BULLETIN 229 (Reprinted 1937) MAY, 1929 Methods of Research in Soil Dynamics as Applied to Implement Design By M. L. NICHOLS ALABAMA EXPERIMENT STATION OF THE ALABAMA POLYTECHNIC INSTITUTE M. J. FUNCHESS, Director AUBURN, ALABAMA Methods of Research in Soil Dynamics as Applied to Implement Design By M. L. NICHOLS Agricultural Engineer BULLETIN 229 (Reprinted 1937) MAY 1929 Table of Contents Page INTRODUCTION -3 DETERMINATION OF SOIL PROPERTIES AFFECTING TILLAGE EXPERIMENTAL SOILS AND METHODS OF ADJUSTING MOISTURE CONTENT----- -_4 7 EXPERIMENTAL METHODS FOR STUDY OF SOIL PROPERTIES_--_S Resistance to Chtion Compression----------------------------------8 ------------------------------------------ 15 Chesion-------------------------------------------------15 EXPERIMENTAL METHODS FOR STUDY OF SOIL-METAL RELATIONSHIPS--------------------------------------------- 20 Friction-------------------------------------------------20 Adhesion ---------------------------------- Effect of Shape of Surface Applying Pressure to Soil APPLICATION TO DESIGN-------------------------------------- ------ 22 24 26 SUMMARY ---------------------------------------------------- 27 Methods of Research in Soil Dynamics as Applied to Implement Design By M. L. NICHOLS Agricultural Engineer TILLAGE is the greatest power-consuming operation on the farm. To date the design of equipment for this purpose has been largely empirical and, for this reason, is subject to question. Moreover, there seems to be little practical connection between the vast amount of miscellaneous information accumulated by soil technologists and the practical problems involved in soil preparation which face the designing engineer. The object of the work in soil dynamics described in this bulletin is to find a basis for design of tillage implements. Therefore, in attacking this problem, it appeared necessary that the properties of soils which affect their reactions to tillage implements be determined and methods for studying them evolved. This has been done and the following pages describe the methods and apparatus used. A sufficient amount of data obtained by these methods are included to allow the reader an opportunity for appraisal of their value. It is quite evident that the soil is a constantly varying material. It varies both in physical structure and in chemical composition in different localities, and even within a field itself. Any given soil varies from time to time in response to the various forces acting upon it. If progress is to be made in implement design these facts must be recognized and the solution sought by the isolation of general or fundamental laws, rather than by attempts at empirical measurement of unknown complexes by the so called "practical" field trial method. This conception is of the greatest importance for only by understanding the cause of a soil's reaction to given force application can the results of any operation be accurately predetermined. Accurate knowledge of reaction to force application is, of course, the essential basis of engineering design. While apparently little attention has been paid to the general principles of a soil's reaction as such, it is quite evident that such principles exist, else there would be no basis for the judgment evidenced by experienced handlers of the soil. Moreover, extensive experimentation has developed certain facts which give a means of approach to these problems. The most The writer wishes to express his appreciation for the assistance of Dr. F. W. Parker and Dr. W. H. Pierre of the Soil's Laboratory of the Alabama Polytechnic Institute for chemical analysis of soils and for many valuable suggestions; and to Mr. R. W. Trullinger of the office of Experiment Stations, Washington, D. C., for valuable assistance in organizing the attack on the problem of Soil Dynamics. impiortant of these is the explanation of the physical properties ot the soil on the basis of soil moisture films (1 ) and the effect of coilloidal material (2) . While chemnicalI comp1osition is important from a crop's standpoint, in general its pihysical effect is relatively unimportant wvhen compared to such factors as p~articl e size andl moisture content. Th is appears to be quite generally agreed1 upon by soil tech nologists, except in the case of the smallest sizedl particles wxhere chemical forces are more evident. Even h ere, in onur com mon field soils, the q uaintitv of colloidal material appears to b)e ot more implortance than the comp losition of the colloid. The most notalIe exc eption to this genera Ivxiewploint applears to be with colloidal materials wNhich vary wxidely in their ratioi ot silica to alIuminum and iron. DETERMINATION OF SOIL PROPERTIES AFFECTING TILLAGE The determination ot physical lprolperties entering into impilement dlesign wxas inade by the following method: A small, nickel plated plow xxwas mounted ini a box ( Figuric 1) so that its point, shin, and heel ran beside a piece of heavy 1)1ate glass. The moxveinent of the soil could be observedl througzh the glass andl measurements taken where desired. A prismatic binocular microscope, mounted to mox e with the plow, madle it p~ossile to observe the soil's action at any p~oint desiredl. The plow was drawn by a cord wvhich Nvoundl aroundl a s1)oo1 so arranged th at a wide variation in sp~eed Could lie obtained. Slow sp)eeds5 «veic usedl for microsco)pic studlies. Vaiious soils xv-eie used and with each soil the plo-wing \vas (lone at different depths. It wxas found that the reactions to the plow were siniilai in all cases wxh ich p~ermitted a general classification of properties. Foi dlitfeient soils. moisture percentages, and structures, lbhe dlifferent soil lopeities v aied in inipoitanice. The theory of p~lowv action, generally used as an (exIIIanation grap waA- takeii. The p]lw lantdti e p& oint aot tact writh the vlass side otf the stoil hox. Att in are in Con- of its pulverizing qualities, is that the lower soil in passing over a curved plow surface traveled farther than the surface soil. It was thought that the soil dividing into different layers, which traveled at different speeds, accounted for the pulverizing action of the plow. This action is shown diagrammatically in Figure 2. Under these conditions steeper curvatures of the plow would produce greater pulverization. The observed action of the plow is quite the opposite of this theoretical action in that the pulverization is at right angles to the tangents of the curve, instead of parallel to them. This is shown diagrammatically in Figure 3. The following description of the soil movements will explain this and show the basis for the selection of soil qualities for study. As the point of the plow advances in the soil its bluntness (compared With soil particles) catches a part of the soil which it drives ahead in the form of a wedge. This soil wedge is compressed until its resistance to compression equals the resistance encountered in driving it into FIGURE 2. Theoretical soil movement by plow. the soil. As this wedge advances through the soil there is a constant rolling motion of the soil along its sides due to the interlocking of particles. This brings into account the force required to roll or slip soil over soil, which may be termed internal frictional resistance or "shear". "Shear" is usually preceded by a compression of the soil when the soil moisture is within what is ordinarily considered the Obviously, plowing range. a determination of the coefficient of internal resistance, compression values of the soil and the friction of soil on particular metal would make it possible to determine the value of sharp FIGURE 3.-Actual soil movement by points on a plow. plow. A. Soil wedge at point of As the point of the plow plow. advances through the soil the B. Shear plane. inclined plane following it compresses the soil upward and forward. When the resistance to compression exceeds the shear value, the block of soil is sheared off and moves upward and forward as a solid unit.* The force required for this action depends upon the pressure of the soil and the friction of the metal and the soil. The forward movement of this block of soil is, of course, due to *The lateral and twisting movements enforced on the soil by the various curve combinations of common plows are not mentioned as this discussion of soil movements is limited to those which may be observed by the use of the apparatus described. the friction of the soil and metal and continues until the resistance of soil ahead of the block becomes equal to, or greater than, the friction between it and the metal when the soil is moved upward. When the plow shin is in the form of a curve, constantly increasing in steepness, the force required to move the soil up the plane is constantly increasing and the forward resistance of the soil must be greater to cause this more rapid movement. As the shin becomes more nearly perpendicular it will be noted that the surface applying force is nearly at right angles to the surface at the point. The pulverizing action of the plow is then due to the angles of the surface applying force being changed through the entire range of the curve of the plow from the nearly horizontal point to the vertical upper shin. This may be more clearly explained by calling attention to the amount of forward travel in relation to the upward movement enforced on the soil by the plow's curvature. As the soil passes over various parts of this curve it is forced to slip over itself at a constantly varying rate. It should be noted that these force applications and resulting reactions are vastly different from, in fact being almost at right angles to, the forces of inversion to which is commonly attributed the pulverizing action of the plow. When the soil particles are cemented together by drying, this wave of shearing forces is still generally apparent, but due to areas of high and low bondage, the soil breaks into irregular lumps which grind over one another producing the same general result. However, in a cemented or dried soil the resistance to compression is high in proportion to the shear value and the pulverizing effect is limited to the fragmentation of the lumps. When a soil is in proper condition for tillage the stresses flow through the entire soil mass. On the other hand when a soil is so filled with moisture that there is free water present, or where the films are large enough to be easily broken, the waves of compression and shear produce the injurious puddling effect shown by the shining furrow slice. The accompanying chart of the variables entering into these reactions was prepared from the above and similar studies. It is apparent that a knowledge of the interrelations of these elements is what is sought for they constitute a cause and effect relationship. The general procedure consists of measuring and evaluating the results obtained by the various elements of implement design on soils of known composition whose dynamic properties have been determined by the methods which follow. CLASSIFICATION CHART Variables entering into soil dynamic studies. Soil structure uniform; cementation zero. Primary Soil Factors (Measurable or Design Variables (ConDynamic Properties of Soil Dynamic Resultants controllable) Particle size Colloidal content Moisture (percentage) Apparent specific gravity (State of compaction) trollable) Kind of metal Polish Bearing area Curvature of surface applying force (Measurable) Coefficient of internal resistance (or shear value) Friction Resistance to compression Cohesion (Measurable) Fragmentation Arch Action Compaction Shear Organic matter Chem. composition of colloid Adhesion Moment of inertia EXPERIMENTAL SOILS AND METHODS OF ADJUSTING MOISTURE CONTENT As previously stated, the soil's reaction is largely due to particle size and moisture content. These properties may either be varied in synthetic soils or determined in natural field soils. Experiments at this laboratory indicate that practically the same results are obtained by either method but the use of synthetic soils is preferred as a desirable range of variation can be more easily obtained. These synthetic soils are prepared by drying and mixing a heavy colloidal clay (amount and composition of colloid known) with sand. The general plan of procedure being followed is to determine the physical reactive forces from a series of soils ranging from heavy colloidal clays to coarse sands at different moisture percentages. The mixtures of 1/3 clay: 2/ sand, and 2/3 clay: 1/3 sand, when taken with pure sand and pure clay, gives a satisfactory range of variation for most work. One of the most difficult tasks in physical studies of this kind is bringing soil to a desired uniform moisture content. After much experimentation a method was evolved which is rapid and satisfactory. The dry soil is placed in a box or large container where it can be stirred continuously. As this stirring proceeds (See Figure 4) the soil is sprayed with a steam jet which passes over an atomizer filled with water. The steam jet and atomizer are so adjusted that the steam will vaporize the spray it throws from the atomizer and as a result the vapor is at so low a temperature that it condenses on the soil particles giving a very uniform moistening. As the moistening goes on the soil's temperature is slightly raised which materially decreases the danger of puddling. The slow, almost imperceptible, and uniform change in soil color gives an indication of the success of this method of moistening. Moreover, soils can be brought to a rather high moisture content and remain unpuddled in spite of the continued stirring. A handful of soil so moistened, which has been compressed into a ball, can again be crumpled into its fluffy, finely divided state without the formation of lumps or puddled particles. .5 I ~ rmnws Ih; FIG . 1i eiStance oel~to Compresson ratCll all of~i our tillage imp ' slens aie giveil levoicae for appllying pressure in different ways. It has b)eeni shown that the reactions of soil to forces applied by a plow surface consist ot ia resistance to comn lression antd, at the end of th is compression, a shear. It was also suggested that each of these reactions under conditions (ot uniform Soil Structure indIicated1 constant prope~rties of a soil at any given moisttire p~ercentage. This is to be expected as these reactions are due to size of particle and moisture content. The reactioin to pressure, therefore, is of the greatest imp lortance in studlies of the dlynamnic pro perties of the soil. The equipment designed for determining resistance to compression consists of a two-inch cylinder (Figure 5) in which moves an air tight piston (A) ;to this cylinder is attachedl a 8 ro(d w\hich p~asses through guid~es to a brass5 1)1u nger two inches in dliameiter (h). TI'is f)1UiioVet compre~~sses the soil in a briass ring (C() or con- tainier 2 t 8 inches in di a meter and .984 inches (250 AI M.) deepi. C'ompressed alir is used in ap)lyIinog the pries- sine. W hich can be iead diiectly tiomi the gauges in pounds(1 per squ are inch as the plunger and piston aie ot the same (diamneter. The priessure is conflt 1 01 1ed( th rough a pressure regulator v'alve (1)) bet ween the piston and the aii ipumpi tank. Al easurern ( ts of the amniouint of compriession are nmade bjv nmeans of a mnicriom etei scriew (I) reading t o t h e hundiredith part of a m illimeter. The micrometer is attached to an Aimes dial (F) indicatoi ini contact with the upper suirface of the lpltilig-ei so that the aimounit of comrn uession) (an iile a1cc uately m easn ied. Mlicronmetei reainilgs are tiaken at intervals oif fi\ e 11 R 5. IA ar itus I naition Co iii pi of i o tI'U 1)1 I i to - Itesjstallc( A. I Air Liston cVlinder IK. Pluner (V as pis (t l o iti, C. Si S n diam E..MicrUoete .cwX V" Anm Diaj jl. pounds lnessuie and the weight of the soil is takeii before or after the seiies of tests. The applarent specific giavitv can: be dIeterminedl at any step) in the compr~essioni as the volume and weight of the soil mass is known. With foreces of the magnitude used in till age the relation of foie appl11iedl to comnpression m ay lie sh own by a smooth curve. See Figures 6 and 7.) Since the state of compriession of test samp1)1es vary, dup1licates ot the soil iun for check wxill give similar curives but uniiiformlIy higher or lower due to varying masses of soil. It is qjuite implortant that curves of this kind be studiedl careflu llv to dIiscover math ematical expriessions for functional ie- lations..Not only does t he discovery ot em piirical foim ul as by .c uirve fitting"' serve as a p~ractical means of calc ulating apprItoxi- mationis to other values of the variables b)ut its foim may suggest hitherto unknown laws connecting the variables which can be established by further investigation. While these studies have not been of sufficient scope to warrant the dirawinhg ot final concI usions, the followving dliscuission ot these curves with tentative conec1lusions is given as an ilustration of metho(d(. 9 All of the curves obtained in these compression studies appear like hyperbolas in that the value of Y approaches a limit as the value of X increases. In attempting to fit a formula to these curves considerable difficulty was experienced in that hyperbola formulas would fit the upper part of the curves closer than those of any other curve considered but would not fit the lower part. This indicated that possibly there were factors other than the film moisture entering into the formation of these curves. Until the pressure had produced a certain maximum contact of particles where laws affecting the elasticity of the films would come into full effect, it was concluded that we should expect no definite law of reaction as the first reaction is due to structural rearrangement. Since soil structure in general is the result of action rather than an inherent property of the soil the lower part of the curve becomes of questionable value so far as general laws are concerned. From a practical viewpoint, however, this part of the curve is quite important, as pressures are far above the average pressure exerted by the plow. Apparently a plow running 6 inches deep and cutting 10 inches wide (60 sq. in. furrow slice) exerts a pressure sufficient to rearrange the structure to the extent of giving full film contact to only a small portion of the soil it turns over even in optimum plowing conditions. Therefore, tilth must be produced by breaking the soil up so that natural agencies can work on it. The question of pressure as it affects design narrows itself into the following parts: (1) pressures should be applied to produce the maximum fragmentation with minimum total pressures, (2) pressures should be applied so as to minimize danger of rupture of moisture films in soils with a tendency to puddle. The first question has to do with RR sharp cutting edges and the arrangement of the plow curves to throw the forces applied into effect at the right time and in the most effective direction. The second question would seem to indicate limitations of these curvatures and definite relationships with soil types.. Further evidence of thevalue of these curve studies may be shown by comparing the curves obtained from field soils. The curve produced by a white coarse sand from Illinois (Figure 7, curve No. 1) rose rapidly at first and flattened off to practically horizontal at 8 per cent compression at a pressure of ( M NNS ° .... O .... e e SH.SO 3 2 I 0 10 'Je FIGURE 6 30 lbs.per 4 square N o° 7 CA-Mo AC 7° inch. W i t h a "push" soil from /Minnesota 2very / -3 ) . Gwith LY Yo o , 3 . otion (Figure 6) which was high in organic matter, 7 per cent compression was obtained at 7.5 lbs. pressure, an d thirty pounds p r e ss u rehad .compacted the soil 24.5 per cent no indicaof the curve flattening. -SAo The Zsand will compact but litit tle before shears while the black mucky soil o a/ /,S o BS.__ 23 _ will o _considerably "ball up" be- an d place. compress The fail- fore shear takes FIGURE 7 ure of the"push" soil curve to flatten out indicates that there is a third phase of compression to be considered: namely, that the soil material itself may be compressible. The measurement of this property of compression resistance is, of course, of little benefit by itself but in studying soil movements and the reaction of soil to forces it is of the greatest importance. For example, in studies of stress distribution by the plaster cast method, the pressure exerted on any point in the soil can be determined by measuring the amount of compression and, as will be shown later, the arch action of a soil in the path of a lug or other moving body is greatly affected by this property. Arch Action The pressure exerted by a body being pushed into the soil, or resting upon it, is distributed out over a considerable area. This arch-like distribution of stress is quite an important consideration in tractor lug action, plowing, and all other soil operations. The resistance to the point and other cutting edges of the plow as previously shown is materially affected by this property. Before entering upon this discussion however, there should be a clear understanding of the difference between arch action and compressive action as used in this report. By arch action is meant 11 the tendency of a soil to dlistribute 01r rector' out a comrnlessiv'e force ;by comrnlession is meant the reduct(in in volumirne iiy a coinpressive force. Two methods of studIyi ng this act ion have been d eve loped at this instiltution. the plaster (cast method (8)) andu a Visual met hod. Inc the pilaster cast method the soil is stratIified inlto layers by mens00 of 0 alifrn leaf oir other' di ciate mateial an n( deli nite pressurie applied-the effect of whic h is to be studied. The soil is th en remov ed one laveir at a time andI a cast madle of t he di(stor'tedl surfares of' the aluminum leaf. When the casts ale ic,mox et their' surtfaces mnay be contoured and studied. A camrna I urida is tisedl foir transfeiring the onitoui's to coori~inate Ippir for study. Figure 8 shows a series of casts madle by this method. Iwo blIorcks of soil are sh own dix ided into six Ilayers each. The arching out of' the soil is ra used in this case by a small tractorlu tg sining lbhiceIi fourths ot an inch into the soil block par'alleI to lbhc stur'-"' -. fare show n. The visual method ( Figcuoe 1)) of stuintvig arch pro0perties of ai soil conlsists essenltiaIlvy iin straifying the soil into layeis by means of thin Italian aluminum foil in a glass fared box (A) and forcing a pilungel' (P) dowvn I into the soil beside lbhe glass. The pr essu re (of the p'1) lnger' is m~eastiredt by means of a callibralted spring (C(') and the inovement of' the soil is ____________________shown by the dlistoi'tion o)f I the aluminum leaf' lavers. Tbhe aluim inum leaf being eery thin and fragile (does not apI)pariently a fleet the movement to all ,,' r apprl~e- _________________soil -where a '2 ________________________________ I GtIiU r,.-Two ,(,ieso plaster cast, ,heox joig the oh. tel 111,1 of soil aot Lotce.-'ixe layer, cau.e boll a r iabl e extent excelpt in cases it is xvr close and1( ~par'allel 1(o a shear' area in which case its e ffec c ain he noited in fragmentaltion at a place oil low bondage. The soil is p~laced in the box one layer' ill a time and( comrlessetl by means oof a craiik diiven wxor'm gearedl plungr' II 6' a. Soil IU ('. C'alibrated .pring 1). (Comprer.100 dev ice F'. Loicatioin ofi Goldheck Gaini'(' F. C'ompresio air app-u!iiiIts for nHhasuifiiig cuiopationi. sure is mleasu red by means of a Goldbec k gauge (h'.) (4), air for which is suppllied and controlled by means of needlle v alv'es and l)ressure regulators (F) . Measurements are madle of the sinking in of the plunger and widlth and dep~th ot the arch. Some of the results obtained with the v isible method app~aratus follow andl are given in order to allow an applraisal of the valute of this method of observation. Studies with the ab~ove described app~aratus show that the full comp~ressive force of the plunger is ouly exertedl on a comnp~arativelv shallow area as the aluminum leaves are not moved far ahecad of the pl unger. This means that the force is vectoredl outwardl andl distri butedl in all directions. This vectoring is apparently caused1 by the friction of soil on soil, by the interlocking of p~articles and by the cohesion of the moisture films. The lines of force v ectoring outward are met by the compr1essive resistance ot the soil and sh ear resistance. Siniue coh esion can be rneasu red andl the relation ot pressure to cornpaction is k nown, and, as wvillI be sh own later', since the sh ear valu oftia soil is p~rop~ortional te to p~ressure, it is possiblIe to (determine with a reasonable degree 13 of accuracy the amount of force being exerted on different parts of the soil as well as its direction. Magnitude and direction being all that is needed to define a force, it would seem possible then to predict the reaction of different soils to various force applications from the data obtained by these ex- ,, periments. It was found that the FIGURE 10.-Pure sand 6.6% H20 compressed 10 lbs. per sq. in. is driven soil ahead of the soil is driven ahead of the plunger in the shape of a cone (See Figure 10) and that the lines of distortion of the soil caused by the advancement of the cone were in the form of an arch closely resembling parabolas in appearance. The soil movement was along the sides of this cone. Experiments were conducted with different widths of plungers and it was found that the width of the arch increased much more rapidly than the width of the plunger. Apparently the cross sectional area of the arch increases with some power of the width of plunger. The parabola changed in depth with the increased width of plunger and the area of soil in which movement was consequently considerably increased. With a uniform soil such as was used in these studies where the arch moves ahead of the plunger at a definite distance and with uniform width, the total penetration was found to be almost directly proportional to the pressure; i. e., if 5 lbs. gives .5 inch, 10 lbs. will give 1 inch penetration. This was only true, however, after the plunger had moved through enough soil to form the arch typical for that soil. It was found with all soils that an arch of the same form and dimensions would move ahead of the plunger through the entire mass of the soil in the container until the bottom or side of the container was reached. This, of course, means that the arch formation is a constant characteristic of a soil. Different compactions (5 lbs., 7.5 lbs., 10 lbs., 15 lbs., per sq. in.) were given soils and arch studies made. With all different compactions the width of arch for any particular soil was practically constant. It was therefore tentatively concluded that structure, at least within the limits ordinarily found in field soils (in what is generally considered a state of good tilth) was not an important factor in arch width but that this would be closely correlated with the cohesive properties of the soil and the interlocking properties of the particles. Studies of the cohesive properties of the soil showed that cohesion varied with pressure up to a point where full contact of the particles was made and then became constant. Cohesion was also found to vary with moisture content. Tests of sandy loam of a wide range of moisture, (2.7, 4.2, 6.7, 10.0, and 11,1 per cent) however, gave practically the 14 same width of arch. Evidently the cohesion is of minor, and the interlocking of particles of greater, importance in this connection. This is logical, as the force of cohesion should be uniform throughout the mass when full contact is established. Although the width of arch was a constant property the sinking of the plunger varied with compression. A fifteen pound weight on the plunger sank 1.7 inches when a sandy loam soil containing 6.6 per cent moisture was compressed by a force of 10 lbs. per square inch; 2.2 inches when it was compressed by a force 7.5 lbs. per square inch and 3.7 inches when compressed by a force of 5 lbs. per square inch. If the thickness of different soil layers beneath a plunger be measured and plotted it will be found that the compression varies inversely with the depth of the layer. This is plotted for four different pressures in Figure 11. If slight variation for original compaction be allowed, the curves may be considered as practically straight lines. When these values are, however, compared with the curve (Figure 7) giving the relation of compression to pressure it will be seen that the amount of compression would indicate pressures far beyond those possible with the apparatus used. This may be explained, however, by the movement of the soil from in front of the arch, and in fact this high resistance to compression apparently accounts in large measure for the soil movements which cause arch action. It has been shown that the arch action is largely due to soil movement in front of the advancing surface. If this is true the arch could be vectored or thrown in any direction desired by slanting or curving the surface so that the soil movement is affected in a desired direction. It was observed that when a plunger, 1.2 inches wide, had a slope of 23' the arch, which was 2.8 inches wide, extended from 1.0 inch to the left and .6 inch to the right of the plunger. At this angle soil was moved to both sides of the plunger in about the percentages indicated. With the plunger set at an angle of 330 the ratio of 1.0 to .3 was obtained and it was observed that the arch itself was carried on one side of the plunger. Cohesion It is apparent that the sticking together of particles of soil is an important physical property affecting every tillage operation. That this property is largely dependent upon the moisture content of a soil is also evident. Theories of the forces exerted by film moisture are so satisfactory in explaining this phenomenon that they may be accepted as a basis for a practical understanding of this property. The force of cohesion undoubtedly is closely correlated with the size of particles, and the size of particle logically would govern the number and curvature of the moisture films. While this simple theory may not be all that is needed to explain the physical effect of colloidal material, it still is sufficiently accurate to be of the greatest assistance in enabling one to judge the properties of a soil by a physical analysis. 15 C.,-7'CEN SN/O r/A'6 OPFFo,e GE. NN. O41 (/" / FAG . L C e 5 /0 N vo.,ej OF Y4*/OU 5 7W3 Cr O.e /NvG 04'7O.2 O/7~/±3 !/T/ON N O F 66N 40 3o I. " it.\ /5. ON PL N~r /O'ONPL UA'OA F 50 /00 /50 200 MM. ! M O 7 FIGURE 11 The apparatus used at this station for measuring cohesion (Figure 12) consists of a modification of the "Tenacity of Soil Apparatus" developed at the University of Illinois (manufactured by Central Scientific Company). The modification which enables the measurement of cohesion or tenacity of such soils as sand consists primarily of substituting a sensitive Jolly balance (A) for the sand weight bag, and filling the soil containers (B) with blocks of wood so that the soil chamber is one inch in all three dimensions. It was found necessary to give the soil a uniform compression so that the square inch of soil being pulled apart would always contain approximately the same number of films. This was done by means of a calibrated spring compressor (C). To avoid jerking the mechanism the Jolly balance was driven by a worm reduction gear (D) which applied the pull to the spring very slowly and uniformly. By careful manipulation the results obtained with this apparatus could be made to check within 5 per cent, which was considered a sufficient gree of accuracy for such a varying material as soil. In each case de- twenty or more measurements were made and the averages taken as the cohesion. In general, cohesion was found to increase with moisture, the curve rising rapidly at first and gradually decreasing in slope until a certain maximum was reached when the curve fell off quite abruptly. Since the falling off in the high moisture ranges 16 wVas evidently clue to free moisture, a soil condition of little interest in tillage, the tests were not generally carriedl much beyondl the lange of lblowing conditions. The noumber of particles in contact, as idicatedl by the a pparent specific gravity, is of the greatest importance with a soil of sufficiently lowv moisture content to stand packing without ru1)tuie of the films. It was found( that the cohesion varied with up to a certain point anda thein buecaume fairly uniform. prssr Up to this point the results werle more or less erratic as checks were difftic ult to obtain with t he most painstakhing mani pu lation of' the a pplaratus. Undou bted ly th is was dlue to cdifferences in structure which dletermine the amount of' contact betwxeen particles, because until sufficient pressure w\as apl 1iedl to insure full contact, the measurement dI e pended as much on he amoulnt of' soil or number of films ashon thetnacitof biof theitnact ehe0 films themselves. Tbhc effect of soil structure is o I considera ble i m p)otance' not oly in cohesion but in the studly of' other dlyna mic prop ertiesI lhe structure of soil is a thingt of' infinite \ ariabuility and it is illogical to expect small such e\ en quantities as gr'am samples to be exactly alike in sti'uctun1'e a nd( consequcently e xa ctly alikec in such physi- A ' J - cal pr1operties as cohesion, shear, re-c sistance t o comthe priession, other etc. h a n cd On A IIRE 12-*.\IO)rIu BA. evr pr'oper )1owmanknows y t h a t b tillage a satlrad'- fo J~ol l Blance lt ).u tijg C'hesin isf'actor\' and lical ly uniform Conclition ot' tilth can C. ll. Spring comtprso Worm gear divi ng appaau. be obtained over large acreages. It is evident, therefore, that there are certain general conditions which must be obtained by the tiller and that these conditions prevail between rather wide limits. There is little doubt that the reactive forces of field soil vary. This suggests the important question: What is a reasonable degree of accuracy in experiments of this kind? The only answer that can be offered to this question is that studies of this nature must be, at least for the present, qualitative rather than accurately quantitative, and that the most accurate meas, urements are of importance simply as they indicate the order of magnitude of the reactive forces being studied. Shear From an engineering viewpoint the shear value of soil is generally accepted as being of the greatest importance. Unfortunately, however, there appears to be little or no information concerning this property of soil or even any generally accepted method of making a determination of the shear value of a soil. It was found necessary to define soil shear and devise a method of measuring it. Observation showed that, except in soils which had dried so that the colloid formed a cementing material, the separation of particles in a definite manner, such as is found in shear of solids, did not commonly take place in soil. The dynamic properties of the soil are such that when one layer is forced by pressure to slip over another layer, the bondage of the particles may be increased instead of decreased due to the increased amount of surface contact, or with high moisture contents, to a rupture of the films and a puddling effect. Moreover, due to the interlocking of particles, there is a rolling effect and the area of shear is indefinite, sometimes extending more or less through a large mass of the soil. This effect would be nearly as accurately described by the term viscosity as by the term shear. The term shear, therefore, was defined as the slipping of soil over soil and its value is that of the internal resistance of a soil to any movement of its particles. To obtain a quantitative value of shear the following apparatus (See Figure 13) was devised. A soil cylinder (A) 5 inches high and 6 inches in diameter was mounted in a press in such a manner that the soil could be confined at any desired pressure; the pressure being applied by means of a screw and plunger (B) fitting the top of the cylinder. The pressure on the soil was measured by means of a Goldbeck gauge (C) which closed the entire base of the cylinder. The cylinder was cut so as to form three rings, the lower being one inch thick and the two upper rings two inches thick, respectively. The lower and upper rings were fastened securely in place and the center section hinged so that it would swing out by means of a lever (D). When the center section was swung out, the confined soil was "sheared" at the upper and lower surfaces of the hinged ring. The lever was pulled by means of a crank and windlass (E) to avoid jerking and 18 i11(l ndsrn 2L4 I I LeL r1 Gol;~dbek ug the pull measured by a calibratedI spring. The shear in p)oundls per sq uare inch could be deteirnined since the leax elage. as Nv ell as the number of squfare inches of shear area, was known. Ev en with this simple apparatus, dlifticulty was encounlteredl in getting exact values since the soil would compress in front of the monog section before it would shear, that is at least u nder p jessures ot the magnitudec to he exp~ected in field soils. Th~is means that the area of actual shear was varying and the valute obtained was too low wxhen calt ulated on the square-inch basis. It was possible, however, to measure quite accurately the exact area over w hich shear xxas taking place as the center ring sw inginug out exposed the edge of the soil column and the dlistance between it and the edge of the ri ng w"as the am on t of comrn )ession . The laction wvas also measuredl by the p~ressure exertedl in this comp Goldbeck gauge. While a fewx exact qua ntitatix e measurements of this kind we re made in the studlies to dlate, this degree of exactness has not been generally attempltedl for the reason that the first obj ective has been to determine wxhat properties are of i mportanc e andl by an applroximate evaluation dietermineit something of their relativ e importance to design. Considering the extreme vai19 ability of field soils and the number of factors affecting the shear of soil the value of a high degree of accuracy in quantitative measurements of this kind is questionable. Tests of shear value have been run with a large number of soils ranging from pure sand to heavy colloidal clays. From these tests certain definite In all cases shear was general facts have been determined. found to vary directly with pressure. The ratio of shear to pressure, however, varied with the moisture percentage. With increasing moisture the shear increased up to a certain point and then fell off as the moisture increased. The connection between force required for shear, pressure and the other variables, and the design of tillage implements is obvious. EXPERIMENTAL METHODS FOR STUDY OF SOIL METAL RELATIONSHIPS Friction Methods of studying friction between soil and metal, and some conclusions drawn from these studies have already been published (5). The importance of this in relation to soil dynamic studies is sufficient however to warrant a repetition of the discussion of the methods used and a summary of conclusions with applications to show their justification. In these studies the primary equipment consists of a piece of flat metal pulled by hand with a calibrated spring balance. The scale reading divided by the weight of the slider gives the coefficient of kinetic friction (U'). For studies of the effect of speed on friction the slider is drawn by a constant speed motor through suitable reducing gears. It was found that there were four different sets of conditions depending largely upon the film moisture and the weight and material of the slider, and that the laws governing friction varied with these conditions. The adhesions of the soil moisture to the metal was found to be a most important factor in the so-called "friction" between soil and metal. When the pressure of the slider or the attraction of its metal for the soil moisture was sufficient to cause the wetting of its surface, the soil would stick or adhere to the metal and the "friction" would greatly increase. This necessitates the determination of the value of the wetting or spreading coefficient of the metal before the general laws of friction can be applied to any surface. The law governing this property was stated by Harkins as S - Wa - Wc, (6). This expression exhibits the extremely simple relation that spreading occurs if the adhesion between the liquid of the films and the metal surface (Wa = Work of adhesion) is greater than the cohesion of the liquid (Wc -Work of cohesion). It is obvious that a positive value of the spreading coefficient corresponds to spreading or wetting and a negative value to non-wetting. From the studies made it seemed possible to lay down tentatively certain fundamental laws for sliding friction between a 20 metal surface and the soil, remembering that these hold only between certain limits (a fact common to all laws of friction). A. Friction Phase.-In a dry soil when the value of S is negative and when the bearing power of a soil is less than the pressure, the coefficient of sliding friction (U'). (a) Varies with the speed, (b) Is proportional to the pressure per unit area, and (c) Varies with the smoothness of the surface and the materials of the surface. B. Friction Phase. -When the bearing power of a soil is greater than the pressure per unit area and the value of S is negative; i. e., the slider does not get wet. (a) The magnitude of the friction is proportional to the total pressure between the two surfaces: (b) The value of U' depends upon the roughness of the surfaces and the materials of the surfaces; (c) It is independent of the area of contact, and (d) It is independent of the speed of sliding. C. Adhesion Phase.-When there is enough moisture present to cause the soil to adhere to the sliding surface (a positive value of S) but not enough to have moisture brought to the surface, then U' varies (a) With the speed, (b) With area of contact, (c) With the pressure per unit area, (d) With the surface tension of the film moisture; i. e., (1) It varies with the amount of colloidal matter present. (2) It varies with the amount of water present, and (3) It varies with the temperature and viscosity of soil solution. (e) With the surface and kind of metal. D. Lubrication Phase.-Where there is enough moisture present to give a lubricating effect, U' varies (a) With the pressure per unit area, (b) With the speed, (c) With the amount of moisture and viscosity of the solution, and (d) With the nature of the surface and kind of material of which it is composed. It will be seen that the coefficient of sliding friction is a constantly varying factor rather than a fixed quantity and that in any soil it is affected by moisture content and particle size. The relation of these factors to the elements of design of tillage equipment is quite important. To show this connection a few applications of the findings to plow design follow. In the A phase the shape that would give the lowest surface speed of soil over the metal surface and the lowest pressure per unit area of contact would give the least frictional resistance. Soils having these 21 / P'R2 5ANo 2 SAO 7 5% CLAY .3 SAND O CAY SAND 2S7 CLAY 3U ,Q2 CLAY Z S7O 7S%~ /oo /O/ "700 .600 /00 .00- -0 -3 /0 s f4 LTE' 7YA FIGURE 14.-Friction Curves: Soil and Chilled Plow Iron. The five curves give the coefficient of kinetic friction for five soils of the composition shown. The curves show the effect of moisture quite distinctly. In each curve the lower moisture ranges have a constant coefficient of friction represented by a straight line which covers the A and B friction phases. The adhesion of C phase is represented by the sudden rise in curve to a maximum.. Following this sloping portion of the curve covers the lubrication or D phase. the the properties are usually worked with a steep moldboard plow which is, exactly the opposite to what frictional laws. would seem to indicate as advisable. The ordinary range of plowing is represented the A and B phase. The B phase would also, seem to indicate gentle slopes of the moldboard. In some soils (push soils) B completely disappears, the value of S. being positive, due probably to a low value of Wc in the formula S equals Wa minus Wc. The possibilities of a practical solution of this problem would appear to depend largely upon the development of metals having a low Wa value. It should be noted however, that the importance of the equation S = Wa is not limited to push soils, the limits of the C and B phase being determined entirely by the sign of S. by -Wc ficient Adhesion Figure 14 gives the relation between moisture and the of kinetic friction (U') for four synthetic soils and chilled coef- 22 iron, and illustrates the importance of the adhesion ph ase. From these curves it is ev ident that adhesion can be dIeterinnedl roughly by the slider method. Various treatments of metal surfaces such as polishing and (hilling against Val'ions so bPtances Wel-e tond(to ca use th em to vary in their ad hesion to soil. The slidIer met hod, however, does not showv with suifficierit accuracy the various adlhes5i\e values of metal. for whenl 5 is p~osit ive, the metal wets and( the increasedl h eight of the cnuive (deplends upon the moisture films alone. In other1 wordls, it' soil sticks to a metal, the( metho 0(plrovid es no( means ot telling h ox h aid it is stic king so as to jnudge the various values which the metal surfaces Or treatments thereof max' have. TPhe (Iiffi Ulty Of lprQnan ong samp1)1es of varlions m etals with (diffeieiit tieatm ents in p)1eces large enough to conduc t slider tests also renders the abov e mnethodl imlpractical . For the above reasons an indIirect method of rapidly measnuring adh esion has been (level opedl. This c(onsists of inoning tests w\ith a ictkel surfacedl slider on the v arious soils and com paring the attraction of this metal tor water wxith the attraction of other metals to be testedl. It dlesiredl, the soil soluntioni may be usedl in plae ofI the wvat er. The attraction for -water is dleteim inedl by (Figure 15) 5n51pend(inog the e(dges ot two pierces of metal in a constant tem- Surface. 13. Apparat faces of metal C. 1le asurin for cn t ol(ling' distalite apart of parallel sur- mtoicroscope. 23 perature bath so that their faces are lparallel and~ a known and controllIable distance apart, and th en measu ring the height of the callillar, '01umn drawn tup between them by means of a micrloI sc(I pe eq ipped with an eyepiec e mic rometer. Only prelimninary stud~ies have as yet been made by this method b~ut it is included in this discussion as the most priomising avenue of attack. Effect of Shape of Surface Applying Pressure To Soil The val ue of methouds otf measuin and st udying the ditffeirig ent soil v'ariablies affecting dIesign dlependls upon establishing the rel ationsh ip betw\een th ese vairiables and the ditterenit el emecuts of dIesign. This c alls for an analysis to d eter'mine what the possibl e elemenots of design are. It h as alr ieadly been pointed out that about the only way force can b~e alppliedI to soil is by lIressure. E'ven stich alppllication as that exer'tedl in the ctittinog action of' the rotary Iplow\ or hiarrow\ resolve themselves into this one essential. D~esign, therefore, being limited to dlevices for piresstire a pIpl~ i c a t i oi n resolves itself into consid erations o f amotint of pre'ssurie andl character of appllying the priessurec. Character' ot suriface dec- stiiface pendls ui pI 0 n the kind of metal, its physical state of hardness, andI its plolish . Methodls of stuidyinig metals in t h i s connect ion have been totiched uiponi briefly und(er the heading of ad(hesion. Therefore, t o n I y remains necessary to sugg' e s t chi.sei 1I. Sd phon methods (If ll(.) aihi VM19g(. sttudying the effect of amount andI dircttonl (It Ipressulre, 24 or from a practical design standpoint, area and curvature of the applying surface. Apparatus for studying the effect of curvature has been devised (Figure 16) and studied. This consists essentially of a box (A) in which various soils having known physical qualities are confined and a device for pushing chisel-shaped pieces of metal through them. Observations and measurements of the reactions of the soil are made to determine how these reactions are affected by different depths, curvatures, etc., of the chisel. The chisel (B) is mounted on a bar of metal (C) moving parallel with the surface of the soil on ball bearings and driven by an electric motor through reducing gears so as to give any speed desired. The force exerted by the chisel in passing through the soil is measured by Sylphon bellows (D) filled with water and connected with pressure gauges. Soil distortion is measured directly by calipers through the plate glass side of the containers which permits continual observation of all soil movement in front of the chisel. It is possible to facilitate the study of soil reaction by means of small objects placed at regular intervals throughout the soil or by means of thin pieces of aluminum leaf placed in layers. To show the application of this apparatus and its value in connection with design, a brief review of a few experiments with the conclusions drawn from them follows: For the purpose of studying the effect of angle of force application, a series of nickel plated copper chisels were made. These were one inch wide and arranged so as to run in the soil at any depth desired up to four inches. The nickel plating was done so as to give a constant low and known coefficient of friction in all cases. The series consisted of four chisels bent to give angles of 22.5 ° , 45 ° , 47.50, and 90 ° with the horizontal surface of the soil. Various means of observing the soil were used but layers of thin aluminum foil placed beside the glass and small objects such as rivets equally spaced throughout the mass were most satisfactory. Observation of soil action was also made with a binocular prismatic microscope equipped with an eyepiece micrometer. With this latter equipment the details of the soils reaction could be observed closely through the plate glass side of the box while the large movements were followed by means of the movements of the marker indicated. At various time intervals graphs of the movements were made on coordinate paper for record and study. It has been previously noted that the normal shear angle of soil reaction to a plow was approximately forty-five degrees. In general this seemed to hold with the soil chisels regardless of the angle or curvature of the advancing chisel. When using a perpendicular chisel the soil would compress a certain amount and then a shear plane would develop running from a point near the top of the chisel upward and forward, following this at regular distances down the chisel other planes would develop until the point was reached, the last one starting from just beneath the point. With a perpendicular chisel one and one-half inches 25 deep, in a Louisa clay containing 12.7 per cent water, four of the planes were developed before the soil shear plane developed from the bottom of the chisel. As the pressure varies uniformly with depth, and as shear value is in direct proportion to pressure, it is to be expected that the shear planes would develop at regular intervals. When a part of the soil shears off, the pressure of .... the advancing chisel forces .the block of soil to slip up the shear plane; when the ' " ..... next lower plane develops the two blocks move as a solid block, the entire sheared area always slipping up the lowest or last shear plane. This means that there is a gradual movement upward and forward so that in front of a perpendicular chisel a row of rivets, placed horizontally in the soil should by the soil movements be rearranged in a vertical row before the chisel. This was found to be the case. Similarly with other chisels at different angles the FIGURE 17.-Soil being forced upward and forward by a soil chisel slips on the shear (X) or inclined plane planes (B) until X and any point X' have similar soil moves in such a manner that a horizontal set of rivets in the soil will arrange themselves in a plane parallel to the surface of the advancing slopes throughout. A set of rivets (F) placed horiface will arrange them- chisel. (Figure 17). If the selves so that they have the same curvature as chisel has a varying curvathen chisel and the moveture, the rivets will arrange themselves to conform with The new shear planes dethe curvature of the chisel. velop at the point (D) as it moves forward in the This gives us an insight into soil, and the movement the effect of curvature of the on these planes over one surface of a plow or other another is at a rate deimplement. The amount that pendent upon the curvature of X. a soil slips over itself depends upon the steepness of the slope of the chisel. With a uniformly advancing curve the rate of slipping is proportionate to the slope. If the curvature is gradually and constantly increasing, the conditions are such that the soil must slip on all the shear planes at once in order to effect this arrangement. As this soil is being forced upward and for26 ward the direction of its movement depends upon the weight of the soil above and the resistance offered in front of it. The resistance in front of the advancing soil gradually diminishes as it is lifted above the top surface. APPLICATION TO DESIGN While it is impossible to make an accurate evaluation of the methods which have been set forth until they have proved themselves by contributing definitely to development of improved design in practice, a few examples of their relation to some important problems of soil mechanics may be of interest. The amount a soil will compress and the relation of this compression to shear value gives a ready means of measuring the "permissible slip" of a tractor wheel. The arch action of a soil should govern the spacing of tractor lugs and their distance apart on the rim. It is possible to compare accurately the effect of different design elements by determining the amount of shear, compression, etc., caused by various surfaces by the variable methods outlined. By studying these effects and their correlation with certain soil qualities, known or controlled, it seems entirely possible to come to the definite conclusions necessary for intelligent design. SUMMARY The object of the study of the soil properties discussed in this bulletin is to obtain a basis for the design of tillage implements. For the purpose of determining the properties affecting design, a small, nickel plated plow was mounted to run beside a glass surface permitting observation of the soils reaction. From this and supplemental studies a chart of the variables entering into these reactions was prepared. The study then resolved itself into determining the interrelationship of the factors through their most probable range of variation. Standard methods of measuring and studying many of the variables were used. New methods were evolved for moistening soil and measuring resistance to compression, arch action, cohesion, shear, friction, adhesion, and the effect of shape of surface applying pressure to the soil. Sufficient result data are included to show the applications of the methods. The study of friction between metal and soil was carried sufficiently far to formulate definitely laws covering this phase of the study. LITERATURE CITED (1) (2) Mosier and Gustafson. Soil Physics and Management, Lippincott. Anderson, M. S. and Mattson, S. Properties of the Colloidal Soil Material, U. S. D. A. Bulletin No. 1452. 27 (3) Nichols, M. L. and Randolph, J. W. A Method of Studying Soil Stresses, Proc. American Society of Agricultural Engineers, Vol. 19, 1925. Goldbeck, A. T. The Distribution of Pressure Through Earth Fills, Proc. A. S. T. M. 1917. Nichols, M. L. The Sliding of Metal over Soil, Proc. American Society of Agricultural Engineers, Vol. 19, 1925. Bogue. Colloid Behavior, McGraw-Hill. (4) (5) (6)