Note: Descriptions are shown in the official language in which they were submitted.
3~5
,
MEASUR1~6 AND CONTRQLLING METHODS AND APPARATUS
This invention relates to methods and apparatus for
measuring and controlling the degree of cur2 of carbonaceous
polymeric materials that are formed from a plurality of chemical
reactants and passed through a curing process, from which they
issue as a traveling product or intermediate in a continuous or
semi-continuous form. More particularly the invention relates
to methods and apparatus for performing the measurements by
directing into the traveling material infrared radiations at
multTple wavelengths, detecting radiations that have been
transmitted or reflected by the material, and computing the
degree of cure in a substantially instantaneousy continuous and
non-destructive manner from instrument responses to the detected
radiations.
The availability of instrumentation which functions or is
constructed in accordance with the invention permits immediate
manual or automatic feedback control of one or more of the
process parameters that affect the degree of cure, thus enabling
the manufacture of a product or intermediate in an optimized
state of polymerization and with the probability of substantial
raw material savings and/or energy savings.
hile the methods and apparatus of the invention are
applicable to the manufacture of many different polymeric
material products, the invention is herein described and
illustrated in an embodiment for measuring and~or controlling
the degrae of cure of the binder resin applied to a mat of glass
fibers, to be used, for example, in building insulation
products.
~ackaround Art
Synthettc resins and natural polymeric compounds are used
in the production of a great many materials and discrete items
whose manufacture includes a curing stage. The manufactured
~Z039~
--2--
material or item may consist of the polymerized matarial itself,
or the polymerized material may be used as a binder to hold
together various layers of aggregates of other materials which
in combination with the binder can impart to the finished
product its desired properties.
The desired properties are often significantly affected by
the degree of polymerization of the resin compound. The
affected properties typically include the flexibility and
toughness of plastics, or the stiffness, abrasion and impact
resistance of laminated sheets.
In various manufacturing processes, the degree of
polymerization has been controlled by resulating one or more
process parameters such as concentration, residence time,
termperature and catalysis. Various instruments have been used
to determine heat liberation, viscosity, density and electrical
conductivity. Spectroscopic instruments of the gas
chromatographic and nuclear magnetic resonance type have been
very useful. Laboratory determinations of molecular weight are
commonly performed as a quality control measure.
An arttcle by Crandall, E.W. and Jagtap, A.N., entitled
"The Near-lnfrared Spectra of Polymers," Journal of _A~pl~ed
Polymer Science, Vol. 21, pp. 449-454 (1977) and its
bibliography have indicated that near-infrared spectroscopy can
be useful in the identification of resins and polymers and in
following the course of polymerization and giving some
indication of the state of cure. Polymers were melted and
pressed between glass plates to give a transparent film, or they
were dissolved and their spectra in solution were run against
the spectrum of the pure solvent. It was observed that 7n
various polymers the process of curing (with heat) affected the
relative intensities of the overtone bands of carbonyl (C=0),
amino and amide N--~ and alcoholic ~--H. The overtone bands of
alkyl and aryl C - H were also observed along with certain C---H
combination bands. The intensities of various bands were set
forth by way oF comparison with that of the C - H stretch band
~2C1 39~S
whose fundamental vibration lies at 3~3-3.5 em in the middle-
infrared and whose first overtone appears 3t 1.7~m in the near-
infrared.
Two of the carbonaceous polymeric materials discussed by
Crandall et al are phenol-formaldehyde and urea-formaldehyde
resins, which are commonly utilized, inter alia, as binders in
glass fiber mats, say, for building insulation. In the
manufacure of these mats, typically glass fiber is spun from
molten glass and is sprayed with uncured binder compound as the
fiber is being showered onto a moving chain conveyor. The
conveyor carries the resulting glass fiber blanket through
curing ovens wherein the polymeric binder material is exposed to
elevated temperatures for an appropriate time period to complete
the curing of ehe binder. After its exit from the ovens, the
mat or blanket is cooled by a stream of air from a fan, and
certain of its properties may be measured, typically with
radiatTon gauyes.
The mass per unit area of the traveling mat has been
measured, with various degrees of success, using beta ray
gauges, gamma ray or x-ray gauges, or infrared radiation gauges
using combinations of wavelengths. The resulting measurements
have been used to automatically control the speed oF the chain
conveyor, thereby determining the amount of coated glass fibers
deposited while a section of the conveyor moves through the
felting chamber, with the objective of maintaining the weight
per unit area of the mat constant along its length.
Various attempts have been made to measure the mass of the
binder material per se, for example by taking advantage of the
fact that glass is substantially transparent to certain optical
(e.g., infrared) and x-ray wavelengths that are significantly
attenuated by the binder materials. The objective of this
measurement is to be able to control the mass of the binder by
regulatTng the amount or the dilution of the spray material
applied.
3 S
--4--
The degree of cure has been measured in laboratories, for
example, by free phenol determination or molecular weight
determination. On the basis of their experience with
laboratory-analyzed samples, line operators typically make a
visual estimate of the degree of cure by inspection of the
"color" of the mat, and adjust the oven temperature accordingly.
However, as an indicator of cure, color has been shown
objectively to be misleading in many cases, as well as
subjective. Moreover, color changes do not exhibit high
sensitivity except when the mat is already over-cured (burnt).
The degree of cure of the binder is believed to have
ccnsiderable economic significance, since it affects the
property of the glass fiber mat which is termed "recovery." The
manufactured mat is usually compressed into rol 15 or bales for
shipment and storage, and "recovery" is the extent Jo which the
mat is able to spring back to its original thickness when the
compression is relieved. Recovery is also related to the
ability of the mat to maintain its shape and thickness for long
periods of time in use as insulation and for other purposes.
Hence, it the binder has an optimum degree of cure, a mat with a
desired thickness and insulation value in service can be
manufactured from a smaller amount of glass fiber and binder.
The avoidance of overcuring can also result in lower energy
costs during manufacture.
The properties of recovery and resistance to sag and
deterioration in service seem to be dependent on an adequate
degree of polymerization. There is evidence, on the other hand9
that overcuring results in depolymerization as well as other
deleterious effects.
It follows that there has been a need for a method and
apparatus which provides an instantaneous, substantTally
continuous and non-destructive indication of the degree of cure
of certain traveling polymeric materials.
~2~3~5
Disclosure of Invention
-
In accordance with this invention, there are provided
methods and apparatus for determining the degree of cure of a
traveling carbonaceous polymeric material -tha-t has been formed
from a plurality of chemical reactants and subjected to a curing
process, comprising combinations of method steps and apparatus
elements for directing into the traveling material a first
infrared radiation from the group thereof adapted to selectively
interact with molecular resonance vibrations at frequencies that
are characteristic of respective terminal functional groups of
atoms involved in reactions that take place in the material during
the curing process, so that the material exhibits an absDrp-
tivity for the first infrared radiation that varies with the
degree of cure of the polymeric material, also directing into
the traveling material a second infrared radiation that is either
of the kind that does not exhibit substantial selection inter-
action with molecular resonance vibrations in the material or of
the kind that is adapted to selectively interact with molecular
resonance vibrations at a frequency that is characteristic of
groups of atoms forming the backbones of the polymeric molecules
in the material, receiving from the traveling material radiations
that have interacted with the material; producing from the
received radiations first and second responses to the first
and second radiations; producing a third response that is
: indicative of the mass of the polymeric material interacting
with the radiations, and substantially independent of the
variations in the absorptivity of the material for the first
infrared radiation which occur as the curring process progresses,
and producing from the first, second ana third responses an
output response that is a function of the changes in -the
absorptivity of the material for the first infrared radiation,
substantially independent of the amount of the polymeric material
interacti~g with the xadiations, and correlated with the degree of
cure effected by the curing process.
Typically the first infrared radiation is selected from
the group adapted to selectively interact with molecular
kh/J~
3g~5
esonance vibrations at respective 0- - H, Al and C=0
vibration frequencles.
method steps and apparatus elements may be provided
for directing into the traveling material a third infrared
radiation of the other kind, producing from the received
radiations a fur-ther response -to the third infrared radiation,
and producing the third response from the further response and
the second response.
The third radiation may be adapted to selec-tively
interact with molecular resonance vibrations at a C H stretch
vibration frequency.
The first infrared radiation may comprise the near-
infrared overtone band having wavelengths in the vicinity of
1.50~, whereas either of the second and third infrared radia-tions
may have wavelengths in the vicinity of either 1.35~ or 1.75~.
A first mathematical func-tion of the ratio of the first
and third responses may be formed, together with a second
mathematical function of the ratio of the second and third
responses, and the first and second functions may be combined to
produce the output response.
The first and second functions may be substantially
linear functions. The ratio of the first and second functions
may be formed in order to produce the output response.
The first function may be indicative of the number of
terminal functional groups present in relation to the number of
groups forming the backbones of the molecules, and the second
function may be indicative of the number of groups forming the
backbones of the molecules that have interacted with the
radiations.
The first and second responses may be used to produce
a fourth response that is indicative of the mass of the polymeric
material interacting with the radiations but which is dependent
on the variations in the absorptivity o the material for the
irst infrared radiation that occur as the curing process
progresses, and the fourth response and the third response may
be utilized to produce the output response.
-- 6
kh/~
~,39~
The fourth response may be compensated for the
radiation path length extension effects of scattering. The
third response may be similarly compensated.
Method s-teps and apparatus elements may be proviaed
for directing into the material further radiations having a
mode of interaction -that is differen-t from that of the infrared
xadiations; these further radiations that have interacted with
the material may be detected to produce an additional response,
and the additional response may be used to effect the scat-tering
compensation. The further radiations may be x rays or gamma
rays.
The carbonaceous polymeric material may be ~lsed to
form a binder coating for the fibers in a mat of glass fibers;
tha curing process may include exposing the mat to elevated
temperatures, and the exposed mat may be passed through a
measuring zone in which the radiations are directed into and
received from the mat.
The exposure of the mat to the elevated temperatures
may be controlled in accordance with the output response.
The first infrared radiation may comprise a near-infrared
overtone band adapted to selectively interact with molecular
resonsance vibrations at one or both of the 0 H and N H
vibration frequencies.
6a -
kh /J I-
~Z(~,3g~5
--7--
Method steps and apparatus elements may be provided for
measuring the weight per unit area of the mat passing through
the measuring zone, and controlling the rate of travel of the
mat in accordance with the weight per unit area measurement;
controlling the rate of application of the binder coating in
accordance with the second mathematical function, and
controlling the temperature of ehe mat during at least a portion
of the curing process in accordance with the output response.
The objects of this invention are to provide methods and
apparatus for accurately and reliably measuring the degree of
cure of traveling carbonaceous polymeric materials in a
substantially instantaneous, continuous and non-destructive
manner; to provide such methods and apparatus which make
psssible automatic feedback control of curiny processes so as to
maintain constant a desired degree of polymerization of the
materia1; to provide such methods and apparaeus that are useable
when the polymeric material has been applied as a binder or
coating for other materials, and to provide an improved
- measuring and controlling system for a glass fiber manufacturing
process including binder cure control.
Other objects and advantages of the invention will become
apparent in the following detailed description of some best
modes for carrying out the invention, taken in conjunction with
the appended drawings.
Brief Description of the Drawing
- Figure 1 is a schematic showing oF a process for producing
a mat of glass fibers that are coated with a controlled amount
of a carbonaceous polymeric binder material and cured to an
optimum deyree of polymerization of the binder by a method and
apparatus according to the invention.
Figure 2 is a schematic showing, including a quasi-section
on the line 2 -2 of Figure 1, of an apparatus for automatically
determining the degree of cure and binder weight of the
polymeric coating materlal.
:.
39~5
--8--
Fiyure 3 is an enlarged and more detailed schematic showing
of a portion of Figure 2.
Figure 4 is a graph depicting near-infrared absorption
spectra of a particular carbonaceous polymeric maeerial in four
different stages of cure.
Figure 5 is a graph depicting near-infrared reflection
spectra of a particular carbonaceous polymeric material ;n four
different stages of cure.
Referring to Figure 1, molten glass from a melter 10 flows
into a forehearth 12 that supplies a stream 14 of melted glass
to a centrifugal spinner 16. Filaments 18 oF 91ass ejected from
the spinner 16 are directed downwardly and partially solidified
by air streams from jets as at 20. The filaments 18 descend
through a felting chamber 22 and are collected on a traveling
conveyor chain 24. Air suction through the conveyor 24,
depicted by arrows 26, aids the formation of a fiber blanket 28
on the conveyor.
The filaments 18 are sprayed, during their descent through
the felting chamber 22, with a binder spray ejected through a
piurality of spray nozzles as at 30. The binder spray comprises
a carbonaceous polymeric material that has been formed $rom a
plurality of chemical reactants 32. Typically the reactants 32
comprise the components of phenol-formaldehyde and urea-
formaldehyde that are mixed in 3 suitable material former 34 andsupplied to an injection pump 36. Pump 36 pressure-feeds the
spray nozzle 30 through a header 38 and distributor pipes as at
40~ As a result of the foregoing process, the blanket 28
comprises a mat of glass fibers that have a binder coating of
uncured carbonaceous polymeric material.
The po1ymeric binder material is subjected to a curing
process which i5 completed by transporting the blanket 28
through ovens as at 42 and 44 on the conveyor 24. The conveyor
moves to the right in Figure 1, as indicated by the arrow 46.
~ILZ~3995
In the ovens 42 and 44, the mat of glass fibers is exposed to
elevated temperatures for a time period appropriate to effect
the cure of the polymeric material that forms the binder coaxing
on the fibers.
. In due time the cured class fiber mat 48 emerges from the
last oven 44 and is cooled by a current of air (indicated by
arrows at 50 and 52) from a fan (not shown). The mat 48 is
picked up by other conveyors indicated generally at 54 whose
movement is synchronized with the operation of a shear 56 and a
windup 58 that forms somewhat compressed rolls of the glass
fiber mat for shipment or storage.
The filamentary glass 18 is commonly ejected from the
spinner 16 at a substantially constant rate (mass per unit time)
and hence the weight (mass per unit area) of the cured mat 48
depends on the raze of travel of the chain converyor 24 as set
by a conveyor speed control 60. The cured and cooled mat is
commonly measured in a measuring zone 62 wherein there is
located a gamma-ray or x-ray gauge having a radiation source
unit 64 and a radiation detector unit 66. Radiations emitted
from the source unit 64 pass through the mat 48 and are
attenuated by absorption in the mat in accordance with its mass
per unit area. The unabsorbed radiation is detected in the
detector unit 66 to produce a detector signal that ts processed
by conventional means not shown to produce a response,
represented by a line 68, that is utilized by thy speed eontrol
unit 6G to control the speed of conveyor 24. The objective of
: this feedback control is to maintain the weight of the mat 48
nominally constant along Tts length at a desired value.
The weight (mass per unit area) of the polymeric binder
materiel contained in the cured mat 48 is controlled by a
sprayer device 7~. Device 70 may regulate the volume per unit
time of the spray material fed by the pump 36 to the spray
nozzles 30, or it may control the dilution ^~ the spray
material.
.. '
~Z~;3~335
--1 o--
The degree of polymerization, or degree of cure, of the
carbonaceous polymeric binder material is determined by the
temperature in the ovens as at 42 and 44. The temperature of
one or more of the ovens may be controlled by a temperature
control unit 72. As previously noted above, the line operator
commonly observes the "color" of the mat 48 issuing from the
last oven 44 and manually adjusts the set point of the
temperature control unit 72 accordingly.
The apparatus so far described in connection with Figure 1
is conventional, and no further explanation is deemed necessary.
Insofar as the present invention is physically embodîed in the
Figure 1 apparatus, it comprises essentia1ly the structure of an
infrared radiation gauging device in the measuring zone 62
including a source unit 74, a detector unit 76, and a data
processor arrangement 78. The data processor 78 may
automatically provide set-point values, or target values, to the
sprayer control device 70 and~or the oven temperature control
device 72, as indicated by the respective arrowheads 80 and 82.
Referring to Figure 2, the infrared source unit 74 and
2û detector unto 76 are mounted on a scanner bracket assembly 84
which permits the units 74 and 76 to traverse back and forth
across the width of the coated and cured glass fiber mat 48.
The units 74 and 76 are also movable to ~ff-sheet positions at
74A and 76A wherein radiations can pass from the source unit 74A
to the detector unit 76A without passing through the mat 48. In
this view (Fig. 2) it is assumed that the mat 48 is traveling
out of the paper, toward the observer.
The infrared source unit 74 contains a radiation source
means 86 and a filter wheel 88 that is rotated by a motor 90.
Figure 3 includes an enlarged showing of the filter wheel 88 and
source means 86. Figure 2 shows a magnetic reluctance sensor 92
(or other appropriate sensor) which detects the passing of small
iron "logic slugs" as shown at 94 in Figure 3 (or other
appropriate indicia). A signal acquisition system 96 (Fig. 2)
99~
connected via line 98 to sensor 92 uses electrical pulses from
the sensor to maintain continually updated information on the
instaneaneous position of the filter wheel 88. The signal
acquisition system 96 is a1so connected via line 100 to an
infrared radiation detector 102 in detector unit 76~
The signal acquisition system 96 is a conventional
arrangement that is described in more detail in U.S. patents
49085,326 and 4~300,049. Accordingly, the disclosure herein
shows and describes only those details necessary to permit the
present invention to be readily understood.
According to the terminology of U.S. patent 4,300,049, the
signal acqui 5 ition system 96 contains a combination of filter
wheel logic and mode switch logic. Filter wheel logic refers to
the identification and routing of signals from the detector 102
according to the angular positions of the filter wheel 88 when
the signals are generated.
As best shown in Figure 3, the filter wheel 88 contains
three filters: 104, 108 and 110. In the embodiment of the
present invention, the filter 104 is termed a reference
wavelength filter. Filters 108 and llO are different from each
other and different from filter 104, and are termed absorption
wavelength filters. As filter wheel 88 rotates, the detector
102 receîves pulses of radiation that have passed respectively
through the three ftlters in sequence.
In response to the detector cutput, the signal acquisition
system 96 delivers suitably amplified and processed signal
pulses to a line 112. In response to the pulses from sensor 9~,
the signal acquisition system 96 operates an electronic
switching arrangement 114 that routes each of ehe pulses on line
112 to the proper one of a group of sample-and-hold circuits
116, 11~ and 120.
Accordingly, the output of circuit 116 will provide a
voltage that indtcates the intensity of the radiation detected
3~95
after passing through the reference wavelength filter 104; the
voltage output or ciruit 118 will indicate the intensity of
radiation detected afer passing through the first absorption
wavelength filter 108, and the output ox circuit 120 will
indicate the intensity of radiation detected after passing
through the second absorption wavelength filter 110.
The eircuits 116, 118 and 120 hold each of their
respective, successive reference and absorption voltage outputs
for a period of time somewhat less than one full revolution time
of the filter wheel 88. Hence, the voltage outputs of circuits
116, 118 and 120 are made available to be "read" at convenient
times by a computer 122.
The computer 122 normally reads the voltages sequentially,
controlling the operation of a multiplexer 124 and analog-to-
digital converter 126 and storing the numerical values obtained.Since the sample-and-hold circuit output voltages are updated
every revolution of filter wheel 88, the signal acquisition
system 96 keeps the computer 122 informed, by flag signals fed
over a line 128, as to when the respective sample-and-hold
outputs are not available to be read.
"Mode switch logic" as contained in the signal acquisition
system 96 refers to the identification and routing of signals
from the detector 102 according to wheeher the infrared
radiation gauging instrument is operating in a measuring mode or
a standardizing mode. In tte measuring mode, the source unit 74
and detector unit 76 are positioned with the glass fiber mat 48
therebetween, so that the infrared radiation reaching the
detector 102 has interacted with the material in the mat 48. In
the standardizing mode, the source unit and detector unit are
driven by a motor 129 to their off-sheet positions at 74A and
76A clear of the mat ~8 so that the infrared radiation reaching
the detector lD2 has not interacted with the material in the
mat. Operation in the st~dardizing mode allows the computer
122 to read unattenuated values for each of the detected
-13-
radiations transmitted by filters 104, 108 and 110, for
compari 50r7 Wi th the coresponding measuring va1ues that are read
when the radiations are passing through the mat 48.
Because each processed pulse signal from detector 102 must
be individually integrated, analog signal processing of these
pulses is preferably used in the signal acquisition system 96.
This avoids the need for a high speed analog-to-digital
converter and eliminates a great many digital computations that
would otherwise be needed for carryihg out the integrations.
The sample-and-hold circuits 116, 118 and 120, used for
short term storage of the integrated analog signal values, and
other analog signal processing elements are prone to develop
extraneous signal components. Accordingly, the extraneous
signal values per se are read by the computer 122 in an offset
mode of operation that is carried out at convenient times while
the signals on line 112 are clamped to ground 130, as depicted
in Figure 2 by a transistor switch 132 controlled by a signal on
lins 134 from the signal acquisition system 96.
Computer 122 is kept informed by flag signals fed over
multichannel line 128, whether the system is in the measuring
mode, the standardizing mode, or the offset mode. In turn? the
computer 122 may exert supervisory control over the operation of
the signal acquisition system 96 as indicated by control signal
lTne 136. For example, the computer may determine appropriate
~5 times for the gauging heads 74 and 76 to go off sheet to carry
out standardization.
By operation of the Figure 2 apparatus as described, the
computer 122 receives, identifies and stores numerical
responses, corresponding to infrared radiations received by the
detector 102, as described in Table 1.
39~S
- 1 4 -
TALE 1
W~UELEN6~H
V~ MODE FILTER I_ _
Al Meas. 108 1.50
A2 Meas~ 110 1.75
R Meas. 104 1.35
A1 (AIR) Stdz. 108 1.5~
A2 (AIR) Stdz. llO 1.75
R (AIR) Stdz. 104 1.35
OFF A1 Offset 108 1050
OFF A2 Offset 110 1.75
OFF R Offset 104 1.37
The first column of Table 1 identifies the response value;
the second column indicates whether the response value is
der i ved i n the measuring (Meas.), 5 tandardizing (Stdz.), or
Offset mode; the third column indicates the filter interposed
: between the souree means 86 and the detector 102, and the fourth
column indicates the nominal wavelength passed by the respective
: filter (a narrow band interference filter).
The computer 122 also receives, as sugyested by a
connection to line 68, a response value WT indicative of the
weight (mass per unit area) of the total mat 48 as measured by
an isotopic x-ray gauge 66 (Fig. 1). The response value WT is
; actually derived by conventional means, including the x-ray
detector head 66 and computer 122, thaw per se form no part of
the present invention.
The memory of computer 122 contains stored constant values
for several parameters desTgnated by the alphabetical letters A
through H and the Greek letter I, as identified below.
Util7~ing the stored values, computer 122 forms
: 3n mathematical functions of the ratio of the A1 and R responses as
follow:
Ratio I =~R' * KS11-l (l)
lA2 J
.
39~5
15-
L Ratio 1 = Ratio 1
1 + l * Ratio 1 (2)
LikewTse the computer forms mathematical functions of theratio of the A2 and Al responses as follows
Ratio 2 = [ ! * KS2] (3)
L Ratio 2 = 1 ~a2 Ratio 2 (4)
In equations (1) to (4~,
R' - R-OFFR, A1' = A1 - OFF A19
KS1 = A2 (AIR) - OFF A2
A2 (AIR) - OFF A2
KS2 Al (AlR) - OFF A1
1 and ~2 are linearization factors, and L RATIO 1 and L
RATIO 2 are substantially linear functions of the respective
ratios. l and ~2 are usually the same, but not necessarlly so.
"Binder weight" is computed from
Blnder weight = L Ratio 1 [C * ]
A * WT B
"Binder weight" is a computed response indicative of the
total mass per unit area of the polymeric coating on the glass
fibers in the cured mat 48. This response is used to provide a
display or recording at 138, Fig. 2. It is also the basTs for
the control signal 80 for the sprayer control 70? Fig. 1. This
response is almost completely insensitive to the degree of cure
of the binder.
It is obvious that any remaining effect of degree of cure
in equation (53 could be removed by cross compensation with
equation (6). The method is essentially a solution of
2 3 9
-16-
simultaneous equations, since both equations (5) and (6)) are
linear responses to binder weight.
Also computed is "cure weight":
Cure weigh 3 L Ratio + [E * WT + F] (6)
A' * WT + B'
"Cure weight" is proportional to binder weight but is
highly sensitTve to the degree of cure.
In equations t5) and (6), the constants A, A' , 8 and B'
are used with the isotopic x-ray gauge signal WT to provide
first order compensation for the radiation path length extension
effects of scaetering, a phenomenon whose basics are discussed
in an article by Overhoff, M.W., "Infrared Gauges - Their Use,
Deficiencies and Appiications for On-Line Control," TAPPI, Vol.
56, No. 2, February 1973, pp. 70-73. A and A' are usually the
same, as are B and B' , but not necessarily so.
The constants C, D, E, and F provide first order
compensation for the spectral effects of glass on the cure
weight and binder weight measurements. The si1ica sand, from
different sources, used in glass manufacture, may have different
amounts in particular of iron oxide ~Fe+~), whose effects are
noted in the wave bands of interest. The effects are not the
same in the cure weight and binder weight channels.
An output response that is correlated with the degree of
cure is produced by combining the binder weight function and the
cure weight function, in particular by forming their ratio
2~ acco-ding to:
Cure =[CUre weight G] H (7)
Binder Weight
Here G and H are constants used to scale the cure signal
nto any reasonable units. At this time no industry-accepted
. .
2 3
17-
units exist7 so a calibration is provided in arbitrary units
from zero to one hundred. Zero designates "uncured'l material
that has been heated at low temperature for a time in order to
drive off moisture, whereas one hundred designates material that
has been heated at a high temperature for a time so that it is
definitely "overcured."
"Cure" is a computed response that is indicative of the
degree of cure, or degree of polymerization, but is
substantially independent of the weight or density of the g1ass
fiber mat 48 and the weight per unit area of binder. This
response is displayed or recorded at 140, Fig. 2, and provides
the basis for the control device 72, Fig. 1, to control the
temperature of the ovens 42 and 44.
Figure 4 shows infrared transmission curves for glass fiber
mat of about 500 grams per square meter, coated with about 50
grams per square meter of the mat, of carbonaceous polymeric
binder material, in four different stages of cure. While these
curves resemble conventional spectrometer traces, it is to be
noted that the maximum transmission is only aboue 0.5 percent.
Accordingly, my commercial spectrometer was unable to provide
any useful signals. Hence, data for the Fig. 4 curves was
derived using apparatus similar to that illustrated in Figs. 2
and 3. Etghteen different narrow band interference f;lters were
used to derive data points for each curve. The curves were then
sketched in by hand to show how I believe they would appear if
the con~inous spectra could be measured and if the absorption
spectrum effects of glass were normalized.
Curve 160 Is believed to be typical of the coated mat 28
that has been heated for one hour at 175 F to drive off
moisture but is essentially "uncured." Curve 162 indicates that
substantial curing took place when the sample was heated at 300
F for ten minutes. Substantial further curing took place when
the sample was heated at 370 F for 1.5 hours (curvy 164).
After ten minutes at 450 F curve t66), the samp1e was
definitely overcured.
~2~3~9~
As is well knowrl, for example, frorn Kendall, David N.,
Ed., Applied Infrared Spectroscopy, Reinhold Publishing Corp.,
New York, N.Y., 1966, pp. 5 and 262-267, the power P trans-
mitted by a beam of infrared radiation passing through a
material is given by
P=POe (8)
where P is the power incident on the material, a is the
absorptivity, or inherent absorbing power of the material at
a specific wavelength, b is the thickness or path length
through the sample and c is the density or concentration of
the material.
However, as appears for example from the discussion in
Kendall particularly of the spectroscopy of polymeric materials,
at certain infrared wavelengths especially the absorptivity of
certain materials undergoing chemical and/or physical changes
is not constant, but depends on the instantaneous chemical
and/or physical state of the material.
The change in the absorptivity of the polymeric binder
material as curing progresses is believed to be clearly
apparent from Fig. 4, especially at the A1 (1.5~l) wavelength.
fence the instructional value of the curves shown ap,~ears to
justify the arduousness of their preparation.
The method and apparatus of the present invention is
adapted to produce an output response to the changes in the
absorptivity a of the material. At the same time, to the
extent possible, the output response is made substantially
independent of the amount of the polymeric material as
represented by the product hc of the overall thickness and
density of the material.
- 17a -
: `
kh/ e
z~
vv~
The practice of the present invention requires direeting
into the traveling material a first infrared radiation from the
group thereof adapted to selectively interact with molecular
resonance vibrations at frequencies that are characteristic of
respective terminal functional groups of atoms involved in
reactions that take place in the material during the curing
process. This group of radiations includes a wavelength
coresponding to the fundamental vibration frequency that
commonly lies in the middle-infrared, the first overtone band
that commonly lies in the near-infrared, certain combination
bands, and possibly other harmonically related wavelengths.
The terminal functional groups of atoms involved in
reactions that take place during the cure of the most common
polymeric materials are apparently those identified in the
Crandall et al article supra, namely the 0---H, N---H and C=0
groups.
For the glass fiber mat binder cure measurement, the first
infrared radiation selected is the first overtone band in the
vicinty of 1c50 . This band is apparently adapted to
selectively interact with molecular resonance vibrations at
frequencies characteristic of both the 0----H and the No
termTnal functional groups of atoms. While the nominal
wavelength selected for filter 10~ appears to be somewhat longer
than the 0~--H and N---H overtone bands seen by Crandall et al,
for example, it is noted that the apparatus of Fig. 3 is an
efficient geometry wherein rays from an incandescent filament
142 are collected by a reflector 144 and directed through the
filters. Similarly, rays as at 148 penetrating the mat 48 at
divergent angles are collected by a reflector 146 and focused on
the detector 102. As is well known, a narrow band interference
filter exhibits a different pass band for rays passing through
at an angle than it does for normal rays. For this and probably
other reasons, the infr3red band described herein is probably
closer to those seen by Crandall et all than it appears to be.
:
~L2~35~
-19-
While the carbonyl (G=0) band just above l~9~ is sensitive
to cure, it is no preferred for the particular binder
measurement described becuse of it5 proximity to the moisture
absorption band.
The practice of the present invention also requires
directing into the traveling material a second infrared
radiation that is either of the kind that does not exhibit
substantial se1ective interaction with molecular resonance
vibrations in the material (e.g., the band in the vicinity of
1.35~ Fig. 4) or of the kind that is adapted to selectively
interact with molecular resonance vibrations at a frequency that
is characteristic of groups of atoms forming the backbones
(using the terminology of Crandall et al) of the polymeric
molecules in the material (e.g., the overtone band in the
vicinity of 1.75 I, Fig. 4). The group of atoms forming the
backbones in the present example, as in most cases, is the C---H
group whose fundamental stretch band lies at around 3.4~ in the
middle-infrared.
It is interesting to note that Crandall et al observed an
0--~ stretch band at 1.34~ in urea-formaldehyde resin, which
band disappeared on heat curing. However, as appears from Fig.
4, 7n ehe combination urea-formaldehyde and phenol-formaldehyde
resin of the present example, at least for purposes of the
binder cure measurement and binder weight measurement, the 1.35
radiation does not exhibit substantial selective interaction
with molecular resonance vibrations in the material.
The 1.75~ band is absorbed in the binder materla1 in a
manner that is substantially independent of the effects of cure,
as appears from Fig. 4. Hence, the ratio of the instrument
responses, as in equation (1), to the 1.35~ and 1.75~ bands
provides a response that is a function of the mass of the
polymeric material interacting with the radiations, and the x-
ray mass signal response WT compensations should not be
necessary except for the complications presented by the glass
-20-
fiber structure (path extension effects of scatter and the
spectral effects of the glass composition.
It now becomes apparent also that a ratio of the instrument
response to the 1.50~ band can be formed with the instrumsn~
response to either the 1.35~ band or the 1.75~ band In order
to provide a response that is proportional to binder welght and
highly sensitive to cure. Again, the gamma ray gauge response
WT as in equation (6~ should not be necessary except for the
complications introduced by the glass fiber structure and
composition. Such a "cure weight" response can be combined with
a mass response representing the mass of the polymeric material
to produce a "cure" response, and the mass response can be
derived either from the ratio of the instrument response to the
1.35~ and 1.75~ bands, from the differential response of two
gamma ray weight gauges operating at different energies, or from
some other kind of binder mass determination.
Figure 5 shows reflection curves derived from commercial
spectrometer traces. A quantity of polymeric binder material,
of the kind that is sprayed onto the glass fibers 18 at 30, W3s
deposited instead on a flat aluminum plate and heated at 145 F
for more than an hour to drive off moisture and other volatile
components. The remaining material is termed "uncured" binder,
and its reflection spectrum is drawn at 150 (the long-dashed
curve). The short-dashed curve 154 was derived from ehe sample
heated at 300 F for a half hour. The short-dash long-dash
curve 152 was derived from the sample heated at 340 F for ten
minutes. The solid curve 156 was derived from the sample heated
at 450 F for ten minutes.
The possibility of making a cure measurement by a near-
infrared reflection technique is demonstrated by the fact thatat certain wavelengths in particular the 1.48~ and 1.72~bands,
the reflected intensity changes in opposite directions with
different degrees of cure. For the reasons set forth
hereinabove, the 1.50~ and 1.75~ filters (A1 and A2 as shown)
~35~
-21-
would be appropriate. From this particular laboratory test
data, a ratio of the A2 and R responses cannot provide a
quantitative measurement of the mass of the polymeric material
that has interacted with the radiations. However, the ratio of
these responses can be correlated with the percent of binder
when the binder is applied to the glass fibers using the normal
process. In this case the ratio could be used along with the A2
and A1 ratio to provide quantitative measurement of degree of
cure by reflection.
While the invention has been described and illustrated in
the form of particular procedures and particular apparatus, the
showing and description are illustrative only and not
restrictive, since many changes and modifications can obviously
be made without departing from the spirit and scope of the
invention.
