ULTRASONIC NONDESTRUCTIVE TESTING SYSTEM AUTOMATION

D. H. Gray D. M. Darby

O A K RIDGE Y-12 PLANT O A K RIDGE. TENNESSEE

DISCLAIMER

This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency Thereof, nor any of their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness, or usefulness of any information, apparatus, product, or process disclosed, or represents that its use would not infringe privately owned rights. Reference herein to any specific commercial product, process, or service by trade name, trademark, manufacturer, or otherwise does not necessarily constitute or imply its endorsement, recommendation, or favoring by the United States Government or any agency thereof. The views and opinions of authors expressed herein do not necessarily state or reflect those of the United States Government or any agency thereof.

DISCLAIMER

Portions of this document may be illegible in electronic image products. Images are produced from the best available original document.

Printed in the United States of America. Available from National Technical Inforr~mtion Service

U.S. Department of Commerce 5285 Port Royal Road, Springfield, Virginia 22151

Price: Printed Copy $5.45; Microfiche $0.95

This report was prepared as an account of work sponsored by the United States Government. Neither the united States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that i t s use would not infringe privately owned rights.

Reference to a company or product name does not imply approval or recommendation of the product by Union Carbide Corporation or the U.S. Atomic Energy Commission to the exclusion of others that may meet specifications.

ULTRASONIC NONDESTRUCTIVE TESTING SYSTEM AUTOMATION

D. H. Gray D. M. Darby

Oak Ridge Y-12 Plant P.O. B o x Y , O a k R i d g e , T e n n e s s e e 37830

Date Issued - September 10, 1973

N O T I C E This' report was prepared as an account of work spnnsnred by the Unitod Statca O G V C ~ ~ I I I I ~ I I ~ . Igelrher the United States nor the United States Atomic Energy Commission, nor any of their .employees, nor any o f their contractors, subcontractors,. or their employees, rnnkos annu war.h.alliy, express or ~mplied. or assumes any legal liability or responsibility for the accuracy, com- pleteness or usefulness of any information, apparatus, product o r process disclosed. or represents that its use would not infringe privately owned rights.

P r e p a r e d f o r the U.S. A t o m i c E n e r g y Commission U n d e r U.S. G o v e r n m e n t C o n t r a c t W - 7 4 0 5 e n g - 2 6

DISTRIBUTION M I S DOCUMENT IS UNUM\T 4r

ABSTRACT

A commercial ultrasonic test facility has been interfaced to a medium-sized multiuser computer. The large volumes of data, formerly correlated to flaw size and flaw location manually, are now automatically accumulated and analyzed, and the results are plotted within a matter of minutes. The control system is capable of positioning the ultrasonic transducer with four degrees of .motion-three linear and one angular. The data handling system offers the flexibility to allow implementation of nearly any future test requirements.

The system i s now operational and is in daily use in a development laboratory. Surfaces of revolution and flat plates have been inspected with good results. Fourier transforms of ultrasonic waveforms are handled routinely. The multiuser computer configuration gives the user the impression of having a dedicated computer a t his disposal.

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . SUMMARY 4

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

DATA HANDLING AND CONTROL SYSTEM . . . . . . . . . . . . . . . . . . . . . 6

Hardware . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Data and Control of Terminal . . . . . . . . . . . . . . . . . . . . . . . . . . 6 Computer Interface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 System Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Basic Software Configuration . . . . . . . . . . . . . . . . . . . . . . . . . . 14 Software Descriptions and Logic Flow . . . . . . . . . . . . . . . . . . . . . . 15 Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

BIBLIOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

APPENDIX A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

Data and Control Terminal Schematics and Logic Diagrams . . . . . . . . . . . . . . . 23

APPENDIX B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

Computer Interface Schematics and Logic Diagrams . . . . . . . . . . . . . . . . . . . 35

APPENDIX C . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

Terminal Commands . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

APPENDIX D . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

Software Flow Charts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

APPENDIX E . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

Operator Instructions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

SUMMARY

A commercial ultrasonic test facility was interfaced to a medium-sized mu1 tiuser computer, and programs were written to accomplish data taking and analysis for all presently used test methods. The system is operational and in daily use. The ultrasonic datalcontrol terminal was designed to incorporate a high degree of flexibility, thus making it easier to add any future tests to the system. In addition to normal test data, the system can supply a digitized ultrasonic R F waveform to the computer for Fourier analysis. All test results can be plotted on a local X-Y plotter or strip-chart recorder. Surfaces of revolution and flat plates have been inspected with good results. The transducer position may be changed between data scans via the computer. The capability exists for tracking a complex geometric shape with the transducer, provided the proper part description program is available.

INTRODUCTION

Ultrasonic inspection methods utilize ultrasonic sound transmitted through a medium to detect density discontinuities internal to a part or a joint between two parts. A portion of the ultrasonic sound wave is reflected at each discontinuity or flaw, with the magnitude of the reflected energy dependent upon flaw size. Because of the small cross-sectional area of the beam, numerous passes must be made at different latitudes or positions of interest on a part. Large volumes of data are generated which must be analyzed and correlated to find the flaws at various locations.

A remotely located medium-sized multiuser computer was interfaced to a commercial ultrasonic test facility to accumulate, analyze, and manipulate data as well as to plot results, all in a matter of minutes. The multiuser computer configuration makes an efficient system in an intermittent-use environment, supplies a high degree of computing power, and gives the user the impression of having a dedicated computer at his disposal. The system is now operational and is in daily use in a development laboratory a t the Oak Ridge Y-12 The test facility will serve as a prototype instrument to develop new techniques to meet present and future design agency component inspection requirements. The capability also exists to position the transducer such that intricate geometric part shapes may be examined under program control. Radio frequerlcy (RF) waveforms are digitized and sent to the computer which performs a Fourier analysis on the waveform. The ultrasonic datalcontrol terminal i s characterized by a high degree of flexibility which will allow the implementation of most future test requirements.

(a) Operated by the Union Carbide Corporation's Nuclear Division for the US Atomic Energy Comniission.

DATA HANDLING AND CONTROL SYSTEM

HARDWARE

An ultrasonic transducer is excited by a pulse and transmits a burst of high-frequency energy through a suitable couplant to the part under test. Density discontinuities. encountered by the energy signal cause a portion of the energy to be reflected back to the transducer. The type of discontinuity will determine the amount of reflected energy. For example, an air pocket in a part whether caused by a crack or a poor bond between two surfaces will cause almost 100 percent of the incident energy to be reflected. The reflected energy signal i s received by the transducer and directed to an lmmerxope which determines the peak amplitude of the signal. 'I he lmmerscope the11 yelierates a gate signal, meaning that the analog 'equivalent of the peak signal amplitude i s ready on the analog output line. Two separate lmmerscope channels with independent windows allow differerst gains to bc employed or portions of the returning signal to be monitored. There is an analog output line and gate signal for each of the channels (A and B). A view of the ultrasonic test system is presented in Figure 1.

Data and Control Terminal

Various test instruments are envisioned to be used in place of the Immerscope. These instruments have varying analog output levels which require the capability of offsetting the signal. A dual-channel analog scaler (Figure 2) was developed to permit the test instrument analog output signals to be offset, scaled, and inverted. The scaler inputs are protected against amplitudes in excess of k 15 volts. Succeeding stages require the scaler outputs to be between 0 and +10 volts. A negatlve output detector prevents negative values from reaching the succeeding stages and notifies the operator to change the scaler polarity switch via a light-emitting diode (LED). The scaler outputs are fed to an eight-channel multiplexer on Channels 0 and 1. Figure 3 is a bloclc diagram which shows the signal flow for thc system. Channels 2 and 3 of the multiplexer are connected to the "sample" and "hold" amplifiers which may be connected to external devices such as position potentiometers on another ultrasonic testing tank used for testing small parts. Channels 4 through 7 are available for future expansion of the system's capabilities. When a multiplexer channel is selected, the analng value on its input is delivered to an analog-to-digital converter (ADC) which converts the analog voltage to a digital value for tfansmissiur-I to the computer.

In order. to define the position on a part a t which data were taken, a data synchronization (sync) mode must be selected to suit the particular test. Each data sync mode selects a different source of sync pulses which enable the terminal to take the next analog datum presented on the analog outputs of the ultrasonic test instrument. The first mode is encoder sync which selects the rotating table encoder. The encoder makes one revolution for each revolution of the table, and each encoder revolution generates 3,600 sync pulses plus a position zero marker. Each sync pulse is equivalent to 0.1 degree of table revolution. Table '

speed and hence the encoder pulse reperition Irequency is much slower than the frequency of the Immerscope, which gives good assurance that the lmmerscope will have an arialog value ready to be read that corresponds to the position on a part represented by the encoder pulse.

1m47 F i r 2. ANALOG SCALER.

The other data modes are internal, 'external, and flat plate. The internal mode selects an internal 50-hertz clock arid the-external mode selects an externally supplied signal. The flat-plate mode qlectp the Y-axis. encoder where'each pulse from' the encoder is equivalent t o four mils of linear travel. Each of these three data sync modes are scalable; that is, a programmable number of pulses must be received from the selected source by the terminal before a sync pulse is issued. The length of a data scan is similarly programmable and is defined as the number o f sync pulses (or data points) which wil l be accumulated in one direction along a part. The programmed number may be any number between 1 and 4,095.

Figure 3. SIGNAL FLOW FOR THE ULTRASONIC TEST SYSTEM.

There are two basic classes of parts presently being tested by the system. One is a flat plate where the transducer is moved across the plate with a scan defined as one excursion across the part. The second is a surface of revolution where a part is mounted on a rotary table with the transducer fixed during each. scan. Here, a scan is defined as one revolution of the part. Figure 4 provides a pictorial sketch of these methods showing the allowable paths of transducer movement. When a scan is completed, the transducer position may be changed under computer control. The X I Y, and Z axes and transducer angle (A) are driven by stepping motors which may be controlled rnariually w by the computer. (See Figures 1 and 4 for axes identification.) Under computer control, the number of steps required for each

Transducer Path -

Transducer

F i r 4. T E M MODE&.

L 0 .a .- x 4

axis movement are loaded by the computer into a 12-bit down counter associated with each motor. When commanded by the computer, a 5QHz clock is fed to both the counter and a translator. The translator supplies the correct pulse sequence to the motor. When the counter overflows, signifying that the motor has taken the required number of steps, the clock is disabled, and the motor stops and holds its last position. This same sequence occurs simultaneously for the other motors. When all axes have been driven to their commanded positions, a "transducer arrived" flag is sent to the computer so that the next data scan may be taken. Each step of a linear axis is equivalent to four mils, while those of the A axis are equivalent to 0.05 degree of rotation. The dashed transducers in Figure 4 represent possible positions the transducer might be required to assume in order to test the indicated latitudes of the part.

1 S .- 4-

2 0 Rulur y 5 ez Table

A data scan is begun by an "initiate" cammarid which, in the w ~ f ! of the internal and external data sync modes, is accomplished via a switch on the front panel of the datalcontrol terminal seen in Figure 5. In the flat plate mode, "initiate" is generated by a minus-to-plus polarity change of the Y-axis position display. The encoder sync mode generates the "initiate" command with the first zero position marker from the rotary table encoder. In all of the modes, a "scan complete" flag is sent to the computer at the end of the data s ca r .

An internal priority system determines the order in which &ta flow to the computer. There are two ways in which data may be requested. One is when the lmmerscope issues the gate s i~ la l signifying that data are ready; the other is when the computer requests a channel of the analog multiplexer to be read. Multiplexer Channels 0 and 1 are self addressing while Channels 2 through 7 are addressed only by the computer. The priority system is set up such that lmmerscope Channel A (multiplexer Channel 0) has precedence over Channel 6 (multiplexer Channel 1) and the computer request Channel 6 is next in line, with the computer request having the lowest priority.

Along with each data word sent to the computer are three identification bits which identify the analog multiplexer channel from which the data originate. Data consist of ten bits, therefore a complete data word consists of thirteen bits which includes the three identification bits Each data word is deposited in a table in the computer memory through

START STOP =SET T

I J 2 I! I

N C G U N T E R

E S E T

@'2E UP/DWN COUNTER

STAR1 STOP RESET

- NTROL TERMiNA

Figure 5. DATNCONTROL TERMINAL.

START STOP RESEr

C

a cycle-stealing- process known as direct memory access (DMA). This process steals a four-m icrosecond cycle from the computer, automatically increments a word counter, determines the memory location of the next-word of the table, and deposits the data word in this memory location. When the word counter werflows, the computer is signalled, via an automatic priority interrupt (API), that the table is full. Thus, the computer is not bothered with the timeconsuming process of servicing an interrupt for each data word deposited.

A peak amplitude detector is available for finding the peak amplitude obtained during calibration. The detector compares the first data word with zero. If the first word is greater than zero, it is stored in place of zero. The second data word is then compared with the first and the second stored in place of the first if it is larger. This sequence continues until the stored word becomes larger than the incoming word. In this fashion, the peak amplitude is found and sent to the computer. Circuitry 1s provided to eliminale false peaks due to such factors as noise. This detector is enabled by the "cal" switch on the terminal front panel, as. smri i r ~ Figures 3 and 5.

Two analog output channels are available which can be used for various control functions. The buffered digital-to-analog converters (DAC) are loaded by the computer with the digital equivalent of the required output voltage. Once loaded, the output will hold this value until updated. At present, the primary function of these outputs is to plot the analysis results on a X-Y plotter or strip-chart recorder.

Fourier analysis of the returning ultrasonic signal can be performed by the computer by employing a sampling oscilloscope to sample the waveform. The sampling unit is scanned by the computer via the DAC outputs, and the analog value of each sampled point on the waveform is fed into the terminal via Channel 0 or 1 for digitization and transmission to the computer. Appendix A provides data and control terminal schematics and logic diagrams.

Computer lnatrerface

In order to obtain bidirectional communication between the datalcontrol terminal and the computer, an intermediary interface is needed. (Schematics and logic diagrams can be found in Appendix B.) The purpose of this interface, which is located at the computer, is to process tlags and data from h e terminal and to ba~ls~ l l i l commands to the terminal. (An explanation of the commands is given in Appendix C.) Figure 6 provides a view of the multiuser computer system to'which the ultrasonic datalmntrol terminal is connected. The interfaces to all external users are located in the second cabinet from the right. The data cable, which can be seen leaving the terminal in Figure 1, enters the computer interface cabinet through a port in the floor beneath the interface cabinet in Figure 6.

The interface must be turned "on line" by the computer before it can accept any flags from the terminal. A light on the terminal panel (Figure 5) indicates the state of the terminal, either "on line" or "local". An "off line" (local) command may be generated either by the computer (via the interface) or by a switch on the terminal panel. This switch also serves as an "emergency stop" since all functions terminate when in the "off line" or "local" condition.

When a flag is received from the terminal, the computer receives an API. The computer identifies the system originating the interrupt and proceeds t o determine which flag in the system is requesting service. When the flag is identified, a unique circuit in the interface automatically clears the flag. It does not, however, clear any other flag which might also be requesting service.

As previously mentioned, the test data are accepted and deposited via DMA in a particular location in memory without interrupting the computer. When a datum is ready, the terminal sends a special flag t o the interface which, i n turn, requests a data channel (DCH) transfer. The inputloutput (110) controller responds with a DCH grant signal. Next, the interface supplies the address of a word-count register which is incremented by the controller. Finally, the datum is deposited in a memory address which is the incremented contents of the current address register. This sequence is accomplished without interfering with the job that the computer is running. By stealing some processor time, acquisition of a datum delays the task in progress by only four microseconds.

SOFTWARE

System Configuration

Software for the ultrasonic nondestructive test system is designed t o operate on a medium-sized multiuser computer. The machine consists o f 16K words o f core storage, 262K words of disk storage, and four magnetic tape units o f 150K words each. The programs operate under the management of a real-time executive (RSX). This monitor allows several users to share the computer simultaneously. Programs ~pera t ing under this executive are chained and werlayed into logical units called "tasks". Each task is designed t o execute in a specific partition, o r particular area of core memory, and operate at one of 512 pr ior i ty levels The assignment of a priority to a task depends upon the relative importance o f the task. RSX is therefore a program-management executive rather than a true timeshare system. The tasks are stored on the disk unti l requested by the executive o r another program. Inter-task communication is implemented through user commons which are external t o the partitions. The ultrasonic software consists of 68 programs divided into 17 tasks. These programs use one partition o f 4K, one system common of 300 words, 7,200 words on the disk for data, one magnetic tape unit, and 4,096 words of auxiliary core.

The software must transmit information t o receive data from the ultrasonic interface. This communicmtion is done in an assembly-level language (MACRO) and is capable of controlling four axes of transducer movement, two digital-to-analog converters ( DAC), eight analog channels, and four data sync modes. The f low of execution must be controlled by the ultrasonic operator. This requirement is met through the interactive use of the Teletype and a high-level language (FORTRAN) which decodes the operator's mmmands and performs the necessary routines.

A l l data gathered f rom an ultrasonic test must be available for analysis at a later date. Therefore, data are stored on magnetic tape, and 35 unique file names are provided for each

tape. Six analysis routines are implemented in addition to Teletype and graphic display of a l l data. Extensive checks are made throughout execution, and the operator is informed of malfunctions. The programs are designed to facilitate modifications.

Software Descriptions and Logic Flow

The first and main task for the ultrasonic test system is initiated from the console Teletype a t the computer. Control of al l the programs is then passed to the remote Teletype (Block 1 of Figure 7) located a t the ultrasonic test tank. The program identifies itself and asks for one of six major modes of operation (Block 2). The first mode (setup, Block 3) allows the operator to either load the two DACs or move the four stepping motors (axes control). These options allow the operator to set the scale on the plotters and position the transducer.

The second major mode (Block 4) allows the operator to standardize the data. It collects the data necessary to convert a relative signal level to a flaw depth which allows comparison of data gathered a t different instrument settings. The computer stores the average signal level together with a slot depth and DB scale factor. This information is stored with all future test data. The next step in the conversion is a curve-fitting routine (Block 5) which fits these data to one of 11 curve types. The coefficients returned from ,this test are saved and kept for any future test.

To implement the actual data gathering from a test scan, the operator requests the test mode (Block 6). He can select one of five different tests:

, . 1. Rotating scan save. This test stores al l of the data from any number of scans taking a word of data every one-tenth degree from the rotating table. This option can use one or both lmmerscope channels. It can increment the four axes between runs with either an operator-supplied index or a contour following X-Y motor pulse file.

2. Rotating scan maximum. This option stores only the largest corresponding value for any number of scans. It has al l the options of "rotating scan save". This test is used to determine the largest flaw in a particular plane.

3. Flat-plate save. This test uses an independently controlled Y axis to gather data using a calculated distance between data points. The X, 2, and Aaxes are moved either from an operator-supplied index or a contour following X-Y motor pulse file. This type of test i s used to gather data from a flat surface.

4. Flat-plate. maximum. This test, saves only the largest corresponding values of Y for all values of X, 2, and A and is used to determine the largest flaws on a flat plate. The same motor-handling capabilities as in Test 3 are available. A composite data file is saved which will allow the operator to quickly determine if a part is within tolerance limits.

5. RF wave form. This test collects data from a waveform displayed. on a sampling oscilloscope. The waveform is characterized by sampling a t X positions which are set by a DAC and are determined by an operator-defined number of points. This option is used to analyze the frequency of an ultrasonic signal rather than its relative amplitude. The information can be used to determine the response of a transducer or the absorption spectrum of a material.

I I =P @ STANDARDS @ ALLRT @ DECTAPE FILES @

I

I 1 Collect Data from Fit Co l i b ra t i ~n Doto File Output. MOVE TRANSDJCER PLOT FROM Standards. to Choice of Curves I

TTY and Sove Ceff ic ients. I I 1

Index X, Y, i or STRIP CHART GRAW W- A Axes from TTY. P o t on P .at on Plot on X - Y List File

X - Y Record,?r. S-r:p-Ctm-3 Recoder. Recorder. on TTY. -

I 1 ULTRASONIC TESTS @]

I r Collect Data from Fourier Transform or Smooth Data Search File for Rotating Turntable. Rrtating Turntable. Log'Fourier Transform. over Ciosen Data Exceeding

Save A l l Data. Szve Maximum Vc,lues. qurnber of Points. Threshold.

Collect Data flom Collect Data from

Flat Plate. S a ~ e Flat Plate. Save Max-mum Values.

Calibration Curve

Fiwre 7. SOFTWARE PROGRAM. (Circl'ed Numbers are the Major Bbcks in lhe S.ym~m)

All of the tests store the data with a header which contains the type of test, . . .

@ t'l 1@7 2 . 2 5 i 9 . d Z 65DEi d20, 2 L & P T = S U S L C number of run, number or words per run, a @, ,,,, .O P B B Vc)HLS,RLid calibration information, and 72 characters of identification which are . -

(IJ, File name supplied the Operator (Figure 8). The @ Identification. Up to 72 characters may be inputted by storage of this header information insures the operotor.

that all pertinent information is available @ Test type. A " - 1 " represents data gathered from a wave form.

a t a later date. The identification is used , Number of runs, Inputted by the only to avoid .confusion between data 8 5 Number of words per run. Supplied by the operator:

files. '

Figure 8. SAMPLE HEADER.

The data analysis routines (Block 7 of Figure 7) give the operator mathematical capabilities in handling the data. He has the options to:

1. Smooth the data by averayil.ty. A variable number 07 points, forming a "window", are averaged over the full data set contained therein. (The process is actually an integration of the area.) This technique removes noise from the signal (see Figure 9).

2. Perform a Fourier analysis on the data. The frequency, amplitude, and phase which are results of the transform can be written on the Teletype or saved on magnetic tape. Figure 10 is a plot showing a sample of the results obtainable from this test.

3. Search the data file for any values above a selected threshold. The initial and final location of a series of values above this threshold as well as the maximum value within this series i s written on the remote Teletype. This procedure allows the operator to relate location, length, and depth of a flaw. This option allows a three-dimensional definition of a flaw if used with several consecutive scans.

4. Convert the signal level to depth in mils using the calibration curve. The coefficients for the curve can come from either the values stored with the test or from the most recent coefficient filc.

5. Compute and store 1,000 points on the calibration curve. Coefficients for this curve can come from either the values stored in a test file or the most recent coefficient file. The output is saved on tape. The purpose of this function is to provide a graphic display of the calibration curve (Figure 11 ).

All of the options, except threshold and possibly the Fourier analysis routines, store the computed data in a scratch file which can be permanently scaled for graphic display. This file i s handled in the same manner as a normal data file.

File output (Block 8) allows the operator to graph any data file on the X-Y plotter or strip-chart recorder. These data can be scaled to magnify an area of interest. The data can also be printed on the remote Teletype. The heading for the data file is printed when any of the five options are used. The headings for all the data files can be listed as a tape directory for the operator. Also,'any data file can be deleted.

(a) Or ig inal Data. (3, 600 points - rotation scan graphed on strip-chart rec6i'del)

(b) One-Degree Integration of the Original Data. (window width of 10 pts)

\

(c) Five-Degree Integration. (window width of 50 pts) ., " - . .-.- - -

.. ....-- . _ _ - ------ -------

A

- ,. r \ I \

v -7 - -..- U w U

(d) Ten-Degree Integration. (window width of 100 pts)

Figure 9. EFFECT OF INTEGRATION O N THE BASIC ULTRASONIC DATA.

Time (pet) (a) Plot of a Woveform from o Sample Using a 10 MHz Transducer. (sampling interval of 0.01 pet)

Frequency (MHz)

(b) Fourier Transform of the Original Waveform.

80

.- - E

-x

20 -

U 5 10 15 20 25 Frequency (MHz)

(c) Lng of the Fourier Transform of the Original Data.

Figure 10. FOURlER ANALYSIS OF THE ULTRASONIC DATA.

Signal Level (volts)

Figure 11. r;AbIBRATION CURVE OF SIX-PUINIS FIT TO A SECOND-ORDER POLYNOMIAL. (Standard b~uiation = 0.237)

All six modes of operation return to the main task upon completion. These modes give the ultrasonic operator a basic operating system which can be expanded as new options are required. A complete set of software flow charts can be found in Appendix O.

Operation

Certain conventions have been followed throughout the program. Major modes identify themselves by typing the mode name followed by three asterisks and a request for an input. All commands are single-letter inputs which, in some cases, are followed by a file name and an optional ruri number (an unrecognizable input will cause the request to repeat). In the case of a numeric input, a l l values are integers (no decimal point), except when a decimal point occurs in the request. An example of the latter is the standards routine request ".LEVELu. Error statements are preceded by three asterisks and, if severe, will cause the

program to restart. A severe error is followed by an integer number. If this number is negative, the error probably cannot be corrected a t the remote device; and, if this error repeats, the computer system operator should be notified. An example of this type of error is 'I*** DISK ERROR IEV=-12".

The emergency stop for this system is the "off-line" switch. I t stops everything under computer control. The switch isolates the interface from the computer; and, if the system is in the process of taking data or driving the motors, the switch terminates the program. Another terminal condition is encountered when turning the power on while the system is "on line".

In order to start the ultrasonic series of programs, the ultrasonic operator requests the computer operator to initiate his program. When started, the remote ultrasonic Teletype responds "ULTRASONICS***" and asks for the first option. The flow of logic, as the operator sees it on his remote Teletype, is given in Appendix E.

The multiuser computer allows efficient use of computer time and supplies more computing power than a minicomputer. It also gives the user the impression of being a dedicated computer. The ultrasonic test system has been operational for six months and has supplied good and fast results.

The datalcontrol terminal and associated hardware were designed to have a high degree of flexibility. A very minimum amount of work should be required to accommodate any new test methods. The major effort would be in interfacing signals to the terminal and in modifying the existing software to accommodate the new test.

Moy er, M. W. and Gray, D. H .; Expanding the Capability of a Laboratory Ultrasonic Testing Facility; to be published in Materials Evaluation, Journal for the American Society for Nondestructive Testing.

APPENDIX A

DATA AND CONTROL TERMINAL SCHEMATICS AND LOGIC DIAGRAMS

' I

JZ- 33

-

~ 5 \ ~ 1 ~ 4 L I ' * u ( -, %::%.E

k ~ k s : %LOCK 1

I- INPUT WOLTaCIE. TO S K MUST P& S E T W E E U 0 S - I O W .

Figure A-1. ULTRASONIC TEST SYSTEM. (Analog Data Control and Logic)

LCD w w =

Figure A-2. ULTRASONIC TEST SYSTEM. (Sync Control - Data Enabling Circuit)

, LEO

LED

Figure A-3. ULTRASONIC TEST SYSTEM. (Miscellaneous Logic) I

I KI

2 PEU OO\VN L

?

>

- 4 u n ~ o c - 5 c A L E E

Figure A 4 . ULTRASONIC TEST SYSTEM. (Miscellaneous Schematics)

L C2ILI

Figure A-5. UkTRASONIC TEST'SYSTEM. (Peak Deiector iogic Diagram)

TPL%CWM1-B DISPLYI'

841 <-> 7m

el -z Y 82

sl-83

TI-W

Bl" - > 26

0.1, - >as

-5k- ---+ 35

el -55

$1 c-> 54

Figwe A-6. ULTRASONIC TEST SYSTEM. (Analog Output and Displays)

Figure A-7. ULTRASONIC TEST SYSTEM. (Axes Motor Control Logic)

N 01~s: I- 5 E MODULE 0 0 PEP TUG D\VL1. IJUU%EPS FOE O N ALL M 236 LNPU &?iE!Z.

3 0 ~ U ~ d ~ ? & 57 u 908 ' y ' ~ * 1 6

I 3T0 , - 1 A 3

r a' -02

14bF YO 5 0 - I I ~TC. 6 n s T l l K A c o u w 5

I Tb

a" - 2 TO

Figure A-8. ULTRASONIC TEST SYSTEM. (Auxiliary Resistor Housiw Logic) I

78 -52 - SPAPE

,T2 -

@ I A u k L o a OUrPUls < 45

C 35

Figure A-9. ULTRASONIC TEST SYSTEM. (Miscellaneous Wiring Auxiliary Housing)

I

M O D U L E UT I 1-12 AT 10 M D k ~ k / C 3 NT POL T.EP-MI M

. -- M O D U L E ~ J T ~ L I Z ~ T I O N AUW U 0 u S l ~ c - l

TOP V I E\V L T A a

Fiere A-10. ULTRASONIC TEST SYSTEM; (Module Utilization Layout) f

I

QUAD 4

1 u PUT S

Figure A-11. NDT INSTRUMENTATION AND SYSTEM. (Quadrature Decoder Card Schematic) - (I

Figure A-12. NDT INSTRUMENTATION AND SYSTEM. (Quadrature Decoder Board)

b z 1 ~ 2 1 92% lAl BZl SCAN COUPL

SZlJZ '

TEST l-l BIZEZ '

SChN COMPL Fsau2

APPENDIX B

COMPUTER INTERFACE SCHEMATICS AND LOGIC DIAGRAMS

I I - p,22Ll, sa I 1-2

BZlFJZ I

Figure 8-1. ULTRASONIC TEST SYSTEM. (eul~~putsr Interface Daviao Soloaterc) I

I

A22 5 2

A 3 0 L I

A3oPI

0 N L I N € COMMAND DEcooEE

Fipure 6-2. ULf RASONlC TEST SYSTEM. (Computer lntet'fam - Flag Processing)

DCM oe H R W T 2

V H Z ~ F I

A3051 110 P w E CLP U 31 - W P W Q CLP U u I

1 376 K Z

m.97, I . I

Figure 8-3. ULTRASONIC TEST.SYSTEM. (Computer Interface - Flag Control)

, API 7. EU I r l U . U 2 U 2 APT 2 BU OUT U h32~Z

1 $2-

I

.Figure 8-4. ULTRASONIC TEST ~ y ~ ~ ~ ~ . ' ( C o m ~ u t e r Interface - Commands processing)

'A'.

if Figure B-5. ULTRASONIC TEST SYSTEM. (Module Utilization) 1

APPENDIX C

TERMINAL COMMANDS

Commands to the data set are delivered via a coded 18-bit word which is loaded into the computer accumulator and transmitted by performing a system write inputloutput transfer (IOT) instruction. The system IOT instructions are listed in Table C-1 and the system commands are listed in Table C-2. The first six bits of the command word designate the particular command and the latter twelve bits carry data or subcommands associated with .the command. For example, the rotating table encoder is selected as the data sync mode with the octal code 4200028. Each command is defined in Table C-3.

Table C-3

Table C-2 ULTRASONIC DATA SET COMMAND DEFINITIONS

X, Y, 2, and 0 Axis Load Allows the number of steps which the respective axis is to take to be loaded.

SYSTEM COMMANDS

-- -

~ e i i ~ n a t i o n Code i

Command (octal) Motor Start Starts the motor designated by the enabled bits.

r?/A - 1 D/A - 2 X Axis Load Y Axis Load Z Axls Luad 0 Axis Load Data Point Definition Data Scan Definition Start Data Taking On Line Off Line Recorder Pen Control Data Sync Mode

M X X X X 02xxxx 14XXXX 34XXXX W X X X X 54XXXX 52XXXX 62XXXX 320001 740000 640000 720000 420022

ZZ = 02 Rotating Table Encoder = 04 External Pulse Source

Determines thc dircction of rotation of a motor as designated by the enabled bits. Any time this command is used, the bits previously enabled must remain enabled to avoid changing the direction of a motor.

Motor U~rection

Table C-l

ULTRASONIC TEST SYSTEM INPUTIOUTPUT TRANSFER INS'TRUCTIONS

Motor stat't artd directiort wrnmands may be wmblned into one 18-bit word.

AID Select Allows the computer to select any one of eight channels for AID conversion. AID select has the lowest priority under an internal prlorlty scheme. Channels 0 and 1 are selected automatically.

Code Function (octal)

System Test System Read System Write

Start Data Taking

On Line

Enables the data set to begin taking data on the next initiate pulse.

Enables flags so that the data set may communicate with the computer. Also enables the local clock for stepping motors, etc. The data set switch must be in the "on line" position before this command will take effect.

= 10 Internal Clock = 20 Flat Plate

12000z Z = 0 Multiplexer Channel 1

= 1 iviiiltlpleker Cnannel 2,

Scan Complete Test 706021 Transducer Arrived Test 706041 Data Channel Overflow Test 706061 Data Channel Overflow Clear 706062

AID Select

Off (local Line emergency stoa) Disablcs the data sat by switching i t the l o u l irsode. T l~ is l u ~ l c l i u ~ ~ can also be accnmplished hy planing thn cia?*-set switch in the "local" pOslriOR which als6 serves as an emergency stop.

Motor Start 24zzzz ZZZZ = 4000 X Axis Motor

= 1000 Y Axis Motor = 0200 Z Axis Motor = 0040 8 Axis Motor

Data Sync Mode Select Specifies for thc data sct the source of pulses that define each point at which to take data.

Dam Point Def in i t io~ Defines the numher of steps hetween data p i n t s when 11sin0 cync ~IIIFR oeureoo other thon tho rotating tablo onaodar (ag, thi3 action will allow data taking a1 cve~.y nllie~. ptllsc! if I~la~lnt l will1 2). MIII;! 11.et.i t.wl~~re "data scan delinition".

'LZLLZZ Purward ZZZZ - 2000,

= 0400, = 0100. - 0020,

Reverse 0000

Moror Direcrion X Axis Motor Y Axis Motor Z Axis Motor 0 Axis Motor Defines the nttrnber of riat* p in t s comprising a scan when using pulse

sources other than the table encoder. Also defines the number of peaks to be found per standard during calibration. (A scan is analogous to one rrrvnl~rtini~ nf the trrhlr! - nhe <el i lf u'lsln )

Data Scan Definition

(1) X = data; Z = subcommands.

APPENDIX D

SOFTWARE FLOW CHARTS

Task ULT 1 1 . Decodes the Major Modes of Operation

and Requests Appropriate Subroutine 2. Move the'* Y, Z, A Axes 3. Plot from the TTY

I L Task ULT 6

Decodes Test Type, Gets Test Parameters and Identification, Finds

Empty Data File, Handles Errors

Tasks ALF I, ALF'2, ,ALF3 Decodes Math 'Function Decodes and Executes Gathers Data from General Purpoie

Type, and Checks for Dectape 1/0 Functions, S tandards, Computes Least Sauares Curve 1 p e r 1 1 Returnt0,uLTI 1 1 A v e r a g e ~ t o s 1 1 F i t ~ o u A , n e ( ~ l i ~ 1 When Finished Data on Tape chart not shown)

. Performs Proper Function

Does a l l Graphing for Performs Test, Saves Data

I on Tape with Header I Scales and Stores Data

Return to t l

,. -....-....I. .. . . . ._.,

l ~ e t u r n to ULT I ] I

I

I Program MOVE Program TALK Program PLOT Task ULTl Task ULTl (main task)

Yes Cal l MOVE \ Task ULTl Dimension J(250), A 1(20)

I I (subroutine)

~ommon/Coml/J, A l t ( c a l l ULTSET) I

Yes Cal l PLOT Turn Ultrasonic (subroutine)

Device "On Line" Ask for X, Y, 2, A I Pulses from TTY

Exit Tnsk, f a l l Iv\OTOR Sewlee arher Tasks

Calculate Total Distance

Pen must Wait for a Travel Significant

Event

Number of l ncremen ts = Distancc/lO

I Size o f lncrements = Length/Number

o f Increments

EJu

Cal l DTOA (X, Y, Doley (0))

I - Woi t 3/60 Seconds for Graph Response

Time

I

Progrom MOTOR Used in Tasks ULTl and ULT2

Call Stac

( X )

Stoc

Get Argument

Set Direction

Return

Call Stoc

Call Stoc

(A)

Program DTOA Used in Tasks ULTl and ULT6

Get Arguments DTOA 1 DTOA 2

Load DAC 1

Load DAC 2

Send "Go" TI Return 2,

Program TTL (contains subroutines TTL, UPDWN, ULTON, ULTOF, ULTSET and the interrupt handling routine) Used i n A l l Tasks

API Interrupt from Ultrasonics

f Save A l l Registers,

Clear Accumulator

Act ive Light

UPDWN

Se'nd Pen

Return

U LTO N 4 - Turn System

Act ive Light On

-..- .

Turn Ultrasonics

O n Line

I Clear lnterrupt

I ~ e r o .Both D A C ' S ~

1-4 ~ e t u r n ] 'iet Event

ULTOF* I Variable- Turn System Act ive Light

O f f

I Turn

Ultrasonic O f f Line

l nterrupted Tos k

Disconnect APi I.-- 4 Return 1

,,., ...

U LTS ET

Clear Flag and

Progrom ERR Task ULT6

Number.of X Mils N o

- Number'of Y Mils Number of 'Z Mi ls Number of A Steps

Total Y Delta Y

Ask for One Line of Identification

i Find on Empty

Data File

Ask i f a M o t o r '

Pulse File i s Wanted . '

Read and Save Motor Pulses tor t

X; Y, Z, A Axes

I Write File Name, Number of Points

Request Main Task (ULTI)

Request ULT2 (test routine)

'

Exit

. .

Program TEST Task ULT2

I Connect Interrupt W ~ i i e Heuder

Line (call ULTSET) Information

on File

1 Call GETDAT

(test arguments) (ULT6)

Allocate 7210 Words of Disk

Storage

Gct 125 Words oi Data

from Disk

Get Test Arguments from Common Area

Upen FI le whose I\lamc

was Passed as an Argument

Yes Exit to t

Routine

(ULT6)

I 'I

Program GETDAT

Used i n Task ULT2

Get Test ! 1 I Setup Disk and

DMA, Select

'1 External Sync '1

Get One Word I

from Each Run I I

1 Setup Step and (from dirk) 4 Scan Defn

! Scan Defn =

L 16 (number of

words/run) Enable Data Taking

Get Pulse for this Largest Data

Run from Aux Core Word (on disk)

Select Table

Overflow lnterrupt

Setup Two Internal Wait for Mo to~s Buffers (250 words each)

I Save this Value

1 \ Setup for

Setup DMA Next Run

4,

lncrement Word Counter and DACl

l ~ n a b l e Data ~ a k i n ~ l

Wait for an Interrupt

I I

Swap Buffers on DMA Write Full I-

Buffet on Dlsk, Count the Number of Word:,

2

I\ i

Wrile oily Re~nuir~iny

Words on the Disk

r P

lncrement Run Counter

I -

Programs CORE, CORE], ICOR, COR Used i n Tasks ULTZ, ULT3, ULT6, ULT7, ULT9

CORE ond CORE 1

Gct Argurnent3: (operation, data (integer array),

number of wol-ds, address, skip code)

I Send Address tn AIIX Cnre

I

Marker Bits

I Operation Been ~ e r f o r m c b N' \ \?lie Reuui~eJ P4u111lre1 ul-, 2-

ICOR COR

(address data word

(read/write)

Re turn 23

Oei Argumenrs (address, data word ( ~ e u l ) upe~ul iun) 1

t Send Address to Aux Core

I

Perform Some . Operation

(second word) -.- ..- ti 1

Return A

Program M A T H A Used in Task ULT3

Program MATH 2 Task ULT8

Exit Tusk I

of Integration "Window!'

Threshold

1 , Fourier Transform

Ask for Number of Data Points, Total Length,

Number of Output Points, Offset, Time, 'Code ( 1 =

write, 2 = save)

Ca l l THRFSH

/+\ e c t i ~ ~ - Fourier ~ r a n s f o r m w Col l FOUR I \ T /-

1

I Set Code = 1 1 (nsstrm~! option =

standard)

I Call COLST (code) I

Program CURVE Used i n Task ULT8

7200 Words of Disk Storage 1

Exit to Error Alldcation Successful? Routine

(ULT6)

Average Data over "Window" Length

Shift Computed Data by ("window1'/2) Number of Points

I

Store Computed Data in Aux Core

I Exit Tosk

Program THRESH Used in Tusk ULT8

Value 2 THRESH?

9 this the First Value? Yes , 4 7 1 Location

Write Beginning Valuc, - End Value, and

Program FOUR Used i n Task ULT8 I

r

Exit to Task ULTl

(cal l COR)

Between 0 - 1,000 I /i

Store Amplitude in Au)( Car0 1

ii (cal l COR) j

I .

Core Aux (cal l COR)

1 Maximum Value ]

Request Tas

Exit Task I

'1

Program COLST Used in Task ULT8

( I = equation)

Decode Equation, Write Equation, Ask

i f this i s the Proper Equation

I Read ~ n s w e r 1

I No Use X = 1 to 1,000

. --

Use X = Data in Equation

call ICOR) ?- I Exit Task I

Program MATH 3 ask ULT9

I [Set Scrotch = " lnt" I Comment:

Fourier Data

are not

to be Seoled

Generate

Heoder

Determine Maximum Scratch = Output

and Minimum of File Noms C

Computed Data

I

Open File

Scrotch

I Write Header

I Write Data

I Close File

Scrotch

Write "Scale Data"? . (-1 = No, D = Default,

1 = Input Maximum, Minimum) Request Main Tosk

(ULTI)

Scole Data

Program HANDLE Task ULT4

I

Reqcest Main Task

Exit Task

Number from Tope, Store in Aux Core (call CORE)

Yes

Write List of Commands on TTY

Yes -C

Set Run Number = 1

Request ULTi (graph)

Exit Task

Program GRAPH

Task ULT7

v I Send (0,O) to DAC's 1

Write "Pause" Read TTY

Send (1,000, 1,000) to DAC's Write "Pause"

Write "Scale Data? (Y, N)" - Read Answer +

7 Y Beg, Y End I

Send Data to DAC 1 Starting a t X Beg, Ending at X End.Using Scale

- Factor ( - 6 msec delay)

tpnrl Inrr~mentc tn DAC 1 Starting at X Beg, Ending

at X End, Scaled to go Full Scale. Send Data to

DAC 2 Using Y Scale Factor to Calculate Y (2 50 msec delay)

Write "Pouse"

Read TTY

Send Pen Up Command

I Rcqucst Task

Exit Tpsk

Program STAN Task ULT5

I L

4 4

! I ,

Wri tQ Program Is(Qnt i f icat i~n I'

Selup III~~IIUIJI Calculdre Average Determine 5mallest I Write "Standard Number?" I / I Channel o f a l l Data Points Level Number i n

I I I Stnndnrd Tnhle

1 ~ e a d Standard umber] ' Setup Ultrasonic Remove from Data Interface to Send those Points that

100 Points on Vary by 1 1 OOh of Store a l l Data at lntcrnal Sync the Average 1 Yes , this Level Number,

("noise") End Dotn with

Marker Word

Received Flag

Remove this Dato '

Write Standard Table from Standard Table

Standard Number -.-.. - , - - - ---.-

, APPENDIX E

OPERATOR INSTRUCTIONS

N o

"Math Functions***

Option, Fi le Name; Run Number

Name (run number)" Name (run number)"

"Threshold Value?" Read Input, Scan Data

Write "Number of Points?" "Beg =" [I61 "End =" [I61 "Maximum =" 161 "Input Number of Data Points, Total Number o f Points,

Output Number o f Points, Offset, Time, Code (1 = write, 2 = save) (one per line)" Read Dota D6]

Write "Fi le Nume? (FTI, FT2, ALl, AL2)"

Yes

I

Write "I Integrate T Threshold E F i t Dntn t n Fqn F Fourier Transform S Setup Calculated Curve C Commands

1 R Return" I

Perform A

I I

7

Scan Computed Data ?Scale Data" (-1 = No, 0 = Default, Write 1 = Input Max, Min)" "Max =I1 061 "M in ='I 061 j/ Read lnput

Name 1 = COEF "Data File Name?"

from Either Named 'vV111u Dola an

t ~ l e of C U t t bi le Tape i n File d INT

b.7 Read Data from t

Data File I I

i Read Coetticients 1 - from Either Data

File of COEF File

Decode Coefficients i

' Write Equation,

"Operate with this i

Eqn (Y, N ) " I I

Read lnput

N o

I N o - N o - I G o to @I

Write Write Write "Standards***" "Test*** plot*+*^*

I Test Type? (M, S, F, L, W)" X, Y?

Write Write

"Standard Number? " "Number of Runs Number of Words/Run Read lnput E5]

Number of X Mi ls F.lv~iibci of Y Mil3

v "Standard Number Depth Level Avera e Noise"

E7.4 P 7 . 4 ~ 8 . $ 88.33

N o

... "Number o f Points?"

Write ,

".Levelu

I Write Write "ldent?" "k i le" PC] D6-J "G~uto PoIntsU Read Input Ready'?" (Y, N )

I P 4 ( ~ 5 ) l

-. . Eroco a!! nnto -b%b 1. - for this Points, Average, I

Find Empty Fi lc Read lrtpul Standard Number Insert Data i n Tnbln -

1 W1.i te "Mntnr PIIIFP Fi le? (Y, N)"

4 \ t

1

I Read Input I Write

' Perform Test i f Test Type = F,

"Init iate" Switch must be 'Pushed nftsr the "Enc" Light Comes on

No :-

-

Write Header and Data on Tape

Go to @ - Unrecognizable Input, Restart Decode Sequence.

Write "Dectape Files*** Opt ion?"

I Between Paints I

Wrlte Write "C Commands - ( G O ~ O @ I L List of Part Filcs D File Name R Return D Delete W Write

Read Data from Tape ' 1

Write Header Write Data 1 . "Fi le" [AS] "ldent"

[14 (As1 "Test Type 151 "Runs" D5] "words/RunU

on TTY

(161

I I Ib0 to 01

Write

"Scale? (Y. N)" Read Input

1 I i

; 1 I II

I !

i !

i

yes

Send Data to X-Y I

t Plotter wlth 1/20

i Sec Delay 1 c

- Send Data to

Strip Chart Recorder, 6 msec Delay Between Data Points

1 I 4

DISTRIBUTION

Atomic Energy Commission - ALO

Vespe, V. C.

Atomic Energy Commission - OR0

Hickman, H. D. Zachry, D. S., Jr

Goodyear Atomic Corporation - Portsmouth

Ostroski, A. S.

Lawrence Liveriiiore Laboratorij

L~pta~, H. ti. McFarland, G. C. Technical Information Division

Los Alamos Scientific Laboratory

Baker, R. D. Briscoe, W. L. Elliott, D. E. .

Oak Ridge Gaseous Diffusion Plant

Barton, J. C. Blake, H. W. McLaren, R. A. Wilcox, W. J., Jr Winke1,R.A.

Onk Ridge Nntinnnl' Lnhrn?ary

Adams, R. K. Borkowski, C. J. McClung, R. W.

Oak Ridge Y-12 Plant

Alvey, H. E. Bernander, N. K. Briscoe, 0. W. Burditt, R. B. Burkhart, L. E. '

Darby, D. M. Davenport, C. M. (2) Denny, A. (2) Dodson, W. H. Ellingson, R. D. Foulk, D. L. Gray, D. H. (10) Jackson, V. C. Jones, F. W. Kahl, K. G. Keith, Alvin Kite, H. T. Lundin, M, I. Marrow, ti. l3. McLendon, J. D. Masorr, D. L. Miskell, R. V. Mitchel, G. W. Moyer, M. W. (2) Noey, J. L. Oliphant, G. W. Schneider, P. G. Smith, J. H. Smith, R. D. Stephens, A. E. (2) . Stoner, H. H. Weathersby, W. E. Whitson, W. K. Wright, C. C. Yaggi, W. J. Y-12 Central Files (5) Y-12 Central Files (master copy) Y-17 Central Files (rn~rte) V-12 Central riles (Y-12RC)

Paducah Gaseous Diffusion Piant

Levin, R. W.

Sandia-Al buquerque

Ballard, D. W. Mail Service Sectinn

Sandia-Livermore

Baker, A. E.. Library Murphey, B. F.

In addition, this report is distributed in accordance with the category UC-37, Instruments, as given in the USAEC Standard Distribution Lists for Unclassified Scientific arrd Technical Reports, TI D-4500.