| 摘要 |
1473920 Photo-electric flaw detectors INTEC CORP 13 May 1974 [8 March 1974] 21008/74 Heading G1A Flaw detection apparatus comprises: a laser arranged to provide a radiation beam; means for scanning the beam across a test surface in multiple lines, the scanning velocity at the test surface varying in dependence upon the scan position; radiation detecting means; a discriminator for providing a flaw signal in response to a predetermined output of the detecting means; and delay means arranged to provide a flaw indication when a flaw signal is first generated at any position on a given first scan and to inhibit an indication due to a flaw detected during a time gate at a corresponding position on at least one subsequent scan, the time gate being of variable duration in dependence upon the scan position. A moving test object 20 is scanned Figs. 2, 3, 4 (not shown) in lines perpendicular to the direction of motion by a beam generated by a laser 12, Fig. 1 and deflected by a galvanometer mirror 28, 30 driven sinusoidally by a 1 KHz drive generator 44 controlled by a 2MHz master oscillator 42. Light received by the object 20 either by transmission, refraction, reflection or scatter, is collected by a receiver 50 and Figs. 3, 5 (not shown) having a combination e.g. Fresnel and cylindrical lenses for imaging only the immediate region of the scanned line on a photo-multiplier 60 so as to minimize noise due to stray light. A small portion of the scanning beam is directed to a centre pick-up 40 provided with a slit for generating a pulse each time the scanning beam passes its precise centre position, and thus irrespectively of any phase lag due to the electro-mechanical characteristics of the galvanometer 20. The photo-multiplier 60 output, after processing in a normaliser which provides standard signals representing respectively different kinds of fault (e.g. a peak or a dip in brightness), is applied to a flaw amplitude discriminator 90 which emits a flaw signal in the form of a pulse whenever the input exceeds a predetermined threshold. A flaw signal first detected at any position during any scan is applied to a multiple scan memory and flaw quantizer 110, illustrated in Fig. 12. Here it passes an AND gate 128, normally enabled during outward scans, to provide a quantized flaw indication, e.g. in a counter. At the same time the flaw signal is applied to the multiple scan memory 112 which stretches its duration and delays it so as to provide a gating signal spanning a corresponding position on at last one succeeding scan. During the delayed gating interval a NOR gate 130 is enabled, so as to disable the AND gate 128 and so inhibit a repeated count of the same fault. As illustrated, inhibition lasts for four scans succeeding the one wherein the fault was first detected. Only after four successive scans wherein no fault is signalled during the gating interval, do all inputs to the NOR gate 130 return to zero, enabling the apparatus once again to count a fault in that region. To ensure that inhibition is not interrupted because of fluctuating signal amplitudes, the zerolevel output from the NOR gate 130 is inverted at 138 and used as a feedback signal 105 to reduce the threshold of the discriminator 90 Fig. 1 and so enhance the sensitivity of the apparatus during the gating interval on the successive scans. To reduce variations in the width of the gated segment due to the sinusoidal variations in scanning velocity, a variable flaw gate control circuit 70 Fig. 1, illustrated in Fig. 10 subdivides the outward scan time into intervals X, Y and Z, providing signals to the flaw gate generators 112, Fig. 12 to alter the amount by which the flaw signal is stretched. |
