(LEVEL

TN no. N-1595PROPELLANT-EMBEDDED ANCHORS:

title: PREDIC :O OFHOLDING CAPACITY INCORAL AND ROCK SEAFLOORS ECTE

author: J. F. Wadsworth, III and R. M. Beard Sdate: November

1980

sponsor: Naval Facilities Engineering Command

program nos: YF59.556.091.01.204

[CIVIL ENGINEERING LABORATORY

NAVAL CONSTRUCTION BATTALION CENTER16J.Port Hueneme, California 93043

Approved for public release; distribution unlimited.

81 8 19 0726 At-.

____Unclassified

SECURITY CLASSIFICATION OF TH4IS PAGE '*W,, Isl.

REPORT DOCUMENTATION PAGE 11IOE OILTNFR1. REPORT NUMer GOVT ACCESSION No, 3 RECIPIEN; ISCA LO MEB

/ / - TN-1595.. DN987005 I ___ .~

4. ItYf (- S.W. TYPE UF REPORT & PERIOD COVERED

PROPELL1ANTr-lM,%BEI)i)EI.) ANCH ORS:PREDIC:TION OF HIOLD)ING CAPACITY IN Not final; Oct 1977 Sep 1979

6 PERFORMING ORG REPORT NUMBER

CORAL AND) ROCK SEAFLOORS~. &E~or.,B CNTRC-R GRAN-T N UMBER.,

J. V. Wadsworth. III and R. NM. Beard

2 PERFORMING O RGANIZATION NAME AND ADDRESS I0 PROGRAM ELEMENT. PROJECT TASK

CIVIL. ENGINEERING LABORATORY AREA & W'TBE I)NIT NUMBERS

Naval Construction Battalion Center 62759N;Port H ueneme, California 93043 YF59. 556.091.01.204

I I CONTROLLING OFFICE NAME AND ADDRESS It REPORT DATE

Naval Facilities Engineering Command / November 1.980

Alexandria, Virginia 22332 1 1S *1IWR 5 Z4OAP'QA& I

14 MONITORING AGENCT NAME & ADORESSIII d'1111-,, (rom (,,,Ir.IIIT 011110) IS SECURITY CL.ASS I. 'h'. I.PII)

Unclassified ______

IS.$ DOECASSIFICATION DOUNGRACINGSCHEDULE

I S. DI S TRIB9u TION STA T EMkE NT I, ?h I ROP,,rt)

Approved for public relcase; distribution unlimited.

17, DI S TRI1BU)TIO0N ST7 AT EM ENmT.(oi lt" h s I I Ir. d in HR81.,A 2 0. I14 d(ifw- .I I- R ApasI)

IS SUPPLEMENTARY NOTES

19. KEY WORDS (C-11--..~ .1 5d. I~ Ios5 -F T d-,, h 151-1, -15WIb,

PlIate anchors, proipellant-embedded anchors, anchors, seafloors, moorings, coral seafloors.

rock scatloors.

20 ABSTRACT (C-,nIIA.. . .,.. I...r I d id-flI0 F. - Ik -6s.0~

11ropel la nt-embedded anchors have becomei ain important asset in the Nav ,I,'s nmoorngequipment inventory. They' offer potential for anchoring in coral aind rock %cafloolrs whereconvenitional anchors will not wolrk or present logistic prl lblems. Most olperatiolnal inistalla-tions of propellant-embedded mncho~rs bave beenr in~ coal scafloors and these anchors have

usak en prl ot -loaded t Io their design ca paci . lb It mat cii 11A C I leIIS of thCSC

I Continued) _

DD I 2,AO"07 1473 EDITION OE I NFOV 695 OBSOLETE UnclassificdSECRIY C ASIIC TIN;F T." PAGEf ,;,. -. d.

Unclassified

SECURITY CLASSIFII ATION OF THIS P AGE .. n nl1.a ssi fidd

20. Continued

installations have not been determined, hence the full potcntial ot prope~lant-cmibeddcdanchors in coral has not been determined. This report presents the results f a programof model and full-scale propellant-embedded anchor tests in coral and rock. [hc significance of the results and analysis of the data are presented. A promising equation totpredicting holding capacity in coral is developed from the data. It is concluded that thedata base of tests in rock is not sufficient for leveloping an equation for prediction ofholding capacity.

Library CardCivil Engineering LaboratoryPROPELLANT-EMBEDDED ANCHORS: PREDICTION OFHOLDING CAPACITY IN CORAL AND ROCK SEAFLOORS,by J. F. Wadsworth, III and R. M. BeardTN-1595 70 pp illus November IQ80 Unclassified

1. Propellant-embedded anchors 2. Mioorings 1. YFS.556.091.OI.204

Propellant-embedded anchors have become an important asset in the Navy's mooringequipment inventory. They offer potential for anchoring in coral and rock seafloov, %hereconventional anchors will not work or present logistic problems. Most operational installationsof propellant-embedded anchors have been in coral seafloors and these anchors have usuallybeen proof-loaded to their design capacity. The ultimate capacities of these installations havenot been determined, hence the full potential of propellant-embedded anchors in coral has notbeen determined. This report presents the results of a program of model and full-scale propeilant- Iembedded anchor tests in coral and rock. The significance of the results and analysis of thedata are presented. A promising equation for piedicting holding capacity in coral is developedfrom the data. It is concluded that the data base of tests in rock is not sufficient for developingan equation for prediction of holding capacity.

UnclassifiedSI'> ,R/IT5 V ASS<I( i ' ;. F , .1' 531* ' tiP,,, . ! re

CONTENTS

Page

INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . I

CORAL ANCHORS. .. .................... .......

Previous Tests. .. ....................... 2Barbers Point Tests .. ..................... 2BARSTUR Tests .. ..................... ... 3Analysis of Test Data......................4Prediction of Holding Capacity .. ....... ......... 4

ROCK ANCHORS .. ..................... ...... 6

Previous Tests. .. .................... ... 6NWC, China Lake Tests .. .................... 7Analysis of NWC Test Data .. .................. 7White Sands Missile Range Tests. ......... ....... 8Analysis of White Sands Missile Range Data .. .......... 9Development of Empirical Holding Capacity Equation

for Rock Seafloors .. ................... 10

SUMMARY .. ........ ...................... 14

CONCLUSIONS .. ........ .................... 14

RECOMMENDATIONS .. ....... ................... 15

*REFERENCES. ........ ..................... 16

A'e'o m T~

r,"

v .

INTRODUCTION

The Civil Engineering Laboratory has developed propellant-embeddedanchors as an alternative anchoring system. These anchors can be usedin situations where conventional anchors are difficult to use or willnot meet mooring design requirements. In the early stages, the anchorswere developed for use by Navy salvage vessels in coral seafloors(NAVSHIPS 0994-002-6010; Smith, 1971). Since that time embedment anchorsof varying designs and sizes have been tested and used in a variety ofseafloors (Taylor, 1976; True and Taylor, 1976; True, 1977; Wadsworthand Taylor, 1976). Propellant-embedded anchors are being used becausethey are easy to precisely install with no chance of the anchor draggingout of position.

Of the anchors that have been installed for actaal operational use,most have been in coral seafloors. These installations have usuallybeen proof loaded to design capacity; however, few data on the ultimateholding capacities of the anchors have been acquired. Consequently, itis not known at what factor of safety these anchorages have been used.The factor of safety may have been small or quite large.

There has been no operational use of the propellant-embedded anchorsin rock. An attempt to place a mooring for a semi-submersible drillingrig in rock using a coral type projectile was not successful. Severalrock flukes for propellant-embedded anchors have been built and testedwith sharply inconsistent results. If the performance of rock flukescould be made more consistent and predictable, it is likely they wouldbe widely used because of their high efficiency and because mooringscopes could be reduced (Valent, 1973 and Wadsworth, 1976).

This report presents the results of a program of model and full-scale embedment anchor tests in coral and rock. The significance of thetests is discussed, and analyses of the test data are presented. Basedon the test data, separate equations are developed for predicting holdingcapacity in coral and rock seafloors. This work was sponsored by theNaval Facilities Engineering Command and funded under the Ocean FacilitiesEngineering Program.

CORAL ANCHORS

The configuration of the coral flukes has undergone many changesduring the development of the various embedment anchors (Figures 1through 6). The present coral fluke design is basically the same foreach of the CEL anchors. The only difference is that the 20K and 10Kcoral flukes (Figures 5 and 6) have a spur on the bottom of the fluke.This spur was added to induce a keying action in soft coral seafloors.It also balances the mass of the projectile so the main fluke plate isclose to the axis of the piston. The spur is not necessary on thelarger 100K anchor.

Previous Tests

Embedment anchors have been tested by both the Navy and the Army.Results of tests conducted in coral seafloors prior to this project arelisted in Table 1. Most of these tests were of limited use in a database because of the variation in fluke configurations and the lack ofdata on properties of the seafloor materials. Samples of coral fromaround the embedded flukes were collected by divers in two of the testseries: the SUPSALV anchor tests at Barbers Point, and the CEL 20Kanchor installations for the Barking Sands Underwater Range Extension(BSURE) mooring. Diego Garcia and Argus Island are mooring and salvagereaction point installations in coral with the CEL salvage and lOOKanchors, respectively. In these cases, no attempts were made to pullout the flukes, but proof-loads were applied, and these proof-load datamark a lower bound holding capacity. Review of the available dataindicated additional coral tests were required to develop a method topredict holding capacities in coral seafloors.

Barbers Point Tests

Barbers Point, Oahu was selected as the site for full-scale testsin coral. Testing at this site offered three advantages: (1) someknowledge of the seafloor properties already existed from the SUPSLAVtests, (2) a direct comparison to previously acquired data would bepossible* and (3) support was available at nearby Pearl Harbor. Inorder to conduct as many tests as possible with the available resources,a lightweight propellant-embedded anchor (CEL 10K anchor, Figure 7) waschosen.

The Barbers Point tests were conducted in June 1978. Harbor ClearanceUnit One provided the supporting personnel and equipment to install theanchors. The craft used was a LWT, an aluminum side-loadable warpingtug. The anchors were raised and lowered using a bow A-frame and a3-ton-rated pneumatic winch. After installation of each anchor, thewinch was used to its capacity to proof load the embedded fluke.

All 12 anchors to be tested were installed in 1 day. The embeddedflukes were then surveyed by an engineer-diver. Figure 8 is a photographof a typical embedded fluke. Four of the flukes were embedded in coralbeneath a sand layer and could not be inspected. Each site was photo-graphed, fluke penetration and size of impact crater measured, sitemarkers placed, and samples of loose coral collected from around thefluke.

The anchors were load tested by the USS BEAUFORT ATS-2. Loads wereapplied by connecting into the anchor downhaul with a wire rope led overa bow roller, taking out the slack with the capstan, placing the line ina carpenter stop, then applying load by backing down the ship. Loadswere measured using an in-line load cell. Eight of the 12 anchorsfailed when the wire rope downhaul failed; only four flukes were extracted.Following the load tests the anchor sites were again inspected by divers.Figure 9 is a photograph of the same fluke shown in Figure 8 afterloading. Note that there is no evidence of the fluke extracting or ofthe surrounding coral displacing; this was typical of the flukes that

2

i..6< .-

could be inspected. Figure 10 is a before and after photograph of oneof the flukes which did extract. Complete information on the results ofthe load tests is given in Table 2.

Review of the test results and procedures indicates that the recordedloads may have been considerably lower than the loads actually appliedat the anchors. First, the load line to the anchors passed over theship's how roller. Friction and the inertia of this roller would resultin different line tensions inboard and outboard of the roller. Second,the swell and sea under which the tests were conducted resulted in veryrapid load applications (snap loads) that probably exceeded the loadrecorder's dynamic response capability. Third, one anchor that had beenproof loaded to a nominal 26 kN from the installation craft was pulledout by the BEAUFORT at a measured load of only 4.5 kN. Fourth, eight ofthe 12 tests ended when either downhaul cables or mechanical connectionsat the fluke failed. All of these failures occurred at on-deck recordedloads much lower than the breaking strength of the failed mechanicalparts (-260 kN).

BARSTUR Tests

The 20K coral embedment anchors that were installed off NohiliPoint at the Barking Sands Tactical Underwater Range (BARSTUR) inAugust 1976 (see data in Table 1) provided an opportunity to conductpull tests on previously installed anchors. The anchors had been usedto moor a cable-laying ship during installation of the shore end of thecable runs coming in from the new, expanded range. Figure 11 is asketch of the design of each mooring leg. Note that two 20K anchorswere used for each mooring leg. Following this use, the buoys wereremoved from three of the five moorings, and the wire rope bridles anddownhauls were laid on the bottom. The other two moorings were used tomoor two large navigational buoys.

CEL requested and received permission from the Pacific Missile TestCenter (PMTC) to perform pull tests on the six unused anchors. Since adetachment from Underwater Construction Team Two (UCT 2) was to performa cable survey at BARSTUR in the summer of 1978, PMTC also granted CELthe use of the detachment to locate and buoy the anchors to be tested.A section of the bridle, the wire rope connecting two anchors at eachmooring, was recovered and returned to CEL for testing. The downhaulsand bridles had been submerged for 2 years with no corrosion protection,but the two sections tested broke at 329.3 kN and 258.1 kN. The ratedbreaking strength for this 7/8-inch-diameter 6x19 wire rope was 354.2 kN.The wire ropes, thus, seemed to be in satisfactory condition to placemeaningful loads on the anchors (the original mooring design had assumedeach anchor to hold 222.5 kN).

Load tests were performed with the USS BRUNSWICK (ATS-3) in thesame manner as was used earlier at Barbers Point. A strap of nylon linewas added to the load line to reduce the snap loads experienced atBarbers Point. Also, sea conditions were better, which allowed a moregradual loading. The anchors were surveyed before and after load testing,with the test results as shown in Table 3. In every case the wire ropedownhaul failed, thereby establishing a lower bound holding capacity.Note that the failure loads are below the breaking loads of two test

3

sections of wire rope. It is thought the reasons for this are similarto those for a similar problem with the Barber's Point tests. Post-testdiver surveys failed to show any displacement of the anchor or thesurrounding coral.

Analysis of Test Data

The data from the Barbers Point 10K anchor tests, the BARSTUR 20Kanchor tests, and the proof loading data for the 100K anchor from DiegoGarcia and Argus Island were plotted as histograms with the normal curveadded. Figures 12, 13, and 14 show these data. The 90% confidencelimit, the holding capacity which will be exceeded 90% of the time, isplotted for each case. The confidence of exceeding the rated holdingcapacity is 95% for the 10K anchor, greater than 99% for the 20K anchor,and 88% for the 100K anchor.

The confidence limits and mean holding capacities given in thefigures are conservative since only 3 of the 11 data points in the 10Ktest data and none of the 20K or 100K test data are actual anchor extrac-tions. The data are either the maximum proof-load applied or the maximumrecorded loads experienced before the wire rope downhaul failed. Actualanchor pullout loads would shift the curves in Figures 12, 13, and 14further to the right than is shown. In addition, the wire rope failuresat BarbL.rs Point and BARSTUR were recorded at levels much less than therated breaking strength. It is likely that much higher loads occurredbut were not recorded on deck due to friction and inertia of the ATS bowsheave or a slow response from the load recorder.

Prediction of Holding Capacity

Data Base. An equation for predicting holding capacity in coralseafloors can be developed from test data only where the anchor flukeswere pulled out. This places a restriction on the data base because iteliminates data from both proof-loading tests and tests that resulted inmechanical failure of anchor hardware. Table 4 lists all the data foundon embedded anchor coral tests listed in Table 1. Where known, thevalues of parameters that are thought to have the most influence onholding capacity are listed in Table 4 for each test. The fluke mass(m) and velocity (v) provide a measure of the energy available forpenetration. Fluke frontal area (F) affects penetration while thefluke's embedded surface area (A) determines the area over which thefluke and coral interact to provide holding capacity. Fluke load pointeccentricity, e, and loading angle, 0, define nonfrictional aspects ofholding capacity. Load point eccentricity is defined as the ratio ofeccentricity of the downhaul attachment point from the main fluke axisto the length of the fluke. Two coral properties, unconfined compressivestrength (a) and the dry unit weight (p), help to define the effect ofthe target material. Also included are the data from two model coralfluke tests performed at NWC, China Lake, in a chalky limestone thatsimulated a soft coral. Nineteen of these tests resulted in pullout ofan anchor fluke and are listed separately in Table 5.

There are two major problems with this data base. First, an exam-ination of Table 5 shows that in many cases parameters are unknown. Ofparticular concern is the absence of coral properties for the first 14

4

tests listed. Second, the load measurements for the three CEL 1OKanchor tests at Barber's Point are not thought to be indicative of theloads applied to the flukes as previously discussed. These 17 testscannot, therefore, be used to determine the relative significance of thevarious parameters affecting holding capacity. There are then only twotests, the model 20K tests at China Lake, that are usable for developinga predictive equation that is a function of fluke and coral parameters.However, this data base is too narrow to achieve a meaningful result.

Development of Empirical Holding Capacity Equation. In spite ofthe difficulties with the data, an effort was made to determine therelative significance of the parameters. This was done by performinglinear regression analyses for different combinations of test parametersusing a method described by Tolson (1970). No distinction was madebetween proof loads, pullout loads, or loads that caused hardware failures.The dominating parameters were found to be fluke mass and velocity.Some coral property, however, seems to be a necessary ingredient as itaffects both penetration and pull resistance.

An equation was then developed for holding capacity, H, as a functionof fluke mass, m, fluke velocity, v, and coral compressive strength, a.

H = f(m, v, a) (1)

From dimensional analysis it can be shown that

3 2 4H f(a • m *v) (2)

or

H = k a1/3 1m v2 2/3Tm 3

Equation 3 is dimensionally correct and shows that holding capacityshould be strongly influenced by the kinetic energy (KE = 1/2 my2 ) ofthe fluke and only moderately influenced by the coral's compressivestrength (since H varies as the cube root of a).

It is not possible to test this equation against the data basebecause of the paucity of information on coral strength for the testswhere anchors pulled out. However, because coral strength is not nearlyas significant a parameter as kinetic energy, a plot was made of holdingcapacity versus just kinetic energy (Figure 15). Every test from Table 4,where the kinetic energy could be determined and the holding capacitywas known (58 data points), is included except for the fifth CEL 10Kanchor test at Barber's Point. This test was excluded because the loadmeasurement was very inaccurate. The circled points in Figure 15 representtests where pullout was achieved. The best fit equation for all thedata is:

H = 0.068 KE0 .5 8 (4)

5

At -

The best fit equation for the data points where pullout occurred is:

H = 0.024 KE0 .68 4 (5)

Both equations are made dimensionally correct by making the units of theconstants the cube root of N/m2 when the kinetic energy is in joules.Single-sided lower bound 90% confidence limits are presented in Figure 15along with each best fit line. What is very significant about both ofthese equations, and particularly Equation 5, is that the power of thekinetic energy term is very near the two-thirds power derived by dimen-

sional analysis. This result also suggests that Equation 3 may be anappropriate equation for including coral properties. However, thedetermination of the empirical constant for Equation 3 requires addi-tional data not presently available.

Equation 4 is more a summary of experience with various coralanchors than a predictor of holding capacity. This is because three-fourths of the data upon which it is based are proof loads and hardwarefailures. It provides a very conservative estimate of holding capacity.On the other hand, Equation 5 is based only on pullout data and doesprovide a true estimate of ultimate holding capacity. On Figure 15,lines have been marked at the kinetic energies of the four CEL propellant-embedded anchors at optimum ballistic performance with a coral fluke.Table 6 is a tabulation of the predicted mean holding capacities and theholding capacities expected to be exceeded 90% of the time for each ofthe anchors.

There are limitations to Equation 5. Because it does not take intoconsideration coral properties or fluke shape, its use is restricted tothe fluke shape and range of properties given in Table 5. Howev(:,. asshown by Equation 5 and as discussed by Taylor (1976) for sediments, theholding capacity of a given propellant-embedded anchor is largely afunction of the energy transferred to the fluke. Therefore, Equation 5should have considerable usefulness in predicting the holding capacityin coral of the CEL propellant-embedded anchors.

ROCK ANCHORS

Previous Tests

The amount of existing data on propellant-embedel anchors in rockis considerably less than the data available for coral. The flukedesigns which have been used have had mixed success (Smith, 1971; Taylorand Beard, 1973; Taylor, 1976). The SUPSALV rock fluke (Figure 16) wassimply a coral projectile modified by increasing the fin's taper aridhardening the projectile tip. The CEL 20K rock fluke design (Figure 17)was based upon the SUPSALV anchor results, the scale model tests conductedin a simulated rock medium (True, 1975), and the rock penetration workconducted by Sandia Laboratories tYoung, 19?0).

After testing of these flukes recommendations were made by theinvestigators to conduct further research aud development on a rockfluke for emhedment anichors. Followirig pieliminary assessmeuts (Wadsworth,1976; Beard and Wadsworth, 1976) it ^aE decided tha! the bes.t approach

6

to developing an understanding of embedment anchor holding capacity inrock was to perform as many large-scale model tests as possible in avariety of rock types.

__C, China Lake Tests

The first series of t-ests was conducted to explore the effect ofdifferent projectile shapes and ballistic parameters upon holding capac:ity.The tests were conducted in exposed outcrops of granite and extrusive

basalt at the Naval Weapons Station, China Lake, Calif. Tests wereconducted on land to allow a greater number of tests to be conductedwith available funds. The anchor projectiles were launched by theMagnavox Company's propellant-embedded anchor test stand which wasdeveloped for firing the 7-pound Magnavox propellant-embeddcd anchorflukes on land. The propellant energy is transferred to the flukes by apiston that mates to the fluke and is inserted into the gun barrel.Using this launcher, a 1:3.75 geometric scale could be used assuming the20K anchor as the prototype. Impact velocity of these models was thesame as that of the prototype.

Eight different fluke shapes were used in these experiments. Theyare designated as Dart, Serrated Dart, Star, Cone, Stake, SerratedStake, Tapered Stake, and Coral and are depicted in Figure 18. The Dartwas considered to be the control fluke as it was geometrically scaleddown from the 20K anchor rock fluke then in use. The Serrated Dart andStar are variations of this shape. The three stake shapes represent asignificant departure from the DART as does the CONE. Each of the STAKEprojectiles was shortened to keep the L/D ratio below 8. The weightlost by removing the fins and shortening the body was made up by using aheavier piston. The Coral was a scaled down coral fluke.

These shapes were all tested in a weathered granite and, with theexception of the coral fluke, in basalt as well. Only the coral flukewas tested in a chalky limestone, and those results are given in thesection on coral anchor holding capacity. All holding capacities weremeasured in direct uplift using a calibrated hydraulic jack.

The results of the NWC, China Lake tests are given in Table 7.

Analysis of NWC Test Data

Eighteen tests were conducted in a weathered granite, and 12 testswere conducted in a vesicular basalt. The properties of these materialsare given in Table 8.

Comparison of the tests, which were redundant with respect to thefluke and rock, indicates considerable variation in holding capacityeven when similar conditions are maintained (compare tests 1, 2, and 6,tests 3 and 9, tests 4, 7, and 18). The most apparent reason for thisvariation is the amount of radial fracturing in the target caused by thepenetration. These fractures relieve the normal stress on the projectileand result in lower holding capacities. The tendency to form radialcracks was greater in the dart-type projectiles than it was in the star,cone, and stake-type projectiles. Cracking was also greater when flukeswere embedded in smaller outcrops or closer to the edge of large outcrops.

7

In basalt, all of the projectiles were damaged and would onlypenetrate if they impacted in an existing crack or joint. Two modifiedtlukes were built. One design was a dart fluke with a blunter nose anda heavier after body, the other design was a heat-trea'ted (one with theoriginal nose shape and a heavier after body. The modif ied darts impactefspalling off large pieces of rock, but did not penetrate. The heat-treated cones penetrated in both cases (tests 29 arid 30) about one-halfof their length. Even with a 4140 steel alloy fluke hardened toRockwell C 34-36 to achieve excellent toughness characteristics, the

nose was slightly bent.The conical fluke was selected as the shape to use for all further

rock tests. The frontal area of this shape increases evenly duringpenetration which continually compresses the surrounding rock. Becauseof this, the conical fluke had the greatest amount of surface areacovered with a comminuted rock coating. This coated surface is the areaof the projectile in which rock is actually bonded to the fluke due tothe heat generated while penetrating. While thir may or may not actuallyestablish a rock-to-fluke bond, it certainly increases the surfaceroughness of the projectile and thus the frictional resistance to pullout.

The even increase in frontal area also eliminates the developmentof local high stress areas which lead to cracking of the rock, as wascaused by the fins of the dart-type projectiles.

The final advantages of the conical projectile are more practical.It can be more easily forged in larger sizes and is amenable to heattreatment. The dart and star-type flukes have too much variation incross-sectional thicknesses to permit heat treatment to the hardnessesrequired.

White Sands Missile Range Tests

Additional scale-model tests were conducted in 1978 using only theconical shaped projectile to establish a data base for developing aholding capacity equation. The tests were conducted by Sandia Laboratoriesat the White Sands Missile Range (WSMR) in outcrops of granite, basalt,limestone, sandstone, and shale. Th., same scale tock projectile wasused as in the NWC tests; however, a Larger piston was used because theSandia launcher had a larger bore than the launcher used at NWC (seeFigure 19).

Five rock flukes were shot into each of the five types of rockusing Sandia Laboratories' pneumatic launcher (Figure 20). The flukeswere then extracted using a calibrated hydraulic dck (Figure 21).Tests were made in direct uplift with displacement measured and plottedversus applied load. Core samples were taken at each test site andtested to determine properties. The properties of the rock at each siteare given in Table 9.

Twenty-two of the 25 shots were embedded. Two of the three whichdid not penetrate were machined from bars which were cracked during heattreatment and so partially split upon impact. The third uhsu'rcessfu]fluke was simply "rolled up" when it impacted the limevt.'no target. Sixof the 22 flukes which were embedded sustained damage to the base orwere embedded at an angle which prevented pullout tests. These six werenot extracted until 6 months Iater. To achievc extraction of theseflukes, an attachment. was welded to the fluke's base and then they werejacked out vertical to the rock surface. There is some doubt as to the

8

meaningfulness of this data because the embedded flukes were exposed towetting and drying and freezing and thawing over the six-month period oftheir embedment. Finally they were heated to 315 0C during welding.These actions should reduce the holding capacity of the flukes. rhis,in turn, would make estimates obtained from equations developed from thedata base conservative.

There is one other possible mechanism, creep, which could reducethe holding capacity over a period of time. The effect of this factorwas not explored in these tests.

The results of the WSMR tests are given in Table 10. Figures 22,23, 24, 25, and 26 show typical fluke embedments and the post-extractionappearance of flukes for granite, basalt, limestone, sandstone, andshale, respectively. Note the relative degrees of damage to the targetrocks caused by the embedment. Radial cracking was absent in all casesexcept the limestone. The granite tended to spall and crater on theweathered surface layer. The basalt damage was limited to the areaimpacted by the fluke and the larger diameter piston. The limestone,the most difficult target to penetrate, showed minimal amounts of surfacedamage. Successful embedments in the rock occurred where the rocksurface was closely jointed or the fluke impacted near a joint. Thesandstone was badly damaged by impact; the rock was shattered into12-inch to 3-inch-sized blocks. The shale, which was very brittle andthin-bedded, was extensively cratered and shattered into 3-inch to1/2-inch fragments. It was impossible to prepare a suitable compressiontest specimen from the shale core sample.

Analysis of White Sands Missile Range Data

If the tests in which the fluke failed to embed and the tests inwhich no effective holding capacity was developed are termed "unsuccessful,"we can say that 100% of the granite tests, 80% of the basalt tests, 60%of the limestone tests, and 40% of the sandstone tests were successful."Success" in this case being an embedded fluke which provided somemeasurable holding capacity. Table 11 lists these successful tests andtLeir measured holding capacities. It also gives the mean holding (H)capacity and standard deviation (o)for each rock type and for the collec-tive data.

It can be observed from Table 1] that the holding capacity one mayreasonably expect by shooting the rock fluke into an unknown type ofrock is highly variable. The mean of all the listed tests is 54.97 kM.The standard deviation is 49.2 kN. One could say with 83% certaintythat the anchor holding capacity would exceed 5.77 kN. The accuracy canbe improved if one knows the test was conducted in granite, basalt,limestone, sandstone, or shale.

The holding capacities even in a single rock type show fairly widevariations. Properties of the rock types were obtained from a singlecore sample taken in the rock outcrop, with the test shots being made inthe same outcrop usually within 15 feet of the core. The granite targetwas an exception. Tests 1-1, 1-2, and 1-3 were performed in the samegranite outcrop but 70 to 80 feet away from the part that was cored.Although the material properties of the various rocks a~e probably thesame within such short distances from the core, the properties of the

9

rock mass must change enough to account for the variation in the testresults. For example, the material properties of the rock at two sitescould be identical, but if one test is conducted near a joint and theother in a fairly intact portion of the outcrop, the test results wouldbe quite different.

la an a(i:,al emplacement of the CEL-type embedment anchors, the

anchors are fired upon touchdown. No provision is made for precisepositioning of the anchor before firing. Thus, a similar variation inpossible holding capacities should be expected.

Another source of the variation of holding capacities within onorock type is the deformation to the fluke. The amount of damage to thefluke in each test is described in T;,lIe 10. It was impossible toassess the influence such deformation has on the fluke's holding capa-cities, so it was ignore-l as an individual factor.

Development of Empirical Holding Capacity Equation for Rock Seafloors

Simple regressions of each target property versus holding capacitywere performed as a first attempt at predicting anchor holding capacities.A computer program was used which evaluates the statistics of eightdifferent forms of equations the simple regression may assume. Theprogram then selects which equation best fits the data based upon thestandard error of estimate and correlation coefficient.

The best evidence of existing relations was found between holdingcapacity (H) and Poisson's ratio (p), H and Youngs modulus (E), and Hand the rock quality designation (R') in decreasing order of correlation.The correlation with unconfined compressive strength (a), unit weight(p), and porosity (p) were low. The simple regression procedure wasalso applied individually to the maximum embedded diameter of the fluke(D), the embedded length (L), and the embedded surface area (A ). Allof these had a low correlation to the holding capacity. TableS12 piesentsthe results of the simple regressions.

Use of any of the equations listed in Table 12 for p, E, or R

results in a lower standard error of estimate than the standard deviationfor all tests given in Table 11. This means that knowing any one ofthese properties and applying the appropriate regression equation allowsone to make a more confident prediction of the anchor's holding capacity.

These equations will also allow some interpolation since the propertiesof targets other than the rocks tested may be substituted into theequation.

Further improvement in the accuracy of predictions may be possibleby combining various independent variables in one function. The predic-tion method should also have some factor or physical fluke dimensionincluded to allow extrapolation to larger size rock flukes. Trialequations were developed using the multiple linear regression method ofTolson (1970). The earlier trials are based upon results of polynomialregressions conducted on the relationship of individual independentvariables to holding capacity. The various powers used are based uponthose results. The combinations of variables used were based uponresults from earlier trials.

Table 13 contains the results of the trials. Note that some of theequations were developed from a set of 14 data points (White Sandstests) and some from 16 data points (White Sands tests plus two data

10

points from the NWC, China Lake tests). The NWC data and the WSMR shaledata could riot be used in some trials because of a lack of data on someof the variables. Trials developed from 16 data points which appearedsignificant were also developed using 14 data points to afford comparisonwith other trials.

Indepelett variables used to develop the trial equations are:

) = maximum embedded diameter of the fluke, cm

L = maximum embedded length of the fluke, cm

a unconfined compressive strength of target, MPa

p = unit weight of target, g/cm3

E = Young's modulus of target, GPa

p Poisson's ratio for target

R = Rock Quality Designation for target

p = porosity of target material, percent

z = depth of crater in target following embedment, cm

m = projectile mass, kg

v = projectile velocity, m/s

Trials number 1 through 11 explored equations that used only theindependent variables which were properties of the target rock. Most ofthese trial equations reduce the range of the 83% confidence intervalgiven in Table 11 for all the data. They have two failings, however:(1) they do not contain any variables which account for the variabilitybetween holding capacity in the individual targets, and (2) they do notcontain any variables which allow the projectile size to be varied.

Trials 12 through 30 use various powers of the maximum embeddeddiameter, D, and the embedded length, L, with the variables describingtarget properties. Trials 31 through 34 use the projectile mass, m, andimpact velocity, v, as the projectile descriptive variables. D and Lare effective in correcting the two failings mentioned above; m and v,however, are the same in almost all of the tests and so do not improvethe predictions of the earlier trials.

Development of trial equations 35-40 followed a different path thanthe previous trials. A trial equation was developed from a theoreticalviewpoint. It was assumed that the holding capacity is a function ofthe residual stress in the rock acting on the embedded surface area ofthe fluke. Dividing the modulus, E, of the target rock by the Poisson'sratio, p, for the rock will give the stress per unit lateral strain forelastic conditions. The amount of lateral strain induced would in turnbe a function of the volume of rock displaced by the fluke.

Holding capacity, H = f (embedded surface area x residual stress)

11

If the fluke is idealized as a simple right circular cone,

H k 12 L2 1/2 [ -tED L

k1E 1 D3 L + L2) 1/2

where kip k' = constants

E = Young's modulus, GPa

p = Poisson's ratio

D = maximum embedded diameter, cm

L = embedded length, cm

This equation can be expressed in a form amenable to the multipleregression methods outlined earlier by taking the log of both sides

logHog L + log _+ L

The preceding considers only elastic conditions. The embedment innatural rock targets is not strictly elastic. Targets which are jointed,fractured, or in which porosity is significant are less likely to developresidual stresses. This effect was included in the above equation bymultiplying by the rock quality designation R' (trial 36). This is thepercent length of a core sample which is composed of sections longerthan four inches. Thus rock with a large number of joints will have alow RQD value. The RQD does not take into effect porosity, p, of therock which also tends to prevent the development of residual stresses.So 1-p was used to factor in this influence in trial 37.

One other parameter was tested in an attempt to combine the effectsof porosity and fracturing. This is the depth of the crater, z, createdby impact (trial 38). The volume of the crater would describe the rockcondition better; however, the only measurements of the crater takenwere depth.

Trials 41 and 42 were based upon the earlier findings of polynomialregressions performed on the relationship of embedded surface area toholding capacity. The best fit was a third order equation. Cubing theexpression for volume and then adding the target properties which deter-mine state of stress results in the combination of variables shown.

Equation 41 had the highest correlation of those tried. The standarderror of estimate is quite high compared to others. The equation withthe lowest standard error of estimate was number 8. The trend seemed tobe that the simpler the equation, the lower the standard error of estimate.The more complex the equation, the higher the correlation coefficient.In fact, none of the trial equations had an error of estimate lower thanthat found in the simple regression between Poisson's ratio, p, andholding capacity.

12

tp

Because the best simple regression was found between Poisson'sratio and holding capacity, the regression was repeated after normalizingthe holding capacities by the kinetic energy of the flukes. This changesthe coefficients in the equation, but not the statistical evaluation.

The trial equation with the best fit to the test data with respectto minimizing the standard error of estimate would be then

2 ±MV2 * P * 1.96 x 10-3 (43)

This equation makes allowance for use with larger size projectilesthrough the kinetic energy term. The trial equation with the loweststandard error of estimate which does include a variable relating toprojectile size is Equation 15:

H = 43.01 - 195.18 D + 21.42 L - 1.00 a + 142.36 p (15)

Equation 41 results in the highest correlation coefficient followedclosely by Equations 26 and 28.

H = -27,799.65 + 1.40 D6 - 935.67 D 2 _ 0.40 L3 + 602.48 L

- 18.25 E + 1,631.57 + + 12,929.01 (-)+ 24 ,54 7.8 6 (1-p) (41)

A 20K embedment anchor rock fluke has been designed and is shown inFigure 27. This fluke will be tested in Fiscal Year 80. Equations 15,41, and 43 were applied to see what holding capacities would be predictedfor this larger fluke. Full embedment of the 20K rock fluke to theprotective collar was assumed. Equations 15 and 41 extrapolated verypoorly to this anchor size. Equation 43 gave results that seem in theright order of magnitude (1.78 MN to 0.250 MN (400 to 56 kips) as therock was varied from granite to sandstone (Table 14)).

Equations 26, 32, and 34 were also tried. Equation 26 gave predic-tions that seem too high and that did not vary with rock type. Equation 34gave predictions that seem too low (94.1 to 4.02 kN). Results fromEquation 32 seem of the right magnitude, but as with Equation 26, thepredictions varied little as rock type changed. This lack of variationseems to be a weakness with Equation 32 because the WSMR tests resultsseemed to be quite strongly influenced by rock properties.

The risks of extrapolating any of the equations beyond the as yetnarrow data base are apparent. At present it seems that Equation 43 ora similar equation offers the best hope for developing a general equationto predict holding capacity in rock. The form of Equation 43, where thefluke's kinetic energy has been normalized into the holding capacity,needs to be tried with other rock parameters, combinations of rockparameters, and with the inclusion of different anchor sizes.

13

k -r -• ,,,

SUMMARY

A review of available information on propellant-embedded anchortests and installations in coral seafloors revealed that with few excep-tions, insufficient data were gathered to allow analysis of results orto develop predictive procedures. Therefore, to develop holding capacity

predictive procedures, a series of tests was conducted in coral atBarbers Point, Hawaii, using the CEL 10K propellant-embedded anchor;anchor, coral, embedment, and anchor pull parameters were measured.Also, a number of CEL 20K anchors that had been previously installed incoral at BARSTUR, Hawaii, were pull tested. In addition to these full-scale tests, two model-scale tests were conducted in a chialky limestone.Most of the pull tests ended in mechanical component failures ratherthan fluke extractions. A series of simple regression analyses of thevarious parameters measured indicated that the most important factorscontrolling holding capacity were fluke mass and velocity. Using onlythe tests where flukes were extracted, an equation foi predicting holdingcapacity was developed as a function of kinetic energy.

Considerably fewer data are available on propellant-embedded anchorstested in rock seafloors than coral seafloors. Success has been mixed,with performance ranging from a failure of the fluke to embed to capa-cities exceeding the mechanical limits of the connective components anddownhaul cables. Work on holding capacity in rock began with a seriesof model tests at NWC, China Lake, designed to study the ballistic andfluke characteristics that affect penetration. Seven different flukeswere tested in a variety of rock types. A conical shaped fluke gave themost consistent performance in these tests. This fluke shape had lesstendency to produce radial cracks in the rock which were observed toreduce holding capacity. Also, this shape has a uniformly increasingfrontal area entering the rock surface which continually compresses the

surrounding rock. As a result, this shape was observed to have largersurface areas covered with a comminuted rock coating than the ,thershapes. This coating is thought to contribute to frictional resistanaceto pullout.

Next, a series of model-scale 'I:sts were conducted :t the WhiteSands Missile Range using the onical fluke Shape in five rock types:basalt, limestone, granite, sandstone, and :hale. Ballisti, and flukeparameters were held constant. Penetration, rock fractaring, and holdingcapacity were observed to be dependent on rock tyriy.

With the data from these tests as a base, trial predictive equationsusing selected anchor and rock parameters were tested, against the database. Some equations that resulted in a good fit to the data did notinclude fluke parameters that would alLow for extrapolating the equatiornbeyond the existing data base. A trial equation in which the holdingcapacities were normalized by the kinetic energy of the flukes offeredboth a best fit to the data and a means of extrapolating beyond the database.

CONCLUSIONS

1. The best equation for predicting the holding capacity of propellant-embedded anchors in coral is:

14

H = 0.024[- m v

where H = holding capacity, kN

m = projectile mass, kg

v = projectile velocity, m/sec

This equation is satisfactory for predicting the holding capacity of allthe CEL propellant-embedded anchors. The single-sided lower 90% confi-dence limit is 45.3% of the predicted value. The use of this equationis limited to the range of parameters included in the data from which itwas developed. It should not be applied to fluke shapes other than theCEL plate-like coral fluke. It should be applied with caution to instal-lations in coral where properties are outside the limits of the database. This equation could be improved by including the effect of coralstrength.

2. An equation of the form

1 2H = 7 m v [f(p, a, p, ... )]

offers the best possibility of developing a general equation for predictingholding capacity in rock. An equation of this form will allow extrapo-lation to other anchor sizes. No trial equation gave suitable predictionsfor anchor sizes beyond the model scale used to acquire the data base.Data from full-scale 10K, 20K, OOK, and 300K propellant-embedded anchorswill be necessary to develop a satisfactory equation.

3. Variations within each rock type tested had a large influence onperformance. Of the 25 model tests performed in rock at White SandsMissile Range, only three flukes failed to embed, but three othersprovided minimal resistance to pullout. In each rock type where a poorresult was observed, good results were also observed. In actual opera-tional scenarios, provisions will be needed to accommodate an expectedpercentage of failures.

RECOMMENDATIONS

1. The equation for predicting holding capacity could be improved byincluding the effect of coral strength. It is recommended that such animprovement be made by:

a. Collecting more data as a part of every future installation

b. Using downhaul cables and mechanical connections strong enoughto achieve pullout during testing

c. Thoroughly documenting test parameters

2. It is recommended that data be gathered from full-scale anchor testsin rock to provide an expanded data base for developing an equation topredict holding capacity in rock. These tests should:

15

a. Include as many anchor sizes as possible

b. Include tests where a given anchor is operated at different

kinetic energies

c. Be done in several rock types

d. Use downhauls and mechanical connections strong enough to

achieve pullout

e. Thoroughly document test parameters

REFERENCES

Beard, R. M., and J. F. Wadsworth (1976). Research and developmentprogram for anchoring in rock and coral seafloors. Civil EngineeringLaboratory, Technical Memorandum M-42-76-8. Port Hueneme, Calif., May1976.

Bradley, W. D. (1963). Field tests to determine the holding capacity ofexplosive embedment anchors. Naval Ordnance Laboratory, NOLTR 63-117.White Oak, Md., Jul 1963.

Ezekiel, M. (1949). Methods of correlation analysis, Second ed. NewYork, N.Y., John Wiley and Sons, 1949.

Mayo, H. C. (1973). Explosive embedment anchors for ship mooring. U.S. Army Mobility Equipment Research and Development Center, Report 2078.Fort Belvoir, Va., Nov 1973.

Smith, J. E. (1971). Explosive anchor for salvage operations - Progressand status. Naval Civil Engineering Laboratory, Technical Note N-1186.Port Hueneme, Calif., Oct 1971.

Taylor, R. J. (1976). CEL 20K propellant-actuated anchor. CivilEngineering Laboratory, Technical Report R-837. Port Hueneme, Calif.,Mar 1976.

Tolson, W. E. (1970). A study of the vertical withdrawal resistance ofprojectile anchors. Thesis for Master of Science in Civil Engineering,Texas A&M University. College Station, Tex., May 1970.

True, D. G. (1977). Use of the CEL 100K propellant anchor to providepulling reactions for toppling the Argus Island tower. Civil EngineeringLaboratory, Technical Memorandum M-42-77-1. Port Hueneme, Calif., Apr1977.

True, D. G., and R. J. Taylor (1976). The CEL 100K propellant actuatedanchor - Utilization for tanker moorings in soft coral at Diego Garcia.Civil Engineering Laboratory, Technical Note N-1446. Port Hueneme,Calif., Jul 1976.

USS FRANK KNOX (DDR-742) Stranding Salvage. Naval Ship Systems Command,NAVSHIPS 0994-002-6010. Washington, D.C., 1968.

16

Valent, P. J. (1973). Design considerations for seafloor foundations onrock. Naval Civil Engineering Laboratory, Technical Note N-1281. PortHueneme, Calif., Jun 1973.

Volk, W. (1958). Applied statistics for engineers. New York, N.Y.,McGraw-Hill Book Co., 1958.

Wadsworth, J. F. (1976). Anchoring in rock - A preliminary study.Civil Engineering Laboratory, Technical Memorandum M-42-76-5. PortHueneme, Calif., Apr 1976.

Wadsworth, J. F., and R. J. Taylor (1976). CEL 10K propellant-actuatedanchor. Civil Engineering Laboratory, Technical Note N-1441. PortHueneme, Calif., Jun 1976.

Young, C. W. (1967). The development of empirical equations for pre-dicting depth of an earth-penetrating projectile. Sandia Laboratories,Report no. SC-DR-67-60. Albuquerque, N.M., May 1967.

17

p.-, L

'o~~ ~ ~ 'o 00 10'C. cLo

01 01' 0000

.I .I .I .I . . ....

C 0 0 .

c CC 0 0 av 0

o o o*o0 00 0 0 0 0a

LI 0u E*~ r r0 cI r I

o 0 ol~

LII0eL 00 00 LIL 0 LIa

It co 00 000 0 0 0 0000C

0 000 Cr e 40 0 000

'o -0 -0 - -

.0 o o o

u 0 0

e4 r: > o>o 000 > > >

L~L7<~ <-

~018

6-- Ak~d

'0 r c~

- 0 -0 -0 o -0 o

- m CL CL c. CL

D -'

r0

t ~ ~ - z- . owoo 0 o o o -t 0 oo o c4 4 'D

0 o000000 o00C 0 0o00oo 0

-3r..

o~~ ~ ~ Cfl 0 0 .tZ 6 .a

I ' rC P ~ r rJjr -1- 1- j j j - 4. 3 j I - 3 - 1. j j j j

.Q w L Q Q w u 4z ;w w w w w - j 1E0 uCu>-u..-.U U ) QU v UU Uw -+ + +

C~ '9 C?. 00 00 r00 '9000 T.--7

- - C9

jiC

E E E

CL 0. CL

0~

U ~ 0 0 0 00 00 0 0

C 0

+ +~0 ~ 0

t4U

40um

000 000 0 00 00 0

uU u u u uu

~e0l 71 UJUW LWW

uUU U QU.

02

u0I

r. i05

ma 00 0'.

0 0 0 0 00 0

0 0' 0 0 0

-4

oa *- 00 0

m~ U

eGr. O 0.- V)

-4 )0 '000000'' '' -

u 00U

m4 %0 0 0 N. m(U

C)00 m 0 ~ C 0 ~ C ~ V'

C3 a00-4 '.4

0 w(4. 04 0 10I 01 o1 00 00

0 wU

u.4 > CA

4)00 'o0000 0 4 O n r

0 0 aa0

Ga (

4)00

.0

M 12 In(U0 O

E-~ -

~ Ga U ~ 21

'V ~0 m 10 0 0

0V 0.0

0 0 diw r 00 0 tl

=~0 0 0 0 'V

"a "o a~- 0

0

0 c

W 4 4) C CL

0 (n0)'

0)0

00 >1 "~0

40 - Cc 0'00 'a -nL). 4 C-4 C1 qrsr ru j

0

000 m MJ,

v)0 4- a) ~ ' U 0co> L 0 0)I4- a)C t t'

4-. 4)

0>

0 Q)

N c7 r--+ 1-- 0

. 41 -4'~ 1 '

.59 00. mmm

u- 04 000

k 4 00 U L)O oC w

(U 0-4 C,4 C4 N 04 N

0)~ 4 . 0)4 4 4 .

L) Q- L)U

Qj 0 'E 0)

Table 4. Data Available on Performance of Embedment Anchors in Coral

Embedded

Type of Fluke Projectile Muzzle Arontal Angle Surface Target, ou TargetTest No. Anchor Test Location Type Mass, M Velocity. V Area, I l.ading. Area, A (MI'a)(kg) (mIs) (cm2 ) (deg) (n 2 ) (g/cm3 )

1 SEASTAPLE Key West MK V unknown unknown unknown 0 0.1277 unknown unknown2a

SEASTAPLE Key West MK V unknown unknown unknown 0 0.1277 unknown unknown

,aMERDC Key West Sediment 102 107 73 0 1.3468 unknown unknown4a XM-50 Key West Sediment 102 107 73 0 1.3468 unknown unknown

1 MERDC Key West Sediment 408 138 271 11 4.3215 unknown unknown

2 XM-200 Key West Sediment 408 125 271 11 4.3215 unknown unknown3 XM-200 Key West Sediment 408 138 271 11 4.3215 unknown unknown4 XM-200 Key West Sediment 408 97 271 45 4.3215 unknown unknown7 XM-200 Key West Sediment 408 97 271 0 4.3215 unknown unknown12 SUPSALV Key West Coral 1 815 46 892 >80 4.3299 not measured unknown

1 3a SUPSALV Key West Coral 2 815 46 892 >80 4.3299 not measured unknown

14a SUPSALV Key West Coral 2 815 46 892 >80 4.3299 not measured unknown22' SUPSALV Barbers Point Coral 2 815 73 892 >80 4.3299 10.4-17.3 unknown23 SUPSALV Barbers Point Coral 2 815 85 892 >80 4.3299 10.4-27.3 unknown24 SUPSALV Barbers Point Coral 2 815 85 892 >80 4.3299 10.4-17.3 unknown25 SUPSALV Barbers Point Coral 2 815 85 892 >80 4.3299 10.4-17.3 unknown

1a CEL 1OK Midway Island Sand 67 120 -100 -70 0.4167 unknown unknown

2 CEL 1OK Midway Island Mud 78 116 -100 -70 0.7327 unknown unknown3 CEL 20K Midway Island Sand 136 120 -150 -70 0.9985 unknown unknowne CEL 20K Midway Island Sand 136 120 -150 -70 0.9985 unknown unknown

5 CEL 20K Midway Island Sand 136 120 ~150 -70 0.9985 unknown unknown

P-I CEL lOOK Diego Garcia Coral 3 725 133 unknown 0 3.4146 unknown unknownP-2 CEL 1OOK Diego Garcia Coral 3 725 131 unknown 0 3.4146 unknown unknownP-3 CEL lOOK Diego Garcia Coral 3 725 127 unknown 0 3.4146 unknown unknownP-4 CEL 1OOK Diego Garcia Coral 3 725 127 unknown 0 3.4146 unknown unknownP-5 CEL 1OOK Diego Garcia Coral 3 725 120 unknown 0 3.4146 unknown unknownP-6 CEL 1OOK Diego Garcia Coral 3 725 120 unknown 0 3.4146 unknown unknownP-7 CEL 100K Diego Garcia Coral 3 725 - unknown 0 3.4146 unknown unknownP-8 CEL 1OOK Diego Garcia Coral 3 725 130 unknown 0 3.4146 unknown unknownP-9 CEL 100K Diego Garcia Coral 3 725 128 unknown 0 3.4146 unknown unknownP-10 CEL 1OOK Diego Garcia Coral 3 725 - unknown 0 3.4146 unknown unknownT-1 CEL 1ooK Diego Garcia Coral 3 725 125 unknown 0 3.4146 unknown unknownT-2 CEL 1OOK Diego Garcia Sand/Clay 884 102 unknown 0 5.2077 unknown unknownT-3 CEL looK Diego Garcia Coral 3 725 125 unknown 0 3.4146 unknown unknownT-4 CEL 1OOK Diego Garcia Sand/Clay 884 108 unknown 0 5.2077 unknown unknownT-5 CEI. 1OOK Diego Garcia Sand/Clay 884 110 unknown 0 5.2077 unknown unknownT-6 CEL 100K Diego Garcia Sand/Clay 884 110 unknown 0 5.2077 unknown unknownT-7 CEL lOOK Diego Garcia Sand/Clay 884 107 unknown 0 5.2077 unknown unknownT-8 CEL 1OOK Diego Garcia Sand/Clay 884 102 unknown 0 5.2077 unknown unknown

A

Data Available on Performance of Embedment Anchors in Coral

kzzle Frontal Angle of Embedded Target holdingiEV. V Area. F Loading, 0 Surface Target, ou Density, p Capacity, 1i Remark%

(-/S) (CM2 ) (deg Area, A (MPa) (g/cm 3

(deg) (m2 ) (kN)

known unknown 0 0.1277 unknown unknown 3.12+ Mar 1961; downhaul failedown unknown 0 0.1277 unknown unknown 2.23 Mar 1961; pulled out

107 73 0 1.3468 unknown unknown 289.3 May 1963; pulled out107 73 0 1.3468 unknown unknown 356 May 1963; pulled out138 271 11 4.3215 unknown unknown 645.3+ Mar 1963; keying flap failed; pulled out

125 271 11 4.3215 unknown unknown 267+ May 1963; keying flap failed; pulled out138 271 11 4.3215 unknown unknown 445+ May 1963; keying flap failed; pulled out

97 271 45 4.3215 unknown unknown 956.8+ May 1963; keying flap failed; pulled out97 271 0 4.3215 unknown unknown 979+ May 1963; keying flap failed; pulled out

46 892 >80 4.3299 not measured unknown 302.6 May 1968; pulled out46 892 >80 4.3299 not measured unknown 534 Jul 1968; pulled out46 892 >80 4.3299 not measured unknown 605 Jul 1968; pulled out73 892 >80 4.3299 10.4-17.3 unknown 289.3 Apr 1969; pulled out85 892 >80 4.3299 10.4-17.3 unknown 334+ Apr 1969; unable to extract85 892 >80 4.3299 10.4-17.3 unknown 334+ Apr 1969; unable to extract85 892 >80 4.3299 10.4-17.3 unknown 668+ Apr 1969; unable to extract

120 -100 -70 0.4167 unknown unknown not measured Sep 1975; pulled out116 100 -70 0.7327 unknown unknown not measured Sep 1975; left embedded120 -150 -70 0.9985 unknown unknown not measured Sep 1975; left embedded120 -150 -70 0.9985 unknown unknown not measured Sep 1975; pulled out120 -150 -70 0.9985 unknown unknown not measured Sep 1975; left embedded

133 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded131 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded127 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded127 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded120 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded120 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded- unknown 0 3.4146 unknown unknown 668+ May 1975; proof test; left embedded

130 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded28 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded- unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded25 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded02 unknown 0 5.2077 unknown unknown 734+ May 1975; proof test; left embedded25 unknown 0 3.4146 unknown unknown 490+ May 1975; proof test; left embedded08 unknown 0 5.2077 unknown unknown 490+ May 1975; proof test; left embedded10 unknown 0 5.2077 unknown unknown 490+ May 1975; proof test; left embedded10 unknown 0 5.2077 unknown unknown 490+ May 1975; proof test; left embedded07 unknown 0 5.2077 unknown unknown 490+ May 1975; proof test; left embedded02 unknown 0 5.2077 unknown unknown 490+ May 1975; proof test; left embedded

continued

23

Table 4. Continued

Projectile Muzzle Froptal Angle of Target-1p fFluke Surface 'Iarget, o urest No. of rest Location Mass, M Velocity, V Area, F Loading, 0 Densiry, p

Anchor ype (kg) (m/s) (cm2 ) (deg) Area, A (MPa) (g/cm3)

(in2)

CIL lOOK Argus Island Coral 1 - unknown 72 5.2077 35-70b unknown

MOD2 CEI lOOK Argus Island Coral 1 - unknown 72 5 2077 35-70 unknown

MO/)3 CEL 100K Argus Island Coral 4 680 unknown 72 2.9193 35-70 unknown4 CEL 1OOK Argus Island Coral 4 657 134 unknown 84 2.9193 35-70 unknown5 CEL 11K Argus Island Coral 4 657 134 unknown 84 2.9193 35-70 unknown6 CE ,L lOOK Argus Island Coral 3 661 133 unknown 84 3.4146 35-7o unknown7 CL 10K Argus Island Coral 4 680 130 unknown 84 2.9193 35-70 uniknown

M-IA CIL 20K BARSTUR Coral 127 122 153 80 1.1288 23.8 2.59M-IB CEL 20K BARSTUR Coral 127 122 153 80 1.288 23.8 2 59M-2A CEL 20K BARSTUR Coral 127 122 153 80 1.1288 23.8 2 59M-2B CFL 20K BARSTUR (:oral 127 122 153 80 1.1288 23,8 2 59M-3A CiL 20K BARSTUR Coral 127 122 153 45 1.1288 23.8 2.59M-3B CEL 20K BARSTUR oral 127 122 153 45 11288 23 8 2 59M-4A CEL 20K BARSTUR Coral 127 122 153 0 1.1288 23.8 2.59M-4B CEL 20K BARSTUR Coral 127 122 153 0 1.1288 23.8 2.59M-SA CEL 20K BARSTUR Coral 127 122 153 0 1.1288 23.8 2.59M-5B CEI. 20K BARSTUR Coral 127 122 153 20 1.1288 23.8 2.59

1 CEL 1OK Barbers Point Coral 52 172 85 68 0.6437 291 1.982 CEL 1OK Barbers Point Coral 52 172 85 68 0.7695 29.1 1.983 CEL 1OK Barbers Point Coral 52 172 85 68 0.7695 29.1 1.)84 CEL 1OK Barbers Point Coral 52 172 85 68 0.6179 29.1 1.985a

CEL 1OK Barbers Point Coral 52 172 85 68 0.6934 29.1 1.986a

CEL InK Barbers Point Coral 52 172 85 68 0,5308 29.1 1.98

7 CEL IK Barbers Point Coral 52 172 85 0 0.6437 29.1 1.986a

CEL 1OK Barbers Point Coral 52 172 85 0 0,7695 29.1 1.98

9 CEL 1OK Barbers Point Coral 52 172 85 82 0.7695 29.1 1.9810 CF L 10K Barbers Point Coral 52 172 85 82 0.7695 29.1 1.9811 CEL IOK Barbers Point Coral 52 172 85 82 0.7695 29.1 1.9812 CEL 1OK Barbers Point Coral 52 172 85 82 0.7695 29.1 1.98

NWC-28 Scale Model NWC, China Lake Coral 2.3 140 12 0 0.0441 2.2 1.50

NWC-29 a CEL 20K NWC, China Lake Coral 2.3 140 12 0 0.0631 2.2 1.50

aData used to develop the empirical holding capacity equation.bAverage: 52.5 MPa.

24

__

Table 4. Continued

Muzzle Frontal Angle of I'mbedded rarget IoldingVelocity, V Area, F Loading, 0 Surface Target, o u Density, p Capacity, II Remarks

(m/s) (cm 2 ) (deg) Area, A (MPa) (g/cm3) (kN)(m

2)

- unknown 72 5.2077 35-70 b unknown not tested Apr 1976; not tested; left embedded

- unknown 72 5.2077 35-70 unknown not tested Apr 1976; not tested; left embedded

unknown 72 2.9193 35-70 unknown not tested Apr 1976; not tested; left embedded134 unknown 84 2.9193 35-70 unknown 556+ Apr 1976; measured loads; left embedded134 unknown 84 2.9193 35-70 unknown 556+ Apr 1976; measured loads; left embedded133 unknown 84 3.4146 35-70 unknown 556+ Apr 1976; measured loads; left embedded130 unknown 84 2.9193 35-70 unknown 556+ Apr 1976; measured loads; left embedded

122 153 80 1.1288 23.8 2.59 75+ Aug 1976; proof load; left embedded122 153 80 1.1288 23.8 2.59 75+ Aug 1976; proof load; left embedded122 153 80 1.1288 23.8 2.59 75+ Aug 1976; proof load; left embedded122 153 80 1.1288 23.8 2.59 151+ Aug 1976; proof load; left embedded122 153 45 1.1288 23.8 2.59 231+ Aug 1976; tested Aug 1978; downhaul broke122 153 45 1.1288 23.8 2.59 178+ Aug 1976; tested Aug 1978; downhaul broke122 153 0 1.1288 23.8 2.59 178+ Aug 1976; tested Aug 1978; downhaul broke122 153 0 1.1288 23.8 2.59 249+ Aug 1976; tested Aug 1978; downhaul broke122 153 0 1.1288 23.8 2.59 142+ Aug 1976; tested Aug 1978; downhaul broke122 153 20 1.1288 23.8 2.59 205+ Aug 1976; tested Aug 1978; downhaul broke

172 85 68 0.6437 29.1 1.98 156+ Jun 1978; downhaul broke172 85 68 0.7695 29.1 1.98 181+ Jun 1978; downhaul broke172 85 68 0.7695 29.1 1.98 140+ Jun 1978; downhaul broke172 85 68 0.6179 29.1 1.98 174+ Jun 1978; downhaul broke172 85 68 0.6934 29.1 1.98 4.5 Jun 1978; anchor pulled out172 85 68 0.5308 29.1 1.98 162 Jun 1978; anchor pulled out172 85 0 0.6437 29.1 1.98 100+ Jun 1978; downhaul broke172 85 0 0.7695 29.1 1.98 105 Jun 1978; anchor pulled out172 85 82 0.7695 29.1 1.98 223+ Jun 1978; downhaul broke172 85 82 0.7695 29.1 1.98 220+ Jun 1978; downhaul broke172 85 82 0.7695 29.1 1.98 - Jun 1978; downhaul broke; load not measured172 85 82 0.7695 29.1 1.98 131+ Jun 1978; downhaul broke; load not measured140 12 0 0.0441 2.2 1.50 21 Apr 1977; pulled out140 12 0 0.0631 2.2 1.50 16 Apr 1977; pulled out

I-

v 0 0 4 w0 0 - 0 0 0w 4) 0 0 0 0

0 0 0 0 0 0 U u U u

N 0 0

N s 0 .fn In .rl In 0, sd 4 'T . N 'D - .SA4 sO 4 m 0 w a ;0 O -

o 44a 4 0 o o o o o o0 00000l(7

ID0 0 0 0 0 0 0 0 0 0 0 0 0 ' 0 5 sA .'

41 -0 o acL

0,-orV~ o4 0 a Q a 0 o o. o c.o a

00 w

VG m~ m*

jd~~~ NoI N N N a N

V 4) Ja. .a~ Ua

bd W*. 44 or- w.0 mf x0? A'- b4 s A 00 N Nw 0

- -A

f-

Table 6. Predicted Mean Holding Capacity and HoldingCapacity Expected to Be Exceeded 90% of

the Time for Each Anchor

AnchorI tern

CEL 10K CEL 20K CEL lOOK CEL 300K

Mean Holding Capacity (kN) 260 360 1,050 2,200

Holding Capacity Exceeded 115 160 470 1,00090% of the Time (kN)

> >-4) *4 r4) s. 5

M 0F 0) m 0o c 0 U) Q V 0 0 v w 0 w V.54 -S 54 w 0 w 0 0 r. 1 0 C

"Q u 0 V w V) .4 V) 0 0 .-4 q 0 0 0 0 .,-4m0 m 0 V 41 "o e 1 0 0 41

0 sx 0 x r.

N~ 10 r- (n r- w4 N r- r- 0 w4 w4 sI m r --4-4 V1 Q V ) Vct C 4 0 CNJ 4) LA, C . N*. ,-4 MV 4--) --4 4.) 4-) 0 0 (N4

C N-4 0 - -f -4 .-4 0 1-i m IV N N 'i$4 x X54 cc w 54 swd w.

40 . 0 W Lm t) 444)0 *4, - 00 a 0 0 0 -S ON4 4~ 0, C: .c

u w) Ln' Ln' Its \ n M0 4t 4 . 00 Ir

-.4

00

to VOC

V1 V4) r. 0 ui o ) Nq F- 0 LO m% 0 uO 00 ' 4t 9r 1'-E.4 .0 Vj 14 4 1-4 -4 -4 -4 -I 4- N- .- 4 -

41~ 4.) 41) 41. 0 i . 4.) 41) 4.-) 4. 4.) 4.) 4. 4.) 4- 4< ~ .4 H 4 4 .,. 1 " 4 -4 -r- ".4 q. ..4 ~Fq rq. ,4 *.14 .,.1

40 00 0 m0 0 0 00 0 0 00 0 0 m 0.- f w- w- w4 s 5.4 W4 54 W. 54 W- 5.4 w4 w~ w4 w4 4 wto 000 0 00 0000 00000000w000000 w

0 LA 4~s0.S- c.I 1 4.4 N 4 N C4 -4 0N M

C4 M.0 01-- M C-Mv m m " .1 " O ' * 0

U- a 4 1I -4 4 4 4 -4 -4 1- - -0 0

-4 ~-4.0 .,0v A.)0' -to- V) N N - N " C' 0 -4 ") - %

E-4 00to00 Nq N ON NO 0% 47S m0 w 0 -0

' '-' N N4 N* N- N- 04 NC-4 N - 1 '004

V) V 0 V041 1 41 41 w4 V 41) w. w 4.1) 4 ) 41 4 41 4.) > V1 4 Q

054w 0 m . 0l w 1 m 5 W- to w4 w. m w1 w i .wAdt V wc 0 m 4.) cc 40 0 54 w 0 0 kv Q VC

J= = W U2 U4) )4 ..

U) * 4 m It) -* If 0 C'- 00 0% 0 '-N)' fn 4 If) '.Q14 . 4 1 .

27

00 10 *,-4 ,4w0 > 00 00 00V

C4 V 0. W4 -A" 1%" 0 .9 10. 03) cM _.4 41 4) 0) 4) JH u (A 4) .,4 u 4 .. 4 u 4) .14 u-H M : 0 010 41C <0 r I I ro)C '4-)C Vz104-1C m 1 C'Vu 0 0 0 .0 0) w c E4 V CA Wm0( m 4j 4 u 4.) 0.,q U 4)-,-1 U Q) -U 41)

Ix$4o x x .0~ .0) 0

.00

-4 -4 Gi 4) cu w) 0)En~ Nt N 41) 4J V 0 4-) +) .1 V

w w 00 00 a ).4 000) U x4 x U U u Z U U U

~0 ~00

0 41 0 0 00 60 0 N 00

V00

4- U4-

0 ca C4 Ln14Cu

00) 0)

4.1 4J 4.) 4.) 4.) 4-J 4.) 4.) 41)4 41)* ~ ) .14 .,k ,.4 ,..4 -1 .4 -4 41 -

0 CO CO CO CO CO CO CO CO 3 mm0 ~ to cc c0 FA 0 (n [a CO C) ( 0 W0

w) Z -4 w4 S-4 CO Co cO CO CO C O00 04 w 0 ~0 ...0 .0 a

E Cl- Nl 0) 0-4 \

0n 0 C1 4 C14 C N N I

U4.) 4.)1.

mOU -W~ CO Z c'n cn' c" 00 - T '0 COa C

0.0o-,. r- Cl V N c"') N n a--4 8" -" - 0

4J CO. 0 0m o

C4N N N4 N4, N C) N N4 N4 N N0w.

41 0) ) VC0)) 1) - 4.0 4.) )4) 4.)b $4 -6) 0) c0 ) ..

$4 hCC 4J W $COCO( CO CO CO .0 0 mO OCO$H 41d. En) w cm p V) ha4. -) E

0 n ) 4)C' Cn o W

4.)U). r 00 42, 0 -4l N\" ~ i' 0 -

04 ' . N Nl N N N

28

0 uC4 8.4 0

'0L) 4) m $00x 00 $4.

4.0 0

L) 0

4 '0 4.) 0)0 -IV0 .

84j4 In 1. 0

00 0 8

414

00

'0 [a3W0

0 J U " 10

".40 00 0J~ ~>. 94-4 c

4j ~ 4 48U u 00 0 0 c 0co

4) ~ e- 0A0dI : C0 .00 4

00

0 '0 '0 '0

0 00

U00 00 0%00 4N). .

~ N N N 49

Table 8. Properties of Rocks at NWC, China Lake Test Site

Unconfined Strain 50%Sample Rock Density Compressive to TangentNo. Type (g/cm3 ) Strength Failure Modulus

(MPa) (%) (GPa)

1 Basalt 2.324 48.8 1.12 5.02 Basalt 2.305 47.8 1.07 5.13 Basalt 2.271 29.7 0.84 4.14 Basalt 2.140 23.1 0.59 5.55 Granite 2.750 41.2 1.37 3.66 Granite 2.592 50.9 1.66 3.47 Granite 2.520 58.8 1.69 5.38 Chalk 1.504 2.2 1.24 2.0

30

•4

p 00

'0.1 a-11. 1 0 -' TO0

AC i a u .. w~ 2 0 o ..) 4) '-~~- u). a N0Z -

No'C '40 .: 4wC w ). 0 ac U ~ 00 A~

., 'a t "U '0 Z, -, U .0)

.- )0. AnO~.. .. CaU'C ~ ~ ~ O -wC- .U UC) ~ U

C4 . C Ca . ~ 4) O C a * CU 0 1)00 A vU .0 ~ Z A4 . "0

4)0 ~ ~ ~ - 00C ) 1 .U 0'0 u. 4) 40 ' C . U . C U

.CCU1 -) v) C -U 4)

4) U) A .0 U 4 .

a~~~~~~ Q. .- U . A - C U C 0)C

ICC .0 v " 00 *- 0l a *A

up) .v .1 0>, ! 4

vJU ,.U. 4, 1C o. w)) c;~ w aca~ a~ 4nOO MC c;0 aowaU, u = .4

4-)~~ ~~ .4).) - . a U . U) a )

U o- CIJ oC o ~ U o 4 . IC o .U AC

U "o o o

-o

9.44-) a 'C>oo o0

ow00 N"" wNC.f cC auC A.' A)

o C~U'-'~)C) IC~N)C ~ U?'. ~ a

'4- ____ ___ ____ ____ ___ ____ __ z

m m o oo m a o%

oK 1U NN 4 N N 0NN NN

A. . . . . ... . ..4.'' o ' 0 0 0 0 0 0

0- 000- 00 000 00 00D

4)9 1 7c 1

In CC-C4) C C

3'

u I,

0 4, -IOU

.14 4.~ a4 :3 41w

~4) - m. a a v

-a 1 4- . 1 .0 C u 0.1 00 u 0 4j Q ~ U 0 0 . U i0-

0v .. 1U4 la 0 0..1 x.'0J~ At.114 4 IV 0

S4 - -0M 0.- xO ) .Id:..v

-M w 1441 0M. 4004 ) M.:3 '0 00 ..I9. .A.w 0.o 00'o 4,

04,. 4,vw 0 0 C0 00ac I..UX.M4 0 0,~4 I . 0 .

4) ..v3 o =0VII z10 V

16, 4,4A 400 .4

44'00 04

'0 CM.

Ili 00 0

4 .

u on

0-~ 00 " %aco00 v.~ 0 t

0

Ql

0432

0 p 4, v~

w 0 A . 0 4.'0 *I 00,C3464 o .064 a ~ 4 .

-0 v4 040 4 441.0 M-.U.4,

,- '0 A 4

41, 4 o o G C

44.110.

134 41* 000w40 0 .w4azu .uz-

.04''N 4 /

41~v4 0 0 10

0.

4 ~ ~ 3 0*

-0

a co

ao 4

F- 0

w --

N 04 010 :

44 1

00

0

0-4 _ ____3

0 0

00 .0 . 0 0 0 0 . 00A0.0 0.. 0 0 0.0 V Q 0

0 0.I 0- v - 1 J

C2 0 0 : 0 z0 4 v

0In a0 w 0Q1 0 v0~.0

0 00 0 0 0O.1a 0 0 C' a0- 0000.

1. 4. 0..- -I wO. 4).J '4 -00

0 00 ~ Q 0 = 0 1 04 00

'00 01 w 40 4-.O04 014J0;x0 0 . 0. 0 0 00

V0.0 7400 10 00404

0.0 .

0 00 cl 000000a

2.~0n -oo-a -

~ .. 0' 0 ~ 0 N Nl

w00m 4 0 ~ 0 0 0

t 0

l 00 0 000A 0

mm n .n 0 OnDnI

0 ,- 0, 1o ' e

340

(A 0 4) .4

0 3 04.4 04

-4

W414

0 oJ c 00

44 00). C

0 0J x~

4

44 4.4 014 *- C .0L

4- 4--j '--

"4J

4J >

-0-4 0C 4: -4 ' 000

04. 0)r cU '0) ' 4

1-4 (N-*.4N L

01 4.

w 0 0C-, F-0 4 to) w4

-44

.0 c

0 0 -. tflO'-4 4'.0'~)CC0~ 35

Table 12. Results of Simple Regression Performedon the Independent Variables

StandardIndependent Form of Best Fit A B Error of

Variable Equation Estimate

H = A p 138.78 38.62 0.669

p H = A + B log p 109.35 44.56 41.32 0.679

E H = A + B/E 13.51 436.86 47.79 0.528

R H = A + B/R 0.77 16.61 47.81 0.527

o H = A + B o 81.30 -0.43 51.58 0.399

1-p H = A + B log (l-p) 64.11 86.12 53.68 0.299

p H = A + B/p 146.52 -219.09 54.99 0.211

A H = A + B A 104.89 -0.67 55.63 0.149s s

D H = A + B D 177.51 -31.30 55.69 0.143

L H = A + B L 118.00 -2.74 55.79 0.129

36

-4

0000 0

0 rw Q

,.4 4d4

4.)m cu Lr) 0 m' w0 00 0 w- w- w0 C) 0 0 4r--f -4 4 I.t 11 LO) 'D0 '0 'D %. .0 %Q 0 r'- - r- C- %.0

4444 ...S144 o 0000000( C 00 ;000 0000

0 0 0

u -4

C: 4 > 0 0 (- u'M m" 0 C'- 1- -I? w r- N' m' w- IT 4 N m0 M w 4A ..-.4 a-e.4 4) -7 'T N L'- '0- '00 0 o ' U') '70, CD* -

6) LI')4

4 4- - ' ON 'O 00 a' 00 00 00?- r- c0 m' a a% 00 r- 00

0~-

-4-4 w0 41 r- C') VI) N. N' 0 0 0 C'4) 0-7 .- 4 N' CY) C')U 44 a' m" '.0 '0 a' C -. 0 C4 r- \. T '. 0 ,t0

04.4 d0i C'- o' C"H \0 C"CM~ 0 ' a 00 00 m ' '

00 0 ____=0

'H0 -4 00 -T cn m" m" 0%) 0 ' c) \.D -4 - CY) 0 ITV 44 % 0 %0 m" '.0 \0 '- .0 ' 00 \'0 '0 r- it' if' VI) 00 00

,-4 0 0 40 .4 01.w

Wo 00 u .. 9- 0 U') (Y) c") Cl) 00 "' -Z7 C") IT a' C 00 a' \.O-4 W 0 w w 00 'D0000 0000 C'- I"- -400 Nq in a) 77-

m 0 ea 40 w k.0 (Y) %0 LO' it) LI' at) '. ) It) en \0 %0 %.0 r- r0.0 0

040

4.000 00

) 044 -4

4)U.0

E- 0

N1. N. (4 . 4w) CL C L L C 0 144

00

-4) -

'44) 0.C4 M T t 0 r 0 M 4 C4 C) -r L e

$44 'V -4 04

0 '-4

4 4 -4r4 -

00 0 u

41.4 f 0000n %0- - 00C 0 C7% .- 0 C4 0000M-4.~ rl4 w 0 r 0 '0 r - r- r- r- r-. r- Cl) r- r'- r- r-. '0

24 44 0 0 0C 000 0 0 0 0 0 0 0 0$4 4)0 0

u 1-493 c r -t 10 00 IT -1 00 00% '00 '-400 00 r'- n CS C7% (7s

L) 444J 00 I'- 00 00 ON 00 c0 0% 00 0D 00 C% 0D 00 0% m% ON 0%ON 0 0+1--

0 4

$404J L M~ -0 0% r- r- (N 00 ON " fn C % sdn '0 '0D-v 0o 0% 0 0% - L() 0% 0% t- 0 M% 00 0 0% r- 10 (N C4 00

10 '0 54 8 C4) 0O. o' 00 '-0000 - - C-J 00 (N 0000000%M:s to5 41j '.-4 IT . I -4 It- It7 It -1 L(I -1 -7? L( -t -7 4 -.It IT I

0 -M W0 I0 N. -T 7'00 0 r- r-- 0% e.J 00 00 MU. 4.4 I '00 Go - 00 ur) u2) iIn Le) i -7 r= -7 ifn 'D Ln' 17- IT 0

04- -4

C O1. C0l v7 w iN n (NO '0 0D 1' o- in 0% 0-7 -.7 '0 r- cnC04 14 w. 00 C1O r- en (N (N4 C4 (N 0S r- 0- '0 in Y 00 P'- 00

0Gz w . ui t'- ON 00 '0 '0 '0 %Q '0 '0 '0 I'- 00 '0 NO '0 '0 in

4)

0l 0 0J t) bZ b

4T4

'N IF vW NMV rW0c oO

H4) N- 00 C%0.. O0 C- nILn Or 00 0.

w Z N NN N N N N CN C4. " 0

a- N-J4-4 - 1r- 4

kv- 1

0 cl

'.4 4.'

0 0 00C ) -CA C 0 0 0n 0 00 000

100 O

v 4., q00 2 0 L 4 0 0.) N' 0'.I~ N Ln M-.4LI

mA4 to a C4 M 0 0 0 0t co m0 w t

m W 4 j ) 4 4 11 IT V LI) '.0 - '.0 r- In0J S4 0(A. - 4 I- 0 @ c

4J-k

0 0 in '0 04 eq r- -4 (14 -4 0Cu. 44 ~ 1 '0 '0 00 r- r- r- 00 @3 It 40

0 -)-

w3 u' C-. 4n LA 04 00 MN-4G~) + n mA' '0 -i 4 M C1 N 4 C- 0'

00

N0 00 0 0 -

-4-4 r-4 4-4 -0to 0

.r4 -4 00c 0--4 P-4 .- 4 .- 4 0. 4 0 N

cc to 00 04 1 -D4 L> 0. + .- .. , +- 0 + *+- C

14 4 -4 r-4 -4 0 N

W cn t Cc I c n T cc cc %-.- + -0 4 t1- 0+ 0+ 0+ .- 4 4 ).~-4 .- .- '' -4 N 0 .. +

L a A Ci4 (A c(1 - O aCL 0- *C -(0 .

4) 0 0 C)0 0 0 g~ 004' * 'fPC 4 e~ -~ 00 -4 000 4 DO0 -4 00 - ' 0% ,- " 4 I-'

b 1~ o 1C I0 ~ 0 1 0 :L . N 4--- 4 4 . r4. -4 1-

> >N - N C'4 N L460'- 0 00 ccD-.c 0 be-,. 00 00 0000 -0 14 0-40 0 -4 0 0 -4 00 0 0 (Dto

S U -, -4 -- 4 r-4 4 -4 '- 4 -4 -4 -4

'-4

N0 Ci 4 LA 'P- 3 0 0 - 0N CY)wA ZA m A m m (n en MA c -T 4 44 T

39

44.4

0.4 830 3t0 .00 -. , (A 0 m

H~. 44 C4

4 - = ~.1-44)

0

2) V IT 0 VU0u W 0 N

C4 :.4 cc-4

0 0 0 .

>4) IT Le".4 ,4

0ui~0-4 >, 4.

41 -4

0o0

a = 4)

,4

44 4)0 V

0 NL

4

NY m

4)0 '-

CA

m 4) %0. C1 c

.-4 0o N1 nc5W :f N

_ I

) .4, I- I

'.3 L1~

- uJ~3

'C

00.

C,-

41

Fi gim 2a MAIRDC( 50-K cxplo~tvc .ini h4r.

'4.

/I t

C.

f>k4 I/

I -~

/ V

/

K]A

C

C

V r-J

V CU.CK

J

43

;/20.500 in. dia.

2.88 in-1

1.00 in.bsj s.t0ockso

34.0 in.stc

82.2 in.in

67.250 in.

/ i.stc

2.7 in.

347/n

9 09

Iigure 3ai ( rigimi StITSA I coral flukv (Coral 1)

2.8 ] F14 0 b &sic

1. 000 in.8 stoc 120 b'asic In. sock3 plces1.000 in. stock

9.00 in.

18.25 in. dia. stock 6.'00 in.

3 places 2.00 inl.

Figure ~ 2 3b. lafce sLPA. (r1flk(Cri2)

4If

00

Ir 0

It"-

do

r4 .

e--

m *~ A

46 4

1-7/8 in- 4- 1-7/8 i1- /8 in A -.

1-7/8 in

1-1/8 in 2tin+0 TYP 10 in.-t2f6in

6 in. i 6in. 6 in-

I lft.O in,

j1 2 ft -.

12 .03A in. 1i in TY PI t6~ift ~1 n 8-9/16 _

3-3/4 in.- 7-3/4*i

(for hole) 3 i.

2 ft.9-%'

129/1

12 2 ft.6 % in.t% t7-/ n

8-9/16

1-3/4'

I, ft. 3 in.- ff.3 in.

Figure 4b. IOOK coral anchor used at Argus Island (Coral 4).

47

10

Obb

48

1-19/32 in.

1-19/32 in. see note 34~6-% in. y n ft. 0i.

I P i n . I i .R r I + ~ -v -!9/3 2 i. se n o e 3

a in.

5/1I 10.5/16 in.3/1 .43

~43/16hoe6 in. -b~- (ab) 112

'7-% 6 in.9/16 in. R-116i.

12 (for hole) 1-/1 in2 ft. 0-in. nteee/nine4

2 ft.ftin

/161-1/6 in.

', l~~i n,:t Ri

11~ 1iin

in '/ain. in. V2in.6 in. -4--6 in. -.

-t SYM ABT CL

Vigure 6, CFA, O0K voral fluke used at Barbers Poinlt.

49

I~~~~ ~ iik -TAI (k

I Iu oaj wi~ 17 ,k

gill Ibarrcj

Gunl A~seinbltx

079

cl fluke

Hi kt. ,\,sctnlIN ; and flukc

I Inuchownl probe(rcuidv Position)

Figure 7. Schematic tt the CH,. 10K propcllant-actuatcd anchor

50

IT

... .

- &

',m 0 4 . d tik 4..-S A c miigt iiuc - hti(iq

.'ev

44

4A

IIL

mooi~rinig firic float mooirinig liUon

BS = 1 17,000 is.6 in. yvit heic iti.c cqtiakliCf SWitCh

bilock %%/%%wivcl)shacklc (BS =94,0)O 1b1))

"ire roipe %i re roipe liridic,

BSLIK heaive and re~t (if~ tup

75 ft, 7/81 in.

duuu iihaul

coral fluke __

Figurc~~~r I 3.BU . moigIg

NI S IIu~u

W~ 99 I~utZl 3u3pjuo', %06 P2P3pt-uO - _ _______

NI SI~ r~r

% .3~ua3ii3 10 Aouanbaid.

54

N~ $IX~! 3~I~o ~6P!pS-3uo - IEU

I II

%'a3u*MNw3o jo A~uanbaiti

55

NJ I

N-l gi-i Xiimd .w imaIIN~I 6ft I1!l 2-lu3P!JU03 IA06 PJP!s 3UO-

% ljy2.Jfl33' jo Aau-,nb~az

-0o

a ~xoz 133

00. 0310113

NI'- Aw'-CSipo

57E

.5oo in,basic.I 33 -__ places-

1.000 j.,.stock-

M9.0 in.

6.000 in. din.k-

.5830 ini3.0.n

9.42~.0 ini2n.in

58.860 in. .500 R

+ch.m10.x2 5Bin

6.in. 16D. 12.20I. in. lue

0 .0 3- pace

fluke and gun barrel

24.250 in.25.850 ~launch vehicle

420i-

-- 50

I. 25.850

4j/8 2 ft. 3 in.

in.1/in

3 t 2 . 5 -al

ft. I I'tn.5-5/8 in. 5 in. 0 -

-Piston -

Assembly

2.92 311)0o

~sYM ABTin. ~VATf v~-1/8 in. Tn-Sn3I

-1/8 in. _5-5/8 in. 90 -30 ft.

ii. 2 1n. '-Ref 7-% 'in.

%45093xi i 2_ 1.// bolh 01o

14ll 'I %i.90-1/Sin.

c3 in.

f t. Dif .l e t a il A "

3 fttn. 2-3/8 in.

5-5/8 2 ft. 0 in. ,/

we \ der"l"A"

7- 3 in.0 in. SYM ABT

20A

Section GC

Figure 17. CEI, 20K rock fluke.

59

L .

dart serrated dart star colic

0 0 0

stake serrated stake tapered stake coral

Figure 18. Fluke shapes tested at NWC, China Lake.

450 90 "X11,2 0 .1 1 - F r- --- ---- ---------- 1 n.80 n. C' 1.6in. di. 1. 1/64 in. ) dii.

dijL I

L-A - --------- - -- -- -- -- -- -- -- -- ---

- F19.83iin. in.18.2 5 in.

fnuke piston

Figure 19. Rock fluke and piston configuration used in all %VSMR tests.

60

SAO:

Figl c 2 . Slli i.1 -1b r'loris, piciman I'lifihilg cll

p ic ~ l~ li kw iki udl~l 11p k~llll l".silL it*I

7-,

Figurc 22a. Ty\pical c~finbdlflcft of s i1. rock- tlukc i rmc

IFigllr( 221), Condition of fluke~ to~lb)owifg extravicion fromi grainitc.

02

: 7 41

, .3p., -,,

'V '

%oC

qot e

,~'- _

silo

i gurc 24a Apcaac of a tLCCSN Cl fl nlt ill 1I~n"tl(St

ftm%

i) t

"' f

~-7

Figiurc 261). 1 picaI ~iditimti ,f fluke and pi~toii .icr ('xmriction~ tioml shi.

C

4-

-z

C E

-i-I-'

d ~(-4 _______ __________

(4 'aff%

67

- V

DISTRIBUTION LIST

AFB CESCH. Wright-Patterson: 11O Tactical Air Cmd/DEMM (Schmidt) Langlev, VA; MAC/DET (Col. 1.Thompson) Scott. IL; Stinfo Library, Offutt NE

ARCrICSUBLAB Code 54, San Diego, CAARMY BMDSC-RE (H1. McClellan) Huntsville ALARMY COASTAL ENGR RSCH CEN Fort Belvoir VA; R. Jachowski. Fort Belvoir VAARMY COE Philadelphia Dist. (LIBRARY) Philadelphia. PAARMY CORPS OF ENGINEERS MRD-Eng. Div.. Omaha NE; Seattle Dist. Library, Seattle WAARMY CRREL A. Kovacs, Hanover NHARMY ENG WATERWAYS EXP STA Library. Vicksburg MSARMY ENGR DIST. Library. Portland ORARMY ENVIRON. HYGIENE AGCY HSE-EW Water Cual Eng Div Aberdeen Prov Grnd MDARMY MATERIALS & MECHANICS RESEARCH CENTER Dr. Lenoc. Watertown MAARMY MOBIL EQUIP R&D COM Mr. Cevasco. Fort Belvoir MDARMY TRANSPORTATION SCHOOL Code ATSPO CD-TE Fort Eustis, VAASST SECRETARY OF THE NAVY Spec. Assist Energy (Leonard). Washington. DC- Spec. Assist

Submarines, Washington DCCINCLANT Civil Engr. Supp. Plans. Ofr Norfolk. VACINCPAC Fac Engrng Div (J44) Makalapa, HICNO Code NOP-964, Washington DC; Code OP 323, Washington DC; Code OP-413 Wash. DC; Code OPNAV

09B24 (H); Code OPNAV 22. Wash DC; Code OPNAV 23. Wash DC; OP987J (J. Boosman), PentagonCOMCBPAC Operations Off. Makalapa HICOMFLEACT, OKINAWA PWO. Kadena. OkinawaCOMNAVBEACHPHIBREFTRAGRU ONE San Diego CACOMNAVMARIANAS Code N4. GuamCOMOCEANSYSPAC SCE. Pearl Harbor HICOMSUBDEVGRUONE Operations Offr. San Diego. CADEFENSE INTELLIGENCE AGENCY DB-4C Washington DCDLSIE Army Logistics Mgt Center, Fort Lee. VADNA STTL. Washington DCDTIC Defense Technical Info Ctr/Alexandria. VADTNSRDC Code 4111 (R. Gierich). Bethesda MDFMFLANT CEC Offr, Norfolk VAHCU ONE CO. Bishops Point, HIMARINE CORPS BASE PWO Camp Lejeune NC; PWO, Camp S. D. Butler. Kawasaki JapanMCAS Facil. Engr. Div. Cherry Point NC; Code S4. Quantico VAMCDEC NSAP REP. Quantico VA; P&S Div Quantico VAMILITARY SEALIFT COMMAND Washington DCNAF PWO Sigonella Sicily; PWO. Atsugi JapanNALF OINC, San Diego. CANARF Equipment Engineering Division (Code 61(X), Pensacola, FLNAS Code 18700. Brunswick ME; Dir. Util. Div., Bermuda; ENS Buchholz, Pensacola, FL; PWD Maint. Div.,

New Orleans, Belle Chasse LA; PWD. Willow Grove PA; PWO Belle Chasse, LA; PWO Key West FL.PWO Whiting Fld. Milton FL; PWO. Glenview IL; SCE Norfolk, VA; SCE, Barbers Point III

NAVACT PWO, London UKNAVAEROSPREGMEDCEN SCE, Pensacola FLNAVCOASTSYSCEN Code 772 (C B Kocsy) Panama City FLNAVCOASTSYSTCTR CO. Panama City FL; Code 715 (J Quirk) Panama City, FL; Code 715 (J. Mittleman)

Panama City, FL; Library Panama City, FLNAVCOMMAREAMSTRSTA PWO. Norfolk VA; SCE Unit I Naples ItalyNAVCOMMSTA Code 401 Nea Makri, Greece; PWO. Exmouth, AustraliaNAVEDTRAPRODEVCEN Technical Library, Pensacola. FLNAVELEXSYSCOM Code PME-124-61, Washington DCNAVENVIRHLTHCEN CO, NAVSTA Norfolk, VANAVEODFAC Code 605, Indian Head MD

68

NAVFAC PWO, Centerville Bch, Ferndale CANAYFAC PWO, Lewes DENAVFACENGCOM Code 043 Alexandia. VA. Code 044 Alexandria. VA: Code WP53 (D. Potter) Alexandria.

VA. Code 1145411 Alexandria, Vii: Codc 04113 Alexandria, VA-, Code 04A1 Alexandria%. VA. Code 100IAlexanittn, VAI (Axle t)l (J. Leliianls) Alexantdrin, VA; ('tide I1113 (M, Carr) Aloxiindrin, VAi Code1113 (T'. Stevenii) Alexandria, VA; Morrison Yap, Caroline 1Is.

NAVFACENGCOM - CHES DIV. Code 407 (D Scheesele) Washington. DC: Code 405 Wash. DC; FPO.I(Kurtz) Washington. DC; FPO-IC (Spencer). Washington. DC; FPO-t Wash. DC

NAVFACENGCOM - LANT DI V. Code 10A. Norfolk V'A, Eur. BR Deputy Dir, Naplcs Italy: EuropeanBranch, New York; RDT&ELO 102. Norfolk VA

NAVFACENGCOM - NORTH DIV. (Boretsky) Philadelphia, PA;, CO. Code 09P (LCDR A.i. Stewart); Code102-8. RDT&ELO. Philadelphia PA; Design Div. (R. Masino), Philadelphia PA: ROICC. Contracts, CraneIN

NAYFACENGCOM - PAC DIV. Code 2011 Pearl Harbor. III; Code 402, RDT&E, Pearl Harbor III:Commander, Pearl Harbor, I-fI

NAVFACENGCOM - SOUTH DIV. Code 90. RDT&ELO. Charleston SCNAVFACENGCOM - WEST DIV. AROICC. Contracts, Twentynine Palms CA: Code 04B San Bruno, CA:

09P/20 San Bruno. CA; RDT&ELO Code 2011 San Bruno, CANAVFACENGCOM CONTRACT' AROICC, Point Mugu CA; Eng Div dir. Southwest Pac, Manila. P1;, 01CC,

Southwest Pac, Manila. P1; ROICC AF Guam:, ROICC. Diego Garcia Island; ROICC. Keflavik, Iceland;ROJCC. Pacific. San Bruno CA

NAVFORCARIB Commander (N42), Puerto RicoNAVNUPWRU MUSE DET Code NPU-30 Port Hlueneme, CANAVOCEANO Library Bay St. Louis. MSNAVOCEANSYSCEN Code 41, San Diego. CA; Code 4473 Bayside Library, San Diego, CA; Code 52 (H.

Talkington) San Diego CA; Code 5204 (J. Stachiw). San Diego. CA; Code 5214 (H. Wheeler). San DiegoCA; Code 5221 (RiJones) San Diego Ca; Code 5311 San Diego. CA: Tech. Library, Code 447

NAVPETRES Director. Washington DC-NAVPGSCOL D. Leipper. Monterey CA: E. Thornton, Monterey CANAVPHIBASE CO. ACB 2 Norfolk. VA; Code S3T. Norfolk VA; Dir. Amphib. Warfare Brd Staff. Norfolk,

VA:. Harbor Clearance Unit Two, Little Creek, VANAVREGMEDCEN Chief of Police, Camp Pendleton CA: SCE (D. Kaye)NAVSEASYSCOM Code SEA QOC Washington. DCNAVSECGRUACT PWO, Adak AKNAVSHIPREPFAC Library, Guam: SCE Subic BayNAVSHIPYD: Code 202.4. Long Beach CA; Code 380, Portsmouth, VA; Code 440 Portsmouth NH; Code 440,

Puget Sound. Bremerton WA; Salvage Supt. Phila.. PA; Tech Library. Vallejo, CANAVSTA CO Naval Station, Mayport FL; CO Roosevelt Roads P.R. Puerto Rico; Code 4, 12 Marine Corps

Dist, Treasure Is.. San Francisco CA; Engr. Dir.. Rota Spain: Long Beach, CA:. PWD (LTJG.P.M.Motolenich), Puerto Rico; PWO Midway Island: PWO. Keflavik Iceland, PWO, Mayport FL; SCE, Guam:SCE, Subic Bay, R.P.; Security Offr. San Francisco. CA

NAVSUPPACT LTJG McGarrah, SEC. Vallejo. CANAVSURFWPNCEN PWO, White Oak. Silver Spring. MDNAVTECHTRACEN SCE, Pensacola FLNAVWPNCEN Code 2636 (W. Bonner), China Lake CANAVWPNSTA Code 092. Colts Neck NJNAVWPNSTA PW Office (Code 09CI) Yorktown, VANAVWPNSTA PWO, Seal Beach CANAVWPNSUPPCEN Code 09 Crane INNCBU 405 OIC. San Diego, CANCBC Code 10 Davisvillc, RI; Code 155, Port Hueneme CA:. Code 156, Port Hueneme, CANCR 20, CommanderNMCB 5, Operations Dept.; 74. CO; Forty, CO-, THREE, Operations Off.NORDA Code 410 Bay St. Louis, MS; Code 440 (Ocean Rsch Off) Bay St. Loui% MSNRL Code 5800 Washington, DC; Code 8441 (R.A. Skop), Washington DC, Code 5843 (F. Rosenthal)

Washington. DCNSC Code 54.1 (Wynne). Norfolk VA

69

NSI) SCE, Subic Bay, R.P.NUSC Code 131 New London, CT. Code EA123 (R.S. Munn), New London CTr; Code 332. B-80 (J. Wilcox)

*New London, CT; Codc TA131 (G. De lit Cruz), New London CTlOCEANAV Mangmt Info Div., Arlington VAOCEANSYSLANT LT A.R. Giancola, Norfolk VAONR Code 485 (Silva) Arlington. VA; Central Regional Office, Boston. MA: Code 481. Bay St. Louis, MS:

Code 7Xff Arlington VA; Dr. A. Laufer, Pasadena CAPH-IBCB I P&E, Coronado, CAPMTC Code 3331 (S. Opatowsky) Point Mugu, CA; EOD Mobile Unit, Point Mugu. CA; Pat. Counsel. Point

Mugu CAPWC CO Norfolk, VA; CO, (Code 101), Oakland, CA; CO, Great Lakes IIL- Code 10, Great Lakes, IL; Code

120, Oakland CA: Code 120C. (Library) San Diego. CA: Code 128. Guam:. Code 154. Great Lakes, IL;Code 2(X), Great Lakes IL; Code 220.1, Norfolk VA; Code 30C. San Diego. CA; Code 4WX, Great Lakes.IL:. Code 4W5. Pearl Harbor, I-11; Code 4(W, San Diego, CA: Code 420. Great Lakes, IL: Code 420J.Oakland, CA: Code 505A (H. Wheeler): Code 7W.) Great Lakes. IL. Code 7W5. San Diego, CA

UCTl TWO 01C. Norfolk, VA; OIC, Port H-ueneme CAUS NAVAL FORCES Korea (ENJ-P&O)USCG G-ECV (C E Smith) Washington, DCUSNA Civil Engr Dept (R. Erchyl) Annapolis MD: Ocean Sys. Eng Dept (Dr. Monney) Annapolis, MD; PWD

Engr. Div. (C. Bradford) Annapolis MD

70