A VERSATILE VANE-SHEAR APPARATUS Edward M. Minugh LIBRARY NAVAL POSTGRADUATE SCHOOE MONTEREY, CALIF. 93940 United States Naval Postgraduate School THESIS A VERSATILE VANE -SHEAR APPARATUS by Edward M. Minugh April 1970 Tklt> document kaub been approved ^on public kc- lccu>t and t>aJL dUViibuXAjon -1& unUmitcd. A Versatile Vane-Shear Apparatus by Edward M„ Minugh Lieutenant Commander, United States Navy B.S., University of California, 1958 Submitted in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN OCEANOGRAPHY from the NAVAL POSTGRADUATE SCHOOL April 1970 \c o L^ ABSTRACT The vane-shear devices currently in use exhibit inherent problems and shortcomings associated with their design. The PGS Vane-Shear Apparatus is designed to eliminate these shortcomings. The unique features of the device include: a. Ability to be used in the laboratory or on board a ship. b. A single unit which is easily calibrated and capable of measuring torque over the entire range commonly encountered in marine sediments . c. A torque transducer which is insensitive to temperature changes and orientation. d. Ability to determine shear strength prior to cutting the core liner, thus reducing the disturbance to the sediment caused by cutting and handling. LIBRARY NAVAL POSTGRADUATE SCffOOB MONTEREY, CALIF. 93940 TABLE OF CONTENTS INTRODUCTION 9 EXISTING EQUIPMENT 14 The Wykeham Farrance Vane-Shear Apparatus 14 NCEL Vane-Shear Apparatus ..... 16 IIT Vane-Shear Apparatus 19 Diver-held Vane-Shear Apparatus 20 DESIGN CONSIDERATIONS 22 General 22 Torque Measurement 23 Power . 23 Recording of Test Results ........ 25 Motor 25 Sample Holders 26 THE PGS VANE-SHEAR APPARATUS 27 General Description ..... 27 Torque Transducer .......... 27 Power Supply and Signal Conditioning Unit ........ 32 Recorder ....... 34 Bracket Arm 34 Swivel Assembly ....... 36 Rack and Pinion Assembly 36 Motor and Motor Mount 41 Vanes 43 Calibration Apparatus „ 43 Core Holder 49 TEST PROCEDURES AND RESULTS ......... . . 50 Test Procedures 51 Results 51 Recommendations for Further Research 59 BIBLIOGRAPHY 60 INITIAL DISTRIBUTION LIST 62 FORM DD 1473 63 LIST OF TABLES Table Page I Ranges of Shear Strength 24 II Torque Required to Produce Shear 24 III Test Results 52 IV Additional Test Results 56 V Paired Student-t Test Results, PCS vs. NCEL Data 57 VI Paired Student-t Test Results, PGS vs. Wykeham 58 Farrance Data LIST OF FIGURES Figure Page 1 The Wykeham Farrance Vane-Shear Apparatus 15 2 The NCEL Vane-Shear Device 17 3 The PGS Vane-Shear Apparatus, shown in the 28 Laboratory Configuration 4 The PGS Vane-Shear Apparatus, shown in the 29 Shipboard Configuration 5 The Torque Transducer 31 6 The Power Supply and Signal Conditioning Unit 33 7 Detail of the Bracket Arm 35 -8 Detail of the Swivel Assembly 37 9 The PGS Vane-Shear Apparatus with Core 38 Holder 10 Detail of the Pinion 39 11 Detail of the Rack 40 12 Detail of the Motor Mount . 42 13 Detail of the Vanes 44 14 Calibration Apparatus with Torque Transducer 46 and Calibration Wheel Attached 15 Detail of the Calibration Stand and Wheel 47 16 Detail of the Arms for the Calibration Stand 48 17 Graph of Shear Strength vs. Length in Core 54 ACKNOWLEDGEMENTS The author wishes to thank Dr. R. J. Smith for his assistance and continued encouragement in the development of this device. Appreciation is also expressed to Mssrs. H. Gill, M. Hironaka, H. Lee, and L. Nunez of the Naval Civil Engineering Laboratory, Port Hueneme , California for their assistance and use of equipment during the evaluation of the apparatus. The author is also grateful for the interest and support provided by the Naval Facilities Engineering Command. Finally, the author is indebted to his wife Rose for her continued support and understanding throughout the entire period of postgraduate studies. INTRODUCTION Recent years have seen a growing interest by both private industry and government agencies in the physical characteristics of the marine sediments on the continental shelf and the deep sea floor. Extensive efforts are being expended to determine such factors as: a. The amount of support sea floor sediments can provide objects placed on the bottom, such as platforms, instruments, and manned habitats . b. The breakout forces required for large-object salvage operations . c. The degree of traf f icability of the sea floor. d. Ability of slopes on the sea floor to resist sliding. All of these factors are directly or indirectly related to the shear strength of the sediment. It can be demonstrated that the shear strength of a marine sediment can be expressed by Coulomb's formula: s = c + N tan i> where s =» total shear strength c ■ cohesion N = effective normal stress (6 = angle of internal friction. The significance of each of these terms are discussed in the standard texts on soil mechanics. Fine-grained, saturated marine sediments stressed without loss of pore water are generally assumed to behave as if they were cohesive materials without any internal friction under normal loading. For these conditions the angle of internal friction will be equal to zero [Keller, 1968]. The shear strength of saturated marine sediments is therefore sometimes alter- nately referred to strictly as cohesion. Measurement of the shear strength of marine sediments may be made by direct shear, triaxial shear, and unconfined compression tests. These tests require the sample to be removed from core liners, and the resulting disturbances may appreciably affect the results of the tests. These shortcomings can be minimized by applying the vane shear test [Smith, 1962], All vane shear testing devices currently in use are adaptations of the vane borer, which was developed simultaneously in Sweden by John Olsson in 1928 and in Germany as evidenced by a patent in 1929 [Osterberg, 1957]. These devices received little attention until Cadling and Odenstad [1950] reported results of comprehensive tests conducted in Sweden on the shear strength of clays. This report described a method of obtaining shear strength of clays in-situ. The equipment used by them consisted of four rectangular shaped vanes, welded at right angles to a rod. The vane assemblage was inserted into a hole bored into the ground and extensions added to the rod until the vane reached the bottom. The vane was then driven into the undisturbed soil below the bore hole and torque applied to the rod from the surface. This torque was measured by means of a cali- brated spring. The vane was rotated until the maximum torque was reached, followed by a decrease to a value which was necessary to maintain a constant rate of rotation of the vane. The shear strength of the soil was then determined from the following equation, which was derived by Cadling and Odenstad: 10 M max s = ^„ ntI D , „ nD2 2 D^ where M s maximum torsional moment required to produce shear IT13.X D - diameter of the vane H - height of the vane. The above equation assumes the surface of rupture is a circular cylinder surrounding the vane, with the height and diameter of the cylinder being equal to the dimension of the vane. It is also assumed the stress distribution at the maximum torsional moment is uniform across the surface of the cylinder including the end surfaces. The friction exerted by the soil on the shaft of the vane is considered negligible. These assumptions are taken to be valid in all vane shear testing devices currently employed. Cadling and Odenstad also report that the rate of rotation of the vane influences the results, with a higher shear strength being associ- ated with the higher rotation rate. They worked with rotation rates from 0.1 to 1.0 degrees per second and note that higher rotation rates may yield shear strengths as much as fifteen percent greater than the shear strength determinations at the lower rotation rate. On the assumption that the reported values of shear strength should correspond to the. most unfavorable case, the rotation rate of 0.1 degrees per second was adopted as standard. Various investigators have abandoned this "standard" in favor of higher rotation rates. Morelock [1967] uses a rotation rate of 0.2 degrees per second, assuming that the resultant increase in shear strength is minimal and that the higher rate is much more practical when examining large 11 numbers of samples. Bouma, Bryant, and Tieh [1968] have app; sed the 0.2 degrees per second rotation rate in their studies of tl tinental shelf of the Gulf of Mexico. Aas [1965] reports fir., significant difference in shear strength by using rotation rates from one to ten revolutions per hour and all intermediate values. The height/diameter (H/D) ratio of the vanes has been studied extensively by Cadling and Odenstad [1950], who use an H/D ratio of two as the standard for their work. Aas [1965] experimented with various vane shapes to determine whether the H/D ratio significantly altered the test results, and concluded the results were not appreciably changed unless the H/D ratio was greater than three. Osterberg [1957] suggests that to avoid disturbance of the soil to be tested, the area of the vane should not exceed 10 percent of the area of the circular section to be sheared. The vane borer of Cadling and Odenstad was the predecessor to all vane-shear devices currently used in determination of the shear strength of marine sediments. While it was designed for testing terrestial soils, those subsequently developed for testing marine sediments are scaled- down models of the vane borer. Although the principles rema i same, the shear strength, torsional moments measured, and the dimensions of the vanes are several orders of magnitude less when dealing with marine soils . The following comparison illustrates the differences between terrestial and marine soils: 12 Typical Marine Sediments, West Coast of North America [Moore, 1961] 2 Shear Strength (gm/cm ) Open continental shelf 11 Continental borderland, basins, and slopes 18 Bays and estuaries 13 Continental slope 7 Deep-sea terregenous 26 Classification of Terrestial Clays [Terzaghi and Peck, 1948] 2 Term Shear Strength (gm/cm ) Very soft < 250 Soft 250 -- 500 Firm 500 -- 1000 Stiff 1000 -- 2000 Very stiff 2000 -- 4000 Hard > 4000 It can be seen from the comparison above that any device which is to be used on marine sediments must be capable of measuring compara- tively low values of shear strength. 13 EXISTING EQUIPMENT There are several variations of vane-shear testing equipment cur- rently used to determine the shear strength of marine soils. The devices most widely used or those with particularly interesting features are mentioned below. The Wykeham Farrance Vane-Shear Apparatus The Wykeham Farrance Vane-Shear Apparatus (Figure 1) is manufactured in England and is in wide use throughout the world. It is available in two versions: the hand-driven model, in which the vane is rotated by manually turning a hand crank, or the motor-driven model. Several vane rotation rates are available, with 6 per minute the standard equipment. The sample container is secured to the base plate to prevent its rotation during the test. The vane is lowered into the sample by means of the top hand crank until the top of the vane is at least 0.75 inches below the surface of the sample. The torque necessary to shear the sample is measured by a spring having a linear response to torque. The spring is calibrated by the manufacturer, who provides the linear spring constant associated with each spring. Two dials are provided on the upper part of the apparatus. The outer dial indicates the degrees of rotation of the vane, while the inner dial shows the degrees of applied torque. The outer and inner dials are initially set to zero, and the test is started by turning on the motor or by manually turning the side hand crank. When the sample shears, the vane rotates at the same speed as the application of torque. The inner dial will remain on the reading at the time of shear, while the outer dial continues to rotate until the 14 Figure 1. The Wykeham Farrance Vane-Shear Apparatus 15 motor is turned off or the cranking stopped. The amount of torque necessary to shear the sample is determined by dividing the reading on the inner dial at the time of shear (applied degrees of torque) by the spring constant. The shear strength is then determined by: torque s = a 2tt r2 (h + 0.667r) where h = height of the vane r = radius of the vane. According to Richards [1961], a vane rotation of at least 20 degrees is required for a valid test. The Wykeham Farrance apparatus apparently yields good results, although very sensitive springs are required for soft soils which shear at the very low torque ranges. In addition, the spring calibration must be checked frequently to ensure the spring has not been stretched beyond its elastic limit. The principal advantages of the Wykeham Farrance apparatus are that it is simple to operate, is relatively light-weight, and can be used aboard ship or in the laboratory. The primary disadvan- tages are that it does not provide a continuous and permanent record of the soil shear, and spring calibration must be checked frequently. To obtain a sequential record of a test, it is necessary for the operator to record the degrees of applied torque every 2 of vane rotation during the test. The values must then be plotted, a process which is not difficult but exceedingly time consuming. In addition, much of the detail of the shear profile is lost by manual plotting. NCEL Vane-Shear Apparatus The NCEL vane-shear device shown in Figure 2 was designed by Smith [1962] at the Naval Civil Engineering Laboratory, to provide a high 16 O) o > Q J-i ca w g -C H d 60 •H fa 17 degree of sensitivity throughout the range of shear strengths normally encountered in marine soils, while allowing a continuous and permanent record of the test. The sample to be tasted need not be removed from the core liner, as the liner segment containing the soil is fastened to a disc which is rotated by a motor mounted beneath the base plate. The vane remains fixed throughout the test while the sample rotates around the vane. The vane is attached to a shaft, which in turn is guided through an upper plate by means of a teflon bushing and ball bearings. The top of the shaft is equipped with cantilevered feeler gauge stock reeds equipped with SR-4 strain gages. The ends of the reeds rest against vertical posts attached to the upper plate assembly. As the sample is rotated, torque is applied to the shaft which in turn bends the reeds. The strain gages, in turn, measure the distortion of the reeds which is proportional to the torque developed. Inter- changeable reeds, having thicknesses of 1/64 and 1/32 in. are provided to allow for variability in the samples being tested. Power to the strain gages and the bridge balance unit is provided by rectified 110 volt a-c source. The imbalance of the bridge caused by the output of the strain gages is amplified, and the output is fed to a strip chart recorder. Thus, a permanent record of the entire test is available. The apparatus is easily calibrated by use of a . calibration wheel which is screwed into the shaft. Known weights are passed over a guide block while attached to notches in the calibration wheel. In this manner, known torsional moments are utilized to adjust the amplifier gain to the desired level. Calibration can be accomplished in five minutes . 18 The NCEL Vane-Shear Apparatus is restricted to laboratory use only, and cannot be used aboard ship. The unit must be precisely leveled on the laboratory work bench to prevent misalignment of the shaft, and attendant binding in the bushing and ball bearings. Binding occurs if the shaft is not maintained exactly vertical. The shaft configuration is such that samples in excess of 3 inches in length cannot be tested. IIT Vane-Shear Test Apparatus An interesting apparatus designed to measure vane-shear strength with the soil sample under high environmental pressures, and patterned after the NCEL vane-shear device, was developed at the Illinois Institute of Technology (IIT) Research Institute by Vey and Nelson [1966]. The soil specimen is placed in a 2.5 x 2.0 inch container fitted with a porous stone at the bottom to provide drainage. The vane, shaft, and transducer assembly are then mounted to the top cover of a pressure vessel. The torque transducer consists of a beam rigidly mounted to the shaft. A rigid post mounted 2.0 inches from the center of the shaft keeps the beam from rotating, thereby providing the bending moment to the beam. SR-4 strain gages attached to both sides of the beam then measure the deflection due to twist from the shaft. The torque transducer is contained in a plexiglass housing filled with oil and has a flexible diaphragm cover to equalize the fluid pres- sure in the housing. Calibration of the torque transducer while under pressure is accomplished by mounting a calibration wheel to the shaft. A known torsional moment is applied to the shaft via the calibration wheel by means of weights attached to a cable. A small weight is always suspended 19 from the calibration wheel. This ensures that the beam is in contact with the rigid post support at the start of the test. The entire vane-shear apparatus, including the soil sample, is placed in a pressure chamber by securing the top cover to the pressure vessel. Vane shear tests may be conducted at environmental pressures from atmospheric to 5000 psi and at temperatures from 1 C to 3 C. Diver-held Vane-Shear Apparatus An inexpensive diver-held vane-shear apparatus capable of in-situ operation in shallow waters has been developed by Dill and Moore [1965]. The device consists of a commercial torque screwdriver with a 3/4 x 3/4 inch vane attached to the shaft. The device is capable of measuring torques from 0-24 inch -ounces . According to Moore [1962], these parameters (3/4 in x 3/4 in vane, 0-24 inch-ounces torque) are adequate for most shear strengths encountered in the upper six inches of marine sediment . The device has proven useful in determining shear strength and residual strength in and around an active slumping area at the head of Scripps Submarine Canyon. It has also been used successfully on board ship to determine shear strength of relatively undisturbed sea-floor sediments obtained by box samplers. The primary disadvantage of a device of this nature is that there is no means of controlling the rate of stress application. In this instance the diver was instructed to gradually build up torque over a period of not less than two minutes, until the sediment sheared. It is difficult, if not impossible, for a diver to duplicate the rate of stress application. Another disadvantage is that the shear pattern cannot be 20 obtained since no permanent record is made. It may be desirable to forsake this record for an in-situ test, but for shipboard and laboratory vane-shear tests, a permanent record is highly desirable. 21 DESIGN CONSIDERATIONS The following presents some of the more important factors which must be considered in designing a vane-shear apparatus. These factors explain, in part, the final design and component selection of the items that constitute the vane-shear apparatus described in the following chapter. General A vane-shear apparatus should be versatile, performing equally well aboard a vessel at sea or in the laboratory. In order to meet this criteria, the device must be capable of being easily assembled and/or disassembled, portable, preferably light-weight, and with no special leveling required. At the same time, the individual components must be sturdy enough to withstand the inevitable jolts incurred during transportation. Vane shear tests are generally conducted on soils contained in a core segment which has been cut from the core liner. It would be more desirable to conduct vane shear tests on a soil prior to cutting the core liner into segments in order to minimize disturbance. The vane- shear test could be run on the uppermost portion of the core. The core liner could then be cut and the vane-shear test is run on the next section while still intact with the remainder of the core. This pro- cedure would minimize disturbance of the sediment sample. When the core is cut into sections prior to testing, each section is subjected to two cuttings except for the top and bottom segment of the core. 22 Torque Measurement The critical component of any vane-shear device is that part which measures the amount of torque on the vanes at the time the soil shears. Regardless of the type of device used to measure torque, it must be capable of measuring the entire range encountered in the determination of shear strength of marine soils. Table I shows the range of shear strengths commonly found in marine soils. Table II shows the torque required to produce the shear strength reported in Table I. Examination of these tables reveals that by proper choice of the vane size, a device capable of accurately measuring 1-250 inch-ounces of torque is adequate except in the most unusual circumstances. Measurements of torque in the higher ranges present no special difficulty, but a very sensitive instrument is required to measure torque in the neighborhood of 1 inch-ounce or less. Sensitive, calibrated springs are easily damaged by straining beyond their elastic limit. The use of electronic, strain-gage transducers is feasible but presents problems of signal discrimination over noise at these low torque values. It is highly desirable to utilize a single unit, rather than inter- changeable units which cover only discrete segments of the torque range. Such a single unit eliminates errors in judgment on the operators part, and thus saves time and produces more accurate results. Once a sepcific area of a soil sample has been sheared, it is impossible to obtain an accurate value of shear strength at the same location on a test re-run. Power If electronic devices are to be used in the apparatus, provision should be made for a regulated power supply. Voltage and frequency 23 TABLE I Ranges of Shear Strengths NCEL RICHARDS HAMILTON & MOORE NCEL HAMILTON HORN AND IAMB Number of cores tested 39 31 10 75 -- 7 Number of tests per core 1-15 1-34 1-4 1-21 -- 1-12 Sample depth (m) Surface Surface 0-168 Surface Surface 10-25 Shear strength (gm/cm2) 1.2-138 4.0-234 610-8000 1.5-380 4.0-192 75-1200 TABLE II Torque Required to Produce Shear (inch-ounces)' 1/2" x 1/2" vane .070-8.2 .24-13.9 36.3-476 .09-22.6 .24-11.4 4.5-71.5 1" x 1/2" vane .12-14.4 .35-24.4 63.6-834 .16-39.6 .42-20.0 7.8-125 1" x 1" vane .57-65.8 1.9-111 290.8-3814 .72-180 1.9-91.6 35.8-572 2" x 1" vane 1.0-115 3.3-195 590-6675 1.3-319 3.3-160 62.6-100 Tables compiled by H. Herrmann [1966] 'Torque calculations are the responsibility of the author correspond to those of Table I. Columns 24 surges are common in electrical circuits, both in the laboratory and aboard ship. Surges of this nature can produce erroneous signals in sensitive electronic equipment. Whenever possible, circuits provided with stabilizing transformers should be used. (Such circuits are not commonly found aboard ship.) The basic power supply should be 115 volt-60 cycle, enabling the unit to be used anywhere. Recording of Test Results The desirability of a continuous graph-type record of the test has been emphasized throughout this study. Such a graph provides a shear profile, which varies with different soil types. These profiles may be obtained by manually plotting torque versus degrees of vane rotation at discrete points throughout the test but are demanding upon the time of the laboratory technicians, especially when a large number of tests are to be conducted. In addition, minute variations in the shear profile are missed by incremental plotting. Motor The ideal motor for rotating the vanes or the sample is one which is variable in speed, can provide the necessary torque, and does not require gear reduction to produce the desired rotation rate. Unfortunately, such motors are not available. The slow rotation rates commonly employed (1-2 revolutions per hour), coupled with the torque require- ments, are too demanding upon the armature of the motor. Consequently, reduction gears must be relied upon to produce the desired rotation rate. Every attempt should be made to use precision gearing in order to eliminate "slop" and the attendant vibrations. Vibrations of this 25 nature can be transmitted through the vanes or sample holder (depending upon whether the vane or the sample is rotated) to the test specimen, resulting in undue disturbance. Variable speed motors should be avoided because of their tendency to "hunt" under changing torque loads. Sample Holders Care must be exercised in the design of sample holders to avoid stress application to the sample. Compressive forces on the side walls of the core liner or sample container should be avoided. Such com- pressive forces result in soil disturbance and erroneous results. Where possible, the forces necessary to prevent the sample from rotating during the test should be exerted on the core liner in a vertical direction. If the vane-shear tests are to be conducted prior to cutting the core liner, as explained in the general discussion section of this chapter, the sample holder should maintain the core as nearly vertical as possible to prevent excess water drainage and separation of the soil from the walls of the core liner. 26 THE PGS VANE -SHEAR APPARATUS The design of the vane-shear device was selected following a review of the literature, conversations with users of the various existing vane-shear devices, and correspondence with numerous manufacturers of component parts. General Description The PGS vane-shear apparatus may be configured for laboratory or shipboard use. The combined weight of all the component parts (less recorder) in the laboratory configuration shown by Figure 3 is 26 pounds and the total weight in the shipboard configuration of Figure 4 is 44 pounds . The PGS vane-shear apparatus consists of the following components: 1. torque transducer 2. power supply and signal conditioning unit 3. bracket arm 4. swivel assembly 5. rack and pinion assembly 6. motor and motor mount 7. calibration stand and wheel. All components except the torque transducer, power supply and signal conditioning unit, and the recorder were constructed by the Machine Facility at the Naval Postgraduate School during December 1969 and January 1970. A description of these parts is contained in the fol- lowing sections. Torque Transducer A review of various torque measuring techniques indicated the desirability of utilizing a torque transducer in this apparatus. This 27 c o •H •U CO S-i 3 M) •■-I m c o u o ■u JO o Mi o CO to •u CO U «0 a a, < u CO CO O CM a) x: H iMttlM^U.,. * CO a) u (JO 28 Figure 4. The PGS Vane-Shear Apparatus, shown in the Shipboard Configuration 29 eliminates many of the disadvantages of the calibrated spring. The torque transducer selected, as shown in Figure 5, is an in-line., semi- conductor strain gage transducer, Model A44, manufactured by West Coast Research Corporation of Santa Monica, California. The range of the transducer is 0-250 inch-ounces, although it may be over-torqued 100 percent without damage. The output of the transducer is .269 millivolts/ volt excitation/inch-ounce, and is linear throughout the entire range. Accuracy of the torque measurement is + 0.1 percent throughout the range. The internal resistance of the transducer is 350 ohms. For all practical purposes, it is insensitive to temperature change with a temperature response of 0.0045 millivolts/degree Farenheit, and 72 F being the calibration temperature. The transducer will measure either clockwise or counterclockwise torque, the polarity of the output signal indicating the direction. The semiconductor type transducer was used because of its ability to discriminate signals over noise in the very low torque ranges. The use of semiconductors enhances signal discrimination at these low out- put levels by a factor of approximately thirty over the conventional strain gage. Excitation to the strain gages is a regulated 5 volt DC signal from the power supply unit. Because of the slow rotation rate involved, slip rings are not needed, which helps to eliminate noise in the output signal. The output is unaffected by the orientation; it may be used horizontally, vertically, or in an oblique attitude. The top of the transducer is provided with a 1/2 inch long l/4"-28 thread stud which is screwed into an adapter when the transducer is used for testing. When calibrating the equipment, this stud screws into the 30 u u 3 XI CD G to U H 0) h o H •r-l CO -C a; Q oo CM QJ U 3 00 37 ""■!-■: ;.;*:. -;■;.. f* Figure 9. The PGS Vane-Shear Apparatus wlth Core Holder 38 Slot for Teeth 3/8 Across Rots "DIA 3/8-16 Drill 43 Thru Tap 4-40 nt 2" Drill 29 Thru Tap 8-32 Drill 43 Top 4-40 Figure 10. Detail of the Pinion 39 12" Slot for Teeth .12" Wide .10" Deep 10" Long Brass Teeth -16 /in. 1/2" D ALuminum Retaining Ring Drill 21 Top 10-32 Figure 11. Detail of the Rack 40 and the rack portion will fall freely by its own weight. The thumb screw on the pinion part of the assembly allows the rack to be locked in place in any position. Both screws are teflon tipped to prevent scoring of the barrel of the rack portion. An aluminum retaining ring fits over the barrel of the rack portion, The retaining ring is slotted to avoid interference with the gear teeth so that the ring may be positioned at any location. The purpose of the retaining ring is two-fold: first, it prevents free fall of the rack portion if the set screw adjusting the drag should inadvertently be turned the wrong direction; secondly, the retaining ring may be positioned so that it is flush with the motor mount when the top of the vane is 0.75 inches below the surface of the sample being tested. Motor and Motor Mount The motor mount provides the means for securing the motor to the rack protion of the rack and pinion assembly. The mount shown was designed for a specific motor installation and would require alter- ation if the motor is subsequently replaced. A drawing of the motor mount appears as Figure 12. The single phase, synchronous, heavy duty motor is manufactured by Hurst Manufacturing Corporation, Princeton, Indiana. It is rated at 150 inch-ounces of torque at 1 RPM and requires 115 VAC, 60 cycle power. The output shaft rotates at one revolution per hour in a counterclockwise direction. The reduction gears are contained in a sealed unit and require no lubrication. 41 j ro CM M %7a 9> C P5 CM fi ^ - — CM -IO tP 1 ro <£> g Drill Tap a 2 Q. d co CO CO •H 46 qu f _J u. Q. Q Z X) c CO *o G CO •u en c o to u CO o (1) J2 H-l o CD o m M 3 60 CVJ LI C CO u XI crj U QJ .c 4-1 u o 03 £ U < -C Q n£> 48 "R Cal" circuit has changed. The calibration apparatus was originally designed to check the output specifications supplied by the manufacturer of the torque transducer. Core Holders The legs of the core holders may be adjusted, in three inch incre- ments, to heights between 2.5 and 4.0 feet. The cross member may be rotated to facilitate placing the core at any desired angle. Adapters are provided to accommodate core liners of 3.5, 3.0, 2.5, and 1.5 inches outside diameter. The assembled core holder is shown in Figure 9. 49 TEST PROCEDURES AND RESULTS The PGS vane-shear apparatus was transported to the Naval Civil Engineering Laboratory at Port Hueneme , California where provision was made for cores and samples from storage and use of the NCEL vane- shear apparatus permitted in order to establish comparative values from each device. Tests were conducted on numerous samples and the results obtained from each vane-shear apparatus compared. In order to eliminate temporal variability of the samples as much as possible, comparative tests were conducted within fifteen minutes of each other. The apparatus selected to test each sample first was varied randomly so that the succeeding test result would not be biased by the result of the first test. It was not possible to obtain comparisons of undisturbed shear strength, although several attempts were made to conduct such tests. The maximum length of core which can be tested on the NCEL apparatus is three inches. The sample was first tested on one apparatus, then inverted and tested on the other, but portions of the vane on the second test were extruding into the cylindrical area previously ruptured by the first test. Since it was impossible to extend the length of the test sample (because of the three inch limitation on sample length with the NCEL apparatus) it was necessary to resort to remoulded samples to obtain a valid comparison. The samples were chiefly dark, clayey sediments containing numerous 1 fragments. Where large discrepancies existed in the test results, I i he sample revealed shell fragments or other irregularities ',(i on or in the vicinity of the shear cylinder. Only those tests which were considered valid are included in the results. Test Procedure After cutting the core liner, the soil was extruded from the liner into a ten inch diameter crucible and thoroughly mixed and kneaded. Visible shell fragments and other foreign objects were removed. In certain instances, water was added to the sample to obtain a more favorable consistency in order to ensure valid comparisons over a representative range of strengths. The soil was then placed in a brass cylinder 3 inches long and 2-1/2 inches in diameter. The cylinder was secured to the sample holder and the vanes lowered to 0.75 inches below the surface of the sample. The rotation rate used during all tests was one revolution per hour. Upon completion of the test, the sample was removed and again mixed thoroughly and replaced in the cylinder. The test was then repeated on the second apparatus. Results Fifteen valid sets of comparative data were obtained, the results of which are shown in Table III. The right hand column of Table III expresses the value of the NCEL apparatus result as a percentage of the value obtained with the PGS apparatus. The largest difference obtained expressed as a percentage of NCEL t PGS, was 19.7 percent. The smallest difference was 1.2 percent. Of the fifteen comparative values, three had a difference greater than 10 percent, three had a difference between 5 and 10 percent, and nine had a difference of less than 5 percent. The average difference of results for all tests is 51 TABLE III Test Results 2 Shear Strength (lbs/ft ) NCEL/ PGS - Sample No. Expressed as a NCEL PGS percentage PP-2: 0-3R 379.8 473.0 80.3 3-6R 214.7 206.3 104.1 0-3R(+H?0) 6-3R(+IC0) 6-10R 51.0 49.1 103.9 58.95 68.8 85.7 101.5 98.2 103.3 13-16R 59.0 66.3 90.0 16-18R 45.8 46.6 98.3 TH-1: 3-6R 85.95 88.65 97.0 6-9R 53.0 55.9 94.8 12-15R 71.9 61.4 117 . 1 TH-2: 0-3R 74.4 69.8 106.6 6-9R 53.0 51.0 103.9 12-15R 35.8 36.5 98.1 TH-3: 12-15R 150.5 152.3 98.8 12-15R(+H20) 47.3 48.3 97.9 52 6.5 percent. If the tests having, a difference greater than 10 percent are discarded, the average difference between results decreases to 3 .9 percent . Nine of the fifteen remoulded shear strength values obtained by the PGS device were greater than the value obtained by the NCEL apparatus, which represents 60 percent of the total number of tests. It was anticipated that this percentage would be higher because of suspected frictional losses in the bushing and bearings on the NCEL device, especially at very low torque values. Additional comparisons may show this to be the case. On 5 February*" 1970 a core sample was obtained by the Naval Oceanographic Research Ship BARTLETT off the coast of Central California, at latitude 36 30 'N, longitude 123 56'W, in a water depth of 4200 meters. The core was cut into three inch sections, and shear strength tests were obtained on each of the sections using the PGS apparatus. The results of these tests are shown in Figure 17, which is a plot of shear strength versus length from the top of the core . Examination of the core prior to cutting revealed what appeared to be a water pocket approximately one inch long at a length of twenty-one inches from the top of the core. Subsequent cutting of the core showed a difference in color and texture of the sediments above the water pocket from those below. The shear strength values obtained verify the existence of the two distinct sediment types, the upper sediment being much stronger than that below the twenty-one inch level. 53 360 t 335„ 285-- 4-1 CO 00 c ' n ' ■ 5 Minugh A versati 1e vane- shear apparatus. thesM632 A versatile vane-shear apparatus. 3 2768 001 89107 0 DUDLEY KNOX LIBRARY