Apparatus And Method For Detecting Streaks In Coated Layers On A Web

Abstract

A detecting head and light source are synchronously traversed on opposite sides of a continuous moving web. Photoelectric detector elements are recessed in rectangular channels disposed symmetrically about rectangular coordinate axes, the principal axis of the rectangular channels is in the direction of web travel, and centered on the detecting head opposite the center of the synchronous traversing light source. A streak defect in a coated layer on the web interferes with the transmitted light to the detecting elements resulting in an electrical characteristic signal across the detecting elements. The characteristic signal is amplified, digitized and electronically processed for verification of the presence of a streak defect. The verified signal can be used to activate an alarm, halt the coating of the web or can be recorded and used for the design of subsequent slitting operations on the web.

Description

BACKGROUND OF THE INVENTION

1. Field of the Invention

The invention relates to an apparatus and method for detecting streak defects in a wet or dry coated layer or layers on a web of material, and more particularly to detecting, indicating, monitoring and recording the location of streak defects in a coated layer or layers of photographic emulsion on a web of photographic support material.

2. Description of the Prior Art

Apparatus and methods for inspecting continuous moving webs for defects are well known in the manufacturing arts and particularly well known in the photographic materials manufacturing art.

However, most of the known apparatus and methods for inspecting photographic materials for defects are not sensitive enough to detect fine line streak defects that appear in multilayer slide coatings of photographic emulsion layers on photographic support materials. These streaks are also called pencil streaks and are caused by oversize silver grains, gelatin slugs or entrapped air bubbles in the feed slot or slide of a multilayer coating hopper as taught in U.S. Pat. No. 3,005,440 and U.S. Pat. No. 3,474,758.

Wright, U.S. Pat. No. 3,286,567, Bunge, U.S. Pat. No. 3,564,265, and Burgo, U.S. Pat. No. 3,206,606, recite apparatus for detecting streaks in sheets of photographic material. The photodetection systems employed by these three patents are based on the simple observation or sensing of signal magnitudes as the photoelectric cells traverse the moving sheet. Signals over a certain value are interpreted as being caused by streaks, and signals below that value are excluded as "noise." The apparatus of the present invention, by contrast, employs at least two very closely adjacent, symmetrically offset, recessed detecting elements which produce a sinusoidal signal comprised of a plurality of electrical pulses of opposite polarity characteristic of a streak. The signal is then processed by characteristic signal verifying circuits. Since both detecting elements look at the same general area of the web due to their close proximity, gradual variations in coating weight, web temperature, etc., produce no signal. Another advantage lies in the ability to determine the type of streak from the characteristic signal and to take proper corrective action to eliminate the streak source.

SUMMARY OF THE INVENTION

It is an object of the invention to detect irregularities in a web or length of sheet material and, more particularly, to detect the formation of streaks in a coated layer or layers of photographic emulsion and associated coated layers on a continuously moving web of photographic support material just subsequent to the coating process. Other objects of preferred embodiments of the invention are to analyze and identify the particular type of streak that is being formed so that timely corrective measures can be taken to minimize the quantity of defective material coated, and to monitor the presence and record the location of streak defects in the web and use this information for the design of slitting patterns to maximize the quantity of sheets obtainable from a web of photographic material.

These and other objects of the invention are achieved by an apparatus for detecting irregularities in a web comprising:

1. a detecting head having at least two symmetrically offset channels with photoelectric detecting elements recessed therein, preferably to a depth of at least 1/2 inch,

2. a light source disposed on the side of said web opposite said detecting head,

3. a traversing mechanism for moving the detecting head across the width of said web, and

4. means for sensing the electrical signal produced by said photoelectric detecting elements.

Apparatus of the type just described is capable of producing an electrical signal characteristic of streak defects in order to accomplish the objects of the invention and is particularly suited for detecting irregularities in a continuously moving web. The means for sensing the signal of the detecting elements may include electronic amplifying and waveshaping circuits for further characterizing the signal for subsequent verification; electronic digital circuits for verifying and categorizing the type of streak defect; and recording instruments for monitoring the presence and location of streak defects in the coated layers on the web.

The invention also includes a method of detecting irregularities in a web of material, especially streak defects in a coated layer or layers of photographic emulsion on a continuously moving web of photographic support material, comprising:

a. transmitting light from a light source through said web to a detecting head having at least two symmetrically offset channels with photoelectric detecting elements recessed therein, said photoelectric detecting elements being adapted to produce sequential electrical pulses of opposite polarity when traversed by a web irregularity,

b. sensing the time sequence and amplitude of said electrical pulses to verify the existence of a web irregularity, and, optionally,

c. recording the electrical pulses indicating the presence and location of streak defects in said coated layer or layers.

The apparatus and method of this invention detect and verify the formation of streaks in the coating process. The verified streak signal may be used to activate an alarm and indicate remedial action to eliminate the cause of the streak formation. The recorded information, i.e., the presence and location of streak defects, may be used in the design of slitting patterns for the coated material and thus increase the yield of acceptable material from the coated material.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a perspective drawing of a traversing mechanism, detecting head and light source of an embodiment of the invention.

FIG. 2 is a detail perspective drawing in partial section of a detecting head and light source of an embodiment of the invention.

FIG. 3 is a schematic drawing of the detecting elements and amplifying and waveshaping circuits of an embodiment of the invention.

FIG. 4 is a representative drawing of a streak defect and a characteristic streak signal of an embodiment of the invention.

FIG. 5 is a block diagram of the electronic signal processing circuits of an embodiment of the invention.

FIG. 6 is a timing diagram showing the signal processing operations of an embodiment of the invention.

DESCRIPTION OF THE PREFERRED EMBODIMENT

The apparatus of the invention employs a detecting head having at least two channels. The channels are preferably rectangular and symmetrically offset about the origin of coordinate axes as shown in FIG. 4. For the detection of streaks the channels may also be symmetrically offset about an axis, e.g., the Y axis of FIG. 4. If such a configuration is used a thin opaque wall is required to separate the channels. Alternatively any intermediate configuration may be used with a thin opaque wall separating the channels where they overlap. The preferred configuration, wherein the channels are symmetrically offset about the origin of coordinate axes, simplifies construction of the detecting head and enables the apparatus to detect "chatter bars." Chatter bars are caused by uneven application of the coating layer as the web passes the coating station and appear as bars laterally across the web. The preferred distance between centers of the channels is equal to the width of the individual channels, and the width of said channels is selected to correspond with the width of the streak to be detected. The preferred depth of the channels, i.e., the distance the photoelectric detecting elements are recessed in the channels, is related to the width of said channels by a ratio of at least 16:1. The length of the channels is not critical, but a ratio of length to width of at least 10:1 is preferred. In the preferred embodiment the width of the channels is 1/32 inch, the depth of recess of the detecting element is 1/2 inch, and the length of the channels is 10/32 inch. The channels function to substantially collimate the light incident on the detecting elements by admitting the transmitted light normal to the web and screening out the diffused light transmitted through the web.

The light source is preferably adapted to be moved synchronously with the detecting means across the web but on the opposite side of the web so that the detecting means and light source are always in register. An infrared source is preferred.

The geometrical configuration and symmetrical characteristics of the channels in the detecting head cause an electrical signal to be produced across the differentially connected detecting elements, recessed therein, when a streak defect is traversed. The electrical signal is substantially sinusoidal with the polarity sequence determined by the type of streak being traversed. The means for sensing this signal is any element, alarm, indicator or circuit which is responsive to the signal and which provides a readout of the signal being sensed or records or uses the signal to govern other circuitry or operations. For example, the detecting elements may be connected to electronic amplifying and waveshaping circuits for further characterizing the signals which are then processed by electronic digital circuits for verifying and categorizing the type of streak defect as further described below. These circuits may be used to supply the input data to recording instruments for monitoring the presence and location of streak defects and/or for operating a web slitter or sorter such as described in U. S. Pat. No. 3,286,567.

A preferred embodiment of such apparatus is described below with reference to the attached drawings, in which the same numerals are used throughout.

The detecting head 15 and light source 12 are disposed, respectively, on upper and lower traversing guide tubes 13 and 14, of a suitable traversing guide apparatus with stanchions 10 and 10' as shown in FIG. 1. A suitable traversing guide apparatus is available from the Du Pont Co., Instrument Products Division, Photo Products Department, Wilmington, Delaware 19898, as a Model 110 Traversing Mechanism.

A continuous moving web W of photographic support material, e.g., polyethylene terephthalate, having a coated layer or layers of photographic emulsion is conveyed by suitable web handling means, e.g., drive rolls, between the detecting head 15 and light source 12, which are synchronously moved back and forth across the width of the web.

Detecting head 15 has narrow rectangular channels 16 and 17 disposed about rectangular coordinate axes X and Y, centered on said head, axis Y being parallel to the direction of web travel. Photoelectric detecting elements 18 and 19 are preferably recessed in said channels away from the web surface, and are located in end-to-end spaced relationship parallel to the direction of web travel (Y) and are separated by a distance d, measured from center to center in the direction of travel of the detecting head (X), as shown in FIG. 2. The symmetrically offset relationship of the channels is defined as either the lateral offset which functions to detect longitudinal streaks in the coated layer on the web, or the longitudinal offset which functions to detect defects in the coated layer running laterally across the web. Although only one type of defect, i.e., longitudinal streaks or lateral bars may be of importance, either may be detected if the channels are symmetrically offset about the origin of coordinate axes. The channels may be symmetrically offset only about a longitudinal axis if only longitudinal streak defects are to be detected. The channels may be symmetrically offset only about a lateral axis if only lateral bars are to be detected. In any of the three configurations above, the symmetrical offset relates to either a single axis or the origin of coordinate axes. Web edge sensors 11 and 11' are disposed on the lower surface of said head and spaced respectively from each channel a minimum distance. The head and particularly the inner surfaces of the channels are painted with black nonreflective paint.

The light source 12 has a lamp 21 suitably mounted in holder 20 and a condensing lens 23. A frosted glass 24 is disposed at the focal point of said lens and a filter glass 25 is suitably mounted in the light source.

Preferably the detecting head, particularly the inner surfaces of said channels, is painted with black nonreflective paint. Lead sulfide detectors have been found to be suitable detecting elements. A suitable lamp has been found to be a Type MAZDA manufactured by Westinghouse. It operates at 28 volts giving 21 candle power, and has a silvered envelope directing radiation only through the front frosted window 22. The condensing lens is a 3-element f/0.62 lens with a diameter of 43.5 mm. Its transmission should cover 2.5.mu.. The filter used is a Type RG1000 manufactured by Schott and Jema, of Mainz, West Germany. Its transmittance is from approximately 1.2.mu. to 2.7.mu.. The frosted glass disposed at the focal point of the condensing lens improves the uniformity of illumination from the light source.

The photoelectric detecting elements are differentially connected, by which it is meant that their combined output signal is proportional to the difference between the absolute values of the signals produced by the elements. Although equivalent arrangements may be employed, a convenient way to accomplish this is by means of the circuit illustrated in FIG. 3. The photoelectric detecting elements 18 and 19 are differentially connected between +15 volt and -15 volt power supplies. Their common node is connected to the plus input of operational amplifier 53 through a differentiating network comprising capacitor 51 and resistor 52. Operational amplifier 53 and the feedback network comprising capacitor 56, resistor 55 and resistor 57 together with the differentiating network constitute an active band pass filter.

The detection of a streak signal and formation of the characteristic waveform can be readily understood by reference to FIG. 4. The web W is traversed by the detecting head and light source traveling synchronously from right to left. Assume the streak S is lighter than the adjacent areas of the web W, when the streak is between the light source and a particular rectangular channel the transmitted light to the photoelectric detecting element recessed in the channel is increased, the electrical resistance of the detecting element is decreased, and the voltage drop across the particular detecting element is decreased. A differential electrical signal Vn is produced at the node between the photoelectric detecting elements.

Rectangular channel 16 traverses the streak 15, the voltage across detecting element 18 decreases; the voltage at the common node goes positive, achieves a maximum value when the streak S is centered under channel 16, and then decreases to zero. Rectangular channel 17 traverses the streak S, the voltage across detecting element 19 decreases, the voltage at the common node goes negative, and achieves a maximum negative value when the streak is centered under channel 17, and then decreases to zero. The voltage signal at the common node Vn is differentiated to form the characteristic signal Vo.

If the streak defect S were darker than the adjacent areas of web W, the characteristic signal would be inverted.

Trim potentiometer 50 is provided to balance the response of photoelectric detecting elements 18 and 19.

An operational amplifier that has been found particularly suitable for use in the active band pass filter is the Model 3064/12C manufactured by Burr-Brown. The frequency of the characteristic signal is determined by the speed at which the detecting head and light source traverse the web and the distance between centers of the rectangular channels. The preferred width of the rectangular channels is 1/32 inch on either side of the Y axis, and the distance between centers (d) is also 1/32 inch. The preferred traversing speed is 3 inches per second. The resulting frequency of the characteristic signal is approximately 50 Hertz, and the pass band of the filter is designed for this frequency. Spurious noise signals and defect signals that are not within the frequency response range of the active band pass filter are attenuated and not processed by the characteristic signal verifying circuits.

In accordance with the process of the invention, the characteristic signal is converted to digital logic level signal pulses from which logic level gating pulses are produced. The signal pulses are then gated with the gating pulses to verify the time sequence of the sequential pulses produced by the detecting elements to produce a verified defect signal. This process is described in greater detail infra with reference to the drawings.

The characteristic signal Vo is applied to the sensor converter circuits 103, 113 and 123. Sensor convertors 103 and 113 operate on the positive and negative portions of a characteristic signal, respectively, to convert the analog signal to digital logic level signals with pulse width equal to the time duration that positive and negative peaks of the analog signal are greater in magnitude than threshold levels. The logic level signals are applied to inverters 104 and 114. The output of inverter 104 is applied to a delaying gate generator comprising delay timer circuits 105 and 106, AND gate 109, and also to the clock input of flip-flop 120. The output of inverter 114 is applied to a delaying gate generator comprising delay timers 115, 116 AND gate 119, and to the clock input of flip-flop 110, all as shown in FIG. 5.

The output of AND gate 109 is applied to the data input of flip-flop 110, and the output of AND gate 119 is applied to the data input of flip-flop 120.

The one state output of flip-flop 110 is applied to AND gate 112 and the data input of flip-flop 125. The zero state output of flip-flop 110 is applied to the reset input of flip-flop 110 through delay timer 111.

The one state output of flip-flop 120 is applied to AND gate 112 and the data input of flip-flop 126. The zero state output of flip-flop 120 is applied to the reset input of flip-flop 120 through delay timer 121.

The one state output of flip-flops 125 and 126 are applied to indicator lamps 136 and 137, respectively, through noninverting amplifiers 127 and 128. The zero state outputs of flip-flops 125 and 126 are applied to isolated AC switch 133 through OR gate 122, inverter 129 and switch 143. Isolated AC switch 133 applies 110 volts to the traversing drive motor (not shown). The direction of the traversing motor is determined by isolated AC switches 131 and 132 and flip-flop 30. The one state output of flip-flop 30 is applied to isolated AC switch 131, and the zero state output of flip-flop 30 is applied to isolated AC switch 132.

The one state and zero state outputs of flip-flop 30 are also applied to indicating lamps 138 and 139, respectively, through noninverting amplifiers 134 and 135.

The output of edge sensors 10 and 11 is applied to sensor converter 124, and the output of sensor converter 124 is applied to the inhibit input of inverters 104 and 114.

The characteristic signal is also applied to sensor converter 123. The output of this sensor converter indicates the presence of a defect, but is not processed for verification as a streak defect. It may be used to drive an indicating lamp.

Potentiometer 143 is mechanically coupled to a gear or pulley (not shown) on the traversing mechanism drive, and is calibrated to provide transverse position data of the detecting head to a recording apparatus.

All of the digital electronic circuits, i.e., sensor converters, inverters, delay timers, AND gates, OR gates, flip-flops and isolated AC switches, are available from the Digital Equipment Corporation as K-Series Control Modules, and their operation is more fully described in the Digital Control Handbook for 1971 published by that company.

The operation of the characteristic signal electronic processing digital circuits can be readily understood with reference to the timing diagram of FIG. 6.

The characteristic signal Vo is applied to sensor converter 103. The positive going portion of Vo crosses the positive threshold L.sup.+ at time T.sub.o causing the output of converter 103 to go to the one state and remain in the one state until time T.sub.1 when the characteristic signal falls below threshold level L.sup.+. The characteristic signal Vo is simultaneously applied to sensor converter 113. The negative portion of Vo crosses the negative threshold level L.sup.- at time T.sub.3 causing the output of converter 113 to go to the one state and remain in the one state until time T.sub.5 when the characteristic signal falls below the threshold level L.sup.-. The characteristic signal Vo again exceeds the threshold level L.sup.+ at time T.sub.7 causing the output of converter 103 to again go to the one state until Vo falls below the threshold level L.sup.- at time T.sub.9.

Logic level signal S103 is inverted producing signal S104 which is applied to delay timers 105 and 106 of a delaying gate generator. The inverted output of delay timer 105 and the noninverted output of delay timer 106 are applied to AND gate 109. The output of AND gate 109, signal S109, is a gate pulse having a pulse width T.sub.4 -T.sub.2 and delayed T.sub.2 -T.sub.o seconds from the time T.sub.o that Vo first crossed the threshold level L.sup.+. This delayed gate pulse S109 is applied to the data input of flip-flop 110, and it may be thought of as a window through which flip-flop 110 may view the occurrence of a negative threshold crossing by the characteristic signal Vo. At time T.sub.3 the characteristic signal Vo crosses the negative threshold level L.sup.-, producing logic level signal S113, and inverted signal S114. The negative transition of S114 falls within the delayed gate pulse of S109 and sets flip-flop 110 to the one state. Signal S114 is also applied to delay timers 115 and 116 of a delaying gate generator. The inverted output of delay timer 115 and the noninverted output of delay timer 106 are applied to AND gate 119. The output of AND gate 119, signal S119, is a gate pulse having a pulse width T.sub.8 -T.sub.6 and delayed T.sub.6 -T.sub.3 seconds from time T.sub.3 that Vo first crossed the negative threshold L.sup.-. This delayed gate pulse S119 is applied to the data output of flip-flop 120, and it also may be thought of as a window through which flip-flop 120 may view the occurrence of a positive threshold crossing by the characteristic signal Vo. At time T.sub.7 the characteristic signal Vo crosses the positive threshold level L.sup.+, producing logic level signal S103. The negative second transition of S104 falls within the delayed gate pulse of S119 and sets flip-flop 120 to the one state.

The coincident one state outputs of flip-flops 110 and 120 applied to AND gate 112 produce a logic level pulse signal S112 indicating a verified streak defect. The verified logic level streak defect signal is supplied to a suitable recording apparatus, along with the lateral position signal from potentiometer 143. Longitudinal position of the verified defect may be obtained by calibrating the recording apparatus with the web speed.

The zero state output of flip-flop 120 is applied to delay timer 121 and to the clock input of flip-flop 125. The output of delay timer 121 is a delayed pulse S121 which is applied to the reset input of flip-flop 120. The one state output of flip-flop 110 is applied to the data input of flip-flop 125, and flip-flop 125 is set to the one state by the negative transition of the zero state output of flip-flop 120.

The one state output of flip-flop 125 is applied to noninverting amplifier 127 which drives indicating lamp 136. The zero state output of flip-flop 125 is applied to OR gate 122 and through inverter 129 and switch 143 controls isolated AC switch 133 and the 110 volt power to stop the traversing motor drive. This circuit may be disabled by opening switch 143.

The one state output of flip-flop 30 is applied to indicator lamp 138 through noninverting amplifier 134, this lamp indicates the detecting head and light source are traversing from right to left.

The diagonal configuration of lamps 138 and 136 indicate the streak signal is due to a light streak.

If the streak defect S were darker than the adjacent areas of web W, the characteristic signal would be inverted, and the sequence of threshold crossing would be L.sup.-, L.sup.+ and L.sup.-. This opposite sequence is similarly processed by the characteristic signal processing digital electronic circuits and lights lamps 138 and 137.

Flip-flop 125 must be manually reset by pushbutton 144 to clear the apparatus after a streak signal has been verified.

The edge sensors 11 and 11' inhibit inverters 104 and 114 when the detecting head and light source traverse beyond the edge of the web and prevent confusing the edge of the web with a streak defect.

Although the preferred embodiment of the invention is described for a transmitted light signal and a translucent web, it will also operate with a reflected light signal on an opaque web. And although the invention is preferably used immediately subsequent to the coating process where the coated layers are still wet, other uses are contemplated where the web may be a textile material or the coated layer or layers may be others than a photographic emulsion.

Although the invention is described with particular reference to a continuous moving web and longitudinal streaks or lateral bars in coated layers thereon, its application is not so limited. With suitable modification to the means for traversing, the apparatus may be used for inspection of stationary sheets. It also may be used for inspecting textile webs or other web material for longitudinal and lateral type defects.

Claims

I claim:

1. An apparatus for detecting irregularities in a web of material comprising:

1. a light source disposed on one side of said web,

2. a detecting head having at least two symmetrically offset channels and photoelectric detecting elements recessed therein, said channels substantially collimating the light transmitted through and normal to said web from said light source and sequentially screening any variation in transmitted light incident on one photodetecting element from another photodetecting element,

3. a traversing mechanism for moving said detecting head across the width of said web, and

4. means for sequentially sensing the amplitude and polarity of the electrical signal produced by said photoelectric detecting elements.

2. The apparatus of claim 1 wherein said light source is moved synchronously with said detecting means across the width of said web.

3. The apparatus of claim 1 wherein said symmetrically offset channels are rectangular, the distance between centers of said channels is substantially equal to the width of said channels, and the depth to which said detecting elements are recessed in said channels is at least 16 times the channel width.

4. The apparatus of claim 1 wherein said means for sensing comprises means for converting the electrical signal produced by said photoelectric detecting elements into recordable form, and means for recording the converted signal.

5. The apparatus of claim 4 wherein said means for sensing comprises a band pass amplifier including a differentiating network for differentiating and amplifying the signal from said detecting elements, and means for converting the amplified signal to a digital logic level signal, and means for recording said digital logic level signal.

6. An apparatus as recited in claim 1 wherein said detecting elements are recessed in said channels to a depth of at least 1/2 inch.

7. A method of detecting irregularities in a web of material comprising:

a. transmitting light from a light source through said web to a detecting head having at least two symmetrically offset channels with photoelectric detecting elements recessed therein, said photoelectric detecting elements being adapted to produce sequential electrical pulses of opposite polarity when traversed by a web irregularity, and

b. sensing the time sequence and amplitude of said electrical pulses to identify the existence of a web irregularity.

8. A method as recited in claim 7 wherein said electrical signal is substantially sinusoidal and said sequential pulses are of alternate polarity, and said signal is differentiated to enhance the characteristics representative of said irregularity.

9. A method as recited in claim 7 wherein sensing includes converting said electrical signal to digital logic level signal pulses, generating logic level gating pulses from said signal pulses, gating said signal pulses with said gating pulses to verify the time sequence of said sequential pulses to produce a verified defect signal.

10. The process as recited in claim 9 including the additional step of recording the verified defect signals and their lateral and longitudinal location on the web.