Historical Document · 1961
The Operations Research Office's SALVO II report (received 31 October 1960, published May 1961). The Hyper-Burst Mechanisms dive cites it for the second field experiment's results on multiple-projectile and burst weapons.
UNCLASSIFIED AD NUMBER AD325385 CLASSIFICATION CHANGES TO: FROM: UNCLASSIFIED CONFIDENTIAL LIMITATION CHANGES TO: Approved for public release; distribution is unlimited. Document partially illegible. FROM: Distribution authorized to U.S. Gov't. agencies and their contractors; Administrative/Operational Use; MAY 1961. Other requests shall be referred to Army Materiel Systems Analysis Activity, Aberdeen Proving Ground, MD 21005. Document partially illegible. AUTHORITY usamsaa ltr, 23 mar 73; usamsaa ltr, 23 mar 73 THIS PAGE IS UNCLASSIFIED ^-~:-:-:v:r:": ■ •" z£Mg*0g3E&sM wsm: SR?S=:' S>S: 325 385 AD KepAoaaced by (Ue ARMED SERVICES TECHNICAL INFORMATION AGENCY ARLINGTON HALL STATION ARLINGTON 12, VIRGINIA l! US WS.. m w-SSpi Jf URS: mk 1 NOTICE: When government or other drawings, specifications or other data are used for any purpose other than in connection vith a definitely related government procurement operation, the U. S. Government therehy incurs no responsibility, nor any obligation whatsoever; and the fact that the Government may have formulated, furnished, or in any way supplied the said drawings, specifications, or other data Is not to be regarded by implication or otherwise as in any manner licensing the holder or any other person or corporation, or conveying any rights or permission to manufacture, use or sell any patented invention that may In any way be related thereto. .. ... THIS DOCUMENT IS BEST QUALITY AVAILABLE. THE COPY FURNISHED TO DTIC CONTAINED A SIGNIFICANT NUMBER OF PAGES WHICH DO NOT REPRODUCE LEGIBLYo SALVO II OPERATIONS RESEARCH v • Rifle Field Experiment (U) I The Johns Hfapkins University o c :/~) ■f;i Operating ! fontrc.it with rt< ! iRMY 7NGRA0EB AI 3 YEAR JNTOfl^fc LASSIBEO ÄFTES 2Si VJ^VBf: n ■> DIR 5200.10 TACTICS DIVISION INFANTRY GROUP Technical Memorandum ORO-T-397 Published May 1961 SALVO II Rifle Field Experiment (U) by J. David Reed Leon Feldman Richard E. Tiller Duncan Love Kenneth L. Yudowitch William Pettijohn Paul Michelsen William W. Walton Affk a£ JIAJt i OPERATIONS RESEARCH OFFICE The Johns Hopkins University Bethesda, Maryland Received (or Publication 31 October 1960 Published May 1961 by The Johns Hopkins University OPERATIONS RESEARCH OFFICE 6935 Arlington Road Bethesda, Md. CONFIDENTIAL ACKNOWLEDGMENTS The field team that conducted the SALVO II experiment was directed by Kenneth L. Yudowitch a'xlwas composed of the following ORO personnel: Leon Feldman, Thomas Hawthorne, Paul Michelsen, William Nalley, William C.Pettijohn,Robert R.Redick, David Reed, Robert Shook, and Richard E. Tiller. The Project Officer at Ft Benning, Ga., was Capt Irvin R. Hirsch, then with the Infantry School. The authors would like to express their thanks to the Infantry School for the excellent support and cooperation extended during the SALVO II experiment. Ordnance members nf the SALVO II team were Robert Hoar of Springfield Armory, and Charles Dickey and Charles Gindhart of Frankford Arsenal. Another member of the team was Art Burns of the Winchester Western Division of the Olin Mathieson Chemical Corp. The authors would like to express their thanks to the Mathematics Division of the Naval Ordnance Laboratory for the use of their telecomputing facility. They are also indebted to Development and Proof Services, Aberdeen Proving Ground; and to Springfield Armory for postexperiment tests of the rifles and ammunitions. The Infantry Board furnished unpublished data that were useful in the analysis of the team's results. The detailed numerical analysis at ORO was done almost entirely by Betty Foster. The multiple correlation routine for the 1103A Remington Rand computer used in the analysis wasprepared by Lawrence Reimer and Howard Roberts of ORO. The authors also would like to thank William C. Suhler of ORO for the contribution of the confidence limits in Tables 5 to 7 in the body of the paper and the description of them in App D. Paul Michelsen and William Walton contributed the material in App E in addition to their services in building and designing the instrumentation. m CONFIDENTIAL CONFIDENTIAL CONTENTS ACKNOWLEDGMENTS iii SUMMARY 1 PROBLEM—FACTS—DISCUSSION— CONCLUSIONS—RECOMMENDATIONS INTRODUCTION 7 BACKGROUND 7 WEAPONS AND AMMUNITION 8 EXPERIMENTAL CONTEXT 9 EXPERIMENT SUBJECTS 14 EXPERIMENTAL DESIGN 14 TYPICAL RUN OR FIRING SEQUENCE 14 DATA COLLECTION 16 INSTRUMENTATION 16 DATA REDUCTION DATA ANALYSIS 18 RESULTS 19 AMMUNITION DIFFERENCES—AIMING ERROR—EXAMINATION OF MAJOR AMMUNITION DIFFERENCES—FIRER ACCURACY AND GAIN FROM SALVO AMMUNITIONS—AMMUNITION PRECISION—RATE OF FIRE-TRAINFIRE, ARMY GENERAL CLASSIFICATION TEST(AGCT) SCORES, AND ACCURACY ON SALVO TARGET SYSTEM—LEARNING—TARGET-SYSTEM EFFECTS DISCUSSION OF RESULTS 34 APPENDIXES A. PERSONNEL B. WEAPONS -\NU \MMI NITION C. DATA REDUCTION P. DATA ANALYSIS [■:. SALVO II INSTRUMENTATION 39 45 59 93 113 CONFIDENTIAL CONFIDENTIAL REFERENCES 123 FIGURES 1. SALVO II TEST AMMUNITION 2. SALVO !I TARGET SYSTEM 3. LOCATION OF EQUIPMENT AND REPRESENTATIVE TAUGET POSITIONS IN SALVO II DURING SERVICING OF TARGET SYSTEM 4. SALVO II FIRING LINE SHOWING MODIFIED PRONE POSITION 5-7. CASUALTY GAIN AS A FUNCTION OF INDIVIDUAL AIMING ERROR FOR 8 10 11 13 THREE AMMUNITIONS 5. .30-cAL DUPLEX 6. .22-CAL DUPLEX 7. .30-CAL TRIPLEX 26 26 26 8. DUPLEX AND TRIPLEX FIRST-BULLET HITS, EXCLUDING RICOCHETS, RELATIVE TO .30-CAL SIMPLEX AMMUNITION TABLES 1. TARGET-EXPOSURE TIMES: SALVO I vs SALVO II 2. WEATHER CONDITIONS DURING SALVO II 3. TROOP QUALIFICATIONS AND DIVISION 4. WEAPON-AMMUNITION COMBINATIONS 5. SALVO II FIRING SCHEDULE 6. TOTAL SHOTS, HITS, CASUALTIES, AND HIT AND CASUALTY PROBABILITIES PER TRIGGER PULL WITH SALVO II AMMUNITION 7. PERCENTAGE OF CASUALTY AND HIT GAINS WITH .30-CAL DUPLEX AND TRIPLEX AND .22-CAL DUPLEX AMMUNITION OVER .30-CAL SIMPLEX AMMUNITION 8. TARGETS HIT PER TRIGGER PULL FOR SALVO II AMMUNITION 9. PERCENTAGE OF RICOCHET HITS FOR SALVO Ii AMMUNITION 10. EXPERIMENTAL AND PREDICTED HIT PROBABILITIES FOR DUPLEX AMMUNITION 11. CORRELATION BETWEEN AIMING ERROR AND CASUALTY GAIN FROM SALVO 28 12 12 15 15 15 20 21 21 23 23 25 AMMUNITIONS 12. CASUALTY GAIN ON MOVING TARGETS VS ALL TARGETS FOR .30- AND .22-CAL DUPLEX AND .30-CAL TRIPLEX AMMUNITION OVER .30-CAL SIMPLEX AMMUNITION 13. PRECISION CHARACTERISTICS OF SALVO II AMMUNITIONS 14. RICOCHET HITS AND FIRST-BULLET HIT PROBABILITY 15. CORRELATION OF TRAINFIRE VS AGCT AIMING ERROR 16. SOURCES OF VARIATION IN SHOTS FIRED AND IN SIMPLEX AND FIRST-BULLET 25 27 28 31 33 HIT PROBABILITIES PER TARGET 17. SUMMARY OF DUPLEX AMMUNITION RESULTS: SALVO I AND SALVO II AND 35 INFANTRY BOARD EXPERIMENTS VI CONFIDENTIAL CONFIDENTIAL SUMMARY PROBLEM To measure the relative hit probabilities of several multiple- and single-bullet nfle ammunitions in combat-simulated aimed rifle fire, to use them to compare ammunition effectiveness, and to examine the effects of various parameters, such as range, target-exposure time, TRAINFIRE qualification scores, etc., on the accuracy of aimed rifle fire. FACTS The SALVO II field experiment was conducted byOROat the request of the SALVO Steering Committee of which representatives of the Office of the Chief of Ordnance (OCO), Office of the Chief of Research and Development (OCRD), and US Continental Army Command (USCONARC) are members. DISCUSSION SALVO II is a further examination of ammunition types previously investigated in the SALVO I experiment reported on in ORO-T-378.1 The SALVO II experiment, like SALVO I, examines the ammunitions in a combat-simulated environment when fired by experimental subjects having various degrees of rifle proficiency. The results of the controlled experiment were analyzed to indicate differences in ammunition effectiveness and also to determine the effects of target-system andother environmental characteristics onthe accuracy of rifle fire. The following ammunitions were examined in SALVO II: (a) Test ammunition: .30-cal duplex in standard ,30-cal cartridge; ,30-cal triplex in standard .30-cal cartridge; .22-cal duplex in necked-down .30-cal cartridge; and 12-gage-shotgun 32 flechettes fired from Remington autoloading shotgun (Model 11-48A). (b) Control ammunition: .30-cal [M2 bali and armor-piercing (AP)| and .22-cal (high-velocity) simplex ammunition. * * Hu- , ontrol .miiiiunilion i- .If »inn ilcl tlinuiiili ,i>- . 10- in.I . JJ-i ORO-T-397 CONFIDENTIAL *\ .HUUIiin! t loll . CONFIDENTIAL SUMMARY CONCLUSIONS General 1. SALVO II confirmed in detail the predictions of the salvo theory. It also firmly supports the major conclusions and recommendations of the SALVO I ^experiment. """^^ Average aiming error for the SALVO II system is 3.1 mils. If ricochet hits are included it is equivalent to 2.8 mils. JJ. Ricochet characteristics of the various ammunitions were widely different and account for deviations from expected ammunition performance. Better ricochet characteristics account for the 10 percent higher number of recorded hits achieved with .22-cal duplex ammunition compared with those with .30-cal duplex ammunition. Af. No significant difference can be observed between the precision of the first rounds of the .30-cal duplex ammunition and that of .30-cal simplex ammunition. The effect of drop and improper hold-off of the former, compared with the latter can be observed at ranges greater than 250 yd when the ammunitions are zeroed at 165 yd. 5r* Multiple correlation results indicate that target-exposure time and presented area (angular target size) are the major factors determininghit probability in the SALVO II experiment. Other factors influencing hit probability, but to a very much smaller degree, are target activities such as movement and simulated firing. 6. No accuracy differences ascribable to caliber (i.e., .22 cal vs .30 cal) were observed. ^*— .30- and .22-cal Duplex Ammunition 7. In the SALVO II experiment .30-cal duplex ammunition achieved over-all casualty gains, relative to .30-cal simplex (M2 ball) ammunition, of 49 percent. For moving targets its gain was approximately 60 percent, and for firers having the largest aiming errors (averaging 5 mils) the duplex ammunition achieved an average of twice as many casualties as simplex ammunition. 8. .22-cal duplex ammunition achieved an over-all casualty gain, relative to .30-cal simplex ammunition, of 44 percent. For moving targets and for firers having average aiming errors of 5 mils it achieved an average of twice as many casualties as simplex. 9. Both the .30- and .22-cal duplex ammunition fired from a modified Ml rifle functioned satisfactorily through the experiment. .30-cal Triplex Ammunition 10. .30-cal triplex ammunition achieved an over-all casualty gain, relative to .30-cal simplex ammunition, of 32 percent. On the moving targets its gain was 75 percent, and for firers whose aiming error averages 5 mils it achieved an average of twice the number of casualties as simplex. ORO-T-397 CONFIDENTIAL CONFIDENTIAL SUMMARY ■22-cal Simplex Ammunition and 12-Gage-Shotgun Flechettes 11. .22-cal simplex and the 12-gage-shotgunflecheite ammunition-weapon combinations were of such poor technical quality that no useful data concerning their combat potential were collected. RECOMMENDATIONS 1. .30-cal and/or 7.62-mm NATO duplex ammunition should be adopted as standard for combat use. 2. The vaiue and feasibility of improving the ricochet characteristics of .30-cal duplex ammunition should be investigated. 3. Cur rent flechette development programs should emphasize the achievement of satisfactory salvo patterns. In this respect research and development (R&D) of multiple-launched flechettes should be emphasized. 4. Accuracy requirements for new shoulder-fired weapons to be used in aimed fire should be based on an aiming error of no less than 3 mils., ORO-T-397 CONFIDENTIAL CONFIDENTIAL SALVO II RIFLE FIELD EXPEMMEM CONFIDENTIAL CONFIDENTIAL INTRODUCTION In this study the problem was to measure the relative hit probabilities of various ammunitions in combat-simulated aimed rifle fire, and to use these hit probabilities to compare ammunition effectiveness. An additional problem was to examine the effects of various parameters, such as range, target-exposure time, TRAINFIRE qualification scores, etc., on the accuracy of aimed rifle fire. BACKGROUND The SALVO II field experiment is part of the salvo program initiated by ORO in 1951. In 1954 a SALVO Steering Committee was set up under the leadership of the Chief of Ordnance. Work in the program has included several studies by ORO1-10 and the development of prototype salvo ammunitions by OCO. The potential gain in combat effectiveness of these ammunitions was examined in a field experiment (SALVO I) conducted at Ft Benning, Ga., in June and July 1956. SALVO I results are reported in ORO-T-378.1 The field experiment (SALVO II) reported in this memorandum is a continuation of the salvo program. It was conducted during December 1957 at Ft Benning by ORO under the auspices of the SALVO Steering Committee. Troops and facilities for the experiment were furnished through USCONARC by the Infantry Center. Previous ORO publications1-10 describe in detail the objectives of the salvo program. In brief they are to increase the firing effectiveness of infantrymen by increasing hit probabilities while maintaining sufficient lethality of individual projectiles. The method by which the salvo program has achieved this increase is through the design and development of weapons and ammunitions that more efficiently distribute energy expended in rifle fire and that compensate for the inherent human error in small-arms fire. The ammunition and weapons developed so far fire more than one projectile per aiming effort (trigger pull) and in the SALVO I experiment proved to offer significant advances in effectiveness in aimed rifle fire under simulated combat conditions. SALVO I examined the effectiveness of .30-cal duplex or tandem-round ammunition and automatic fire as an approximation of the salvo principle and furnished limited information on .30-cal triplex and 12-gage-shotgun 32-flechette ammunition. In SALVO I, firing was conducted under conditions of limited visibility as well as in daylight, and from two firing positions, standing and sitting. Since the SALVO II experiment was an extension of SALVO I requested by the SALVO Steering Committee, its main purposes were to furnish a final check ORO-T-397 CONFIDENTIAL CONFIDENTIAL on duplex ammunition prior to its submission for user tests and possible adoption, to collect more data on triplex and flechette ammunition effectiveness, and to examine .22-cal duplex ammunition. These objectives are detailed below. (a) The evaluation of .30-cal duplex ammunition in standard cartridge cases. The SALVO I duplex ammunition utilized a long-necked cartridge case that required an elongation of the standard rif. chamber. The second bullet in the SALVO II ammunition rests in the powder charge. (b) The evaluation of .30-cal triplex ammunition in the standard cartridge case. The SALVO I triplex ammunition, in addition to being in an elongated case, blew up on One occasion and the experimentation was stopped at that point.1 (c) The evaluation of .22-cal duplex ammunition in a necked-down standard case. In the .22-cal duplex it was hoped to combine advantages of a smaller caliber with those of a duplex round. (d) The evaluation of 12-gage-shotgun 32-flechette ammunition, which in SALVO I was limited to 700 rounds. The results of SALVO I indicated a full-scale examination would be justified, and 3000 rounds were procured for SALVO II. (e) More detailed measurement of weapons effects and their relation to man and target-system variables was desired than was achieved in the SALVO I experiment. WEAPONS AND AMMUNITION Weapons and ammunitions for the SALVO II experiment were furnished by OCO. The weapons were fabricated by Springfield Armory and the ammunitions (see Fig. 1) were fabricated by the Olin Mathieson Chemical Corp., Frankford Arsenal, and Aircraft Armaments Inc. The weapon-ammunition combinations included: - •' .30-cal 30-cal »implex (M2 Ball) duplex 30-cal triplex Fig. 22-cal simplex .22 -cal duplex 12-gage 32 flechette 1-SALV0 II Test Ammunition ORO-T-397 CONFIDENTIAL CONFIDENTIAL (a) Standard Ml rifle firing :30-cal simplex ammunition (standard issue) and .30-cal duplex ammunition (controlled dispersion*) made by Olin Mathieson. (b) Standard Ml rifle with modified rifling firing .30-cal triplex ammunition (random dispersiont) made by Olin Mathieson. (c) Ml rifle modified with .22-cal bore firing .22-cal simplex ammunition made by Frankford Arsenal. (d) Ml rifle modified with .22-cal long-chamber barrel firing .22-cal duplex ammunition made by Olin Mathieson. (e) 12-gage autoloading shotgun, Remington Model 11-48A, firing a 32-flechette load made by Aircraft Armaments Inc. EXPERIMENTAL CONTEXT The derivation of the target system used for evaluating the salvo concept in the SALVO I and II experiments is described in detail in App D of ORO-T-378.1 This system consists of 22 E and F silhouette pop-up targets at ranges of from 70 to 340 yd. This layout is shown in Figs. 2 and 3. Also shown in Fig. 2 are the distributions of disclosing fire, target concealment, and target movement. Another feature is the simulation of combat stress through the use of electronic shocking devices attached to the firers. Although the simulation of stress, range to the targets, and general layout of the target system remains the same from SALVO I to SALVO IJ, there are certain differences between the two target systems: (a) Light. All firing during SALVO II was conducted in daylight, using only the daylight target positions of SALVO I. The system in SALVO II is generally oriented toward the north, as opposed to a general orientation toward the south in SALVO I. The effect of this change was more uniform visibility conditions from run to run than those obtained in SALVO I. (b) Exposure times. Exposure times for the pop-up targets were reduced by roughly one-third from SALVO I. This reduction was not uniform, and the specific amounts are shown by target in Table 1. The exposure times were shortened at the suggestion of OCO and OCRD. It was felt that the target-exposure times collected in interviews, which formed the basis of the SALVO I system, were overestimated by the interviewees. Since a good common-sense argument can be made for this point of view and since a quick trial of time estimates revealed such overestimation, ORO concurred in the suggestion to shorten the exposure times by about one-third. (c) Fatigue of firers. The firers were double-timed for 5 min before each run. This was an attempt to introduce fatigue, and its subjective effect is discussed in App A. (d) Weather conditions. In contrast to SALVO I, which was conducted at Ft Benning in June and July, SALVO II was conducted on the same post during December. The weather conditions during the experiment exhibit considerable variation. This variation is shown in Table 2. Kor controlled-dispersion duplex ammunition used in SALVO II tlir second bullet deviates at about 2'2 mils from the path of the first. Ti'or the random-dispersion triplex ammunition the paths of the second and third bullets are, within certain limits, random with respect to the lead bullets. \ complete description of random and controlled dispersion is ^iven in \pp 1! of ORO-T-378. ORO-T-397 CONFIDENTIAL CONFIDENTIAL \3±*/ \ DIO 3J ' °- \ "*/ 300 09 V 200 DISTANCE, YD Fig. 2-Solvo II Target System d Small demolition > Target group 0 Large demolition -** Moving forget (consecutive) V Partiol concealment 10 ORO-T-397 CONFIDENTIAL CONFIDENTIAL o o E n. o c \s E Q. o> C Ü > O 11 ORO-T-Ö97 CONFIDENTIAL < CONFIDENTIAL TABLE TARGET-EXPOSURE TIMES: 1 SALVO I vs SALVO II Exposure time, sec Hange, yd Target type SALVO I 74 77 86 89 111 F F E 127 139 152 162 164 165 169 176 216 218 245 259 267 269 334 336 339 FF F F E (moving tgt) E (mo\ ing tgt) E (moving tgt) E E E F F E E E SALVO 11 4.5 3.0 15.0 10.0 4.5 3.0 15.0 19.5 10.0 13.0 9.0 4.5 9.0 6.0 6.0 3.0 8.0 6.0 15.0 31.5 11.0 18.0 3.0 4.5 4.5 9.0 6.0 3.0 2.0 3.0 4.0 6.0 4.0 7.0 3.0 F F F 25.5 16.0 7.5 7.5 5.0 4.0 F 21.0 12.0 235.5 157.0 10.5 Total 1 A, fl \ , \, - WEATHER CONDITIONS DURING SALVO II Day Runs conducted °F Mean wind velocity, mi/hr Mean temperature, n I Direction 1 2 8 6 50 .12 15.5 Cross range Cross range 3 11 35 3 Cross range Weather condition Clear Intermittent light snow Clear ORO-T-397 12 CONFIDENTIAL CONFIDENTIAL The most extreme weather conditions were experienced on the second day. The wind and snow plus the freezing temperature made firing conditions very unpleasant. The number of firers who wore gloves while firing was much higher than on the first and third days (see App A). The .22-cal simplex and duplex runs (two each) and two shotgun runs were fired on the third day, and the weather conditions may account for the slightly lower rate of fire on these runs compared with the rest of the experiment (see subsection on learning). (e) Firing line and firing position. Whereas SALVO I firing was conducted from sitting and standing positions, the SALVO II firing line was constructed so that firing was done from a modified prone position (see Fig. 4). An earthen breastwork was constructed, and firing was conducted from this parapet. This change was made as the result of suggestions by the Army. Its effect was to provide a very much more stable firing position than those used during SALVO I and hence reduced aiming error. ; „, -f**' ■ pM«v AST Fig. 4—SALVO Ü Firing Line Showing Modified Prone Position In addition the firing line for SALVO II was equipped with devices that detonated electric blasting caps among the firers. The devices protected the firers from the blast but were designed to add noise and confusion to the firing line. The firers, however, were seldom aware of their detonation (see App A). (f) Weapon zeroing. Each rifle was zeroed by theman firing it. (SALVO I rifles were zeroed by,experts.) The sights were then adjusted to a battle setting that yielded the least-miss distance for the total range complex of the target system. This process is described in App B. The shotguns were prezeroed at the Development and Proof Services (D&PS), Aberdeen Proving Ground (APG), and were not changed during the experiment. ORO-T-397 13 CONFIDENTIAL CONFIDENTIAL EXPERIMENT SUBJECTS The SALVO II experiment utilized 20 subjects selected from the trainees assigned to the Human Research Unit of the Human Resources Research Office at Ft Benning. They had just completed basic training and their rifle marksmanship qualifications on the TRAINFIRE range totaled 3 experts, 7 sharpshooters, and 10 marksmen. They were organized into two balanced 10-man firing orders. (A detailed description of the firers is found in App A.) These firers differed in the following respects from those in the SALVO I experiment: (a) They had received TRAINFIRE rifle-marksmanship training. (b) Ninety percent of the subjects were enlisted reservists whereas 75 percent of the SALVO I subjects were Regular Army. (c) They were from a special test unit and had participated in other experiments of various kinds. From the point of view of realism, there were both advantages and disadvantages in using the TRAINFIRE troops. In that the rifle training they received is being implemented in the Zone of Interior (ZI) by USCONARC, it adds to the realism of the experiment. The fact that they were drawn from a special test unit, however, detracts from realism. The SALVO I troops were more typical of the over-all Army population in motivation and experience. EXPERIMENTAL DESIGN The simplification of SALVO II, as compared with SALVO I, consists of the use of only one position for firing and the exclusion of night firing. This permitted a shorter experiment—24 runs in SALVO II vs 68 runs in SALVO I. The troop qualifications and test-material specifications involved in the experiment are shown in Tables 3 an^. 1. A total of 24 runs were fired as specified in Table 5, each weaponammunition-squad combination being f'red twice on the target system (2 x 6 x 2). TYPICAL RUN OR FIRING SEQUENCE A typical run or firing sequence followed a set pattern. Appropriate rifles and ammunition were placed at the firing positions (1 to 10 on Fig. 2). The firers then took their places at their assigned firing positions. After the stress simulators (electric shockers) had been placed on each firer's leg the firers took up a comfortable position on the earth parapet on the firing line (see Fig. 4). As soon as the firing line was ready, the programed target-system sequence was initiated in the contrui and recording center. At that point the target-system events began, i.e., target appearances, demolitions, electric shock for the firers, etc., as described in the programs in App C. The targets appeared sequentially, all men firing at every target seen. The electronic chronological recording system made possible the identification of shots fired and hits by each individual. The program ran for 300 sec. 14 ORO-T-397 CONFIDENTIAL CONFIDENTIAL TAHI.K 3 TROOP QUALIFICATIONS AND DIVISION Squad Kxperts Sharpshooters Marksmen A li 1 2 4 3 5 5 TABLE 4 WEAPON- AMMUNITION COMBINATIONS Ammunition Weapon .30-cal simplex (M2 ball) ,30-cal duplex .30-cal triplex ,22-cal simplex .22-cal duplex 12-gage flechette Fired from standard Ml rifle Fired from standard Ml rifle Fired from modified Ml rifle Fired from modified Ml rifle Fired from modified Ml rifle Fired from Model 11-48A Remington autoload shotgun TADI.K 5 SALVO II FIRING SCHEDULE Date Weapon-ammunition combination 10 December .30-cal triplex (Ml) .30-cal triplex (Ml) .30-cal triplex (Ml) .30-cal triplex (Ml) .30-cal duplex (Ml) .30-cal duplex (Ml) .30-cal simplex (M2) APb .30-cal simplex (M2) AP .22-cal simplex (.22-cal Ml) .22-cal simplex (.22-cal Ml) .22-cal duple: (.22-cal-long-chnmber Ml) .22-cal duplex (.22-cal-long-chamber Ml) 12-gage flechette (shotgun) 12-gage flechette (shotgun) 11 December 12 December Squad Progrot \ .! 1! 1! A 3 li \ li A A 1! \ li 4 4 5 5 6 d 1 1 2 2 \ :i if .! \ .22-cal duplex (.22-caMong-chamber Ml) .22-cal duplex (.22-cal-long-chamber Ml) .22-cal simplex (.22-cal Ml) .22-cal simplex (.22-cal Ml) 12-gage flechette (shotgun) 1! 12-gage flechette (shotgun) ..'10-cal simplex (M2ball) (Ml) .30-cal simplex (M2 ball) (Ml) .30-cal duplex (Ml) .ilO-cul duplex (Ml) \ \ \ f li r> n \ li 1! i i i i \ •> li o '"The programs [i.e., the sequence in which croups of targets appeared (sec | IC 2). demolitions were detonaterl, and electronic shock was administered] were prepared !i\ randomizing the sequence of their occurrence to prevent learning. \s is noted in I .ilde 1 six different programs were used. These programs appear in \pp (. "AP ammunition was brought to the field In mistake ll.ilher than miss .1 .!.!\'s firine it was used. Its characteristics differ slighth from \i.' ball mm nniti'm. ORO-T-397 15 CONFIDENTIAL CONFIDENTIAL DATA COLLECTION The following types of data were collected in the experiment: (a) Hits on targets recorded on paper-target faces. (b) Hits on targets recorded electronically in time. (c) Trigger pulls recorded electronically in time. (d) Target movement and up times recorded by an elapsed-time camera (also electronically recorded). (e) Weapon malfunctions and the time during which the weapon was out of action recorded electronically by an observer. (f) Hits on targets by flechettes recorded by AN-N/ 6 gun cameras at the targets. (g) Ammunition expended recorded by ammunition count before and after each run. (h) Meteorological data (previously described). (i) Subjective information concerning the firers collected in debriefing interviews after each run. INSTRUMENTATION The objectives of the instrumentation of the SALVO It experiment were twofold: First to ensure reproducibility from run sequence to run sequence, and second to collect the data mentioned in items b to f above. The central aspect of the reproducibility function of the instrumentation is the sequence controller or programer, which is described and illustrated in App E. This unit, by means of a paper punch tape, permitted the precise reproducibility in time of all scheduled events on the target system. The main problem in the data-recording functions of the instrumentation was to record target hits and trigger pulls with a resolution time of 50 msec. Resolution time of fMs magnitude permitted the identification of first-, second-, and third-bullet hits and single, double, and triple hits. It also made possible relating trigger pulls to hits on the target, thus permitting the analysis of the data by man, although all men were able to fire at each target when it appeared. The only serious problem encountered in reducing the data was the result of varying ammunition velocities. Although the average time of flight from each firing position to each target was known accurately, the identification pt hits from the second and third projectiles from a single trigger pull depended on the fact that these projectiles traveled more slowly than the first projectile. Hence they could be identified by the amount of time it took them to reach the target. In actual practice it was found that bullet velocities varied from round to round and the time of flight of a second bullet from one duplex round might be shorter than the time of flight of a first bullet from another duplex round. The problem was even more severe in triplex ammunition and is illustrated in Tables C30 and C31 in App C. This was not a problem that the ORO instrumentation could solve, but it did not affect the identification of double hits from a single trigger pull. These could be determined by the characteristic time separation between bullet strikes. It did affect the accuracy, however, of the identification of first- vs second- vs third-bullet hits. 16 ORO-T-397 CONFIDENTIAL CONFIDENTIAL The complete description of the salvo programing and data-recording system is found in Apps D and E of ORO-T-3781 and App E of this paper. Taken together these appendixes furnish sufficient information for the reproduction of the system. The SALVO II system performed its functions of control and data collection with almost perfect reliability. From the point of view of comprehensiveness, SALVO II accurately measured two factors that could only be estimated in SALVO I: (a) the number and duration of weapon malfunctions, and (b) the effect of malfunctioning targets. In SALVO I one of the major difficulties in the analysis was the fact that it could only be roughly estimated how long a weapon malfunction kept a given weapon out of action. To remedy this situation in SALVO II a monitor system was instituted. It consisted of observers at each firing position who could signal the duration of weapon malfunctions. The system is described in App E and the monitors are shown in Fig. 4. An allied problem concerned the malfunctions of targets. Occasionally, a target did not appear, did not stay up for the entire duration of its programed appearance, or stayed up too long. This type of malfunction occurred more frequently in SALVO I, but only observational records were taken. Two methods were used in SALVO II to measure the actual duration of target exposure: (a) an electronic record was kept of the up and down times for the targets, and (b) the lapsed-time cameras (one frame per second) synchronized with the system program afforded a visual indication of target exposure. This was used as a detailed check on all aspects of the operation of the field layout of the target system. The photographic hit recording of the flechette runs did not prove very successful because identification of flechette hits from the film data was in most cases impossible and camera film speeds could not be adequately controlled. This type of recording was required because of the lack of time separation of flechettes from one trigger pull and the tendency of the flechettes to short-circuit the aluminum sandwich targets. Both faulty data recording and poor ammunition performance render the flechette results almost unusable. The results that are available are included in Table 5 and are detailed in App D. Data Reduction The principal problems of data reduction concern the relating of hits on targets to individual trigger pulls and, in the case of multiple-bullet ammunition, the identification of each hit as a first-, second- (in the case of duplex and triplex), or third- (in the case of triplex) bullet hit. The hit-trigger-pull relation and the identifications depend on accurate knowledge of the time of each bullet strike on the target, the time of the trigger pull, and the time of flight from the firing line for each type of projectile. The first two pieces of data were measured as described previously and times of flight were determined by examining a large number of individual hits with simplex ammunition, double hits with duplex ammunition, and triple hits with triplex ammunition fired from each firing position. Time data were recorded on Esterline-Angus tapes in the field and later transferred to International Business Machines (IBM) cards through the use of a Telereader analog-to-digital computer. This information was then printed out, and the numerical time relations were individually examined. ORO-T-397 17 CONFIDENTIAL CONFIDENTIAL In relating hits to trigger pulls, the accuracy with which the data were collected permitted the almost complete allocation of hits. Only 1 percent of the hits could not be ascribed with certainty to a given trigger pull. In the case of the identification of first-, second-, and third-bullet hits, however, less accuracy was achieved. This is due not so much to the recording as to the variation in time of flight or bullet velocity from round to round. This is illustrated in Fig. Cl, where it can be seen that triplex first- and second-bullet times of flight overlap, i.e., in some cases a lead triplex bullet's velocity was such that its time of flight to the target was the same or slower than the second bullet of another round. Thus there are cases where it is impossible to determine for a given single bullet strike whether it was a first or second bullet. The same is true to a lesser extent for the duplex ammunition. The magnitude of the error for the four analyzed ammunitions is shown in Table C31. In actual practice in the data reduction every effort was made to identify bullet hits from the logical context as well as by using time-of-flight data. Cutoff points were used at the points where time-of-flight distributions intersect. Since the overlapping times of flight presumably occurred randomly and affected a very small percentage of the data, they can be assumed to have little effect on the analysis described below. DATA ANALYSIS The first problem in the analysis of the data after they were reduced was dealing with cases where data were missing. In SALVO I there was a considerable amount of missing data, and compensating for it was one of the most difficult problems in the analysis. In SALVO II, however, owing to the improved and more comprehensive instrumentation and to better weapon and target performances, it was not a serious problem. Missing data in the SALVO experiments occurred for one of three reasons. First a target failed to appear, second a weapon malfunctioned and the firer did not fire his usual number of shots, or third the target appeared but the electronic recording did not function properly. There was only one target that did not appear in the course of the 16 runs that were used in the major comparisons {.22-cal simplex and flechette runs excluded), and for it an average value determined on the basis of the other runs was filled in. The second case, that of weapon malfunction, was extremely important in the SALVO I analysis. In SALVO II, however, only 20 minor weapon malfunctions occurred during the 16 runs. These malfunctions were of such short duration (they are shown in detail in Table B7) that they are ignored in the SALVO II analysis. In the third case, where the data were not recorded electrically because of a target shorting out or a ricochet failing to register properly, alternative data from the paper target faces were used.* Serious malfunctions of this type occurred 16 times during these 16 runs (or 352 target appearances) and affected 329 out of the 4252 hits (see Table C22). These 8 percent of the hits Kiroc hets always perforated the targets but occasionally, because of their broad aspect of presentation to the target, would not properly broach the insulating rubbc layer in time to record (see App K). 13 ORO-T-397 CONFIDENTIAL CONFIDENTIAL on the paper target faces were prorated as to first-, second-, and third-bullet hits on the basis of average values of other runs and as to the firers on the basis of the individual firer's aiming error determined in the other runs. This process is described in detail in App C. The only other discrepancy between electronic-record and target-face count encountered in the data analysis was the result of a target failing to drop because of mechanical failure. Because the electronic recording ceased at the programed drop time, it was accepted as correct and the target-face count was ignored. However, because of this malfunction the following target that appeared did not receive as many hits as it normally would have, and again average numbers were used in adjusting its hits. In summary it may be said that adjustments to actual hits on the targets, i.e., holes in the paper faces, were made for less than 1 percent of the total hits. Adjustments to the electronically recorded data to bring them into line with the holes in the paper faces were made for 6 percent of the total hits and always on the basis of average values determined on other runs or on aiming errors computed for the man and ammunition involved. RESULTS This section considers two types of experiment results: (a) those pertaining to ammunition differences and (b) those pertaining to the general nature of rifle fire on the salvo target system. Ammunition Differences The effectiveness criterion used in this analysis is casualty gain per trigger pull, i.e., test-ammunition score minus control-ammunition score, divided by control-ammunition score. Ammunition and weapon weights for those ammunitions on which usable data were obtained were essentially the same. In addition production costs of the ammunitions are comparable. Hence comparisons by weight and cost are not of primary pertinence. Another criterion, gain in number of targets hit per trigger pull, is included in the major tables. The effects of overkills for double and triple hits from one trigger pull are computed in App O of ORO-T-378,1 and this same method is used here. The specific casualty probability used for single hits is .7; for double hits, .91, i.e., .7 + [0.7 (1 - 0.7)]; and for triple hits, .97 (similarly deduced). The detailed analysis of overkills is presented by man, by target, and by run in Tables Cl to C16. The casualty probability for triplex hits is degraded by a factor of 18 percent at ranges of 200 yd and beyond. This is based on the fact that at those ranges the triplex ammunition used will not penetrate helmets, and the equivalent approximation of the decrease in lethality based on App B of ORO-T-3781 is 18 percent. In contrast to SALVO 1 ammunitions, the .30and .22-cal duplex ammunitions penetrate helmets at a 400-yd range (see App B). Majo