SOME ASPECTS OF KINETIC FRICTION OF GRAINS ON SURFACES Thesis for II" Degree 0" Din D. MICHIGAN STATE UNIVERSITY William George B‘ickert 1964 IIIIIIIIIIIIIIIIIIIIIIIIIII . . ‘ M’" 3 ‘sfrf'TL’V ..-: «'1‘- l 200 9323 MACH:- 2 IMAR161999 ‘.. £38 ABSTRACT SOME ASPECTS OF KINETIC FRICTION OF GRAINS ON SURFACES by William George Bickert Reliable kinetic coefficients of friction are es- sential for analyzing the forces occurring in a grain han- dling device and for designing the most efficient and eco- nomical handling equipment. Apparatus was assembled to study kinetic friction of grains. Basically, this appara— tus consisted of a table that moved horizontally and a sample container above the table for holding the grain. The force necessary to restrain the grain sample from movement when the table was in motion was the force of friction. Factors other than moisture content, normal load and velocity of sliding affect kinetic coefficients of fric- tion of grains. Deposits from the cuticular layer of the kernels on a surface that has been washed with a volatile solvent cause coefficient of friction to increase. These deposits increased coefficient of friction as follows: 29 percent with shelled corn (10 percent moisture content, wet basis); 50 percent with shelled corn (10 percent) on glass; and 97 percent with barley (10 percent) on sheet steel. In some cases the friction force increases during the first portion of the pass of a grain over a conditioned surface. The hypothesis was presented that this increase William George Bickert is partially due to the liquefaction of fatty acids that are a component of the cuticular layer of the kernels. Another contributing factor may be the variation in deposits from the kernels along the sliding path. Water vapor adsorbed on a glass or a steel surface affects coefficient of friction considerably. However, indications were that the effects of water vapor contrib— uted by the atmosphere would be approximately the same up to relative humidities of at least 80 percent. A recommended testing procedure was presented indi- cating the importance of conditioning the surface, changing the grain in the sample container each cycle, preventing outside contamination of the surface and limiting the loss of moisture from higher moisture content grains while testing. Normal load, in the range studied, does not have a significant effect on the coefficient of friction of shelled corn on sheet steel or plywood, or of barley on sheet steel. Moisture content begins to affect coefficient of friction of shelled corn at about 19 percent, wet basis. According to the Brunauer, Emmett and Teller theory of ad— sorption this is the point at which kernel pores exceeding 35 Angstroms in diameter begin to fill. Moisture content begins to affect coefficient of friction of barley on sheet steel at 17.5 per cent, wet basis. The effects of velocity on coefficient of friction were not studied because of limitations of the testing ap— paratus. William George Bickert Coefficients of friction of shelled corn sliding on sheet steel are higher than those of shelled corn slid- ing on plywood. Also, the coefficient of friction of shelled corn on sheet steel is increased more by increasing moisture content in the higher ranges than coefficient of friction of shelled corn on plywood. ApprovedflMu/ Ma or Professor my 2 f; 1%! SOME ASPECTS OF KINETIC FRICTION OF GRAINS ON SURFACES BY William George Bickert A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Engineering 1964 To Suzzanne, Juliet and Gretchen Mr. and Mrs. R. G. Bickert Mr. and Mrs. J. I. Shepard ACKNOWLEDGMENTS In the course of this study I have profited greatly from the counsel of Dr. F. H. Buelow (Agricultural Engineer— ing) under whose guidance the study was carried out. I am indeed grateful for his help and for his sincere interest in my progress and accomplishments. I would like to extend my thanks to Dr. C. W. Hall (Agricultural Engineering), to Prof. D. J. Renwick (Mechan- ical Engineering) and to Dr. C. P. Wells (Mathematics) for serving on my guidance committee and for their time spent on my behalf. A special thanks is offered to Dr. F. W. Bakker— Arkema (Agricultural Engineering) for his counsel on that portion of the study involving biochemistry, and to Mr. R. D. Fox (Agricultural Engineering) for his constructive criticisms of the manuscript. To my wife, Suzzanne, goes my deepest appreciation. Without her loyal support and endless faith I could not have reached this point. William G. Bickert II. III. IV. INTRODUCTION. TABLE OF CONTENTS 1.]- Objectives 0 O O O O O O O O O O O O O O O 0 REVIEW OF LITERATLIRE O O O O O O O O C C O O O O O 2.1 Laws of External Friction . . . . . . . . . 2.2 Friction Of Metals. O O O O O O O O O O O O 2 o 3 PriCtiOIl Of NOD-Metals o o o o o o o o o o 0 EXPERIMENTAL. 3.1 Apparatus 3.11 Apparatus used by other investi— gators . . . . . . . . . . . . . . . 3.12 Apparatus used in this study . . . . 3.2 Procedures. 0 O O O O O O O O O O O O O O 0 RESULTS AND DISCUSSIOPJ. O O O O O O O O O O O O O 4.1 Preliminary Tests . . . . . . . . . . . . . 4.2 Exploratory Investigations. . . . . . . . . 4.21 4.22 4.23 4.24 4.25 4.26 4.27 4.28 4.29 Description of recorder traces pictured in Figure 4 . . . . . . . . Effect of moisture present on the sliding surface on the friction force. . . . . . . . . . . . . . . . Increase in friction force during the conditioning process . . . . . . Changing friction forces during the first part of the pass over a surface. . . . . . . . . . . . . . . Conditioning of a sheet steel surface with shelled corn. . . . . . Conditioning of a plywood surface with shelled corn. . . . . . . . . . Microscopic examination of the surfaces . . . . . . . . . . . . . . Miscellaneous investigations . . . . Recommended testing procedure. . . . 18 2O 26 28 28 29 29 34 37 50 54 56 59 6O Page 4.3 Determination of Coefficients of Friction . 62 4.31 Coefficients of friction of shelled corn on plywood. . . . . . . . . . 62 4.32 Coefficients of friction of shelled corn on sheet steel. . . . . . . . . 68 4.33 Coefficients of friction of barley on sheet steel . . . . . . . . . . . 73 V o SUMI’IARY AND COI‘ICLUS IONS o o o o o o o o o o o o o 7 7 5 .1 summary 0 O O O O O O O O O I O O O O O O O 77 5.2 Conclusions . . . . . . . . . . . . . . . . 80 REFERENCES 0 O O O O O O O O O O O O O O O O O O O O O O 83 APPENDIX A. O O 0 O O O O O O O O O O O O O O O O O O O 88 LIST OF FIGURES FIGURE 1. Friction testing equipment. . . . . . . . . . . . 2. Grain sample container with weight holder removed and partially filled with shelled corn. . . . . . 3. Calibration of the force sensing system . . . . . 4. Recorder traces of relative friction force versus distance moved for shelled corn sliding on a glass surface 0 O O O O O O O O O O O O C O O O O 5. Coefficient of friction versus distance moved for rubber sliding on a glass surface at various veIOCitj-es (ROth -e—t a. , 1942) O O O O O O O O O O 6. Recorder traces of friction force versus distance moved for shelled corn sliding on a sheet steel surface I O O O O O O O O O O O O O O O O O O O O 7. Recorder traces of friction force versus distance moved for shelled corn sliding on a plywood surface 0 O O O O O O O O C O O O C O O O O O O O 8. Photomicrograph of steel surface in original condition (lOOX). . . . . . . . . . . . . . . . . 9. Photomicrograph of steel surface after testing with shelled corn and barley (lOOX) . . . . . . . 10. Photomicrograph of harder grained portion of plywood surface in original condition (lOOX). . . ll. Photomicrograph of softer grained portion of plywood surface in original condition (lOOX). . . 12. Coefficient of friction versus moisture content for shelled corn on plywood showing experimental data at a velocity of 3.5 inches per second . . . 13. Desorption isotherms for shelled corn at 80° F. according to experimental data of Rodriguez- Arias (1956) and the 3—constant equation of Brunauer, Emmett and Teller (1938). . . . . . . . Page 21 21 24 31 47 51 51 57 57 57 57 69 14. 15. Coefficient for shelled mental data second. . . Coefficient of friction versus moisture content corn on sheet steel showing experi- at a velocity of 3.5 inches per of friction versus moisture content for barley on sheet steel showing experimental data at a velocity of 3.5 inches per second . . Page . 71 . 74 TABLE A.l LIST OF TABLES Page Kinetic coefficients of friction from Lorenzen (1959) O O O O O O O O O O O O O O O O O O O O 0 l3 Kinetic coefficients of friction from Burmistrova 22 21. (1956). . . . . . . . . . . . l4 Static coefficients of friction from Brubaker and POS (1963) O O O O O I O O O O O O O O O O 0 l6 Summarized descriptions of recorder traces pic— tured in Figure 4. I I O O O O O O O O O O O O O 30 Regression equations and standard errors of estimate for coefficient of friction of shelled corn on plywood as a function of moisture con- tent . . . . . . . . . . . . . . . . . . . . . . 65 Regression equations and standard errors of estimate for coefficient of friction of shelled corn on sheet steel as a function of moisture content. . . . . . . . . . . . . . . . . . . . . 7O Regression equations and standard errors of estimate for coefficient of friction of barley on sheet steel as a function of moisture content 75 Tables in Appendix Page Friction forces for shelled corn sliding on plywood at various moisture contents, normal loads and a velocity of 3.5 inches per second. . 89 Friction forces for shelled corn sliding on sheet steel at various moisture contents, normal loads and a velocity of 3.5 inches per second. . 91 Friction forces for barley sliding on sheet steel at various moisture contents, normal loads and a velocity of 3.5 inches per second. . . . . 93 I. INTRODUCTION As the degree of mechanization in the handling of agricultural products rises it becomes increasingly impor- tant to know the basic characteristics and physical proper- ties of these products. In particular, a knowledge of the coefficients of friction of grains will allow more accurate analyses to be made of the forces occurring in a handling device. This will permit the design of more efficient and economical handling devices which will better suit the par- ticular handling requirement and will better fit into the overall system. The majority of friction data of grains presently available are given in the Agricultural Engineers Yearbook (1964). In all cases but one the data are based on the results of tests conducted prior to 1906. This does not mean to imply that researchers of that era were not capable of producing reliable data, but little information was given as to the conditions of testing or the effects of variables. Moisture content especially is one variable that is recog- nized as having a pronounced effect on friction of grains. Further, the values given in the Agricultural En- gineers Yearbook (1964) are static coefficients of friction. These are unrealistic from the standpoint of the design of materials handling equipment where the grain may be slid— ing. The force required to keep a body or material moving over a surface is usually considerably less than the force required to initiate the motion. A limited number of kinetic coefficients of fric- tion are available for forages and grains on various sur- faces. But in many cases the studies from which they re- sulted were undertaken as a preliminary step in solving another problem. For example, Hintz and Schinke (1952) needed the friction coefficients of chopped alfalfa and chopped corn sliding on steel when designing a component for a forage harvester. They found values to suit their particular need by experimentation. The effects of slid- ing velocity within a particular range were studied since this was of concern in their design. However, they were not interested in determining the effects of such variables as normal load and moisture content. As a matter of fact the moisture contents of the materials used for testing were not even specified. The picture of friction of grains becomes more per- plexing when a comparison is made of some of the friction data obtained by different investigators. As an illustra- tion, Burmistrova gtflgl. (1956) and Lorenzen (1959) have reported kinetic coefficients of friction of shelled corn at 13 percent moisture content on steel. Burmistrova g£_§l, found a value of 0.25. Lorenzen gave his results as a range, 0.44-0.62, with an average value of O.53--more than 100 percent greater than the value from Burmistrova gt_gl, While it is granted that the above example is possibly an extreme case, nevertheless, it very dramatic- ally points out the need for a basic understanding of kin- etic friction of grains. Also, it illustrates the need for recognizing all of the factors which determine the mag— nitude of friction forces. Only then may friction data of grains be made reliable and useful for design purposes. 1.1 Objectives The objectives of the study were to gain an under- standing of the basic mechanism of kinetic friction of grains and a knowledge of the factors affecting the magnitude of the friction forces. Having obtained an appreciation of the above, the next objective was to develop a testing pro- cedure and technique that would give friction data of a reliable and a reproducible nature. This testing procedure would be used to determine some actual kinetic coefficients of friction of grains. II. REVIEW OF LITERATURE "And indeed, external friction, which was observed first, is apparently the most complicated as to its mechan— ism. While it has been possible to present a fairly satis- factory, although far from complete, theory of gaseous, liquid, non-Newtonian, plastic, and internal solid fric- tion, this is by no means possible in the field of external solid friction." (Gemant, 1950) 2.1 Laws of External Friction G. Amontons, a French engineer in the latter part of the 17th century, is credited with the development of the fundamental law of external friction; i.e., the tan— gential force of a sliding body is directly proportional to the normal load. This law is expressed as F = LL N Where: F = tangential force y.= proportionality factor N a normal load. The proportionality factor, ,u. , in Amontons' law is referred to as the coefficient of external friction and, as the law implies, is independent of the pressure per unit area of the surfaces. Although Amontons is customarily given credit for discovering the proportionality between friction force and normal load, he only rediscovered something Leonardo da Vinci (1452-1519) had expressed earlier (MacCurdy, 1955). In the latter part of the 18th century, Coulomb added independence of the coefficient of external friction with respect to sliding velocity to Amontons' law. Amontons' and Coulomb's laws for external friction are then: 1. The tangential force of a sliding body is di- rectly proportional to the normal load. 2. The proportionality factor is independent of the pressure per unit area. 3. The proportionality factor is independent of the velocity of sliding. An interesting account of the history of friction has been composed by Courtel and Tichvinsky (1963a, 1963b). The first part of this series describes the experiments and conclusions of some of the early investigators of friction. The events leading up to the present day concepts of fric- tion are dealt with in the second part. 2.2 Friction of Metals Extensive studies with metals have been conducted by a number of researchers. A partial list includes Bowden and Leben (1938), Bowden and Hughes (1939), Morgan 2E.2$3 (1941), Hunter (1944), Dies (1945), Dokos (1946), Parker 25:21, (1950), Bowden and Tabor (1956) and Schmidt and Weiter (1957). Bowden and his associates have probably contributed the most to the knowledge of friction. It is generally agreed upon by investigators that the coefficient of friction is not a physical constant of a given material, but is the result of a combination of factors. The Varia— tion of any of the factors influences the coefficient of friction considerably. This has been especially emphasized by Schnurmann (1940), Claypoole and Cook (1942), Dies (1945), Marks (1951), Bowden (1955) and Bowden and Tabor (1956). It is also generally thought that friction between solids is largely a surface phenomenon, but there is still no com- plete agreement as to the origin of the friction force (Bonis, 1964). The state of cleanliness of the surface appears to have a far greater effect on coefficient of friction than does the roughness of the surface. This fact would tend to rule out the possibility of advancing a theoretical con- cept of friction from the standpoint of the interaction of the asperities on the two surfaces as suggested by Coulomb (1785) and used by Gemant (1950). Bowden (1955) states that the coefficient of friction normally observed for metal surfaces depends primarily upon the oxide films and adsorbed gases present on the surface. When metal surfaces are placed in a high vacuum the films are removed. Two surfaces placed in contact in this condition will so intimately contact one another that it is almost impossible to move them. Bowden and Hughes (1939) found that for metals treated in this manner the coefficient of friction was approximately 20 times the coefficient of friction under ordinary condi- tions. This is explained by the fact that two surfaces placed in contact under normal conditions actually touch one another only at the small points or asperities on the surfaces. The actual area of contact may be less than one ten-thousandth of the apparent area (Bowden, 1955). Contact is limited by the presence of oxide films and adsorbed gases so that only localized adhesion occurs. Removal of these films and gases will result in complete or almost complete adhesion. Since the energy produced by friction is dissipated mainly in the form of heat, large temperature rises may occur at the surface because of the very small area of con- tact (Bowden and Tabor, 1956). The temperature reached is normally limited by the melting point of the solid since localized melting at the regions of contact is expected to take place. If, however, the actual surfaces in contact are oxides, much higher temperatures may be reached if the melting point of the oxide is high. Bowden and Tabor re- port momentary temperature flashes of 500 to 1000° C. last- ing for only a few thousandths of a second or less. Peterson and Murray (1963) say that there will be either no surface damage or severe surface damage during unlubricated sliding depending upon the particular combina— tion of frictional parameters. These frictional parameters include such things as pressure per unit area, sliding ve- locity, physical properties of the materials, impurities present on the surfaces and atmospheric conditions. If the combination of frictional parameters causes sliding to take place in the bulk of the material, rather than at the interface, surface damage will result. Once surface damage begins a process referred to as 'fretting corrosion,‘ a particularly serious variety of wear, may occur (Dies, 1945). Much of the work in friction of metals has been on the effects of lubrication of the interface of two ma- terials in contact in reducing wear and transfer of metal during sliding processes. The lubricant, of course, sep- arates the sliding surfaces and the friction force then depends primarily upon the nature of the lubricant. Sur- face roughness begins to be of more importance when the lubricant film decreases in thickness to a point where con— tact between the two surfaces commences (Clayton, 1945). In applications where the environmental pressure is extremely low, lubricants that are usually used tend to disappear because of evaporation or sublimation. Gross seizure may then arise because of the intimate contact of the surfaces. Various dry lubricant materials have been evaluated for use in this situation (Bowen and Hickam, 1963; Bowen gt;§1,, 1963). Results show that the combination of a filler ma- terial, such as polytetrafluoroethylene, with a dry lubri- cant powder, such as tungsten diselenide, and a metal binder, such as silver, yields a material which exhibits desirable wear characteristics in a vacuum. According to Coulomb's law of friction, the coef- ficient of friction is independent of sliding velocity; but, there are numerous instances where this law is not valid. Gemant (1950) discusses the characteristics of the effect of velocity on friction force. According to Gemant, fric- tion force versus velocity may have 1) a positive character- istic, or 2) a negative characteristic. The effects of these are noted when two surfaces, either of which is elas- tically connected, are slid over one another. In the case of a positive characteristic, friction force is either independent of or increases with velocity. If discontinuity of motion occurs when sliding begins, it will continue only for a short period of time until a stable equilibrium is reached. With a negative characteristic, friction forces decrease with increasing velocity. Here, no stable equi— librium is reached if discontinuity of motion occurs at the beginning of the sliding process. As sliding begins and the relative velocity increases, the friction force decreases. This results in an acceleration of motion be- tween the two surfaces owing to the elastic connection. After acceleration the relative velocity decreases, the friction force increases, and the process begins again. This phenomenon is referred to as 'stick-slip' and is the subject of an article by Rabinowicz (1956). The dependence of friction force on velocity re- sulted in two basic classifications of coefficients of fric- tion--static coefficients of friction and kinetic coeffici- ents of friction. Static coefficient of friction is 10 ordinarily defined as the ratio of the maximum frictional force on a body resting upon another to the normal force pushing the two surfaces together. Actually, this is Amontons' first law where the proportionality constant is found for the maximum tangential force. Static coefficient of friction has been found to increase with time of contact (Hunter, 1944) and also to increase with the rate of appli- cation of the tangential force (Parker gtflal., 1950). Parker £3 31, found that the latter was more pronounced between a non-metal and a metal than between metals. Kinetic coefficient of friction is the ratio of the force of friction to the normal force during sliding. The bulk of the discussion prior to this has been concerned with this situation. 2.3 Friction of Non-Metals Friction of grains on various metal and non-metal surfaces are of major interest here. At the present time, the number of published works in this area is woefully few. The Agricultural Engineers Yearbook (1964) gives values of friction coefficients of grains taken from Airy (1898), Jamieson (1904), Pleissner (1906) and Kramer (1944). All of the values given are static coefficients of friction which are not realistic from the standpoint of design of grain conveying equipment in which grain is moved over a surface. Yet, these are all that the design engineer has to work with. Even if the designer wishes to use the values 11 for a static situation, he is given no indication of effects of moisture content, normal load, surface roughness or time of contact. Other researchers have obtained values for friction coefficients of agricultural materials, both static and kinetic. However, in many cases the investigation was un- dertaken in order to find values for a particular use rather than to analyze the character of the friction phenomenon itself. For example, Richter (1953) found it necessary to determine friction coefficients of various forages and si- lages prior to designing equipment for the reduction of dairy chore labor. He later published the results of his friction studies (Richter, 1954). The friction coeffici— ents of chOpped hay and straw and corn and grass silage on galvanized steel were found for various velocities and normal loads. Considerable difficulty was experienced in obtaining consistent results on the galvanized steel sur— face; the friction coefficient decreased as the number of passes of the material over the surface increased. Richter finally polished the galvanized steel with abrasives to a mirror-like appearance, at which condition the zinc coat- ing was completely worn off the surface in two small areas. He reports only small effects of velocity of sliding or normal load on the coefficient of friction of any of the materials tested. Hintz and Schinke (1952) needed values for friction coefficients of forages in connection with the design of 12 a forage harvester. They found that the friction coeffici- ent for green chopped alfalfa and chopped corn on steel is approximately 0.5 in a velocity range of 1000 to 6000 feet per minute. The moisture contents of the samples used for the tests were not determined, nor was the effect of normal load studied. In another analysis of the component power require- ments of a forage harvester Blevens (1958) found a need for coefficients of friction of forages. Tests were run but sketchy information was given as to the condition of the forage and experimental conditions. Velocities of slid— ing were less than 500 feet per minute. Lorenzen (1959) studied the moisture effect on fric- tion coefficients of small grains. Values were recorded for wheat, corn, milo, barley and paddy rice on plywood and steel. Normal unit pressure was varied from 0.2 to 1.4 pounds per square inch with velocity of sliding being held constant at 161 feet per minute. In general, Lorenzen noted that friction coefficient increased as the normal load was decreased. The overall effect of an increase in moisture content was to increase coefficient of friction above moisture contents of 14 percent, but this was not always true. The data, as presented, show quite a wide range of friction coefficients at any particular moisture content. Table l is a sampling of Lorenzen's data as read from his graphs. 13 Table 1 Kinetic Coefficients of Friction from Lorenzen (1959) MOISTURE FRICTION COEFFICIENT PRODUCT CONTENT % SURFACE RANGE MEAN Shelled Corn 7 Steel 0.44-0.62 0.53 Shelled Corn 13 Steel 0.41-0.56 0.48 Shelled Corn 16.2 Steel 0.42-0.61 0.51 Shelled Corn 19.5 Steel 0.43-0.65 0.52 Shelled Corn 23 Steel 0.45-0.76 0.60 Shelled Corn 7 Plywood 0.22-0.40 0.31 Shelled Corn 13 Plywood 0.24-0.41 0.32 Shelled Corn 16.2 Plywood 0.22-0.43 0.32 Shelled Corn 19.5 Plywood 0.23—0.46 0.36 Shelled Corn 23 Plywood 0.22-0.46 0.33 Barley 8 Steel 0.31-0.50 0.40 Barley 11 Steel 0.32-0.49 0.39 Barley 14.2 Steel 0.33-0.52 0.41 Barley 16.5 Steel 0.31-0.49 0.39 Barley 19.6 Steel - -0.49 0.39 Balis (1958) found a 10 percent increase in the friction coefficient of shelled corn continuously sliding on galvanized steel at a velocity of 100 feet per minute for 15 minutes. The galvanized steel surface was mounted on a 48 inch turntable and the shelled corn sample was held in a stationary position above it. ‘This increase was at- tributed to the accumulation of fragments or ”wear particles" from the shelled corn on the surface. Cleaning the surface with a dry, clean cloth caused the friction coefficient to return to approximately its original value. The effect was less pronounced with oats. Balis also emphasizes the effect of surface roughness on coefficient of friction. Investigations conducted in Russia by Burmistrova 14 gtflgl. (1956) suggested that friction coefficients of agri- cultural materials are affected by numerous factors. The more important of these are moisture content, state of the friction surfaces and pressure, while speed, area of con- tact, duration of contact and atmospheric factors are of lesser importance. Both static and kinetic coefficients of friction of grain, chaff, straw, ears, stalks and other various plant parts on a number of surfaces were found. Kinetic values for corn-crop products were found using a normal pressure of 10 grams per square centimeter and a velocity of sliding of 2-3 meters per second. Kinetic val— ues for flax parts were obtained with velocities from 1.2 to 1.6 meters per second and normal pressures of 20 to 30 grams per square centimeter. Kinetic coefficients of fric- tion that are of interest here are presented in Table 2. Table 2 Kinetic Coefficients of Friction from Burmistrova gtflgl. (1956) SUI—RFA—CE MOISTURE GALVA- CONTENT SHEET MACHINED NIZED PRODUCT % PLYWOOD STEEL STEEL STEEL CANVAS RUBBER Kernels 15.0 0.24 0.21 0.29 0.33 0.31 0.54 Corn Ker- nels at 36.5 0.34 0.57 0.49 0.39 0.31 0.62 Waxy Stage Flax 19 0.15 - 0.23 - - — 15 Brubaker (1962) reports static coefficients of fric— tion of some grains on various structural materials. A significant observation was that time of contact did not effect the force required to initiate movement for times of contact from 0 to 48 hours. Normal unit pressure did affect coefficient of friction. Friction coefficient in- creased with increasing normal load up to a maximum where it began decreasing with normal load. The maximum coeffi- cient of friction was observed at a normal unit pressure of 0.61 pounds per square inch. Brubaker also found that rate of application of tangential force had little effect on the static coefficient of friction for rates of appli- cation of 0.045 inches per minute to 0.285 inches per min— ute. Coefficients of friction of winter wheat on various surfaces were relatively constant up to about 13 percent moisture content. For soybeans the static coefficients of friction increased with moisture content from the lower value used of 7 percent to the upper value of 15.4 percent. Brubaker selected the moisture contents for test- ing according to equilibrium moisture contents at particu- lar temperatures and relative humidities. Data presented by Hall (1957) was referred to. In view of the relative humidities and temperatures given it is believed that the values of moisture content presented by Brubaker are on a wet basis rather than on a dry basis. Examples of static coefficients of friction as taken from a publication by Brubaker and P03 (1963) are given in Table 3. 16 Table 3 Static Coefficients of Friction from Brubaker and P03 (1963) MOIS TU RE FRICTION PRODUCT CONTENT , % SURFACE COEFFICIENT Shelled 7.5 Cold—rolled 0.23 Corn 9.9 Steel 0.20 12.2 0.20 13.9 0.24 Shelled 7.5 Plywood, parallel 0.27 Corn 9.9 to Grain 0.31 12.2 0.33 13.9 0.37 Barley 10.7 Cold-rolled 0.20 12.3 Steel 0.25 14.3 0.23 16.4 0.21 A particularly interesting piece of work on the sliding friction of agricultural materials was done by Wieneke (1956). Wieneke suspended various fibrous products (straw, hay, fresh grass, sisal grass and wool) over a ro- tating shaft and measured friction forces. Probably his most significant finding was that the coefficient of fric- tion increased with time of rubbing. It was concluded that this increase was due to the transfer of oils and fats from the product to the surface of the shaft forming a sticky film on the shaft. Also, the presence of water on the sur- face and its mixing with the oils and fats greatly increased friction coefficient. Higher coefficients of friction were found on the smoother surfaces because of the greater in- fluence of the oil and water film. Normal pressure was found to have almost no effect on the friction force. A 17 slight increase in coefficient of friction with velocity was observed up to velocities of 1.0 to 3.0 meters per sec- ond but there was no apparent influence of velocity above this range. Coefficient of friction increased with mois- ture content of the material up to moisture contents of 30 to 40 percent. Beyond this range friction coefficients were relatively constant. III. EXPERIMENTAL 3.1 Apparatus 3.11 Apparatus Used by Other Investigators Many different concepts have been employed in the design and assembly of equipment to determine friction co- efficients of agricultural crOps and also other materials. The basic differences in equipment stem from whether static or kinetic coefficients were sought. For determination of static coefficients of fric— tion a tilt table has been used often. The tilt table con- sists of a plane surface normally at rest with the horizon- tal. Kramer (1944) used a tilting-top drafting table for his work with rice. He placed the surface under investi- gation on the top of the table. A frame filled with rice was placed on the surface and the table was raised until sliding began. The tangent of the angle the table made with the horizontal was given as the static coefficient of friction. To study the effects of normal unit pressure, weights were placed on the rice in the frame. Other vari— ations of the tilt table were used by Airy (1898) and Jamieson (1904). One exception to the tilt table method for deter-- mining static coefficients was Brubaker (1962). His ap- paratus consisted of a positively driven horizontal table upon which the desired surface was mounted. A sample 18 19 container was placed on the surface and restrained from movement. Weights were placed on the sample to vary normal pressure. A disadvantage of this type of apparatus is that the sample container as well as the sample contacts the surface, although actually less than 10 percent of the nor- mal load was transmitted to the bottom of the container. A popular method of determining kinetic coeffici- ents of friction has been the use of a turntable. This is a horizontally rotating circular table upon which the surface is mounted. A restrained sample container either rests upon or is partially suspended above the surface. This basic apparatus was used by Hintz and Schinke (1952), Richter (1953) and Burmistrova gtflgl. (1956). .As with Brubaker's equipment, if the sample container rests upon the surface it contributes to the total force of friction and must be accounted for. If the sample container is sus- pended above the surface, a portion of the normal load may be transfered to the sidewalls of the container and also it is difficult to maintain a given clearance between the bottom of the container and the surface due to variations in the level of the revolving table. In investigations of friction of grains the clearance space is critical. Even minor deviations in clearance result either in contact between the container and the surface or in leakage of the grain sample through the clearance space. 20 3.12 Apparatus used in this study The friction testing equipment used in this study is shown in Figure 1. The equipment was ready for a fric- tion test at a normal load of 16.2 pounds when the photo- graph was taken. The horizontal table upon which various surfaces were mounted was supported by six V-shaped casters which rolled on the angle iron track. The track was rig- idly supported by a channel iron and pipe framework. To increase the rigidity of the six-foot long table a six-foot section of six-inch channel iron was fastened to its under- side. A surveyor's level was used to level the table and the track. Final adjustments brought the overall level of the table to within 0.05 inch from level between the two ends. The table was positively driven by a ball-bearing screw assembly. This screw assembly utilized a six-foot long screw and a mating ball nut. The ball nut was secured to the underside of the table at the end opposite that end shown in Figure l. The connection of the ball nut to the table was somewhat flexible, allowing small movements in all but the direction of motion of the table. The screw was mounted on the channel iron and pipe framework. It was driven by a reversible electric motor through a system of pulleys. The rotary motion applied to the screw was converted to horizontal motion of the table. Each revolu- tion of the screw caused the table to move one inch. Table velocity was changed by using different sized pulleys. An 21 Figure 1. Friction testing equipment '\ ‘ .\ “I Figure 2. Grain sample container with weight holder removed and partially filled with shelled corn 22 electric tachometer was attached to the screw to indicate table velocity. Initially, the table was drawn across the track by a chain and sprocket system but this connection was too elastic and gave an irregular motion to the table. The cylindrical grain sample container was eight inches in diameter and was fitted with a two-inch square grid of vertical plates inside (see Figure 2). The purpose of the grid was to limit movement of the grain with respect to the sample container and also to act as a leveling guide when filling the sample container. The weight holder was placed on the grain in the sample container and clamped against the grain with the aid of two adjusting screws so that the sample container was lifted slightly from the sur- face. Thus the total normal load was completely supported by the grain sample. The normal load was varied by plac- ing different weights on the weight holder. The sample container was restrained by two horizon— tal arms connected to the centers of two simple beams. The simple beams were fastened to angle iron uprights. The points of attachment of the horizontal arms to the sample container and to the simple beams were in the plane of the grain-surface interface. This eliminated any overturning moments on the container. Four strain gages were mounted on each beam to respond to bending of the beams. The bend- ing of the beams was directly related to the friction force which was that force necessary to keep the grain in the sample container from moving when the table was moved under 23 it. Of the four strain gages on each beam, two were mounted on the front of the beam and two were mounted directly op- posite on the back of the beam. When a friction force was being measured, the gages on the fronts of the beams were in tension and those on the backs of the beams were in com- pression. The eight strain gages were wired into a Wheatstone bridge and connected to a Westronics Model SllA Strip Chart Recorder through a Westronics Model SG-l AC Strain Gage Module. The strip chart had a usable width of eleven inches and was modified so that the horizontal friction force was recorded directly in pounds. Full scale deflection was 20 pounds. To calibrate the equipment the pins from the points of attachment of the horizontal arms to the sensing beams were removed and the sample container with the horizontal arms was taken from the table. Small cables were connected to these same points of attachment and extended horizontally over pulleys to the calibrating weights (see Figure 3). The glass surface used in the exploratory investi- gations was cut from a sheet of good quality window glass. The steel surface used in the exploratory investigations and kinetic friction tests was from a cold-rolled sheet. When received it was covered with an oil film and was free from visible scratches. The plywood was 3/16 inch, exterior type, A-A grade, Douglas fir. Surface roughness was measured with a Micrometrical 24 Figure 3. Calibration of the force sensing system 25 Profilometer. The tracing speed of the stylus tracer was 0.3 inches per second. The roughness—width-cutoff was 0.030 inches; i.e., only irregularities of 0.030 inches or less in width were included in the roughness-height average. Irregularities of the surface which are of greater spacing than roughness are classified as waviness (Kelly, 1957). The average roughness-height recorded was a root-mean—square average. This is the same type of average commonly used in electrical measurements. The corn was picked and shelled at approximately 40 percent moisture content, wet basis. It was then placed in sealed plastic bags, quick-frozen and stored at 0° F. until it was to be used. Then it was dried to the desired moisture content by blowing unheated air through it. After drying, the corn samples were held in a closed container for at least 72 hours before tests were run. The yellow dent shelled corn was a hybrid variety grown on the Mich- igan State University campus. The barley was of the Traill variety and was at approximately 10 percent moisture con— tent, wet basis, when acquired. At least 72 hours prior to time of use sufficient moisture was added to each barley sample to bring it to the selected moisture content. Moisture contents of the grain samples were deter- mined with a Steinlite Model 4006 Moisture Tester. Rela- tive humidity was determined with a wet-bulb, dry-bulb psy- chrometer. Air was moved past the thermometer bulbs by a motor driven fan. A sling psychrometer was occasionally 26 used to verify the values. Later in the study (prior to performing the tests of Section 4.3) a polyethylene covering was added to pro- tect the surface on the friction table. This was a long polyethylene sheet that covered all of the surface except that in the area of the sample container when the friction table was in its beginning position (see Figures 1 and 2). The end of the polyethylene sheet nearest the sample con- tainer was fastened to a window-shade roller upon which the polyethylene sheet was rolled when the friction table was in motion. The polyethylene sheet was particularly helpful in protecting the friction surface from dust and dirt from outside sources. It was also found to be help- ful when working with high moisture grains. The polyethy- lene limited evaporation of the moisture deposited on the surface by the grain. 3.2 Procedures Certain procedures relative to operation of the friction table were common to all investigations. The com- bination of these procedures will be referred to as a cycle. One cycle consisted of: 1. Placing a portion of the grain under investigation in the sample container. 2. Placing the weight holder on top of the grain in the container. 3. Establishing the clearance space between the sample container and the surface. 27 4. Placing sufficient weight on the weight holder to make up the desired normal load. 5. Manually aligning the sample container with its normal sliding position on the surface and taking up all slack in the force sensing system. 6. Starting the recorder and setting the table in motion. 7. Stopping the table at the end of the pass, stopping the recorder, removing the weight from the weight holder and reversing the table. 8. Removing the grain from the sample container and repeating steps 1 through 8, or leaving it in if more than one cycle was to be made with the same sample, and repeating steps 4 through 8. When the grain was removed from the sample container it was mixed with a main grain sample and a specific portion of this main sample was then used to refill the sample con- tainer. IV. RESULTS AND DISCUSSION 4.1 Preliminary Tests Extensive preliminary tests were conducted with shelled corn sliding on steel and plywood surfaces in order to determine what effects the moisture content of the shelled corn, normal pressure and velocity had on the friction co- efficients. Friction coefficients obtained under the same apparent sets of conditions varied greatly from one day to the next or even during a single day. Differences of up to 100 percent or more were noted. Also, the coeffici- ent of friction at the beginning of a particular testing period was 15 to 25 percent higher than values obtained after testing had been in progress for 10 to 15 minutes. Another unexplained phenomenon was the change in the fric- tion force during a single pass on a surface. Even though the table was moving at a constant speed the friction force increased at the beginning of the pass and did not become constant until after two to three feet of movement. This was observed with both the steel and plywood surfaces and also with both surfaces turned end for end so that sliding was in the opposite direction on a surface. The preliminary tests indicated a need for estab- lishing a dependable testing procedure and technique that would yield friction data of a reliable and a reproducible nature. They also showed that variables other than moisture 28 29 content, normal load and velocity had a profound effect on friction coefficients of grains. 4.2 Exploratory Investigations 4.21 Description of recorder traces pictured in Figure 4 A glass surface was chosen for initiation of this portion of the study because it would more likely have uni- form characteristics throughout its length than sheet steel or plywood. It also presented a relatively stable surface to the atmosphere. It was found that by washing the glass surface with a volatile solvent, such as carbon tetrachlo- ride, a readily reproducible friction force would occur. This presented the possibility of treating the surface in various ways while still being able to return it to a ref- erence condition. A notable example was the addition of various amounts of moisture to the glass surface which would have caused drastic rusting of a steel surface or would have been adsorbed by a plywood surface. The glass surface also allowed observations to be made of the shelled corn in contact with the glass during sliding. Figure 4 shows reproductions of recorder traces of relative friction force versus distance moved for shelled corn sliding on a glass surface subjected to various treat- ments. The numbers on the vertical scales are intended only as an indication of the order of magnitude of the fric— tion force. Trace A in Figure 4 shows a full-width section from the recorder chart. It represents the friction force of shelled corn on glass when the glass Surface was what 30 H momma mcmpmoomm mmuam mmaoho 0H Hooo on omzomam mommmSm mm momma mcmomoomm mmumm mopscme om mom mommem mcm>os m>onm pmomaa mQEmH ummm w momma mamomoomm mmumm mmmomo mm mmsos mm mom mmmerOEum mam on ommoexm vmmm ocm mommoanommumu conmmo cums pmnmmz amp ou omonHm ocm mommoanommumu conmmo sum: omnmm3 Q momma memomoomm mmumm mmaoho Ram U momma memomoomm mmumm mmuscma om mow mommmSm mcm>os m>onm omomHQ mQEmH #mmm m momma mo mcmpmoomm mmumm mmaoao m>mm suoao amp cmmHo m rum: amp ommm3 ocm mmumz omammpmmo sums pmzmmz cmnu “mommoarommumu conmmo rams pmzmmz mumum uwcomumccoo mommmSm mmmmm ou ucmEummma momma v mmsmmm CH ommswoma mmomm» mmpmoomm mo mcompammommp omNHmmEEsm .w magma 31 n" t ‘ mommmau madam e co unmommm :moo pondonm mom mots: mocmummo mammm> oomom somuoamm m>mummmm mo mmommu mmomoomm d--~...,. PIE. 8235.90 . v mmfimmm W I .‘vnmzfll ‘,.. H. ."”..." 1"). “(xix .pt... :ifffiRitzixf. “git: 93925;}; ‘15 lb Hi- GD W D III! ‘ fi 2 “.h\\ I “. dI.‘ II. I; I. WWII H. u w .0. H I ., I I. H rquxlw ,l\ .H rfi H . I A .x ‘ .. I H NHHMTH - H , in w m7 .‘I I I I 1 32 will subsequently be referred to as a conditioned surface.. The conditioned surface was attained by making repeated passes with shelled corn of 10 percent moisture content, wet basis. Conditioning of the surface took place over a period of several days and was continued until a relatively constant friction force was obtained. The vertical load was approximately 100 pounds and sliding velocity was 3.5 inches per second. The length of the sliding pass was four feet. The shelled corn in the sample container was changed every two passes. All other traces in Figure 4 were obtained with the same moisture content, vertical load and velocity. The reference condition obtained by washing with carbon tetrachloride is illustrated by Trace F in Figure 4. After the surface was washed with the solvent it was allowed to air-dry before the pass was made. The carbon tetrachlo- ride was expected to remove impurities such as dust and oil from the surface and also to drive off a significant amount of moisture from the surface. The first point of interest upon comparison of Trace F with Trace A is that a 50 percent increase in friction force occurred during the conditioning process. Secondly, whereas the friction force was relatively constant during the entire sliding period after the surface was washed with the solvent, the friction force after the conditioning process did not reach a constant value until approximately two feet of sliding had taken place. Trace B in Figure 4 was obtained after the glass surface had been washed with carbon tetrachloride, allowed 33 to dry, then washed with distilled water and wiped with a clean, dry cloth until one would normally assume the sur- face to be dry. This caused approximately a threefold in- crease in friction force over that obtained after washing the surface with carbon tetrachloride only. The effect was the same whether distilled water, cold tap water or hot tap water was used for the second wash. Six cycles after the recording of Trace B, Trace C was obtained. Along with a lessening of the erratic nature of the friction force, a general decrease of 15 percent occurred. Immediately after Trace C was recorded, two 250- watt infrared lamps were placed 14 inches above the glass surface for 20 minutes. The table was passed back and forth under the lamps to give more uniform heating of the glass. Following the heating period the glass was allowed to cool for a five minute period and Trace D was recorded. The erratic nature of the friction force returned along with about a 25 percent decrease. The friction force indicated by Trace E in Figure 4 was found six cycles after Trace D. Although the fric- tion force was still somewhat erratic it was particularly interesting that it had decreased to a value less than that found with the glass in a conditioned state. As mentioned previously Trace F in Figure 4 repre- sents the force of friction after the glass surface had been washed with carbon tetrachloride and allowed to dry 34 for a short time. Trace G also indicates the friction force after washing with the solvent except in this case the sur- face was left exposed to the atmosphere for 14 hours between the washing operation and execution of the cycle. Fourteen cycles later Trace H was recorded. Immediately after Trace H, the two infrared lamps were directed at the surface in a manner similar to that following Trace C. They heated the surface for 10 minutes and the surface was then allowed to cool for five minutes after which Trace I was recorded. Heating decreased the friction force to nearly the same value found after the solvent wash. Trace J was recorded 10 cycles after Trace I. 4.22 Effect of moisture present on the sliding surface on the friction force It is obvious when comparing Trace B through Trace E to Trace F in Figure 4 that the amount of moisture pres- ent on the surface plays a key role in establishing the magnitude of the friction force. It also appears that only very small amounts of moisture are necessary to drastically affect the friction force and that these small amounts of moisture are tightly bound to the surface. An indication of the order of magnitude of the thick— ness of the water films may be gained from measurements of the thickness of the rigid water film at a quartz-water interface made by Eversole and Lahr (1941). They concluded that the thickness of the rigid multimolecular layer adsorbed on quartz was of the order of 100 A (Angstroms). They 35 suggested that the rigidity of the structure of the water stemmed from the surface forces and the potential gradient at the quartz-water interface acting on the water dipoles. Just how rigid this multilayer would be during a frictional process remains a question. However, since Eversole's and Lahr's measurements were of the rigid film necessary to support a quartz disk, certainly the water film remaining after washing the glass surface with distilled water and wiping it dry with a clean cloth would be less than 100 X or at the most 100 A thick. Ries (1961) points out that in almost every process of a physical, chemical or biological nature involved with thin films, the monomolecular layer plays the most impor- tant role. The monolayer is held much more tenaciously to the surface than are succeeding layers (de Boer, 1953). If the diameter of a water molecule is taken as 2.9 A, the thickness of the water film after washing with water and drying with the cloth may be inferred to have been between 2.9 X and 100 K. Bowden and Tabor (1950) found that the adhesion between two smooth glass surfaces decreased with a decrease in the thickness of the adsorbed layers of water on the surfaces. In addition, the roughness of the surfaces was a major factor in determining the magnitude of adhesion. They expected that a decrease in adhesion occurred when the height of the asperities became approximately the same as the thickness of the adsorbed film. The root—mean-square 36 average roughness-height of the glass surface used for the cycles illustrated in Figure 4 was 0.2-0.4 micro-inches. The mean of this range, 0.3 micro-inches, is equivalent to 76.2 A. When converted to an arithmetical average height this amounts to 68.6 A. This means that, for all practical purposes, the average height of the peaks was 68.6 A and the average depth of the valleys was also 68.6 A. The friction force indicated by Trace B in Figure 4 was something less than the friction force found if the glass surface was not wiped as diligently with the dry cloth after washing with water. Therefore, if Bowden's and Tabor's suspicions that adhesion begins to decrease when the asperities become of the order of magnitude of the adsorbed film, the thickness of the water film on the surface when Trace B was recorded was less than 68.6 A. The thickness of the water film was then decreased by running several cycles or by heating the surface. Dur- ing sliding the grain could have adsorbed moisture from the surface or moisture could have evaporated into the at- mosphere. Heating of the surface hastened the evaporation process. The temperature of the surface was increased sig- nificantly during the heating period and the friction force did not decrease as much as was expected. A comparison of Trace G with Trace F in Figure 4 illustrates the increase in friction force resulting from contamination of the surface by the atmosphere. Possible contaminants were water vapor and dust. Because repeated sliding did not decrease the friction force (Trace H) but 37 a 10 minute heating period did (Trace I), it was concluded that the major contaminant was water vapor. This indicates that changes in the relative humidity of the atmosphere will affect the coefficient of friction. However, data presented by McHaffie and Lenher (1925) on the number of molecular layers adsorbed on a glass surface as a function of relative humidity indicate that the thickness of the water film does not begin to increase rapidly until rela- tive humidity exceeds 85 percent. Also, Bowden and Tabor (1950) found that adhesion between two smooth glass surfaces did not rise rapidly until humidity exceeded 80 percent. This implies that even though water films resulting directly from moisture in the atmosphere contribute to the total friction force, the effects of these water films are essen- tially the same in the range of relative humidities normally encountered. Only at very high relative humidities or con— ditions where the surface is contaminated by moisture do significant variations in the thickness of the water film occur. 4.23 Increase in friction force during the conditioning process Probably the most unexpected event illustrated in Figure 4 was the increase in friction force during the con- ditioning process. The constant value obtained with the conditioned glass surface was 50 percent higher than the friction force found after a solvent wash. A first conclu- sion might be that this increase was the result of water 38 layer buildup on the surface. But the water film formed by moisture from the atmosphere was shown to increase the friction force by only 15 percent. And the shelled corn was at an equilibrium moisture content with the atmosphere so probably was not losing moisture to the surface. The other possible sources of contamination were dust from the atmosphere, dust and dirt from the shelled corn sample and finally oils, fats and waxes from the contact surface of the corn kernels themselves. Wiping the surface with a clean dry cloth after each cycle did not decrease the fric- tion force appreciably. Heating the surface caused a small decrease in friction force. A major decrease occurred only when the surface was washed with carbon tetrachloride. This indicated that oils, fats and waxes were being depos- ited on the surface by the corn and the film formed was actually causing an increase in friction force rather than acting as a lubricant. The outermost structural part of a corn kernel is the pericarp. Frequently the pericarp may be differentiated into two or three or even four distinct layers (Wolf gt_§1,, 1952; Esau, 1961). Although the pericarp varies in thick- ness from one part of the kernel to another it is usually composed of an average of 20 to 22 layers of cells, the major differences in thickness arising from compression of the pericarp at various points. The total thickness of the pericarp at regions where sliding of the kernel would 39 normally occur is on the order of 100 microns (Kisselbach, 1949). The outer layer of the pericarp consists of a cuti- cle which is composed chiefly of cutin. The cuticle is somewhat impervious to water and can be flaked off by me— chanical means. It varies in thickness from approximately 0.7 to 10 microns. Little seems to be known about the ex- act composition of cutin. Baker gt_al, (1964) say that it is a polyester of hydroxylated fatty acids and other substances and that its composition varies according to its origin. Fruton and Simmonds (1958) state that the cu— ticle waxes contain long—chain fatty acids and long-chain primary and secondary alcohols, ketones and paraffin hydro- carbons. Several writers (Adam, 1938; Bowden and Tabor, 1950; Gemant, 1950) have reported that the addition of fatty acids to various lubricants actually brought about a considerable reduction in friction and even fatty acids themselves were good boundary lubricants. Since fatty acids are a compo- nent of cutin it seems strange that the deposits of cutin on the glass surface caused an increase in friction force. On the other hand, Wieneke (1956) found an increase in fric— tion force when straw and wool were repeatedly rubbed over a surface. He concluded that oils and fats were transferred to the surface where they formed a thin layer and that this thin layer was what caused the friction force to increase. The addition of moisture to the surface caused a particu- larly sticky film which increased the friction force sharply. 40 Hardy and Hardy (1919), in working with glass on glass and glass on copper, found a threefold increase in friction force when a thin film of water was put on a lubricant film. No indication was given of the order of magnitude of the friction force with only a water film between the surfaces. Adding moisture to the conditioned glass surface and then wiping the surface dry increased the friction force to approximately the same value as when the surface was washed with carbon tetrachloride, then washed with water and dried. This indicated that moisture on the surface affected the friction force about the same whether the cu- tin deposits were present or not. However, it must be noted that the procedure for adding moisture to the surface in- volved no critical control of the amount. As far as the effect of the cutin itself is con— cerned, apparently the proportions of the components of cutin are such that when deposited on the surface it acts as an adhesive rather than as a lubricant. The presence of moisture on the surface also contributes to adhesion. Since the friction force increased to a constant value dur- ing the conditioning process, it is apparent that a point is reached where either there are no further deposits on the surface or further deposits on the surface have no ef- fect on friction force. The latter case seems most likely since the corn in the sample container was changed repeat- edly during conditioning. Possibly a point was reached in the buildup of molecular layers on the surface beyond 41 which the attraction by the surface for additional layers was very small. 4.24 Changing friction forces during the first part of the sliding period One more interesting aspect of Figure 4 is the in- crease in friction force during the first part of the pass when sliding shelled corn on the conditioned glass surface. This was not evident when sliding on the glass surface after having been washed with carbon tetrachloride. A first at- tempt to explain this phenomenon might be that the deposits of cutin were greater during the latter part of the pass, thus creating a higher friction force at this stage. How— ever, certain evidences indicated that this was not the primary cause. When the glass sheet was turned end for end and sliding took place in the opposite direction on the surface, the trace recorded was almost identical to Trace A. The pass was then begun at the halfway point of the surface. The trace recorded could almost be superim- posed on the first portion of Trace A. Care was taken to take up all the slack in the force sensing system before starting the pass so as to reduce any vibration that could have affected the friction force. Also, a check of the force indicating system showed that the increase during the first part of the pass was not originating there. A pad was formed from several thicknesses of cloth. This pad was placed on the surface beneath the sample con- tainer and a normal load applied. With cloth sliding on 42 the conditioned glass surface the increase in friction force during the first part of the pass was virtually eliminated. This indicated that the increase was originating with the shelled corn. Increases in friction force during the first por- tion of sliding have been found by Roth 25 El. (1942) with rubber sliding on glass and steel and Hall (1962) with wheat sliding on sheet steel. A possible explanation as set forth by Roth £5 31, was that the water molecules became oriented on the surface and long chains of the oriented water mole- cules were formed. The rise in friction force existed be- cause a finite time was required for formation of these chains after sliding began. The long chains were supposed to extend from the surface into the interior of the liquid film. This may be a possible explanation for the rigid water film observed by Eversole and Lahr (1941). Although Roth 23.21, found an increase during the first part of the pass in the majority of cases they observed that the fric- tion force decreased with sliding when using a very low velocity of 0.0001 centimeters per second. Hardy and Doubleday (1923) refer to a latent period that accompanies static friction of lubricated surfaces. This latent period was defined as the time required for the friction force to decrease to a constant value after applying a lubricant to the solid face. Latent periods of 10 to 60 minutes were observed. Hardy and Doubleday felt that the latent period was the time required for the 43 molecules of the lubricant to become oriented at the solid face. The latent period could also have been the time re- quired for the lubricant to spread out as a continuous film over the surface. Of particular significance was the fact that Hardy and Doubleday (1923) detected a latent period only when using lubricants whose molecular structure was a chain of carbon atoms with a carboxyl or hydroxyl group at one end. No latent period was found with lubricants whose molecules were composed of a chain of carbon atoms and both ends of the molecule were alike. What makes this so interesting is that fatty acids of plant origin are typically a chain of carbon atoms with a carboxyl group at one end (Fruton and Simmonds, 1958). It is interesting to speculate on whether or not there is any connection between the decrease in static coefficient of friction found by Hardy and Doubleday (1923) and the increases in kinetic coefficient of friction during the first portion of the sliding period found in this study and the studies of Roth gtflal. (1942) and Hall (1962). At first it would not seem that rubber has anything in common with grains or fatty acid lubricants; however, the rubber compounds used by Roth gtnal. contained a small amount of stearic acid, an unsaturated fatty acid commonly found in nature. Stearic acid has a carboxyl radical at one end. Of course, it must be recalled that the exact composition of the cuticular layer of a shelled corn kernel 44 is not known but it is known that the cuticle contains nu- merous fatty acids with carboxyl radicals. At this point the available evidence indicates that fatty acids act in peculiar ways with regard to friction. An attempt to set forth a rigorous explanation of the various effects of fatty acids on friction of grains is handicapped by the lack of knowledge of the exact com- position of the cuticular layer on the kernels. However, a few general ideas may be set forth. First, it is known that a temperature rise occurs at the interface of one solid body sliding over another and that this temperature rise may be quite significant (Bowden and Tabor, 1956). Possi- bly this temperature rise is associated with the increas-1 ing friction force during the first part of the sliding period. A thermocouple was embedded in a corn kernel as close to the surface of the kernel as possible. This ker- nel was then placed in the sample container with the thermo- couple next to the grain-surface interface and the container was filled with grain. The remainder of the cycle was ex- ecuted in the usual manner except that the friction table was permitted to run back and forth continuously for a total of 31 times. The sliding velocity was three inches per second. An overall temperature rise of 2° F. was measured by the thermocouple in the kernel. This indicated that friction did cause a temperature increase but was by no means an accurate indication of the temperature rise 45 occurring at the interface. The large temperature rises that Bowden and Tabor spoke of were only a few thousandths of a second duration. These temperature flashes would be sufficiently long to greatly affect the surface of the ker- nel but since the thermocouple was a finite distance from the kernel surface and had a finite mass, the effect of the temperature rises on it would not be so great. The melting points of some of the fatty acids that might be found in cutin are around or somewhat above room temperature. If a melting point were above room tempera- ture, a temperature rise at the surface of the kernel might cause the fatty acid to liquify. Since the friction force increased during the first part of the sliding period the friction force must be greater with the fatty acid at the interface in liquid form than if it were in solid form. One observation made during the continuous running of the table that is worth noting was that a temperature rise was indicated by the thermocouple during the first part of most passes at approximately the same time as the increase in friction force occurred. The temperature measured by the thermocouple then reached a constant value. When the table was reversed the temperature would decrease indicating that the position of the kernel on the surface was being changed and the thermocouple was moved farther from the interface. If the thermocouple measurements can at least be taken as an indication of the temperature trend at the interface the increase in temperature during the first part of the 46 sliding period may be associated with the increase in fric- tion force in this same interval. Why did Hardy and Doubleday (1923) find that static coefficient of friction required a certain time to decrease to a constant value after application of a fatty acid lubri- cant and why did Roth gtflal. (1942) find a decrease in fric- tion force during the first part of the sliding period at a velocity of 0.0001 centimeters per second? In the first case the lubricant was applied to the solid surface in vapor form. If some of the fatty acid constituents had melting points or at least condensing points above room temperature, which seems likely, a portion of the lubricant would pass into the liquid or even the solid phase. In order to find agreement with the hypothesis that a change‘of phase of fatty acids from a solid to a liquid causes an increase in friction coefficient, the static coefficient of friction should have decreased, which it did. Some results from Roth 25‘31, (1942) for rubber sliding on glass are shown in Figure 5. At the very low velocity of 0.0001 centimeters per second where a decrease in friction force during the first part of the sliding period was observed, the temperature rises would have been very small and may indeed have been negligible. Thus it might be assumed that very little or no liquefaction of the fatty acid component of the rubber compound occurred. Also, prior to each test the rubber specimens were cleaned with acetone. The decrease in friction force during sliding 47 Figure 5. Coefficient of friction versus distance moved for rubber sliding on a glass surface at various velocities (Roth £2 21,, 1942) 5 Lo cm/sec. 0.I cm/sec. 4 / ‘ 0.0l cm/sec. g / If». 3 / a: / "5 2% :§ Qflmcmfime E 2 Q I 0.0001 cm/sec. 0 0 IO 20 30 Distance Moved, cms. 48 may have been the time required for the fatty acid to form a thin film and then act as a boundary lubricant at the points of contact of the rubber specimen on the glass and steel surfaces. As mentioned previously, Adam (1938), Bowden and Tabor (1950) and Gemant (1950) indicated that fatty acids are good boundary lubricants. At the velocities where Roth'gtual. (1942) found an increasing friction force during sliding the increase was more significant at the higher velocities (see Figure 5). Also, the time required for the friction force to in- crease to a constant value became less as velocity was in- creased with the exception of 0.001 centimeters per second. Larger temperature rises would be expected at the higher velocities. So more of the fatty acid would be expected to liquefy and also liquefaction would be accomplished in a shorter period of time. At the velocity of 0.001 centi- meters per second the friction force reached a constant value sooner than at the higher velocities. At this veloc- ity the temperature rise would have been less. In addition, the times of contact of the peaks of the rubber specimen with the peaks of the glass surface would have been longer because of the lower velocity so that more of the heat of friction would have been conducted away by the glass sur- face. This would reduce liquefaction. It seems then that the amount of fatty acid lique- faction that may occur is a function of the time of contact of the asperities as well as the temperature rise occurring 49 at the points of contact. These two variables are a func- tion of velocity. At very low velocities the amount of liquefaction may be considered negligible and the fatty acid acts as a boundary lubricant. As velocity is increased the temperature rises become of consequence and cause lique- faction of the fatty acids which in turn causes friction force to increase. As velocity is increased farther the temperature rises become more significant and at the same time the times of contact of the asperities decrease caus- ing a greater portion of the heat of friction to be trans- ferred to the surface of the sliding body. Because of these two reasons the friction force will experience a greater increase during the first part of the sliding period at the higher velocities and the increase will take place in a shorter period of time. The question may now arise as to why the increase in friction force during the first part of the pass was not observed when sliding shelled corn over the glass sur- face after washing the surface with carbon tetrachloride. A plausible explanation may be based on the following sup- position. The glass surface in a clean state would have a much greater attraction for the cutin on the kernels than after a film of cutin had been deposited on the surface. As sliding takes place a certain amount of the heat of fric- tion would be transferred to the cutin at the points of contact. But, rather than this cutin remaining on the ker- nel, it would be attracted to the glass surface taking the 50 heat it had adsorbed with it. Even if a sufficient tempera- ture rise developed to cause a change in phase of the fatty acids the fatty acids which underwent the change would be left behind as a film on the glass surface. As the conditioning process progressed the attrac- tion by the glass surface for the cutin on the kernels would become less because of the cutin already on the surface. More of the cutin would remain at the points of contact of the kernels thus allowing greater concentrations of heat. The friction force would continue to increase during the first part of the pass because of more fatty acids changing phase. 4.25 Conditioning of a sheet steel surface with shelled corn Some recorder traces from the conditioning process with shelled corn on sheet steel are reproduced in Figure 6. Conditioning was carried out with shelled corn at 10 percent moisture content, wet basis. The sliding velocity was 3.5 inches per second with a normal load of 32.4 pounds. The roughness of the steel surface varied little with con- ditioning, being 33—38 microinches, rms, before and 30-35 microinches, rms, after. The steel surface when received was covered with an oil film. This was removed by washing with hexane. The steel surface was then washed with acetone, allowed to dry and Trace A in Figure 6 was recorded. The use of carbon tetrachloride as a solvent was discontinued because 51 A4 momma aamoho own .amamm omcommdocoo a am aoammanlo mamoho nmmmmcmzmlo memoho rmmaomnm mamoao mummMIav moammam voozhma a no mamommm cmoo oamamnu mow om>oa ooaammmo namme> aomom aommommm mo umoamm mmomoomm .a mmSOmm Illv wuuz amxu. r Mr .. 1.3.x: L30 _ o .. I . .. Tllllrl. AT 4 . a . . r . 7. .I . . . o .4 1T. I I. u ~ . Ivy .I . . . r o « Y. f1. . . . . I .. . . . . iv . . . v . .1 . . v ¢ 3 a . . o .. . . . . illil o.o.~v L o..- . 'U..m .‘.,.. . . . a v a . v . . ~ . . . a . c 4 . v . v . . . . . . . I . . o . . r i . . . . r . . . o u . . o u v . ‘ I9 v 0‘ e o 9 f . . Arte? I u . a n I 4 ¢ . 4 Ammama omcommmocoo a am moammau .4 momma mamoao ooelm no momma namoao main no momma maon eco mom amannmoema mam om oaaomxm mummlu m4 momma ammoho um um mmcommoa :mmz momma unmanas mammaloa moaammmo mammm> momom commommm mo mmoamm mmomooem .m mmsumm .o . ... a . . .A I . .. .... .I‘ . . . . . .. I.. . . . c I . . ..- A . o . . 4 ... .AY . ,. . . I . . . .. .... . .. . . . ...o ..I4. oIo.o 9.1. A ..VI .ov Io. . VIIIeo. II. D . . n7 .. .1 . . . . . . . . . . 1 . . . . . . . i u I. ., . . . o 0 iv o .V . . . . . . . . 1 . . . . . . . .. ~ . . . .1 . . ‘o . . . . . . i . . . . . I ,. . . . . i I . . o. . . . . . . . . . o o. . . . . . . . . . v. . . o . . . . . . o . a e a. o . . o . . o ,. c . . c . . . . . . a . . . . e u 9 o 00 s v ,w I a A . . . . ... . . . . .. . . . - . o . Y. . . . . . . ov- . . . . . . o . . . It 52 of its possible toxic effects. Trace B was recorded 12 cycles after Trace A. During this period the shelled corn in the sample container was changed each cycle, as it was during the entire conditioning process. The friction force indicated by Trace B does not differ appreciably from that of Trace A other than a small difference in the form of the trace. After recording Trace B the sheet steel surface was left exposed to the atmosphere for one hour and Trace C was then obtained. This exposure caused a 25 percent increase in the friction force. The friction force indi- cated by Trace D was found 13 cycles after Trace C. A de- crease in friction force can be seen; however, the friction force did not return to the value found before exposing the surface to the atmosphere. Trace E represents the friction force of shelled corn on the conditioned steel surface. It was obtained 400 cycles after the condition- ing process began. As was the case with shelled corn on glass, the increase in friction force during the first part of the pass was absent when the steel had been washed with acetone. With the steel surface in a conditioned state the increase was indeed present--to a much greater extent than was found with shelled corn on glass (see Figure 4, Trace A). The overall increase in friction force during the conditioning of glass with corn was 50 percent, whereas the condition- ing of sheet steel with corn resulted in less than a 30 53 percent increase. The deposits seem to have had less of an effect on friction force on the rougher steel surface than on the smoother glass surface. This greater effect on a smoother surface is in accordance with the findings of Wieneke (1956), who found that deposits from fibrous mate- rials had a greater effect on friction coefficient on the smoother surfaces. A 25 percent increase in friction force resulted from the one hour exposure of the surface to the atmosphere. This was probably due to the accumulations of moisture and dirt on the surface from the atmosphere, since with further sliding of the corn over the surface the friction force decreased. Another possible cause could have been the pres- ence of drying oils on the surface. Drying oils, as dis— cussed by Fieser and Fieser (1957), form into dry, tough and durable films when spread on a surface exposed to air. They are used as components of paints and varnishes. If cutin contains drying oils, even in small amounts, the in- crease in friction force during exposure to the atmosphere may have been due to their presence on the surface. How- ever, it seems likely that their effects would not be solely responsible for the increase, but would be additive to the effects of moisture and dirt accumulations from the atmo- sphere, since a similar increase was found when exposing a clean glass surface to the atmosphere. Another difference between sliding shelled corn on the steel surface and on the glass surface both in 54 conditioned states was that of the relative amounts of de- posits along the sliding path. Previous evidence indicated that the deposits of cutin were the same all along the glass surface. However, it appeared that the deposits were some— what greater over the latter part of the sliding path on the steel surface. The increase in friction force as slid- ing took place was still evident when starting the pass at the halfway point on the steel surface. But a constant friction force was reached more quickly than when beginning the pass at the normal starting point. This difference was likely due to the greater roughness of the steel sur- face having more of an abrasive effect on the kernels' points of contact. When the conditioning process was completed and a constant friction force had been found, the steel surface was again washed with acetone. The subsequent recorder trace was essentially the same as Trace A in Figure 6. 4.26 Conditioning of a plywood surface with shelled corn The progress of the conditioning process with shelled corn on a plywood surface is illustrated by the recorder traces reproduced in Figure 7. The shelled corn was at 10 percent moisture content, wet basis. Sliding velocity was 3.5 inches per second. The normal load was varied from 16.2 to 48.6 pounds with 16.2 pounds being used in the ma- jority of the cycles. All recorder traces depicted in Fig— ure 7 were obtained with a normal load of 16.2 pounds. The 55 surface roughness of the plywood varied considerably through- out its length. Also, some smoothing resulted from the conditioning process. The surface roughness of the softer grained portions decreased from 70-110 to 50-80 microinches, rms, and of the harder grained portions from 40—70 to 30-50 microinches, rms. All of the readings were taken parallel with the grain of the plywood which is the direction in which sliding took place. Surface roughness measurements perpendicular to the grain were considerably greater (150- 220 microinches, rms). The plywood surface was not washed prior to condi- tioning because of the possible effects of the solvent on the surface. Trace A in Figure 7 represents the friction force of shelled corn on the plywood surface just as re- ceived. Trace B was recorded three cycles after Trace A. A considerable decrease in the maximum friction force oc- curred with only four cycles. The friction force represented by Trace C was found 16 cycles after recording Trace B. Trace D was recorded with the plywood surface in a condi- tioned state 380 cycles after the conditioning process began. The peak in friction force evident in Trace D of Figure 7 occurred as the shelled corn slid over a section of the plywood surface that was made up of a higher percent- age of harder grained material than the remainder of the surface. This indicated that the friction coefficient in- creased as surface roughness decreased. Wieneke (1956) also found this in his work with fibrous materials. Further 56 confirmation of this phenomenon is based upon comparison of the coefficients of friction of shelled corn (10 percent moisture content) on conditioned plywood and steel surfaces. The coefficient on the smoother steel surface was observed to be higher than on the rougher plywood surface. Notably absent is the increase in friction force during the first portion of the pass over the surface with shelled corn on the conditioned plywood surface (see Figure 7, Trace D). This was rather disappointing in view of the hypothesis set forth to explain the increases during the first part of the passes with shelled corn on the condi— tioned glass and steel surfaces. However, earlier in the study an increase had been found with shelled corn sliding on a different plywood surface. Unfortunately the signif- icance of this was not realized at the time and no rough- ness measurements were made on the first plywood surface. Undoubtedly the variation found was associated with differ- ences in the characteristics of the two surfaces. 4.27 Microscopic examination of the surfaces Figures 8 through 11 are photomicrographs of the sheet steel and plywood surfaces. The figures are 100X actual size. Figure 8 is representative of the microscopic appearance of the steel surface as it was received and after the oil film had been removed. The contrast in the photo- micrographs resulted from variations in reflectance of light due to the irregularities of the surfaces. As can be noted, 57 Photomicrograph of steel surface in original condi- tion (lOOX) Figure 10. Photomicrograph of harder grained portion of plywood surface in original condition (100x) Figure 9. Photomicrograph of steel surface after testing with shelled corn and A barley (100x) Figure 11. Photomicrograph of softer grained portion of plywood surface in original condition (100x) 58 the surfaces are very rough on a microscopic scale. The photomicrograph in Figure 9 was taken after the portion of the testing program in Sections 4.31 and 4.33 with shelled corn and barley on sheet steel had been completed. This was not necessarily the same spot on the surface shown in Figure 8. The tracks made by the kernels are evident in Figure 9. The gross effect of these tracks upon viewing the surface in actual size was a very shiny appearance. The tracks themselves were an integral part of the surface; i.e., washing with acetone did not remove them. If they were actual scratches they were very small in magnitude since a surface roughness measurement parallel to them was identical to one perpendicular. This was also the case with the steel surface in its original condition. Figures 10 and 11 are typical of the plywood sur- face in its original condition. Before taking these photo- micrographs it was necessary to allow black ink to soak into the surface in order to obtain sufficient contrast. The irregularities of the plywood surface with its original coloring could be distinguished when viewed through the microscope; however, the variations in color did not pro- vide sufficient contrast for the black and white film. The direction of the grain of the harder grained portion of the surface is obvious in Figure 10. This orientation was less evident in the softer grained portion as shown in Figure 11. No difference in surface appearance was noted after testing had taken place on the plywood surface. 59 4.28 Miscellaneous investigations Some investigations that were conducted on a lim- ited scale at various times throughout the study are in— cluded in this section. The first of these was a determi- nation of the effects on friction coefficient of major de- viations in the overall level of the surface on the fric- tion table. This was carried out with shelled corn sliding on a plywood surface. A l/4-inch spacer was placed under the plywood surface at various points giving the surface a concave shape or a convex shape. The various positions of the spacer did slightly alter the form of the recorded trace of friction force versus distance moved. However, when considering that the table as used for testing varied not more than 0.05-inch in vertical displacement through- out its length, the effect would seem insignificant. The next point considered was the effect of temper- ature of the shelled corn sample on the friction force of shelled corn sliding on a plywood surface at room tempera- ture. Corn samples were tempered at 0° F., 40° F., 70° F. and 100° F. for a period of 12 hours and trials were then run. Little variation in friction force occurred among the different temperatures. At one point it was thought that a reorientation of kernels may have been taking place in the sample con- .tainer at the sliding surface. In an effort to determine this shelled corn kernels were glued to an 8-inch diameter plywood disk. The disk was fastened to the underside of 60 the sample container with the kernels down. Very erratic results were obtained. Sliding took place in stick-slip fashion producing a widely varying trace. No conclusions could be drawn. Later observations from underneath a glass surface showed no visible reorientation or movement of the kernels with respect to the sample container. The actual area of contact between the grain and the surface was a particularly intriguing issue. Efforts to determine this were not successful. It would be desir- able to know the initial area of contact as well as the variation, if any, in area of contact during sliding. 4.29 Recommended testing procedure The recommendations in this section were formulated on the basis of knowledge gained during previous investi- gations in the study. Of course, they are primarily appli- cable to the equipment used in this study, but, in general, are appropriate for most any kinetic friction testing equip- ment. It is believed that friction coefficients of a re- liable and a reproducible nature can be obtained by follow- ing these recommendations. Conditioning of the surface is an essential com— ponent of the recommended testing procedure. Whenever pos- sible the surface should be washed with a volatile solvent before beginning the conditioning process. This gives a clean surface with which to begin conditioning and also hastens the conditioning process. It is advisable to change 61 the grain in the sample container every cycle during con- ditioning. (The procedure for executing a cycle has been given in Section 3.2.) This limits the possibility of wear- ing through the pericarp of the kernels. The conditioning process should be continued until a constant friction force is obtained. It is best to carry out the conditioning process over a period of several days. This allows any drying oils that may have been deposited on the surface to harden (see Section 4.25). Every effort should be made during conditioning, as well as during sub- sequent testing, to protect the surface from contamination by external sources. Examples are touching the conditioned surface with one's fingers, allowing water to splash on the surface, etc. The surfaces used during the portion of the study covered by Section 4.3 were protected by the polyethylene sheet described in Section 3.12. It is desirable to measure the roughness of the surface before and after conditioning. Measurement during the conditioning process may also be desirable if it can be done conveniently. When a conditioned surface has been obtained the determination of coefficients of friction may be started. The grain in the sample container should be removed and mixed with the main sample every cycle. At least ten rep- 1ications should be made with each combination of variables. If a general increase or decrease in friction force is noted during these ten replications, the replicating process should 62 be continued until a constant friction force is found. This is particularly important when testing higher moisture content grains. When testing higher moisture content grains it is important to limit the loss of moisture from the grain. A controlled temperature and relative humidity chamber in which the testing apparatus could be located is the ideal way to limit this loss. Equilibrium moisture contents at particular temperatures and relative humidities could be used. In the event such a chamber is not available, the higher moisture content grain being used for testing should be kept in a closed container as much as possible. Also, the main sample should be large in comparison with the por— tion required to fill the sample container. Moisture content, relative humidity and temperature should be determined periodically. Also, the calibration of the force sensing and recording system should be checked before and after running the required number of replications with a particular combination of variables. 4.3 Determination of Some Kinetic Coefficients of Friction 4.31 Kinetic coefficients of friction of shelled corn on plywood Kinetic coefficients of friction of shelled corn on plywood were determined considering the effects of nor- mal load, velocity of sliding and moisture content. Normal loads of 16.2, 32.4 and 48.6 pounds were used. These loads correspond to one, two and three foot depths of shelled 63 corn at 56 pounds per bushel, respectively. Also, they are equivalent to normal unit pressures of 0.31, 0.63 and 0.94 pounds per square inch, respectively. Velocities con- sidered were 3.5 and 7.0 inches per second. Data were taken at all normal loads and moisture contents using the velocity of 3.5 inches per second. Data at the velocity of 7.0 inches per second were taken only at normal loads of 16.2 and 48.6 pounds and moisture contents of 10 and 20 percent, wet basis. Moisture content of the shelled corn was varied from 10.5 to 26 percent, wet basis. Above 26 percent moisture con- tent excessive vibrations occurred owing to stick and slip. The testing procedure of Section 4.29 was followed. The plywood surface reached a conditioned state after ap- proximately 380 cycles. Surface roughness of the plywood for the tests was 30-80 microinches, rms. The experimental results of the friction tests with shelled corn on plywood are presented in Figure 12. Only data taken at 3.5 inches per second were plotted since it was felt that vibrations occurring in the apparatus at 7.0 inches per second rendered data at this velocity invalid. Also shown in Figure 8 are the least square regression lines and standard errors of estimate calculated from the data. These were found by grouping the 10 replications at each normal load at a particular moisture content into an aver- age. This was justified by the fact that the normal loads used had no definite effect on coefficient of friction. The equations for the least square regression lines 64 ocoomm mme mmnocm m.m mo >mmooam> a ma ammo HmmcmEmmmaxm mcmsonm ooozhmm co :moo omHHmsm mom mammcoo mmsmmmoa mammm> commommm mo mammommmmoo .NH mmammm 33m $3 o\o .2280 95562 8 a a 8 e e z N. e _ 8° ”:35 a :85 8325/ / a Ha I \ \ U I H H lw \. \ “I ma \\ x V \ \ \ “z: gamma“; . A s. . as 3.33 33 .2532 n. .2: ¢.~n u 925 .2352 G .3. «.2 u o m mm ammo Hmmcmfimmmmxm mcm3onm Hmmmm mmmnm co cmoo omaamnm mom mammcoo mmammmoe mammmb commommm mo mammommmmoo .ma mmzmmm 38m $3 o\o €2.30 22%: cw Nu cu m. o. 2 N_ o. o «we 36 m2... 22353: I/ Emir—.8 “.o 82.5 2.3245 I / '7 8.6 71 _ \m III... as D uouogu g0 Iuagomaoo %\ 33.3. 23 .2352 D n .s. can . 23 .2352 a .s. 3. u 23 3:52 o and end 72 estimate are plotted as solid lines in their particular moisture ranges. They are then extrapolated as broken lines into the range of 17.5 to 20 percent moisture content. Moisture content in the lower range has more of an effect on coefficient of friction of shelled corn on sheet steel than of shelled corn on plywood. However, the effect illustrated by Figure 14 is within the standard error of estimate so was considered insignificant. Coefficient of friction of shelled corn on sheet steel began to increase significantly at a moisture content of 19 percent, wet basis. This is the same moisture content at which the desorption isotherm of Brunauer, Emmett and Teller (1938) for six mo- lecular layers intersected the experimental desorption iso- therm of Rodriguez-Arias (1956) (see Figure 13). This sup- ports the hypothesis that coefficient of friction of shelled corn is affected by moisture content when kernel pores greater than approximately 35 Angstroms in diameter begin to fill with moisture. Shelled corn at 80° F. has an equilibrium moisture content of 19 percent, wet basis, at approximately 92 per- cent relative humidity (see Figure 13). This indicates that, as far as the corn itself is concerned, friction co- efficient will not be affected by relative humidity until humidity reaches 92 percent. However, Bowden and Tabor (1950) found that adhesion between two glass surfaces showed a marked increase at relative humidities exceeding 80 per- cent. The relative humidity at which friction coefficient 73 is affected by adhesion due to the presence of moisture then would seem to be dependent on the adsorption charac- teristics of the grain as well as the adsorption character- istics of the surface upon which it is sliding. Bowden and Tabor also point out that the effects of adhesion are less when dealing with rougher surfaces. This seems to be sub- stantiated in this study by the fact that coefficient of friction increased at a greater rate with moisture content in the case of shelled corn on sheet steel than with shelled corn on the rougher plywood surface. 4.33 Kinetic coefficients of friction of barley on sheet steel Kinetic coefficients of friction of barley on sheet steel were determined for two normal loads, two velocities and five moisture contents. Normal loads were 16.2 and 32.4 pounds on an overall area of 51.9 square inches. The friction force of barley on sheet steel at a normal load of 48.6 pounds exceeded the strength of the force sensing beams. Velocities were 3.5 and 7.0 inches per second, but again excessive vibration occurred at the latter velocity. Moisture content was varied from 10 to 23 percent. The testing procedure of Section 4.29 was followed. Approximately 500 cycles were made before the steel surface was deemed to be in a conditioned state. The surface rough- ness of the steel sheet was 30-35 microinches, rms. The results of testing barley on sheet steel are presented in Figure 15 together with the least square 74 onoomm mme monocm m.m mo hmwoon> m mm ammo HmmcmEmmmmxm amazoSm Hmmmm mmmnm so hmmman mom mammcoo mmdmmAbE mammm> commommm mo mammommmmoo .ma mmsmmm £25 a; o\o .2380 95582 a a 8 e o. s m. e mono one WEE—mu mo 32.5 3324.5 \ ' \ \\III C I mW and \ x 4 II \ \ xx \ II I I Ir .\ \ \. m. I I. \ L m. .\\\\CV\H\.\ IIV Locomw \ .5... 5335mm \\ m: J... a“ as m. 2:3» . 23 35:82 a w. .33.! ... Q43 .2582 0 etc etc 75 regression lines and standard errors of estimate. The re- gression equations were calculated for data in the moisture content ranges of 10 to 16.5 percent and 21 to 23 percent, wet basis. The equations and their standard errors of esti- mate are given in Table 7. Table 7. Regression equations and standard errors of esti- mate for coefficient of friction of barley on sheet steel as a function of moisture content Moisture content Standard error (percent, wet basis) Regression equation of estimate lO-l6.5 CF = 0.388 — (5.8 x 10‘4) MC 0.007 21—23 CF = 0.248 + (7.5 x 10’3> MC 0.005 As in Sections 4.31 and 4.32 the regression equa— tions and standard errors of estimate are plotted as solid lines in their respective moisture content ranges and are extrapolated as broken lines into the range of 16.5 to 21 percent. Coefficient of friction decreased slightly with increasing moisture content in the lower moisture content range. However, the decrease was within the standard error of estimate in that moisture range so was considered insig- nificant. Moisture content began to have a significant effect on coefficient of friction when it reached approx- imately 17.5 percent, wet basis. This is somewhat lower than the 19 moisture content at which coefficient of fric- tion began to increase with shelled corn on sheet steel. 76 It is unfortunate that the constants for the B—B-T three- constant equation are not available for barley. An anal— ysis for barley, similar to that made for shelled corn in Section 4.31, on the number of molecular layers theoretic— ally present at 17.5 percent moisture content could be very enlightening. A notable difference was found in comparing the conditioning process of barley on steel with that of shelled corn on steel. The recorded trace of friction force versus distance moved obtained when sliding barley on the steel surface washed with acetone was of approximately the same form as obtained with shelled corn sliding on the steel surface in the same condition (see Figure 6, Trace A). However, the average friction force for barley on the washed steel surface was about 10 percent lower. This is rather surprising in view of the fact that the friction force for barley sliding on a conditioned steel surface was approx— imately 40 percent higher than the friction force for shelled corn on a conditioned steel surface. Whereas the friction force increased 29 percent during the conditioning process with shelled corn on steel, the increase during the condi- tioning process with barley on sheet steel was 97 percent. V. SUMMARY AND CONCLUSIONS 5.1 Summary A knowledge of the kinetic coefficients of friction of grains allows more accurate analyses to be made of the forces occurring in grain handling equipment. The results are more efficient and economical handling devices which will better suit the particular handling requirements and will better fit into the overall system. Apparatus was assembled to study kinetic coeffici— ents of friction of grains. Extensive preliminary tests conducted with shelled corn on sheet steel and plywood in- dicated a need for establishing a dependable testing pro- cedure and technique that would yield friction data of a reliable and a reproducible nature. Factors other than moisture content, normal load and velocity of sliding were found to affect coefficient of friction. A glass surface was chosen for initiation of explor— atory investigations of friction of grains because of the stability it presented to the atmosphere and the likelihood of its having uniform characteristics throughout its length. One of the results from the work with glass was that water vapor adsorbed on the surface affected the friction coef- ficient of shelled corn on glass considerably. However, the effects of water vapor deposited on the surface from the atmosphere were not too great and indications were that 77 78 the effect would be approximately the same up to relative humidities of at least 80 percent. A large number of cycles were required to obtain a constant friction force with shelled corn on glass. The friction force increased during the conditioning process. With the glass surface in a conditioned state, the friction force increased during the first portion of the pass over the surface; then reached a constant value. The hypothesis was formed that the increase during the first portion of the pass was due to liquefaction of the fatty acids that are a component of the cuticular layer of the corn kernels. It was indicated that the magnitude and extent of this in- crease were a function of velocity. The increase during the first part of the pass was not present when sliding shelled corn over the acetone washed glass surface. The conditioning process of a sheet steel surface with shelled corn closely followed that of the glass sur- face with shelled corn, except the overall increase in fric- tion force during conditioning was 29 percent with sheet steel as compared to 50 percent with glass. The deposits from the cuticular layer of the corn kernels seemed to have less of an effect on the rougher steel surface than on the smoother glass surface. Four hundred cycles were required to condition the sheet steel surface with shelled corn. A plywood surface was conditioned with shelled corn with approximately 380 cycles. The increase in friction force during the first portion of the pass over this 79 conditioned surface was absent. But an increase was found using a different plywood surface. Photomicrographs taken of the sheet steel and ply- wood surfaces showed them to be very rough on a microscopic scale. Miscellaneous investigations indicated the follow- ing: 1) major deviations in the overall level of the sur- face on the friction table had no effect on coefficient of friction; ii) the coefficients of friction of shelled corn at 0° F., 40° F., 70° F. and 100° F. on a plywood sur- face at room temperature did not differ appreciably; and iii) shelled corn kernels held in the sample container showed no visible reorientation at the surface during sliding. A recommended testing procedure was presented for obtaining reliable and reproducible coefficients of fric- tion. These recommendations included the importance of conditioning the surface, changing the grain in the sample container each cycle, preventing outside contamination of the surface and limiting the loss of moisture from higher moisture content grains while testing. Kinetic coefficients of friction of shelled corn on plywood and sheet steel and of barley on sheet steel were determined considering the effects of moisture content and normal load. The effects of velocity were not studied because of limitations of the testing apparatus. Normal load, in the range studied, did not affect coefficient of friction significantly. 80 Moisture content began to affect coefficient of shelled corn at approximately 19 percent, wet basis. Ac- cording to the B-E-T theory of adsorption this is the point at which kernel pores exceeding 35 Angstroms in diameter begin to fill. Coefficient of friction at moisture contents exceeding 19 percent, wet basis, increased at a greater rate on sheet steel than on the rougher plywood surface. Also, the coefficients of friction were generally higher on the smoother steel surface than on the rougher plywood surface. Moisture content began to affect the coefficient of friction of barley on sheet steel at about 17.5 percent, wet basis. The conditioning process with barley on sheet steel required approximately 500 cycles. The increase in friction force during conditioning was 97 percent compared to 29 percent with shelled corn on sheet steel. 5.2 Conclusions 1. Friction coefficients of grains are affected by variables other than moisture content, normal load and velocity of sliding. 2. Materials deposited by grain on a solvent washed surface cause coefficient of friction to increase. For this reason it is necessary to condition a surface with a particular grain before determining coefficients of fric- tion. 3. A greater increase in friction force occurs 81 during the conditioning process with barley on sheet steel than with shelled corn on sheet steel. 4. Friction force increases during the first por- tion of the pass of shelled corn over a conditioned glass or steel surface and of barley over a conditioned steel surface. The hypothesis was presented that this increase is partially due to the liquefaction of fatty acids that are a component of the cuticular layer of the kernels. Another contributing factor, in some cases, is the varia- tion in deposits from the kernel along the sliding path. 5. Water vapor adsorbed on a glass and steel sur- face affects coefficient of friction considerably. However, indications were that the effects of water vapor contrib- uted by the atmosphere would be approximately the same up to relative humidities of at least 80 percent. 6. Major deviations in the overall level of the surface on the friction table do not affect coefficient of friction. 7. The coefficients of friction of shelled corn at 0° F., 40° F., 70° F. and 100° F. on a plywood surface at room temperature do not differ appreciably. 8. Normal load in the range of 16.2 to 48.6 pounds on an overall area of 51.9 square inches does not have a significant effect on the coefficient of friction of shelled corn on sheet steel or plywood. Neither does it have a significant effect in the range of 16.2 to 32.4 pounds on coefficient of friction of barley on sheet steel. 82 9. Moisture content begins to affect coefficient of friction of shelled corn at about 19 percent, wet basis. According to the B-E-T theory of adsorption this is the point at which kernel pores exceeding 35 Angstroms in di- ameter begin to fill. 10. Coefficient of friction of barley on sheet steel begins to be affected by moisture content at 17.5 percent, wet basis. 11. Coefficients of friction of shelled corn slid- ing on sheet steel are higher than those of shelled corn sliding on plywood. 12. The coefficient of friction of shelled corn on sheet steel is increased more by increasing moisture content in the higher ranges than coefficient of friction of shelled corn on plywood. REFERENCES Adam, N. K. (1938) The Physics and Chemistry of Surfaces. Oxford: Clarendon Press. Agricultural Engineers Yearbook (1964) St. Joseph, Mich- igan: American Society of Agricultural Engineers. Airy, Wilfred (1898) The pressure of grains. Minutes of Proceedings of the Institution of Civil Engin- eers (London), 121, 347. Baker, E. A., Batt, R. F., and Martin, J. T. (1964) Stud- ies on plant cuticle. VII. The nature and deter— mination of cutin. The Annals of Applied Biology, 53, 59. Balis, J. S. (1958) Methodology on the Determination of Friction Coefficients of Agricultural Materials. Unpublished Master of Science Thesis, Purdue Uni- versity. Blevens, F. Z. (1954) Some of the Component Power Require- ments of Field-type Forage Harvesters. Unpublished Master of Science Thesis, Purdue University. Bonis, L. J. (1964) Surface analysis. Industrial and Engineering Chemistry, éé, 7, 43. Bowden, F. P. (1955) Recent studies of metallic friction. Institution of Mechanical Engineers Proceedings, 169, 7. Bowden, F. P., and Hughes, T. P. (1939) The friction of clean metals and the influence of adsorbed gases. The temperature coefficient of friction. Proceed- ings of the Royal Society of London, Series A, 112, 263. Bowden, F. P., and Leben, L. (1938) The nature of slid- ing friction and analysis of friction. Proceedings of the Royal Society of London, Series A, 169, 371. Bowden, F. P., and Tabor, D. (1938) The area of contact between stationary and between moving surfaces. Proceedings of the Royal Society of London, Series A, 169, 391. 83 84 Bowden, F. P., and Tabor, D. (1950) The Friction and Lubrication of Solids. Oxford: Clarendon Press. Bowden, F. P., and Tabor, D. (1956) Friction and Lubri- cation. New York: John Wiley and Sons. Bowen, P. H., Boes, D. J., and McDowell, J. R. (1963) Wear and friction characteristics of self-lubri- cating composite materials. Machine Design, 359 19, 139. Bowen, P. H., and Hickam, W. H. (1963) Outgassing char- acteristics of dry lubricant materials in a vacuum. Machine Design, fig, 16, 119. Brubaker, J. E. (1962) Determination of Static Coeffici- ents of Friction of Some Grains on Various Struc- tural Materials. Unpublished Master of Science Thesis, University of Toronto. Brubaker, J. E., and P05, J. (1963) Determination of static coefficients of friction of some grains on various structural surfaces. American Society of Agricultural Engineers Paper 63—828. Brunauer, S., Emmett, P. H., and Teller, E. (1938) Ad- sorption of gases in multimolecular layers. Jour— nal of the American Chemical Society, 69; 309. Burmistrova, M. F., et.gl. (1956) Physicomechanical Prop- erties of Agricultural Crops. Translated from Rus- sian. Published for the National Science Founda— tion by the Israel Program for Scientific Transla- tions, January, 1963. OTS 61-31216. Available from Office of Technical Services, United States Department of Commerce, Washington 25, D.C. Claypoole, W., and Cook, D. B. (1942) The nature of static friction. Journal of the Franklin Institute, 233, 453. Clayton, D. (1945) A short review of surface finish in relation to friction and lubrication. Institution of Mechanical Engineers Proceedings, 153, 332. Courtel, R., and Tichvinsky, L. M. (1963a) A brief his- tory of friction--Part one. Mechanical Engineer- ing, g5, 9, SS. (1963b) A brief history of friction--Part two. Mechanical Engineering, g5, 10, 33. de Boer, J. H. (1953) The Dynamical Character of Adsorp- tion. Oxford: Clarendon Press. 85 Dies, K. (1945) Fretting corrosion, a chemical—mechanical phenomenon. Engineers Digest (American Edition), 2, 14. Dokos, S. J. (1946) Sliding friction under extreme pres- sures--I. Transactions of the American Society of Mechanical Engineers, 68, A-l49. Esau, K. (1961) Anatomy of Seed Plants. New York: John Wiley and Sons. Eversole, W. G., and Lahr, P. H. (1941) The thickness of the rigid water film at a quartz-water interface from a measurement of Newton's rings. Journal of Chemical Physics, 23 686. Fieser, L. F., and Fieser, M. (1957) Introduction to Or- ganic Chemistry. Boston: D. C. Heath and Company. Fruton, J. 3., and Simmonds, S. (1958) General Biochem- istry. New York: John Wiley and Sons. Gemant, A. (1950) Frictional Phenomena. Brooklyn, N.Y.: Chemical Publishing Company. Hall, C. W. (1957) Drying Farm Crops. Ann Arbor, Mich- igan: Edwards Brothers. Hall, G. E. (1962) Instructor, Ohio Agricultural Experi- ment Station, Wooster, Ohio, personal correspond- ence, December 29. Hardy, W. B., and Doubleday, I. (1923) Boundary lubrica- tion:--The latent period and mixtures of two lubri- cants. Proceedings of the Royal Society of London, Series A, 1249 25. Hardy, W. B., and Hardy, J. K. (1919) Note on static fric- tion and on the lubricating properties of certain chemical substances. Philosophical Magazine, 389 32. Hintz, O. E., and Schinke, R. (1952) Coefficient of slid- ing friction for corn and alfalfa on steel. Report 3-A186G-52R1. Hinsdale, Illinois: International Harvestor Company. Hunter, M. C. (1944) Static and clinging friction of pivot bearings. Institution of Mechanical Engineers Pro- ceedings, 151, 274. Jamieson, J. A. (1904) Grain pressures in deep bins. Engineering News,.§l, 236. 86 Kelley, D. V. (1957) New standard for surface roughness, waviness, and lay. Instruments and Automation, 30, 86. Kisselbach, T. A. (1949) The structure and reproduction of corn. Research Bulletin 161. Agricultural Ex— periment Station, University of Nebraska College of Agriculture. Kramer, H. A. (1944) Factors influencing the design of bulk storage bins for rough rice. Agricultural Engineering, 25, 463. Lorenzen, R. T. (1959) Moisture effect on friction coef- ficients of small grain. American Society of Agri— cultural Engineers Paper 59—416. McHaffie, I. R., and Lenher, S. (1925) The adsorption of water from the gas phase on plane surfaces of glass and platinum. Journal of the Chemical Soci- ety (London),'1gl, 1559. MacCurdy, E. (translator and arranger) (1955) The Note— books of Leonard da Vinci. New York: George Braziller. Marks, L. S. (1951) Mechanical Engineers Handbook, Fifth Edition. New York: McGraw Hill Book Co., Inc. Merchant, M. E. (1940) The mechanism of static friction. Journal of Applied Physics, 11, 230. Morgan, F., Muskat, M., and Reed, D. W. (1941) Friction phenomena and the stick-slip process. Journal of Applied Physics, 133 743. Parker, R. C., Farnsworth, W., and Milne, R. (1950) The variation of the coefficient of static friction with the rate of application of the tangential force. Institution of Mechanical Engineers Proceedings, 163, 176. , Peterson, M. B., and Murray, 8. F. (1963) How to use prop- erties of materials to aid lubrication. Machine Design, 32, 21, 200. Pleissner, J. (1906) Versuche zur Ermittlung der Boden und Seitenwanddrucke in Getreidesilos. Zeitschrift des Vereines Deutsche Ingenieure, 30, 976. Rabinowicz, E. (1956) Stick and slip. Scientific Amer- ican, 194, 5, 109. 87 Richter, D. W. (1953) Equipment for the Reduction of Dairy Chore Labor. Unpublished Doctor of Philosophy Thesis, Cornell University. Richter, D. W. (1954) Friction coefficients of some agri- cultural materials. Agricultural Engineering,.3§, 413. Ries, H. E., Jr. (1961) Monomolecular films. Scientific American, 204, 3, 152. Rodriguez-Arias, J. H. (1956) Desorption Isotherms and Drying Rates of Shelled Corn in the Temperature Range of 40° to 140° F. Unpublished Doctor of Phi- losophy Thesis, Michigan State University. Roth, F. L., Driscoll, R. L., and Holt, W. L. (1942) Fric- tional properties of rubber. Journal of Research of the National Bureau of Standards, 283 439. Schmidt, A. D., and Weiter, E. J. (1957) Coefficients of flat-surface friction. Mechanical Engineering, '_7_9_, 1130. Schnurmann, R. (1940) Angle of repose of snow on solids. Nature, 145, 553. Schnurmann, R. (1940) Braking high-speed trains as an engineering problem and the problem of solid fric— tion. Mechanical Engineering, 63, 408. Schnurmann, R. (1940) The mechanism of static friction. Journal of Applied Physics, 11, 624. Wieneke, F. (1956) Reibungswerte von Pflanzen und Faser- stoffen. Landtechnische Forschung, Q, 146. Wolf, M. J., Buzan, C. L., MacMasters, M. M., and Rist, C. E. (1952) Structure of the mature corn kernel. II. Microscopic structure of pericarp, seed coat and hilar layer of dent corn. Cereal Chemistry, 22, 334. APPENDIX A Table A.1 Friction forces for shelled corn sliding on ply- wood at various moisture contents, normal loads and a velocity of 3.5 inches per second Moisture Relative Content humidity Temperature Normal load (lbs.2 (% wet basis) (%) (° F.) 16.2 32.4 48.6 11.2 11.2 11.2 11.3 11.2 11.3 11.2 11.2 11.2 11.2 10.5 45 83 14 43 85 11.5 11.4 11.3 11.4 11.3 11.3 11.3 11.4 11.4 11.3 0 O I C . COCO 0000000000000090041 00000000000000000000 00000000000000me 17 56 80 11.5 11.5 11.4 11.5 11.5 11.4 11.5 11.5 11.5 11.5 12.3 12.2 12.3 12.4 12.2 12.0 12.2 12.3 12.0 12.2 21 S7 81 CHNOKOOl-‘MHN OIONONONQUIOIO‘OSON qmmmmmmmmm Uth‘lohU'lU'ILfloD-U'Iw ##3## wwwwwwwwww wwwwwwwwww wwwwuwwwww O (Doommqmmmmm \IQQQQQQQQQ \IQQQQQQQQQ QQQQQQQQQQ 89 90 Table A.1 (continued) Moisture Relative Content humidity Temperature No ma 0 d bs. (% wet basis) (%) (° F.) 16.2 32.4 48.6 23.5 48 82 4.0 8.3 12.7 4.2 8.3 12.7 4.2 8.4 12.6 4.1 8.3 12.6 4.1 8.3 12.6 4.2 8.3 12.4 4.3 8.4 12.5 4.2 8.4 12.5 4.2 8.4 12.5 4.2 8.2 12.5 26 56 83 4.3 8.8 13.6 4.4 9.1 13.7 4.3 8.9 13.9 4.4 8.9 13.6 4.3 9.1 13.4 4.3 9.1 13.6 4.3 9.0 13.5 4.4 9.0 13.4 4.3 8.8 13.4 4.4 8.9 13.4 Table A.2 Friction forces for shelled corn sliding on sheet steel at various moisture contents, normal loads and a velocity of 3.5 inches per second Moisture Relative content humidity Temperature Normal load (1bs.) (% wet basis) (%) (° F.) 16.2 32.4 48.6 12.9 13 .0 13.1 13.2 13.2 12.9 13.7 12.8 12.8 13.0 10 19 80 13.5 22 74 14.2 14.0 13 .8 14.4 13.6 14.1 14.3 13.7 13.7 13.7 13.8 13.7 13.5 13.5 13.9 13.9 13.7 13.3 13.7 13.7 17.5 18 72 20 18 72 14.1 14.5 14.5 14.3 13.9 14.6 13.9 13.9 13.9 14.2 OWOQHNNQWN NHNbwawOH mwpqmqmbpp mmooqqmmmoom mummqoomoomq WDHwWWDbanP- WWNUWWWNND) mwbmmpbbwp hub-hbb-hbnbnbb ubnbubobobb-hD-bnb bfihbbbbhfib ubfinbubnh-h-h-b-hub KOKOKDQKOKDQODCDKO \D\D\O\D\O\D\O\OO\O 00000000000000000000 00000000000000000000 91 92 Table A.2 (continued) Moisture Relative content humidity Temperature Normal load (lbs.) (% wet basis) (%) (° F.) 16.2 32.4 48.6 22 18 74 4.9 9.7 15.5 4.8 9.6 14.7 4.8 9.4 14.9 5.0 10.3 15.4 4.7 9.8 15.1 4.6 9.7 15.1 4.5 10.0 15.0 4.8 9.8 15.1 4.6 10.1 15.2 4.7 9.8 14.8 Table A.3 Friction forces for barley sliding on sheet steel at various moisture contents, normal loads and a velocity of 3.5 inches per second Moisture Relative content humidity Temperature Normal load (lbs.) (% wet basis) (%) (° F.) 16.2 32.4 10 44 80 12.2 12.2 12.2 12.2 12.5 12.3 12.4 12.4 12.3 12.5 14 43 80 12.5 12.6 12.6 12.5 12.4 12.2 12.4 12.5 12.5 12.5 16.5 54 75 12.4 12.3 12.2 12.2 12.2 12.3 12.3 12.3 12.2 12.4 13.2 13.0 13.3 13.3 13.3 13.2 13.3 13.1 13.3 13.3 21 57 75 mmmmmmmmmm mmmmmmmmmm mmmmmmmmmm mmm'mmmmmmm pmmbwpmpmm WNOOQOHDNI-‘N OONNNONOOH wNNNNl-‘l-‘HHN 93 94 Table A.3 (continued) Moisture Relative content humidity Temperature Normal load (lbs.) (% wet basis) (%) (° F.) 16.2 32.4 23 61 73 6.9 13.4 6.8 13.6 6.7 13.6 6.9 13.7 6.8 13.8 6.7 13.6 6.8 13.6 6.7 13.9 6.8 13.5 6.8 13.7 QNS TEUNV "filfiujlnzmgmjfl Intqnyrrlnfllumjflflfljai“ 2