Note : Les descriptions sont présentées dans la langue officielle dans laquelle elles ont été soumises.
CA 02046387 1999-O1-14
1900066
Title: METHOD AND APPARATUS FOR DETERMINING THE
FUNDAMENTAL VISCOELASTIC PROPERTIES OF A MATERIAL
FIELD OF THE INVENTION
This invention relates generally to an apparatus and method for measuring the
fundamental viscoelastic properties of a material, and more particularly, to
improvements in a Dynamic Stress Relaxometer.
BACKGROUND OF THE INVENTION
Viscoelastic materials, such as elastomers, rubbers, plastics, and the like,
can
be described by a number of fundamental viscoelastic properties which vary
between
different material types and, in fact, may vary between separate batches. of
the same
material type. Consequently, these fundamental properties, such as the loss
and
storage moduli, the complex viscosity and the loss tangent, may be used to
identify
materials either by type or by grade. Further, these properties may also be
used to
predict the processing characteristics of a material.
However, heretofore many methods of determining the fundamental
viscoelastic properties of materials have be~'n time consuming, i.e., often
requiring
several hours, and have relied on the participation of a skilled technician to
perform
the test. Moreover, with many tests it is possible only to describe a few
characteristics of the material, such as die swell or stress relaxation, and
not actually
to determine the fundamental viscoelastic properties of the material. These
drawbacl~s
have made these tests unsuitable for use on the factory floor or for use in a
control
system wherein the results of the test are employed to modify the composition
or
processing of the material.
A Dynamic Stress Relaxometer (DSR), as described in U.S. Patents 3,693,421
and 3,818,751, has provided a relatively quick way to predict some
characteristics of
a material and also to at least roughly distinguish between materials and
grades of
materials. In a DSR test a sale of a visooelastic material is sub-
jetted to a prescribed angular deformation thus creating shear stresses
wer time is measured and plotted to provide a stress relaxation curve
for the material. As the relaxation curve follows a generally
- 1 -
~o~~~~~
exponential decay path, a number of standard measurement points along the
curve are
selected which have been found to yield significant information about the test
specimen in a relatively short time, i.e., less than ten minutes. This
infornnation is
then used to predict the die swell, processability or type or grade of the
material
tested. While DSR testing has been quite an advance over previous methods, it
still
lacked the ability to determine quantitatively the fundamental viscoelastic
properties
of the material.
It would be desirable to develop an improved accuracy DSR device and to
combine such a DSR device with the means to interpret the results to determine
the
fundamental viscoelastic properties of a material.
As used herein the term "viscoelastic" is meant to include synthetic or
natural
rubbers, plastics, thermoplastics, elastomers and any other material that
exhibits
viscoelastic properties.
SUMMARY OF THE INVENTION
It is an object of the present invention to provide a fast and accurate way to
measure the fundamental viscoelastic properties of a material.
It is another object of the invention to substantially automate the process of
performing a dynamic stress relaxation test.
According to one aspect of the present invention, an apparatus for determining
the fundamental viscoelastic properties of a viscoelastic material includes
means for
imparting a torsional stress in a viscoelastic material; means for measuring
the
relaxation of the torsional stress over dme and converting the relaxation
stress to a
representative waveform; and processing means for determining the frequency
dependant fundamental viscoelastic properties of the material based on the
shape of
a portion of the waveform. Preferably, the processing means includes a first
dedicated processor for producing a digital representation of the waveform and
a
second general processor, such as an IBM compatible microcomputer for
determining
the fundamental viscoelastic properties from a portion of the digital
representation.
To facilitate automating the determination of the fundamental viscoelastic
properties
34 of the test material, the first processor examines the waveform and aborts
the
determination if the digital waveform is atypical of an expected waveform for
a
viscoelastic material subjected to a torsional stress.
-2-
~o~~~~~
According to another aspect of the present invention, a method of determining
the fundamental viscoelastic properties of a viscoelastic material includes
the steps of
imparting a torsional stress in a viscoelastic material; measuring the
relaxation of the
torsional stress over time and converting the relaxation stress to a
representative
analog waveform; digitizing the representative analog waveform to produce a
representative digital waveform; and, determining the frequency dependant
fundamental viscoelastic properties of the material based on the shape of a
portion of
the representative digital waveform. To facilitate automating the method, the
shape
of the waveform may be analyzed to detect the start of a test, the end of the
test, and
whether the test is a valid test. Preferably, the fundamental viscoelastic
properties
of the test material are determined by converting the torque based amplitudes
of the
waveform to shear relaxation modulus values, such as by multiplying the torque
amplitudes by the geometric form factor of the test material, and then by
transforming
the shear stress modulus values to frequency dependant fundamental
viscoelastic
properties through the known Yagii/Maekawa approximation.
These and other objects, advantages, features and aspects of the present
invention will become apparent as the following description proceeds.
To the accomplishments of the foregoing and related ends, the invention, then,
comprises and features hereinafter fully described in the specification and
particularly '
point out in claims, the following description and the annexed drawings
setting forth
in detail a certain illustrative embodiment of the invention, this being
indicative,
however, of but one of the various ways in which the principals of the
invention may
be employed. It will be appreciated that the scope of the invention is to be
determined by the claims and the equivalents thereof.
_3_
BRIEF DESCRIPTION OF THE DRAWINGS
In the annexed drawings:
Figure 1 is an illustration of the Dynamic Stress Relaxometer system of
the present invention;
Figure 2 is an oblique partially cutaway view of the DSR device;
Figure 3 is a front view of the internal components of the DSR device
with the cabinet removed;
Figure 4 is a graph of the Torque v. Time curve generated by PROGRAM
I (there are several operational programs or computer programs employed in the
DSR,
as are described below);
Figure 5 is a graph of the Log-Torque v. Log-Time plot generated by
PROGRAM II;
Figures 6' through 9 are graphs representative of those produced by
PROGRAM III;
Figure 10 is a partial side view of the DSR device;
Figure 11 is a rear view of the DSR device;
Figure 12 is a cross-sectional view of the stator assembly with the rotor
shown at partial closure;
Figure 13 is an illustration of the area including the stator and rotor
assemblies with the assemblies shown in cross-section;
Figure 14 is a top view of the rotor table showing the connections with
the load cell and pneumatic cylinder;
Figure 15 is an electric schematic diagram of the DSR device control
electronics;
Figure 16 is a data block diagram representation of the dedicated
processor electronics;
Figure 17a-f are an electrical schematic diagram of the dedicated
processor;
Figures 18 and 19 are flow charts illustrating the functioning of general
software within the general processor including parts of PROGRAMS I, II and
III;
Figures 20 through 23 are flow charts illustrating software functions of
the dedicated processor including parts of PROGRAMS I, II and III;
Figures 24 and 25 are flow charts illustrating the functioning of code
within the general processor related to PROGRAM III data analysis;
-4-
Figure 26 is a close-up view of the rotor and stator shown at closure;
Figure 27 is a close-up view of the rotor and stator as mathematically
modelled;
Figure 28 is a sectional view of the rotor and stator illustrating the
individual mathematical modelling regions; and Figures 29 through 34 are flow
charts
illustrating PROGRAM III data comparison functions within the general
processor.
DETAILED DESCRIPTION OF THE INVENTION
Referring now to the several figures in which like reference numeral
depict like items, and initially to Figures 1-3 there is shown a Dynamic
Stress
Relaxometer (DSR) System 10. The DSR system 10 includes a dynamic stress
relaxometer device 12, a dedicated processor 14, a general purpose computer
16, a
printer 17 and a plotter 18 which function together to provide both
quantitative and
qualitative information about viscoelastic material subject to the test.
Briefly, the DSR device 12 performs a test of a viscoelastic material (test
specimen) by subjecting the test specimen to a prescribed impulsive angular
deflection,
generally 1l2, 1 or 2 degrees. (Other deflections may be employed, for
example,
depending on the properties of the material, the capability or design of the
DSR
machine, etc.) This creates a torsional stress in the test specimen resisting
the angular
deflection. The DSR device 12 then measures the relaxation of the torsional
stresses
in the test material over time and provides the data, in the form of an analog
signal,
to the dedicated processor 14. The dedicated processor 14 scales and filters
the signal
and digitizes it for digital processing in the dedicated processor and, in
some cases, in
the general processor 16 also.
The DSR system 10 preferably provides a selection of three testing
procedures, called PROGRAM I, PROGRAM II, and PROGRAM III, for manipulating
the data, for controlling the test, and optionally for controlling ancillary
or external
operations, such as in a feedback or feedforward control system. In a feedback
control
system the DSR system may be provided with a sample of a prepared viscoelastic
material of which it will determine the fundamental viscoelastic properties or
other
measured characteristics, as will be discussed later. The values of some or
all of these
properties or measured characteristics are then provided to a suitable
controller which
may adjust the quantities of resins and fillers or other additives supplied to
a continuous
mixing apparatus preparing the viscoelastac material which was sampled. In
another
example of a DSR system in a control system, control can be had over the
processing
-5-
of a previously prepared viscoelastic material. In this instance the DSR
system ,
determines the fundamental viscoelastic properties or other measured
characteristics of
a sample from a batch of material, such as in forming a tire. The values of
some or
all of these properties or measured characteristics are then provided to a
controller
which may, for example, adjust the time or temperature of the next process.
While
this is an example of a feedforward control system, the DSR system could
similarly be
useyd in a feedback control system used in controlling the processing of a
material based
on the analysis of a sample of processed material.
The PROGP,AM I test procedure, performed in the dedicated processor
14, provides a torque vs. time response curve (see Figure 4), measures the
maximum
torque and the time for the stresses in the test specimen to relax beyond
certain set
amounts (time constants), and computes a summation of the relaxation torque
over a
certain time period, such as a two second period. The PROGRAM I test results
provide a convenient and efficient way of grading materials, such as for
quality control
purposes, and for predicting the suitability or processability of a material
for a certain
use. When the PROGRAM I test is used as a quality control mechanism, the
numerical results of several tests performed over a work shift or other period
of time
may be transferred to the general computer 16 for statistical analysis, such
as to
provide a mean or standard deviation of the results.
The PROGRAM II test procedure, also performed in the dedicated
processor 14, produces a log-torque vs. log-time curve of the relaxing
stresses in the
test specimen (See Figure 5). The PROGRAM II log plot provides a justification
or
explanation of the PROGRAM I test results, and also facilitates the
determination of
certain characteristics of the test specimen, for example, whether the test
specimen is
a branched or linear polymer. The PROGRAM II plot could, of course, be other
than
log based; the log-based plot, however, provides readily usable information in
a
convenient format and is thus employed.
Unlike the PROGRAM I or PROGRAM II tests, the PROGRAM III test
is preferably performed partly in the dedicated processor 14 and partly in the
general
processor 16. The dedicated processor 14 produces a preselected number of
coupled
torque and time data pairs which are transferred to the general processor 16
for further
processing. The general processor 16 determines from those pairs a number of
fundamental viscoelastic properties of the test specimen as a function of
frequency,
such as the storage modulus, the loss modulus, the complex viscosity and
another
-6-
property called tan-delta. These values are presentable in numerical and/or
graphical
format and may be presented in some cases as a function of one another, or
with
multiple tests and/or properties on one graph. (See Figures 6-9).
The numerical and/or graphical output of the PROGRAM I, PROGRAM
II or PROGRAM III tests may be displayed on a video monitor 20 or in a hard
copy
format on the printer 17 or plotter 18. When displayed on the video monitor 20
the
general processor 16 allows much of the graphical output to be massaged, such
as
through changing the degree of fit of the curve following the actual
coordinate data
points. The outputs of PROGRAMS I, II and III may be used to control various
apparatus, such as that used to make the material being tested, that employing
the
material being tested, etc.
Turning to a more in depth discussion of the DSR system 10 and its
components, and referring again to Figures 1-3, the general processor 16 is
preferably
an IBMTM compatible microcomputer, such as that manufactured by COMPACTM,
although other generalized computers, processors or the equivalent may be
substituted
with the appropriate modifications to the communication protocol. The general
processor 16 is provided with a video display device 20 or the like and a
keyboard 22
for interaction with an operator.
The dedicated processor 14 includes analog components for conditioning
an analog input signal from the DSR device 12 and digital components for
analyzing
and recording the digitized signal and providing digital output to the general
processor
16. The center of the digital components is preferably an INTELTM 8085
microprocessor although other microprocessors may be employed, as would be
apparent to one of ordinary skill in the art. A more complete description of
the
dedicated processor 14 is provided below.
The DSR device 12 includes a number of panels 30 forming a cabinet 32
which partly encloses the stator assembly 34, the rotor assembly 36, the
control
electronics 37 (shown in Figure 15), and a supporting structure 38 for
maintaining all
the elements in their respective places.
Fundamentally, when the DSR device 12 is used to test a material, a
quantity of such material is placed in the stator, the material is heated, the
stator and
rotor are brought together to apply compressive force to the material to
obtain a sample
having a constant thickness, and the above-mentioned deflection is applied.
Data is
taken, especially as the material relaxes after being deflected (subjected to
torsional
_7_
stress), and the data is processed and used in various ways, for example, as
are
described herein. Further details of the operation of the DSR are presented
below.
' On the front face of the DSR immediately below the stator assembly is an
upper control panel 40. The upper control panel 40 has a tripole switch 42 for
placing
S the DSR in MANUAL, AUTOMATIC or OFF mode, a power switch 43, and a pair
of tripole switches 44a and 44b; one located at the extreme left side and the
other at
the right side of the panel. The switches 44a and 44b open or close the stator
and
rotor assemblies 34, 36 as described below. The switches 44a and 44b are
biased in
a null position. To close the assemblies 34 and 36, which is actually
accomplished by
elevating the stator assembly 34 to a position slightly below the rotor
assembly 36, both
switches 44a and 44b must be held in the CLOSE position simultaneously. This
is a
safety feature which prevents an operator from accidently closing the
assemblies and
thus helps to prevent operator injury. A pair of indicator lamps 46 and 48
visually
indicate whether the rotor and stator assemblies are in the CLOSE or OPEN
position,
1S respectively. Gauges S0, S2 and S4 located near the top of the DSR device
14 provide
a visual readout of the closure, stroke and retract pressures, respectively.
An lower control panel S6 directly below the upper control panel 40
includes the rotor and stator temperature controllers S8, 60, which allow the
temperatures of the rotor and stator to be set, and a presetable stroke timer
62, which
controls the time period between closure of DSR and the deflection of the
viscoelastic
material to be tested.
The supporting structure 38, as shown in Figure 3, includes a base 64
from which extends vertically four support columns 66. Affixed atop the
vertical
support columns 66 is a secondary base 68. The base 64, vertical support
columns 66,
2S and the secondary base 68 provide a stable structure upon which the support
table 70
is secured.
Four cylindrical rods 72 arranged in a rectangular configuration and
extending vertically from the support table 70 maintain the rotor table 74 and
stator
table 76 in correct relative positioning. The rotor table 74 from which the
rotor
assembly 36 is mounted is non-movably secured at or near the top the rods 72.
Stator
table 74 upon which the stator assembly 34 is mounted is slideably secured on
the rods
72 by bushings 78 that allow the stator table 76 to slide vertically along a
portion of
the length of the rods 72. Vertical movement is achieved by a hydraulic
cylinder
assembly 80 positioned between the secondary base 68 and the stator table 76.
The
_g_
blind end of the cylinder assembly 80 is suitably secured to the secondary
base 68 and
the cylinder rod 82 extending from the cylinder is secured to the bottom of
the stator
table 76. Actuation of the hydraulic cylinder assembly 80 is accomplished
through the
control of pressure through inlet and outlet ports (not shown).
The maximum length upward of travel of the stator table 76 along the
cylindrical rods 72 is determined by two rods 84 extending vertically downward
from
the stator table 76 through clearance holes in the support table 70, as shown
in Figure
10. Shims 86 mounted to the bottom face 88 of the support table 70, such as by
machine screws 90, provide for precise control of the closure height of the
stator
assembly 34. Stops 92 secured to the end of the rods 84 by a nut 94 contact
the shims
86 when the stator table 76 has reached its maximum height to prevent further
elevation. The maximum elevation is further determined by abutment rods 96
which
extend from the stator 'table 76 along the vertical axis of the rods 84. The
abutment
rods 96 contact the vertically adjustable screws 98 which extend downwardly
from the
bottom face 100 of the rotor table 74.
A pair of closure gauges 102 are provided for visual measurement of the
closure height of the stator table 76 and attached stator assembly 34. The
closure
gauges 102 include downwardly extending gauge pins 104 which contact
vertically
adjustable elements 106, such as machine screws, mounted to stator table 76.
Precise information as to the vertical location of the stator assembly 36
is determined by limit switches, shown in Figure 11, and is provided to the
DSR
control electronics 3? (Figure 15). Lower limit switch 108, secured to the
stator table,
contacts the support table by pin 110 when the stator table 76 has reached its
lowest
position. Upper limit switch 112 and fine upper limit switch 114 are mounted
to a
common support 116 which is secured to the stator table 76. Upper limit switch
112
includes a vertically extending pin 118 which contacts the bottom face 120 of
the rotor
table 74 when the stator assembly 34 has reached a location .050 inches below
its final
closure height. Fine upper limit switch 114 includes an upwardly extending pin
122
which contacts a vertically adjustable contact 124 when the stator table has
reached a
location .005 inches from its final closure height. Electrical signal outputs
from the
limit switches 108, 112 and 114 are provided to the DSR device 12 control
electronics
37 which are discussed below.
-9-
Outer support elements 126 secured to the rotor table 74, the support table
70, and the base 64 provide an outer structure upon which the housing panels
30 are
secured.
Referring now to Figures 12 and 13, there is shown a cross-sectional view
of the stator and rotor assemblies 34 and 36. The stator assembly 34 includes
a stator
base 128 which is bolted to the stator table 76 with bolts 130. An insulating
member
132 interposed between the stator base 128 and the stator table 76 prevents
heat
transfer from the stator base 128 to the stator table 76. The stator base 128
includes
an annular recess 134 located in its bottom face 136 for the insertion of a
500 watt
ring-type electrical heating element 138, and a lateral recess 140 for the
insertion of
a temperature sensing device 142, such as a thermocouple. The stator base 128
further
includes a conical recess 144 in its top face 146 into which a specimen
container 148
is removably secured, such as with bolts 150.
The specimen container 148 includes a truncated cone shape cavity 152
in its top face 154 for the reception of the generally conical shape lower
portion 156
of the rotor 158 of the rotor assembly 36. The conical shape lower region 160
of the
cavity 152 is truncated at the upwardly extending cylindrical face 162. The
conical
shape lower region 160 may be provided with radially extending channels or
grooves
or some other suitable surface (not shown) which prevents rotation of the
specimen
contained therein relative to the specimen container 148.
The cone shape lower portion 156 of the rotor 158 and the cone shape
lower region 160 of the cavity 152 are of an equal conical angle.
Consequently, when
the stator assembly is raised to the closed position, the volume of specimen
material
sandwiched between the rotor and the specimen cavity will be of a constant
thickness
throughout. This facilitates the calculation of accurate test results without
requiring
that the specimen be subject to a constant stress throughout the specimen.
The stator base 128, the specimen container 148 and the rotor 158 are
constructed of a material having a high coefficient of thermal conductivity
and a high
degree of resistivity to chemical interaction with the specimen material, such
as a
chromium steel. The high thermal conductivity of the stator base 128 and the
specimen
container 148 in conjunction with the low thermal conductivity associated with
the
insulating member 132 provide a heat transfer path whereby heat generated by
the
heating element 138 is efficiently transferred to the test specimen located in
the
specimen container 148 without substantial loss to the stator table 76.
-10-
~~46~8~
The rotor assembly 36 aligned directly above the stator assembly 34, as
shown in Figure 13, includes an annular housing 164 extending through and
secured
to the rotor table 74 by the bolts 166. A rotor shaft 168 is disposed within
the annular
housing 164 and maintained in correct relative positioning by annular bearings
I70 and
I72. The bearings 170, 172 allow for angular rotation of the rotor shaft 168
within
the annular housing 164 while preventing axial movement. The outer surface 174
of
the rotor shaft 168 is stepped with each subsequent lower step increasing in
diameter.
This stepped outer surface 174 of the rotor shaft 168 allows the upward forces
on the
rotor shaft to be distributed among the bearing elements 170, 172.
Secured to the bottom of the rotor shaft 168 is the rotor 158. The rotor
includes a lower conical region 156 for insertion into the specimen container
148 and
a cylindrical shaft portion 176 broadening into a flange 178 for mounting to
the rotor
shaft 168. The conical lower portion 156 of the rotor 158 may be provided with
radially extending channels or grooves 180 to prevent the specimen from
rotating
1~ relative to the surface I56 upon deflection. The rotor 158 further includes
a vertical
passage I82 for the insertion of a 200 watt electrical heating element 184 and
an offset
vertically extending passage 186 for the insertion of a temperature sensing
device 188,
such as a thermocouple. The electrical leads 190 of the heating element 184
and the
thermocouple 188 extend upwardly through an axial passage 192 located in the
rotor
shaft I68 for connection to the DSR device 12 control electronics 37 described
below.
At or near the tog of the rotor shaft I68 is connected the rotor arm 200
as is shown in Figure 14. The rotor arm 200 extends perpendicularly to one end
of
the longitudinal axis of the rotor shaft 168 and is connected to a universal
joint 202.
At the other end of the universal joint 202 is connected an extension arm 204
which
is in turn connected to the load cell 206. The Universal joint 202 connection
between
the rotor arm 200 and the extension arm 204 allows the torque transferred from
the
rotor I58 through the rotor shaft 168 to the rotor arm 200 to be translated
into a linear
force through the extension arm 204 to the load cell 206 without imparting
stresses
involved with a bending moment on the load cell 206. Consequently, the load
cell 206
accurately measures torque exerted on the conical face 156 of the rotor 158 by
the test
specimen.
The load cell 206 may be a conventional device which produces an output
that is electrical in nature or is able to be read, sensed, detected or used
in ar with an
-11-
electrical device. One example is a resistor, the resistance of which varies
as a
function of force applied thereto or deformation thereof. Other examples are a
piezoelectric device, a variable capacitor, etc.
The load cell 206 preferably is coupled in a whetstane bridge
configuration (not shown) and will be referred to below as including or being
a .
whetstone bridge type load cell. Of course, other Load cell devices and
circuits may
be used to detect or sense mechanical input and to convert such input to an
electrical
output or response of desired resolution. Thus, a similar element or elements
capable
of converting mechanical stress to electrical energy, which generates an
electrical signal
linearly proportional to the resisting torque on the rotor surface 156 may be
employed.
The electrical signal is conducted to the dedicated processor 14 over the
wires 208.
The load cell 206 is also connected to the pneumatic cylinder 210 through
the cylinder rod 212. The pneumatic cylinder 210 includes two ports 214 and
216
which are connected to pneumatic supply/exhaust lines (not shown). Applying
pressure
to port 216 while exhausting pressure from port 214 forces the cylinder rod
212 to
extend from the pneumatic cylinder 210. This in turn causes the rotor 158 to
rotate
a predetermined angular distance in the counterclockwise direction. This
angular
distance is controlled by the axial distance between an annular stop 218
mounted to the
cylinder rod 212 and an abutment element 220, secured to the cylinder base
224, when
the cylinder rod 212 is in its fully retracted position. The angular position
of the rotor
100 when the annular stop 218 is abutting the abutment element 220 is the null
position
of the rotor 158, corresponding to the position of the rotor at the start of
the DSR test.
Conversely, applying pressure through the port 214 while exhausting the
port 216 forces the cylinder rod 212 to retract into the pneumatic cylinder
210, thus
causing the rotor 158 to rotate clockwise an angular distance corresponding to
the axial
distance between the abutment element 220 and the annular stop 226, mounted to
the
cylinder rod 212, when the cylinder rod 212 is in its fully extended position.
Preferably this axial distance corresponds to a 2 degree rotation of the rator
158. It
takes approximately 5 milliseconds to effect the complete 2 degree rotation
and thus .
the deflection of the specimen.
The degree of angular rotation of the rotor 158 may be decreased by
mounting a shim (not shown) of appropriate width to the face of the abutment
element
220 proximate the stop 226. Consequently, the travel of the cylinder rod 212
from its
-12-
20~~~8'~
fully extended position towards its retracted position will be shortened by
the width of
the shim. It has bin found that for some materials it is advantageous to
effect a
deflection of only 1 or 1/2 degree. To affect a 1 degree deflection, for
example, the
inserted shim would be of a width corresponding to one-half of the axial
distance
between the stop 226 and the abutment element 220, without the shim in place,
when
the; cylinder rod 212 is in its fully retracted position; thus shortening the
travel of the
rod to one-half of its normal travel.
Referring now to Figure 15 there is shown a diagram of an electrical
circuit 300 for the DSR device 12 control electronics 37. Basically, the
circuit 300
controls the upward or downward movement of the stator assembly 34, the
impulsive
rotation of the rotor 158, and the heating and temperature control of the
stator base 128
and the rotor 158. The circuit 300 is connected to a 115 volt 20 amp
alternating
current power supply 302 through the power switch 43. A 15 amp fuse 304
provides
overall overcurrent protection for the circuit 300. When the DSR device 12 is
in an
ON state, such as when switch 43 on the indicator panel 40 is in the ON
position,
switches 43a and 43b are closed. This provides power to the remainder of
circuit 300
including the auxiliary power outlets 306.
While power is applied to the circuit 300, the stator temperature controller
308 continuously monitors the temperature of the stator assembly 34 as
indicated by
the output signal of the thermocouple 142. Based on the indicated temperature,
the
stator temperature controller 308 will cause the solid state switch 310 to
assume an
opened or closed position. When the solid state switch 310 is in its closed
position,
supply power from line 312 is conducted to the 500 watt ring type electrical
heating
element 138 in the stator base 128. A 10 amp fuse 314 provides overcurrent
circuit
protection. When the stator temperature controller 308 determines that the
stator
assembly 34 has reached its desired temperature, the solid state switch 314 is
opened
interrupting power to the 500 watt stator heater 138.
Likewise, the temperature of the rotor 158 is controlled by the rotor
temperature controller 316 based on the output signal from the thermocouple
188 in the
rotor. When the rotor 158 is below the selected temperature, the solid state
switch 318
closes providing supply power from line 312 to the 200 watt rotor heater 184
to heat
the rotor. A 10 amp fuse 320 provides overcurrent safety protection for this
area of
the circuit.
-13-
v'; .: : ; , ;:.. ; ,,. ., .. . ': ~ '>. .. . . . . _ ., , , . . . >
2~~~38'~
When the DSR device 12 stator and rotor assemblies 34 and 36,
respectively, are in the fully open position, the lower limit switch 108 has
pin 110
depressed against the support table 70 thus closing the switch 108A and
permitting
power from line 312 to flow to and light OPEN indicator lamp 48 located on the
front
face control panel 40. The switches 44A and 44B are biased in a null position
and do
not conduct power absent simultaneous actuation by an operator.
The lower limit switch 108 is also equipped with a switch 108b which
engages contact 322 when the pin 110 is depressed against the support table 70
or
contact 34 when stator table 128 is elevated so pin 110 is not depressed by
support
table 70. With the DSR device 12 in its fully open position, switch 108b is in
position
to engage contact 322 thus supplying power from line 312 through the closed up
latch
contact 326 to activate the down solenoid 328 and down latch 330 to maintain
the stator
assembly 34 in its fully lowered position. The activated down latch 330 opens
the
normally closed down latch contacts 332 preventing current flow to the up
solenoid 334
absent action by an operator.
To start the test an operator simultaneously holds the switches 44a and
44b in their up, or CLOSE, position, thus power is supplied from the supply
line 312
to the up solenoid 334 and the up latch element 336. Energizing the up latch
element
336 opens the normally closed up latch contacts 326 thus preventing current
flow to
the down solenoid 328 and down latch 330, and closing the down latch contacts
332.
This permits the upward movement of the stator table 76 while pin 110 of the
lower
limit switch 108 is still depressed and switch 108b is in position to engage
contact 322.
With the switches 44A and 44B held in their CLOSE positions, power is
continually
applied to the up solenoid 334 to raise the stator assembly 34 towards the
rotor 158.
When the stator assembly 34 reaches a location approximately .050 inches
from the final closure height, the upperly extending pin 118 of the upper
limit switch
112 will be depressed against the bottom face 120 of the rotor table 74
closing the
switch 112a and allowing power to flow from line 312 to light the CLOSED
indicator
lamp 46 located on the control panel 40. Simultaneously, the switch I 12b of
the upper
limit switch 112 will also close and power will flow from line 312 through the
closed
down latch contacts 332 through the upper limit switch 112b to the up solenoid
334,
bypassing operator switches, 44a. After once the stator assembly 34 has
reached a
height .005 inches from the final closure height, the pin 122 of the fine
upper limit
-14-
~~~6~8'~
switch 114 will be depressed by the adjustable contact 124 thus closing the
switch
114a. In the automated test case the switch 42 on the control panel 40 will be
in the
AUTOMATIC position (engaging contact 338) and power will be supplied from the
supply line 312 through the upper limit switch 112a through the tripole switch
42 and
contact 338, and through the fine upper limit switch 114a to the presetable
preheat
timer 340. The preheat timer 340 will begin to count down to zero from a
preset value
that is chosen to ensure that stator and rotor temperature controllers 308 and
316,
respectively, have had sufficient time to bring the test specimen to the
desired
temperature. The up solenoid 334 will continue to elevate the stator assembly
34 until
it reaches its final closure height.
When the preheat timer 340 times out, the timer contacts 342 are closed
thus supplying power from the supply line 312 to the exhaust solenoid 344 and
the
exhaust timer 346. Activation of the exhaust solenoid 344 vents the pneumatic
cylinder
port 216 to ambient. The exhaust timer 346 invokes a 5 second pause, ensuring
that
the pressure in the pneumatic cylinder reaches ambient. At the end of the 5
second
time-out, the exhaust timer contacts 348 are closed supplying power from the
supply
line 312 to the stroke solenoid 350 causing 70 psi of pressure to be applied
to port 214
of the pneumatic cylinder 210. This causes the cylinder rod 212 to retract
which, in
turn, causes the rotor 158 to rotate the preset amount. At this point the test
proceeds
with the load cell 206 generating a signal proportional to the torque exerted
on the
rotor 158 by the test specimen and sending that signal to the dedicated
processor 14 for
processing as described below.
If the DSR device 12 is operating in manual mode, the operator actuates
switch 42 on the upper control panel 40 to the MANUAL position once the stator
assembly has reached the final closure height. This causes switch 42 to engage
contact
352 thus supplying power to the exhaust timer 346 and exhaust solenoid 344 via
line
354. The operation from this point on is as described above relative to the
automatic
mode. Note that the DSR device is preferably only operated in the manual mode
to
perform set-up operation on the device and not during an actual material test.
After the test is completed, the operator simultaneously holds both
switches 40A and 40B in the OPEN, or LOWER, position to lower the stator
assembly
34. When the switches 40A and 40B are held in the OPEN position, power is
conducted from supply line 312 through the lower limit switch 108b, which is
in
position to engage contact 324, to the down solenoid 328 and down latch 330.
The
-15-
2~~6~~'~
down latch 330 latches the down latch contacts 332 in an open position thus
preventing
eu:rrent flow to the upper solenoid 334 and up latch 336 regardless of the
position of
the upper limit switch 112a and contacts 332. Simultaneously, the down
solenoid 328
is energized and the stator assembly 34 continues to lower while the switches
44a and
44b are both held in the OPEN position.
One safety feature of the circuit 37 is that travel of the stator assembly is
prevented absent direct action by the operator. When the stator assembly 34 is
in its
fully open, or fully lowered, position the up latch contacts 326 are in a
closed position
and the lower limit switch 108b is in a position to conduct current to the
down latch
328 and down solenoid 330 thus preventing any movement of the stator assembly
until
both switches 44a and 44b are simultaneously held in the up or CLOSE position.
At
any time during elevation of the stator assembly 34 prior to reaching a
location .050
inches from full closure height or at any point during lowering of the stator
assembly,
movement of the stator assembly is stopped if either of the switches 44a or
44b is not
held in position.
Referring now to Figure 16 there is schematically illustrated an electrical
circuit data flow diagram for the dedicated processor 14. The dedicated
processor 14
electronics include analog circuitry for processing an analog signal from the
load cell
206 prior to digital conversion, and digital circuitry for analyzing the
digital
information to determine values indicative of various viscoelastic properties
of the test
specimen. The digital electronics include a computer processing unit 402, such
as an
Intel 8085 microprocessor, read-only program memory 404, random access storage
memory 406, an analog to digital converter 408, plus a number of other
components
as will be described more fully below.
Discussing initially the analog components or analog component package
of the circuit, the electrical components of the load cell 206, such as a
whetstone
bridge, are shown schematically in Figure 16 at 206. The torque signal
generated by
the load cell 206 is a function of the excitation signal voltage 410, as well
as the axial
force on the cell. Accordingly, span circuitry 412 is provided to allow the
supply
power 414 to the load cell 206 to be adjusted, within a range of +/-5 percent
of 5.25
volts, to accommodate inevitable sensitivity variances among different load
cells.
Reference numeral 414 may be used here and elsewhere to designate an
electrical line, group of lines, connection or output or to designate the
signal thereon.
Generally, if there is any significance to the difference betw~n such
designation, it
-16-
20~fi38°~
will be clear from the context. Similar uses of reference numerals occur with
respect
to other lines and signals thereon, as will be evident.
Upon angular deflection of the test specimen, the load cell 206 generates
an analog signal on line 416 which is provided to a differential amplifier 418
along
with the zeroing signal on line 420 from the conventional manual zero 422 or
offset
device. The amplifier 418 sums the zero signal 420 with the load cell signal
416 to
provide for calibration of the load cell 206. The resultant signal is
amplified with a
gain of 166 since the excitation output voltage of the load cell is relatively
small.
The output signal 424 of the amplifier 418 is provided to a low pass filter
426 with a cutoff frequency of 50 hertz to filter out high frequency
components of the
output signal 424. The resultant signal 428 is summed in a summer 430 with an
offset
signal 432 from a program offset circuit 434 relating to whether the CPU 402
is
executing PROGRAM I or PROGRAM II/PROGRAM III as will be described fully
below. The output signal 436, having a correlation of 20 millivolts to one-
inch pound
of torque, may be tapped for use by a strip recorder or for use as a test
signal 438.
A scaling unit 440 scales the output signal 436 of the summer 430 with
a gain adjustable between zero and 1.5 to provide the analog signal 442. The
gain of
the scaling unit 440 is adjusted to absorb errors imposed on the analog torque
signal
through the cumulative tolerances of the various components of the analog
component
package and to provide scaling to the A/D converter 408 of one count equals
0.1 in
Lbsr torque. In this manner any selection of analog components can be adjusted
to
perform to consistent specifications, thus allowing the interchange or
replacement of
analog component packages without affecting the performance of the DSR. The
scaled
and filtered torque signal 442 is now suitable for digital conversion by the
12 bit analog
to digital (A/D) converter 408.
The AJD converter 408 and scaled signal 442 are connected through a test
switch 444 which allows the scaled signal 442 to be replaced by the test
signal 446
under appropriate conditions. The test signal 446 is a known waveform which
mimics
an actual DSR test torque versus time curve. Since the waveform, and thus the
data
points along the waveform, are known, the test results are also known and can
be used
to determine whether the digital components of the dedicated processor 14 are
functioning properly by whether they compute the expected known values.
The test signal 446 is an analog recreation of a series of data points stored
in program memory 404. The CPU 402 accesses the locations in locations memory
-17-
2~~~~~°~
404 containing the test digital data points via an address bus 448. The data
points are
returned to the CPU 402 over a data bus 450. The CPU 402 then transfers the
test
data to a 12 bit digital to analog (D/A) converter 452 via a 12 bit latch 454.
The D/A
converter 452 converts the latched 12 bit test data into an analog test signal
446. The
test signal 446 is provided to the test switch 444 through the buffer 456,
which
increases the amplitude of the test signal commensurate to that of the scaled
signal 442,
for connection to the A/D converter 408 during test conditions.
The position of the test switch 444 is transparent to the operation of the
A/D converter 408 which converts an analog input signal to a digital word
regardless
of whether that signal represents actual torque relaxation data from the load
cell 206 ;;
or recreated test data. The A/D converter 408 having twelve information bits
has a
resolution of 4096 nondimensional counts. In the exemplary embodiment a 500
Lbf
load cell 206 will, under operating conditions, experience a maximum of 200
pounds
of force, and the desired resolution is .1 inch-Lbf; therefore, it is
convenient to allocate
4,000 of the 4,096 available counts in the A/D converter 408 for the
conversion of
positive torque. Consequently, 1 count equals .1 inch-Lbsf. For a PROGRAM I
test
the measured torque is always positive, therefore, the remaining 96 counts go
unused.
When performing a PROGRAM II or PROGRAM III test, however, it is possible to
have a measurement indicating a small negative torque. In these cases, the
remaining
96 counts are used as an offset, equalling 9.6 inch-Lbsf, to compensate for
any small
negative torque values.
The rate at which the A/D converter 408 converts the input analog signal
to a 12 bit digital data word is either 1,000 or 4,000 conversions per second.
(Other
rates may be employed equivalently.) The rate at which the A/D converter
operates
is determined by the precision timer 458. The timer 458 produces a 1,000 or
4,000
hertz clock signal which is sent to the A/D converter to produce the desired
conversion
rate over the line 460.
Each 12 bit digital data word converted by the A/D converter 408 is
passed to a programmable peripheral interface device 462 via the 12 bit data
bus 464.
The programmable peripheral interface 462 temporarily stores the digital data
words
in a data buffer section of the interface 462 prior to access by the CPU 402.
The CPU 402 accesses a data word from the programmable peripheral
interface 462 by addressing the proper location in the data buffer via the
address bus
448. The data word is returned to the CPU 402 over the data bus 450. Each data
-18-
2~4~38°~
word obtained from the programmable interface 462 is examined by the CPU 402
and
then placed in a location in the static random access storage memory 406.
Forty
kilobytes of accessible memory are provided for the accumulation of torque and
elapsed
time data for one or more DSR tests. The storage memory 406 is provided with a
battery backup 466 to prevent the loss of stored data when power is
interrupted to the
memory. Alternatively, the storage memory could be implemented with chips
using
EI3PROM technology.
The CPU 402 may provide information to the plotter 18 or printer 17 via
an opto-isolator 468. The CPU 402 is also provided with a calendar clock 470
through
an 8 bit latch 472 to allow data to be tagged with the time the test occurred
and to
permit the CPU 402 to transfer test data at a predetermined time. The CPU 402
may
transfer stored data from the storage memory 406 to the general computer 16
through
a UART 474 and opto-isolator 476.
The CPU 402 interfaces with the 14 kilobyte program memory device 404
via the address bus 448 and data bus 450. The program memory 404 aside from
storing the test data, also stores the executable computer program code for
PROGRAM
I, PROGRAM II, a portion of PROGRAM III, and other related computer program or
operating codes necessary to control the dedicated processor 14. The program
memory
device 404 further includes a number of ports for the control of several LED
displays
478 indicating such information as whether a test is running 480, whether the
first time
constant has been reached 482, whether the second time constant has been
reached 484,
if data is being stored 486, and a ready LED 488 which indicates when a
PROGRAM
II or PROGRAM III test is complete. Another port operates a 4 digit digital
display
490 through a display driver 492 to provide a periodic visual display of
torque
information for use in calibrating the load cell 206 via the manual zero 422.
Further
ports interface with function switches 494 such as an additional keyboard 496,
a reset
switch 498, a zero switch 500, and a program switch 502 the functions of which
are
described below.
A digital power supply 504 provides 5 volt power to the digital
components. An analog power supply 506 provides +/- 15 volt power to the
analog
components of the dedicated processor 14.
Referring now to Figures 17a-17f there is shown a schematic electric
circuit diagram for the analog and digital electronics of the dedicated
processor 14.
The drawings may be placed side by side to depict continuation of various
electrical
-19-
lines or connections from sheet to sheet to complete the circuit. The load
cell 206
(Figure 17f) provides an analog signal to a series of operational amplifiers
550, 552,
554 (Figure 17e) which perform the functions of the differential amplifier
418, the low .
pass filter 426, the summer 430 and the scaler 440 described above with
respect to
Figure 16. The resulting signal is provided to the A/D converter 408 (Figure
17d) r
through the test switch 444 (Figure 17f). The A/D converter 408 provides 12
bit '
words to the programmable peripheral 462 which is implemented as a scratch pad
'
RAM having two 8 bit ports and one 6 bit port. The information in these ports
is
accessible by the CPU 402 (Figure 17a), which is preferably an Intel 8085
microprocessor, over the combined address/data bus 616. The CPU 402 operates
at
approximately 3 megahertz.
Preferably three Intel proprietary 2 kilobyte electronically programmable
read-only memory (EPROM) integrated circuits or chips 556a, 556b and 556c form
the
program memory 404. Each chip has two 8 bit input/output ports for interface
with
the function switches 494, and the LED displays 478 (Figure 17a), the D/A
converter
452 and 4 digit display 490. ('The terms integrated circuit, chip, etc., may
be used ,
equivalently herein.) These EPROM chips 556a, 556b and 55tx (Figures 17a and
17b)
are accessible by the CPU 402 over the combined address/data bus 558. Each
Intel
EPROM chips 556a, 556b and 556c contains means for latching out the lower
eight
address bits from the combined address/data bus 558. However, if the functions
of the
EPROM chips are implemented with other commercial EPROMs, it may be necessary
to include additional hardware to latch out the lower eight address bits, such
as is
described below relative to EPROM chips 568a and 568b (Figure 17e). A chip
select
decoder 559 (Figure 17c) enables the EPROM 556a, 556b or 556c allocated with
the
storage and port addresses to be accessed by the CPU 402.
One 8 bit port of the EPROM chip 556a provides digit information for the
four digit LED display 490 (Figure 17a) through the display driver 492 (Figure
17b).
The other 8 bit port of EPROM 556a monitors the program switch 502 and the
zero
switch 500 (Figure 17a). One of the 8 bit ports of the EPROM chip 556c
controls the
LED display lamps 478 (Figure 17a). A number of bits of the other port
controls the
display driver 492. A single chip 560 (Figure 17c) provides the 12 bit D/A
converter
452 and the 12 bit latch 454 functions to produce the analog test signal. The
chip 560
is provided 12 bits of digital data from the output ports of the EPROM chip
556b over
the bus 562. 'Ifie generated analog test signal is sent over the lines 564 to
the scaling
-20-
201~6~~'~
operational amplifier 566. The resultant signal is coupled to the A/D
converter 408
through the test switch 444 (Figure 1?fj.
Two standardiaed commercially available 4 kilobyte EPROM chips 568a
anal 568b (Figure 17e) add an additional 8 kilobytes of read-only memory, thus
totaling
the fourteen kilobytes of program memory 404. These EPROM chips 568a and 568b
are also accessed across the address/data bus 558. However, since these are
standardized commercial EPROM chips, a latch 570 (Figure 17d) is needed to
latch out
the Iower eight address bits from the address/data bus 558. These lower eight
address
bits are sent to the EPROM chips 568a and 568b over the address only bus 572.
A
chip select decoder 574 (Figure 17b) determines which EPROM chip 568a or 568b
contains the sector of memory being addressed.
Five 8 kilobyte by 8 bit static random access memory (RAM) chips 576a,
576b, 576c, 576d and 576e (Figures 17b-17d) provide the 40 kilobytes of random
access memory comprising the storage memory 406. Preferably these RAM chips
576x-a are standardized commercially available chips, and thus, the addressing
scheme
described above relative to the EPROM chips 568a and 568b is again necessary.
Accordingly, when the CPU 402 addresses a memory location in one of these
chips
across the address/data bus 558, the latch 570 will latch out the lower eight
address bits
which are provided to the chips 576a-a over the address only bus 5?2. A chip
select
decoder 578 enables the appropriate RAM chip containing the address to be
accessed.
The chip select decoder 578 is provided conditioned power through the opto-
isolator
580 (Figure 17a) which disables the RAM chips 576a-a when power is low, thus
preventing data changes in the RAM chips when power is turned ON or OFF.
The timing signals for the EPROM chips 556a, 556b, 556c, 568a and
568b, the RAM chips 576x-c and the UART 500 is generated by the timer chip 582
(Figure I7e).
The specific implementation of the functions of the dedicated processor
14 in hardware, as shown in Figures 17a-f and described above, is but one way
in
which the dedicated processor could be constructed. Other implementations of
the ,
dedicated processor and arrangements and selection of specific chips and
components
which accomplish the same objective will be apparent to one skilled in the
art.
OPERATION
Referring now to Figures 18-25 and 29-34 there are shown several
flowcharts illustrating the operation of the general processor 16 and
dedicated processor
-21-
Y
~o~s~~~
14. Various steps in the operation will be referenced in the following
discussion by
a number contained within parentheses Q which corresponds to an identical
number in
the figures. Note that as used in the figures an arrow pointing toward a
letter or
number indicates that the routine will jump to the step on the flowchart
indicated by
another arrow extending from the same letter or number encircled at another
location
on the flowchart. The various flowcharts illustrate a preferred structure
that, given the
ensuing discussion, one of ordinary skill in the art could reduce to computer
program
code suitable for execution by the resident processor or some other computer
program
executing device or system.
To begin a DSR test the power must first, of course, be turned on via
power switch 43 if it has not been turned on already. Upon power up, the
general
processor 16 and dedicated processor 14 will perform their initialization
routines. The
general processor 16 will perform those functions conventionally performed at
power
up by an IBM compatible microcomputer. The dedicated processor 1~ wm execute a
software reset of the CPU 402 (step 800, Figure 20) thus zeroing its program
counter
and causing it to perform its initial functions, such as clearing its memory
and setting
up its ports (802), and preparing for communication with the general processor
16
(804).
The operator will then weigh or measure out an approximate amount of
the test specimen and place it in the cavity 152 of the specimen container
148. The
rotor and stator temperature controllers 58, 60, respectively located on the
lower
control panel 56 on the front face of the DSR 12 are set to the temperatures
that it is
desired that the test specimen be heated. Preferably, both temperature
controllers will
be set to approximately the same temperature to ensure an even temperature
throughout
the test specimen. The stroke timer 62, also located on the lower control
panel 56, is
set with a preheat time selected large enough to ensure that the test specimen
will reach
the temperature selected on the temperature controllers 58, 60 before being
deflected.
The operator will also set the external program switch 502, located on the
front face
of the dedicated processor 14, to a position indicating whether PROGRAM I, or
PROGRAM II or PROGRAM III is to be performed.
Once the general processor 16 has completed its initialization functions,
the operator may then command the general processor to begin execution of the
desired
program. The executable code within the general processor is generally divided
into
or considered as two segments: the first, which executes PROGRAM I or PROGRAM
II, is run by entering, or typing the characters, DSRI or DSR2 at the
operating system
level; the second, which executes PROGRAM III, is run by entering DSR3 at the
' operating system level. The reason for the distinct segments of code is that
since
PR OGRAM I and PROGRAM II may be executed solely within the dedicated
processor
14, the DSR may be configured to operate under certain embodiments without the
action of the general processor. However, to perform the functions required by
PROGRAM III, and to allow a high degree of operator interaction, if desired,
PROGRAM III requires the use of the general processor. Depending on the
command
entered, the general processor will display main menu 1 (Figure 18, 600),
corresponding to PROGRAM I or PROGRA11Z II, or main menu 2 (601),
corresponding to PROGRAM III. Notice that many of the selections provided by
the
menus are identical, such as calibrating the DSR system, changing the system
configuration, performing an operation verification function, or returning to
the
operating system.
Before examining the functioning of the code which performs the selected test,
PROGRAM I, PROGRAM II, or PROGRAM III, the common functions will be
discussed. Exiting to the operating system, in this case, DOS, is self
explanatory. It
allows the operator to switch from performing PROGRAM I or PROGRAM II tests to
a PROGRAM III test, or vice-versa. It also allows other functions or programs
typical
of a microcomputer to be executed. The DEAD WEIGHT CALIBRATION option
(generally, 602) provides a method of calibrating the DSR system. Briefly, a
known
weight is suspended and connected to the system in a fashion which causes it
to exert
a known force on the load cell. Through interaction with the dedicated
processor 14,
the corresponding torque measured by the DSR system is displayed on the 4
digit LED
display 490. Based on that display a technician can calibrate the system, such
as by
adjusting the manual zero 422 and span 412 to produce the correct value. The
VERIFY option (604) allows the operator to interact with the dedicated
processor 14
(see Figure 23) through the general processor 16, such as to set the start
level or to
perform diagnostics. Similarly, the CHANGE CONFIGURATION option (606, Figure
18) allows the operator to modify system default values, such as degree of
curve fit,
rotor height, shift times, etc., some of which are also transferred to the
dedicated
processor I4.
An option available from menu 1 only is the INPUT PROGRAM II DATA
FILE option, which as is described more fully later, allows the operator to
enter a
-23-
20~~ 3~"l
previously stored PROGRAM II data file for graphical manipulation, such as by
changing the curve fit or scaling factors. An option available from the
PROGRAM III
menu, main menu 2, only is the CHANGE DSR FIXTURE DIMENSI01~1S option.
This option allows an operator to change critical geometric values, such as
those which
comprise the form factor. Since these values are necessary only to the
conversion
performed in PROGRAM III from torque to shear stress relaxation values, and
the
determination of the fundamental viscoelastic properties derived therefrom,
modification of the values are not permitted when running PROGRAM I or
PROGRAM II.
To perform a DSR test of a viscoelastic material the operator will select the
specific test desired to be performed, PROGRAM I, PROGRAM II or PROGRAM III,
from applicable main menu (600 or 601 Figure 18). If PROGRAM I or PROGRAM
II has been selected, the general processor 16 will jump to a routine that
controls the
test functions of the dedicated processor 14 (Figure 19). That routine then
prompts the
operator to enter information identifying the material to be tested and
various test
parameters (702), such as the test temperature (i.e., the temperature entered
into the
rotor and stator temperature controllers 58, 60). The routine will then
attempt to
establish communication with the dedicated processor 14 over the RS-232 data
bus
(704). If communication is not successful, the processor will instruct the
operator to
reset the dedicated processor 14 (706). This executes a hardware reset of the
dedicated
processor 14 (see Figure 20, step 806) and causes the dedicated processor to
reset its
memory and to set its ports (802) and to prepare for communication again
{804). The
general processor routine then attempts communication again (704, Figure 19).
If
communication is established, an appropriate message will appear on the
monitor (708)
and the routine will begin to serially transfer a command to the dedicated
processor 14
indicating whether the PROGRAM I, or PROGRAM II test routines are to be
executed
(710).
If PROGRAM III has been selected, the general processor will immediately
establish communication with the dedicated processor 14 and begin to serially
transfer
to the dedicated processor the command to ex~ute PROGRAM III code and the
number of data points to collect.
The dedicated processor circuitry will obtain the command or commands from
the general processor 16 in a bit string format over the data bus (808, Figure
20), and
the bits will be assembled to foxm individual input commands (810). Based on
the
_24-
204~3~~
assembled command, the dedicated processor 14 will transfer program control to
the
appropriate routine to execute the PROGRAM I, PROGRAM II or PROGRAM III test
routines (812).
If the PROGRAM I test routine has been selected for execution (see Figure
21), the routine will request and obtain the header information from the
general
processor (810-812) and store that information in its storage memory 406
(814). The
routine will then command the precision timer 458 to begin generating a 4000
hertz
clock signal which is set to the A/D converter 408 to establish the digital
conversion
rate of 4000 12 bit words per s~ond (816). The routine then clears a flag to
indicate
that a valid test has not yet begun. (Herein the phrase "valid test" is used
to denote that
the DSR device has actually deflected the test specimen.) The CPU 402 then
begins
accessing words from the programmable interface device 462, checking the start
flag
to determine if a valid test is in progress (818) and comparing each word
against a
preset value to determine whether a valid test has begun (820), as will be
discussed
more fully below. Since a valid test should not have yet occurred, the routine
will stay
in a loop (818-820) accessing words and checking the start flag (818), which
will
indicate that a valid test is not in progress, comparing the newly accessed
word to a
start level (820), which the word should not exceed since a valid test is not
in progress
(818-820), and then accessing another word to begin the loop again.
If PROGRAM II or PROGRAM III has been selected (see Figure 22), the
routine handling these tests will also request and obtain necessary
information from the
general processor (852-854) and store that information again in the storage
memory
406 (856). The precision timer 458 will now, however, be commanded to begin
generating a 1000 hertz clock signal which is sent to the A/D converter to
establish the
digital conversion rate of 1000 12 bit words per second (858). The start flag
will be
cleared to indicate that a test is not in progress. The CPU 402 then begins
accessing
words from the programmable interface device 462, checking the start flag to
determine if a valid test is in progress (860) and comparing each word against
a preset
value to determine whether a valid test has begun (862), as will be discussed
more fully
below. Since a valid test is not in progress the routine will loop (860-862),
as is
described above relative to PROGRAM I, waiting for the test to begin.
From the time that the dedicated processor 14 has been commanded by the
general processor 16, through an operator selection from the main menus, to
perform
a PROGRAM I, PROGRAM II or PROGRAM III test, the dedicated processor will
-25-
2~~63~'~
continue to convert the input signal from the load cell 206 to corresponding
digital
words which are analyzed by the selected routine to determine whether the test
sample
has been deflected.
It should be noted that to this point many of the functions performed by the
operator, such as setting the temperature controllers, selecting the test to
be performed
from the main menu, placing the test specimen in the specimen cavity, etc.,
may be
performed in a different order than that presented above. The order depicted
above is
illustrative only; other sequences which accomplish the same end will be
apparent to
one skilled in the art and are included in the scope of the invention.
Once the operator has entered all information requested by the general
processor 16 and that information has been transferred to the dedicated
processor 14,
the general processor will display a message on the monitor indicating that
the DSR
is ready to begin a test (712, Figure 19). Assuming all other operator
functions have
been performed as outlined above, the operator may then raise the stator
assembly by
simultaneously holding the switches 44a and 44b in the CLOSE position. This
actuates
the hydraulic cylinder 80 causing it to elevate the stator table 76 and stator
assembly
34 toward the rotor 158.
As the stator table 76 reaches a position approximately .050 inches below
final
closure height, the upper limit switch 112 contacts the bottom surface 120 of
the rotor
table 74, whereby automated control of the elevation of the stator assembly 34
will
raise it to a predetermined distance below the rotor 158. At this point the
closure
gauges visually indicate to the operator that the stator assembly 34 has
reached the
height where continued elevation is performed automatically and the operator
may then
release the switches 44a and 44b.
From this point on the DSR test is completely automated and requires no
operator participation until the test has been completed. This greatly
improves the
reliability and accuracy of the test results and the calculated fundamental
viscoelastic
properties by eliminating any variables attributable to human interaction,
Also, since
there is no operator interaction required during the critical period of the
test (i.e., the
time period from just prior to the deflection of the test sample until the
output of the
results) the level of skill and attention required of the operator is at a
minimum.
As the stator assembly 34 approaches its final closure elevation, the conical
surface 15C of the rotor 158 enters the cavity 152 of the specimen container
148 and
compresses the test specimen contained therein. This forces discontinuations,
such as
-2tr
2~4~3~'~
voids or bubbles, out of the test specimen and extrudes excess
specimen material
between the outer edge of the surface 156 of the rotor 158, and
the cylindrical surface
162 of the specimen container 148. The heating element 184 in
the rotor 158, now in
a
intimate contact wikh the test specimen, thus begins contributing
to evenly heating the
test specimen sandwiched between the rotor surface 156 and cavity
152.
Once the stator assembly 34 reaches a position just below its
final elevation,
" a second upper limit switch, the fine limit switch 114, contacts
the adjustable contact
124 and electrically closes thus starting the preheat stroke timer.
This causes the D8R
to pause for a preset preheat period (the time entered into the
stroke tamer 62) to
ensure that the temperature of the specimen material has reached
its preset temperature.
. During this period the stator assembly will reach its final
elevation,
and the residual
stresses, created during the compression of the test specimen
material contained
between the conical surface 156 of the rotor 158 and the surfaces
160, 162 of the
specimen container 148, will substantially dissipate.
Air pressure maintaining the cylinder rod 212 in its extended
position is then
exhausted through port 216 and five seconds are allowed to elapse
to ensure that the
pressure in the cylinder has returned to ambient. At that time
an impulse of
approximately 70 psi of pressure is introduced to the cylinder
210 through port 214 to
retract the cylinder rod 212. The axial retraction of the cylinder
rod 212 is translated
into angular rotation by the universal joint 202 and rotor arm
200 through the rotor
shaft 168 thus causing the desired rotation of the rotor 158.
' Upon the impulsive rotational deflection of the rotor 158 the
test specimen
contained between the surface 156 of the rotor and the surfaces
160, 162 of the
specimen cavity 152 will deform and exert a resisting torque upon
the surfaces 158,
160 and 162. This torque is translated back through the rotor
shaft 168 and rotor arm
200 to the universal joint 202 where the torque is converted to
an axial force exerted
through the extension arm 204 to the load cell 206. The stress
sensing element or
elements of the load cell 206 will convert the mechanical stress
developed by the axial
force into an analog signal that is proportional to the axial
force. This analog signal
generated in the load cell 206 is sent over lines 208 to the dedicated
processor 14 for
processing in accordance with the previously selected test, PROGRAM
I, PROGRAM
II, or PROGRAM III, as is discussed individually below.
PROGRAM I
Digressing briefly to the period in time just before the test specimen is
deflected and with reference to Figure 21, it is seen, as mentioned above,
that the
PROGRAM I test routine has already requested and obtained header information
from
the general processor 16 (810-812), stored the information in memory (814),
set the
digital conversion rate to 4000 12 bit words per second (816), and set up a
loop to
await the beginning of an actual test (818-820). This processing prior to the
start of
a valid test is necessary because the dedicated processor 14, which is now
functioning
in a completely automated mode, must itself determine when the actual
deflection of
the test specimen has occurred and begin to store and to analyze the test data
immediately thereafter.
It is possible for the routine to determine when the test specimen has been
deflected by comparing the digital data word representation of the scaled and
filtered
analog signal against certain criteria to which a torque-time response curve
of an actual
test conforms. To explain, it is known that during an actual DSR test (meaning
when
the test specimen has actually been deflected) the torsional stress developed
in a
viscoelastic material subjected to the previously discussed angular deflection
will rise
to maximum at some point during a 40 millisecond window, usually within 5-10
milliseconds of deflection, and then relax along a generally exponentially
response
curve. For a given material it is known that this developed stress, and thus
the signal
generated by the load cell, will exceed a certain known value, called the
START
LEVEL, which is chosen well below the approximate maximum stress expected to
be
developed within the test material but above most signals that might be
encountered due
to stresses developed by compressing the sample to a constant thickness or
other
anomalies. It is also known that the developed stress will not relax to a
value of less
than one-half of the START LEVEL at the end of the 40 millisecond window.
Consequently, the digitized response curve can be compared to these known
characteristics of an actual curve to determine definitely whether the test
specimen has
been deflected.
Until a data word has been obtained from the programmable interface 462 that
exceeds the START LEVEL value it is known that test specimen has not yet been
deflected and, consequently, each previous word written to the storage memory
403
will be overwritten. Once a word is examined that exceeds START LEVEL (820)
there are two possibilities: 1) that the test specimen has been deflected and
the data
_28_
204387
word is the digitized equivalent of the stress created in the sample; or, 2)
that an
anomalous signal has occurred on the input line to the dedicated processor.
Since the
routine cannot ascertain at this time whether an actual test has occurred or
whether the
signal is an anomaly, the routine assumes the data to be part of an actual
test.
Therefore, a 40 millisecond window is initiated and the start flag is set to
irAdicate that a valid test is in progress. The first word that exceeded the
START
LEVEL then forms the beginning of the two-second integration sum, and the 40
millisecond timer is incremented (822). The accessed data word is then
compared to
the maximum recorded torque value, which is zero in this case, and thus
becomes the
temporary maximum torque value (824). The timer is checked to determine if 40
milliseconds has elapsed since START LEVEL was exceed (826). Since 40
milliseconds has not elapsed, no other action is taken and the next data word
is
accessed and sequentially stored in the memory 406 along with the elapsed time
thus
beginning the accumulation of a digital representation of the relaxation
torque vs. time
curve in memory.
The routine will then check the start flag, which will now indicate that a
valid
test is in progress (818). The data word is then added to the two second
integration
sum and the 40 millisecond timer is incremented (822). The data word is
compared
to the temporary maximum torque value and if the data word exceeds that value,
the
maximum torque value is updated (824). The timer is then checked again to
determine
if the 40 millisecond window has expired (826); if not, the next data word is
digitized
at the 4000 data word per second rate and the loop repeats (818, 822-826).
Once it is determined that the 40 millisecond window has expired (826), the
last accessed data word is compared to a value of one half of START LEVEL
(828).
If this data word is less than one-half of the START LEVEL, the assumption
made
above (that since a data word exceeded START LEVEL, the test must be a valid
test)
is deemed to be in error and the test is declared to be an anomaly. The two
second
integral sum, the 40 millisecond timer, and start flag are then cleared (830),
and the
test is started anew in a loop (818-820) again searching for a data word that
exceeds
START LEVEL. Once a valid test is later begun, the new sequentially stored
data
words will be written over the words previously stored in the storage memory
406
during the invalid test.
If the last accessed data word has not decayed to a value of less than one
half
of START LEVEL, the test is declared valid and conkinues. Since the maximum
-z9-
2~~6~~'~
torque for a valid test is known to occur within the 40 millisecond window,
the
temporary maximum torque value obtained during the window (step 824) must be
the
maximum torque TM. The test continues by dividing the present two second
integration sum by 4 (832) to equally weight it with values subsequently
accumulated
during the lesser 1000 data word per second rate at which the A/D converter
408 is
now set (834).
Since the stresses in the specimen material decay exponentially, the time it
takes to decay to approximately zero will be very large and difficult to
accurately
measure due to the limited 12 bits of resolution of the A/D converter 408 and
the
comparatively small rate of change of the torque signal at times near the end
of the
test. For these reasons it is advantageous to base system measurements on well
known
time constants which are a measure of the time that it takes a signal or
function to
decay to a certain percentage of its maximum amplitude. The first time
constant Ti of
a generalized system is defined as 1/e' or 36.79 percent (rounded to the
nearest
hundredth for convenience of discussion) of the maximum amplitude and the
second
time constant TZ is defined as 1/e2 or 13.53 percent of the maximum amplitude.
Consequently, for the DSR system the first time constant T, will be the time
that it
takes the digital representation of the torque signal to decay to a value of
36.79 percent
of the maximum measured torque TM, and the second time constant T2 will be the
time
it takes the digital representation to decay to 13.53 percent of TM.
The latest accessed data word, which has been previously added to the two
second integral sum (822) and checked to determine whether a valid test is in
progress
(828), is now compared to a value equalling 36.79 percent of the maximum
measured
torque TM (mathematically, 1/e*T11~, called TC 1 (836). If the data word has
decayed
2S to a value less than TC1, then the first time constant Tl has been reached
and the time
in milliseconds since the START LEVEL was surpassed (as determined thus far by
step
822) is stored in the storage memory 406 as Tl, and a flag is set to indicate
that event.
At this time the CPU 402 also writes to a port in the program memory 404 to
light the
LED lamp 482, located on the front face of the dedicated processor 14, to
provide a
visual indication that the first time constant T1 has been reached. If the
data word
examined has not decayed to less than TC1, then no action is taken.
Similar to the above step relevant to the first time constant Tl, the routine
then
determines whether the second time constant TZ has been reached (838). The
data
word is compared to a value equalling 14.53 percent of the maximum measured
torque
-30-
~o~~~s~
TM (mathematically, I/e2*TM), called TC2. If the data word has decayed to a
value
less than TC2, then the second time constant TZ has been reached and the time
in
milliseconds since the START LEVEL was surpassed (as determined thus far by
step
822) is stored in the storage memory 406 as T2, and a flag is set to indicate
the event.
At this time the CPU 402 also writes to a port in the program memory device
404 to
light the LED lamp 484, located on the front face of the dedicated processor
14, to
provide a visual indication that the second time constant TZ has been reached.
Again,
if the data word Pxamined has not decayed to less than TC2, then no action is
taken.
On the first pass through these steps (836, 838) it is unlikely that the
digitized
representation of the torque vs. time response curve will have decayed to the
value at
first time constant Tl, since the data word compared to TC1 is the first data
word
accessed at the end of the 40 millisecond window and the test has been
determined to
be valid. (And, of course, since the value at the second time constant TZ is
less than
that at the first time constant TI, the second time constant TZ also should
not have yet
been reached.) Accordingly, it is unlikely that the first or second time
constants will
have been reached during the 40 millisecond window when the data words were
not
being compared to TC 1 or TC2.
Next, the routine checks the time constant flags and the test time to
determine
if the test has been concluded (840). If both flags are not set, indicating
that at least
the second time constant TZ has not yet been reached, or the test time is less
than two
seconds, then the test is not over and another data word is accessed for
processing and
the test time is incremented (842). If the test time is still less than 2
seconds, the data
word is added to the two second integration sum (844). The data word is then
compared to TC1 and TC2 to determine if the first or second time constants
have been
reached (836, 838) and the time constant flags and test time are again
examined to
determine if the test has been concluded (840). The routine remains in this
loop (840,
842, 844, 836, 838, 840) accessing data words, adding them to the two second
integral
sum and comparing the words to TC1 and TC2 until both time constants have been
reached and the two second integral has been calculated.
~ Once the test is completed, the data (Tl, T2, the torque v. time coordinates
and the two second integral) are stored (846) and sent to the general
processor 16 over
the RS-232 data bus (848). The data registers are then cleared, all flags and
the test
timer are reset, and the LED displays are turned off (850). The routine then
sets the
-31-
20~6~8~1
A/D converter 408 to the 4000 data word per second rate (816) and begins a
loop
searching for the start of a new test (817, 818, 820).
The general processor 16 on the other end of the RS-232 data bus receives the
test information from the dedicated processor 14 and displays the data in
numerical
farm on the display monitor 20 (714, Fig. 19). That data is then stored on the
hard
disk or floppy disk of the general processor 16 and printed on the attached
printer 17
(716). The routine executing within the general processor 16 then checks to
see
whether the reset switch has been depressed (718). If so, the test running
within the
dedicated processor 14 has been aborted, and the operator is prompted to again
enter
information for a new test (702). If the reset button has not been depressed,
the
routine will determine whether the operator has chosen to alter the test
information,
which is indicated by pressing the function key F10 (720). If the function key
has been
depressed, the display monitor is cleared and the PROGRAM I test in progress
in the
dedicated processor 14 is aborted (722), and the operator is again prompted to
enter
new test information data (702). If the function key has not been depressed,
then the
routine again waits to receive new PROGRAM I test data from the dedicated
processor
14 (714).
As can be seen from the above, if it is not necessary to re-enter test
information data and a reset has not occurred, the dedicated processor 14 and
general
processor 16 are autonomously performing repeated PROGRAM I tests,
accumulating,
analyzing and storing data for numerous tests, without operator intervention.
Consequently, once PROGRAM I has been selected from the main menu (700 Fig.
18), ,
and the operator has entered the requested test information, such as at the
beginning
of a shift, no interaction between the operator and the general processor 16
or
dedicated processor 14 is necessary during subsequent DSR tests. The operator
simply
places a desired amount of the test material in the specimen cavity, closes
the DSR
machine, and then waits for the DSR to perform the test itself, when the test
is over
the operator opens the DSR machine, replaces the used test sample with a new
test
sample and again closes the DSR machine to perform the next DSR test. This
process
may be repeated indefinitely until it is desired to perform a test on a
different type of
material or to change the testing parameters or testing information.
-32-
204638
PROGRAM II/PROGRAM III
As is discussed above, the routine executing the PROGRAM II and
PROGRAM III tests is already running prior to the actual deflection of the
test
specimen. It has requested and obtained header information from the general
processor
16 (852-854, see Figure 22), stored the information in memory (856), set the
digital
conversion rate to 1000 12 bit words per second (858), and set up a loop to
await the
beginning of an actual test (860, 862, 864). Whether the routine is executing
the
PROGRAM II test or performing the data collection aspect of the PROGRAM III
test,
the routine performs the same operations in detecting when the test specimen
has been
deflected and in determining whether a valid test is in progress.
As is described more fully with respect to PROGRAM I above, it is known
that for a valid test the torsional stress developed in a viscoelastic
material subjected
to the previously discussed angular deflection will rise to maximum at some
point
during a 40 millisecond window, exceeding a certain known value, called the
START
LEVEL, and then relax along a generally exponentially response curve. It is
also
known that the developed stress will not relax to a value of less than one-
half of the
START LEVEL at the end of the 40 millisecond window. Consequently, these
criteria
are again used to evaluate whether a signal corresponds to an actual test, or
is the
result of an anomaly.
The routine thus accesses a data word from the programmable interface 462
(860) which is collecting data words from the A/D converter 408 at a rate of
1000
words per second. Assuming the routine has not already determined that a test
is in
progress, data words are accessed and examined in a loop until it is
determined that
START LEVEL has been exceeded (860, 862, 864). When START LEVEL has been
exceeded, the start flag is set and a 40 millisecond window is opened (866).
It should
be noted that the value for START LEVEL is offset by 96 digital counts, thus
corresponding to 9.6 inch-lbsf, from that entered by the operator prior to the
start of
the test. This relates to the 9.6 inch-lbs~ of analog offset added to the
analog input
signal received by the A/D converter 408 by setting the program switch 502 to
the
PROGRAM II/PROGRAM III position.
As further words are accessed (860), the routine will check the start flag
(862)
and note that a test is in progress. Consequently, the routine will increment
the test
timer (866) and examine it to determine whether 40 milliseconds have elapsed
since
START LEVEL was surpassed (868). The routine will continue in this loop (860,
862,
-33-
~~4~~8'~
866, 868) until the test timer has been incremented sufficiently to exceed 40
milliseconds. Once the 40 millisecond window has closed, the routine checks
the last
data word accessed to determine whether it has dropped below one-half of START
LEVEL (8?0). If it has, the test is declared an anomaly, the test timer and
start flag
are reset (872), and the routine again sets up a loop wherein it repeatedly
accesses data
words and compares them to START LEVEL to detect the start of a new test (860,
862, 864). If, however, the data word accessed at the end of the 40
millisecond
window was above one-half of START LEVEL (870), the test is declared to be
valid
and the CPU 402 will turn on LED lights 482 and 484 (corresponding to the time
constant LEDs of PROGRAM I) by writing to the appropriate ports in the program
memory 404 to indicate that a valid test is in progress (874). At this point
the routine
also determines whether the operator had selected either the PROGRAM II or
PROGRAM III test to- be performed, as are described separately in the
subsections
below.
A) gROGRAM II
If PROGRAM II has been selected, the routine will next collect 21 data points
at evenly logarithmically spaced intervals between .04 seconds to 400 seconds
and
temporarily store the data along with the corresponding times (876). When all
of the
data has been collected, after 400 seconds, the CPU 402 writes to a port in
the
program memory device 404 which lights the ready LED 488 to provide a visual
indication to the operator that all data has been accumulated. The sample is
then
removed. The operator then presses the zero switch 500 on the face of the
dedicated
processor 14 which instructs the CPU 402 to read the output of the load cell
at zero
torque. This zero value indicates the drift of the load cell 206 over the
comparatively
long PROGRAM II test. The routine then corrects the collected data points
based on
the zero value to improve the accuracy of the results, performs a logarithmic
based
conversion of the data points and restores the corrected log based data points
(878).
The results are paired log-torque vs. log-time data points which are sent to
the
plotter 18, such as to produce the representative graph of Figure 5, and to
the general
processor 16 for numerical and graphical display on the attached monitor 20
(880).
1fie routine then clears the data registers, resets the start flag and the
test timer, turns
off the LEDs (882), and begins looping to find the start of a new test (860,
862, 864).
Once the general processor receives the PROGRAM II data from the dedicated
processor 14 (714, Fig. 19), the numerical data is printed on the attached
printer 17,
-34-
2~4~3'~'~
and saved on a hard or floppy disk (724). The general processor 16 then plots
a log
torque versus log time stress relaxation curve on the display monitor 20 (726)
and
displays a number of continuing selections (728) for the operator to chose
(728) such
functions include printing or plotting a copy of the stress relaxation curve
(730, 732,
respectively), or changing the degree of fit of the curve drawn through the
torque
versus time coordinates (734). In each of these circumstances once the
function is
performed, the routine returns to the menu to allow the operator to select
another
function. Once all desired functions have been performed, the routine
continues to the
second menu which allows the operator to select from a separate set of
additional
functions, such as changing the preset curve fit, returning to the PROGRAM II
test,
saving the log torque versus log time stress relaxation curve on disk,
entering new test
information, or exiting the program to the main operating system, such as DOS
(736).
As with PROGRAM I, PROGRAM II requires only a small amount of
operator interaction. In fact, if the operator desires to simply continue
running
repeated PROGRAM II tests, he need only provide responses to two prompts (728,
736), selecting the "ontinue option in the first instance, and in option 2 to
re-run
PROGRAM II with new data in the second instance.
PROGRAM II test data which has been stored on disk can be recalled from
the main menu 700 by selecting the appropriate choice. In this case the
routine
prompts the user to enter the file containing the data to be retrieved (see
Fig. 19, step
738). The routine then opens the file, retrieves the data (740) and displays
the log
torque versus log time stress relaxation curve on the display monitor 20
(742). The
routine then allows the operator to select the functions provided through
menus 1 and
2 described above (728, 736) including printing or plotting the curve (730,
732),
changes the degree of fit of the curve (734) or performing PROGRAM II test run
with
new data, etc.
B) PROGRAM III
If the PROGRAM III test has been selected, the routine executing within the
dedicated processor 14 will have been previously provided with the number of
data
points that were selected for collection (884, Fig. 22). The routine then
collects that
many evenly logarithmically spaced data points and temporarily stores the data
points
in memory with a corresponding time (886). Once all of the data points have
been
collected, the CPU 402 writes to a port in the program memory device 404 which
lights the ready LED 488 to provide a visual indication to the operator that
all data has
-35-
2p46~87
been accumulated (888). The operator then removes the sample from the DSR and
presses the zero switch S00 on the face of the dedicated processor 14 which
instructs
the CPU 402 to read the output of the load cell at zero torque. This zero
value
indicates the drift of the load cell 206 over the comparatively long PROGRAM
III test.
S The routine then determines if data correction is necessary (890), and, if
so, the routine
corrects the collected data points to improve the accuracy of the results
(892).
The torque vs. log time data points are now transferred to the general
processor 16 over the RS-232 data bus for further processing by the general
processor
in accordance with PROGRAM III. The routine executing in the dedicated
processor
then clears the data registers, resets the start flag and the test timer,
turns off the LEDs
(896), and begins looping to fmd the start of a new test (860, 862, 864).
The remainder of PROGRAM III data analysis is executed in the general
computer 16 with the log-torque vs. log-time data pairs computed by and
received from
the dedicated processor 14 over the RS-232 bus (Figure 19, 714). The PROGRAM
III
1S routine of the general processor 16 stores the torque data points and the
log time points
in separate one dimensional arrays (see Figure 24, step 900), called
TORQUE(array)
and X(array), respectively. Hereinafter, both in the discussion and in Figures
24-2S
and Figures 29-34, software variable names are denoted by all capitals.
Further, those
variables that are also arrays are denoted by the use of "(array)" appended to
the
variable name with the number of uses of "(array)" indicating the
dimensionality of the
array. For convenience, only the introductory use of an array variable in the
discussion and in the figures will include the "(array)" appendix.
The routine then performs a logarithmic conversion of the torque relaxation
values in TORQUE and places the results in another one dimensional array,
Y(array)
2S (902). The log torque values in Y can then be converted to obtain the
logarithm of the
relaxation modulus which is restored in Y (904). The log relaxation modulus
values
in Y are then converted to relaxation modulus values which are stored in
G(array) and
their corresponding times are stored in TIM(array) (906). The relaxation
modulus,
commonly abbreviated as G, is used to in the well known Yagii/Maekawa
approximation to yield approximations of the fundamental viscoelastic
properties of the
test material.
The conversion from log torque relaxation values to log relaxation modulus
values is possible from a knowledge of the geometry of the test specimen at
deflection
and the amount of deflection. The geometry of the deformed test specimen is a
-36-
~0~0~~7
function of the dimensions of the rotor 158 and the specimen cavity 152, as
well as the
closure height, all of which have been measured for a particular DSR device 12
and
stored in a data file in the general processor 16. A review of the
mathematical basis
for determining the relaxation modulus is discussed below.
The torque, represented in equation 1 as T, measured during the DSR test is
dependent upon the stress generated within the test specimen (confined between
the
conical surface 156 of the rotor 158 and the surfaces 160 and 162 of the
specimen
container 94) and transmitter to the conical rotor surface, is described by
Equation 1:
lO
a ~ R ~ f ~ dA
where r = shear stress (1)
R = moment arm
A = surface area o! rotor
In Equation 1 the shear stress 'r can be replaced by the product of the
specimen
deformation or strain Y and the shear relaxation modulus G as shown in
Equation 2:
w
' J R'G°~'~ (2)
If it is assumed that the relaxation modulus G is independent of strain, i.e.,
that the behavior of the test specimen can be described by linear viscoelastic
theory,
then G can be considered as a constant with regard to the integration,
although it varies
as a function of time. Removing G from the integrand yields Equation 3:
A
- G R . Y . dA (3)
J
-37-
2fl~638'~
The integral, which relies on an accurate description of the test specimen
strain, represents the reciprocal of the geometric form factor between the
rotor 158 and
sp~imen container cavity 152. Since the actual geometry of the rotor 158 and
specimen container cavity 152 are known, the form factor can be calculated and
used
to solve equation 3. Since the relaxation torque T is measured during the
test, it is also
a known value at a given time. Consequently, the shear relaxation modulus ~ is
the
only remaining unknown in Equation 3 and it can thus be solved by multiplying
the
form factor by the measured relaxation torque T.
It is noteworthy that Equation 3 does not consider the effects of strain rate
history during the deformation. Typically, classical stress relaxation
experiments try
to impose a nearly instantaneous deformation on the test specimen.
Realistically,
however, stress relaxation data generated during the initial time period
following the
formation are discarded since it is known that data obtained at times greater
than 5 or
10 times the time period required to impose the deformation are essentially
equivalent
to those obtained from an ideally instantaneous deformation. The data captured
by the
PROGRAM III routine executed within the dedicated processor 14 meet this
criterion.
Thus, Equation 3 above provides an adequate starting point for the development
of a
model.
A close-up view of the actual geometry of the rotor I58 and the specimen
container cavity 152 at closure is shown in Figure 26. To reduce the
complexities
involved in mathematically modeling this geometry to arrive at its form
factor, a few
minor approximations are incorporated to arnve at the test geometry as shown
in
Figure 27. Also shown in Figure 27 are the z-r coordinate system and the
variables
used to define the test geometry. Both the radius of rotor, R, and the radius
of the
specimen container recess, Ro, vary along the z axis. The vertical distance H
between
the rotor 158 and the surface 160 of the specimen container cavity 152 and the
angular
deflection, OC , by which the strain is imposed are operator selectable and
thus are
treated as variables.
In order to facilitate modeling the strain throughout the test specimen, a
transformation to a "local" coordinate system is beneficial. For this purpose
the Y-axis
is positioned normal or perpendicular to the conical surface 156 of the rotor
158. The
local X-axis lies tangential to the rotor surface 156 and points in the
direction of the
-38-
~04~~~°~
angular deflection ~. The local strain of an element of "fluid" is thus
described by
Equation 4:
ex
a -
eY
where eX = tangential movement of a point (4)
on the rotor surface resulting
from the a angular deflection.
eY = S, the separation between the
rotor and stator along the Y-axis.
The above transformation to a local coordinate system is the basis of the
known "lubrication approximation" which permits the description of local
strain (as is
shown) of an element of fluid confined in a narrow gap as long as the
dimensions of
the gap are much smaller than the radius of curvature of the gap. The use of
the
approximation for the DSR becomes questionable at small values of z (values
near the
points of the conical surfaces); however, the contribution of the stress in
this region
compared to the total torque is minimal due to the small moment arm, R in the
region.
As z increases the radius of curvature and consequently the moment arm R
increases,
thus the accuracy of the overall approximation becomes less questionable.
Figure 28 shows schematically the manner in which a Y or S, the separation
between the conical surface of the rotor 100 and the surface of the specimen
cavity 94,
varies as a function of z. It becomes apparent from the figure that the DSR
test
geometry requires the consideration of three separate regions. Specifically,
the total .
torque consists of the sum of the torque components for each of the three
regions;
yielding Equation 5.
T = T1 + TZ + T3 (5)
A derivation of the strain and torque components for each region yield the
following
results:
~~a
Region 1 strain, ~l
H sin ~9
-39-
20~~387
Ti = G . B . (A - H)t
4
Row" + H (sin ~)2
where A =
tan
2r a (taa ~)3
B =
H (sin ,Bxcoa ~)
aztan~9
Region 2 strain, 7s _
sin~(A-z)
z(A-z)2
T= = GHB + 2Az (A-z) - Azr In (A-z)
2
_ 3
- (A z) + A(A-z)Z + A~(A-z) (1 - tn[A-z~ Za (9)
6 Z~
~6ero zt, and z= ue tho limits of integration.
a R"" (10)
Region 3 strain, ys = R~ - ~
where R",x = maximum rotor radius
Ro""~ = maximum stator radius
G2:R3""~aL
Ts '~ ~ (tt)
Ro"s,i - R"s~t
where L = vertical length of Region 3
Substitution of Equations 7, 9, and 11 into Equation 5 provides the
mathematical model of the DSR test. Examination of the torque components
indicates
s: ~ that G, the stress relaxation modulus, appears as a single multiplication
factor (the
w:: unknown variable) in each; the remaining terms are all known quantities
obtained from
'~' either manufactured dimensions of the rotor 158 and specimen cavity 152 or
operator
:' -40-
~~4638'~
selected options (i.e., deflection angleoc and closure height H). Thus, the
sum of the
terms, with G removed as a multiplier, is the reciprocal of the geometrical
form factor
far the DSR test. The stress relaxation modulus G as a function of time is
then
obtained by multiplying the measured torque relaxation T as a function of time
by the
form factor.
Consequently, the log torques values in the array Y are multiplied by a form
factor, calculated in accordance with the above mathematical analysis, and
again stored
in Y to yield the log stress relaxation modulus G (904). These values are then
wnverted back to base 10 to obtain the stress relaxation modulus which is
stored in the
array G (906).
A least squares polynomial curve fit of a preselected degree is then
performed,
such as by using Crout's reduction technique, on the stress relaxation modulus
and time
coordinates, as found in the G and X arrays respectively (908). The polynomial
coefficients yielded by the curve fit routine are stored in C(array) {910) for
later use
in determining the fundamental viscoelastic properties of the test specimen.
The routine then determines the minima and maxima of the log stress
relaxation modulus and log time to obtain the appropriate scaling factors
(912), and
plots a log stress relaxation moduius vs. log time curve on the display
monitor 20
(914). The operator or technician is then provided with a number of options
from ..
which to select (916), such as printing or plotting a copy of the curve,
changing the
degree of curve fit of the curve so as to more reasonably match the flow of
the curve
with the actual data points, manually entering the scaling coordinates of the
plot, or
continuing. (Figure 6 illustrates typical log stress relaxation modulus vs.
log time
curves for three different materials.) If a selection other than continuing is
made the
routine will jump to the appropriate place in the routine to perform the
function and
then will eventually return to this menu to allow selection of another or even
the same
function. Note, as mentioned above, that as used in the figures an arrow
pointing
toward a letter or number indicates that the routine will jump to the step on
the
flowchart indicated by another arrow extending from the same letter or number
encircled at another location on the flowcharts.
Once any resealing, printing, etc., functions are performed as selected from
the menu (916), the stress relaxation modulus vs. time data is tabulated and
printed on
the attached printer (918).
_41-
2~~~3~'~
The polynomial coefficients as determined above (908) with the preselected
or modified degree of fit are then used in an implementation of the known
Yagii-Maekawa approximation to generate fundamental frequency dependent
viscoelastic propexty information, including the Loss modulus GPP(array), the
storage
modulus GP(array), the complex viscosity VIS(array), and the loss tangent
TAND(array); each as a function of angular frequency W(array) (920).
A least square curve fit is then performed for the log base value of each of
these viscoelastic properties, GPP, GP, VIS, TAND, as a function of the log
based
frequency W (922). The maxima and minima of the viscoelastic property data are
determined for the purposes of scaling the graphical results (924), and the
property data
is printed in a tabulation form on the attached printer 17 and displayed on a
display
monitor 20 (926). The operator is also given the option at this time of saving
any of
the viscoelastic properties as well as relaxation modulus versus time data and
the curve
fit coefficients in a series of data files (928).
A graphic plot of the loss modulus GPP, the storage modulus GP, and the
complex viscosity VIS is then displayed as a function of angular frequency on
the same
graph on the display monitor 20 (930). The operator is then given a series of
options
from which further functions may be selected, such as printing or plotting the
screen
plot (see Figure 8), changing the degree of fit of the curves, manually
scaling the data,
or simply continuing. In each case, except for when continuing is selected,
the routine
will jump to the appropriate instructions to perform the desired function.
Those
functions will be performed and eventually the options will be once again
displayed
(932). A Cole-Cole plot, which is a plot of the loss modules as a function of
the
storage modules, is then performed and displayed (934) (see Figure 9.) The
operator
is then provided with a second set of functions from which to select,
including a
change in degree of fit again, performing the PROGRAM III test again with new
data,
entering new test information, or exiting the program all together (936).
As with PROGRAM I or PROGRAM II, a series of PROGRAM III tests may
be run sequentially with very little operator interaction. For example, once
the
operator has closed the DSR device to the required height, the test is
performed and
results are generated with no further action by the operator. Once the test
has been
completed, the operator need only answer a few simple prompts, which among
other
things store the data for later analysis, to re-run the PROGRAM III test with
new data.
-42-
~o~~~~~
In this manner a number of tests can be performed and the results saved for
future
analysis by a skilled technician or engineer.
Once one or a series of PROGRAM III tests have been completed, the results
may be reviewed and compared through software manipulation of stored data.
Initially, this section of PROGRAM III code performs basic initiation
functions such
as determining the communication parameters for interface with the printer 17
or
plotter 18, and characteristics of the display monitor 20 (see Figure 29, Step
1000).
The routine then prompts the user to select the comparison options desired,
such as
comparing the relaxation modulus or storage modulus curves for a number of
previously performed tests (1002). The user is then prompted to select the
data files,
corresponding to previously performed tests, desired for comparison (1004).
These
files contain identifiers such as the file name as well as the relaxation
modulus and time
coordinates, the number of data points collected during the test, the degree
of
polynomial fit and the related coefficients for the test, and the minima and
maxima of ;
the coordinates. Depending on the options selected from Menu 3 (1002) the
routine
will then jump to a corresponding subroutine to perform the required
functions. Note
that in Figures 29 through 34 the comparison option selections and
corresponding
subroutines are identified by double letters, such as AA, BB, etc.
If the user choose to compare the relaxation modulus curves far a number of
previously performed tests, the routine will jump to the appropriate
subroutine (AA)
and, initially, determine the maximum and minimum values amongst all the
selected
files to perform their appropriate scaling (1006). The routine then plots the
axis and
appropriate labels on the display monitor (1008) plots the relaxation modulus
vs. time
data coordinates for the first selected file on the graph, and then plots the
relaxation
modulus curve, determined by the polynomial fit coefficients on the graph
(1010).
Note that while the data points identified on the graph with appropriate
symbols, such
as squares, circles, etc., are actual collected data points from the test, the
curves are
a function of a least squares fit routine, and thus may not pass through the
actual data
points.
Once the data points and associated curve for a selected file has been
completed, the plotting color and symbol are changed (1010) and the routine
determines whether all the selected files and related coordinates have been
plotted
(1012). If not, the step of plotting the coordinates and associated curve is
then
repeated (1010) until ail the files have been plotted on the graph. The user
is then
-43-
~~4638'~
provided with the option of printing or plotting the graph shown on the
display
monitor, manually rescaling the graph or performing no function at all (1014).
If the
user chooses to manually rescale the graph, program control will return to
Step (1010)
to again plot the selected files. Once all the desired graphics options have
been
selected and completed, a second comparison options menu (Menu 3b) is
displayed to
allow the user to compare other properties, select new data files for
comparison, or
exit the program all together (1016).
If the user selected comparisons of the storage modulus, loss modulus, or
complex viscosity data from the previously performed tests in either Menu 1 or
Menu
2, the routine jumps to the appropriate subroutine to handle these comparisons
(BB, see
Figure 31). This routine provides the user the option of plotting one property
for all
of the data files on the same graph. Based on the selection the routine will
then
determine the appropriate minimum and maximum values based on the selected
files
(1018), and then, based on the scaling factor, calculate and plot the scaled
points
(1020). The routine then checks to determine whether all the files have been
plotted
on the graph (1022) and repeats the process until the graph is completed. The
user is
then again provided with graphics options from which to select, such as
printing or
plotting the displayed graph or manually rescaling the graph (1024). Upon
completion
of the graphics options, the second comparison options menu is then
redisplayed
allowing further comparisons by the user (1016).
If the user has selected a COLE-COLE plot, which is a plot of the shear loss
modulus as a function of the shear storage modulus, the routine will jump to
the
subroutine associated therewith (CC, see Figure 32). The routine first
determines the
maximum and minimum values for both the loss modulus and storage modulus data
points for scaling of the curve (1026). The axes are then plotted and
identified with
the appropriate labels (1028). Since the actual values of the loss modulus GPP
and the
storage modulus GP yielded from the YAGII/MAEKAWA conversion are a function
of angular frequency, the least squares curve fit polynomial coefficients of
each is used
to calculate GPP as a function of GP (1030). A sufficient number of these
values are
determined and plotted to yield a relatively smooth curve. The points
identified by
symbols along the curve, consequently, are not actually test data points, but
rather are
convenient points along the calculated curve. This process is repeated for all
the
selected files until the graph is complete (1030, 1032). The user is then once
again
provided with graphics options from which printing, plotting or rescaling may
be
-44-
selected (1034). Once the graphics options have been selected and completed,
program
control is returned to the second comparison options menu to allow the
selection of
further comparisons, etc. (1016).
Another option which the user has from either of the main menus is to select
the tabulated comparison criteria. This option allows the user to select a set
of criteria
for comparing the various samples at either constant property values or
time/frequency
values. The responsible routine (DD, Figure 33) first determines whether a
criteria file
has already been developed (1036). If so, the user is prompted to enter the
name of
the criteria file (1038) and the routine will retrieve the file (1040) and
then proceed to
tabulate the values as discussed below. If a criteria frle has not been
constructed, a
suitable file will be opened (1042). The user is then prompted to enter the
number of
criteria to be tabulated (1044), and to enter the type of parameter types from
which the
tabulation will be made (1046). Examples of the parameters are the relaxation
modulus
vs. time, storage or loss modulus vs. frequency, the complex viscosity vs.
frequency
or the loss modulus vs. the storage modulus. The user is then prompted to
enter the
basis of criteria, in other words, whether the property value to be held
constant is an
independent or dependent variable in the parameter type (1048). The user is
then
prompted to enter the numeric value of the criteria to be held constant
(1050). If the
total number of criteria selected have not all been entered (1052) the user is
again
prompted to enter the next type of parameter (1046), the basis of criteria
(1048), and
the numeric value of the constant criterion (1050) and all the criteria are
stored (1060).
The basis for tabulating values based on the constant criterion depends on the
parameter type and whether the criterion is a dependent or independent
variable. If the
criterion is a constant independent variable of the relaxation modulus, the
storage
madulus, the loss modulus, or the complex viscosity, the numeric value of the
criterion
is simply entered into the function of the parameter type as determined by the
coefficients of the least squares polynomial, and the result is the tabulated
value (1062).
If the constant criterion is, however, a dependent variable of the relaxation
modulus,
storage modulus, loss modulus, or complex viscosity parameters, then the
tabulated
value for each PROGRAM III test is determined by Newton's well-known iterative
method (1064). Consequently, the least squares polynomial of the selected
parameter
type, i.e., G, GP, GPP or VIS, is solved for the independent variable
iteratively. The
iterative solution is continued until the desired independent variable is
calculated within
-45-
2~4638'~
a preset tolerance or the number of iterations exceeds a preset maximum, where
upon
a warning is issued.
If the constant criterion selected is the independent variable of the loss
modules vs. storage modules parameter type, i.e., GP, then a two-step process
is
performed (1066). First, the least squared polynomial of the storage modules
vs.
frequency is iteratively solved for the frequency at the constant criterion GP
using
Newton's iterative method. Secondly, the solved for frequency value is then
entered
into the least squared polynomial for the loss modules vs. frequency, yielding
the
desired loss modules tabulation.
If the selected constant criterion is the dependent variable of the loss
modules
vs. storage modules parameter type, i.e., GPP, then a two-step process is
again
followed (1068). In this case the least squared polynomial for the loss
modules is
iteratively solved for the frequency corresponding to the selected constant
GPP
criterion using Newton's iterative method. Secondly, the least squared
polynomial of
the storage modules vs. frequency is then solved using the frequency
determined
above, thus yielding the tabulated storage modules.
One of the above processes (1062, 1064, 1066, 1068) is repeated for all the
data files and all criterion until the tabulation is complete (1070). The
tabulated data
is then placed in the appropriate format and displaced on the display monitor
(1072).
The user is then again provided with the second comparison options menu to
perform
further comparisons, begin a new data analysis, or exit the program all
together (1016).
While specific embodiments of the mechanical and electrical hardware, and
the computer software structure are recited herein it will be appreciated that
other
embodiments and implementations that accomplish the spirit of the invention
are
ZS possible and that the invention is not in any way limited to the specific
embodiments
discussed. For example, in some instances it may be possible to implement a
system
performing many of the same functions without separate dedicated and general
processors, or to employ similar processing elements such as optical computers
or
parallel processors. In such instances it would be apparent to modify the
software to
operate most effectively and efficiently within the constraints and
capabilities of the
specific processing element or elements implemented. Further, the described
embodiments, mechanics, and geometries of the DSR machine could be modified in
other like ways which would impart an impulsive rotational deflection on the
test
specimen and detect the reactive torsional stresses developed in the test
specimen.
-46-
