_:_ __::___:_____12:21:3333:: ___:__ «HF.» m. . (\ ~ o.HJ .16. § . v Q at? .- . L ... C» k r w ~‘:’ wm .. L1 .Iha w . .‘(I‘ .A. '1 (I an» 1‘ “a W. '0’. f; 3‘ S, "‘ o .37., r \ n.‘.u. .1. \ a. . .2 .. F 4 /'~ '1 ‘ \s 5% ... .. {k 9 J. ‘— THESIS 5-.-. This is to certify that the thesis entitled THE CORRELATION OF A PENETROMETER TO LOAD BEARING TESTS ON PREPARED PLOTS r—r- presented by ROBERT H . KEYSER ’ has been accepted towards fulfillment of the requirements for M. 8. degree in CIVIL ENGINEERING - , ‘. M 0. WW 1 Major professor mew 7.4- . O-l69 THE CORRELATION OF A PENETROMETER TO LOAD BEARING TESTS ON PREPARED PLOTS by Robert Henderson Keyser A THESIS Submitted to the school of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Civil Engineering 1951 AC KN OWLE DGME N T Profound gratitude is eXpressed to Professors L. S. Robertson and C. M. Hanson, Department of Soil Science and Agricultural Engineering, Michigan State College, for the loan of their Recording Soil Penetrometer. Special appreciation is extended to Professor G. C. Blomquist, Department of Civil Engineering, Michigan State College, for his liberal contributions of time, Knowledge, and constructive comments, without which this work could not have been formulated. INTRODUCTION STATEMENT OF PROBLEM EQUIPMENT SCOPE OF PROBLEM PLOTS . ROCEDURE DISCUSSION RESULTS CONCLUSIONS SUMMARY. TABLE OF CONTENTS page . 12 . 46 INTRODUCTION The soil types encountered vary not only with the geo- graphy and the geology of the areas in question, but it may vary from place to place and almost from spot to spot. Thus it is seen that the load bearing capacities of soils may be almost infinite in number. load bearing tests vary from complex tests dependent on tests for cohesion, internal friction and shear, to simple, on the spot, loading tests. The type of test required will depend on several conditions. Some of these are the soil in question, the size of the structure, the equipment at hand, and the degree of perfection required. To date, the tests used are not standardized either to method or specification of equipment. It is not the purpose of this investigation to delve into the conflict of the con- troversial factors governing the determination of the load bearing data and the methods of attaining such data. The fact that such altercations exist lends credence to the pro- blem under discussion. All structures, mobile or permanent, must at some time or other depend upon the soil (or rock) as a means of support. No matter how precise the design, or the degree of perfection attained in the production of the component materials, it is axiomatic that a structure is only as stable as the foundation upon which it rests. It is pertinent then that the soil data is as accurate as present means of attainment deem possible. V STATEMENT OF PROBLEM The correlation of a penetrometer to load bearing tests on prepared plots. EQUIPMENT The equipment used for the load bearing tests consisted of a twelve ton truck for the required load, a hydraulic jack to transmit the desired increments of load with a dynamometer to indicate the load produced, and strain gages to measure the deflection. (figure 1) A one hundred square inch round bearing plate was used. It is believed by many that the round bearing plate was used. It is believed by many that the round bearing plate produces a more uniform distribution of stress around the perimeter, thus eliminating the highly or over-stressed sections produced by the corners of the square plate, also the stress pattern of soil in a horizontal section is circular, and the square plate stresses cannot conform to this pattern. Time is an element which must be taken into consideration in the selection of the size of the bearing plate. The larger the plate, the more is the time required to attain the deflection corresponding to the increment of load in question. Thus it may be seen that days or even months may be required to obtain one set of load bearing datum. method of conducting bearing test. A view of plate loading device. figure 1 - 3 _ A recording soil penetrometer, (figure 2), is an in— strument which measures physical soil conditions. This is accomplished by determining the pressure required to force a probe into a soil to a given depth. The new instrument was designed in an attempt to provide a piece of equipment having the following requirements: 1. Compact - so that it might be tran ported in an ordinary automobile; 2. Light weight - for one-man Operation; 3. Rapid performance - so that adequate data can be obtained from a given soil in mini- mum time and with minimum effort; 4. Versatile - so that it can be used for a number of soil conditions; 5. Inexpen- sive - to construct and maintain; and 6. Functional - so that the user can obtaig a picture of soil conditions to a depth of twelve inches. The instrument is mechanically accurate within nine per cent. The graph is a component of the pressure required to force the probe into the soil, and the depth of penetration of the probe. The abcissa is drawn by the pressure of the soil transmitted to a calibrated coil Spring, and the ordinate is produced by the differential between the probe head and a float rod foot which rests on the soil surface. Two pulley systems are required to produce the desired four inch graph and to compensate for the spring compression. 1. L. S. Robertson & C. 4. Hanson, "A Recording Soil Penetro- meter", Reprinted from Michigan Agriculture Experiment Station Quarterly Bulletin, vol. 33. No. 1. Aug. 1950. -4- Recording Soil Penetrometer figure 2 Probe points used in the problem were: a pointed tapered shaft, and flat heads with circular areas of 0.15 square inch, 0.25 square inch, 0.50 square inch, 0.75 square inch and 1,00 square inch. The rate of application of the force was not controlled mechanically but a speed indicator is contemplated in later develOpments. The maximum size of the head to be used is limited by the type of soil and the weight of the Operator. Sizes larger than one square inch will require either a person of larger than average size, or a mechanical loader. Either condition will defeat the purpose of the project and simply produce an— other cumbersome load bearing machine. The penetrometer itself is not at present refined to per- fection by industrial methods. It is still in the hand made develOpment state. It will perform with sufficient accuracy its essential functions. Any elaboration on the need for eloquence of performance or ap-earance need not be considered herein, as long as results are contained within tolerable working limits. The equipment use will contain many of the attributes and deficiencies attributed to such apparatus by various authorities on the subject. This has no deterrent influence upon the original problem. SCOPE OF PROBLEM It may be seen from the discussion of the equipment used for load bearing capacity tests that the results are obtained only after much tedious work, with cumbersome equipment, and at considerable expenditure of time and money. It is thus desirable to obtain comparable results with equipment more easily handled, with less expenditure of time, and with a reduction in monetary expense. Some soil engineers consider the practicability of the above stated problem as remote, and it may well be especially when one is dealing with material as heterogeneous as soil, yet in a limited way this may be applicable. If one is work- ing with a limited set of constructed materials, such as may be encountered in highway or airfield construction, then one is not encountered with the problem of presenting a "cure all" for obtaining the load bearing data, but rather a method with practical application in a limited field. Thus this problem is reduced to that state, to present an economical, convenient, and accurate method of obtaining load bearing data to supplement material obtained by conventional load bearing tests on constructed structures of known soil material. PLOTS In accordance with Michigan State Highway Department practices, well drained granular soils including bank-run gravels, sands, and loamy sands having no plasticity, require no subbase. When plastic soils are encountered such as clays, sandy clays, and silty clays, a fifteen inch granular cushion is constructed on the prepared subgrade. The grade for the finished pavement is set and sufficient selected binder soil is next added to a minimum depth of six inches, after which the shoulder is thoroughly consolidated. The soil material may be bank-run sand or gravel, or dune sand of approved granular mix. The material must all pass a two inch sieve, sixty to one hundred per cent passing a one inch sieve, and zero to twenty-five per cent passing a No. 4 sieve, the loss by washing shall be less than ten per cent of the entire sample. ApprOpriate substitutes may be used on basis of laboratory tests. It may be comparable to P.R.A. classification A-3. ' Salvaged tOpsoil may be used for stabilization of the granular material. Binder soil for stabilizing 22-A material may consist of clay, sandy clay or loam with a P.I. of be- tween one - six and nine - fifteen, this corresponds to P.R.A 2 soil A-2. 2. E. A. Finney, "Shoulder Construction Practice in Michigan" Report to the Highway Research Board, Roadside Development Subcommittee on Shoulders, Department of Design, Washington, D. 0., December 6, 1948. 22—A Material Sieve size 3/4" 3/8" #10 #200 3 Construction of Plots: Per cent passing 100 60-80 25-40 0 - lO 1. Remove tOp soil to a depth of eight to ten inches. Grade bottom to provide prOper drainage. 2. Fill excavated area with twelve inches of 22-A or suitable material simulating subbase at shoulders. 3. Consolidate area to the prOper density. Lay out plots ten by eight feet using boards to confine tOp mixtures to the prOper areas. 4. Entire tOp area shall be leped from center to out— side at same gradient as regular highway shoulders. 5. TOp soil to be mixed with base material as outlined in Table I. 6. Area to be compacted in accordance with standard practice. 3. Michigan State Highway Department, "Outline for an Investigation of Turf Growth on Highway Shoulders," Project 42 E-9, Research Laboratory Testing and Research Division, August 31, 1943 ADMIXTURE 'AELS I 8| 8| 8| 8! 8! 8| OUTLINE 03 21035 BASE Incoherent Graded Sandy-Gravel 22-A Sand-Dune Sand Fox Bellefontaine 117 (13) 1257 (37) 20% 20% 20% 20% H ‘(2) T147 (267* (33) ~ g 35% 35% 35% 35% q-i :1 #13) (15) (27) (39) A 50% 50% 50% 50% (4) (16) (257 (£0) E 20% 20% 20% 20% {3; (5) (177 4 (29) (41)” .3 35% 35% 35% 35% A? (6) (is) (307 (42) 50% 50% 50% 50% (77 7197 (31) 743) Clay Muck Clay Much Clay Muck Clay, Muck a;§. 10% 5% 10% 5% 10% 5% 10% 5% (D £8 (a), (20) (32) (2+1), 4 .33 25% 10% 25% 10% 25% 10% 25% 10% 3‘3" (9) (21) , T33) (45), 35% 15% 35% 15% 35% 15% 35% 15% *(10) J (22) fl (34) (467 2 50% 50% 50% 50% «H as (11) J (23) (35) fl, (47) Pg 75% 75% 75% 75% 3 T12T (24) , (36) (48) '3 100% 100% 100% 100% m 10' 10' 10' 10' _ 10 _ 6.2:: no on we 5...... a 3 mag—3:0 73...... a... v:- oeoo. i=9 .8339 633...... we moaaoxm .2312: 20 1:65 «uni-8......“ ERR“... an? 20 29.22592. £29 .39..- ue 2:55: :25. a up .8935... ova—n .56.. 9.652 ate-a £2293. mm 53.5 .... Rafa... EOE mmxwmmu we QmmD . mm Hm mm 4 H8 is .5... 3.5 a... .25. a... 38... 9|. Sol 3 .983 Eon. .9: £3...— (20400 X0... .83.. .28: 932... 839:. snipe—Eh do... .9899. .c 5...... a c. 258?. 59:2, 2... Us. .3... use!” as 3.3.2:. .5: to»... .35... .95: 5.3 .39.». .5: .89. u: no.8 we 41:73.83...— :4. m: .399 .332... 90 3 mane—v u 3 .683: w . :23: .529 .3: 283.95 3898.19 a he. no: he gonna A r 6.3282. .929. .39.. 58:09 :2 8’9... .8. 32....» 5.3 .5238 .529 has: 3 2.9 93.38 ha..- 5...... 9.59 519... «o 2:38: :55. a .3 83.509 22.8 .83. .385. «9.9309 235... a I c a a no.2...»— ‘ A r1 m: 6.3.992. 1 i 3 .8. 5.2.3....— .589 1.. ~83- ».33 no 589 z 838.. 8.1.2.9? .9395 can I .9» 5.3 .8338 .519 5.83 “9- 809 he}. :89 .83. 5.3!. 031.... EE'lllll'l'l'l'lrl'I'E'A'T‘l'l'l'l‘I'I‘l'l'I'I'I'I'L'I'I'I’I‘E 11111.1 :2: SIM)- .88 Be. :33; 5:32....” .82.. 0.42... o. .95 no.5... «a 3 .95... E .3339... {an n . I. do. . 4.3 88.5 can is: 3:5..- Uflu 0539‘ Oil 293 3 .5334 u 303 HIS £033 9—68 ~12 .hUua—A UZ .(PZOuMJJm—m ZOmeOOKm . 25 5 ll- PROCEDURE Load bearing tests were conducted on each of the forty- eight plots. The equipment as mentioned above was a truck for load, a dynamometer to measure the load increments of five hundred pounds, and a hydraulic jack to transmit this load to a round bearing plate of one hundred square inches in area, and strain gages to measure the deflection of the plate. The deflection recording dials were set and the dial setting was recorded. The above stated increment of load was applied by tle hydraulic jack when the deflection had ceased to change, for all practical purposes, the strain gage dial reading was recorded. The load was then increased by five hundred pounds and again at the cessation of the deflec- tion change, the gage reading was recorded. The above pro- cess was repeated and recorded until a load of about seven thousand pounds had been applied, the load was then released and the elastic rebound of the soil was recorded. The dif- "zero" strain gage reading and ference between the original any successive reading will give the total deflection for the corresponding load. The above data is reproduced in part in Table II. The soil recording penetrometer originally had two heads, one a tapered point probe and the other a flat head with a circular area of 0.15 square inches. A place large enough to accommodate tle probe and the float rod foot was cleared to loose surface material such as a Dab ass. as am aaa." . x P Y 1 .176 2000 .228 2500 2250 2 .176 2000 .235 2500 2250 3 .174 3000 .215 3500 3300 4 .182 1500 .266 2000 1600 5 .178 2000 .232 2500 6‘ .128 1000 .211 1500 1450 7 .200 2000 2000 8 .131 1000 .222 1500 1400 9 .200 1500 1500 10 .200 2500 2500 11 .188 2500 .232 3000 2700 12 .141 2000 .260 2500 2250 13 .178 2000 .228 2500 2250' 14 .190 2000 .232 2500 2100 15 .200 2500 2500 16 .174 2000 .224 2500 2250 17 .164 1500 .232 2000 1750 18 .200 1500 1500 19 .160 1500 .247 2000 1750 20 .129 2000 .216 2500 2410 21 .183 1500 .233 2000 1670 22 .196 2500 .236 3000 2550 23 .168 1500 .212 2000 1865 24 .188 4500 .209 5000 4800 25 .182 5000 .216 5500 5265 eel-5% TABLE II (Continued) Plot Depth Load at Depth Load at Load at No. X(1) Depth X Y£1) Depth Y. Depth 0.2" '26 .191 2500 .213 3000 2700 27 .194 3500 .225 4000 3600 28 .173 2500 .206 3000 2910 29 .188 2000. .224 2500 2165 30 .173 3000 .205 3500 3420 31 .171 3500 .210 4000 3870 32 .200 2500 2500 . 33 .182 2500 .214 3000 2800 34 .190 4500 .215 5000 4700 35 .192 4000 .227 4500_ 4100 36 .195 3000 .219 3500 3100 37' .200 6000 6000 38 .200 6000 6000 39 .190 5500 .216 6000 5700 40 .193 5000 .214 5500 5165 41 .191 6000 .207 6500 6250 42 .130 3500 .205 4000 3900 43 .193 4000 .222 4500 4125 44 .193 5500 .222 6000 5625 45 .200 1500 1500 45 .200 7000 7000 .47 48 .200 5500 5500 (l) X.and Y are the depths and corresponding loads which, when interpolated, give the load at the depth of 0.2 inches. -14.. stones, leaves, and branches, such that a firm smooth plain was produced on the soil surface prOper. Care was taken not to disturb the turf roots, tufts, or soil other than this effort to provide a datum plain for the machine. Manual pressure was applied to the penetrometer in such a manner so as to produce, within reason and without benefit 01 gages, a slow, uniformly increasing pressure, thus the load produced by impact was eliminated. The graph (resultant of this pres- sure was produced in the coil spring and the depth of pene- tration was produced by the_differential between the float rod foot and the probe head) was recorded automatically on prepared graph paper. Six different probe heads were tried, a tapered point, and one each with an area of 0.15 square inches, as listed above. An attempt was made to have at least three curves from which a-composite "average" curve 'may be produced.for each plot. The preliminary investigation in the application of the penetrometer was to determine which‘head was best adOpted for what base material. The larger the head, the closer the trial curves will approximate one another, and the erratic nature of the curves will be lessened. As was stated previously, the- maximum size of head for a given base material will be depen- dent on the size of the Operator. The primary problem was to determine which head was to be used on each given base mater- ial, such that the maximum size which would produce the best curve, was also of such size that a depth of at least three or four inches of penetration was produced. Numerous trials were conducted in the fall months and also in the spring - 15 - .._k A “1-11 l 4 I I I u | I 1 I ‘1 Y I 3.7 l I 1 r p. I 1 I I I 4 I I 1 I . 1 I 1.1 1 1 1 v A II I 1 1 « I I 1 ‘ L ‘\ I I I A 1 I a I 1 I; ‘.‘{‘1 .fl. 3 .1 :11: I L_1 1 1.“ N k 1W . 1. 1“ :.I.1 11I.II. . .Itl11 . . . . (.151, TII...¢..I.2!III.. 'fl—yv "-- _. , H" u months to determine a tentative maximum size. This size was used and the resulting data, Table III, compiled. Moisture content tests were performed during the tests. Table V. - 17 - TABLE III DEPTH IN INCHiS Area P13? Gu§60 0“ 1" 2" 3" 4" 5" 6" 7" 8"‘ -2 42:3 at Load in Lbs. for above Depths. in. Depth 2" 1 1 0 62 76 84 88 96 98 99 99 .75 1 2 0 105 138 140 140 148 154 155 158 .75 131 1 3 0 120 154 168 174 .75 1 4 O 140 158 172 .75 1 5 0 140 168 .75 1 6 0 154 171 .75 2 1 0 93 121 150 158 .75 2 2 0 108 135 145 158 .75 136 2 3 0 131 152 153 170 I 3 1 0 85 127 150 180 .75 3 2 O 131 185 .75 3 3 O 148 152 165 174 .75 133 3 4 0 149 169 172 .75 1 O 87 121 135 153 174 .75 2 0 108 127 141 156 174 .75 132 4 3 0 148 152 160 174 .75 5 1 0 55 88 128 141 150 158 165 .75 5 2 0 91 114 126 137 158 .75 5 3 0 100 130 148 170 .75 128 5 4 0 130 141 152 165 .75 5 5 0 160 168 170 170 171 ..75 18 TABLE III Continued) DEPTH 1N INCHES P12? cu§67 0" 1" 2": 3" 4" 5" 6" 7" 8"—AEE: 1:26 at Load in Lbs. for above Depths in. Depth 2" 6 1 0 110 131 152 156 .75 6 2 0 118 132 143 154 .75 135 6 3 0 120 141 152 172 .75 1 0 104 120 136 141 150 .75 2 O 128 134 141 150 158 .75 129 7 3 0 130 134 142 160 176 .75 8 1 O 85 96 127 150 166 .75 8 2 0 90 106 115 128 142 162 .75 104 a 3 0 96 109 125 150 165 .75 9 1 0 102 138 150 162 .75 9 2 0 119 138 152 152 .75 146 9 3 O 123 156 172 180 .75 9 4 0 130 152 180 .75 10 1 0 120 149 157 157 157 157 .75 10 2 0 116 125 132 141 146 146 146 .75 148 10 3 o 155 157 154 154 170 .75 10 4 0 I155 161 162 171- .75 11 1. 120 152 160 160 160 161 161 .75 11 2 O 140 158 160 166 167 167 .75 157 11 3 0 148 161 161 161 161 161 .75 * 19 * TABLE III (Continued) DEPTH IN INCHES P18? cu§g?0" 1" 2" _3" 4" 5" 6" 7" 8" [fifga I922 at ~Load in Lbs. for above Depths. in. Depth 2" 12 ' 1 0 110 122 122 122 122 128 .75 V 12 2 0 118 133 131 123 123 .75 129 12 3 0 131 131 131 130 123 .75 13 1 O 65 88 115 139 160 .75 13 2 0 72 124 154 158 .75 118 13 3 0 87 118 138 158 .75 13 4 O 118 140 141 158 .75 14 1 O 79 98 138 144 159 175 .75 14 2 0 84 .98 131 144 170 175 .75 114 14 3 O 91 118 136 152 169 175 .75 14 4 0 127 141 150 158 168 175 .75 15 1 0 96 140 146 146 158 172 .75 15 2 0 103 130 168 .75 139 15 3 0 123 148 168 171 .75 16 1 0 61 92 124 131 152 162 171 .75 16 2 0 79 111 129 138 147 156 158 158 .75 101 16 3 0 79 100 126 138 152 170 .75 17 1 0 62 68 87 121 138 .75 17 2 O 68 92 121 130 142 .75 93 17 3 0 70 101 122 139 152 .75 17 4 _‘0 80 112 128 141 162 .75 20 - TABLE III (Continued) DEPTH IN INCHES P83? C§EY80" 1" 2" 13“ 4" 5" 6" 7“ 8" 1:38 1:32 at Ioad_;n Lb3,_for above Depths in. Depth 2" 18 1 O 70 84 99 122 140 .75 18 2 0 80 98 126 140 154 .75 106 18 3 0 91 112 126 134 144 .75 18 4 0 108 130 132 138 146 .75 19 1 0 46 70 88 118 134 152 .75 19 2 0 50 72 96 127 140 152 161 .75 82 19 3 0 68 92 110 127 131 150 .75 19 4 0 72 92 114 131 150 .75 20 1 0 36 74 100 128 142 154 .75 2O 2 O 48 78 95 120 138 151 .75 20 3 0 68 92 123 126 130 132 140 152 .75 95 20 4 0 92 112 123 140 158 178 875 20 5 0 92 119 132 144 170 .75 21 1 O 38 62 81 100 120 138 .75 21 2 0 44 82 112 135 141 148 .75 87 21 3 O 44 86 136 152 .75 21 4 0 98 118 129 148 162 .75 22 0 88 110 126 141 .75 22 2 0 108 127 138 139 154 .75 126 22 3 0 121 140 144 150 150 .75 -21 - TABLE III (Continued) DEPTH IN INCHES P13? C§gYegflr1" 22:43" 4" 5" 6" 7" fife: 83:3 at Load in Lbs. for above Depths inm Depth 2" 23 1 0 128-129 140 152 .75 23 2 0 137 138 140 140 .75 139 23 3 O 148 149 149 149 .75 24 1 0 110 110 112 119 120 130 .75 24 2 0 118 118 129 121 120 130 .75 125 24 ‘0 129 129 129 135 .75 24 4 O 140 143 151 153 164 .75 25 1 0 38 100 100 118 133 160 .50 25' 2 0 57 84 122 144 161 .50 112 25 3 0 80 131 140 144 .50 25 4 O 126 133 140 .50 26 1 0 85 96 121 139 150 154 .50 26 2 86 118 140 155 164 .50 112 26 3 0 91 122 122 141 .50 27 1 0 79 89 98 112 136 150 160 .50 27 2 0 87 128 134 148 .50 117 27 3 O 98 115 134 .50 27 4 O 100 136 154 176 .50 28 1 0 64 88 115 122 144 .50 28 0 80 108 108 128 140 .50 105 .__28 3 O _90 128 128 128 140 .50 - 22 TABLE III (Continued) DEPTH IN INCHES P13? C§EY90" 1" 2" 39 4" 5" 6" 7? 8" ffga 8:23 at peed in Lbs. for above Depths. in. Depth 2" 29 1 0 66 94 128 .50 29 2 O 66 108 118 .50 29 3 0 80 111 118 141 155 .50 113 30 1 0 60 81 110 126 141 156 .50 30 2 0 68 85 120 .50 85 30 3 0 70 90 130 131 131 .50 31 1 0 71 85 129 139 .50 31 2 0 68 91 .50 104 31 3 O 77 118 140 .50 31 4 0 68 122 121 128 145 .50 32 1 0 43 72 100 140 140 .50 32 2 56 80 85 95 137 141 .50 78 32 3 O 61 82 87 94 .50 33 1 O 38 68 78 92 137 158 .50 33 2 0 48 68 82 98 130 .50.. 71 33 3 o 58 77 93 140 155 .50 33 4 O 120 131 131 131 .50 n.g. 34 64 72 132 .50 34 2 65 76 88 13C .50 76 34 3 0 71 80 80 80 82 130 .50 ¥ .2} - TABLE III (Continued) DEPTH IN INCHES __ - 24 - P83? C§§Yeo" 1" 2" 3" 4" 5" 6" 7" 8" ffga E223 at“ Load in Lbs. for above Depths. in. Depth 2 35 1 o 68 7o 80 .50 A 35 2 O 68 78 78 88 101 116 144 .50 35 3 O 86 87 87 88 125 .50 101 35 4 o 86 142 142 135 .50 35 5 o 96 130 119 119 119 140 .50 36 1 O 68 68 7O 7O 55 .50 36 2 O 77 77 77 77 77 67 87 .50 88 36 3 o 82 82 82 82 82 82 82 130 .50 36 4 o 130 125 116 .50 37 o 91 125 145 168 .50 ' 37 2 119 128 144 168 .50 130 37 3 o 131 137 150 152 168 .50 38 1 O 75 132 139 158 .50 38 2 o 115 128 132 .50 130 38 3 O 126 130 133 150 .50 39 1 o 100 142 158 .50 39 2 o 115 134 142 143 165 .50 138 39 3 O 128 138 148 168 .50 TABLE III (Continued) Plot Curve DEPTH IN INCHES Area Mean No. NO. 0" 1" 2" 3" 4" 5" 6" 7" 8" -—2 Load at Load in Lbs. for above Depths. in. Depth 2" 4o 1 o 78 118 137 158 .50 3 40 2 0 88 132 158 .50 136 40 3 0 99 131 155 .50 4o 4 o 120 162 .50 41 0 6o 94 90 138 .50 41 2 72 94 128 148 .50 41 3 78 94 132 144 154 .50 42 1 O 72 148 158 .50 42 2 o 75 78 90 110 133 140 170 .50 42 3 o 90 130 150 .50 133 42 4 o 141 144 .50 42 5 o 140 168 .50 43 1 74 90 120 150 .50 43 88 115 128 .50 . ,102 43 o 91 120 138 150 .50 44 1 o 71 118 146 165 .50. 44 2 o 76 86 122 131 140 .50 44 3 O 76 130 160 .50 44 4 O 89 128 170 .50 126 '44 5 o 140 170 .50 - 25 - TABLE III (Continued) Plot Curve NO. 0" >1" DEPTH IN INCHES. ___Area Mean No. 2fi——3“_'4“"Efi'6" 7" 8" IE? ggaihagfl _;oad iniébs. for above Depths P 45 1 o 78 99 130 144 152 .50 45 2 o 88 120 120 125 140 .50_ 125 45 3 'o 118 122 133 133 140 .50 45 4 o 120 141 152 152 152 .50 45 5 o 131 142 150 .50 46 1 o 70 80 120 135 160 165 ' .50 46 2 O 74 102 155 .50 46 3 o 95 124 145 .50 107 46 4 To 110 110 110 112 122 142 .50 46 5 o 120 120 119 121 .50 47 1 O 68 81 81 94 .50 _47 ’2 o 76 90 9o 90 .90 .50 47 3 o 81 90 92 108 .50 101 47 4 o 120 120 119 119 130 .50 47 5 o 122 120 130 148 .50 48 1 o 138 42 42 42 42 42 42 42 .50 48 2 o '40 58 58 6o 60 60 6o 80 .5o 48 3 o 68 68 122 122 ..50 73 48 4 0 7o 70 .72 84 .50 48 5 o 73 73 73’ 73 73 .50 48 6 o 129 129 .50 - 25 - TABLE III (Continued) DEPTH IN INCHES Plot Curve Area Mean NO. NO. on 1" 2" 3" 4" 5" 6" 7" 8" ___2 Load at! Lgad in Lbs. for above Depths. in. Depth 2" 1 1 o 52 61 87 116 132 157 168 169 .75 g 1 2 0 95 118 122 130 130 130 121 121 .75 114 1 3 0 95 127 159 167 178 .75 2 l O 73 121 135 141 162 174 .75 2 2 0 80 114 136 158 .75 2 3 O 122 160 180 .75 3 1 O 111 171 171 171 .75 3 2 O 122 153 180 .75 168 3 3 O 167 180 .75 5 1 O 92 120 135 145 158 .75 5 2 O 97 127 136 147 169 .75 127 5 3 O 121 134 154 171 159 .75 11 1 O 102 127 127 127 127 125 .75 11 2 O 130 151 150 160 .75 146 11 3 O 136 151 152 152 152 155 .75 16 1 0 78 98 126 137 157 171 ..75 16 2 O 87 110 127 138 159 173 .75 106 16 3 O 88 110 128 141 160 175 .75 19 1 0 35 58 79 112 138 151 .75 19 2 O 47 67 85 125 140 168 .75 65 19 3 O 48 69 85 125 152 .75 TABLE III (Continued) W IN INCHES P42? 018390» 3.. 4.. 5.. 6w if? 42:: .1. Load in Lbs. for above Depths. in. Depth_g: 22 1 o 82 121 136 170 .75 ’ 22 2 o 98 130 138 146 151 .75 129 22 3 O 117 135 158 178 .75 1 1 110 .75 140 l 2 150 .75 1 3 160 .75 2 1 140 .75 2 2 - 142 .75 142 2 3 144 .75 3 l 134 .75 3 2 i 162 .75 156 3 3 172 .75 1 120 .75 2 122 .75 124 3 130 .75 5 1 102 .75 5 2 140 .75 137 5 3 153 .75 5 4 154 .75 - 28 _ TABLE III (Continued) Plot Aaurve Load in Lbs. Area Mean No. No. at 2“ Depth -2 Load at in. Depth 2" 6 1 137 .75 6 2 141 .75 145 2 103 .75 113 8 l 114 .75 8 2 141 .75 138 8 3 158 .75 9 1 152 .75 9 2 _ 156 .75 156 9 3 160 .75 10 l 154 .75 10 2 158 .75 156 10 3 11 1 ' 153 .75 11 2 165 .75 162 11 3 168 .75 12 1 140 .75 12 2 152 .75 146 -29.. TABLE III (Continued) Tiot Curve _*Load in Lbs. Area Mean No. No. at 2" Depth -—2 Load at in. Depth 2" 13 1 128 .75 13 2 142 .75 137 13 3 142 .75 14‘ 1 123 075 14 2 127 .75 131 14 3 143 .75 15 l 125 .75 15 2 128 .75 130 15 3 138 .75 16 l 93 .75 16 2 ' 112 ‘ .75 110 16 3 125 .75 17 2 112 .75 114 17 3 120 .75 18 ' 1 128 .75 18 2 128 .75 129 18 3 130 .75 19 l 80 .75 19 2 80 .75 93 19_, _3 118 275 - 3o - TABLE III (Continued) _P10t Curve Load in Lbs. Area Mean No. No. at 2" Depth -—2 LOad at in. Depth 2" 2o 1 117 ' .75 127 20 2 137 .75 21 1 . 115 .75 116 21 2 118 ' .75 22 1 127 .75 130 22 2 133 .75 23 1 146 .75 151 23 2 156 .75 24 1 160 .75 24 2 153 .75 156 24 3 155 .75 - 31 - TABLE IV DENSITY TESTS AND MOISTURE CONTENT November 17, 1949 Plot Density No. fi/ft? 1 96.5 12 96.4 19 105.0 37 109.2 48 102.8 ’ ‘ ' ' ‘ ‘ Enamér' i949 """"" Plot No. Density Moisture Content #/ft3 percent 1 90 10.4 2 96 1 9.7 3 93 ’ 13.0 4 101 9,2 5 91 .5.3 6 94 5.3 7 97 5.4 8 90 5.3 9 90 5.3 10 95 5.3 11 85 5.6 12 97 5.4 13 101 4,0 14 101 4.0 - 32 - TABLE IV (Continued) DENSITY TESTS AND MOISTURE CONTENT Summer 1949 -. - EDensity Moisture Content Plot No. #thB per cent ' 15 -101 3.8 16 101 4.0 17 99 l 4.1 18 97 4.0. 19 97 3.8 20 100 3.32 21 95 4.0 22 99 5.3 23 91 5.3 24 97 5-4 25 122 1 3.0 26 120 3.0 27 119 3.2 28 120 3.0 29 116 3.1 30 117 2.61 31 116 3.0 32 112 3.5 33 115 3.4 34 116 2.0 35 101 3.75 .36 91 4.0 ..33- TABLE IV (Continued) DENSITY TESTS AND MOISTURE CONTENT Summer 1949 Density Moisture Content Plot No. fi/ft3 per cent, 37 116 3.0 38 110 2.83 39 109 3.0 40 112 2.81 41 114 3.0 42 106 3.0 43 115 2.5 44 114 3.1 45 108 2.2 46 108 3.2 47 87 3.3 48 85 4,5 Plot No. April 6‘4-1aTe9351Content % Over Dry. 3 13.25 5 10.20 11 12.45 16 4.56 19 ‘ 6.10 22 8.10 29 13.65 30 13.25 36 17.41 - 34 a TABLE IV (Continued) April 7, 1951 Plot No. Water Content % Over Dry. 42 . 16.6 48 18.12 -35... DISCUSSION It may readily be seen that a tapered point probe would be greatly influenced by local fluctuations due to the dif- ference in the size and arrangement of the interstices, and by the size, packing and structure of the material encounter- ed. This was found to be the case. The tapered probe did not penetrate in a straight line, but rather it slipped off the larger pieces of material into the less resistant voids or finer materials. This action was most noticed in the gravely material and was very evident to the Operator. The 0.15 square inch head functioned much as the taper- ed probe head, but to a lesser degree. Maximum penetration of twelve inches could be obtained in all materials encounter- ed but the curves for any one plot did not approximate one another to any marked degree. The next probe used was the one Square inch head, This head allowed no penetration on any of the plots in question. It is therefore evident that the maximum head for this pro- blem must be contained in a head size less than one square inch. It must be kept in mind that the maximum load, not including impact, will be produced by an average sized Operator and will be somewhere around one hundred and sixty pounds. The probe with a head of 0.75 square inch was the next size to be tried. This probe did not give maximum penetration, but rather a penetration of about six inches or less., 0c- casionally a depth of eight to ten inches was obtained but - 35 _ but this was the exception rather than the rule, and the depth of penetration was considered excellent if it reached a depth of four to six inches. This probe performed very well on the sandy base material, irrespective of admixture or type of turf material, it was tentitavely selected as the probe to be used for more comprehensive tests. The gravel based material produced curves of a very erratic nature. On one plot (48) the curves produced by this head varied in maximum penetration from one at depths each of one—half, one, six and eleven inches. Many of these plots gave maximum penetration of less than three inches. It was, therefore, concluded that this 0.75 square inch head was applicable only to the sand based plots. (plots 1-24). A probe with a head size of 0.5 square inches was then tried on the gravel based plots. Penetrations were confined, as a rule, to a depth of six inches, more or less. This was encouraging, but the most noticeable feature of these curves was the complete lack of order. Any two curves, in the same plot, which approximated one another was purely accidental. ‘ The eccentricities exhibited in the curves of the gravel based plots may be due to the great range of heterOgeneity of the soil and rock material encountered. It may be seen that the head of a specified area may rest on a rock particle of almost any size, which is contained within the specified soil fractions. This may produce a pseudo-head of unknown cross- sectional area, with this pseudo-head changing at each trial, thus producing a set of curves that are not even remotely -37... related. 0n the basis of the results obtained on the gravel based plots, and the above discussion, eXperiments on the plots 25-48 were discontinued and the remaining work con- ducted on the sand based plots. The formula for the modulus of subgrade stiffness, "k" is: k 3 P AZ Where: k : modulus of subgrade stiffness in #/cu. in. P 3 load in # A = bearing area in sq. in. Z = penetration in inches. As the penetrometer is a supplement, rather than a sup- planter to the conventional load bearing capacity method, it was thought advisable to solve for only one unknown, P, and then to use that unknown in the above formula to obtain "k", and to use the load bearing plate area and a predetermined "Z". Thus the "k" obtained would conform to existing "k" values. The area of the load bearing plate is constant, (100 square inches) and the "Z" penetration used was 0.2 inches. The load "P" must be obtained from the penetro- meter curves, and toward this end the_remaining work was directed. After but little experience with the penetrometer one may "feel" the various resistances to penetration encountered. Turf roots are, as a rule, the first major resistance en4 countered, penetration seldom exceeds one inch at this . point, while resistance to penetration, in pounds of pressure, increases rapidly. As the root mat slowly yields to the pressure the lepe of the curve changes and when the root mat has been completely passed through, at a depth of about one to two inches, the slope of the curve changes abruptly and should remain fairly constant to a depth of six inches, where the probe passes from the stabilized material to the subbase material. These transitions are not only readily felt by the Operator, but they are reflected clearly in the graphs. The ability of the various base materials and admixtures to support plant life, the ability of the various turf mater- ials to survive in their environment, and the state of growth or decay due to seasonal variations will be reflected in the first part of the curves. During the period of turf degeneration little effort was required to penetrate this turf root mat with the 0.75 square inch head, yet during rejuvenescence the mat was penetrated only with the utmost difficulty. It is self-evident that the problem under dis- cussion must be carried through a seasonal cycle before definite conclusions may be drawn. Due to the above facts, and to the increasing deviation of the curves as the depth increases, it was thought that a depth of two inches, and its corresponding load, was an appropriate point to conduct the correlation problem. At this point the turf roots will not have a fluctuating affect on the points in question, and the curves approximate one another to a better degree. -39.. 6 117.35 .5 ~ 5' I Plot 6 Head size 0.75 sq. in. Spring '51 Sr of” - 1.80 o ' \F‘ 160 O l 40 _ U‘ 8» 1 .— (9 El? 2 (r.- ” [n2 3‘89. . 41-. U4 2 ‘22 E: 2: {Y1 e: 54': 2 are rs CL.- 1:19 . ,. .. 'JJ 23 c: 6125456789101112 DEPTHIBIRCHES Plot 48 Head size 0.50 sq. in. Spring '51 .-. v «.p... ......- q. . . 4.....- N -.-m —- «v. o a..- —- -1 -'-..' The problem is further reduced to the solving of the equation: M - P PO Where: P : load in lbs. from the load bearing tests at the depth of 0.2". P0 = load in lbs. from the penetrometer curve at a depth of penetration of 2.0". M = the tentative constant, which when multiplied by PC will give the load P to be used to solve for the modulus of subbase stiffness. The arithmetic mean was found to be more consistent in the determination of the composite curve, than either the mode or the median curve, therefore, the mean curve was used. To determine "M" in the above formula, the mean load for any one plot at the depth of two inches was computed and the load bearing test load at a depth of 0.2 inches was recorded, and "M" computed. To illustrate: Plot 1. P = load at 0.2" = 2250 lbs. - from 1oad bearing tests PO= mean load at 2.0" = 131 lbs. from penetrometer V curve. M - P/Po I 2250/131 i 17.2 If another set of tests are performed on plot 1 and the mean load at depth two inches is multiplied by the above "M", P should be obtained. This was not the case. Assuming an "M" Of 20, and a machine error of 10%, assume one set Of curves may be ten pounds high and the next set ten pounds, - 41 - which is not at all unreasonable, but most likely it is the rule, that will give a total machine error of twenty pounds, but when multiplied by "M" it produces an error in the answer of some four hundred pounds. This includes only the mechanical error, not the human error or the local soil ir- regularities. It may be seen then that the small errors con- fined, if possible, to the second place to the left of the decimal, will automatically be shifted to the third place by the simple application of the multiplying factor "m". To reduce the induced error caused by the movement of the decimal point to the right, it was thought that a common or Briggian logarithmic system may be employed. The small original error, when converted to a log, may be eliminated all together, or at least greatly reduced by its position in respect to the decimal point. P0 is computed from the penetrometer graph composite curve, and is then set in common lOg form. The multiplying factor "M" is then found by the above formula, and P found as stated above. It was found that an error in the unit or tens digit of P0 , when converted to lOg form and multiplied by "M", was still contained in the original place with respect to the decimal pOint. To illustrate: Plot NO. 11, trial 1 and 2. P0 = 157 lbs. P0' = 146 lbs. a difference of 11 lbs., P . 2700 lbs. M = 17.2 P a MP0 lst trial: 2700 lbs. P : (17.2)(157) -42.. 2nd trial: P = (17.2)(146) = 2520 lbs. The 11 lbs. error thus gives a 180 lbs. error in the final results. By common logs: Same conditions as above M = 1230 error : 11 lbs. Po ' 157 lbs. P0' = 146 lbs. log 157 = 2.195 log 146 = 2.164 P = (log.Po)M P = (1230)(2.195) = 2700 lbs. P - (1230)(2.164) = 2670 lbs. The 11 lbs. error is thus kept in the second place to the left of the decimal and the final figures differ by but 30 lbs. Tentatively, then the log system is the one to be employed. The tentative multiplying factors "M", are listed in TABLE V. - 43 - TABLE V Plot Tentative No. M 1 1070 2 1055 3 1500 4 760 5 1180 6 675 7 965 8 670 9 690 10 1145 11 1230 12 1050 15 1070 14 1010 15 1175 16 1110 17 875 18 720 19 930 20 1180 21 830 22 1210 23 865 424 2235 - 44 - RESULTS The results are presented in tabular form in Table IV. ‘Representative composite graphs are shown on page 40. An original graph is shown on page 16. For the plots containing a base of sandy material (plots 1-24), the three-fourths square inch head functioned satisfactorily. No decision on the prOper head size for the plots containing coarse gravel material (plots 25-48), was obtained. _ 45 - CONCLUSIONS Results from the plots containing coarse gravel material (plots 25-48), were erratic and inconsistent. It is not thought that further work on the problem in- volving these plots will produce a correlation that will be reliable. The correlation involving plots containing sandy materials (plots 1-24), produced excellent results and it is thought that work on this problem is worthy of further consideration. - 46 - SUMMARY To obtain the load bearing capacity for a given con- structed project of sandy material and a binder, at a given location, for example an airport, a load bearing test may be conducted by conventional means. From this test "k" may be computed either from a specified depth of deflection or from a specified load P. The penetrometer may then be Oper- ated in the immediate area of the load bearing test spot, and from the graphs, a composite PC at a specified depth of penetration may be obtained, and "M" computed. The penetro- meter may then be run in the area contained in the given project, of course assuming the same base materials and binder, and constructed in the same manner. Thus after one or two load bearing tests have been conducted, a quick check of the complete project may be performed with nothing more than a slide rule (for calculations and logs) and a penetro- meter. Any deviations may be quickly rechecked by the penetrometer and, if need be,a load bearing test may be conducted on that spot.in question. For a conservative estimate, including the running of the penetrometer and the computations, a person should be able to check twenty-five spots, more or less, in a half a day. - 47 - BIBLIOGRAPHY Campen, w. 5., and Smith, J. R., "An Analysis of Field Load Bearing Tests Using Plates", Omaha Testing Labor- atories, Highway Research Board, Division of Engineer- ing and Industrial Research, National Research Council,Vol. 24,1944. Field Manual of Soil Engineering, Revised Edition, 1946, Michigan State Highway Department, (Michigan State Highway Department, Lansing, Michigan). Finney, E. A., "Shoulder Construction Practice in Michigan", Report to the Highway Research Board, Roadside DevelOpment Subcommittee on Shoulders, Department of Design, Washington, D. 0. Dec. 6, 1948. Highway Practice in the United States of America, Public Roads Administration, Washington, D. C. 1949. Michigan State Highway Department, "Outline for an Investi- ation of Turf Growth on Highway Shoulders, Project - 2-E-9, Research Laboratory, Testing and Research Division, August 31, 1943. PCA Soil Primer, Portland Cement Association, Chicago, Ill. 1950. Rooertson, L. S. and Hanson, C. M., "A Recording Soil Penetro- meter" ,Reprinted from Michigan Agriculture Experi- ment Station Quarterly Bulletin, vol. 33, No. 1, August. 1950. Taylor, D. W., "Fundamentals of Soil Mechanics", John Wiley and Sons, New York, 1948. Terzaghi, Karl, and Peck, R. B. "Soils Mechanics in Engineer- ing Practice", John Wiley and Sons, N. Y., 1948. Urquhart, L. 0., "Civil Engineering Handbook", McGraw-Hill 0o.. Inc. N. Y., 1950. as . ‘ b 1 ‘ I . . . , ». ‘- - "‘5 k x. .4 V." ' E “111111111 @1111 141111111?!”