PNEUMA'HC FRUIT HARVESTING AND ASSOCIATED FRUIT CHARACTERISTSCS Thesis fee the Donne of M. S. MICHIGAN STATE UNWERSITY Hamid Eugene Quackonbush 1961' This is to certify that the thesis entitled Pneumatic Fruit Harvesting and Associated Fruit Characteristics presented by Harold Eugene Quackenbush has been accepted towards fulfillment of the requirements for M_°S°—degree in A iCUJ-tural Engineering LA A 31er Major professor Date January 18, 1961 0-169 PLACE IN RETURN BOX to remove this checkout from your record. TO AVOID FINES return on or before date due. MAY BE RECALLED with earlier due date if requested. DATE DUE DATE DUE DATE DUE FEB 2 7 2011 0624 11 2/05 p1/CIRClDateDue indd-p.1 PNEUMATIC FRUIT HARVESTING AND ASSOCIATED FRUIT CHARACTERISTICS by HAROLD EUGENE CUACKENBUSH ABSTRACT Submitted to the Colleges of Agriculture and Engineering of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN AGRICULTURAL ENGINEERING Department of Agricultural Engineering Approval 6 A Shut ,I/M/i: _ 11 Harold Eugene Quackenbush The main objective of this study was to investigate a possible method of harvesting tree fruits. The present method of mechanically harvesting fruit consists of shaking the main branches of a tree with a mechanical shaker to remove the fruit. The fruit after being shaken loose falls to a catch- ing device beneath the tree. Because considerable damage results to some species of fruit when this method is used, another method of harvesting is desired. The possibility of using air movement to lower the falling fruit slowly to a catching device without bruising was investigated. Fruit removal from the tree was accomplished by hand-shaking the tree limbs. Preliminary tests were conducted on various species and varieties of fruit to determine some of the physical charac- teristics. These data were used to determine terminal veloc- ities of the test fruit which provided design values for the air velocity necessary to suspend or float the fruit. A test duct was constructed and connected to a fan system in the laboratory. A device was constructed above the duct opening to measure impact forces of falling fruit. The impact measuring device consisted of SR-A strain gages in a Wheatstone bridge arrangement mounted on a cantilever beam. An amplifier and oscillograph recorder were connected to the strain gages to record the results. Fruit was connected to the cantilever beam by a nylon 111 Harold Eugene Quackenbush line. Tests were conducted by attaching various species of fruit to the line, pulling the line and fruit upward to dif- ferent heights and allowing it to drop. At the bottom of the fall, a knot in the line came into contact with the beam and produced a deflection which was measured as an impact force. Several species of fruit were tested by dropping them in air streams ranging in velocity from zero to the terminal velocity of the test fruit. The impact force data provided a method of determining a height of drop in still air that would give an impact force equivalent to that obtained from dropping the fruit in a moving air stream. The laboratory data were used in designing a field test machine. This machine represented only a portion of a full scale machine, but the principle involved was the same as for a full scale machine. The results from this machine were not completely satisfactory; however, some basic principles were established. This study revealed that heavier fruit, dropped in a con- fined air stream, required greater air velocities for suspen- sion than lighter weight fruit of the same species. Greater air velocities were necessary because the velocity required to suspend fruits was disclosed to be dependent upon the ratio of fruit weight to projected area. Tests indicated that this ratio increases with increasing weight. iv Harold Eugene Quackenbush Laboratory data indicated that air velocities necessary to effectively handle fruits by pneumatic methods were very‘ nearly equal to the terminal velocities of the fruits. Theo- retical calculations gave the terminal velocity for a McIntosh apple weighing 0.432 pound as 8,180 feet per minute. Labora- tory tests for this apple revealed that an air velocity of 6,500 feet per minute reduced the impact force for a five foot drop to an equivalent value received for a drop of A.2-inches in still air. The results from the field test machine indicated that fans capable of providing sufficient air velocities are essen- tial and may need to be of special design if further work proves that fruit can be harvested by pneumatic methods. PNEUMATIC FRUIT HARVESTING AND ASSOCIATED FRUIT CHARACTERISTICS by HAROLD EUGENE QUACKENBUSH A THESIS Submitted to the Colleges of Agriculture and Engineering of Michigan State University of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE IN AGRICULTURAL ENGINEERING Department of Agricultural Engineering ACKNOWLEDGEMENTS The author wishes to express his appreciation and thanks to the following people who contributed to this investigation: Dr. Bill A. Stout, who as my major professor provided inspiration, constant supervision, and assistance throughout the entire investigation and preparation of this manuscript. Dr. Arthur H. Farrall, Dr. Merle L. Esmay, and members of the guidance committee for administration of the graduate program. Dr. Stanley K. Ries, Horticulture Department, and staff at the horticultural farm for providing fruit and trees used in the experimental investigation. Dr. Clement Tatro, Applied Mechanics Department, for the use of the Brush Strain Analyzer and oscillograph. Mr. James Cawood, foreman of the research laboratory, for providing facilities and assistance on construction of test apparatus. Messrs. Glen Shiffer and Harold Brockbank, employees of the research laboratory, for assistance in construction of the field model harvester. Mr. Max Austin, Horticulture Department, for assistance in testing and evaluating fruit. Messrs. Dale Marshall and Donald Pettengill, student assistants,for their help in the laboratory and field tests. vii My wife, Dolores, for her suggestions and assistance in preparing this manuscript and especially for typing it. The author also deeply appreciates the financial support, made available by the Department of Agricultural Engineering, through a graduate research assistantship. TABLE OF CONTENTS Page INTRODUCTION ..................... 1 LITERATURE REVIEH .................. u PHYSICAL CHARACTERISTICS OF VARIOUS SPECIES 0? FRUIT - 15 Apparatus ----------- ' ......... 15 Procedure -------------------- 16 Results --------------------- 18 Discussion ------------------- 23 THEORETICAL ANALYSIS - - -. -------------- 29 Horsepower Consideration ------------ 3h EXPERIMENTAL INVESTIGATION -------------- 38 Laboratory Tests ---------------- 38 Apparatus ----------------- 38 Procedure ----------------- #2 Results ------------------ 50 Discussion ----------------- 61 Analysis of Error Involved --------- 63' Conpressibility Consideration ------- 67 Field Tests ------ » ------------- 68 Design of the Machine ----------- 68 1 Air velocity Measurements -' -------- 73 Test Procedure - - - - - - - - 1 ------ 76 Evaluation of the Machine --------- 78 TABLE OF CONTENTS, Continued Page COST ANALYSIS -------------------- 85 CONCLUSIONS --------------------- 89 RECOHIENDATIONS FOR FUTURE STUDY ---------- 91 SUMMARY ----------------------- 93 REFERENCES --------------------- 95 APPENDICES --------------------- 99 I ----------------------- 100 II ----------------------- 109 III ---------------------- 114 VII VIII N XII LIST OF TABLES Page Critical forces for apples ----------- 5 Average values fro-.50 fruit for each species and variety -------------------- 19 Results of regression analyses --------- 22 Terminal velocities -------------- 35 Theoretical terminal velocities computed with the results froa.sanples of 50 fruits ----- 59 Equivalent drop distances and corresponding impact forces ----------------- 60 Reduction of impact forces and equivalent heights 1 of drop when using air to lower the fruit - - - 62 Acceleration of fruit in free fall with line attached -------------------- 65 Percent of error in velocity ---------- 66 Classification of apples bruises -------- 80 lumber of Jonathan apples contained in each bruise category for various drops ------- 118 Number of Northern Spy apples contained in each bruise category for various drops ------- 119 Figure 10 11 12 13 14 LIST OF FIGURES Page Relation between weight and projected area for Stanley Prune plums -------------- 20 Relation between weight and projected area for Red Raven peaches --------------- 21 Relation between weight and projected area for Northern Spy apples -------------- 101 Relation between weight and projected area for Jersey blueberries -------------- 102 Relation between weight and projected area for Elberta peaches ----------- - - - - - 103 Relation between weight and projected area for Montmorency cherries ------------- 10A Relation between weight and projected area for Nontgamet apricots -------------- 105 Relation between weight and projected area for Jonathan apples ---------------- 106 Relation between weight ”and projected area for Cortland apples ---------------- 107 Relation between weight and projected area for McIntosh apples ---------------- 108 Relation between the weight and the force to remove McIntosh apples from.the tree ----- 115 Relation between the weight and the force to remove Red Haven peaches from the tree - - - - 116 The engine, fans, and duct assembly used for laboratory tests --------------- 43 Upper portion of the duct extending from floor below and some of the equipment used ----- A3 Figure 15 16 17 18 19 2O 21 . 21a 22- 22C 23 23a 2” 2ha xii LIST OF FIGURES, Continued Relation of engine drive speed to air velocity in duct .................... Cantilever been with strain gages mounted above the duct outlet ................ V—shaped trough above the duct outlet shielding the cantilever beam from the air stream - - - - Device used for attaching apples to the canti- leverbeam ----—--—.' ......... An apple hanging inside the plastic duct from the cantilever beam -------------- Section of oscillograph tape with recorded results of a typical drop ----------- Relation of impact force to height of drop for a McIntosh apple ............... Relation between the actual drop in an air stream and an equivalent drop in still air for a lchntosh apple --------------- Relation of impact force to height of drop for a Hontguet apricot -------------- Relation between the actual drop in an air stream and an equivalent drop in still air for a Hontgamet apricot -------------- Relation of impact force to height of drop for a Red Raven peach - - ~ ------------- Relation between the actual drop in an air stream and an equivalent drop in still air for a Red Raven peach ------- . ........ Relation of impact force to height of drop for a Stanley Prune plum ............. Relation between the actual drop in an air stream and an equivalent drop in still air for t Stanley Prune plum ------------- Page 2&6 1&6 118 A8 A9 53 54 55 57 58 Figure 25 26 27 28 29 30 31 32 33 3h 35 36 37 xiii LIST or FIGURES, Continued Position of fans beneath the fruit catching and conveyor belt ............... Pneumatic fruit harvester with a limb inside the confining walla .............. Over-all view of the pneumatic fruit harvesting machine with the author viewing a limb inside - Method of transporting the harvesting machine - Measurement of air velocity inside the walls with a pitot tube ............... Placement of a limb inside the confining walls Limb completely enclosed by use of a canvas flap over the opening ------------ v- - - Apples being delivered to a box from the con- veyor belt ------------------ Apples that were harvested with the pneumatic fruit harvester ................ Number of apples receiving bruises when shaken loose by hand and lowered pneumatically - - - - Number of bruises Occurring on the apple when shaken loose by hand and lowered pneumatically Percent of total apples (80) receiving bruises when shaken loose by hand and lowered pneu- matical 1y ------------------- Percent of total apples (50) receiving bruises -when dropped manually from various heights onto the machine with the fans not operating - - - - Page 71 72 72 ‘71: 7a 77 77 79 79 81 81 83 83 LITERATURE REVIEW Fruit growers and research workers have partially solved the fruit harvesting problem by providing successful methods of mechanically harvesting some species of fruit, but for others the only efficient method is still hand-picking. Some fruits are not readily adaptable to mechanical har- vesting, due to their susceptibility to bruising. Catching and handling is difficult since most fruits cannot be dropped from great distances onto sharp edges or hard surfaces. Experiments performed by Gaston and Levin (1951) indi- cated that drops of three inches or more caused skin breaks when tender-skinned apples struck sharp edges or wires. When McIntosh apples three inches in diameter were dropped onto a hard surface from a height of one inch, the resulting bruises averaged one-half of an inch in diameter. The larger the apple the more susceptible it was to mechanical injury. They also found that apples may be bruised by pressures as well as drops. When a gradually increasing forCe was applied to a portion of the surface area of an apple, no apparent damage occurred until a point was reached at which a number of cells collapsed and a visible bruise resulted. The magnitude of force causing the cells to collapse was designated as the critical force. Tests made on 80 apples showed that none of the apples in the trials were bruised by forces of less than It is apparent that mechanization of fruit harvesting has become necessary for growers to maintain a profitable operation. The importance of mechanization can be seen not only in reducing harvest costs and solving the labor problem but also in providing increased production. Mechanizing the fruit harvest has been left mainly to the growers, small manu- facturing companies, united States Department of Agriculture, and universities although a few major machinery companies have shown some interest. Methods that have been tried in mechanizing the fruit harvest include mobile platforms, mobile ladders, hydraulic booms, and picking tubes. These techniques have not been entirely successful because of the complexity of problems involved. Fruits, unlike other mechanically harvested pro- ducts, bruise easily and are highly perishable; therefore, they must be handled quickly and carefully. Tree structure is a problem as it varies with age, variety, and method of Pruning employed making some fruit hard to reach. A machine that will effectively harvest tree fruit must not cause injury to the tree because it will seriously impair the following year's production. ‘ The objectives of this research study were to: (1) Collect data pertaining to fruit in order to establish basic physical characteristics. (2) Investigate the use of air as a possible method of harvesting fruit by shaking it loose from the tree and allowing it to fall without bruising. This investigation included: (A) Laboratory tests to determine (B) (C) (D) l) the capabilities of a high velocity air stream in reducing the descent velocity of fruit. 2) flotation (terminal) velocities of various species of fruit. 3) impact forces received by fruit after fall- , ing in a high velocity air stream. Calculation of horsepower requirements, fan size, and number of fans required. The design and construction of a field machine which would demonstrate pneumatic possibilities which might be adopted in principle to a com- mercial machine. Testing and evaluation of the field machine under actual conditions in an orchard. INTRODUCTION In production of tree fruits in the United States in 1958, Michigan ranked third in apples, fifth in peaches, fourth in pears, second in plums and first in sour cherries. The total value of Michigan production of these crops was $35,423,000 and for the united States was $421,472,000. Pro- duction costs for these fruits have increased while the aver- age price per unit received by Michigan growers from 1953 to 1958 has decreased from six to 28 percent.1 A major portion of the production costs can be attributed to harvesting. Levin (1959) reported that harvesting and handling costs amounted to over 50 percent of the cost of pro- duction for some species of fruit. In addition to high har- vest costs, recruitment of labor to harvest the fruit has become a major problem. Much of the available labor lacks experience. They are demanding higher wages and improved housing while in return, the growers sometimes receive a poor quality of harvested fruit due to carelessness of the workers. Adrian, Fridley, and Kaupke (1959) reported that hand harvest- ing of tree fruits in California required 60 to 100 man-hours per acre while in the highly mechanized harvest of small grains only three man-hours were required. 1Michigan Agricultural Statistics, July 1959. 7.5 pounds. Tests with different varieties of apples two and one-half inches in diameter produced the results given in Table I. TABLE I Critical forces for applesa (For each force a three-eights inch diameter bruise resulted on a two and one-half inch apple.) Variety Forceb . 1b. Wealthy ' 7.5 McIntosh ' . 8.5 Northern Spy 12.0 Jonathan 18.0 a. Table reproduced from "How to Reduce Apple Bruising," Gaston and Levin, 1951. b. Data based on 160 bruise measurements made on 80 apples. To determine whether apple picking could be successfully mechanized and what characteristics an effective mechanical picker should possesa.a time and motion study was conducted by Gaston and Levin (1953). Their study revealed the follow- ing pertinent factors: (1) The motions involved in picking apples are selective, diversified and complex. (2) Considerable mechanical injury takes place during conventional picking operations. (3) Approximately 40 percent of the fruit are picked from the ground and 60 percent from ladders. (h) Seventy-three percent of the workers time is spent picking apples. (5) Nineteen percent of the workers time is spent mov- ing fruit to and placing it in crates, and return- ing to picking position. (6) Three percent of the worker's time is spent moving his ladder. (7) Five percent of the workers time is spent in smok- ing, eating apples, and in other non-productive activities. These factors indicated that 22 percent of the workers time (excluding rest) was spent on unproductive activities; there- fore, a mechanical device which could eliminate a part of this time should enable a worker to pick more fruit. They stated that many growers have tried to eliminate this time loss by using mobile platforms and ladders. Development of these methods led to the design of the "steel squirrel." This machine permitted the operator to position himself almost anywhere in the tree to facilitate Picking. Hill and Brazelton (1955) reported that five years of field tests indicated that a worker on the steel squirrel could do from one and one-half to two times as much as he could do on a ladder. They stated that costs for owning and operating the machine ranged from 21 cents to 33 cents per hour with the difference being attributed mainly to engine life and availability of repairs. The machine was widely adaptable to various climates and type of jobs. Harvesting methods employing hand—shaking, pole, pneumatic and hydraulic shakers were tried on Michigan grown fruits and evaluated by Gaston, Levin, and Hedden (1958). Experiments included harvesting trials with red tart cherries, sweet cher- ries, plums, pears, and blueberries. Chemicals, to loosen the fruit so that mechanical separation would be easier, were also used. In conjunction with the shakers, a catching device was used under the tree to catch the fruit as it fell. Many types of catching devices were constructed and tested to find one which would work effectively with a shaker mechanism. Tests indicated that 80 to 90 percent of the crop could be harvested with the hydraulic shaker and a cloth covered semi-circular catching frame on cherries and plums. This method was, how- ever, objectionable for use on pears since considerable injury occurred. Comparison of harvesting costs for red tart cher- ries revealed that for mechanical harvesting it cost one cent POP pound as compared to the usual grower cost of two and one- half cents per pound. Bruising studies indicated that tart cherries could be harvested with no more bruising than occurs When they are hand-picked. Development of hydraulic shakers and catching frames as a method of fruit harvesting has been continued by a number of fruit growers constructing and testing catching devices in an attempt to increase the efficiency of operation. Among these growers is Friday Tractor Company, located in south- western Michigan, who during the 1960 season developed a catch- ing device which conveyed the fruit up sloping sides to a con- veyor that moved the fruit to a tank of water. Fork-lift equipment was adapted for handling the water tank. Gaston and Levin (1960) reported that blueberries could be harvested mechanically by means of a hand held vibrator that moves metal fingers through an amplitude of one-fourth to one-half of an inch at the rate of 700 to 800 cycles per minute. To remove the berries, the fingers of the vibrator are held against the fruit bearing stems. A vibrator could separate blueberries from five to ten times as fast as the work could be done by hand. Levin, Gaston, Hedden, and Whittenberger (1960) reported that a tractor-mounted hydraulically-activated boom-shaker was developed in 1958, by Gould Brothers Incorporated of San Jose, California, for harvesting nut crops. Tests conducted on red tart cherries with this shaker during the 1959 season indicated that 95 percent of the cherries could be separated from the tree. The grade of unsorted mechanically harvested cherries varied from 70 to 95 percent U. S. No. 1. Total mechanical harvesting costs varied from one-half to over two and one-half cents per pound. Under conditions existing in many orchards, mechanical harvesting enabled seven men to do the work of 33 hand-pickers and reduced harvesting costs by one-half. They also stated that several hundred bushels of "juice" apples were harvested with machines and placed in con- tainers at a per-crate cost of approximately three cents. They concluded that improved collecting units may make it pos- sible to reduce the amount of bruising so that apples which are to be made into baby food or apple sauce could also be harvested mechanically.“ A time and motion study was made by Adrian, Fridley, and Kaupke (1959) on boom type tree shakers coordinated with a low-profile self-propelled catching frame. Prune harvesting tests with this equipment indicated that it was possible to harvest 30 boxes per man-hour with a shaker speed of 30 trees per hour. This was six times the rate of the average hand harvest. Bruising damage to the prunes amounted to about six percent. A harvest rate of 60 trees per hour with a three man crew was obtained in 1960. Bulk handling of the fruit was required for this harvest rate. .This method also was tried on peaches, apricots, and olives, but to minimize fruit on fruit impact, baffles were needed to decelerate the fruit before it fell on the conveyor area. With these fruits, two additional problems were apparent. These consisted of damage to the fruit before it separated from the tree and damage to the fruit by limbs as it dropped through the tree. Levin (1958) reported that time and motion studies revealed that on the average a human picker separates apples from the tree and places them in a field box at the rate of one every 10 three seconds. From this fact, it is apparent that a mechan- ical harvesting method for apples will need to be at least this efficient and must accomplish the job without causing mechanical injury to the fruit. * Hydraulic boom type shakers in combination with catching frames have contributed considerably to a solution of the har- vesting problem for some tree fruits, but a problem still remains for others. An annual report by Adrian (1958) stated that boom type shakers have not proven satisfactory for use in the coastal areas where normally three to four harvests are required. As a solution to this problem, it was felt that a pulsating air blast might shake the trees sufficiently to remove the fruit. Tests were conducted on prunes in 1958 with a John Bean speed sprayer on which the nozzle had an oscillator nattachment to produce a pulsating air stream. The unit was moved down the rows at ground speeds of one and two miles per hour. Fruit removal was nearly the same as that for hand- shaking. Adrian reported that 1959 trials included a blower on which the rate of air flow and frequency of cycling could be varied. A Buffalo centrifugal fan was used which had noz- zle areas of one-half and one square foot. The nozzles were driven through an arc of about 90 degrees in order to subject the tree to several sharp blasts of air. Ground speed was about one-half mile per hour. The nozzle oscillated at speeds that varied from 60 to 120 cycles per minute. Air velocities ranged from.5,hh0 to 9,150 feet per minute with the higher ll velocities being more effective. The controlling factor on fruit removal was the ratio of force required to remove the fruit from the tree to the weight of the fruit (F/W). This ratio (F/W) represented the number of g's acceleration required for fruit removal. In any event, it was found that a higher percentage of fruit must be removed for this method of harvest to be practical. The citrus industry is also experiencing labor problems in harvesting citrus fruit and, therefore, are looking toward mechanization as a solution.' Coppock and Jutras (1959) reported that preliminary tests of a boom type tree shaker for harvest- ing grapefruit and oranges proved unsatisfactory. Consider- able bark damage occurred from the limb grasping device on the shaker and only 50 to 80 percent of the fruit were removed with oranges being the most difficult to shake loose. A mobile 'pickers platform was constructed for use in studying the design requirements and economics of this general type of machine. It was found for picking valencia oranges that the operational picking rate of an inexperienced operator using the pickers platform was increased four percent over that for conventional methods. A new method being developed by Coppock (1960) is a “picking spindle." This device consists of spindles three inches in diameter and 2u-inches long mounted parallel on four and one-half inch centers. The spindles have tapered flights of rubber which form an auger. The flights rotate and gently pull on the fruit until it is detached from the tree. 12 The development of bulk bins for use in harvesting and handling fruit has helped tremendously in paving the way for mechanical fruit harvesting. ‘McBirney (1959) reported that about 186,000 bins were used in the Pacific Northwest area in 1958. Research revealed that apples and pears could be har- vested and handled in bins with no more injury than when har- vested,in standard boxes. Bulk bins usually hold about 25 to 27 apple boxes of fruit. The Northwest standard apple box has a volume about one percent greater than a standard bushel. Bulk bin size is commonly AT-to 48—inches square, 29-to 33— inches high outside and 24-to 28-inches high inside. They weigh from 100 to 150 pounds and their gross weight when filled with apples is approximately 1,000 pounds. There are many advantages associated with their use, some of which are, faster picking, total cost savings, labor savings, improved fruit quality, and more storage capacity. It was reported that bulk handling saved Washington State about half a million dollars in 1959. McBirney also reported on some of the picking aids tried in the Pacific Northwest. Among these was a vacuum type pick- ing unit developed by a New York manufacturing firm for apples, oranges, and perhaps other fruit. This device consisted of a vacuum chamber on the end of a long supply tube attached to a vacuum supply. Fruit was placed in the vacuum chamber by an operator on the ground maneuvering the vacuum chamber to the fruit. The vacuum power unit exerted a pull on the fruit, 13 separated it from the branch and allowed it to fall to the ground through a cushioned spiral conveyor tube attached to the vacuum.chamber. Tests in 1956 with the unit indicated that apples could be picked without bruising, however, the unit was slower than hand-picking and it was very tiring to operate. Another method was a picking funnel and tube. The funnel was attached to the picker with the tube extending down to a bin. The worker was positioned by a mobile plat- form. Tests of this type of equipment indicated an increase in picking rate of about 40 percent. McBirney reported that one of the newest ideas is hedgerow planting of dwarf or semi- dwarf trees planted in 12 foot rows and pruned to give a four foot width. Picking will be accomplished by pulling a pick- ing machine through the space between the rows. The picking machine will consist of pickers on elevated platforms pick- ing about two feet into each tree row. An article appearing in the Farm Journal (1960) stated that dwarf trees are becoming popular throughout the country from Maine to Washington. Besides cutting costs, these trees start producing two or three years after planting, produce higher quality fruit than standard trees because of better spraying and pruning, produce up to 2,700 or 3,600 bushels of fruit per acre and eliminate dangerous and expensive ladder work. They are planted as a row crop six feet apart in the row with 12 feet between rows, giving 605 trees per acre. A grower in Grant county, Washington reported that his 1959 costs 14 with dwarf apple trees averaged $300 an acre. Aside from 14.2-cents per 40 pound lug for picking,his highest cost was $30 per acre for hand weeding. Similar costs in standard trees may run $500 to $700 per acre. Thus, dwarf trees reduced the cost per acre by approximately one-half. PHYSICAL CHARACTERISTICS OF VARIOUS SPECIES OF FRUIT Large variations among fruit species were evident, there- fore, it was necessary to establish basic values for some of the physical characteristics. These characteristics were necessary to facilitate theoretical calculations and provide information applicable to pneumatic fruit harvesting as described in this manuscript. The following information was obtained for each species and variety: (1) Average diameter. (2) Average weight. (3) Ratio of weight to projected area. (A) Average area. (5) Force required to remove the fruit from the tree. (6) Relation between force to remove the fruit from the tree and the weight. (7) Firmness of the flesh for some species of fruits. Apparatus sassy. A Hettler Precision Balance, model K7T, was used to weigh the fruit. This balance had a range of zero to 800 grams with scale divisions calibrated in tenths of a gram. Caliper and engineers scale An outside caliper was used to determine the fruit diameters 16 and an engineers scale used as a standard of comparison for obtaining the diameter measurements from the calipers. Pressure tester A Magness-Taylor pressure gauge was used to determine firmness of some of the fruit. This device consisted of a plunger, with a changeable tip for use with different fruit. The tip of the plunger was pressed into the flesh of the fruit where the skin had been removed. The resistance of the fruit to penetration of the plunger tip was registered on the handle of the tester in pounds. The plunger tip contained a line which indicated the depth that the tip should penetrate the fruit. Spring scale Spring scales with ranges of zero to 64 ounces and zero to 25 pounds were used to pull on the fruit and measure the force required to separate the fruit from the supporting limb. A small wire hook was constructed, for each type of fruit, as a means of attaching the fruit to the scales. Procedure Different species and varieties of fruit were selected from those available on the Michigan State University Horti- cultural Farm. The fruit was removed from the tree by the aPring scale for each species and variety chosen. This con- sisted of attaching a specially designed wire hook to the scale and then slipping the wire hook around the fruit. The 17 book was arranged in a stable position to permit application of a pulling force through the scale in a direction parallel with the fruit stem. Force was steadily applied to the scale manually until the fruit separated from its supporting limb. The magnitude of force occurring at separation was observed~ and recorded for a total of 50 fruits for each species and variety. Each fruit was given an identifying number by plac- ing the number on masking tape and than securing the tape to the fruit. Each of these fruits was weighed. Diameter measurements were taken with the caliper by measuring the maximum and mini- mum diameters. The maximum and minimum diameters occurred at different orientations on the fruits for the various species, but in most cases, for varieties of the same species, these measurements occurred at the same orientation. Firmness tests were made on peaches and apples with a Magness-Taylor pressure gauge. These measurements were obtained by removing a thin slice about the size of a nickel from the surface of the fruit. By holding the fruit in the palm of the hand, the plunger tip was placed on the cut surface and force applied to the gauge handle with the other hand to cause the tip to penetrate the flesh of the fruit. For peaches, a five-sixteenths inch diameter tip was used and a seven-six- teenth inch diameter tip used for apples. Two measurements were made on each fruit. 0n peaches, these measurements were obtained from the cheek and suture. The suture is defined as the seam extending from the stem end to the blossom end and 18 connecting the cheeks of the peach. The cheeks are located on either side of the suture. 0n apples, the firmness readings were obtained from the blushed and unblushed areas. The blushed area is the area where more color change has occurred and the unblushed is the area which lacks color compared with the blushed. The unblushed area is usually yellow in color and is called the "groun color or underlying color. The fruit selection for each species and variety, here- after called samples, was made at random from the tree and an attempt was made to choose fruits of various size, shape, and maturity on the same tree. Results The results are presented in Table II as average values for the samples of 50 fruits each.- The average fruit diameters appearing in Table II were obtained by averaging the maximum and minimum.diameter measurements of each fruit and then aver- aging these values for the entire sample to obtain single average values. The values of projected area and the ratio of weight to projected area were computed using the average values for diameters and weights of the various fruits. Graphs were constructed to determine the relation between weight and projected area for each sample of fruit. Two of these graphs are presented in Figures 1 and 2. The remaining graphs are presented in Figures 3 through 10 and are included in Appendix I. The graphs shown in Figures 1 and 2 were selected from the graphs for the entire group of samples as . . . . . m ocshm m H mmo 0 mm H mm a mao o >\ sofleeem mzmqm m.e a.» s.m ooo.o so.o me.m eoe.o onxo seasons . . . . . Ni.m mmeo H\® C0>fl~m 60m H m H 0H 0 m moo 0 mm m m m m mmmuema one a so . so. no.0 m .o oHo.o mm\e mosesosueoz a m x no mm o m o w mmHmmmmo mm.o «so.o 0H.o mm.o Hoo.o mwxm season mMHmmMQMDAm mm.o mmo.o mm.H ee.a ewo.o m\m ossomasoz do maoonmm< H.sH e.se m.m moo.o HH.m m>.m mos.o mh\oH nan chonchoz >.:H =.mH 3.: mmo.o 0H.m mm.m mwm.o mfi\oH senescou m.sfi s.mH m.m moo.o mm.m me.m mam.o wm\m cesarean o.mH o.mH m.: moo.o mm.m m>.m Ham.o mm\m snocho: . mmqmm< censuses: , sarcasm .oH .ss\.oa m.ss Aoomev o>oEom swam .nomm .cH .nH uno>som hoodns> .cH on .u: sos< no one zueam oumpommm op venom bosom no capsm .nosm haveamdm aswaoz even neaoodm avenues use someone some sou pasta om soon nesas> ewsne>< HH mamas 20 4 _ so . 1 as u mw m 0. ultimo +4 1 x l m 3 1: a O H O O. _ .. +. _ 5 m m 4;- w 0. w : M 1 V; l _ w .3 o no a 2m / p L l q {mime . M M l “S S M L a R e e _ e r, v p s a I? _. 4. “4 MW no no 4. n4 4 4 .3 3 3 3 nu in no no my no my .2. .0m \ .mmJ .4mm4 omkomfiomn. O._. F1003 no Cram GRAMS WEIGHT, Relation between weight and projected area for Stanley Prune plums. Figure l. IN. LBS. / SQ. RATIO OF WEIGHT T0 PROJECTED AREA, .080 .076i« .072 - .064 1060 1356 21 - /( Y = .0438 + .05338 x i .00205 ..:/i ,_ , -1 A,-______1 ./ I . ‘ ‘ {771/1 J l l l . l 3 .4 .5 .6 .7 ' l2/5/60 WEIGHT, POUNDS H. E. 0. Figure 2. Relation between weight and projected area for Red Haven peaches. 22 representing those with the greatest and least scatter of points from a line. The data on these two graphs were sub- jected to correlation and'regression analyses and the method of least squares used to obtain the regression lines through the scatter of points. Also, the standard error of estimate was computed and is shown on each graph by the dashed line on either side of the regression line. For clarity, only the results of these computations are shown in Table III. The details are presented in Appendix II. TABLE III Results of regression analyses ’— _:_‘ ‘—s Correlation Standard gquation of Regression Line Coefficient Error of y I Ratio of Fruit wt. per (r) Estimate proj. area Se) x : Weight of Fruit ‘7 ** i; _ 4 A = 0.028 +—0.000304 (X) Fig. 1 0.8M:M 0 00095 Y 5 83:3'30095“ .i A : 0.0u 8+—o.os33 K Fig. 2 0.995 0.00205 y 3 1:0.00205 (**) Highly significant. 9 = An estimate of y. A statistical analysis was not applied to those graphs appear- ing in Appendix I since the graphs of Figures 1 and 2 were chosen as representative of the extremities of all the samples. The data on the graphs in Appendix I are expected to give a correlation coefficient of 0.8A4 S r 5-0.995- RESP°831OB lines drawn on these graphs were not calculated, but were estimated by averaging a group of points at each end of the 23 plot and then constructing the line through the average of these points. Discussion Theoretical calculations which are presented in the fol- lowing section (THEORETICAL mums) disclosed that air velocities necessary to suspend fruit in an air stream are dependent upon the ratio of fruit weight to projected area. It was revealed by the data collected that this ratio varied among fruits of the same species. Therefore, the relationship of this ratio with fruit weight was established. The graphs presented in Figures 1 through 10 give the joint distribution of fruit weight and the ratio of fruit weight to projected area. _Corre1ation analyses were conducted with the data from two of these graphs (Figures 1 and 2) to determine the degree of association between these variables. Also, regression analyses on the same two samples were con- ducted for the "regression of the ratio of fruit weight to pro- jected area on weight." In statistics, the word "regression" means average relationship, thus the analyses presented gives the relation between the mean value of the weight to projected area ratio for a given value of weight. The regression lines for the graphs of Figures 1 and 2 permit estimating the ratio of weight to projected area when the weight is known. These P‘sressions should not, however, be used to estimate the weight with a known weight-projected area ratio. DOints on the graphs for all samples (Figures 1 through 10) The scatter of indicated that the relationship between these variables was linear. To establish whether or not a linear relationship could be assumed for calculation purposes a theoretical equation was established. The derivation of this equation consisted of representing the fruit weight as we 5’17 (1’) where: V? = specific weight of the fruit, pounds per cubic foot V = volume of the fruit, cubic feet Assuming the fruit as spheres permitted representing the volume as V’: E Z7 r3 (2’) Substituting eqiation (2’) into equation (ll) gives w a g ”'2? r3 and solving for (r) gives r = .EV/F‘ H;_fl_ 2‘- (3’) fi'd The ratio of fruit weight to projected area can be written as E- 24 flrr3 .1w (u’) A 377r‘L ,3 , and substituting, equation (3 ) into equation (4 ) gives 3 3 / led’ui /"‘3'w‘ nod/w (5) ‘ '3' ZIrnr - where: 4 LI \j‘ 3 3 3 717? Equation (5’) gives the relation between fruit weight and the Ca: ratio of fruit weight to projected area and reveals that instead of a linear relationship, a cubic relationship exists. 24a Values of the ratio (%) were computed with equation (5/) using the fruit weights encountered in this study. Graphs made with these data revealed that for the range in fruit weights for a species very little change in slope occurred and for all practical purposes a linear relationship could be assumed. The computations involved in the regression analyses were therefore based on a linear relationship. The coefficients of correlation (r) for the data contained in the graphs of Figures 1 and 2 are presented in Table III with double asterisks attached to the values indicating that each value was highly significant. The correlation coefficient for a sample of this size (N a 50) with N-2 degrees of freedom need only be equal to or greater than 0.279 to be significant at the five percent level and equal to or greater than 0.361 to be significant at the one percent level.2 These values of correlation coefficients to be significant, mean that when there are #8 degrees of freedom, only five percent of the time would a correlation coefficient as large or larger than 0.279 occur "by chance" if the true or population correlation coeffi- cient was zero. And only one percent of the time would a correlation coefficient as large or larger than 0.361 occur "by chance" if the true or population correlation coefficient was zero.3 Hence it can be concluded that the probability of being wrong, in using the regression line to estimate the weight- area ratio, is less than one percent. Therefore, the cor- relation coefficients obtained from the analyses can be 2Values of correlation coefficients obtained from Table 11, Walker and Lev. 3Garrett, 1956. 25 labeled highly significant indicating that there was a high degree of association between the two variables on each graph. The equations of the regression lines can be used to com- pute an esthnated weight-area ratio if the weight is known. These equations, however, provide an estimated value that falls between the value obtained, plus or minus the standard error of estimate. This same type of analysis could be applied to those graphs appearing in Appendix I. The author feels, however, that. the results obtained with the graphs given in Figures 1 and 2 would include within these limits any results obtann- able fron.the remaining samples. These analyses revealed that the ratio of fruit weight to projected area does not remain constant along fruit of a species but increases with increas- ing fruit weight. Since a high degree of association exists between this ratio and fruit weight, the weight-area ratio can be estimated from the regression lines if the weight of the fruit is known. Graphs were made for the force required to remove the fruit frou.the tree versus the weight of the fruit for all the samples. The points were highly scattered and did not readily suggest any relationship. Correlation analyses were made for .the data of two samples (peaches and apples) which were chosen to represent the samples with the least and most scatter of Points. These analyses and the graphs (Figures 11 and 12) are presented in Appendix II. As an approximation for a line 26 through the scatter of points, a linear relationship was assumed. The analyses gave negative correlation coefficients for each sample with one correlation coefficient at the border- line of being significant at the five and one percent levels. The other correlation coefficient was nearly zero, indicating that very little relationship existed. To eliminate any error involved in the method of obtain- ing these forces, the pulling force data were grouped and graphs constructed for groups versus fruit weights. These graphs did not present any significant change from the pre- vious graphs. It was, therefore, concluded that the results of the correlation analyses for these data were valid and there was no linear relationship between the force to remove the fruit from the tree and the weight of the fruit. It was also concluded that fruit maturity was probably the most predominant factor in causing the variation in pull- ing forces since fruit of various maturities were selected. The force required to puncture the fruit's flesh, obtained by a pressure tester, was presented in Table II for apples and peaches. Fruit growers usually employ this tester as an aid in determining the maturity of fruit to allow them to select the proper time for picking. This tester was used in a study conducted on bruising of McIntosh apples in a packing house by Burt (1959) to determine the firmness at which apples could be handled mechanically. The evidence found indicated a firm- ness index around ten or eleven pounds below which apples 27 could not be safely handled mechanically on a packing line. They found that firmness of the fruit was probably the most important single factor in determining the amount and severity of bruising incurring to an apple in the packing operation. The firmness data given in Table II was obtained only to establish typical values of the relative firmness between different varieties and species of fruit. No attempt was made to use this data to determine the degree of handling permis- sible in harvesting. Smock and Neubert (1950) stated that there are numerous limitations of the pressure tester. These limitations should be considered before trying to evaluate any pressure test data. Some of the factors are: (l) The firmness of a given variety varies from season to season. (2) The firmness of a given variety varies from one location to another. (3) The pressure tester usually gives a higher reading on the blushed side of the fruit than on the unblushed side. (4) hature well-colored apples on the outside of the tree may have a higher pressure test than less mature apples on the inside of the tree. (5) Soil fertilization with nitrogen fertilizer applica- tion may affect the firmness. (6) Temperature of the fruit and its moisture content have an effect on the pressure test reading. (7) Fruit size has an influence on firmness. Large fruits are usually but not always softer than small fruits. The primary usefulness of the test is to tell the_dif- ference in firmness between two or more lots of the same variety on a given date or to tell the general degree of ripeness. THEORETICAL ANALYSIS To aid in designing laboratory and field apparatus for conducting tests on various species of fruit, an analysis of a particle in an air stream was made. The principal items of investigation were: air velocities needed for flotation of fruit, fan size and output, and horsepower requirements. A particle in free fall will reach a steadyestate veloc- ity that depends upon the physical characteristics of the particle, the fluid in which it is falling, and the accelera- tional force. The steady-state velocity (terminal velocity) is also the air or liquid velocity required to suspend or float a particle. The net force acting on a particle in a given direction (in this case vertical) is the sum of the frictional force and the external force. The following analytical procedure is adapted from a treatment of this particle characteristic by Lapple and Shepherd (1940)- For a particle falling in a vertical air stream, THHF 30 the forces involved are m'dVE .W-F (1) where: W 2 particle weight, pounds F = frictional drag force, pounds but considering the buoyant force of the air gives W = KPH/é - 31/1,; and by definition, the drag force is F - c v22” is therefore equation (1) becomes :11 dV -NP(6P-?f)- c v2sA 2‘s 37:2 OF dV = g/U-U>- c v22”. (2) p . 4—— at“ \ KP 2 n where: V 2 relative velocity (Va-t Vb), feet per second V = velocity of the particle, feet per second Va: velocity of the air, feet per second t = time, seconds w : particle weight, pounds 6’: fluid specific weight, pounds per cubic foot '35: specific weight of particle, pounds per cubic foot 0 = particle aerodynamic drag coefficient, dimensionless 31 £2 H volume of particle, cubic feet A : projected area of particle, square feet g = acceleration due to gravity, 32.2 feet per second per second m. mass of the particle This equation must be solved by a method of approtha- tions since it cannot be solved explicitly. Solving this equation by the method of approximations will not be dealt with here as this analysis pertains to finding a steady state velocity. For steady state conditions de/dt is zero and equation (2) can be solved for terminal velocity giving v- ngé—K' t004w> m3 700 .- . 44 _ nvlv Al 80C) GCK) ENGINE DRIVE SHAFT SPEED, 50C) CC) I2/5/60 H.Ell RPM Relation of engine drive speed to air velocity in duct. Figure 15. Figure 16. Cantilever beam with strain gages mounted ibove the duct outlet. Figure 17. V-shaped trough above the duct outlet shielding the cantilever beam from the air stream. 47 threaded through a small opening in the trough, through a hole in the beam and secured by placing a knot in the end. The other end extended down into the duct for attachment of the fruit. Figure 18 shows the device used for attaching apples. The needle formed from a nail with an enlarged head was inserted through the apple from blossom end to stem end. Figure 19 shows an apple attached to the nylon line with this device. For those fruit with a pit in the center, a sewing needle was used to insert a short length of nylon line through the fruit on either side of the pit and equip it with a small disk on the bottom side. The disk prevented the line from PHIling into the flesh of the fruit. Tests were conducted by pulling the line with the fruit attached upward with the line placed inside the small plastic tube mounted above the beam. The small tube served to hold the line in a vertical position above the beams When the fruit was released from different heights marked on the small plastic tube and allowed to fall, it was brought to a stop by contact of the knot in the line with the beam. The beam was manually given an initial deflection before releasing the line to allow the fruit to fall. The purpose of this deflection was to indicate on the recorder an initial point from which the time required for the fruit to fall could be measured. Figure 20 shows a section of oscillograph tape with recorded results of a typical drop. The time required for the fruit to fall from.the raised position to the position #8 Figure 18. Device used for attaching apples to the cantilever beam. Figure 19. An apple hanging inside the plastic duct from the cantilever beam. “9 a mo .QOhv Havana» muasmma cocaooou and: was» sasnwoaaaomo no Loaaowm .om magmas l.’ c. I i - . i M - . , l a \ n N l. hwwo Mammfi A WWZM—Qwvwhmiu Em I may we... w E / i . 3%.: w j. . 3E .: S. ._ W k. w "Hf; h M. w w MRHW a w w W w... W- m .kaw ww «W W - . -i- M m M M w W a -m MW\W . ,- - .m ..w M h . . .. c ”.1“ a - M m w . a. 50 occupied when fully supported by the beam was computed from the oscillograph tape. The oscillograph was operated at a speed of five inches per second which is one of the three standard speeds of an oscillograph. This speed was checked for accuracy with a stop watch. The time required for the fruit to fall was measured from the release point, shown in Figure 20, to the impact point. The magnitude of the impact force was determined by the lines of deflection from the reference line to the top of the deflection curve. A hard- rubber damper was mounted with slight contact to the beam at the free end. This damper eliminated beam vibration between the release point and impact point. Experiments were performed with different species of fruit by drOpping them.from one-half, one, two, three, four, and five foot levels. Each fruit was dropped from these heights in air velocities ranging from zero to 6,500 feet per minute. Results The results obtained are presented on the graphs in Figures 21 through2ha. The graphs showing coordinate points were plotted directly from experimental data. The graphs showing curves with no coordinate points were obtained from the graphs of the experimental data. . The theoretical terminal velocities of the test fruit were computed by equation (A) and are shown in Table V. 51 LBS. \NEKfliT '4“ AVERAGE DIAMETER: 2.90 IN. ' >0 FI/ MIN. .432 L83. — Mm-..‘ ——-..—o~-_ '4000 Figure 21. HEIGHT OF DROP, FT. 14.2.56” Relation of impact force to height of dr0p for a McIntosh apple. FT. EQUIVALENT DROP IN STILL AIR, 52 0| A __P_‘.._4 3 2 _.J I _ e___ O O m/sxeo ACTUAL DROP IN AIR STREAM, FT. H.E.Q. Relation between the actual drop in an air stream and an equivalent drop in still air for a McIntosh apple. Figure 21a. FORCE . LBS. I M PACT 53 3 4, "WV—Ia _ I , _ AVERAGE DIAMETER = L57 IN. IC) FT/KMN. WEIGHT = .084 LBS. 1/ l2/ HEIGHT OF DROP, FT. H.543, Figure 22. Relation of impact force to height of drop for a Montgamet apricot. DROP IN STILL AIR, FT. EQUWALENT 54 03 U" A 01 ‘1 4000 a: 4500 -4 5000 “ 5500 O | ACTUAL Figure 22a. 2 s .4 5 2 5 60 DROP IN AIR STREAM, FT. L’s/Q Relation between the actual drop in an air stream and an equivalent drop in still air for a Montgamet apricot. LBS. 55 '6 I I I I '4 WEIGHT = .350 L88. AVERAGE T DIAMETER = 2.53 IN. I I2 IL_--____.I-___.. ___ IT“ --__.L ___ 1.0 FT/MII! I / .4000 l 5000 5500 '6000 ..-__in___. t. _-_“i _i I o x. E—l6500 I__..___+ ._ 4 5 I2/5/60 HEIGHT OF DROR FT. H.E.Q. Figure 23. Relation of impact force to height of drop for a Red Haven peach. 56 FT. FT./ MIN; IN STILL AIR, EQUIVALENT DROP 2 3 4 5 I2/5/60 ACTUAL DROP IN AIR STREAM, FT. H. E. 0. Figure 23a. Relation between the actual drop in an air stream and an equivalent drop in still air for a Red Haven peach. LBS. FORCE, IM PACT 57 7 I I I 6 WEIGHT = .092 L83. AVERAGE DUMWETTH? = |f70 IN. I 0 FT/MIN; 5 I “ . I / I I 4.- _I--... ’ --_.-__ /, 3500 5 — / I /+ p 4000 II I + I 2 ~—+-— -_._I_.__- Ame _._ 4.» I I I I - 4500 I _ i ‘ _ - . _ _. _. .L....____.....___ T "'5000 r I x O k 4 5 O I 2 3 I2/5/50 HEIGHT OF DROP: FT. H. E_Q_ ‘ f F ure 24. Relation of impact force to height 0 ig drop for a Stanley Prune plum. FT. IN STILL AIR, EQUIVALENT DROP 58 0 C) I 2 3 4 5 '2l'5/ 60 H.EL O. ACTUAL DROP IN AIR STREAM, FT. Figure 2ha. Relation between the actual drop in an air stream and an equivalent drop in still air for a Stanley Prune plum. 59 TABLE‘V Theoretical terminal velocities computed with the results from samples of 50 fruits. Fruit Weight Ratio of Weight Terminal to Projected Velocity lb. Area* lb./'in.2 ft./min. APPLE McIntosh I 0. 432 0. 0662 8180 APRICOT Montgamet 0.08h 0.0432 6610 PEACH Red Haven 0. 350 0. 0624 7940 PLUM Stanley Prune 0.092 0.0flll 6520 and 10. *Values obtained from.Figures l, 2, 7, Air conditions of 70 degrees Fahrenheit at one atmosphere (0 : 0.075 pounds per cubic foot) were used in computing the terminal velocities in Table V. Table VI has been prepared to concentrate the informa- tion provided by the graphs of Figures 21 through 24a. Only five foot drops have been considered since these represent the extreme case for this investigation. The equivalent drop values in Table VI are the distances which fruit must be dropped in still air to produce an impact force equivalent to that obtained when the same fruit is dropped in air streams of the velocities shown. 6O .udonv poou o>du :0 comes mamas 02.0 ba.o mm.m om.H om.: m essam Saddam. spam oe.m mm.o ow.s mm.o 0:.w mm.a om.» ms.H o>.w mm.m o:.ma m co>am com mo.m m poaawaeo: aooumma. os.m mm.o oo.m No.0 oo.m m:.a om.HH om.m om.ma m snowman: mumm¢ .nH .vu .nH .pu .na .uu .nH .pu .nH .9“ .AH .uw venom moan eoaom dean coach moan eosom moan eonom moan oonom moan weedsH.>Hsom;mosnaH.basumgmosdaH .bHsumposasHI.>fisom powdsH.ammwmmvosaaH.>«scm ommIqm oooIm 8m 6 oooImI _ 80.: o I 33m .fis\.3 33303, s? snoopou needed undononuoaaoo cos soocsuuao nous unedsaasam Hb mundm no woman mamas ensam heficsum I I I I am mm 33a mm Hm ms we mm mm no>sm com _ scams .II II. hm mm mm mm easemenoz m... . 982.2 mm we. I I we. ms . £323: . mqaa< Hasmmwo .noomdm Haammwo noomrm Hasmumo seamen 23 on :33 :35 poems...“ I Smash p39: oom.m - o oom.m I o . ooo.m I 0 sense newsasoonom enp.mmfiusmmoo ca coup 6.:H£\wauv acquaoofleb h«<. seesaw on» nosed op has mean: can: done no npnwaon unodsbaaoo can «vouch assess no coaposoom HH> mqm / g 4 \\ L” 2 /\\ WM 111 ' 5 O Q: (I) L 4 1 L 4 1 4 g 4 l 1 O 5 IO I5 20 25 30 35 4O 45 5O 55 PERCENT OF TOTAL gig/g? Figure 37. Percent of total apples (50) receiving bruises when dropped manually from various heights onto the machine with the fans not operating. 84 was present, 21.5 percent more fruit were contained in the "cull" category. The drop-test fruit (Figure 37) shows an increasing percentage of fruit occurring in bruise evaluations of 12 down to 2 while for the harvested fruit (Figure 36) the opposite was true. This fact indicates that many of the apples harvested with the machine received an impact force of sufficient magni- tude to only bruise the apples and not break the skin. The drop-test fruit received the entire impact force and as a result caused more skin breaks. This comparison indicates that the air from the machine was capable of reducing the number of "cull" fruit, however, the remaining fruit contained considerable bruising and was not acceptable. ‘ Comparison of the results given in Figures 36 and 37 is not completely Justified since both samples of fruit were not harvested on the same date. The machine harvested fruit were harvested October 1“ and the drop-test fruit were harvested October 19. Because of these different harvest dates, the drop-test fruit were more mature and softer, thus making them more susceptible to bruising. The percentages given in Figure 37, for fruit in each bruise category should be decreased to Justify a comparison between Figures 36 and 37. The changes would probably be mmall and not affect the original compari- son appreciably; therefore, this comparison is representative of the capabilities of this machine. COST ANALSBIS The economics of a pneumatic type fruit harvester were analyzed by a procedure given by Bainer, Kepner and Berger (1955) and are presented in this section. They stated that the total cost of performing a field operation includes charges for the machine, for the power utilized, and for labor. These costs are grouped under the headings of overhead costs and operating costs. Overhead costs include depreciation, inter- est on investment, taxes, insurance, and shelter. Operating costs include repairs, maintenance, lubrication, fuel and oil, and labor. Numerous assumptions were necessary for this analysis. They were based upon material given by Bainer, Kepner and Berger and material pertaining to semi-dwarf apples. The assumptions made are as follows: (1) A pneumatic fruit harvester would have a service life of five years. (2) There would be no salvage value. (3) An acre contains 86 trees (semi-dwarf type). (4) Each tree would require three minutes to harvest. (5) The harvest season would be of 35 days duration. (6) The harvester would operate ten hours per day. (7) Fuel consumption would be 8.5 horsepower-hour per gallon. (8) The horsepower requirement would be 1,000 and would 86 be utilized one-half of the total time. (9) Field efficiency would be 60 percent. (10) The initial cost of a machine would be $15,000. (11) Interest on investment would amount to six percent. (12) Fuel would cost $0.20 per gallon. (13) Repairs, maintenance and lubrication would amount to 3.4 percent of the initial cost. . (lb) Taxes insurance and shelter would amount to 1.5 per- cent of the initial cost. (15) The cost of oil would be three precent of the fuel cost. (16) Labor would cost $2.00 per hour and three men would be employed. (17) Hand harvest would cost 14.2 cents per 40 pounds of apples. (18)/k/acre of semi-dwarf apple trees would yield 1,000 bushels of apples. (19) A bushel of apples weigh 50 pounds. (20) The price received per bushel of apples would be $1.50. The following values were computed with data from the assumptions: Yield per tree - 11.6 bushels Field capacity = 0.1h0 acres per hour Annual use = 350 hours Total acreage per year = “9 87 The annual overhead charges are: Depreciation (15,000 - 0) 8 $3,000.00 Interest 0.06 (15,000+0) = 450.00 Taxes, insurance d shelter 0.015 (15,000) = 225.00 Total annual overhead $3,675.00 charge The costs per acre are: Overhead 3,6;3.00 : $ 75.00 Repairs, maintenance and lubrication 0.034§l§,000) = 10.04 Fuel, 8.5 hp-hr/gal. 2 84.00 Labor, 3 men 8 $2.00/hour = 42.80 011 cost (0.03) (84.00) : 2. 2 $214.36 Total cost per acre Thus the mechanical harvesting costs are given by this analysis as $214.36 per acre. The annual cost for hand harvesting would be equal to to. .142X 0 1b. %trees 11.6 bu. Eu. acre x tree 3 $177.00 per acre The value received by the grower for his apple crop would be equal to 8§_trees 11.6 bu. 1. 0 acre ‘ tree 1 u. t $1,500.00 per acre 88 The harvest costs expressed as a percent of the total value received per acre are given as follows: (1) Mechanical harvesting: 56 = 214. 6 (100) = 14.3 when three men are utilized. 9 (2) Hand harvesting: gs : 1;; (100) = 11.8 Comparison of harvesting costs for the conditions assumed indicate that harvesting apples with a pneumatic harvester would cost 2.5 percent more than hand harvesting. Even though the results of this analysis do not favor pneumatic fruit har- vesting, the cost analysis has disclosed that a pneumatic fruit harvester may be economically feasible. With a few improve- ments, mainly in overhead and fuel consumption, mechanical harvesting costs could possibly be reduced below hand har- vesting costs to make pneumatic fruit harvesting practical for commercial use. cowcmslows The conclusions derived from this study may be stated as follows: (1) An air column moving at a velocity very nearly equal (2) to the terminal velocity of the fruit is necessary for pneumatic harvesting to be effective. It was found that an air velocity of 6,500 feet per minute reduced the impact force received from a drop of five feet to a value equivalent to that obtained from a drop in still air of 4.2-inches for an apple, and three inches for a peach. An air velocity of 5,500 feet per minute gave an impact force for a drop of five feet equivalent to a drop of 1.8-inches in still air for an apricot. An air velocity of 5,000 feet per minute gave an impact force for a drop of five feet equivalent to a drop of 2.04Qinches in still air for a plum. These velocities ranged from 750 to 1,580 feet per minute below the theoretical terminal velocities of these fruits. A greater air velocity is needed to float heavier fruit than is needed to float lighter fruit of the same species because the ratio (H/l) increases for heavier fruit (Equation 4). 9O (3) Fans capable of supplying air at a velocity greater (4) (5) than the velocity needed to effectively lower the fruit are necessary to overcome head losses. A survey of fan companies indicated that fans of this size are not readily available and would prob- ably require special design. Horsepower requirements for fans capable of provid- ing the terminal velocities discussed in the manu- script were found to be large. This means that the initial cost of a machine of this type, covering the complete tree, would be large; therefore, high efficiency would be a necessary requirement. A theoretical cost analysis indicated that a pneumatic fruit harvester for apples has possibil- ities of being economically Justified on a cost per acre basis. RECOMMENDATIONS FOR FUTURE STUDY Further study would be advisable to determine if fans other than propeller fans could be used for harvesting fruit with air. An investigation, made with the Magness-Taylor pressure tester to determine what value it may have toward han- dling techniques used in harvesting fruit, is suggested. It is suggested that a study be conducted to determine fruit velocities when falling in an air stream and relate these by an equation to impact forces. With this infor- 'mation, an exact air velocity could be computed which would lower various species and varieties of fruit effectively. The conclusions from-this study indicated that it would be desirable to decrease the air velocities needed to float various species of fruit. Based on this fact, the recommendation is mmde that the possibility of coating the fruit with a lightweight material should be investi- gated. Bince terminal velocity is dependent upon the square root of the ratio of fruit weight to projected area (Equation 4), it is evident that if this ratio can be decreased the terminal velocity will decrease. The material should have a low density and add considerable bulk to the fruit to provide a large projected area. It 92 should be non-toxic, easy to apply and remove and possess enough rigidity to withstand the required air velocity. A material of this nature would also greatly reduce chances of damage occurring to fruit from impact forces while fall- ing. A foam material, produced by Dow Chemical Company, used to spray on plants for frost protection was investi- gated by the author for use on fruit. A letter from Dow Chemical indicated that this material was undesirable. The letter also stated that most of their products meet- ing these specifications would be toxic to the fruit and, therefore, could not offer a product of this nature. Another recommendation involves a method of harvesting fruit with flexible fingers. This method would employ a multiplicity of slender rods mounted very closely together on a single frame. The rods would be of sufficient length to protrude into the branch area to the center of the tree. With the fruit positioned among the rods, the tree would be shaken to loosen the fruit and allow it to fall to the rods to be caught from only a two or three inch fall. A slight tilt to the rods and possibly some vibration would cause the fruit to slide toward the rod mountings where they could be deposited in a storage container. SUMMARY Samples of 50 fruits each were obtained for various species and varieties of tree fruits. Basic data for dis- meters, weights, firmness snd forces to remove the fruit from the tree were recorded. An air duct was constructed and used in the laboratory to determine the effect of a high velocity air stream in reducing impact forces received to fruit from a fall. A theoretical analysis was made for a particle in an air stream. ~The analysis provided informationregarding the ter- minal velocities of the fruit and horsepower requirements for the required volume flow rate of air. Using the information obtained from the laboratory tests and theoretical analysis, a field machine was constructed and tested on a Jonathan apple tree. A correlation and regression analysis of data obtained from the various species and varieties of fruit indicated a high degree of association between fruit weight and the ratio of fruit weight to projected area. A correlation analysis, conducted for Red Haven peaches and.flclntosh apples, for the fruit weight and force required to remove the fruit from the tree indicated a definite association did not exist between these variables. Also, graphs of fruit weight versus force to remove the fruit from the tree for each species gave very 94 little indication of any relationship existing. It was con- cluded frost the scatter of points on the graphs that other factors were present and probably the most predominant one was maturity of the fruit. ' The theoretical analysis indicated that horsepower require- ments to move air at a velocity of 7,000 feet per minute through a lS-foot diameter duct would be 992 if the efficiency of the fan was 60 percent. Additional horsepower would be required to overcome head loss from the duct system.and turbulence. The reduction in velocity due to turbulence and harvester design for the field machine constructed in this study was 27.2 percent. The laboratory tests with various species of fruit indi- cated that air velocities very nearly equal to the terminal velocities of fruit are necessary to appreciably reduce impact forces occurring to falling fruit. Terminal velocities for large fruit such as apples and peaches were calculated to be in the range of 7,000 to 8,000 feet per minute while for snaller fruit the range was from 3,h00 to 7,000 feet per minute. These velocities are theoretical and were based on assumptions. Tests. conducted with the field machine indicated that ' an air velocity of 3,020 feet per minute was not effective for Jonathan apples. The number of apples with skin breaks was reduced considerably as compared with still air conditions on the machine, however, a.naJority of the apples received large bruises that were undesirable. REFERENCES Adrian, P. A. (1958). Annual Report. 11. s. D. A. ( ) (1959-1960). Annual Report. 0. s. D. A. Adrim, P. A. and a. s. Pridley (1958). wechenicel fruit tree shaking. California Agriculture. 12(10):3 and 15. October, 1958. Adrian, P. A., R. s. Fridley and c. R. Kaupke (1959). Low profile catching equipment for tree fruit. A. S. A. E. Paper lo. 59-601 presented at Winter meeting, 1959. Anonymous, (1959). ’Storage of apples. Importance of harvest- ing maturity. Agricultural Gazette of new South Wales. 70:237-hl. May, 1959. Anonymous, (1959). Mechanizing small tree-fruit harvest. Agriculture Research. 7( ):10. February, 1959. Anonymous (1960). Picking spindle. American Fruit Grower. 80(33 :30. larch, 1960. Bainer, R., R. A. Kepner and E. L. Berger (1955). Princi les of Farm Machinery, John Wiley a Sons, Inc., New YbrE. 5'71 pp. .Burt, S. H. (1959). An experimental packing line for McIntosh apples. Agr. Mktg. Serv., Hktg. Res. Div. August, 1959. Coppock, G. E. and P. J. Jutras (1959). PrOgress in mecha- nizing citrus fruit harvesting in Florida. A. S- A. E. Paper No. 59-118. Presented at 1959 annual meeting of Ame Soc. of Ag. Engr., Ithaca, New Ybrk, June 21-2h,1959. Coppock, 0. E. and P. J. Jutras (1960). Spindle type citrus harveggar. Letter received by.B. A. Stout dated August 19, l . Crane, J. H. (1956). Pressure drop due to the pneumatic con- veyance of grains and forages. Thesis for degree of H.S., lich. State Uhiv., East Lansing. (unpublished). Della Valle, J. a. (1948). wioroueritice. 2nd ed. Pitman Publishing Corporation, New YorE. 555 DP. Davis, R. P. (1935). The conveyance of solid particles by fluid suspension. Engineering. lk0:l-3. July 5, 1935. 96 ( ) (1935). ‘The conveyance of solid particles by fluid suspension. Engineering. lhO:l2N-l25. August 2, 1935. Duncan, A. J. (1955). alit Control and Industrial Statis- tics. Richard D. rv ,- omewoSd,'IIlinoi§. 663 pp. Fridley, P. s. and P. A. Adrian (1959). Develo nt of fruit and nut harvester. A. S. A. E. Journal. 0(7):386-7, 391. July, 1959. Garrett, H. E. (1956). ElementEgz Statistics. Longmans, Green and Company, ew or , oronto. ‘16? pp. Gaston, H. P. and J. H. Levin (1951 . How to reduce apple bruising. Special Bulletin 37 , Mich. State College. September, 1951. ( ) (1953). Time and motion studies of apple picking mad? to determine the possibilities of mechanizing har- vest. Mich. Agr. Exp. Stat Quar. Bul. 36-h(l):18-23. August. 1953. ( ) (1955). Experiences with box pallets in harvest- ing and handling apples. Mich. State Horticultural Society, Annual Report. pp. 9h-97. Gaston H. P., J. H. Levin and s. L. Hedden (1958). Mechanized harvesting of Michigan grown fruits. Proceedings, Mich. State Horticultural Society. p. 19. December, 1958. Gaston, H. P. and J. H. Levin (1959). Handling apples in bulk boxes. East Lansing, Mich. Agr. Exp. Sta. Special Bulletin #09. Gaston, H. P., J. H. Levin and S. L. Hedden (1960). .Blue- berries, plums, and cherries can be harvested with maghines. American Fruit Grower. 80(h):14-15. April, 19 0. Gaston, H. P., s. L. Hedden and J. H. Levin (1960). Mecha- nizing the harvest of plums. Michigan Quarterly Bulletin. fl2(#):779. May, 1960. Hedden, s. L., H. P. Gaston and J. H. Levin (1959). Harvest- ing blueberries mechanically. Michigan Quarterly Bul- letin. “2(2). August, 1959. Henderson, 8. M., and Re L. Perry (1955). Agfiicultural Pro- cess Engineering. John Wiley and Sons, ew or . 352 pp. Hill, F. L. and a. w.:srazelton (1955). The steel squirrel. Agriculture Engineering, St. Joseph, Michigan, 36:17-19. 97 Jutras, P. J. and G. E.Coppock (1958). Mechanization of citrus fruit picking. Proceedings from Florida State Horticultural Society. 71:201-h. Lamouria, L. H., R. w. Harris and others (1957). Harvester {gr canning fruit. California Agriculture. ll(8):ll-l2, Lapple, C. E. and C. B. Shepherd (l9h0). Calculation of parti- cle tra eotories. Industrial and higr. Chem. 32 :605-616. "J: 19 0. Levin, J. H. and H. P. Gaston (1952). How to reduce losses caused by apple bruising. Extension Folder P-l72. June, 1952. Levin, J. H. (1958). Summarg of apple harvest research. U. S. D. A. November, 19 . Levin, J. H., S. L. Hedden and H. P. Gaston (1959). Experiments in harvesting cherries mechanically. Mich. Agr. Exp. Sta. Quar. Bul. 1:805-11. Levin, J. M. H. P. Gaston, S. L. Bedden and R. T. Whitten- -berger (1960). Mechanizing the harvest of red tart cher- ries. Reprint from.Mich. Quarterly Bulletin, Art. h2-60. haw). May, 1960. Lorang G. (1960). Grow apples like a row-crop; dwarf trees. Farm Journal. 83:46—7. April, 1960. Madison, R. D. (Ed) (1949). Fan Engineering. 5th. ed., Buffalo ‘Forge Company, Buffalo, New Ydrk. McBirne S. W. 1958). Use of pallet bins for the 1958 apple halvest in Sashington. Reprint from Wash. State Hort. Assoc. Proceedings. 54:153-155- ( 1 . Mew developments of fruit harvesting and .handligggig)the Pacific Northwest. A. S. A. E. Paper No. 59-139. Presented at 1959 annual meeting of Am. Soc. Agr. Engr. Ithaca, New York. McBirney, s. w. and A. Van Doren (1959). Pallet bins for har- vesting and handling apples. Washington Agr. Experiment Sta. Station Circular 355. April, 1959- Ower, E. (1927). The Measurement gg_Air Flow. Chapman & Hall, Limited, London. pp. p - 98 Schomer, H. A. (1957). Bruising of apples: Where does it occur and how can it be minimized. Reprint from Wash. State Hort. Assoc. Proceedings. 53:129-131. Smock, R. w. and A. M. Heubert (1950). A les and A le Products. Vbl. 11. Interscience Pu s ers, Igc., New YorE and London. 486 pp. Tidbury, G. E. (1958). The bulk harvesting of orchard fruit. Commonwealth Agricultural Bureaux, Digest Ho. 1, Parmham Royal, Bucks, England. 38 pp. 0. s. D. A. (1959). Michigan Agricultural Statistics. Agr. Mktg. Service, Mich. Dept. of Agr. July, 1959. ( ) (1960). Michigan Agricultural Statistics. Agr. g. Service, Mich. Dept. of Agr. July, 1960. Walker, H. M. and J. Lev (1953). Statistical Inference. Henry Holt and Company, New YorE. SIG pp. Whittenberger, R. T. and C. H. Hills (1958). Cherry harvest- ing made easier. Agriculture Research, Washington. 7(1):7. APPENDICES APPENDIX I Figures 3 through 10 are presented in this section. The reader is cautioned that the regression lines presented on the following graphs have not been computed by established methods, but have been estimated. ‘ LBS. / 80. IN. 1 RATIO OF WEIGHT T0 PROJECTED AREA 101 .070rs 068’ 06th— .064*” (362*~‘— I (360 ' Figure 3. .60 .70 POUNDS LIZ/3’89 .50 .400 WEIGHT, .30 Relation between weight and projected area for Northern Spy apples. LBS. / SQ. IN. RATIO OF WEIGHT TO PROJECTED AREA, 102 .OZCI - ; .CHB (M6 .Ol4 .CH2 .CHO I I I I I L .40 .60 .80 LOO [20 WEIGHT, GRAMS 3’38? Figure A. Relation between weight and projected area for Jersey blueberries. LBS. / SQ. IN. RATIO OF WEIGHT TO PROJECTED AREA, 103 1376» 1372 1368 LN54- — .060 - - .056 ,- , t . I JLVI I I I I I i I I l .30 .35 .40 .45 2 0500 WEIGHT, POUNDS E’g’ s. Figure 5. Relation between weight and projected area for Elberta peaches. NV. LBS./ SQ. RATIO OF WEIGHT TO PROJECTED AREA 104 .022 ——-—i JDZI .02C> .CHQ .OWB» .OI7 / ‘ I 1 I i I l I I I 315 413 ‘15 SE) 535 WEIGHT, GRAMS 3’38? Figure 6. Relation between weight and projected area for Montmorency cherries. 105 -&)44 1342* LBS. / SO. 1340' .038 ’- (136' 1134 RATIO OF WEIGHT TO PROJECTED AREA, 032* / . a 1 I I l I 1 l I 20 25 30 35 'mlegs WEIGHT. GRAMS H. 5.0. Figure 7. Relation between weight and projected area for Montgamet apricots. IN. LBS. / SQ. RATIO OF WEIGHT TO PROJECTED AREA, 106 .062 .060 .058 . .056 .054 .052 .050 1le .20 Figure 8. .25 .30 .35 .40 0 WEIGHT, POUNDS 3’33 Relation between weight and projected area for Jonathan apples. IN. LBS. / SQ. RATIO OF WEIGHT TO PROJECTED AREA, 107 .072 _070 . iDGGw .064 .062 136C) .058 > 1356I~——~ °+~ I I I ~ ; Figure 9. .25 .30 35 .40 .45 .50 . 0 WEIGHT, POUNDS 3’33. Relation between weight and projected area for Cortland apples. IN. LBS. / SQ. RATIO OF WEIGHT TO PROJECTED AREA, . 070 1368 1066* 1364-* 1362 ~ .060 1358 .056 - 108 I) v 0 I I I | I I ' . . . I . . ‘17, .25 .30 .35 .40 .45 .50 WEIGHT, POUNDS glee/g? Figure 10. Relation between weight and projected area for McIntosh apples. APPENDIX II The regression and correlation analyses for Figures 1 and 2 are presented as follows: Let: y - ratio of weight per projected area x = weight of the fruit. The regression analysis using the symbols above con- sists of finding the regression of "y on x." The following equations are solved simultaneously to obtain the regression line. Zy - Na+b Ex (11) ny = a Zx+b 2x2 (12) where: a = the intercept 0n y—axis b : slope of the regression line H = number in sample The correlation coefficient is given by the formula r = P (13) W '/ < r < / where: P = Zx - (2x) (2 ) 'III'z T 1:1 0:? = 2x2 - ‘3'an Standard deviation in N ( N x-direction (75' = 2x2 - §1 '4 Standard deviation in J// N (I > ’ y-direction The following information was computed from the original data which is on file with Dr. B. A. Stout. 110 For plums (Figure l): 21: 8 1761.59 Zy = 1.9602 ny 3 69.5217 :ixz = 63.57h.08 E y? - 0.07701m H = 50 For peaches (Figure 2): Ex : 19.928 Zy = 3.2568 ny = 1.3263 5x2 = 8.4721 Zyz = 0.2136u7 N 3 50 The couputations for plums (Figure l) are: WW 0‘“va “a: (Zr-X11) @2511 (flexes) r: P ’ arrsgi'gggfi‘nw'm" Uin - 9 Substituting values into equations (11) and (12) gives 1.9602 : 50 a + 1761.59 b (1n) 69.5217 = 1761.59 a+ 63,571I.08 b (15) 111 Multiplying equation (14) by 1,761.59 and equation (15) by minus 50, then adding these equations algebraically gives “23000 : -75’505e67 b b : 23.00 = 0.000304 9 e Substituting the value of (b) into equation (in) gives 1.4241 0.0285 1.9602 : 50 a +-176l.59 (0.0003ou) ETD 50 a : 1.9602 -o.5355 I ' I The equation of the regression line for Figure l is y = 0.0285-+ 0.000301 (x) To establish the limits of this equation, the standard error of estimate (So) was calculated as follows: Se: ‘8 " EX Substituting values into this equation gives 8 : . - . . - . . ) ’ 50 - 2 0.000954 The final equation of the regression line is given as follows: 9 - 0.0285-+ 0.000304 (X):t 0-00095“ (15) where: A y 3 an estimate of y The computations for peaches (Figure 2) are: 2 0.1029 112 “Es-(596814611 (assess—8) = 0.000566 r = P s 0.000366 = 0.995** a} a} . . Substituting values into equations 11 and 12 gives 3.2568 = 50 a-+ 19.928 b (17) 1.3263 = 19.928 a-+ 8.4721 b . (18) Multiplying equation (17) by 19.928 and equation (18) by minus 50, then adding these equations algebraically gives - 1.4135 : - 26.4798 b b = 1.4133 : 0.0534 Substituting the value of (b) into equation (17) gives 3.2568 = 50 s +-19.928 (0.053u) 50 a = 3.2558 - 1.0638 a = 2.1930 s : 0.0438 The equation of the regression line for Figure 2 is y s 0.0u384—0.053h (x) To establish the limits of this equation, the standard error“ of estimate (Se) was calculated as follows: Substituting values into this equation gives 113 0.00205 The final equation of the regression line is given as follows: 9 = 0.0438+ 0.0534 (x) 1: 0.00205 where: A y an estimate of y (19) Figures 11 and 12 and a correlation analyses for these APPENDIX III data are presented in this section. The regression of "force to remove fruit from the tree on weight of fruit" was investigated by letting y = force to remove the fruit from the tree and x = weight of the fruit The following values were obtained from the data collected on these samples. For McIntosh apples (Figure 11) 23x = 18.539 Zy = 224.0 ifxyjs 81.169 2x2 : 7.1121 Eye = 1,118.50 For Red Haven peaches (Figure 12) Zx : 19.928 Zy = 301.50 ny -.- 119.59: 2:2 = 8.4721 25y? = 2,000.25 For Figure 11: The'standard deviation in the x-direction is” 0'5: 0.069 X 115 mQZDOQ e i . c . i . . i _ .. e _ 0 . 0 e a i _ _ . P O 6 O f P _ . . I I I . I . t aIIII e..IIIII III ETIIII I I,II- . 5 / . 8 II . _ . n u e _ 5 E h e i _ _ m . _ . / . t m . . _ M .1 m H d , _ F _ __ n m _ i _ I - w_. - - I I- 5 a m I II+ l “IIII I -HI I|III 4. tf . , . _ h a . 1 ab 8 _ e i . __ 1 e . _. M _ e 1 . . e e W p . . a _ A m e w. H s, x w 09 I I . t I .._ I . S h - . e . _ e . D t .fl _ _ e .. . _ w n o _ . e t _ _ 0 e n . /W I eI.I.+I I I IN II IIIII “III I IIIII. 111111111111111 {In /m I u Seamed , 4 _ i .503 23 9.31031 A . . e . 5m} 00 Fenced one ”£03800 00515.» 05 3 vouofico .. ./ m _ . e e .8030 NEH. is; 03m .003A 00 308.3033 £00086; .3 .. e E 3.33 boomed no .023 pa «0 808 05 GMB 00530.. I -III I . . a -II I II I E 09.8% 00 32 anemone .2805 p088 been 0:» 0.53: 801 II- I - III. II I e w e _ 6032 soon 93 vogue 80A «.3 883800 his 4033 89¢ “.03 _ , .e a 05 £023 .035 .6: 05 am pone—same 330:0 . .. 600 .060 was 83803 . 05 .0 octwuovmeaoo an _ + I I I H. II .. I... mUZdJHDwZH $.52 _ .1 w . i . _ _ “I E F i p . 8 7 6 5 4 e3 . . 1111111111 v .Eomoa v.8... mama—€00.30 0062: 08 5.5 so heeded 05 $5.138 093.03 05 E mono—060 0n 0. 0.8 *‘ a .83.. sees as 883. 83 as scents as 823 see ASH v 05 £0.23 some: .Gzon .I.I. . v v -5300 “EH. 4.82 3.0m .8an no 7 am 0:33 5.8003 :0 .Dzoa E 0Ma8dp 00 32 a «0000;3033 .wouuammmfim 3 than: Eu... mo ado. 05 .33 0053008 . a 6.6.08 p088 been 05 0.32: «00p 05 E 10839.» «serum 600 0050 was 82503 1%! 05 mo 60.580238 5 v v v . / . a _ i I ., .1 . at. ’4')?” _/ /- ,’ .’ POUNDS TO REMOVE, FORCE 116 '2 I I I I I I I '0 I i + v “—0 - . ! . I 8 Lh— ‘ . 9' I’ + -O— A‘fl _, ._.___.__I 6 I-w -o co. — —— 4 .- - _ -TI... k g m -_ . . + o I 2 -- e + If - ”n—e ° I 0 1 l 1 l I ’ .3 .4 .5 .6 .7 l2/5/60 WEIGHT, POUNDS H. E. _ Figure 12. Relation between the weight and the force to remove Red Haven peaches from the tree. ,5 ”a. ‘1“ .L. 117 The standard deviation in-the y-direction in WWWW : 1.516 P = £1 - £1: E I: 81 15 -( 11 (T) (11) 1812 (22“ = - 0.037714h The correlation coefficient is F04 I P = P = -0.0 7 : - 0,3605 5x U? . For Figure 12: The standard deviation in the x-direction is i . 0.1029 The standard deviation in the y-direction is W I/W 6%") (11) Eel-Iisg—Xee) - 110.0122368 3, The correlation coefficient is r : P = -o.0122 = - 0.0623 0x0? 0 0 APPENDIX IV Drop-test data for Jonathan and northern Spy apple varieties are presented in Tables XI and X11 below. TABLE XI lumber of Jonathan apples contained in each bruise category for various drops ‘_. Bruise 6 - Inch 1 - Foot 2 - Foot 3 - Foot 4 - Foot Evaluation Drop Drop Drop , Drop Drop 0 1 5 2 6 1 l4 2 6 1 2 12 1 cull 1: 5 8 12 119 TABLE XII Number of Northern Spy apples contained in-each bruise category for various drops #— Bruise 6 - Inch 1 - Foot 2 - Foot 3 - Foot h - Foot Evaluation Drop Drop Drop Drop Drop 0 1 2 8 1 II, 1 5 II u I 12 2 1 2 cull 3 5 8 .0- flé- L, y F. '1‘”! / ”“5 ESE Quay 3 1293 50026 4845