-.i I . Technical Note N-1082 ir material in a more peraanent emplecement. 3. To investigate the holding properties of the indigenous coral send. In addition, the parameters of cone diameter (anchor area), cone angle and jetting nozzle size were vart.ed in order to investigate their effect on the ease of emplacement and the holding power. ~ j 4--FIRE HOSE CONNECTION I i 10 ft. X 2h in. Pipe j JSHEET METAL ANCHOR CONE ONE ANGLE 1 6 to iS In. Dia. - 450 JETTING NOZZLE______ I in. Dia. FIGUR 1. ILLSTRATION OF JETTE ANCHOR iI EMPLACEMENT AND TESTING PROCEDURES Emplacement The twenty-three anchors emplaced during the test program were in a plot 40-ft X 40-ft on a flat bottom in 25 feet of water. The plot was divided into 10-ft X 10-ft sections and stakes were driven at the corners. Thus, 25 anchor sites were laid out within the test plot. The equipment used to emplace the anchors consisted of the fully equipped (SCUBA and shallow water gear) diving barge moored at the site, a standard P-250 Gas Driven Fire Pump, 2 -in. fire hose, hose adaptor, safety strap, signal line, and the anchors themselves. The anchors were jetted by two divers to a depth of 6 to 9 feet into the r'and bottom. The divers were equipped with SCUBA or lightweight surface-supplied life support equipmet according to their own prefer- ence. The SCUBA divers were essentially neutrally buoyant and wore fins, whereas the surface-supplied divers used from 10 to 20 lb. weight belts and wore no fins The jetting procedure is shown in Figure 2. Cement Slurry Injection The purpose of these tests was to determine the effect of a cement slurry injected either beneath the anchor cone or just abbve it in order to bind the anchor more permanently into the surrounding sediment. A cement slurry made of one part Portland cement, one part sand, and one part water was injected on four of the anchors emplaced at the test site. These were 9" in diameter. The equipment used, ii addition to that used to jet the anchor, consisted of a slurry pot (see Figure 3), an L. P. air compressor, hose, assorted pipe fittings and valves. The procedure consisted of mixing the slurry, filling the pot with approximately 16 gallons of mix, sealing it, and injectifig the slurry by pressurizing the pot to 35 psi (24 psi over bottom pressure) with air and opening the gate valve at the bottom of the pot. For 'o of the four anchors slurried, the slurry was injected thrbugh te anchor shaft and out the tip of the cone. The other two were slurried through a sepa- rate probe placing the cement just above the anchor cone. Some problems were encountered during the slurrying operation. The fixst anchor that was slurried through the anchor shaft was done with the .ilurry pot attached to the top of the anchor, then jetted as shown in Figure 4. Although this setup was satisfactory for injecting the slurry$, it* was found to be unwieldy due to 'the bulkiness of the pot, and because its buoyancy changed drastically as the slurry was forced out. This procedure also proved to be very costly in terms of use of diver bottom time. Therefore, the other three anchors were slurried by keep- ing the pot on the diving barge and pumping the slurry to the anchor through a fire hose.f 3 IJETTING OPERATION: Figure 2. AllI anchors were jetted by two divers. A water let was used to emplace the anchors six feet to nine feet into the sediment. Equipm. ent used and jetting sequence is depicted below. firehose to surface =, 1arid P-250 portable adptrsafety fire pump (I ) Divers connect hose and diver'ssafety strap, and erect anchor. air Diver signals surface with signal line. "Start pump" dive (2) 'Water jet starts. Divers start pushing. anchor into (3) Anchor jetted about half- way. Water flowing out of 'hole carries soil and builck mound around hole. Same soil stays in 4 JETTING OPERATION Figure 2. Continued. I (4) Anchor 6' to 9' down. J "~~~ S t o p p u m p " i e (5) Divers disconnect hose and refill hole with soil. (6) Diver vibrates-loosi soil with 10 raps from an 8 lb. hammer. ii 5$ I 11 I I ~1. *1I I II L~ I 4 3 V Figure 3. Slurry pot. II (7 6 4 ___________________ __________ _________ ____________I -S. ';- - V '- - - . . . - -- - - -, 'H - - - - - ~-r~~C------t--~ 2: Use N at d/B 2 q *For loose sand dN<4: Use N at that d/B q d/B>4: UseNq at d/B = 4 For dense sand A' 10: Use N at that d/B d/B >1:UseN atd/I 10 *This value was assumed. H I .-'---.- S 4 - -I 5- - I NI - S I I [ I U -0 Ge. 'A, :1 I IL. I- I I - I 19 -- 9 p ~ -~ ~ ------- -- -~------------~---- _________________________________ lI This modification is necessary because available experimental evidence from experiments on 3" 0 plates suggests that the critical relative depth D/B above which embedded objects should behave as rhallow anchors depends upon the relative density of the soil. Thi. limiting depth increases from D/B 2 for a very loose sand to D/b = 10 for a dense sand (Vesic5). ANCHOR PULLOUT RESULTS The field tests indicated that increasing the cone diameter generally increased the anchor holding capacity. Varyfng the cone angle and nozzle size had no apparent effect on the anchor holding capacity. In addition, increased depth of burial increased the holding power, as would be expected. Those anchors that were jetted with a cement slurry generally had a higher holding capacity than those without the cement slurry. Unslurried Anchors Figures B-i to B-5, Appendix B, depict graphically the holding capa- city versus the depth of embedment for the unslurried anchors. An exami- nation of the figures shows that the anchors seemed to fail by two distinct mechanism. 1. The anchor displaced at almost constant load until break- Iout occurred. In this case, the pullout force required was small. 2. The sustained load which the anchor held inc:eased uni- formly to a maximum and then dropped off uniformly with increasing displacement. In this case, the required pullout load was somewhat higher. These two cases re illustrated in Figure 10. A comparison of this figure with Kalajian's data , Figure 11, which shows results of similar pullout tests in loose (relative density, Dr, < 40) and dense sand (Dr < 80), indicates that there is a definite correspondence between in-situ density and force-displacement curve shape. Therefore, it should be possible .o infer the in-situ relative densities from the shape of the field force-displacement curves. This inference has been made, as no measurements of in-situ density were possible. Data from the anchor tests and corresponding theoretical predictions of anchor holding capacity are presented in Table I. The qualitative descriptions of relat-ve density were inferred from the field force-dis- placement relationships, Equations (1), (2), (3) and the average soil properties determined during the direct shear tests were utilized in making the theo retical predictions indicated. E. . KalaJian and S. N. Imben conducted an investigation of the verti- cal pullout capacity of marine anchors embedded in seand by vibration. He also notes that there appears to be two mechanism cf failure within the soil uass. ~ 2D 0A -c c1 CC <1 G %00 0 C4~0 0l t W 0 N G 21 1 TABLE I ACTUAL AND THEORETICAL HOLDING CAPACITIES - - DEPTH STATIC SITE OF CONE Qmax Qmax Qmax qmaI RELATIVE* NUMBER BURIAL DIAMETER MEASURED CYL TORICAL DEW WT. DENSITY FEET INCHES POUNDS POUNDS POUNDS jtOUNDS S4 7.71 6 1100 993 735 79 Loose 5 8.17 6 600 1046 308 77 V, Loose 14 8.02 6 1000 1079[ 772 83 Loose is 8.04 05 2,300 !2979 I 4820 517 Loose 6 7.94 9 1900 1574 1710 184 Loose 7 7.90 9 1900 1553 1700 183 Loose 13 8.67 9 100 1815 740 185 V. uoose 25 8.63 9 800 1744 736 184 V. Loose 9 7.68 9 2800 1760 8200 200 Dense 11 7.28 9 2000 1600 7800 190 Dense 12 8.71 9 3300 2248 9120 228 Dense 16 7.67 9 900 1413 650 163 V. Loose 19** 8.54 9 3700 - - - 20** 8.50 9 7400 .. .. 23*** 8.96 9 2800 .. .. 24*** 8.75 9 4000 - 6.29 12 1800 1489 2410 259 Loose 6.88 12 1200 1743 2630 283 Loose 15 8.15 12 900 2249 I 1236 309 V. Loose 17 7.75 12 3300 2570 110,000 360 Dense 8 8.54 12 1600 2621 3264 351 Loose 10 8.00 12 800 2073 1210 303 V. Loose *Relative density inferrtd from general shape of holding power vs displacement curvr for each anchor I'.] **Cement slurry used under cone ***Cemnt slurry used over cone 22 41' Values of Ko used in the calculations were assumed as follows: Very loose K0 - .56 Loose K - .52 0 Dense K - .43 0 Explanation of Data From Table I, it is readily apparent that the dead weight approach to predict holding capacity is very conservative. This occurs because the soil is assumed to exhibit no shear strength which is in contradic- tion to the results obtained from the direct shear tests previously presented. Results of the Torical and Cylindrical analyses are also presented in Figure 12. Actual holding capacity is plotted against theoretical holding capacity. Ideally, the data should fall on a 450 line through the origin but it is clearly evident that this is not the ca&e. The results of the Torical solution for dense sand were much greater than the actual results and could not be plotted realistically with the reat of the data. A possible explanation is that the bacfil!-Id soil was not actually in a dense state and therefore the coefficient Iq used in hold- ing capacity calculations was much too large and resulted in much greater predicted than actual values of holding capacity. There are fallacies involved in using either the Cylindricse or Torical failure criteria for all values of d/B for this problem. The cylindrical failure theory assumes a general shear type failure at all values of d/B which is simply not the case. A localized or punching type failure occurs in very loose sand at d/B > 2 and in a dense sand at a d/B > 10 (Vesic The Torical failure theory for the particular boundary conditions imposed on the anchor is not applicable. Torical theory assumes that the soil through which the failure q irface will form is uniform. Because the soil is backfilled with very little control over its final density, it is doubtful whether the jetted out soil could be emplaced at its in-situ density. If the backfilled soil is denser than the surrounding soil, then its failure mode should be controlled by the surrounding soil. However, if the backfill is less dense than the surrounding soil, its failure mode should be controlled by the backfill. Since in-situ density was not determined, Torical theory should not be used. 23 3500 ~ 3000 _ __ I-- : 2500 . / I J 6 / ___" 1500 f- w ______ A-l Val zed data fit _00_ _Torical 7 0 Cylindrical 0 1 0 500 1000 1500 2O00 2500 3000 Actual Holding Capacity, Ib Figure I2 Theoretical ve-ss actual holding capoc;ty for wnbedment anchors. 24 24 Cement Slurried Anchors Tabulated also in Table I is the approximate depth of embedment versus the holding capacity for each anchor upon which the cement slurry was injected. The four anchors slurried showed markedly greater holding powers than the non-slurried anchors. The mean maximum sustained load of these 4 was 4,475 lbs whereas the corresponding mean for unslurried anchors of the same size was 1,838 lbs. Two of these anchors were slurried through the apex of the cone. After testing, traces of concrete were found adhering to the surface of the cone and concrete in the shaft was protruding from the tip of the cone which apparently had broken loose from the concrete under the anchor. The other two slurried anchors were slurried by injecting the mix- ture alongside the anchor shaft above the cone as discussed above. The slurry was injected approximately 2 feet above the top of the anchor cone and formed upward displacing the loose sediment above. Evidence of some bonding with the surrounding sediment structure was noted but the primary increase in holding power for this case is attributed to the increased friction surface between the anchor and the side of the anchor hole as the anchor was extracted. For the fourth anchor slurried, only about half the full slurry load was emplaced, as discussed earlier. This anchor exhibited increased holding power, although no cement was found adhering to the anchor after testing. The use of a cement slurry to form additional holding power for the jetted anchor was thus a limited success. The experimental evidence indicates increased holding pow-r due to slurry use; however, the small number of tests performed precludes the possibility of drawing qualita- tive conclusions. A more controlled experiment, using a large number of anchors and an adequately engineered emplacement procedure, is indicated in order to obtain conclusive evidence of the value of the cement slurry for increased holding power. FINDINGS AND CONCLUSIONS From cost and manhour requirement considerations, the jetted-in cone anchors tested here seem to be a feasible means of obtaining easily em- placed, light duty, bottom tie-downs and anchorages . The anchors required approximately 1.2 man hours each to emplace using a six man crew. Several means have been suggested which cculd reduce the em- placement time and, in particular, considerably reduce the diver bottom time required per anchor. In general, the anchors proved easy to handle and emplace. No special skills, other than diving, were required of the Seabee enlisted personnel which emplaced the.. 25 Results of the holding cepacity tests and the resulting analysis indicate that the jetted cone anchors can develop holding powers within the desired range of 2,000 to 10,000 lbs The primary deter- mining parameters of holding power are the anchor size, the depth to which the anchor is jetted and the compaction obtained during backfilling of the hole. The variation of cone angle or nozzle size showed no measurable effect on the holding capacity of the anchors. Correlation of actual with predicted values of holding capacity was not very good. Until better control of the backfill density is realized, it will be very difficult to predict holding capacity with any of the available theories. The attempt to increase the anchor holding power by injection of a cement slurry proved to be time consuming in terms of topside and diver manhours. Encouraging, although non-conclusive, results were obtained indicating the use of such slurry injection may be a most effective means of increasing the holding power of the anchors tested. The addition of the cement slurry may have increased the resistance to pullout by: (1) increasing the dead weight of the anchor; (2) increas- ing the projected area of the anchor; and (3) penetrating into the undisturbed soil. Mechanical and procedural difficulties during the injection of the slurry accounted for the excessive manhour requirement and limited the usefulness of the results. I RECOMMENDATIONS Further tests using the jetted anchor described here are recommended. these tests should be conducted in both sand and clay soils using a larger number of each size anchor tested. Emphasis should be placed on determination of the effects of anchor size, depth of emplacement and soil properties on the holding power. In addition, further design work is recommended to develop the equipment and procedures for injection of a cement slurry on the anchors. This should be followed by further testing of the concept of increasing the holding power with the slurry. Specific suggestions concerning the implementation of these recomenda- tions follow: 1. All experiments should consider diver human factors and the implementation of procedures which will reduce the effort required to emplace the anchors. 2. The backfill soil, if required as in the present case, should be compacted adequately. Either simple rodding or some form of hand vibrator device should be used. This should cause the haldin* capacity to approach the values predicted by the Torical theory explained above. Some attempt at compaction of the surrounding soil should be made in any case. 26 3. Measurements of in-situ soil density should be made before anchor pullout in order to facilitate mathematical analysis of the anchor failure mechanism. It is suggested that a cone penetrometer or similar device be used for this purpose. 4. Larger cone diameters should be tested and more variation in depth of burial should be used. This would bring the ratio of embedment depth/cone diameter down into a range covered more adequately by previous tests. 5. The use of flukes on the anchors which would open after emplace- ment should be invyestigated, 6. When cement slurry is used, procedures for obtaining accurate placemenc of the slurry must be instituted. Excavation around a few slurried anchors may be useful in determining by what mechanism the cement is increasing the holding power. 27 REFERENCES 1. U. S. Navy, NAVSHIPS 250-538, U. S. Navy Diving Manual, July 1963 2. Lambe, T. W., Soil Testing for Engineers, The Massachusetts Institute of Technology, 1951 3. Larbe, T. W., and Whitman, R. V., Soil Mechanics, Wiley, New York, N. Y., 1969 4. Kalajian, E. H., and Bemben, S. M., "The Vertical Pullout Capacity of Marine Anchors in Sand", Project Themis, University of Massachusetts, April 1969 5. Vesic, A. S., "Breakout Resistance of Objects Embedded in the Ocean Bottom", Naval Civil Engineering Laboratory, Report CR .69.031, May 1969 6. Esquirel-Diaz, P. F., Pullout Resistance of Deeply Buried Anchors in Sand, Thesis, Duke University, 1967 28j I! I 28 ACKNOWLEDGMENTS The authors gratefully acknowledge the work of Commander WJ. Eager of the Naval Facilities Engineering Command on this experiment. The jetted-in cone anchor concept, which originated at the Naval Electronics Laboratory, San Diego, California, was developed further by Commander Eager and he ran preliminary on,,site tests of several anchors. Commander Eager originated the concept of using a cement slurry to in- crease holding power and developed the static test apparatus used. He also developed the besic procedure for the experiment and set forth the idea of running a parametric study of the factors affecting the anchor holding power. Without his efforts before and during the testing, this experiment would not have been possible. Senior Chief Utilitiesman R. Miller provided much assistance in organizing and supervising the Seabee Divers who tested the anchors. His efforts aDd those of the Seabees who constructed and tested the anchors are greatly appreciated. The assistance of Messrs. H. G. Herrmann and Homa J. Lee of the Naval Civil Engineering Laboratory is gratefully acknowledged. They provided much needed advice and assistance with the soils testing and the use of the holding power prediction theories. 29 APPENDIX A LABORATORY SOIL TESTS 3D .LHOI3M AS WISMYOD INIDU~dS too0000 10%0 w.o S~ H. 1111I11M I .. E 8 001 a'0 09 *v W6 9. a ~ g ~ ~ cc S IH§W A~ NH IIfit IND13M AG WISWVO3 ±N90kf3d * I* I I N h ojr to M E N w 1L0 IiL T- I Ci! j A a 2 LABORATORY DETERMINATION OF VERY LOOSE DENSITY A. PROCEDURE A cylindrical glass container of known dimensions was partially filled with fresh water. The weight of the water and container was noted. Dry sand was then carefully spooned into the cylinder, and the volume of the sand; and the combined weight of the sand, water, and container were noted. B. DATA Diameter of cylinder .734 inches weight of weight of cylinder, cylinder & water water & sand Height of 3and 66.11 gm 81.62 gm 1 53/64 In 71.04 gm 82.96 gr. 1 34/64 in 70.45 gm 103.19 gm 3 57/64 in C. CALCULATIONS density - weight/volume (wt of sand [gm]) (.002205 l ] (height of sand [in. ) (dia. of cyl. [in.] 2 72 i n3 • " 1728 3i 0. RESULTS Density [ib/Ft3j 75.96 75.54, 75.34 ! ~33 r CALCULATION OF DRY DENSITY, POROSITY, VOID RATIO, AND BUOYANT UNIT WEIGHT A. PROCEDURE During each shear test, three volumes were noted: the initial volume, relaxed volume, and loaded volume.. At the end of each test, the test sample was weighed using only the sand that remained in the test cylinder. B. CALCULATIONS ' wviuht of sample 3] DRY DENSITY d = wolune of sample ~Yd POROSITY (n) I 1 -- Gsyw n VOID RATIO (e) - BUOYANT UNIT WEIGHT s- 1 1 b/Ft WHERE = density of water G 3 specificty gravity of sample (2.76) (2.76) Is 34 K TABLE A-i DIRECT SHEAR AND DENSITY ANALYSIS RESULTS # 2 NORMAL 2 NORMAL 2 NORMAL 244 /Ft LOAD 7710/Ft LOAD 1484#/Ft2 LOAD DENSE LOOSE DENSE LOOSE DENSE LOOSE DENSITY 1 3 I lb/Ft 3 INITIAL 95.44 78.24 90.46 80.24 89.18 80.81 LOADED 95.45 78.39 92.53 81.21 90.14 87.41 RELAXED 89.57 78.49 91.02 80.5 86.57 81.28 POROSITY I INITIAL .4458 .5457 .4748 .5340 .4822 .5308 LOADED .4458 .5448 .4628 .5285 .4766 .4925 RELAXED .4799 .5443 .4715 ,5326 .4973 .5281 VOID RATIO INITIAL .8044 1.2012 .9040 1.1459 .9312 1.1312 LOADED .8044 1.1968 .8615 1.1209 .9106 .97044 RELAXED .9227 1.1944. .8921 1.1394 .9892 1.1191 PK. STRENGTH lb 20.6 10.8 43.2 30.0 60.5 44.1 PK. STRESS lb/Ft2 585.4 306.9 L228. 852.6 1719. 1253. UL. STRENGTH lb 11.6 10.4 25.0 25.0 44.0 43.6 UL. STRESS lb/Ft 2 330. 296. 710. 710. 1250. 1239. 35 600 Direct shear tests 245 lb/ft2 normal fod 0Dense 6~Loose 300 ii 100 8.0 t~4.0 0.0 -2.0 I 0.0 0.1 0.2 Shear Displacement OIn.) 36 1300 4 Direct shear test 771 lb/ft2 normal load 1100 0 Dense 1100 0 I .0 SO J 00 W1 8.0 1 6.04 -24.0 0.0 I0.1 0. Shear Displacement (in.) L 37A Direct shear tests 600 1400 C 6. . . . 400 f 8.0 6.0 -2.0 0 S. 0.1 0. i2 SerDisplacement (in.) 38 APPENDIX B PULL-OUT RESULTS d I Il I 39 $! S 04) '00 40' 1'41 El 0 1 0 *0 CD 0 00 I...D $1 00 (91) Io~ssB nl ii 42 06 -- _ I 044 4 0,I 7 Io _ _ - I 4 1_0 4 44 APPENDIX C PULL-OUT FACTORS CHART 'I *11 %I $** -% IN, 0 . ~~ ~ C ,•" %, AA S*,A % ,,4 -.\ .. * , 4, ,4, 4 , 4. ") % " % % 0ens. Sond' ID I % %| i% 6 - 7 9.0 20 40 O GO 80' 00 1 FACTOR Nq !i Anchor Breakthrough Factors Nq ?46 4.4 . Unclaosified -A St.cunty Classification DOCUMENT CONTROL DATA. R & D S'rurity classhte alto, ol tilll, body .,t aib ltur , ,,,,I.d. . ,fi,, ti,ol , ,n .,t .1 t,, Iti.,e t d h, :thin Ili' c,.,ll rt.p ,t f. , ! s i lI) I ONIGINA TING AC TIVI TV (Corporate A41tghor) 20. HE '-R - SEJRI TY C a 5 1 It . ATI.):I Naval Civil Engineering Laboratory Unclassified Port Hueneme, California 93041 2,. couP LTJG H. S. Stevenson and W. A. Venezia (Authors) 3 REPORT TITLE JETTED-IN MARINE ANCHORS 4 oCSCOIPTIVE NOTES (Type ol report and Inchisive dates) Final S AU THOIQ(SI (First name. middle initial, last name) LTG He S. Stevet.son (Principal) W. A. Venezia 6 REPORT OATE i. TOTAL NO OF PAGES 17b. NO. OF RCFS February 1970 156 6 _ 8. CONTRACT OR GRANT NO Ga. ORIGINATOR'S REPORT NUMBI PIS) , ! b. PROJECT NO YF 38.534.006.01.001 __TN-1082 C. hb. OTHER REPORT NOIS) (Aniy other number, that may be assigned this report) ~d. 10 OlSIRIBUTIOth4 STATEMENT This document has been approved for public release and sale; its distribution is unlimited. I SUPPLEMENTARY NOTES 12. SPONSORING .ILITARY ACTIVITY Naval Facilities Engineering Command I. Z ASTRACT -Twenty-three lightweight anchors consisting of a 10-foot X 2! -inch pipe with a metal cone welded at the tip end were emplaced and tested in twenty-five feet of water at Lameshur Bay, St. John, U. S. Virgin Islands. The anchors were jetted into the coral sand bottom by forcing water through the pipe and out an aperture in the cone. Two divers guided the anchors into the sediment using the jet of water to excavate the soil beneath the cone. Fcur of the anchors were emplaced with a cement slurry to increase the holding power. It was found that the emplacement procedures were straightforward and posed no problems to the divers. However, the injection of the cement slurry was very time- consuming due to numerous problems. The pullout results and theoretical analysis showed the jetted anchors to be capable of developing 2,000 to 10,000 lbs holding capacities in the soil at the test site. The holding capacity may be increased by increasing any of the following: anchor cone diameter (area), empiec-ment depth, compaction of the overburden sediment, or use of a cement slurry. "'es--s indicate that the use of these light weight jetted-in anchors may be of practical use where bottom tie-downs and light anchorages are required. Further testing is recommended. FOR (PAGE 1) Unclassified SA 0 I01. 87l- 680 1 Securttv C|l'ins-fication Unclassified Security Classification -4 KEY WORDS I.NK A LCNK D INK C ROLE W ROLE WT ROLr WT Anchors (structures) Foundations Jetting - Ocean bottom I i ! Placing j 1 Marine sediments I I Slurries Loads (forces) Performance tests JII I-I I ; I I a: I i4 ! DD ,.o,.1473 (BACK) Unclassif ied (PAGE 2) -Security Ciaq-til'ction 52 "I.