samples which the test is determined by individual inspection
of each. The temperature difference between the oven (or the hot bath)
and cold water required to produce fracture is taken as a measure of thermal
shock resistance. Thermal shock resistance is evaluated by the following
test procedures depending upon the purpose of test; (1) pass test, (2)
progressive test (to a predetermined percent of breakage) and (3) progressive
test (total).
|
Viscosity, Softening point, Annealing point, Strain point |
For the viscosity of glass spanning roughly 14 orders of magnitude, the measurement has to be performed using various techniques applicable for limited ranges of viscosity. These are summarized in the following table.
Range |
Method |
Viscosity value
|
Melting |
Falling sphere or drawing-up sphere
Rotation viscometer
(a) Rotating spindle and fixed crucible.
(b) Rotating crucible combined
with fixed inner cylinder.
Sinking bar viscometer |
<105 dPa.s
<106 dPa.s
=104 dPa.s
|
Softening and annealing |
Parallel plate
Penetration viscometer
Fiber elongation
Softening point
Annealing point and strain point
Beam bending
Annealing point and strain point |
105 < viscosity < 109 dPa.s
105 < viscosity < 109 dPa.s
105 < viscosity < 1016 dPa.s
=107.65 dPa.s
=1013 and 1014.5 dPa.s
107 < viscosity < 1014.7 dPa.s
=1013.2 and 1014.7 dPa.s
|
Glass transition temperature |
Dilatometer
DTA, DSC
|
|
|
|
Falling sphere or drawing-up sphere
|
The moving velocity of a platinum ball in molten glass is constant due to the balance of gravitational force or draw-up force with the viscous drag force acting on the ball according to Stokes’ relation for a viscous fluid. The viscosity is determined by the velocity measurement.
|
|
Rotation viscometer
|
|
|
(a)
|
Rotating spindle and fixed crucible
|
The determination of the viscosity of glass is conducted through the use of a platinum alloy spindle immersed in a crucible of molten glass. During the spindle rotates at a constant angular velocity, the torque of the spindle is measured. The angular velocity and the torque are used to calculate viscosity. (ASTM C1351M-96)
|
|
(b)
|
A rotating crucible
combined with a fixed inner cylinder |
During the crucible or the outer cylinder rotates at a constant rate, the torque developed by differential angular velocity between the crucible and the inner cylinder (or shaft) is measured. (ISO 7884-2, ASTM C1351M-96,)
|
|
Sinking bar viscometer
|
A vertically positioned narrow metal rod (the bar) is allowed to sink under gravity (its own weight) into the melt in a crucible. The bar is made of platinum alloy and is 0.5 mm in diameter, 20 mm in length and 0.902 g in mass. The sinking time of the bar corresponding to specified depth is measured. The viscosity is calculated from the time and geometrical conditions. This method is used for determining the viscosity within the range of approximately 103.7 to 104.5 dPa.s. (ISO 7884-5)
|
|
Parallel plate viscometer
|
A disk of glass approximately 6-8 mm diameter and 3-5 mm thickness, is sandwiched between two horizontal parallel plates inside a well-insulated furnace. The specimen is pressed through a load rod vertically, and the rate of sag is recorded as a function of time. The viscosity is calculated from the rate of sag, the applied load, the geometrical condition of the specimen and the thermal expansion coefficient of the glass. The viscosity range of determination is from 105 to 109 dPa.s. (ASTM C 338-93)
|
|
Penetration viscometer
|
A ball or rod of defined geometry is loaded and allowed to penetrate a given distance into the glass. Viscosity is calculated from the rate of penetration. (Ball : Douglas, R. W. et al., Glass Technology 6, 52 (1965), Rod : Kunugi M. et al., Zairyou 15, 567 (1966)) )
|
|
Determination of softening point
|
A glass fiber of 0.55 to 0.75 mm in diameter is drawn. One end is fused to form a ball. The fiber is cut to 23.5 mm in length, and is suspended inside a specified furnace, which only covers the top 100 mm. The furnace is heated at 5 C/min. The fiber softens and elongates under its own weight upon heating. The sagging of the lower end is observed using a telescope and measured as a function of time. The temperature at which the elongation rate of the fiber is 1 mm/min is identified as the softening point. The density of the glass acts as a pulling force and the surface tension of the glass acts as a shrinking force on the fiber. For soda-lime glass, 1 mm/min corresponds to approximately 107.65 dPa.s viscosity. For glasses of other compositions, the viscosity values vary somewhat. (ISO 7884-6)
|
|
Fiber elongation
|
To measure high viscosities, loading a weight at the lower end of a fiber is needed to pull the fiber. In this way, viscosities have been measured to the practical highest limit of approximately 1016 dPa.s. (ISO 7884-3)
|
|
Beam bending
|
A simply supported specimen beam is centrally loaded and its rate of sag at the center is recorded. The equation to calculate the viscosity is derived by extending viscoelastic analogy to the three-point elastic beam bending method. This method is beneficial to the glass hard to form fiber by lampworking such as green glass of glass-ceramics because a beam can be shaped from glass block by cutting and grinding. (ISO 7884-4)
|
|
Determination of annealing point and strain point by beam bending
|
The beam bending measurement is carried out under a cooling condition, 4 C/min. The annealing point is defined as the temperature at which the viscosity of glass corresponds to 1013.2 dPa.s, and the strain point is identified as the temperature at which the viscosity of glass corresponds to 1014.7 dPa.s by extrapolation of the sag rate to lower temperatures. At the temperature corresponding to annealing and strain points, the viscosity of glass is highly time-dependent. Hence, any viscosities determined by this procedure cannot be assumed to represent equilibrium structural conditions. The temperature corresponding to annealing and strain points should not be used for the VFT equation. (ISO 7884-7)
|
|
Determination of annealing point and strain point by fiber elongation
|
A glass fiber of 0.55 to 0.75mm in diameter and 508 mm in length is suspended from the top of a furnace 368 mm in length (method A), or a glass fiber of 100 to 200 mm in length with the same diameter is suspended in the center of furnace using two supporting rods (method B). A weight of 1 kg mass is applied to fiber as a pulling road. The fiber elongation measurement is carried out under a cooling condition, 4 C/min. The sag rate is measured by a suitable mean. The annealing point is defined as the temperature at which the viscosity of glass corresponds to 1013 dPa.s, and the strain point is identified as the temperature at which the viscosity of glass corresponds to 1014.5 dPa.s by extrapolation of the sagg rate to lower temperatures. (ASTM C336-71)
|
|
Glass Transition Temperature, Tg (Transformation Temperature) |
Significant changes take place in some specific physical properties at Tg. The observed temperature can vary significantly depending on the property chosen for observation and on the details of the experimental technique. The observed Tg should be considered valid only for that particular technique and set of test conditions. It is said that Tg corresponds to the temperature at which the viscosity value is 1013.3 dPa.s for soda-lime glass, but Tg has no relation to the annealing point and the strain point.
|
Dilatometric method
|
Tg is normally defined by the intersection of the tangent of the solid side and the tangent of the steep portion in transition range of the dilatometric expansion curve. The measurement is conducted after a specified annealing treatment. (ISO 7884-8)
|
|
Differential thermal analysis (DTA) or differential scanning calorimetry (DSC)
|
Glass powder is put into a test cell, and then the instrument is heated at a constant rate. Tg is defined by the intersection of the tangent of the base line and the tangent of the steep descent portion in the endothermic region by the glass transition of the DTA or DSC curve.
|
|
Linear Thermal Expansion Coefficient |
There are three methods of measurement; pushrod dilatometers continuous, pushrod two temperatures and interferometric. In every case, the glass test piece has to be annealed according to the specified schedule.
|
Pushrod dilatometers continuous
|
There are three type of arrangements, vertical, beveled and horizontal. The test piece, which is butted against the flat wall of the measuring system, pushes on a spring-loaded pushrod upon heating. The measuring system and the pushrod are generally made of fused silica or alumina. Sometimes, an equal-length platinum sample is used as a standard. The temperature of the furnace is raised at a constant rate, and both the expansion and the temperature are recorded continuously. The coefficient of the mean linear thermal expansion is calculated from a expansion difference between two specified temperatures in the solid glass range.
|
|
Pushrod two temperatures
|
A glass test piece 5 mm in diameter and 100 mm in length, both ends of which are shaped like a pencil, is prepared. The test piece is set in a vertical holder made of fused silica, and then a pushrod made of fused silica is inserted and pushes the test piece. A dial gage is set on the pushrod. An ice-water bath (0 C) and a vertical tubular oven maintained at 300 C are prepared. After being placed in the holder, the test piece and pushrod are heated in the oven, and then immersed in the ice-water bath. The difference in the scale of the dial gage is recorded. The mean thermal expansion coefficient, 0 to 300 C is calculated from the length of specimen, the length difference and the temperature interval. (JIS R 3102-1978)
|
|
Interferometric
|
Three carefully ground and polished pieces, two of a known and one of the unknown glass, separate two optically flat and parallel surfaces in a furnace. As the temperature rises, the separation also increases, which can be monitored interferometric methods. (JIS R 3251-1995)
|
|
Liquidus Temperature and Crystallization |
|
Temperature gradient furnace
|
Crystallization behavior is observed visually by this method. A glass sample set on a platinum holder is heat-treated in a furnace maintaining a specified constant temperature gradient. After the treatment, crystals grown in glass are observed by a microscope, and the liquidus temperature and temperature dependence of crystal growth are determined. Two types of sample holders are used. One is a trough-type platinum container which is filled with finely screened particles of test glass, and the other is a platinum tray in which a number of holes are perforated in a line. Larger screened particles of test glass are positioned one per hole on the plate. (ASTM C 829-81(1995))
|
|
DTA, DSC
|
The crystallization behavior of glass that is easy to crystallize is studied. The onset of crystallization and the peak of the crystallization rate are determined from the exothermic curve of DTA or DSC.
|
|
Fluorescence |
A UV light beam is produced when the light beam of a super-high-pressure mercury lamp is filtered by a UV filter glass which transmits light of 330 to 400 nm and is opaque outside of this wavelength range. The spectral intensity of visible light emitted from an optical glass sample exposed under the UV light is measured by a spectrometer in the direction perpendicular to the UV beam. Glass is classified by the relative intensity of the fluorescence. (JOGIS 07-1994)
|
Solarization |
An optical glass sample is exposed under the light of a super-high-pressure
mercury lamp at 100 C for 4 hours. After exposure, the light transmission
spectrum is measured. The difference in light transmission before and after
exposure is determined as the solarization tendency. (JOGIS 04-1994)
|
Chemical Durability |
The Chemical durability of glass is measured by conducting simple experiments;
such as immersing a glass specimen in a given quantity of fluid contained
in, for instance, a covered, nonreacting vessel at a controlled temperature
in a shaker bath. The quantity of material leached is determined either
by measuring the mass change, by the acid titration of a leachate or by
flame photometric/atomic absorption/induction-coupled plasma analyses (ICP)
of a leachate. The experiment can be conducted on either a bulk specimen
or a pulverized specimen. The latter experiments require less time. Since
the glass constituents are likely to affect the attacking-medium characteristics,
the experiment can be performed in a closed (fixed quantity) medium, or
in an open (flowing) medium where the fluid is continually replenished.
The followings are typical industrial standards related to chemical durability.
Water Durability
Standard
|
ISO 719-1985
|
ISO 720-1985
|
DIN 12111
|
Test specimen
|
Powder
300-500 um,
2 g
|
Powder
300-500 um,
10 g
|
Powder
315-500 um,
2 g
|
Attacking medium
|
Distilled water
50 ml in a SiO2 glass flask.
|
Distilled water
50 ml in a SiO2 glass conical flask
|
Distilled water
50 ml
|
Temperature
|
98 C
|
121 C
in autoclave
|
98 C
|
Time
|
60 min
|
30 min
|
60 min
|
Determination
|
Titration with 0.01N HCl
|
Titration with 0.02N HCl
|
Titration with 0.01N HCl
|
Report
|
ml HCl/ mg grain
|
ml HCl/ mg grain
|
ml HCl /mg grain
|
Class
|
Hydrolytic class:
HGB 1
HGB 2
HGB 3
HGB 4
HGB 5
|
Hydrolytic class:
HGA 1
HGA 2
HGA 3
|
|
Standard
|
ASTM C 225-85
|
JIS R 3502-1995
|
JOGIS 06-1999
|
Test specimen
|
Bottle
|
Powder
300-425 um
10 g
|
Powder
250-420 um
mass in grams corresponding todensity
|
Powder
425-600 um
same grams as specific gravity
|
Attacking medium
|
|
High-purity water 50 ml in Erlenmeyer Flask (250 ml)
|
Distilled water
50 ml in a glass flask.
|
Hight-purity water with pH6.5-7.5
80 ml in flask with cooler
|
Temperature
|
121 C
in autoclave
|
90 C
|
121 C
in autoclave
|
100 C |
100 C |
Time
|
1 h
|
24 h
|
1 h
|
60 min
|
60 min
|
Determination
|
Titration with 0.020N H2SO4
|
Titration with0.005 mol/LH2SO4
|
Spacific loss in mass |
Report
|
ml of 0.020N H2SO4
|
As mg Na2O
|
Hydrolttic class: 1-6
1: <0.05wt%, 6: >1.1wt% |
|
Acid Resistance
Standard
|
ISO 1776-1985
|
ASTM C 225-85
|
DIN 12116
|
JOGIS 06-1999
|
Test specimen
|
Plate 30-40 cm2
|
Bottle
|
Plate 300 cm2
|
Powder
425-600 um
same grams as specific gravity
|
Attacking medium
|
6 mol/L HCl,
25 mL
|
0.0005N or
0.0002N H2SO4
|
6N HCl
|
0.01N HNO3
80 ml in flask with condenser
|
Temperature
|
100 C
|
121 C
in autoclave
|
Boiling
|
100 C |
Time
|
3 h
|
1 h
|
6 h
|
60 min
|
Determination
|
Na2O by Flame photometer, atomic absorption, ICP
|
Titration with 0.020N NaOH
|
Loss in mass
|
Spacific loss in mass |
Report
|
Na2O ug/dm2
|
Milliliters of acid consumed in the te
|
mg/dm2
|
Acid class: 1-6
1: <0.2wt%, 6: >2.2wt% |
Remarks
|
|
|
Acid class:
1 (lowest), 2, 3 and 4
|
|
|
Alkali Resistance
Standard
|
ISO 695-1991
|
ASTM C 225-73 |
DIN 52322
|
Test specimen
|
Plate 10-15 cm2
|
|
Plate 10-15 cm2
|
Attacking medium
|
Equal volume of1mol/L NaOH and
0.5 mol/L Na2CO3in silver vesselwith condenser
|
|
Equal volume of1mol/L NaOH and
0.5 mol/L Na2CO3in silver vessel
with condenser
|
Temperature
|
102.5 C
|
|
102.5 C
|
Time
|
3 h
|
|
3 h
|
Determination
|
Loss in mass
|
|
Loss in mass
|
Report
|
mg/dm2
|
|
mg/dm2
|
Remarks
|
Alkali class:
A1
A2
A3
|
|
Alkali class:
A1
A2
A3
|
ISO 695:1991
|
Glass Resistance to attack by a boiling aqueous solution mixed alkali - Method of test and classification
|
ISO 719:1985
|
Glass Hydrolytic resistance of glass grains at 98 degree C. Method of test and classification.
|
ISO 720:1985
|
Glass Hydrolytic resistance of glass grains at 121 degree C. Method of test and classification.
|
ISO 1776:1985
|
Glass Resistance to attach by hydrochloric acid at 100 degree C. - Flame emission or flame atomic absorption.
|
ISO: 4802-1:1988
|
Glassware Hydrolytic resistance of the interior surface of glass container - Part 1: determination by titration method and classification.
|
ISO 4802-2:1988
|
Glassware Hydrolytic resistance of the interior surface of glass container - Part 2: determination by flame spectroscopy and classification.
|
ISO 7086-1:1987
|
Glass ware and glass ceramic ware in contact with food-release of lead and cadmium.
|
ISO 9686:1990
|
Raw optical glass Resistance to attack by aqueous alkaline phosphate-containing detergent solutions at 50 degrees C
Testing and classification. |
ISO 10629:1996
|
Raw optical glass Resistance to attack by aqueous alkaline solutions at 50 degrees C-Testing and classification.
|
ASTM C225-85
|
Resistance of glass containers to chemical attack
|
JIS R 3502-1995
|
Test method of glass apparatus for chemical analysis
|
JOGLS 06-1999
|
Measuring method for chemical durability of optical glass (powder method)
|
DIN 12111
|
Testing of glass Powder method for testing of water-durability of glass product at 98 C and classification of glass in hydrolytic class
|
DIN 12116
|
Testing of glass Determination of acid-resistance (gravimetric method) and classification of glass in acid class
|
DIN 52322
|
Testing of glass Determination of alkali-resistance and classification of glass in alkali class
|
|
Weathering |
Weathering experiments generally involve measurements of the changes in
the surface property of a specimen kept in a humid chamber or in a cyclic
humid-dry atmosphere for a specified period of time. The extent and the
mode of chemical attack is evaluated by light scattering (or haze) and
may also be analyzed using surface characterization instruments such as
IRRS, AES, SIMS and EDAX. According to JOGIS 07-1975, a polished glass
plate is held in a chamber saturated by water vapor at 50 C for 24 h. Then
the glass plate is dried and the intensity of scattered light from the
surface is measured by a hazemeter. |
|