ISO 7884-4:1987

INTERNATIONAL STANDARD
INTERNATIONAL ORGANIZATION FOR STANDARDIZATION
ORGANISATION INTERNATIONALE DE NORMALISATION
MEXjJYHAPOAHAR OPrAHM3AL&lFl fl0 CTAH~APTM3AYMM
Glass - Vkcosity and viscometric fixed Points -
Part 4 :
Determination of viscosity by beam bending

---------------------- Page: 1 ----------------------
Foreword
ISO (the International Organization for Standardization) is a worldwide federation of
national Standards bodies (ISO member bedies). The work of preparing International
Standards is normally carried out through ISO technical committees. Esch member
body interested in a subject for which a technical committee has been established has
the right to be represented on that committee. International organizations, govern-
mental and non-governmental, in liaison with ISO, also take part in the work.
Draft International Standards adopted by the technical committees are circulated to
the member bodies for approval before their acceptance as International Standards by
the ISO Council. They are approved in accordance with ISO procedures requiring at
least 75 % approval by the member bodies voting.
International Standard ISO 7884-4 was prepared by Technical Committee ISO/TC 48,
Laboratory glassware and related apparatus.
Users should note that all International Standards undergo revision from time to time
and that any reference made herein to any other International Standard implies its
latest edition, unless othetwise stated.
0
International Organkation for Standardkation, 1987
Printed in Switzerland

---------------------- Page: 2 ----------------------
ISO 7884-4 : 1987 (E)
INTERNATIONAL STANDARD
Viscosity and viscometric fixed Points -
Glass -
Part 4 :
Determination of viscosity by beam bending
viscosity is possible only with Newtonian or linear-viscoelastic
0 Introduction
behaviour of the glass. The procedures are sensitive to interference by
devitrification of the Sample. With viscosities above 1012 dPas the ad-
International Standard ISO 7884, Glass - Viscosity and
justment of the structure equilibrium within the glass is perceptibly
viscometric fixed Points, consists of the following separate
delayed with respect to the temperature setting. For tests within this
Parts:
range it should be agreed whether it is necessary to wait for the final
equilibrium viscosity at a given temperature or to take the viscosity
Part 1: Principles for determining viscosity and viscometric
value corresponding to a conventional temperature-time Programme
fixed poin ts.
(sec 6.3).
Part 2: Determination of viscosity b y ro ta tion viscometers.
Part 3 : Determination of viscosity b y fibre elonga tion 2 Reference
viscome ter.
IEC Publication 584-1, Thermocouples - Part 1: Reference
Part 4: Determination of viscosity by beam bending.
tables.
Part 5: Determination of working Point by sinking bar
viscometer.
3 Principle
Part 6: Determination of softening Point.
3.1 Beams
Part 7: Determination of annealing Point and strain Point by
For this test procedure, rod-shaped test specimens, called
beam bending.
beams, are prepared from the Sample. Along their length I they
o f (dila tome tric) transforma tion
Part 8: Determination
have a constant Cross-sectional area (see figure 1) which may
temperature.
be
a) rectangular, of thickness h, and width b;
1 Scope and field of application
b) circular, of diameter d.
This part of ISO 7884 specifies a method of determining the
dynamic viscosity of glass on a rod-shaped test specimen
3.2 Supports, span
. (called a beam) supported at its ends. The viscous deflection
The beam is placed horizontally on two supports; the beam axis
rate of the beam is measured under a given load at the midpoint
and the supports are perpendicular.
between the supports. In addition the viscosity-temperature
relationship and the dependence of the viscosity on the thermal
For rectangular Cross-section beams the supports are horizon-
history of the Sample tan be determined.
tal and have straight edges.
The viscosity range covered by this method extends from 109 to
For circular Cross-section beams the support edges may be
1015 dPas*, corresponding to measuring temperatures be-
semi-circles or notches.
tween about 900 and 400 OC for all glasses of normal bulk-
production compositions.
The distance Zs between the supports is called the span. The
beam juts out only little beyond the supports, satisfying equa-
The procedures are limited to small deflections and to small
tion (1):
deflection rates (see 3.6).
1
NOTE - During beam bending, elongation flows of both signs occur
- < 1,15 . . .
1,l < (1)
(Zero passage within the neutral plane). The determination of shear
k
dNs
*
1 dPas = 1 z = 1 P
(P is the Symbol for Poise)

---------------------- Page: 3 ----------------------
ISO 7884-4 : 1987 (E)
Cross-sections of the beams
A supports
bending edge, unloaded
BO
bending edge, loaded (elastically deformed beam)
Bl
bending edge, sagged Position Af after measuring time At (Af is the interval between B, and Bz)
Bz
beam length
span; Position of bending edge at ls/2
k
forte of the load
FO
b beam width
(rectangular Cross-section)
beam thickness
h
d beam diameter (circular Cross-section)
Figure 1 - Principle of viscosity determination by beam bending
3.5 Flow
33 . Load, loading pieces, bending edge
The load consists of all Parts of the measuring device on which When the forte of the load (disregarding the dead-weight) acts
upon a beam free from defects and showing Newtonian or
gravity acts to produce a forte on the beam by means of the
bending edge, i.e. the loading pieces (variable) and the loading linear-viscoelastic behaviour, and all elastic deformations after
applying the load have faded out and thereupon the sag is suffi-
rod together with yoke and bending edge (given for the in-
The load exerts a vertically ciently small, the flow is described by equation (2) as follows:
dividual measuring device).
downward directed forte F. upon the central Cross-sectional
3
area of the beam (distance ls/2 from both supports). The bend-
df 1, Fo
. . .
ing edge is horizontal and parallel to the supports. (2)
dt=
144 r,r
3.4 Dead-weight
The dead-weight Sterns from the beam; it tan be taken into ac-
dt is the midpoint deflection rate, with which the bend-
dfl
count by calculation - see equations (13) to (15). Within the
ing edge moves downward (sec figure 1);
span, the forte of the dead-weight acts in the same sense as
that of the load. The dead-weight of the overhanging Parts of
is the Cross-sectional moment of inertia of the beam;
the beam produces a forte component opposed to the forte of
the load; this part of the dead-weight tan be neglected if equa-
is the span;
tion (1) is respected.
k
2

---------------------- Page: 4 ----------------------
ISO 7884-4 : 1987 (E)
of the beam caused by the load and - if necessary - the
is the dynamic viscosity of the glass.
YI
deflections during previous flows.
the Trouton ratio 3 and some in-
NOTE - The factor 144 comprises
Measuring devices with 4 < 13 shall be checked by means of a
tegration factors.
beam made from a reference glassl) and the results empirically
with a rec-
The Cross-sectional moment of inertia for beams
corrected, if necessary.
tangular Cross-section is described by equation (3)
Relative deflections z > 0,05 are not permissible. Beams
h3b deflected down to this limiting value may be turned over for a
. . .
r, = - (3)
further run (see also 6.3.3).
12
NOTES
and that for beams with a circular Cross-section, by equation
1 The correcting calculations known from the statics of an elastic
(4) :
beam with moderate support ratios q = 10 are subject to the condition
nd4
that supports are freely movable against one another in the direction of
. . .
rc = - (4)
the span. Using the test set-up this is not possible for the flow;
64
therefore mathematical corrections are not available.
During the measuring time At the beam sags below the bend-
2 The dimensions and loads recommended in ISO 7884-7 are taken
ing edge for a distance Af. The viscosity is calculated according into account. In view of a more uniform temperature distribution,
shorter beams are proposed. The essential differente in comparison
to equation (5) :
with ISO 7884-7 is that:
l:Atm
a) viscosities tan be calculated from the bending rates (therefore
. , .
=681- (5)
r
only considerably smaller relative midpoint deflections are admit-
Vf *
ted);
b) the viscosity of the delivered Sample having its own thermal
history is determined, if necessary (therefore the Sample is not
is the viscosity in decipascal seconds;
heated up to 1012 dPas, and furthermore no viscosities are deter-
mined for decreasing temperatures).
in millimetres during measuring
Af is the sag of the
time At;
4 Apparatus
is the Cross-sectio nal mo ment of in milli to
4
The requirements for components of the beam bending testing
the fout-th power;
device are given in 4.1 to 4.6. Figure 2 Shows an example of a
At is the measuring time in seconds;
testing device.
is the mass of the load in grams;
m
4.1 Viscometer furnace
is the span in millimetres.
k Electrically heated furnace for temperatures up to about
900 OC. The introduction of thermocouples for the determina-
When calculating the viscosity it may be necessa ry to take cor-
tion of temperature and temperature distribution along the
7.3)
rections into account tsee 7.1 to
beam shall be possible. Temperature differentes within the
beam shall not exceed 1 OC.
3.6 Range of applicability of the simplified
The furnace shall be controlled by a device for maintaining a
calculations
constant temperature within & 1 OC or better within the work-
ing space of the furnace and for the adjustment of linear
Equations (2) and (5) hold only for very thin beams and very
temperature-time Programmes with heating rates up to
small deflections. That range is characterized by the support
6 OC/min.
ratio 4 according to equation (6) or equation (7) :
The furnace and its control device for the temperature-time
Programme shall be such that the furnace, starting from a con-
=-
. , . (6)
4 ;
stant temperature level, reaches the required heating rate at the
latest 5 min afterwards and maintains it to L- 10 %.
121s
=
. . . (7)
4.2 Temperature measuring and indicating
q d
instruments
and also by the relative midpoint deflection z:
4.2.1 The alumina-insulated platinum-10 % rhodium/plati-
num (type S according to IEC 584-1) thermocouples or nickel-
f
=-
. . .
(8)
2 I
1,
chromium/nickel (type K according to IEC 584-1) thermo-
‘
s
couples shall exhibit low thermal inertia (the diameter of the
In equation (8) f is the total midpoint deflection of the beam, wires should be not greater than 0,5 mm). The wires shall have
i.e. f is the sum of the deflection Af during the measuring time a sufficient length within the furnace (with respect to heat con-
At according to equation (5) together with the elastic deflection duction along the wires).
See for example ISO 7884-1 : 1987, annex B, “Examples of certified reference glasses for viscometric calibration”.
1)

---------------------- Page: 5 ----------------------
ISO 7884-4 : 1987 (El
-
-
X
1
4 B 5 6
l
/
I I
.
I
_~-
1 1 4
1 Support stand, made from vitreous silica 6 Test specimen (beam)
2 Frame, made from a suitable temperature-resistant low- 7 Locking rod, made from vitreous silica
expansion metal alloy
8 Upper part of viscometer furnace: vertically movable cap
3 Supports
9 Loading rod, made from vitreous silica
4 Locking counterpoise, made from a suitable temperature-
A and B : Hot junctions of thermocouples (see 4.2)
resistant Iow-expansion metal alloy
5 Yoke with bending edge, locking edges and Suspension of
the loading rod, made from a suitable temperature-resistant
low-expansion metal alloy
Example of a testing device for the beam bending method
Figure 2 -

---------------------- Page: 6 ----------------------
ISO 7884-4 : 1987 (E)
4.2.2 Control thermocouples should be located as close as 4.5 Loading device
possible to the furnace windings for fast response. The hot
junction of the measurement thermocouples, however, shall be
4.5.1 Yoke and loading rod
placed in the immediate vicinity of the beam (see A in figure 2).
The axial temperature distribution along the beam shall be
The yoke and bending edge are made from chromium-nicke1
monitored by further thermocouples (see B in figure 2). In ac-
alloy or hard porcelain. The radius of curvature of the bending
cordante with ISO 7884-1, the measurement thermocouples
edge tan vary Getween 0,5 and 2 mm; the cylindrical surface is
shall be calibrated and the calibration checked regularly.
ground and polished.
4.2.3 The electrical output of the thermocouples shall be To prevent sticking of the beam to the bending edge, Strips of
determined at zero current by means of potentiometers, or
platinum or nicke1 foil (about 0,Ol mm thick) may be inter-
high-resistance electronie amplifiers having a sensitivity of 1 PV posed.
for type S (according to IEC 584-1), or 4 PV for type K (accord-
ing to IEC 584-1) thermocouples. Precautions shall be taken NOTE - A greater radius of curvature is advantageous multiple
for use
that the ice-bath for the cold junction is maintained at 0 OC of the beam after turning it over.
throughout the test. If the temperature measuring equipment is
fitted with automatic cold junction compensation, the ice-bath The loading rod connects the yoke within the viscometer fur-
nace to the loading pieces underneath. The loading rod shall be
tan be omitted.
made from the same material as the support stand tube (see
4.4.2) with respect to similar thermal expansion characteristics.
Loading pieces
43 .
A set of loading pieces with masses from about IO to 200 g (for
4.5.2 Locking the yoke
arrangements according to 6.2.3, up to 1 000 g) made from
brass, nickel-plated or equivalent material. The masses of the
A device is needed for lowering the bending edge onto the
loading pieces shall be determined to 0,Ol g. The mass of the
beam and for lifting it after the test. The device shall ensure
loading rod including the core of the displacement pick-up
that the parallelism between the lowered bending edge and the
together with the yoke and the bending edge tan be limited to
supports is better than Io and that the edge is less than 0,5 mm
about 10 g.
from the median plane.
NOTE - A detailed device whic h is able to fulfil these requirements is
4.4 Beam support
given as an example in figure 2.
4.4.1 Frame
4.6 Equipment for the determi nation of the midpoint
deflection rate
The frame shall be sufficiently rigid against bending and torsion
and be made from a suitable temperature-resistant low-
expansion metal alloy or hard porcelain. The front sides of the
4.6.1 Moving indicator (including transducer core) as Point
frame bear sufficiently broad (IO to 15 mm) supports with a
of Observation for the determination of the deflection rate,
radius of curvature of about 0,5 mm, the surfaces of the sup-
placed beneath the viscometer furnace.
ports being ground and polished. The span, i.e. the distance
between the two lines of contact to the bottom surface of the
4.6.2 Device for the determination sf the midpoint
beam, shall be determined to 0,05 mm. Parallelism deviations
deflection Af during the measuring time At according to
of the two lines of contact should not exceed 0,05 mm, after
equation (5). Deflections of 0,l mm shall be determined to 1 %
the frame has been annealed. After prolonged use, span and
(see also tables 1 and 2).
parallelism shall be checked.
To prevent sticking of the beam to the support, Strips of
4.6.3 Device for the determination of the total beam
platinum or nicke1 foil (about 0,Ol mm thick) may be inter-
deflection f monitoring the limiting condition for the relative
posed.
deflection z < 0,05 with a sensitivity of 0,l mm.
NOTE - Linearly variable differential transformers (LVDTs) with a
4.4.2 Support stand
removable core are suitable for the deflection determination. A
measurable elevation adjustment of the coil is then sufficient for the
The support stand bears the frame upon its upper front surface.
determination of fi whilst the measurement of Af (sec 4.6.2) is achiev-
In the example shown in figure 2 it is set up separately from the
ed by means of the electronie meter of the LVDT having several sen-
furnace. The stand shall be equipped with an adjustment
sitivity ranges. With this testing method, recording of the values
device for the horizontal support of the beam.
measured should be aimed at. Alternatively, a measuring microscope
with scale micrometers (for Ajl and mounted upon a cathetometer
The support stand is made from vitreous silica. If temperatures
base (for J) may be used.
between 750 and 900 OC are often applied, and/or if alkali con-
tamination is suspected, alumina refractory as a material for the
4.6.4 Measuring device for time intervals ranging from 10
support stand is a useful alternative.
to IO 000 s for the determination of the measuring time At ac-
NOTE - Another example for a possible construction of the beam sup- cording to equation (5). Systematic deviations of the measur-
port is shown in ISO 7884-7. In that case the supports are machined ing device shall be determined to 0,2 % and shall be taken into
directly into the top of the stand tube.
account.

---------------------- Page: 7 ----------------------
ISO 7884-4 : 1987 (E)
5.4 Special requirements
4.7 Devices for measuring the beam dimensions
Special requirements concerning the treatment of Sample and
4.7.1 Sliding gauge (vernier division l/lO is sufficient) for
beam shall be agreed upon, especially in the following cases:
the determination of the beam length 1.
a) for samples delivered in the form of grains, the condi-
tions for melting and annealing the rod las-drawn or cast
4.7.2 Micrometer caliper for the determination of beam
from the melt), from which the test specimen will be
diameter d or beam thickness h and width b.
prepared;
b) for specially annealed samples, the highest tempera-
tures to which the beams tan be exposed without affecting
5
Preparation of test specimens
this annealing treatment.
5.1 State of delivery
6 Procedure
The supplied glass Sample shall be uniform, bubble-free,
homogeneous and annealed. lt shall consist of pieces large
6.1 Calibration of the testing device
enough to permit the preparation of the test specimens.
Generally, the beam bending measurement is an absolute
5.2 Preparation of the beams
determination of the viscosity, i.e. the viscosity is calculated
(see 3.5 and 7.2) from the dimensions of the finished test
Rectangular beams shall be made from the Sample by cold
specimen and from the span Is. An examination is recommen-
working, e.g. diamond-saw tut and mill ground. Cylindrical ded by means of a beam of equal dimensions and made from a
beams shall be either flame drawn or centreless ground.
reference glass of known viscosity. This is necessary for sup-
port ratios 4 < 13 (sec 3.6 and 7.3).
The beams shall not have any scratches or defects. The dimen-
The calibration includes the adjustment of the frame for
sions shall be within the ranges specified in 6.2.1, 6.2.2 and
6.2.3 and figures 3 and 4. horizontal positioning of the plane defined by both supports.
Subsequently the highest temperature is set that might occur in
this testing device, and after cooling down
...