jwv U 5 W83 HARVESTING ASPARAGUS MECHANICALLY —- ECONOMIC ASPECTS AND PHYSICAL PROPERTIES By David Roger Knicely AN 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 1902 Approval 3%; //? CE;Z{7«:N David Roger Knicely Michigan ranks first in the midwest and fourth in the United States in the production of asparagus for fresh market and processing, a crop worth $40 million annually in the country. The expense of hand harvest and handling averages 30 to 45 percent of the gross income. This expense plus the uncertainty of the transient labor for harvesting had led to a need for mechanical harvesting. The objectives of this study were to Supply information which would lead to the development of a mechanical aSparaguS harvester by: (1) an economic analysis of mechanical har— vesting. (2) determining the physical properties of the Spears. and (3) determining the harvesting characteristics of a vertical axis paddle—wheel harvester. A review of literature revealed some of the aSparaguS growth characteristics and harvesting methods. It presented a history of the mechanical harvesters which had been de— veloped prior to the investigation. An economic analysis of mechanical aSparaguS harvesting provided an equation for net return in terms of growing, harvesting. and handling costs which can serve as a guide in the production of a profitable harvester. Theoretical results showed that the snapping force was proportional to the Spear cross sectional area raised to the 1.5 power and the snapping angle was proportional to the reciprocal of the average Spear diameter at the point of breakage. The average of the results for the entire season confirmed the theoretical analysis for both relationships. David Roger Knicely Data were taken three times daily for 13 days through— out the harvest season of 1962 to determine the effects of temperature, relative humidity. Spear Size, time of day. and time of Season on the snapping force and snapping angle. A snapping device was constructed to measure and record the bending angle and force required to Snap the asparagus Spear. Force was measured by SR-4 strain gages in a Wheatstone bridge arrangement mounted on a cantilever test arm. Angles were measured by a timer wheel in a D—C electrical circuit. The two circuits were connected to amplifiers and an oscil- lograph recorder to record the results. Studies of the snapped Spears revealed that the size of the Spear was the only significant factor affecting both the snapping force and snapping angle. A fixed level vertical axis paddle—wheel mechanical harvester, using forward Speeds of one, two. and three miles per hour at platform levels of two and four inches. was used to obtain mechanical harvesting results. The percentage of marketable Spears was greatest at two miles per hour at both platform levels. The average harvesting percentages were 57.0 and 50.4 for the two—inch and four—inch platform levels, respectively. An aSparaguS variety comparison of the two—inch plat- form level showed the Viking variety yielded a higher per— centage of marketable Spears than the Mary Washington. Spears contacted but passed over by the harvester plat— form decreased in damage as the forward Speed increased at the two—inch platform level. Damage increased with forward David Roger Knicely speed at the four-inch platform level. Damage was greater for all Speeds at the two—inch platform level with the taller Spears receiving the most damage. High Speed moving pictures of the harvester action re— vealed many spears were contacted by two or three paddles before being snapped or passed over. It indicated the bend- ing angle was not great enough for adequate snapping of the Spears. HARVESTING ASPARAGUS MECHANICALLY —— ECONOMIC ASPECTS AND PHYSICAL PROPERTIES By David Roger Knicely 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 630005 ‘W%%¢ ACKNOWLEDGMENTS The author wishes to express sincere thanks and ap- preciation to the following people who contributed to this investigation: A Dr. Bill A. Stout, Who as major professor provided guidance and inspiration for the investigation and preparar tion of this manuscript. Dr. Frederick H. Buelow, who as temporary advisor in the absence of Dr. Stout, provided assistance in developing the apparatus and editing the manuscript.' Dr. James S. Boyd, Dr. Merle L. Esmay, and Dr. George H. Martin, the other members of the guidance committee. for their many helpful suggestions. Dr. Arthur W, Farrall and the Agricultural Engineering Department for arranging the assistantship and providing the personnel and facilities for graduate study. Dr. Stanley K. Ries, Horticulture Department, and the staff at the horticultural farm for providing the aSparagus plots used in the experimental investigation. Dr. Klaus Daniel, formerly of the Statistics Depart- ment, for providing assistance in the statistical analysis of the data. Mr. James Cawood, foreman of the research laboratory, and the employees of the research laboratory for providing ii facilities and assistance on the construction of the test apparatus. Mr.ThomasIBurenga, research laboratory assistant, for assistance in conducting field tests and construction of the test apparatus. Members of the computer laboratory for providing assistance in the computer analysis of the test results. iii TABLE OF CONTENTS INTRODUCTION REVIEW OF LITERATURE Growth of Asparagus Asparagus Harvesting Mechanical Harvesters ECONOMIC ANALYSIS OF MECHANICAL HARVESTING ENGINEERING ANALYSIS OF SNAPPING FORCES \ Snapping Force versus Area Relationship Snapping Angle versus Diameter Relationship EXPERIMENTAL INVESTIGATION Physical Properties of Asparagus Mechanical Harvester Test CONCLUSIONS RECOMMENDATIONS FOR FURTHER STUDY. SUMMARY REFERENCES iv Page ll 10 22 68 71 73 Table 10 LIST OF TABLES Regression values for the relation of Spear snapping angle to physical properties for Six variates Analysis of variance in regression: m variates Analysis of variance in regression: six vari- ates. Test of the hypothesis Bl = Bg = B3 = B4 = B5 = db = O. . . . . . . . . . . . . . Regression values for the relation of Spear snapping angle to physical properties for four variates Regression values for the relation of Spear snapping angle to physical properties for three variates . . . . . Regression values for the relation of Spear snapping angle to physical properties for two variates Mechanical aSparaguS harvester results for average of eight tests on 90—foot rows Mechanical aSparagus harvester results for the Mary Washington and Viking varieties at a two-inch platform height Condition of aSparagus Spears which made contact with the mechanical harvester plat— form. Average percentages for nine tests on Six rows 'Condition of aSparaguS Spears Which made contact with the mechanical harvester plat— form. Average percentages for nine tests at the two platform heights 36 38 3Q 54 56 61 Figure 10 11 12 13 LIST OF FIGURES Relation between the initial harvester cost and the net return for a 3—row harvester on 30 acres per year Relation between the acres harvested per year and the net return with growing costs of $120 per acre for a 3-row harvester Relation between acres harvested per year and the net return for a 3-row harvester costing $4000 Tractor and trailer used to tranSport Snapping device and equipment to the field Snapping device used to measure the snap- ping angle and snapping force of an asparagus Spear . . . Snapping device with labeled components Amplifiers and oscillograph used to record the snapping information Section of oscillograph tape with recorded results of a typical Spear snapping Asparagus Spear being bent by the snapping device test arm . . . . . . . View of aSparagus Spear broken by the snap- ping device . Relation of the Spear snapping angle to the reciprocal Spear diameter for all the snapped Spears Relation of the Spear snapping force to the cross-sectional area raised to the 1.5 power for all the snapped Spears Relation for one test of snapping force to the Spear cross—sectional area raised to the 1.5 power vi Page 16 17 23 23 24 28 28 31 31 4O 42 Figure 14 15 16 17 18 19 2O 21 22 23 24 25 26 Relation of snapping angle to the reciprocal Spear diameter for one test Experimental vertical axis paddle—wheel aSparagus harvester View of vertical axis paddle—wheel aSparagus harvester Experimental aSparaguS harvester with high Speed moving picture camera and apparatus View of aSparagus Spear as it makes con— tact with the harvester Experimental harvester passing over aSparaguS Spears too short for harvesting A non-damaged Spear and a Spear with broken tip Showing relative growth after two days Condition of Spears passed over in re- lation to forward Speed Condition of Spears passed over in re- lation to their height above the platform bottom . . . . Angular measurement of Spear being bent by the snapping device Angular measurement of Spear being bent by the mechanical harvester Sequence Showing an aSparaguS Spear being snapped by the mechanical harvester Sequence showing Spear being snapped by the second paddle of the mechanical . harvester vii Page 47 4Q 49 51 51 58 58 59 6O 63 63 65 66 INTRODUCTION In the production of aSparaguS for processing in the United States in 1961. Michigan ranked fourth with a gross value of $2,091.000 of a total production value of $31,201,000 in the United States. An additional value of $16.70l.000 was grown for the fresh market (Anonymous. 1961). Production costs for aSparaguS have increased yearly while the average price per unit received in 1960 and 1961 was only 6 percent higher than the 1950-1959 average. A large portion of the production costs can be accred— ited to harvesting. Harvesting and handling costs represent 30 to 45 percent of the gross returns (Carncross, 1956). The uncertainty of adequate labor for harvesting has become a major problem. The diminishing number of migrant workers are demanding higher wages and improved housing while. in return, the growers sometimes receive a poor Quality of harvested Spears owing to carelessness of the workers. As a result. many attempts have been made to mechanize the aSparagus harvest. Research on mechanical harvesting is presently being conducted by several companies and univers— ities. Selective and non—selective methods of mechanical harvesting have been investigated. Kepner (1959) and Downes (1961) reported that non—selective methods caused a sub— stantial decrease in the yield per acre of 35 to 50 percent and a Significant reduction in aSparaguS quality as compared to hand harvesting. Kepner estimated the "break even” point with non—selective machine harvest was about 65 percent of hand harvest return. Various attempts at building selective harvesters have been made, but none have been successful for commercial use. The physical properties of the aSparaguS Spear and its effect upon mechanical harvesting were not known. Stout and Ries (1962) reported on some preliminary tests conducted in 1961, but the results were not complete. Additional infor- mation on the physical properties of aSparaguS is needed be— fore the design of a mechanical harvester is undertaken. The objectives of this Study were to: 1. Increase the available information which would lead to the development of a selective mechani- cal harvester for snapping aSparagus Spears. 2. Determine the aSparagus physical properties as to snapping force and snapping angle as affected by Spear Size, weather conditions, and time. 3. Evaluate a vertical axis paddle-wheel aSparagus harvester. 4. Study the damage to Spears contacted, but passed over, by the mechanical harvester. 5. Conduct an economic analysis of aSparaguS pro— duction with mechanical harvesting. REVIEW OF LITERATURE The design criteria for a mechanical aSparaguS har— vester depends upon the nature of the crop and the method by which it is harvested. This review will therefore be divided into three topics: the growth of aSparaguS. asparagus har— vesting, and mechanical harvesters. Growth of Asparagus ASparagus is one of the most valuable of the early vegetables and perhaps the most important of the perennial crops (Thompson, 1954). Carolus and Brown (1961) express that the American public did not become familiar with the nutritive value of green asparagus until about 1925. Carn- cross (1961) Showed that the total acreage of aSparagus in- creased to 156,000 acres in 1960, a 140 percent increase Since 1925. Over 92 percent of the total crop was grown in California, New Jersey, Washington, Michigan. and Illinois in 1960. ASparaguS has been a native of Europe for over 2.000 years and was consumed during the Greek and Roman Empires. The aSparaguS plant is botanically classified in the lily family. It is an herbaceous perennial with perennial fleshy roots, annual fibrous feeding roots, and an annual vegetative top (Carolus and Brown, 1961). Succulent Shoots arise from the short rhizomes of the crowns. The crop is harvested 3 before any appreciable amount of photosynthesis can occur to supply the food for growth. Therefore. the development of the harvested spears is almost entirely dependent on the foods stored in the roots. Asparagus plants are started from seed in a nursery and grown in a seedbed for one year. The desired crowns with two or more well developed bud clusters are tranSplanted in the location of the permanent planting (Ward and Ellis. 1948). The crowns for green aSparaguS are placed 9 to 24 inches apart in the row. four to eight inches deep. in rows four to Six feet apart (Apple and Barrons. 1945; Johnson. 1900; Wiebe. 1959; Thompson. 1954). Crowns for white, or blanched. asparagus are placed from o to 18 inches apart in the row. 8 to 14 inches deep. with 7— to 8—foot row spacings (Under— hill, 1962). Moran and Isaacs (1960) reported the average yield and number of Spears per acre increased as Spacing in the row decreased. although the mean Spear weight was Significantly greater for increased Spacings. Crowns are placed in deep. loose. well drained. fertile soils with a pH range of Six to seven (Ward and Ellis. 1948; Apple and Barrons. 1945). No Spears are harvested until two years after tranSplanting to allow for root growth. Probable life of an asparagus bed varies from 10 to 18 years. Asparagus Harvesting The harveSt period of established beds is from six to eight weeks in most areas and up to 12 weeks in California (Thompson. 1954). Early season Spears may require harvesting Ul only every third day, but more frequent harvests are required as the temperature rises and the growth becomes more active. Blumenfield (1961) found the Spear growth rate increases linearly with the temperature and also with the Spear height. Results also Showed an average of 59 to 79 percent of the daily growth occurred from 7:00 a.m. to 7:00 p.m. Early emerging plants were found to be more productive than later emerging plants (Ellison and Schermerhorn, 1958). Three classes of asparagus are marketed. Spears may be entirely green, green with white butts, or entirely white (Thompson. 1954). Green Spears are those grown above the ground, whereas, white aSparaguS is grown and cut underground. The method of growing white aSparagus is more costly than for green aSparagus Since complete absence of sunlight must be maintained to assure whiteness of the spears (Honma and McArdle, 1958). Due to this reason and the increasing demand for green aSparagus for processing, only the harvesting of green aSparatuS will be considered hereafter. The two accepted methods of harvesting green aSparagus are cutting and snapping (Ward and Ellis, 1948; Barrons, 1945; Hanna, 1950; Schermerhorn, 1947). Field cut aSparaguS is cut in the field with a knife at or below the ground sur— face with the fibrous portion being removed in the processing plant (Anonymous, 1957). Barrons (1945) observed that only 55 percent of the cut aSparagus was usable for processing. Field snapped aSparagus consists of green Spears snapped so as to be free of non-Soluble fiber or that portion which is tough (Anonymous, 1957). Snapping is accomplished by gripping the Spear just below the tip and bending it forward in an arc. The break will occur at the junction of the high-low fiber region (Carolus, 1949). The relation of Spear length to the snapped length was studied by Huyskes (1959). Tests on cutting versus snapping,conducted in Cali- fornia, Showed a 12.9 percent increase in cutting yield over snapping (Hanna, 1950). Feathering, branching below the break, was pronounced in the snapped stubble toward the end of the season. New Jersey tests indicated that snapping would not work consistently because cfi' interaction between the tur— gidity of Spears and weather conditions (Moore, 1961). The yield of edible material increased up to 15 percent in Michigan tests by using the snapping method while the har- vesting cost was reduced by nearly 50 percent (Carolus, 1948 and 1949; Downes, 1958). As a result of these findings, growers on the east coast and west coast continue to harvest asparagus by cutting the Spears. Most aSparagus in the midwestern states grown for processing is harvested by the snapping method. Properly snapped aSparaguS is preferred by processors and a premium price is paid for it. Results of a questionnaire returned by 29 processors throughout the country to the author revealed hand harvest was used almost exclusively at present. The average price paid for hand cutting was 30 percent of the gross income, while an average of 25 percent was paid for hand Snapping. \I This was for an average cost of 3—1/2 to 4 cents per pound for both harvesting methods. Many processors also indicated a need for mechanical harvesters to replace the frequently unreliable labor and, at the same time. increase the present acreage and production. Mechanical Harvesters A mechanical harvester could be based on either the cutting or snapping principle. Several attempts have been made using either principle and the Selective or non—selective method of harvest. A selective machine will harvest only the Spears of a marketable length and leave the shorter Spears undisturbed. Kepner (1954, 1957a, 1957b. 1959) developed a non- selective harvester which cut the Spears just below the ground surface by means of a bandsaw blade. A gripper unit held the Spears while cutting and then lifted and dropped them in a conveyor which deposited them in a box. The machine/hand yield ratio in 1957 was 0.55 and in 1958 was 0.63. An esti— mated break even ratio of 0.65 must be obtained before the mechanical harvester could be considered practical for com— mercial use under current economic conditions. Downes (1960, 1961) reported on an experiment comparing a tractor mounted cutter bar mower to hand snapping for a 16— day test in 1959. On a weight basis, only 50 percent as much aSparaguS was collected by mowing as by hand snapping with a decrease in the quality of the mowed aSparagus. A set level horizontal axis paddle—wheel harvester. developed by the Friday Tractor Company of Hartford. Michigan. was tested in 1960 (Downes. 1961). Spears were sheared be— tween the wooden paddles and the set level plate. They were carried rearward and upward by the paddles and dropped into boxes. The machine/hand ratio was 0.59 for the 24—day har— vest period. A non—selective mechanical harvester with a bandsaw type device to cut spears at ground level or Slighly above was developed at Rutgers University. Tests in June 1959. comparing frequent and infrequent machine harvests to hand harvest. yielded machine/hand ratios of 0.58 and 0.57, respectively (Mears. 1960). Without exception. the non-selective harvesters have greatly decreased yields and are not economically feasible for commercial use at this time. Several researchers have attempted to develop a selec— tive harvester. Matteoli (1952) developed a machine which separated the aSparagus row into channels with a lever action feeler on each channel set at the desired marketable Spear height. The feeler. connected to a microswitch and electric Selector valve. put into motion a hydraulic cylinder and attached knife which cut the Selected Spear. A series of tines above the blade impaled the Spear Slightly above the cut to bring the Spear to the knife starting position where it is Stripped from the tines and dropped into a container. A Speed of 1-1/2 miles per hour attained efficient selection and cutting of the Spears in the ridged rows. Marihart (1954) invented a machine which used probe devices adapted to be thrust into the ground adjacent to the aSparagus Spears and cut the Spear at a given distance below the ground Surface. The severed Spear was retracted with the probe. The probes were controlled by photoelectric de— vices carried ahead of each probe. This machine is applic- able to the harvesting of aSparaguS or other vegetable Sprouts. A machine invented by Lafferty and Miller (1955) utilized two photoelectric light beams to select a marketable Spear and to indicate when a vertical knife carrier was positioned correctly for cutting the Spear. The actual knife operation was controlled when the Spear broke both light beams at the same time. Several knives were used to cover the entire row width. An invention by Turkington (1956) utilized a feeler arm and microswitch sensing device for selective harvesting of Spears. The machine consisted of a series of rotating arms on a horizontal axis perpendicular to the aSparagus row. When a marketable sized Spear was sensed, the arm was activated and moved downward in a circular path, cutting and retaining the Spear, then backward and upward to an upright stop position. MomentUm caused the Spear to be thrown for— ward into a container. No performance data have been pub— lished. Moore (1961) developed a selective mechanical harvester utilizing an air cylinder to actuate the cutting knife link- age. The piano wire knife haS an additional forward velo- city at the time of contact with the Spear. Each cutting ‘1 [Elli-llllllll..ll'llu|'|’ Al!" I} ' ii. I'll! ’- l 10 cycle of the knife is capable of harvesting a four—inch square area when operated at the arbitrary speed of two miles per hour. Several methods of sensing Spears have been an air stream directed against a vane on a microswitch, a device using feelers where the Spear itself completed an electrical circuit, and a photoelectric cell in which the Spear closed a relay when the light beam was interrupted. The photo- electric cell showed the most promise in preliminary tests. No performance data were available. A mechanical snapping unit harvester was constructed in 1960 at Michigan State University (Stout and Ries. 1962). The snapping units consisted of two parallel 3/8-inch rods, 30 inches in length, and connected at the ends to an endless chain. All Spears of harvestable length extended through one of the snapping units. AS the unit traveled along the bottom horizontal plane, a Spur gear on the unit end engaged in a gear rack which rotated the unit. The motion snapped the Spear and held it against a Sponge rubber bar. The chain conveyed the Spear to the harvester top where it was released into a container as the unit was rotated to its original position. Due to mechanical difficulties, no per- formance data were collected. A An evaluation of present selective harvesters is Some- what difficult because no performance data were available to compare. Expenses involved in a series of independent cutters and sensing devices will require the machine/hand yield ratio to be much higher for the selective mechanical harvesting than the value reported by Kepner. No cost analyses were con— ducted for any of the mentioned Selective harvesters. ECONOMIC ANALYSIS OF MECHANICAL HARVESTING Although the first requirement of a machine is that it satisfactorily perform its intended function. the economic aSpect of the machine application is also of great importance. The design problem is largely controlled by anticipating costs of production and farm operating costs. The economics of mechanical aSparagus harvesting were analyzed by using a procedure given by Bainer. Kepner. and Barger (1955). The total cost for field operation includes charges for the harvester, power utilized, and labor. The costs are grouped under the headings of overhead costs per year and the fuel and labor costs. Overhead costs include depreciation, interest on investment, taxes, insurance, Shelter, repairs, maintenance, and lubrication. Values for the growing costs prior to harvest and the handling and marketing costs are taken from a study by Carncross (1956). Machine capacity is expressed as: speed x width x efficiency = SW .AcreS} 825 825 .hour 1 The assumptions for the analysis are as follows: 1. Interest rate on average machine value —- 6 per- cent per year. 2. Taxes, insurance. and Shelter —— 1.5 percent of initial cost per year. 3. Repairs, maintenance, and lubrication —- 4 ll U] 10. 11. 12. 12 percent of initial cost per year. Fuel consumption -— 8.5 horsepower hours per gallon. Fuel cost —— 20 cents per gallon. Labor charge for tractor operator —- $1. “\1 U1 '0 (D '1 hour. Service life of mechanical harvester —— 5 years. Salvage value of harvester —- none. Service life of tractor -— 10 years. Salvage value of tractor —— 10 percent of initial cost. Annual use of tractor —— 600 hours. Tractor and harvester depreciation is linear. The following values were computed from the above assumptions: A. Harvester overhead costs per year. \ Depreciation ‘ 0.20 of initial cost Interest on investment 0.03 Taxes. insurance. Shelter 0.015 Repairs, maintenance. lubrication 0.04 Total 0.285 of initial cost Tractor overhead costs per year. Depreciation 1 — .10 0.09 of initial cost 10 Interest .06 x 1_j .10 0.033 2.... Taxes, insurance, Shelter 0.015 Repairs, maintenance. lubrication 0.04 Total annual costs 0.178 of initial cost 13 Fuel and labor costs. Fuel cost per acre 825 x $.20 per gallon SWE 8.5'Hp hours per gal. h se r = 19.4 H x or powe ‘SWE—_ Labor cost per acre 825 X $1.75 er hour = 1445 /Dollars’ S—_WE ' p "SW—E— \‘A'c're ) Harvesting yield loss. Yield loss = Acres x potential yield x price per unit x harvesting loss. Harvesting machine efficiency = 100 - harvesting loss. Variables which have been considered and their notations are as follows: (I) QH'Uv-<1t'*t'n Acres harvested per year. Initial cost of mechanical harvester. Initial cost of tractor. Tractor horsepower (75 percent maximum from Nebraska Test). tractor hrSu Fraction of total tractor use (based on 600 hrs. ' Forward Speed, miles per hour. Rated width of implement, feet. Field efficiency, percent. Loss of marketable weight, percent of total. Potential yield, pounds per acre. Selling price per pound. Number of harvests per year. Growing costs per acre prior to harvest (cost of establishing the bed, fertilizer, labor, use of land, tractor and equipment, and miscellaneous costs). 14 M = Marketing costs per acre. The expected net returns per year from mechanical aSparaguS harvesting are: of: (lOO-L) YPA _(GA + 0.285C Net return = Gross return - total expenses G SS ret n = (100-L) YPA to ur 100 Growing costs = GA Harvesting costs = 0.285 Ch Tractor costs = 0.178 CtU Labor and fuel costs = (1445 + 19.4 H) AT SWE Handling costs = MA Combining the above terms will produce the net return 100 h + 0.178C U + (1445 + 19.4H) AT + MA t swg ) AS an example for illustrative purposes, the following values are given to the variables: 0 (+5 I ' $2500 (average local price) 22 horsepower .005 per acre = 3 miles per hour = 15 feet (three 5—foot row5) = 75 percent 20 percent loss 1600 pounds per acre (Michigan average) $0.13 per pound (Michigan average) 2 8 harvests per season 3 |-] "U I-< 1"" tn 2 m C! I II = $8.00 per acre 15 Gross return = (10200 20) x 1000 x 0.13 A = $100.50 A Tractor costs = 0.178 x 2500 x .005 A $2.22 A. (1445 + 19.4 X 22) 8 A = $4.45 A b +f1= La or ”e 15 x 3 x 75 Marketing costs = $8.00 A The unknown variables remaining are pre—harvest grow— ing costs, initial harvest cost, harvested acreage. and net return. A comparison between any three of these variables can be made by holding the fourth constant. Figures 1. 2. and 3 Show the results of holding constant the acreage. har— vester cost, and growing costs, reSpectively. The graphs show the types of curves which result from the assumed conditions. A designer can compute the expected net return by the use of the developed equation and then design a harvester which can be used profitably. Net return from hand harvest can also be computed by substituting hand harvesting costs for harvester, tractor, fuel, and labor costs. A comparison of the two could determine the merits of mechanical harvesting as compared to hand harvesting. 1 T I I PREHARVEST r IOOO T 800* O) C) C) fl, 200 ” NET RETURN, DOLLARS b o o T “200 0 I000 2000 3000 4000 5000 INITIAL HARVESTER COST, DOLLARS Figure 1. Relation between the initial harvester cost and the net return for a 3—row harvester on 30 acres per year. I 2500 INITIAL COST OF HARVESTER 2000 T I500 U I000 500 ~ NET RETURN , DOLLARS "500 I 1 1 1 l l 1 1 1 l l 20 4O 60 80 I00 ACRES HARVESTED PER YEAR Figure 2. Relation between the acres harvested per year and the net return with growing costs of $120 per acre for a 3-row harvester. 5000 T PREHARVEST GROWING COST 5 I 4000 T 3000 T 2000 IOOOt NET RETURN , DOLLARS U 1 ‘HDCHD 1 1 1 1 L 1 1 1 1 1 20 4O 60 80 I00 ACRES HARVESTED PER YEAR Figure 3. Relation between acres harvested per year and the net return for a 3-row harvester costing $4000. ENGINEERING ANALYSIS OF SNAPPING FORCES Before comparing the physical properties of the aSpara— guS Spears, a theoretical relationship of the snapping force and snapping angle to the cross—sectional area of the Spear at the point of snapping is needed. This was accomplished by using equations for homogeneous strength cantilever beams. The following assumptions were made for the analysis: 1. The Spear cross-sectional area is nearly circular. 2. Maximum tensile strength at snapping is constant for all SpearS. 3. The Spears snap due to bending only. 4. Maximum tensile strength is constant over the entire cross-section. Snapping Force versus Area Relationship The following is an analysis of the relationship of the snapping force to the cross-Sectional Spear area: 2 - Area (A) = _%92 sac—3%— (1) where: d is the major Spear diameter b is the minor Spear diameter Snapping stress (S) = gF-= constant (2) 3 and l = 77d for a circular beam c 32 where: I is the moment of inertia 19 20 c is the distance from neutral axis to surface M is the bending moment 3 Bending moment (M) = is =nd S c 32 (3) or M = Kd3 if K =.JL§ = constant 32 The bending moment consists of a force (F) times a moment arm (X). M = FX. Assuming the moment arm to be constant: FOCCI3 (4) orFocA1~5 (5) Snapping Angle versus Diameter Relationship Assume the aSparagus Spear is a circular cantilever beam with a concentrated load applied at one point. . . pL3 Max1mum deflection (D) = ___ (6) where: P is the concentrated load L is the distance from the applied load to the snapping point E is the modulus of elasticity I iS the moment of inertia Snapping Stress (S) = g§.= constant (2) where: M = PL c = d/2 - PLd then S TZT' and 31: ___ 2S (7) I 7T Substituting into equation (6) 21 2 23L (8) 9:517:— If the distance L remains constant and S and E are constant terms. then R D = _ d 28L2 where R = 3E and is constant Therefore, DOC-clI (9) (The deflection, or snapping angle, is inversely pro— portional to the Spear diameter.) EXPERIMENTAL INVESTIGATION Physical Properties of ASparaguS Before attempting to design and construct an aSparagus harvester, tests were conducted to determine the physical properties of aSparagus and their relation to harvesting by the snapping method. An experimental asparagus Snapping device was developed to obtain these physical character— istics. The objectives of the physical property tests were to determine the relation of the snapping force and snapping angle to: 1. Temperature Relative humidity Size of Spear #010) Time of day 5. Time of season Apparatus An Allis Chalmers Model G tractor was used to pull a trailer tranSporting the experimental aSparagus snapping de- vice and instrumentation as shown in Figure 4. The snapping device (Figure 5) was used to record both the snapping force and snapping angle of each Spear tested. Experimental snapping device.——Components of the snap— ping device (Figure 6) are described as follows: 22 23 Figure 4. Tractor and trailer used to transport snap— ping device and equipment to the field. Figure 5. Snapping device used to measure the snapping angle and snapping force of an aSparagus Spear. 24 I _— ~ I” - H fluVil- Figure 6. Snapping device with labeled components. Gear reducer Control box with D-C rectifier Magnetic clutch Magnetic brake Timer wheel with microswitch Test arm mmuow> 25 1. Test arm. Calculations based on formulas of strength of materials were used to determine the Size of the test arm. The three-inch aluminum test arm was one—half inch wide and one—fourth inch thick. The one—half inch driving shaft centerline was located two and one-half inches from the bottom of the snapping unit (Figure 6, part F). 2. Bridge circuit. Type A-7, SR-4 Strain gages were mounted on the cantilever test arm to measure the bending forces of the aSparaguS Spears. Four gages were used to form a Wheatstone bridge arrangement for connection to the Brush recording equipment. The gages were placed seven-eighths inch from the driving shaft centerline, Two gages were placed on each of the two arm faces to obtain maximum bridge sensitivity and temperature compensation. The gages were waterproofed with a thin layer of wax and covered with plastic tape to protect them from being damaged. Gage lead wires were connected to a four conductor Shielded cable to conduct the signal to the recording equipment. The shielded cable was securely tied to the test arm to prevent damage from forces applied to it. 3. Timer wheel circuit. A two-inch diameter timer wheel was placed on the driving shaft to indicate the angular displacement of the test arm. One Side of the timer wheel was divided into lO-degree increments with alternate incre- ments coated with clear nail polish to form insulated sur- face sections. A D-C circuit was then completed through the timing wheel to the amplifier by using a small dry cell to 26 provide the necessary power. The alternate open and closed electrical circuit produced a square wave recording (Figure 8) and indicated the test arm position. 4. Magnetic clutch-brake System. A Dyna-torQ Model 303 magnetic clutch and brake (Figure 6, parts C and D) was used to obtain nearly instantaneous starting and stopping of the test arm. Power for the magnets was supplied by a 90 volt D—C circuit. A contact switch and a rectifier (Figure 6. part B) were used to provide the correct power to the cir- cuit. A microswitch. operated by a cam arrangement on the edge of the timer wheel (Figure 6, part E). controlled the action of the clutch and brake. When the microswitch arm was in a released position, the field coil of the magnet on the clutch armature was excited, causing the uniform magnetic attraction of the rotor for the armature to assure complete contact of the friction surfaces and permitted the test arm to rotate. AS the cam caused the microswitch arm to be Sup— pressed, the circuit was switched. causing the clutch magnet to be released and. instantaneously, the brake rotor was at- tracted by the stationary armature causing the test arm to Stop and hold its position while the circuit was closed. A lock—in torque of 5.0 pound-feet was Specified on both the clutch and brake when engaged. 5. Gear reducers. A 50:1 ratio gear reducer was placed in series with a 9.67:1 gear reduction unit to form a total reduction of 483:1 between the electric motor and the test arm driving Shaft (Figure 6. part A). A roller chain 27 connected the units together. The large reduction was neces— sary to eliminate undesirable vibrations of the test arm while it was accelerating to operating Speed. A flexible coupling connected the reduction unit to the magnetic clutch armature. The reduction unit was connected to the electric motor by a flexible rubber tube. 6. Bearings. The snapping unit components were held in alignment by four Fafnir PB ball bearings and secured to the base by pillow blocks. A handle was attached to the unit to aid the tranSporting of the unit to each Spear. Electric motor.——A one—fourth horsepower. 115 volt, capacitor electric motor was used to operate the snapping device. Maneuverability was increased by attaching a handle to the motor base. Recording instruments."—Two Brush strain analyzers and a two-channel Brush oscillograph were used to record the snapping force and snapping angle of each Spear (Figure 7). The bridge circuit Signal was conducted to one amplifier while the D—C signal was conducted to the second amplifier. The resulting signals were conducted to the oscillograph for recording. The recorder was calibrated for snapping force by adjusting the pen deflection when a two-pounds weight was applied to the test arm while the arm was locked in a hori— zontal position. The deflection for angular diSplacement was relative and needed almost no adjustment. The amplifiers and recorder were strapped to a board 28 Figure 7. Amplifiers and oscillograph used to record Snapping information. Figure 8. Section of oscillograph tape with recorded results of a typical Spear snapping. 29 mounted on rubber Springs to assure safe transport. A one— half inch Sheet of polyethylene foam was placed between the instruments and the board to reduce Shock. Generator.-—A 3000 watt, 115 volt, portable generator was used to provide the electrical energy to operate the snapping devices and recording instruments. It was powered by a gasoline engine. An auxiliary 1500 watt. 115 volt generator was mounted on the tractor. It was powered through the power-take—off drive of the tractor. Psychrometer.-—A Sling psychrometer. containing a wet-bulb and a dry—bulb thermometer, was used to obtain the wet and dry bulb air temperatures. The dew point temperature and the relative humidity were then determined from a psychrometric chart. Procedure The experimental aSparagus snapping device was con— structed, as previously described, and tranSported to the field for tests on the standing Spears. Three tests were conducted during the day: 5 a.m}, 10 a.m., and 3 p.m. The 20 Spears selected for each test had a minimum base diameter of one-fourth inch and a height of seven to ten inches. Prior to the snapping for each test, the 20 Spears were selected. dry-bulb and wet-bulb temperatures were re— corded, and weather conditions were noted. The recorder was O calibrated for snapping forces by adjusting the amplifier so I that each line of deflection of the oscillograph recording 30 represented 0.1 pound force applied to the test arm. During each test the snapping device was placed beside the selected Spear, without regard for the major Spear diameterorientation, and aligned with one-half inch between the test-arm and the Spear. After recording the initial Spear height to the nearest eighth inch. the recorder was Started and the magnetic clutch was engaged. The Spear was bent by the test arm until it was snapped or until the mag— netic brake was engaged (Figures 9 and 10). Braking oc- curred after the test arm had rotated about 125 degrees be— yond the vertical position. The test arm rotated at 3.6 rpm. providing a contact Speed of 1.13 inches per second. Spears which did not snap were also included in the 20 Spear test. Snapped Spears were measured for snapped length. major diameter, and minor diameter. Diameters were measured at the point of snapping to the nearest sixteenth inch. The information for eaCh Spear was recorded on the oscillograph recording paper with the deflection lines. The oscillograph was operated at a standard recorder speed of 0.5 centimeters per second. The information for each spear observation was placed on a separate coded punch card. Recorded information in— cluded time of day and season, weather conditions, temper— ature, relative humidity, spear lengths and diameters. and test and Spear numbers. A parameter tape was made which described the variables to be used in the analysis. The punch cards and parameter tape were fed into the MISTIC digital computer for analysis. Results were obtained by 31 Figure 9. ASparagus Spear being bent by the snapping device test arm. kJ’ su‘s Figure 10. View of aSparagus Spear broken by the snapping device. . . 411:; 1r .iiL-r FALL" 32 Lising a library routine program (K l6—M) for multiple re— ggression. The taped results were then printed out by the Lise of a teletype. Ilesults and Discussion The results and Significance of the physical property tests were determined by the method of least squares. The rnethod of least squares gives the most probable values of the unknowns that render the sum of squares of the differ- ences a minimum. where the most probable values of the un— knowns are those that most nearly satisfy the observation equations. The result is a straight line relationship called a multiple regression equation. The basic regression equation is: V = 60x0 + 01x1 + 82X2 + 03x3 + 04x4 + b5x5 (10) where: Y = Calculated snapping angle from regression plane bO - Y—intercept value for constant XO b1 - Regression coefficient for temperature. X l 2 Regression coefficient for relative humidity. X2 0" ll b3 = Regression coefficient for l/diameter, X3 CT II 4 Regression coefficient for time of day. X4 b5 = Regression coefficient for time of season, X5 If X0 is defined as one, giving bOXO a constant value 0(0, then Y =‘Xo + b1X1 + b2X2 + b3X3 + b4X4 + b5X5 (11) The general form for observation t is: Yt *‘1 - blxtl - b2Xt2 - P3Xt3 ‘ b4Xt4 - b5Xt5 = e (12) 33 A . where e is the error or difference between Y and Yt' The basic equation for the sum of squares of the difference is: 2 2- - (Yt ‘ 0“ letl ‘ b2Xt2 ‘ b3Xt3‘ b4Xt4 ' bSXtS) ' e ‘ Z The desired values of (x. bl, b2, b (13) 3, b4, and b5 are those which will render the function a minimum. This is ob— tained by equating to zero the derivative of the function with reSpect to that variable. following values: 51.2.=0, dd dZ - 21—52 0 dz= at. 0 It is desired to have the dZ _ as. " 0 d2 _ €03 - 0 dZ _ as, 0 Differentiation of equation (13) with respect to each of the variables yields: dZ dZ dZ 2 (Y1: - 0< - bixti- b2Xt2- P3Xt3‘ b4Xt4- b5Xt5)(’Xt1?Il) 2 (Yt - or — 131th — b2Xt2— b3Xt3— b4Xt4— b5Xt5)(-do<) (14.1) 14.2) 2 (Y1: - o< — b1Xt1- b2Xt2- b3Xt3— b4Xt4- b5Xt5)(-Xt5db5) (14.6) Setting the derivatives equal to zero and simplifying: dz: = an 0 d2. .. 3151-0- s}.=0= db5 (Yt-CX- blxtl- b2Xt2— b3Xt3— b4Xt4— b5Xt5)(-1) (15.1) (Yt-rx— b1Xt1- b2Xt2- b3Xt3- b4Xt4- b5Xt5)<-Xt1) (15.2) (Yt-cx- b1Xt1— b2Xt2— b3Xt3— b4Xt4- b5Xt5)(—Xt5) (15.6) For N observations the equations become: 34 r, 2 . r .- , , c- r r <- 2 ‘r r 1' V' ’ 0(4th + blzxtlktZ + DZéXtZ + b3“"\t2‘\t3 + b4gkt2AL—4 + b52Xt2X‘t5 :EYtXtZ (16.3) , r , , 2.2 , OZXt3 + b15-1t14‘fit3 + bzfxtzkw + b3‘9‘153 + b451t3xt4 + bjth3Xt5 = ZYtXt3 (16.4) .. , , -. .. ,2 alXt/J + blEthxt4 + bZEXtZXtcl + b3£.\t3xt4 + b4cxt4 4” O5ZXt4Xt5 = zYtXt‘l ‘ (16.5) yzxts + blfxtlxts " IDazxtaxts +b3‘ixt3Xt—5 + b42Xt4Xt5 + bszfo = Zthts ' ‘ (10.0) The development of the equations for each of the re- gression coefficients is performed by the use of matricies. These calculations were performed by‘the MISTIC digital com— puter and will not be further developed here. All the values for the symbols in equations (16.1) through (16.6) were printed out on the computer results. Values for regression equations with fewer variables were computed from modifi— Cations of the above equations. Computer results of the regression values for the snapping angle are presented in Table l. A test was con- ducted to determine the validity of the regression coef— ficients for the snapping angle equation. The method ex— plained by Rao (1952) Shows how far the concommitant variables are helpful in prediction. If these variables are of no use, then the prediction formula does not depend upon them so that Bl = 32 = B3 = Bn ='0. Provided this hypothesis is true, then the minimum value of 2(Yt —cx)2 in equation (13) is ZY% - nyZ, which is the total sum of squares with (n — 1) 35 degrees of freedom. The reduct squares is due to regression. ion in the above Sum of The general analysis of vari— ance in regression by Snedecor (1946) is presented in Table TABLE 1. Regression values for the relation of Spear snap- ping angle to physical properties for Six vari— ates. Standard Partial Corre— Regression Error of lation Coeffi— Variate Slope Value b's cients of b‘s o( .1_/ +07.5100 11.0741 .000 bl(temperature) — 0.04504 .108976 -.01628 b2(relative humid- ity) — 0.0115 .07080 —.00507 b3(l/diameter) + 0.0309 1.8422 +.20100 b4(time of day) + 0.8250 1.8786 +.01730 b5(time of season) 0.0137 .07382 +.00732 l/q is a constant with the value being the y—intercept. TABLE 2. Analysis of variance in regression: m variates. Degrees of Sum of Mean Freedom Squares Square F—test Source (df) (SS) (MS) Result Regression m — l RZLZVZ MSR , . F: 32. . 2 M Res ReS1dua1 n — m (l-R ) 13' ° Total n — l .Eyg number of variables number of observations ‘4de observation Snapping angle multiple correlation coefficient 3 (.4 . 36 An analysis of variance in regression applied to six variates in Table 3 Shows the hypothesis must be rejected Since F is greater than Fa. even at the 1 percent level of significance. Hence. the variables considered above are useful in prediction. TABLE 3. Analysis of variance in regression: Six variates. Test of the hypothesis Bl: 82: B3: B4: Bj=f7o= 0. Source df SS MS F Regression 5 14539 2907.8 F = 3607.3 ' ,_ ’12 ReS1dual 044 323569 5ng_44 3‘ .44 Total 649 333103 = 5.18/ Fa;5. 044 = 2.25 for a = 53 level of Significance 3.10 for a 13 level of significance It is now desirable to test whether the removal of a variable will decrease the accuracy of the prediction. This iS equivalent to testing whether Bn = 0. Using the basic re~ greSS1on equat1on, Y =CXO + Ble + BZX2 + B3X3 + B4X4 + B5X5. and the testing hypothesis Bn = 0. the test statistic is: 2 2 = Ian = Ian Variance (b ) 2 n Sbn Reject the hypothesis if F is greater than Fa where: bn regression coefficient Sbn standard error of bn Testing the hypothesis ;10 = 0 37 = (07.5100)? = F (11.0741)Z 33'44 r Fa;l, 644 = 3.85 for the 5% level of Significance. There- fore, the hypothesis was rejected Signifying that u should not be disregarded in the equation. Testing the hypothesis B1 = 0 .04504)2 .108976)3 A F = = 0.171 A Fa;l, 644 = 3.85 for the 5% level of significance. There— fore. the hypothesis was accepted and signified that removal of b1 from the equation would not decrease the prediction accuracy. The remainder of the hypotheses and F values are: B2 = 0 F = .0207. B3 = 0 F = 27.384 B4 = 0 F = .193 B5 = O F = .0345 Since the F values for b2 and b5 were the smallest. they were assumed to be zero and the regression coefficients were recomputed using the four variates. Partial corre- lation coefficients for b2 and b5 Shown in Table l approached zero which further indicated they could be neglected. Table 4 presents the results for four variates. A com— parison of the regression coefficients in Tables 1 and 4 Showed the values were nearly identical with little change in prediction. The partial correlation coefficient and F value for b1 remained small. Table 5 indicates the results after the elimination of b1 from the equation. 38 TABLE 4. Regression values for the relation of Spear snap- ping angle to physical properties for four variates. .Standard Partial F - values Regression Error Correlation for Variate Coefficient of b's Coefficient B = 0 <1 +66.587 6.6296 .000 100.88 b1 (temper— ature) — .04196 .10041 —.01644 0.175 b3 (l/dia— meter) t 9.6333 1.8145 +.20447 28.187 b4 (time of day) + 1.0111 1.5035 +.02645 0.452 Fa;l, 646 = 3.85 when a = 5% level of Significance. TABLE 5., Regression values for the relation of spear snap— ping angle to physical properties for three variates. Standard ,Partial F — values Regression Error Correlation for Variate Coefficient of b's Coefficient B = 0 ‘3 +64.567 4.5348 .000 202.73 b3 (1/dia- . meter) ' + 9.6583 1.8123 .20506 28.40 b4 (time of day) + .57010 1.0703 .02094 .284 Fa;l. 647 = 3.85 when a = 5% level of significance. The remaining regression coefficients still Showed only minor changes. The F value of 0.284 for b4, much below the rejection value of 3.85 for Fa. was also eliminated from the 39 equation. Table 6 shows the F values of ~;and b3 much larger than 3.85. therefore. no more elimination can be done without decreasing the accuracy of the prediction. TABLE 6. Regression values for the relation of Spear snapping angle to physical properties for two variates. Standard Partial F — values Regression Error Correlation for Variate Coefficient of b'S Coefficient B = 0 '1 65.659 4.0430 .000 263.74 03 (l/dia— meter) 9.6818 1.8108 .20556 28.59 Fa;l. 648 = 3.85 when a = 5% level of Significance. An analysis of variance in regression was computed after the elimination of each variate. The resulting F values were much larger than the Fa values. thus rejecting the hypothesis that all the variates equal zero. The resulting expression for the Spear snapping angle iS the linear equation: 9 = 05.059 + 0.0010x3 1 22.3545 where: Y is the computed snapping angle with a Standard error of estimate of 22.3545 X3 = reciprocal of average Spear diameter 65.659 = y - intercept ‘ A graph of this equation is shown in Figure 11. A Similar analysis was conducted to determine the equation expressing the Spear snapping force. The result 40 .mumoam pummmcm 0:» Han How ummemflc umvdm Hmooudfiuou use 0p mamas mcfimaacm Madam 0:» mo :ofiuwaum .HH unammm meoz. 2. mm...ms.<.o m< ll (area)1-5 at the snapping point 0‘ ll Slope of the curve 44 a = y-intercept The slope is expressed as: ZXy— EEEX b = N 2X2 - (2)02 N where N = number of Spears in the test. The y-intercept is: = EX _ bEX N N a The standard error of estimate (Standard deviation from the line for a given X value)is: synaglibpxyeigx] N — 2 SE = f The same equations can be used for the relation of the snapping angle to the reciprocal of the spear diameter by substituting for y and X. where: snapping angle X reciprocal of Spear diameter Solving for the snapping force in Figure 11: .3 ='—.0202 + 9.753x : .2202 where: a = -0.0262 b = 9.753 SE= 1.2202 Similarly, Y = 45.17 + 19.247X 1 24.98 for the snapping angle. The linear regression curve in Figure 13 follows the observation points closely for all values. Thus. the direct Quezon m.H 0:» Op panama «one mom...uz:s 21m Hmcowuoumummouu umodm 0:» o» ouuow mchanm no away one you cemumaom .ma unamwm 0....2. .8 .35: _. O: . 3 a mu - J u d d 1 a 4 010 id 1 4 O \\\\ t\\\\\\ .\\\ + + + \\ . \\ m r + + 10.0 + \ \ V + .d \\\ \\\+ Hm \ + \ N \\\ + \\\ nu . \\\+. + \\\ .o; _\\\ mw \ + \ H \ 0 \\\M\ \\\ . a. .+\ m 83.0 ... 382.8 .. 88.0- n» .0. .d \ + O .. m C\\\\\. .nem Mw mvN 46 5 held true for proportionality of snapping force to (area)l‘ this test. Similar results occurred for each of the other tests during the season. The linear regression curve in Figure 14 is an ex— ample of the scattered values when comparing the snapping angle to the reciprocal of the average Spear diameter. A distinct difference exists between the individual regression curve values of tests throughout the season. However. the overall regression for the season Showed a trend with the snapping angle being dependent upon the Spear Size. Field conditions of uneven Snapping heights and non—uniform Spear cross—sections were not considered in the theoretical analysis or results. Mechanical Harvester Test No attempt was made to design a completely new mechan— ical aSparagus harvester. Some modifications were made to the vertical axis paddle~wheel harvester described by Stout and Ries (1962) and Stout (1962). Tests were conducted to determine the results of a semi—selective harvester using the snapping method. The objectives of the mechanical harvester tests were to: 1. Determine the relation of harvesting efficiencies to forward Speed. 2. Determine the relation of harvesting efficiencies to the platform height. .umuu 626 now umpmamflp umodm HmuOHQwuwu on» Ca vamnm mcwmamnm mo cofipmaom .vH munmfim meoz. 2. KMHMESQ m52 23 9. OD p L b L p b p — b Pl - .P p P L OV— 8338930 ‘ 310W 9NlddVNS 43:4; 3. Compare the harvesting efficiencies of two Spear varieties. 4. ‘Compare the harvesting results to the estimated Snapping percentages. 5. Determine the relation of Spear damage to forward Speed and platform height. 6. Compare the desired Snapping action to those re— vealed by high Speed moving pictures. Apparatus , A Ford Model 601 Select-o—matic tractor was used to operate the mechanical aSparaguS harvester. The harvester was mounted to the tractor 3—point hitch and powered through a 1000 rpm power-take—off shaft. The harvester consisted of fwo Synchronized vertical axis paddle—wheels with axes lo— cated 21 inches apart. Each wheel contained eight paddles four inches high and 14 inches long. resulting in a wheel diameter of 28 inches (Figures 15 and 16). The Side of the paddle contacting the Spears was covered with 1/4—inch thick polyethylene foam which extended l/4—inch below the paddle to reduce Spear bruising and breakage. One paddle turned clockwise and the other counter— clockwise above a fixed level platform having an M—Shaped leading edge. The platform outer edge extended radially from the axes of the paddle—wheels forming a vertex at the center of the aSparagus row. A V—notch four inches wide and extend— ing back one inch was cut from the vertex to ensure better contact between the Spear and platform. The entire leading 49 \ x 9, K7 Figure 15. Experimental vertical axis paddle-wheel aSparagus harvester. 35:555. vase Figure 16. View of vertical axis paddle-wheel aSparagus harvester. edge was cushioned by l/8-inch thick rubber. The 3/8—inch thick platform extended 20 inches to the rear and had side panels to hold the harvested Spears. The forward motion of the platform and the rearward motion of the paddles provided the action desired to snap the asparagus Spears. A two—inch distance was maintained between the paddles and the platform. Spears taller than the bottom of the paddle would be contacted by both the paddle and platform and snapped (Figure 18). Spears con— tacted only by the platform were bent and passed under the platform (Figure 19) or broken. All Shorter Spears were un— touched by the harvester. The distance between the platform and the ground was maintained by gage wheels. A beam balance was used to deter— mine the weight of the harvested Spears. A high speed moving picture camera was mounted on the harvester to obtain pictures of the aSparagus spears being snapped at various Speeds (Figure 17). Power for the camera Iwas provided by a Storage battery. Two flood lamps provided additional lighting for the focused area. One paddle—wheel was removed to permit a better view of the snapped action. Procedure Six harvester teSt conditions and one hand harvest check were established as Shown in the following table: 51 “1353-...“ "9% \fl‘ ’ I. v‘~ $‘1. ‘ -‘D Figure 17. Experimental aSparagus harvester with high Speed moving picture camera and apparatus. Figure 18. View of aSparagus Spear as it makes con- tact with the harvester. Test Platform Forward Paddle Condition Height. Speed. Speed, Paddle Tip Speed Number in. mph rpm *Forward Speed" 1 2 1 86 6.14 2 2 2 107 3.80 3 2 3 129 3.07 4 4 1 86 6 l4 5 4 2 107 3.80 6 4 3 129 3.07 ‘1 Hand harvest The test plot consisted of 23 90—foot rows with four- foot row Spacing. The two end rows were used as guard rows. Rows 2 through 8 were random selections of the Seven condi— tions. The remaining 14 rows were divided into seven rows each of Mary Washington and Viking varieties of aSparagus. Harvesting efficiency test.—-Each test row was har— vested at the platform height and Speed indicated by the condition number assigned to the row. The total number of Spears snapped and collected were placed in one paper con— tainer. the snapped Spears remaining on the ground were placed in a second container. and the Spears of marketable Size which had not snapped were snapped by hand and placed in a third container. The harvested Spears were weighed and the results recorded. Remaining Spear damage test.-—One row of each Speed and platform height was used for damage test checks of mech— anical harvesting. The Spears ranged in height from the 53 bottom of the platform to the bottom of the paddle. For ex~ ample. all Spears with forward Speeds of l. 2. and 3 miles per hour at a two—inch platform height were divided into three groups and marked before the harvest. The Spear heights were two to three inches. three to four inches. and four to five inches. Spears of these heights Should all be remaining after the harvester passed over them. The condi— tion of the Spears was observed and recorded immediately following the mechanical harvest. Observations and height measurements were also recorded 24 and 48 hours after the first observation. High Speed photography.-—High Speed moving pictures were taken on 16 millimeter film at the rate of approximately 400 frames per Second. A 100—foot roll of film was used for recording the snapping action for each of the three forward Speeds. A platform height of two inches was used for all the tests. Results and Discussion Harvesting results obtained from eight tests are pre- sented in Tables 7 and 8. The mechanical harvester snapped between 66.7 and 84.4 percent of the marketable aSparaguS by weight. However. only 47.7 to 58.6 percent of the total weight was of marketable quality while an average of 13.5 percent of the asparagus remained loose on the ground and 8.5 percent was damaged or broken into small pieces. A comparison of efficiency to platform height in Table 7 Shows 82.1 percent of the Spears were snapped at the 54 TABLE 7. Mechanical aSparaguS harvester results for average of eight tests on 90—foot rows. Snapped by Damaged Loose on Marketable Condition Machine 1/ Spears 2/ Ground 3/ Asparagus 3/ Number Percent Percent Percent Percent 1 (2-1)* 78.69 9.32 15.22 54.15 2 (2-2) 84.45 11.35 14.50 58.60 3 (2—3) 82.67 11.37 13.58 57.72 4 (4—1) (66.68 5.04 11.54 49.50 5 (4—2) 71.19 5.48 12.39 53.32 6 (4-3) 66.86 5.98 13.14 47.74 .Average at two inches 82.13 10.75 14.41' 56.97 Average at four inches 68.37 5.67 12.29 50.40 Total . Average 76.02 8.50 13.47 54.05 l/ Snapped by machine = Marketable wt: harv.+wt. loose+wt. dam. total weight offlaSparagus 2/ Damaged = Weight of damaged Spears Total weight of asparagus 3/ Loose on ground = Weight loose on ground Total weight ofTaSparaguS fl/ Marketable aSparagus = Marketable weight harvested Total weight 5f aSparaguS Total weight of aSparaguS = Spears snapped by machine + remainder of total harvested by hand * Platform height and forward Speed are listed, reSpectively. in parenthesis. ' . 55 two—inch platform height. Percentages of marketable Spears were 57.0 and 50.4, reSpectively, providing a 54.05 percent overall average. However, a higher percentage of damaged Spears and Spears dropped on the ground resulted at the two— inch height with a total of 25.16 as compared to 17.96 at the four-inch height. The two miles per hour forward Speed produced the highest snapping percentage and percent marketable aSparaguS for both platform settings. An average of 13.76 percent more aSparaguS of marketable Size was missed at the platform height of four inches. A comparison of the Mary Washington and Viking vari- eties of asparagus Shows the percent snapped and percent loose on the ground were nearly the same for the average of both varieties as presented in Table 8. The Viking variety had a higher percent marketable weight due largely to the. lower percent loss of damaged Spears. The two and three miles per hour Speeds produced Slightly better harvesting percentages than the one mile per hour rate. One difficulty encountered was the speed of the paddle contacting the Spear. The paddle Speed was directly pro- portional to the distance from the center axis of the wheel. Misalignment of the harvester with the row or Spears several inches from the row centerline would cause marketable sized Spears to be contacted by the paddles at a lower velocity and could decrease the snapping ability of the harvester. Results from the remaining Spear damage test, pre— sented in Table 9, Show an average of 67.4 to 85.7 percent 56 TABLE 8. Mechanical aSparagus harvester results for the Mary Washington and Viking varieties at a two- inch platform height. 'Forward Snapped by Damaged Loose on Marketable Speed, Machine, Spears. Ground. Asparagus. mph Percent Percent Percent Percent Viking 1 80.72 6.67 15.63 58.42 2 . 87.04 7.97 16.20 62.87 3 79.99 10.33 10.93 58.73 Average 82.97 8.36 14.34 60.27 Mary Washington 1 80.32 13.89 15.17 51.26 2 . 84.29 15.23 12.26 56.80 3 86.71 13.95 15.17 57.59 Average 84.01 14.41 14.08 55.52 of the spears under observation received no apparent damage when they were bent and passed over by the mechanical har- vester platform. Spears which were Skinned, Slightly Skinned, or bruised had average growth rates of up to 50 percent less than the non-damaged or normal Spears for the first day. How- ever, growth rates during the second day after injury were nearly equal to the normal rate. Spears with broken tips and other damages attained growth rates of 10 to 50 percent of the normal rate (Figure 20). Downes (1962) also observed the growth rate was greatly reduced for broken tipped Spears. ‘3] U1 TABLE 9. Condition of aSparaguS Spears which made contact with the mechanical harvester platform. Average percentages for nine tests on Six rows. (Platform height, incheS—-miles per hour) Test Condition (2—1) (2—2) (2—3) (4—1) (4—2) (4—3) No Apparent Damage 67.4 73.6 76.7 85.7 82.8 81.0 Bruised 0.0 3.6 0.0 1.3 0.0 0.0 Slightly Skinned 13.0 8.2 8.3 3.9 6.4 3.5 Skinned 2.2 0.9 1.7 ' 3.9 0.6 4.9 Other Damage 1/ 1.4 0.9 0.0 0.0 0.0 0.0 Broken Tip 4.4 4.6 5.8 3.9 5.7 7.8 Snapped 11.6 8.2 7.5 1.3 4.5 2.8 Average of ' first four 82.6 86.3 86.7 94.8 89.8 89.4 Average of last three 17.4 13.7 13.3 5.2 10.2 10.6 l/ Spears damaged before the test. The results were averaged into two groups, those having normal or nearly normal growth and those with less than 50 percent normal growth. An average of 82.6 to 94.8 percent of the Spears in a given row were classified in the first group and Should attain marketable height for the next regu— lar harvest. Table 10 shows the no damage percentage to decrease with increased Spear height. The percent of Spears snapped and damaged increased with Spear height (Figure 22). 58 Figure 19. Experimental harvester passing over aSparagus spears too Short for harvest— ing. .- c r \ 7 n J I N \ ' .- . . K A . “ ‘.‘N n-‘ ‘ . ’ ‘. . . -. O \y ‘ _ .- A . 9 - .- . . .»-~ m-n *tg«-. - u t . . - . .~ ._ 1p, ‘ ’4 .v» H.- Figure 20. A non—damaged Spear and a Spear with broken tip showing relative growth after two days. PERCENT OF TOTAL CROP I00 . ' ' A\ & DAMA a SLIGHT DAMAGE 9C)* '-——QE.3_..__..___.____.__13 .— ——o— — — —'0 M NO 9! DAMAGE 1 80* 7‘“ -—O 70 I- 0/07 60 - 50 - O———0 2 INCH PLATFORM HEIGHT 40- A—-—A 4 INCH PLATFORM HEIGHT 30 - 20 t O~\‘ IO .. ~ _AEEZEB-EEQ-EAE:3 Ar’ ’7 0 ' ‘ ‘ l 2 3 FORWARD SPEED , MPH Figure 21. Condition of Spears passed over in relation to forward Speed. PERCENT OF TOTAL CROP ICHD T I r N0 DAMAG R A \ 907 \ \ \ \A 80*- ‘0 70* 60 t 50' 0—0 2 INCH PLATFORM HEIGHT 40- A——-A4 INCH PLATFORM HEIGHT 30' a/"’£> 2C)* ,,r // /=\ EL DpdflhggivC)’ i” IOF SEALER”, / 0" 9 ,” Ab-""“"_'-—'-_' O J 1 L O-I I'2 2'3 INCHES ABOVE PLATFORM BOTTOM Figure 22. Condition of Spears passed over in relation to their height above the platfo rm bottom. 61 Taller Spears must be bent to a greater angle than the shorter Spears to be passed over by the harvester plat— form. Since the bending angle of the taller Spear iS greater. a larger percentage of the Spears may be broken by the platform. Much the same reasoning can be applied to the higher Snapping percentage at the two—inch platform level. Spears two inches above the two-inch level are re— quired to bend through a greater angle than Spears two inches above the four—inch level platform. thus accounting for the difference in snapping rates. TABLE 10. Condition of aSparaguS Spears which made con— tact with the mechanical harvester platform. Average percentages for nine tests at the two platform heights. Two-inch Platform Four-inch Platform Test 2 - 3 3 — 4 4 — 5 4 — 5 5 — 6 6 — 7 Condition in. in. in. in. in. in. No Apparent Damage 85.8 73.6 57.4 91.2 89.9 65.6 Bruised 0.0 1.4 1.7 0.0 0.0 0.8 Slightly Skinned 6.2 8.6 15.7 3.4 2.8 8.4 Skinned 0.0 2.9 1.7 0.7 0.9 7.6 Other Damage l/ 0.0 2.1 0.0 0.0 0.0 0.0 Broken Tip 5.3 4.3 5.2 4.0 3.7 10.9 Snapped 2.7 7.1 18.3 0.7 2.7 6 7 Average of first four 92.0 86.5 . 76.5 95.3 93.6 82.4 Average of last three 8.0 13.5 23.5 4.7 6.4 17 6 1/ Spears damaged before the test. 62 Table 10 indicates the Spears near the height of the platform and those two inches above the platform received the greatest broken tip damages. This can possibly be attri— buted to the direct contact of the platform and paddle with the tip. A higher rate of impact or speed also increases the damage (Table 9). Spears received a much higher damage rate if they were two inches taller than the platform. Some of these Spears made momentary contact with the paddle forcing them more severely against the leading edge of the platform. This additional force was Sufficient to cause skinning, bruising, and snapping to occur more readily. The determination of the angle an aSparagus Spear is bent by the mechanical harvester Should be directly related to the angle measured by the aSparagus snapping device. The angle 0 recorded by the snapping device in Figure 23 has its origin at the center of the test arm driving Shaft 2-1/2 inches above the soil surface. although the Spear is free to bend from the soil surface. The mechanical harvester in Figure 24 bends the Spear over the cushioned leading edge of the platform at a height of about 2—1/2 inches for the two-inch platform height. The Spear is not free to bend from the soil surface until the harvester moves forward and bends the Spear to a greater total angle 81 as in Figure 24b. If the Spear does not snap at this point, it will be bent forward by the platform and either passed over or broken on the ground. ASPARAGUS SPEAR SOIL SURFACE . -* if :5 JJE#!E'E= 2a —4 gang .. T Figure 23. Angular measurement of Spear being bent by the snapping device. 2" f PLATFORM f) CUSHIONED 2T" SOIL , EDGE SURFACE .’ EI—(llf'difl E = __J ('5; [5.5. I ”IIIF-FI - 9 :33.- III =’(7=§m‘§=~fl £4 5' O. Fégure 24. Angular measurement of Spear being bent by the mechanical harvester. 64 The angle 0 can be computed as: _ d (distance from Spear to paddle edge) T 7 t - angen d 2 (paddle clearance from platform) or d = Arc Tangent % All Spears over 4—1/2 inches high will be contacted by the paddle; however. the taller spears will be bent through a greater angle because the distance d will be greater. In comparing the angles 0 and d for Spear snapping. 0 repre~ sents only the test arm angle while d is the actual Spear angle at its bending point. Therefore, if 0 = d in degrees, the actual bending of the Spear by the harvester is equal to or greater than the Spear bending by the snapping device. A greater difference would exist when the harvester moved to the position of Figure 24b. The sequence of pictures in Figure 25 illustrates an aSparagus Spear being snapped by the mechanical harvester. The time t under each picture designates the time elapsed from the first picture of the sequence. Figure 26 illus— trates how the bending of the Spear by the first paddle doesi not snap it. but the Spear straightens up and iS broken by the second paddle. A study of the high Speed motion pictures revealed that many Spears were not snapped by the paddle making the first contact with it. Many of these Spears were pulled under the platform after being bent by the paddles several times. The above events were observed at each of the three forward Speeds. A greater snapping angle between the platform edge and the paddles would enable a greater percentage of the 65 t2 = .012 sec. t3 = 0.022 sec. t5 = 0.050 SEC. Figure 25. Sequence showing an asparagus Spear being snapped by the mechanical harvester. Photo- graphs taken from rear of machine looking forward. 66 t3 = 0.032 SEC. t4 = 0.075 SEC. t5 = 0.090 sec. t6 = 0.105 sec. Figure 26. Sequence Showing Spear being snapped by the second paddle of the mechanical harvester. Photographs taken from rear of machine look- ing forward. Spears to be snapped by the first paddle. Spear damages could also be reduced by the reduced number of paddle con— tacts. It was also noted in the moving pictures that Spears contacted near the periphery of the paddle—wheel snapped better than Spears Contacted nearer the axis of the paddle— wheel where the velocity was lower. CONCLUSIONS The conclusions derived from this study may be Stated as follows: 1. [‘x) An economic analysis. using the equations developed in the thesis. is needed before the design of a feasible mechanical harvester is completed. Ac— cording to the assumptions made in this Study with growing costs of $120 per acre on 20 acres har— vested. a positive net return can be made only if the initial harvester cost does not exceed $2200. Experimental results confirmed the theoretical analysis that the Spear snapping force is pro— portional to the cross—sectional area to the 1.5 power. The theoretical analysis Showing the Spear snap— ping angle is inversely proportional to the aver— age Spear diameter was confirmed for the average of all tests during the Season. Results of in~ dividual tests did not always confirm this re- lationship. The Significant factor affecting the asparagus snapping force was the Spear cross-sectional area. Temperature. relative humidity. time of season. and time of day were statistically insignificant for the entire test results. 68 A O L) The only Significant factor affecting the aspara- gus snapping angle of the snapped Spears was the inverse of the average Spear diameter. Factors of temperature. relative humidity. time of season. and time of day were proven Statistically insig— nificant for the entire teSt results. The harvesting efficiency of marketable Spears was the greatest at the two miles per hour rate for both the two- and four—inch platform settings on the paddle—wheel mechanical harvester. The harvesting efficiency was greater for the two— inch platform height than the four—inch level for each of the forward Speeds. This relationship held for the percent Spears snapped and the per— cent marketable Spears. although the percent damaged Spears and Spears dropped on the ground were greater at the two-inch level. The machine/ potential yield ratio at the two—inch level was 0.570 and was 0.504 at the four—inch level. Spears missed by the machine were hand harvested. A comparison of the snapping ability of the Viking and Mary Washington varieties at a two—inch plat— form height revealed no Significant difference. The percent marketable Spears for Viking was 4.75 percent higher than the Mary Washington. The machine/potential yield ratio was 0.603 for Viking and 0.555 for Mary Washington. 10. 70 The relation of the damage of Spears passed over by the platform to the forward Speed was that damage decreased with increased Speed at the two— inch platform level and increased with Speed at the four—inch platform level. The damage at the two—inch platform level was greater than at the four—inch level for each of the three forward Speeds. The greatest damage occurred to Spears between two and three inches taller than the plat- form level as compared to Shorter Spears. High speed moving pictures revealed the need for a greater snapping angle by the paddle-wheel har— vester to insure a higher Snapping percentage by the first paddle contacting the Spear. [‘L) RECOMMENDATIONS FOR FUTURE STUDY Conduct further tests for the spear physical properties by including the soil moisture content and Soil temper— ature as additional factors affecting the Spear Snapping angle and Snapping force. The test of the irrigated and non-irrigated plots revealed a difference in the Snapping forces which Should be considered in future tests. Consider recording the Spear diameter in the direction of bending as well as the major and minor diameters for the physical property test. This Should provide a more accurate relationship between the Spear Size and the snapping angle and snapping force. Also test the differ— ences incurred in the properties when the Spears are snapped about the major and minor diameters. Spear orientation may yield more predictable results. Use an extension in the test arm driving Shaft of the snapping device so that bending a Spear over it would more nearly Simulate the bending of Spears over a Stationary level platform. A more valid comparison could be made between the snapping angle of the Snapping device and the paddle—wheel harvester. Higher snapping percentages could also be obtained for the physical pro— perty tests and ensure more accurate relationships for snapping angle and snapping force. 71 The paddle—wheel principle of mechanical aSparagus har— vesting merits further study with the following improve— ments: (a) increase the bending angle of the Spears by the harvester paddle—wheel to increase the snapping efficiency, and (b) use some type of paddle action which would yield a constant velocity over the entire row width such as a horizontal axis arrangement with a straight leading edge on the platform to replace the present M— shaped edge. Future Studies with the paddle—wheel harvester should in— clude: (a) tests at the three-inch platform level to provide a better platform height comparison Study. (b) comparing harvesting results with time Of day to deter— mine the effects of damage throughout the day. and (c) tests conducted frequently enough to avoid the necessity of hand harvest between tests. thus comparing the effect of continued machine harvest on the yield for the season. \I SUMMARY The physical properties of aSparagus related to har- vesting by the snapping method and the economic aSpectS of mechanical aSparagus harvesting were studied. Samples of 20 asparagus Spears were tested three times daily for 13 days throughout the harvest season. Basic data for temperature, relative humidity, Spear Size. time of day. time of season. snapping force, and snapping angle were recorded. A snapping device was constructed and used in the field to measure and record the snapping force and snapping angle of each Spear. Theoretical analyses were made for the snapping force and snapping angle which provided the relation of snapping to the Spear cross—sectional Size at the snapping point. Average results for the entire season confirmed the theo— retical analyses with the experimental results. Multiple regression analyses and analyses of variance of the 650 snapped Spears showed the Spear cross sectional size at the snapping point was the only factor considered which Significantly affected the snapping force and snapping angle. An economic analysis of mechanical aSparagus production yielded an equation for net return in terms of growing, har— vesting. and handling costs which can serve as a guide in the production of a profitable harvester. 73 74 A set level. semi-selective. vertical axis paddle— wheel mechanical harvester was used to obtain harvesting re— sults. Platform heights of two and four inches at forward Speeds of one. two. and three miles per hour were used. The most efficient harvesting Speed was two miles per hour. The average percent of Spears snapped at the two- and four—inch platform levels were 82.1 and 68.4 while the average market— able Spears were 57.0 and 50.4 percent, reSpectively. Tests conducted between two varieties showed the Vik— ing yielded a higher percentage of marketable Spears than Mary Washington. although the overall snapping percentages were nearly equal. Tests conducted on selected Spears contacted but passed over by the harvester platform Showed total damage decreased with increased forward Speed at the two—inch plat— form level while the total damage increased with forward Speed at the four—inch platform level. Damages were greater for all Speeds at the two—inch platform level with the taller spears receiving the greatest damage. The percent broken tips also increased with forward Speed. High speed moving pictures taken at 400 frames per second Showed the bending action of Spears by the paddles over the platform. Many Spears were contacted two or three times by the paddles before Snapping. thus indicating a need for a greater bending angle of the harvester. -.“--'I-. H“ x— .- -_I 1.?“ at ‘ Wm _‘m ._ x... fi\/- REFERENCES Apple, S. B. and K. C. Barrons (1945). ASparagus production in Michigan. Mich. Agr. Exp. Sta. Hort. Cir. Bul. 194. 23 pp. Anonymous (1957). Recommended standards of green asparagus for processing. Mich. Can. & Freezers A550. and Mich. Frozen Food Packers A550. 2 pp. Anonymous (1961). Vegetables—Processing, 1961 annual sum— mary of acreage, production, and value of principal commercial crops. USDA Stat. Reporting Serv. Vg. 3~2(61). 28 pp. Bainer. Roy, R. A. Kepner. and E. L. Barger (1955). Principles of Farm Machinery. John Wiley & Sons. Inc. New York. 571 pp. Barrons, K. C. (1945). The field snapping method of harv— esting aSparagus. Mich. Agr. Exp. Sta. Quar. Bul. Vol. 28(2):1ll-ll4. . Blumenfield. D. and others (1961). Field study of asparagus growth. Am.Soc. Hort. Proc. 77:386-392. Carncross. J. W. (1956). What it costs to grow asparagus. N. J. Farm and Garden. 27:30—32. Carncross. J. W. (1961). Production and price trends of asparagus for fresh market and processing. Rutgers U. Dept. of Agr. Econ. A. E. 161. 10 pp. Carolus, R. L. (1948). Reducing aSparagus harvester costs. Section of Horticulture. Mich. State C. 4 pp. Carolus, R. L. (1949). Yield and quality of aSparagus harvested by the field snapping method. Mich. Agr. Exp. Sta. Quar. Bul. 31(4):370-377. Carolus. R. L. and L. D. Brown (1961). We've been over— feeding aSparagus. Am. Veg. Grower. 3:9. 52—53. Dixon, W. J. and F. J. Massey. Jr. (1957). Introduction to Statistical Analysis. 2nd. Ed. McGraw—Hill Co.. Inc., New York. 488 pp. 76 Downes. J. D. (1938). The effect of harvest method on the yields of market value asparagus. Mich. Agr. Exp. Sta. Quar. Bul. 40(3):543-548. Downes, J. D. (1960). 1959 Annual report to the Sodus Horticultural Experiment Station. Mich. State U. 18 pp. Downes, J. D. (1961). Growth characteristics of aSparagus in relation to yields from mechanical harvesting. Mich. Agr. Exp. Sta. Quar. Bul. 44(2):272-28l. Downes. J. D., D. P. Watson. and R. R. Dedolph (1962). Stem elongation of aSparagus. Mich. Agr. Exp. Sta. Quar. Bul. 44:772-780. Ellison, J. H. and L. G. Schermerhorn (1958). Selecting superior asparagus plants on basis of earliness. Am. Soc. Hort. Sci. Proc. 72:353—359. Hanna, G. C. (1950). Harvesting aSparagus-—comparative effects on yield of cutting and snapping Studied. Calif. Agr. 4(5):13—14. Honma. S. and F. J. McArdle (1958). Producing blanched aSparagus. The Canner and Freezer leaflet. 2 pp. Huyskes, J. A. (1959). A method for selecting aSparagus varieties with a long fiberless top. Inst. of Hort. Plant Breeding, Wageningen, Netherlands. 151:21-28. Johnson, W. B. (1960). How to grow better Spears. Am. Veg. Grower. 8(6):9, 31. Kepner, R. A. (1954). Asparagus harvester improved. Cal. Agr. Exp. Cal. Agr. 8(9):3-4. Kepner, R. A. (1957a). ASparagus harvester. United States Patent No. 2,791,878. Kepner, R. A. (1957b). Improved aSparagus harvester. Cal. Agr. Exp. Cal. Agr. ll(12):4—6, 13—14, 20. Kepner, R. A. (1959). Mechanical harvester for green asparagus. Trans. ASAE. 2(1):84-87. 91. Lafferty, J. W. and H. L. Miller (1955). ASparagus har— vester. United States Patent No. 2,709,330. Marihart, L. A. (1954). Asparagus harvesting machine. United States Patent No. 2,690,043. Matteoli, A. J. (1952). ASparagus harvester. United States Patent No. 2,581,119. ‘1 \l Mears, D. R. (1960). Principles of harvesting green aSpara- gus. Progress report. N. J. Agr. Exp. Sta. to the N. J. Asparagus Industry Council. Moore. M. J. (1961). Development of a mechanical harvester for asparagus. Am. Soc. Agr. Engr. Paper No. NA 61—205. 8 pp. Moran, C. H. and R. L. Isaacs (1960). Effect of crown Spac- ing on the yield of aSparagus. Am. Soc. Hort. Sci. Proc. 75:416-418. Rao, C. R. (1952). Advanced Statistical Methods in Biometric Research. John Wiley & Sons. Inc., New York. 390 pp. Schermerhorn. L. M. (1947). IS field snapping of aSparagus practical in New Jersey? Hort. News (N. J.). 28: 1956-1957. Snedecor. G. W. (1946). Statistical Methods. 4th ed. Iowa State College Press. Ames Iowa. 485 pp. Stout. B. A. (1962). Basic principles of harvesting. Re— port for Agr. Exp. Sta., MSU. 13 pp. Stout, B. A. and S. K. Ries (1962). Mechanical asparagus harvesting. Am. Soc. Agr. Engr. Paper No. 62—157. 22 pp. Thompson. R. C. (1954). Asparagus culture. USDA Farmers Bul. No. 1646. 22 pp. Turkington, J. O. (1956). ASparagus harvesting machine. United States Patent No. 2,767,544. ‘ Underhill, J. (1962). California's deltaland turns ”grass" into profits. Am. Veg. Grower. 10:12—13, 55. Ward, W. B. and N. K. Ellis (1948). Asparagus production. Purdue U. Agr. Ext. Bul. 339. 11 pp. Wiebe, J. (1959). ASparagus. Ontario Dept. Agr. C. 5 pp. "11111111111111 IS