HalrtuuwuzllylumlmwW This is to certify that the thesis entitled DEVELOPMENT OF A BUNCHING MECHANISM FOR LEAFY GREEN VEGETABLES presented by Donald Lee Peterson has been accepted towards fulfillment of the requirements for ph.D. ficgreein Agr. Eng. Major professor .Date g,/6,/79 0-7 639 v-” 44‘ -4“. EHHV OVERDUE FINES ARE 25¢ PER DAY PER ITEM Return to Book drop to remove this checkout from your record. DEVELOPMENT OF A BUNCHING MECHANISM FOR LEAFY GREEN VEGETABLES BY Donald Lee Peterson A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Agricultural Engineering 1979 ABSTRACT DEVELOPMENT OF A BUNCHING MECHANISM FOR LEAFY GREEN VEGETABLES BY Donald Lee Peterson A mechanism was developed to form uniform size bunches of leafy green vegetables such as turnip and mustard greens. The mechanically formed and tied bunches were of sufficient quality and composition to be acceptable for fresh market use. Tying efficiency was greater than 97 percent in the final design. A stochastic model was developed to simulate in-row leaf mass distribution based on the mean, standard devia- tion, and autocorrelation coefficient of the field sampled data. Computer models of machine functions were developed to predict machine performance in forming uniform size bunches based on parameters of both actual and simulated data. The models were used to establish the feasibility of machine concepts. ACKNOWLEDGMENTS The author wishes to express his sincere appreciation to the following persons and organizations who contributed to this study: To the U.S. Department of Agriculture, Science and Education Administration, Agricultural Research, for its financial assistance. To Dr. Ajit Srivastava, my major professor, for his advice and guidance throughout my entire graduate program. To Dr. Galen Brown, committee member and research leader, for his advice and assistance in identifying, developing, and testing my research project. To Dr. Bill Dean (Horticulture Department), Dr. George Martin (Mechanical Engineering Department), and Dr. Larry Segerlind (Agricultural Engineering Department), for serving on my guidance committee. To Fred Richey for providing land at the Horticulture Research Farm. To George Hogaboam, USDA sugar beet project, for providing greenhouse space. To Joe and Jim Bihl, green growers, Portmouth, Ohio, for providing advice on production and cultural practices and plant material for testing. ii TABLE OF CONTENTS Page LIST OF TABLES . . . . . . . . . . . . . . . . . . . . . V LIST OF FIGURES. . . . . . . . . . . . . . . . . . . . .Vii Chapter 1. INTRODUCTION . . . . . . . . . . . . . . . . . . . l 2 LITERATURE REVIEW. 3 3. OBJECTIVES 5 4 PRE-DESIGN ANALYSIS. 7 4.1 Machine Concepts 7 4.2 Field Sampling . 9 4.3 Computer Models of Machine Bunching Concepts. . . . . . . . .13 4.4 Stochastic Simulation Model of In- -row Leaf Mass . . . . . . . . . . . .18 4.5 Machine Simulation Using Simulated Data . . . . . . . . . . . . . .26 5. THE INITIAL DESIGN . . . . . . . . . . . . . . . .30 6. TESTING AND MODIFICATIONS. . . . . . . . . . . . .40 6.1 Test Procedures. . . . . . . . . . . . .40 6.2 Initial Design Testing . . . . . . . . . .41 6.3 Second Design. . . . . . . . . . .43 6.4 Testing of the Second Design . . . . . . .46 6.5 Third Design . . . . . . . . . . . . . . .47 6.6 Testing of the Third Design. . . . . . . .50 6.7 Final Design . . . . . . . . . . . . . . .57 6.8 Testing of the Final Design. . . . . . . .61 7. CONCLUSIONS. . . . . . . . . . . . . . . . . . . .68 8. RECOMMENDATIONS. . . . . . . . . . . . . . . . . .69 iii Page LIST OF REFERENCES . . . . . . . . . . . . . . . . . . .70 APPENDIX A DATA. . . . . . . . . . . . . . . . . . . . .72 APPENDIX B COMPUTER PROGRAMS . . . . . . . . . . . . . .79 iv Table 1 2 q o: (n e 10 11 12 13 14 15 16 LIST OF TABLES Summary of Hand Harvested Turnip Greens Bunches. Simulated Machine Formed Bunches for Machine Concept One. . . Simulated Machine Formed Bunches for Machine Concept Two. Values of Parameters for Test of Randomness. Model Parameters Based on Field Observed Data. Simulated In-Row Plant Distributions Simulated Machine Formed Bunches from Simulated In-row Distributions of Turnip Greens. Simulated Machine Formed Bunches from Simulated In-row Distributions of Mustard Greens Performance Data for the Second Design Using Turnip Greens. . . Performance Data for the Third Design Using Turnip Greens. Analysis of Variance of Bunched Turnip Greens for Design Three Test for Homogeneity of Variance for Bunched Turnip Greens for Design Three Performance Data for the Third Design Using Mustard Greens . . Analysis of Variance for Bunched Mustard Greens for Design Three Test for Homogeneity of Variance for Bunched Mustard Greens for Design Three. Test Summary of Machine Bunched Turnip Greens for Final Design Page 10 16 18 21 24 24 27 28 46 51 51 52 55 55 56 62 Table 17 18 A1 A2 A3 A4 A5 A6 Bl B2 BB Analysis of Variance for Final Design Test Test for Homogeneity of Variance for Final Design Test. Hand Harvested Turnip Greens (6-20-78) Leaf Mass per Consecutive 2.54 cm of Row. Field Sampled Leafy Green Vegetables. (6/20-21/1978) . . . . Machine Bunched Turnip Greens, Design Two. Machine Bunched Turnip Greens, Design Three. Machine Bunched Mustard Greens, Design Three Machine Bunched Turnip Greens, Final Design. Simulation Program of Machine Concept One. Simulation Program of Machine Concept Two. Simulation Program of In—Row Plant Mass. vi Page 64 64 72 73 75 76 77 78 79 81 83 LIST OF FIGURES Figure Page 1 Flow Chart for Bunch Forming Steps for Machine Concept One. . . . . . . . . . . 8 2 Bunch Size Probability Distribution Function for Hand Harvested Turnip Greens . . . . . . . ll 3 Mass/Interval Sampling Device. . . . . . . . . . . 12 4 Interval Mass Probability Distribution Function for Field Sampled Turnip and Mustard Greens. . l4 5 Bunch Size Probability Distribution Function for Simulated Machine Bunching of Turnip Greens by Concept One. . . . . . . . . . l7 6 Bunch Size Probability Distribution Function for Simulated Machine Bunching of Turnip Greens by Machine Concept Two. . . . . . . . . 19 7 Interval Mass Probability Distribution Function for Both Field Sampled and Simulated Data. . . 25 8 Bunch Size Probability Distribution Function for Machine Concept One Using Simulated Data . . . 29 9 Schematic Diagram of Initial Design. . . . . . . . 32 10 Schematic Diagram of Rectangular Motion Generator. 34 11 Motion of the Tip of Packing Fingers . . . . . . . 36 12 Leaf Holding Clamp . . . . . . . . . . . . . . . . 41 13 Schematic Diagram of Design Two. . . . . . . . . . 44 14 Design Two: A. Feed Belts; B. Packing Fingers; C. Rectangular Motion Generator; D. Accumu- lating Pocket Disk . . . . . . . . . . . . . . 45 15 Design Three: E. Tying Mechanism; F. Accumulat- ing Pocket "Wheel" . . . . . . . . . . . . . . 49 16 Tying Mechanism: G. Bunch Compressor; H. Bunch Control Rods . . . . . . . . . . . . . . . . . 49 vii Figure 17 18 19 20 21 22 23 Bunch Size Probability Distribution Function for Design Three, Treatment One Using Turnip Greens. Disoriented Bunch Presented to Tying Mechanism . Final Design Prototype Force Sensor and Bunch Positioning Device: I. Packing Fingers; J. Microswitch; K. Extension Spring; L. Accumulating Pocket; M. Roller Chain On-track; N. Transition Rods; O. Camfollower; P. Four Bar Linkage; Q. Cam; R. Bunch Holding Rod. . . . . . . . . . . Tying Mechanism and Bunch Positioning Devices: M. ROIIer Chain On- track; S. Stem Support Plate; T. Knotting Mechanism . Bunching Size Probability Distribution Function for Final Design for Treatment One Typical Mechanically Bunched and Tied Turnip Greens . . . . . . . . . . . . viii Page 54 56 58 58 6O 65 67 1. INTRODUCTION Leafy green vegetables are specialty crops that include turnip greens, mustard greens, collards, and kale. In 1974, the total production of these crops in the United States was grown on more than 12,000 hectares involving 4,000 farms (5).l/ Approximately one-third of the total production was sold through fresh market channels with a farm value of over $12,000,000.g/ Leafy green vegetables for fresh market consumption are hand harvested and field packed (l, 9).g/ No mechanical harvester for these vegetables currently exists. The greens are sold in bunches that ideally have a length of 30 cm and mass of 340 g (1). Hand harvesting usually consists of cutting groups of leaves at the desired length, gathering them into a bunch of desired size, and securing the bunch with either a rubber band or twist tie. In some parts of the United States bunches of loose leaves are packed in cellophane bags (9) or loose packed in bushel containers. However, loose packed leafy greens present a problem for the retailer in displaying and repacking ll Numbers in parentheses refer to appended references. 2/ Personal correspondence with Extension Horticulturists from 15 states, November 1977 to April 1978. the produce for consumers. Current retail preference is for consumer size bunches supplied by the grower. Traditional harvest costs of 2-3 cents per bunch have recently risen to 4 cents per bunch (1). Harvest labor costs are 10-25 percent of the wholesale price at the farm (1). Scarcity of a skilled and reliable hand harvesting labor supply is forcing many growers to consider discontinu- ing production of leafy green vegetables (1). In addition to domestic migrant labor, many growers depend on high- school-age youth to harvest their crop. Harvestable crops are often left in the field after school resumes in the fall. A successful mechanical harvester for leafy green vegetables for fresh market use is needed because of the present high labor requirements per acre, the lack of a reliable labor supply, and the increasing cost of labor. 2. LITERATURE REVIEW" Porter-Way Harvester Mfg. Co.§/ (ll) manufactures a mechanical harvester that is widely used to harvest leafy green vegetables for processing. These harvesters handle the crop in bulk with no precise orientation of the crop. Spinach and other leafy green vegetables for fresh market use are sometimes harvested with a Porter-Way harvester, but are then packed into consumer size packages by hand in a packing shed (1). There is no packing equipment that will orient the leaves to form uniform size bunches and package the bunch (8). The Tawco leaf crop harvester (17) was a single-row— self—propelled machine designed for harvesting leafy green vegetables. It used a pair of inclined parallel belts to gather and elevate the greens as they were cut from the plants. These belts are known as gathering belts. From the gathering belts the leaves were deposited in a bushel basket. These harvesters were not widely accepted by the leafy greens industry and are no longer manufactured. 2/ Trade names are used in this paper solely to provide specific information. Mention of a trade name does not constitute an endorsement of the product by the author to the exclusion of other products not mentioned. Many commercial mechanical harvesters (6, 20) use gathering belts to gather and lift vegetable crops. The plants are either dug or pulled out of the ground, or cut off above the ground with a rotating disk. Similar gather- ing belts and cut-off disks would be adaptable for gathering and cutting leafy green vegetables. K. F. Roth (13) developed a machine to harvest and bundle nursery plants. The plants were mechanically separ- ated into bundles of a certain number. 'This technique would not apply to separating leafy green vegetables into uniform size bunches since they do not grow as discrete plants. Grain and corn binders (19) that harvest and bundle plant material have been used for many years. It was decided that fresh leafy plant material would not stay properly oriented in a grain binder type machine. Fresh plant material would not slide properly into the bundle forming cavity. Since bundle size was determined by weight, machine vibrations would not permit accurate sensing of the relatively light—weight bunch size required for leafy green vegetables. On grain binders there was no positive mechanism for separation of plant material that was inter— twined. It was also decided that the corn binder principles would not provide good separation of intertwined plant material encountered with leafy green vegetables. The chain conveyor system of the corn binders would not handle fresh plant material without damage. 3. OBJECTIVES The need for a harvester for leafy green vegetables for fresh market has been established during the foregoing discussion. The harvester would conceivably harvest and form bunches in the field. Gathering belts have been used to successfully gather vegetables after they are cut and to elevate the crop to bulk bins. These belts have the advantage of maintaining the natural crop orientation. However, no suitable mechanism for forming and securing uniform size bunches was available. The objective of the research reported here was to design, develop, and test a machine which could take the leafy green vegetables from the gathering belts, form them into bunches of uniform size and then secure the bunches. The constraints on the design were that the uniformity of the bunch sizes be comparable to that of hand-harvesting while keeping crop damage to a minimum. The machine was to be adaptable to a field unit. To achieve the objective, the research was conducted in two phases. Phase one utilized field data and simula- tion models to predict the performance of different proposed machine concepts for forming uniform size bunches. Phase two consisted of the design and testing of the bunching machine based on the results of phase one. 4. PRE-DESIGN ANALYSIS 4.1 Machine Concepts Leafy green vegetables are grown in rows. The in-row plant spacing is such that the individual plants are often indistinct except at ground level. Thus the row appears as a continuous mass of leaves. Two machine concepts were envisioned for forming bunches of the desired size. The first concept involved accumulating all leaves growing in sequential intervals of fixed row length until the proper bunch size was obtained. The steps to form bunches would be as follows and are represented by the flow chart in Figure l. 1. Separate a constant interval of leaves from the gather- ing belts. 2. Pack these leaves into an accumulating pocket. 3. Check bunch size by measuring the force developed in the accumulating pocket. If proper size bunch (force) is obtained or exceeded, proceed to step four. If not, repeat steps one through three. 4. Remove bunch from accumulating pocket to provide an empty pocket so that steps one through three can con- tinue without interruption. 5. Secure the bunch. .wco pmooqoo weaned: sow mmoum waHEsom nocsm new usdso Bon .H os=Mam :pmcoH gonzo poxooo ossm gonzo poxoom Boy 90 Hd>sop:H muaon ossoo o>o€om Homo“ :Hpaasezoow pampmcoo Eosm :wHoSpww Scum pmafimwa Moono owefi Roam mo>soH opasdnom mm>qu J1 Li P W HHmEm 00» 50:59 am The second machine concept involved accumulating all leaves growing in a constant length of row, regardless of the in—row leaf mass distribution, to form the bunch. This concept eliminated the need for sensing proper bunch size and would simplify the machine design. It was felt that with field data on in-row leaf mass distribution, computer models could be developed to simulate the effectiveness of these machine concepts in forming uni- form size bunches. These models could also be helpful in determining machine design parameters. 4.2 Field Sampling Information regarding the uniformity of hand formed bunches and the distribution of leaf mass within sequential intervals of row length was collected from a commercial farm. It was decided to collect data on turnip greens for the development of the bunching mechanism due to the limited time available. Data for mustard greens were used to a lesser extent. Four boxes were randomly selected from a truck load of hand harvested turnip greens. Each box was supposed to contain 24 bunches. The data obtained is summarized in Table l. The first two boxes did not contain the proper number of bunches. Their average bunch mass was more than the 340 g desired. The other two boxes contained the proper number of bunches and their average bunch masses were closer to the desired mass than those of the first two boxes. The range in the individual bunch size for 10 Table 1. Summary of Hand Harvested Turnip Greens Bunches. Bunches Average1 Standard Box number per box bunch mass deviation CV (g) (g) 2% l 15 557.7 114.41 20.5 2 18 446.2 136.81 30.7 3 24 327.3 43.93 13.4 4 24 355.4 76.36 21.5 Reference2 48 341.3 63.24 18.5 1Mass was determined by a balance beam scale assuming the acceleration of gravity to be 9.8 m/s . 2Combination of box 3 and 4. the last two boxes was much less than that of the first two boxes as indicated by the smaller values of the standard deviations. Boxes three and four were selected as repre- sentative of hand harvested turnip greens. The data obtained from these boxes would be used to compare the effectiveness of machine bunching. The probability distribution function of bunch size for these two boxes is shown in Figure 2. This figure shows the probability of a bunch's mass falling in any 10 g interval. Even though this was limited data, there was not sufficient time to permit more sampling. The selected data was considered to be a good estimate for typical hand harvesting (1). Field data needed for input to simulation models of the machine concepts described in section 4.1 were leaf mass 11 .mcoosc aficsse cosmo>smm comm 90m qofipocsm :oszanpmHQ mpflawndnosm oNHm noczm .N osswfim Ame mufim noasm own own can con owv owv.ovv omv oo - I D b b V L b b < << v own 0 mm own can con ow - b n n b N omN OWN ONN P P n R o 1/[I\\d 10H. .ON. rom. eoueJJnooo go Aitttqeqoxd 12 per sequential interval of row length. It was decided that the smallest interval length of row that would be practical to separate would be 2.54 cm. Leaf mass per 2.54 cm interval of row length was obtained by using a comb-like device shown in Figure 3. This device had 4 mm diameter rods spaced 2.54 cm on center and was 1.2 m in length. This device was inserted into a row of turnip or mustard greens about 10 cm above the ground, The leaves were then cut approximately 30 cm from the average height of the leaves' top. The mass of each sequential interval was then determined and recorded. Figure 3. Mass/Interval Sampling Device. The selected sampling sites were representative of healthy normal turnip and mustard greens. Sections of the field had been harvested and other sections were very non- uniform and did not permit random sampling. Thirty meters 13 of row length of turnip greens and 7.3 of mustard greens were sampled (Appendix A). The probability distribution functions of leaf mass per 2.54 cm interval of row length are shown in Figure 4. The average leaf mass per interval for turnip greens was 19.46 g. The median was 17.36 which indicated that the majority of the data points were skewed to the left of the mean. The average leaf mass per inter- val for mustard greens was 12.43 g. The median was 8.94 g which showed that for mustard greens the majority of the data points were also skewed to the left of the mean. 4.3 Computer Models of Machine Bunching Concepts Computer models (Appendix B) were developed to simulate the machine functions of both machine concept one and two described in section 4.1. The model for machine concept one was constructed to sequentially accumulate the leaf mass from fixed intervals of row lengths to form bunches based on a minimum acceptable bunch size (MABS). IABC was defined as the interval of row length added to the bunch between comparisons with MABS. For any run of this model, IABC would be a multiple of 2.54 cm. When the accumulated bunch mass was greater than or equal to MABS, the bunch mass was recorded and a new bunch was started. Simulated bunch distributions were obtained for IABC = 2.54, 5.08 and 7.62 cm for MABS = 300 to 320 g by 10 g increments for turnip greens. For mustard greens the MABS range was the same but IABC = 5.08 and 7.62 cm. Input data 14 .waooso panama: was assume noHQEmm naofim sow zofiuoczm :ofipsnflspwfln zpwawnmnosm mmwz Ha>soch .v ossmfim Ase em.m\mv mmmz H3>nmqu memoso csmpmsz .lnam .m. eouexxnooo go AittthQOJd mnoosm Qflqsse 15 of leaf mass per 2.54 cm interval of row length was the field collected data described in section 4.2. Results are summarized in Table 2. Two representative probability distribution functions for bunch size using turnip greens are shown in Figure 5. As expected, the varia- tion in the uniformity of bunch size, as denoted by the standard deviation and the coefficient of variation, increased as IABC increased. However, for all three values of IABC the uniformity of machine simulated bunch size was better than that of the hand harvested bunches. This can be seen in the probability distribution functions of Figure 2 and Figure 5, and by the fact that the coefficient of variation for any simulation run is at most 48 percent of the coeffi- cient of variation for the reference hand harvested bunches. Average bunch size was controlled by adjusting the value] of MABS. This machine concept seemed feasible since uniformity of simulated bunches was more uniform than those harvested by hand. Machine concept two was also simulated using the same field data of leaf mass per 2.54 cm interval of row length for turnip greens. The row was divided into equal length intervals. For each interval the mass of plant material within that interval was added together and considered a bunch. Four runs were simulated using four different values for lengths of row per bunch. This length varied from 35.56 cm to 50.8 cm by an increment of 5.08 cm. 16 Table 2. Simulated Machine Formed Bunches for Machine Concept One. Average Bunch Standard Coefficient Crop IABC MABS Mass Deviation of Variation (cm) (s) (g) (g) (91:) Turnip Greens 2.54 300 313.9 11.78 3.8 5.08 300 324.2 17.22 5.3 7.62 300 336.4 29.43 8.8 2.54 310 327.0 14.24 4.4 5.08 310 340.0 22.37 6.6 7.62 310 353.0 29.32 8.3 2.54 320 334.7 10.32 3.1 5.08 320 347.8 21.67 6.2 7.62 320 360.8 28.64 7.9 Mustard Greens 5.08 300 314.7 10.49 3.3 7.62 300 316.7 19.60 6.2 5.08 310 329.4 17.55 5.3 7.62 310 337.5 21.39 6.3 5.08 320 342.1 13.77 4.0 7.62 320 339.9 18.35 5.4 17 .oco Homoeoo so maoouo assume mo mcwnocsm masses: copsHsefim cow soapocsm aosusnfiepmfio spssfinanoed musm noasm .m mesmfim own own can con oww omv o Ame musm sonsm g omv oov own 0mm own cum oom owm com ovN CNN l p b P Jib szh L, a \\ ’ (\ a V (\./l/ i z . (. fl 4 AH <1, . .N. r Nm.mm m.mmm Cam N®.b III: hm.NN o.owm can mo.m Ill! . m. Amy Amy va AEoV .>wQ .Upm .w>< mm<2 0 1.96 for Z defined as: 2 N N l 2 R ' (N1 + N2 + 1) Z = - [1] 2N1N2(2N1N2 - N1 - N2) 2 (N1 + N2) (N1 + N2 - 1) 21 where: Z = variable with u = 0 and 02 = l u = mean 2 _ . o - var1ance R = the number of runs N1 = number of data points below the median N2 = number of data points above the median To determine the number of runs a plus (+) sign was placed by every datum point above the median and a minus (-) sign when the datum point was below the median. When the data were in the natural order as sampled, a run was counted whenever a sign change was encountered in the sequence. Values of test parameters are given in Table 4. The null hypotheses of randomness was rejected for both turnip and mustard greens. Therefore it was assumed that the sampled data was auto-correlated. The stochastic simula- tion model developed was based on this assumption. A comparison of simulated and actual data would be used to determine the validity of this assumption. Table 4. Values of Parameters for Test of Randomness. Crop Median N N R Z 1 2 Turnip Greens 17.36 581 571 470 -6.31 Mustard Greens _8.94 140 148 127 2.11 22 Llewellyn (7) described a simulation model to generate auto-correlated data having a specified mean, standard deviation, and cumulative distribution function (CDF). Brown (4) used an equivalent model to generate an average daily wind velocity, which he found to be auto-correlated. An analogous model to predict in-row leaf mass distribution is given by: - 2% GMPIi - p + p (GMPIi_1 — u) + 0(l-p ) X1 [2] where: GMPIi = correlated leaf mass for interval i GMPIi__l = correlated leaf mass for interval i-l u = mean of leaf mass per interval p = auto-correlation coefficient 0 = standard deviation of leaf mass per interval X. = standard random variable generated from the CDF for leaf mass per interval of row length Since a standard theoretical distribution for leaf mass per interval of row length did not exist, the empirical distributions obtained from field sampling were used to estimate model parameters. Watson (18) defined the auto-correlation coefficient. For leaf mass per sequential interval of row length the auto-correlation coefficient can be calculated by: N 1 E (GMPIi - ‘fifiT) (GMPI1+1 - EMF?) - i=1 ' [3] p- N ____ 2 2 (GMPI. — GMPI) =1 1 i 23 where: GMPIi leaf mass per interval 1 GMPI1+1 leaf mass per interval 1+1 GMPI mean leaf mass per interval Values of estimated parameters are shown in Table 5. Four simulation runs of 1200 intervals were made using Equation 2 for both turnip and mustard greens. Results are summarized in Table 6. The chi-square test for goodness of fit (15) was used to compare the simulated distributions with the empirical distributions. The data was segregated into class intervals of 5 g size. The equation used to calculate x2 was: 2 x2 = .i‘ “’1 3“ [4] i=1 1 where: 01 = relative frequency of the ith interval of the simulated frequency distribution E. = relative frequency of the ith interval of the empirical frequency distribution N = number of class intervals into which the frequency distribution was divided The null hypothesis was that the simulated and empirical dis- tributions represented the same distribution. The critical value for rejecting this assumption wasx2 = 22.4. None of .05 the values of x2 exceeded this value. Therefore the null hypothesis was not rejected and Equation 2 appeared to be an acceptable model. Figure 7 shows how closely the simulated 24 Table 5. Model Parameters Based on Field Observed Data. Mean Leaf Mass Per Interval Standard Auto-correlation Row Length Deviation Coefficient (g) (g) ‘ p Turnip Greens 19.46 13.36 0.24 Mustard Greens 12.43 12.36 0.27 1Interval Length = 2.54 cm. Table 6. Simulated In-Row Plant Distributions Mean Leaf Mass Per Interval Standard 2 Row Length Deviation X (g) (g1 Turnip Greens Simulation 1 19.95 14.32 7.479 2 19.70 14.11 6.723 3 20.37 14.46 8.229 4 19.84 13.83 6.605 Mustard Greens Simulation 1 12.25 12.53 10.181 2 12.09 12.28 13.533 3 12.71 12.67 11.793 4 12.20 12.04 10.897 1Interval length = 2.54 cm. 25 .mpan oopaaaesm new coHQEmm caoflm zoom sow cofipoasm :oHuannpmHn sowaflndnosm mmw: Hm>sopsa Ase em.m\wv mmez Hm>eopeH vowaasefim mcomso campus: uaowm mucosa csdpmsz coomaaefim mcoonu maesse oHosm mucosa nausea .b madman aouexanooo jo Aitttqeqoxd 26 distributions matched the empirical distributions. 4.5 Machine Simulation Using Simulated Data The simulated in-row leaf mass distributions generated by Equation 2 were used as inputs for the machine bunching simulator for concept one. This was done to determine the effect of longer length of row "runs" on machine bunching performance. Bunches were formed for IABC = 5.08 and 7.62 cm and for MABS = 300 to 320 m by 10 m increments for each set of simulated data. Simulation results are presented in Tables 7 and 8. Representative probability distribution functions for bunch size are shown in Figure 8. All simulated bunch distributions were more uniform than hand harvested bunches as indicated by the coefficients of variation. Machine simulated bunching was comparable for both field and simu- lated input data. All simulation results supported the feasibility of the machine concept. 27 Table 7. Simulated Machine Formed Bunches from Simulated In-Row Distributions of Turnip Greens. Average Coefficient Simulation Bunch Standard of Run IABC MABS Mass Deviation Variation # (cm) (g) (g) (g) (%) l 5.08 300 329.7 18.39 5.6 l 7.62 300 337.2 27.73 8.2 l 5.08 310 334.7 20.10 6.0 l 7.62 310 367.0 28.48 7.8 1 5.08 320 347.1 21.52 6.2 1 7.62 320 356.7 29.58 8.3 2 5.08 300 326.7 17.52 5.4 2 7.62 300 340.0 27.91 8.2 2 5.08 310 339.6 21.06 6.2 2 7.62 310 347.6 26.76 7.7 2 5.08 320 346.7 20.20 5.8 2 7.62 320 353.3 23.67 6.7 3 5.08 300 324.7 17.51 5.4 3 7.62 300 338.5 29.73 8.8 3 5.08 310 333.6 16.87 5.1 3 7.62 310 352.0 28.94 8.2 3 5.08 320 345.9 20.18 5.8 3 7.62 320 363.1 30.73 8.5 4 5.08 300 323.5 17.10 5.3 4 7.62 300 334.8 26.07 7.8 4 5.08 310 339.2 21.33 6.3 4 7.62 310 350.2 27.90 8.0 4 5.08 320 346.1 20.43 5.9 4 7.62 320 353.2 26.00 7.4 28 Table 8. Simulated Machine Formed Bunches from Simulated In-Row Distributions of Mustard Greens. Average Coefficient Simulation Bunch Standard of Run IABC MABS Mass Deviation Variation # (cm) (s) (g) (g) (%) 1 5.08 300 324.2 21.51 6.6 1 7.62 300 324.6 19.12 5.9 1 5.08 310 324.2 13.15 4.1 l 7.62 310 336.7 23.29 6.9 l 5.08 320 340.5 _16.14 4.7 1 7.62 320 344.8 22.52 6.5 2 5.08 300 318.2 15.41 4.8 2 7.62 300 324.4 21.16 6.5 2 5.08 310 325.4 15.04 4.6 2 7.62 310 336.9 22.45 6.7 2 5.08 320 336.9 11.57 3.4 2 7.62 320 349.2 23.47 6.7 3 5.08 300 322.8 19.65 6.1 3 7.62 300 325.5 28.42 8.7 3 5.08 310 328.4 18.18 5.5 3 7.62 310 343.3 25.63 7.5 3 5.08 320 341.7 18.15 5.3 3 7.62 320 352.5 22.92 6.5 4 5.08 300 314.6 12.52 4.0 4 7.62 300 322.2 22.38 7.0 4 5.08 310 327.5 16.96 5.2 4 7.62 310 335.8 24.69 7.4 4 5.08 320 341.4 14.26 4.2 4 7.62 320 345.2 23.09 6.7 29 .san copaasefim wcsmb oeo pmooaoo oaHnomz sow coflpocsm cofiusnwsumwn zufiawanOAQ oNHm soczm .w osswfih Awe musm soasm own own ONm oom owv 00v owv ONv 00v own com ovm ONm oom owN omN owN ONNRJ P p h b b n b b b p P p D n n 1‘ 1':- ' ma.w.n v.wNm OHM wo.m mummhc “Exams: III! / mo.HN ®.Nmm on No.0 mammhu aficsae as a: as 2.3 < a .>mc.eom .m>< mm wuflm = _ socfion made lJvn 4|! oMHQM mo>moH msomsflm mcHxOMQ 1L. yo comm .. muoxooq H kasdefisesoo< woofisu fl MWW \r coflufimcase m mime women... ) llllll Au IIIIIII < l... P whomcflm mqflm0dgilv somnmm meson ) 33> doe 33 length to provide more leaves for the fingers to separate. The length of this interval was adjustable. The desired path of the packing fingers was essentially a rectangular motion as shown in Figure 9. It was felt that minimum leaf damage and cleaner separation would occur if the packing fingers could be inserted perpendicular to the path of the leaves, since contact with the leaf surface would be minimized. The advantage of the rectangular motion was the elimination of relative motion between the leaves and the packing fingers during transport to the accumulating pocket. If the packing fingers were also withdrawn perpen- dicular to the path of leaf travel the accumulated bunch would not be disturbed. No mechanism was found that would satisfactorily produce this rectangular motion. A device was designed that would mechanically change circular motion into a near rectangular motion. The schematic diagram of this device is shown in Figure 10. A scotch-yoke mechanism was used to change rotation into rectilinear translation. The crank drive pin was extended beyond the scotch yoke mechanism to drive the slider bar. The packing fingers were attached to this slider bar. In top view A the packing fingers were in position ready to be inserted into the leaves in the feed belts. The drive pin contacted the extended arm of block 8. Although block B was pinned at point C, it could not rotate since it was held in position by bearing A. Translation of the slider bar was initiated as block B slid on bearing A. 34 an: ill“ ‘1'!" M I l)-i+( fl Crank drive pin ‘—‘_* IHF= {-1 L .:. Side View ul Guide bearings Slider bar Guide bearings . B1 ck D h f 7-‘ \J Block B Scotch—yoke f—\ l A .3.’ Q \ /‘C r“) r. /‘ ”4L r< >3 \J / l x I (N . \ I. _; Top View B l Figure 10. Schematic Diagram of Rectangular Motion Generator. 35 Block D was rotated 90 degrees counterclockwise about E by contact with bearing A and was transported along with the slider bar. Note that guide bearings controlled the path of both the scotch-yoke mechanism and the slider bar. After the crank rotated clockwise through the angle a a position shown in top view B was reached. Block B rotated 90 degrees counterclockwise since it was no longer restricted by bearing A. The extended arm of block B folded out of the path of the crank drive pin. Translation of the slider bar stopped and would not move relative to the scotch-yoke mechanism again until the crank had moved through the angle 6. Then the actions of blocks B and D were reversed to move the packing fingers out of the path of the leaves. Figure 11 shows the motion of the packing fingers. The length of the stroke in the direction of D was selected to be 20.2 cm. It was hoped that this distance would provide space for adequate separation of the leaves. It was felt that 5.7 cm movement in the direction perpendicular to D would allow the packing fingers to sufficiently penetrate the path of the leaves. During the insertion of the packing fingers, only 4.1 mm of side travel in the direction of D would occur. The resulting motion of any point on the pack- ing fingers was very close to a rectangular motion. Flat rectangular plates were used as guides to control the transition of leaves from the feed belts to the accumu- lating pocket. The accumulating pocket functioned as a temporary storage container, where the discrete bundles of 36 SE H.v .msomcflm mafixoam mo age was mo :oHpoz .HH cheese 37 leaves delivered by the packing fingers were formed into bunches. Two accumulating pockets, 180 degrees apart, were cut out of a disk as shown in Figure 9. The cross-sectional area of each was slightly greater than that of the cross- section of the desired size bunch. A force sensing device was used to gauge the size of the accumulated bunch. It was assumed that the mass of the bunch would be directly related to the size. A lever arm was positioned above the disk and was extended across the back of the accumulating pocket. A normally open micro-switch was positioned against the lever and was triggered by the movement of the lever. The triggering force was adjustable by use of a compression spring. The force on the lever increased as more leaves were packed into the accumulating pocket. It was expected that control- ling the triggering force would control the bunch size. The normally open switch was connected to the solenoid of a partial revolution c1utch-brake.3/ This clutch-brake mechanism was of the wrap-spring design and had a collar that allowed 180 degree rotation of its output shaft per solenoid activation. This output shaft was connected to the accumulating pocket disk. This disk was then rotated 180 degrees when the proper size bunch had been accumulated. This rotation removed the bunch from the packing site and provided an empty accumulating pocket so that the bunch 3/ Model CB-6, Warner Electric Brake and Clutch Co., Beloit, WI 53511. 38 forming process could continue. The rotation of the disk was completed when the packing fingers were on their return stroke. This action made possible the formation of bunches without interrupting the incoming flow of leaves. A switch in series with the force detection switch prevented the disk from indexing unless the packing fingers were at the extreme left side of their stroke. There were two basic ways of securing a bunch. One way was to place a band around it and the other was to place the bunch in a bag. Both methods are used in hand harvesting; but bagging to a lesser extent. It was decided that a device to band the bunch would be simpler in design than one that would have to gather the leaves and deposit them in a bag and then seal the bag. An idea was formulated to secure the bunch with a strip of pressure sensitive tape using the principles from a common tape-bag sealer (2). The rotating disk containing the accumulating pockets could be used to provide the drive action, and thereby minimize the additional parts required. A conventional tying mechanism could be used but would require more space and more moving parts than the envisioned tape sealer. Narrow string could possibly cause excessive shear loads on the bundle of leaves. Twist ties would be an acceptable banding device but machines to secure ties around bunches were too expensive to use on an in-field harvester. It was decided to band the bunches with tape. 39 Pressure sensitive tape was positioned around the outside of the disk as shown in Figure 9. The adhesive side of the tape faced outward. The end of the tape adhered to a strip of spring steel. This strip held the tape against the outer surface of the disk and was free to travel con- centric about the disk. The tape was pushed into the accumu- lating pocket and around the bunch as the bundles of leaves were accumulated. Tape was supplied from both the roll of tape and side B by pulling the tape holder toward the pocket. When the desired size bunch was formed, tape from side B was positioned along the front of the bunch as the disk was rotated clockwise as shown. Tape from sides A and B then overlapped with their adhesive sides compressed together to secure the bunch. The tape for the seal was stripped off the tape holder on side B and replaced by the oncoming tape. A knife then cut the tape to the left of the seal. The tape was then in its initial position and was ready for the next bunch. 6. TESTING AND MODIFICATIONS Extensive testing and modifications were required before a satisfactory workable bunching mechanism was developed. 6.1 Test Procedures It was thought that the leaves should be fed to the bunching mechanism as they are distributed in the row in order to have an accurate evaluation of the machine's ability to form uniform bunches. It was important to main- tain the mass distribution and intertwining of leaves. A clamping device was designed to meet these requirements and is shown in Figure 12. This clamp made possible feeding of a continuous row of leafy green vegetables to the bunching mechanism. The clamp was made of two 61 cm long boards that were jointed at one end with a hinge. The boards were spaced 1.5 cm apart by a strip of urethane foam that was attached to one of the boards. To gather a section of a row, the boards were spread apart, slid along the row on either side and then positioned the desired length from the top of the leaves. The free ends of the boards were then clamped and the leaves out under the boards. The distribution and 40 Figure 12. Leaf Holding Clamp. position of the leaves were maintained. Up to 20 clamped sections of leaves were used for each run. Feeding the leaves from successive clamps into the feed belts simulated the leaves coming from the gathering belts of a down—the—row harvester. The clamps were released when all the leaves in each clamp were securely held by the feed belts. 6.2 Initial Design Testing Tests were conducted with feed belt speeds of 0.11 m/s. The leaves from 7.62 cm of row length were packed into the accumulating pocket for each cycle of the packing fingers. The following problems were encountered during the initial testing: 1. Inability of packing fingers to remove the leaves from the feed belts without disorienting the leaves. 42 The holding force from the foam covered feed belts on the midsection of the leaves was too high to allow the packing fingers to remove the leaves without either bending the tops or bottoms of the leaves. Inability of packing fingers to provide adequate separation of the leaves. Inability of unit to maintain the leaves under control and in an upright orientation. The leaves in the accumulating pocket fell to reach a more stable position. A compact bunch could not be maintained. The leaves were thrown out of the accumulating pocket during rotation of the disk due to the action of centrifugal force. Attempts to constrain the leaves along their path of travel only resulted in more drag and disorientation. Failure of the taping unit to secure the bunch. The loosely formed and disoriented bunch inter- fered with the taping mechanisms. Leaf material entered the path of the tape and prevented the tape ends from sticking together. The force to bend the leaves was less than the force required to pull the tape off its storage roll. This action caused the tops and bottoms of the leaves to be bent by the packing fingers instead of being pushed, along with the tape, into the accumulating pocket. 43 Substantial modifications were required to correct these problems. 6.3 SecOnd Design It was decided that many of the problems of the initial design could be corrected by handling the leaves in a more stable position. The complete bunching mechanism was rotated 90 degrees so that the leaves would be lying in a horizontal plane in the feed belts. This position of the leaves could be achieved in a field harvester by twisting the gathering belts 90 degrees after the leaves had been gathered and cut in their normal growing position. The decision was made to initially concentrate efforts on forming uniform size bunches. Development of mechanism to secure bunches was postponed until such time as the bunches were successfully formed. The schematic diagram of the modified design is shown in Figure 13. The prototype is shown in Figure 14. The top feed belt was shortened by 10 cm in addition to being rotated 90 degrees. With this shortening, the leaves would not be held firmly when the packing fingers were inserted in the path of the leaves. Secondary support bars were added along side the feed belts. These bars helped to support the tops and bottoms of the leaves. The lower sides of the accumulating pockets (filling position) were extended to provide support for the leaves as they were transported from the feed belts to the accumulating pocket. The exten- sions are labeled as transition rods in Figure 13. The 44 .039 :mflmon m0 Edpwdwn ospdEonom .MH ossmHm amH> mcsm v! sen psomgzm e r s woos sofipflmcmse mpfiwn swam H. muoxoos \H whomsfim msfiumasssoo< . mcfixodm has whoamsm.l¥ somcfim mcwxodm i0 TLAII sowcom bosom -— —----- —- - _- --‘ ——-_ --- fl-— -— —-‘ _ o L, _ _ cos :oflgflmcdsh IA Jul. J 30H> Q09 45 tops of the accumulating pockets were also extended and bent upward to provide a guide for the leaves entering the pocket. The force sensor of the initial design used a rigid lever to contact the leaves. This lever interfered with the rotation of the bunch and caused leaf damage. The force sensor in this design consisted of a strip of flat spring steel acting against a microswitch. The position of the sensor is shown in Figure 13. The trigger force was set by adjusting distances a and b. It was hoped that the flexible spring steel would cause less damage to the leaves than the rigid lever. Figure 14. Design Two: A. Feed Belts; B. Packing Fingers; C. Rectangular Motion Generator; D. Accumulating Pocket Disk. 46 6.4 Testing of the Second Design Field grown turnip greens available for testing were past their normal harvest date. The leaves were taller and more dense than normal leaves. It was felt that these turnip greens would subject the machine to a more severe test than those at their normal harvest maturity. Test parameters and results are presented in Table 9. Variation in bunch size was greater than that of hand har- vested bunches as indicated by the higher values of the coefficients of variation. Interval of row length accumu- lated per packing finger cycle and feed belt speed had no consistant effect on uniformity of bunch size. Most bunches Table 9. Performance Data for the Second Design Using Turnip Greens. Feed No. of Average Treatment 1 Belt Bunches Bunch Standard Number IABC Speed Formed Size Deviation CV (cm) (m/S) (EL) (93) l 3.81 0.084 15 338.3 109.3 32.3 2 3.81 0.056 22 379.8 116.5 30.7 3 3.81 0.056 15 271.2 75.7 27.9 4 5.08 0.084 16 345.0 113.8 33.0 5 5.08 0.066 10 370.7 122.2 33.0 Hand Harvested 341.3 63.2 18.5 Reference 1IABC = Interval of row length accumulated per packing finger cycle. 47 were in the desired size range. However, for every test there seemed to be one or two bunches that were either very small or very large and increased the standard deviations of the bunch size above that for the hand harvested bunches. Apparently the force sensor for bunch size was not func— tioning consistently. Separation of the leaves from the feed belts was better and with less disorientation than the initial design. Trans— ition from the feed belts to the accumulating pocket had no points where the leaves could become tangled or blocked. It became evident during the testing that it would be difficult to control the bunch sufficiently during packing and movement to allow the previously discussed taping mech— anism to operate successfully. It was unlikely that all the leaves could be kept out of the path of the tape to allow proper sealing. Accumulation of dirt and moisture in field operation would cause more problems. It was decided to investigate an alternative method of securing the bunches. 6.5 Third Design A commercially available string tying mechanismé/ was added to the bunching mechanism to secure the bunches, Figure 15. It was learned that string of different widths could be used if shear damage to the leaves became a problem because of narrow string. Elastic string was also Saxmayer Corporation, Blissfield, MI. Model EM. 48 available which would hold the bunch secure even if shrinkage occurred. The support for the accumulating pockets had to be changed to accommodate the tying mechanism. A "wheel" with eight equally spaced accumulating pockets was designed to position the bunches. Eight pockets, instead of a smaller number of pockets, were used to reduce the centrifugal force on the bunches as they were removed from the forming area. Since the time allowed to remove a bunch from the forming area was determined by the packing finger cycle, centrifugal acceleration on the accumulating pocket could be kept to a minimum by minimizing the distance traveled in a fixed time interval. Eight pockets were the maximum number possible because of space limitations. The accumulating pocket rotated 45 degrees each time a new bunch was formed. Each pocket was positioned 30.5 cm from the center drive shaft to allow space for the tying mechanism. When a bunch had been accumulated the wheel was rotated counter- clockwise to provide an empty accumulating pocket. When the next bunch had been accumulated and removed from the forming area the previous bunch was presented to the tying mechanism. The string was positioned to the inside of the accumulating pocket. As the wheel positioned the bunch, it activated the tying mechanism and the bunch was tied, Figure 16. A spring tensioned lever, positioned by the action of the needle of the tying mechanism, compressed the leaves to form a compact bunch before the tie was made. 49 Figure 15. Design Three: E. Tying Mechanism; F. Accumulating Pocket "Wheel". . . Id:?v.¢, . 1‘2." m4; / , 7 7 , ~ . 1» .3 l' Figure 16. Tying Mechanism: G. Bunch Compressor; H. Bunch Control Rods. 50 Two stationary rods held the leaves in the accumulating pockets during rotation. An additional force sensing switch was added on the opposite side of the accumulating pocket. This sensor was connected in parallel with the previous sensor so that either switch could trigger the indexing sequence. This was done to increase the precision in determining bunch size. 6.6 Testing of the Third Design Leafy green vegetables available for testing were past their normal harvest date, but were the last field grown leafy greens available for the season. Three different treatments were performed using turnip greens. The effect of feed belt speed and interval of row length accumulated per packing finger cycle was studied. Test parameters and results are presented in Table 10. Each treatment had four replications which corresponded to the four rows of a bed. Enough row length was fed to the bunching mechanism to form 25 bunches, i.e. 25 samples per replications. A site in the field was chosen that had a bed of uniform leaves. The analysis of variance (15), Table 11, of this data showed no significant difference between treatment means. The chi-square test for homogeneity of variance (15), Table 12, showed that there was no reason to conclude that the treatment variances were different. These results indicated that the interval of row length accumulated per 51 Table 10. Performance Data for the Third Design Using Turnip Greens. Average Treatment Feed Belt 1 Bunch Standard Number Speed IABC Size Deviation CV (m/S) (cm) (a) (g) (%) 1 0.112 7.62 325.15 68.09 20.9 2 0.056 7.62 350.89 71.39 20.4 3 0.056 3.81 316.63 78.85 24.9 1 Interval of row length accumulated per packing finger cycle. Table 11. Analysis of Variance of Bunched Turnip Greens for Design Three. Source df ss ms F tmt 2 66629.52 31814.76 2.704 N.S. error 9 105881.69 11764.63 sample 288 1489002.l6 total 299 1661513.37 Critical F 5.71 .05(2,9) = 52 Table 12. Test of Homogeneity of Variance of Bunched Turnip Greens for Design Three. df 2 2 l tmt (N-l) Si (N-l) 1n Si (N-l) 1 3 16483.7 29.13 .333 2 3 13746.0 28.59 .333 3 3 5064.0 25.59 .333 2 2 2 2 H0: 01 = 02 = 03 H1: Not all oi equal x2 = 2(N-l) 1n sp2 - 2(N-l) 1n 312 = 1.0459 N.S. Critical X205(2df) = 5.99 No reason to reject HO packing finger cycle had no significant effect on the varia- tion in bunch size. This unexpected result might be ex- plained by two factors. One factor might have been the length of the cut leaves. Since the leaves were taller than their optimum harvest length, only the top 30 cm of the greens were cut for testing. The leaves were often lodged and made accurate gauging of leaf length difficult. Varia- tion in length of cut may have masked differences in bunch variability due to interval of row length accumulated per packing finger cycle. The second factor was the continued inconsistent operation of the force sensors for determining bunch size. However, the bunch size was more uniform than that of the previous design as indicated by the reduction 53 in the coefficients of variation as seen in Tables 9 and 10. The uniformity of bunch size was approaching that of hand harvest turnip greens, This fact can be seen by comparing Figures 2 and 17. Design three was also tested on mustard greens. Two treatments were performed. Test parameters and results are presented in Table 13. The available mustard green field was very non-uniform. Partial rows were selected that contained uniform plant material. Each treatment was run until 75 bunches were formed. Each bunch was considered a replication. The analysis of variance, Table 14, showed no significant difference in treatment means. The test for homogeneity of variance presented in Table 15 showed no significant differences in treatment variance. Bunch size had an acceptable range and uniformity of bunch size was approaching that for hand harvested bunches. The efficiency of securing the bunches was only 50-60 percent. Tying efficiency was defined as the number of bunches securely tied divided by the total number formed and then multiplied by 100. The string was placed around the bunch on the stem end as shown in Figure 16. The stems should have been parallel to the sides of the accumulating pocket and away from the knot forming mechanism. However the stems would become disoriented during the packing and positioning operation and the bunch was often presented to the tying mechanism as shown in Figure 18. For this condition, the needle that placed the string around the 54 .maoosu mwssse wean: 0:0 pcoEucose .oosse :wwmon sow :OHpossm sofiusnwppwwo zpfiafinmnosm onm noesm Amo mnfim socsm .sH mesmfim eoueJJnooo go Aittgqeqogd 55 Table 13. Performance Data for the Third Design Using Mustard Greens. Feed Average Treatment Belt 1 Bunch Standard Number Speed IABC Size Deviation CV (mIS) (cm) (g) (g) 1%) 1 0.056 3.81 315.92 66.66 21.1 2 0.112 7.62 321.28 72.57 22.6 1Interval of row length per packing finger cycle. Table 14. Analysis of Variance of Bunched Mustard Greens for Design Three. Source df ss ms F tmt 1 1077.36 1077.36 .2204 N.S. error 148 728292.64 4887.90 total 149 729370.00 Critical F 5.02 .05(1,193) = 56 Table 15. Test for Homogeneity of Variance for Bunched Mustard Greens for Design Three. df tmt 2 2 1 l 74 4443.19 621.54 .0101 2 74 5267.38 634.13 .0101 Pooled 148 4920.90 2 _ 2 , 2 2 Ho 01 - 02 H1. 01 f 02 2 _ 2 2 _ x - 2(N—l) 1n Sp — £(N-1) 1n Si - 2.515 N.S. Critical x205(1df) = 3.84 No reason to reject H0 Figure 18. Disoriented Bunch Presented to Tying Mechanism. 57 bunch forced loose plant material into the knot forming mechanism. This plant material would prevent a secure knot from being tied. It was discovered that there were several ways in which the leaves could become disoriented. Some even fell out of the accumulating pocket. They could be displaced side to side in the accumulating pocket during repeated cycles of the packing fingers since they were only supported by the middle third of their length. Disorientation could occur during the indexing of the wheel due to centrifugal force, interference with the force sensors, and the resis- tance of the leaves to slide against the overhead bunch control rods. Nearly all the bunches could be securely tied if the stems were gathered together by hand and held up and slightly forward of the accumulating pocket as it was positioned' to the tying mechanism. When this was done, tying efficiency approached 100 percent. ,6.7 Final Design Modifications were incorporated into design three to improve the tying efficiency and reduce the variation in bunch size. The prototype is shown in Figure 19. It was observed that the bunch size sensors were not functioning properly. Their location seemed to be interfer- ing with the movement of the bunches out of the accumulating position. To avoid this, the sensing device was moved to the packing fingers as shown in Figure 20. This change 58 Figure 19. Final Design Prototype. Figure 20. Force Sensor and Bunch Positioning Device: 1. Packing Fingers; J. Microswitch; K. Extension Spring; L. Accumulating Pocket; M. Roller Chain On— track; N. Transition Rods; O. Camfollower; P. Four Bar Linkage; Q. Cam; R. Bunch Holding Rod. 59 cleared the path for bunch movement and eliminated the possibility of plant material falling into and interfering with the working members of the sensors. The packing fingers were rigid levers that could sense the force over the full face of the accumulating pocket. The longer moment arms of the packing fingers were sensitive to changes in packing force. It was felt that the extension spring could be more accurately adjusted for setting minimum triggering force than was possible with the flat spring steel. Changes made to hold and control the bunch during accumu— lation and positioning are shown in Figures 20 and 21. The accumulating pocket width was extended on the side opposite to where the ties were made. This extension would give more support to the top of the leaves. A roller chain with an extended support plate located across from each accumulating pocket was added to index in sequence with the accumulating pockets. This chain and support plate acted as an extension to the accumulating pocket to provide support for the stem end of the leaves. A space was left between the support plate and the accumulating pocket to allow string from the tying mechanism to be positioned around the bunch. In the bunch accumulating position, the support plate was parallel to the transition rods of the accumulating pocket. This position helped to keep the leaves in proper orienta- tion during the packing process. A track was designed to support and guide the roller chain during indexing. The chain and support plate were positioned by the track in 60 such a way that the stems would be held away from the knot forming mechanism as shown in Figure 21. Figure 21. Tying Mechanism and Bunch Positioning Devices; M. Roller Chain On-track; S. Stem Support Plate; T. Knotting Mechanism. A cam actuated rod was designed to hold the leaves in the accumulating pocket during indexing of the bunch. A cam follower, acting on a cam through a four bar linkage, rotated the inside transition rod 90 degrees counterclockwise as the accumulating pocket wheel rotated two degrees from the bunch forming position. This rod was rotated to firmly hold the leaves in the pocket until after the bunch was tied. Then the cam would end and the transition rod would be returned by a spring to its initial position. The bunch compressor and overhead guide rods of design three were no longer needed. 61 6.8 Testing of the Final Design Modifications for this design were completed in January 1979. Turnip greens were grown in a greenhouse for testing. Leaf mass distribution in the row was similar to field grown turnip greens. Turnip greens were at their proper stage of maturity when harvested. The test parameters selected and results obtained are presented in Table 16. Only treatments one, two, and three were used to evaluate uniformity of bunch size. Treatments four and five were reruns of the leaves from the first two treatments. They no longer had the proper leaf distribu- tion to simulate in-row conditions. However treatments four and five were valuable for determining tying efficiency. Feed belt speed was 0.11 m/s for all runs. The interval of row length accumulated per packing finger cycle was 7.62 cm. The force to trigger the packing finger switch was set at three different levels. Triggering force was measured at the tip of the packing fingers using a spring scale. The data from treatments one, two, and three were tested for normality using the Kolmogorov-Smirnov test for goodness of fit (14). The calculated value for the test statistics for treatments one, two, and three were 0.0804, 0.1687, and 0.3154, respectively. Critical values of the test statistics for treatments one, two, and three were 0.234, 0.391, and 0.454, respectively, for a = 0.05. Therefore, it was concluded that there was no reason to 62 m oHnmoflHmmc #09 m .d.av H :59 mo assomm vocsooos #0: m .s.cN N can mo essom .oosow waflsowwflsp m0 mHo>oH money no poowmo map cosmqeoo 0:8 soapsnfispmfic peafia scales acomosmms mans omen» zaso ooefim m can .N .H mess 0p whomom .paosowmfic hausmowmficwfim won one whoppofi beam spas mecca cam H oumefipmm powwo>sdm m.wH vm.m® m.va we .w.s comm ooH powwozfiom w.vH mm.wv v.wmm m.o~ 0H .m.: on vm ofipmcam m.NH mm.ov v.m®m m.0H SH v.s.= mv ooH sopmozaom o.HH mm.Hm sv.wwm m.m w N.n.c m Hm ospmcam ®.oH vm.mm Ham.mHm m.oH Ha m.> N cos nmpmmssom s.ms em.sv o.mem w.oH mm m.mm H s s ”so awe musm sz A.o:v Ase secoHOHmmm Hassopwz cofiwafism> coflucs>on nocsm bosom dossom cam m0 900832 mcwme wcflue pcofiowwmooo csmocmwm owmso>< Howmflse monocsm cameo; psoEuwosB .cmeoQ awash MOM mucosa chsse cosossm ossnocz so musessm “mos .®H wands 63 assume that the distributions were not normal. The analysis of variance for treatments one, two, and three is shown in Table 17. The F test indicated a significant difference in at least some of the treatment means. The least significant difference (LSD) method (15) was used to compare treatment means. The LSD (.05) for comparing treatments one to two, one to three, and two to three were 30, 34, and 40, respectively. There was a significant difference between treatment one and two, and one and three. This indicated that average bunch size could be controlled by adjusting the triggering force. The lack of significant difference between two and three may be due to the small sample size. The test for homogeneity of variance, Table 18, showed that there was no reason to conclude that the treatment variances were different. The uniformity of bunch size was better than that of both previous designs and that of the hand harvested bunches. Coefficients of variation in bunch size was under 15 percent for all treatments which was less than the 18.5 percent for the hand harvested turnip greens. The probability distribu- tion function of bunch size for treatment one is shown in Figure 22. It was felt that the re-designed sensors for determining bunch size was the reason for the increase in this uniformity. The inconsistency in bunch sizing was no longer evident. The mechanisms to control the leaves during bunch formation and positioning were effective. Tying efficiencies 64 Table 17. Analysis of Variance for Final Design Test. Source df ss ms F Treatments 2 28121.2 14060.6 7.57* Error 48 89108.1 1856.4 Total 50 ll7229.3 Cr1t1cal F.05(2,48) = 5.35 Table 18. Test for Homogeneity of Variance for Final Design Test. df 2 2 Treatment (N-l) Si (N-l) 1n Si 1 31 2288.7 239.8 2 10 1111.6 70.1 3 7 1001.7 48.4 Pooled 48 1856.4 2 _ 2 _ 2 Ho 01 ’ O2 ‘ 03 H : Not all 012 are equal x2 = 2.96 Critical X205(2df) = 5.99 No reason to reject Ho' 65 .ocC peoEpmmsB sow swfiwon awash MOM aoflpocsh :ofiusnfispmwn Npflafinmnoum oNHm coasm va mNHm socsm CCm Cvm CNm Com va com 03 CNv Cow Cwm Cmm Cvm CNm CCm CwN CCN CvN CNN b I P P D D D D P p v E .mm mesmsm D L 1‘ eouegxnooo go Aggttqeqoxd 66 for all runs were over 90 percent. When polyester string (SAX-TIE #833) was the tying material all the bunches were securely tied. It was expected that the tying efficiency for longer runs would not stay at 100 percent, but would approach it. The tying efficiency for elastic string (1 mm dia., spring constant k = 0.165 N/cm) was 93 percent and could probably be increased by adjusting the tying mechanism. The bunches mechanically formed and tied were of sufficient quality and composition to be acceptable for fresh market use, Figure 23. Leaf damage did not appear to be a problem. The stem-end of the leaves in the bunches were not always the same distance from the string. This variation seemed to be caused by improper placement of the leaves on the feed belts rather than disorientation by the bunching mechanism. All phases of the design functioned satisfactorily. 67 Figure 23. Typical Mechanically Bunched and Tied Turnip Greens. 7. CONCLUSIONS This study has led to the following findings: It is possible to mechanically separate intertwined leaves of mustard and turnip greens into discrete bundles. It is possible to mechanically form bunches of mustard and turnip greens by using packing force to determine bunch size. Uniformity of bunch size is as good as uniformity of hand harvesting. Bunches can be mechanically secured by tying without damaging delicate leaf material. A stochastic model can be used to simulate in—row leaf- mass distribution based on parameters computed from field data. A computer model of the bunching machine can be developed and used to predict its performance based on parameters computed from simulated or actual field data. A computer model can be used to determine functional machine design parameters. 68 8. RECOMMENDATIONS The following research should be undertaken to fully utilize the knowledge gained in this study. 1. Develop a complete mechanical harvester for fresh market leafy green vegetables using the bunching mechanism developed. Study the parameters of other leafy green vegetables such as collard greens and kale to determine bunching feasibility using the mechanism developed in this study. Explore the possibility of using the bunching mechanism on other crops such as green onions and flowers. 69 LI ST OF REFERENCES 10. 11. 12. 13. LIST OF REFERENCES Bihl, Joe, Portmouth, OH, Personal correspondence and interviews, November 1977-February 1979. Better Packaging Inc., Shelton, CT, Nike Bag Sealer. Bradly, James V. Distribution Free Statistical Tests. Prentice Hall Inc., 1968, pp. 260-263. Brown, Galen Kent. Mechanical Harvest System Simula- tion and Design. Ph.D. thesis, Michigan State University, East Lansing, MI, 1972. University Microfilms, Ann Arbor, MI (DISS. ABSTR. 73-12682). Census of Agriculture, U.S. Department of Commerce, Bureau of the Census, Washington, DC, 1974. FMC Corporation, Inc., Scott red beet and carrot combine, Jonesboro, AR. Llewellyn, Robert W. Fordypj an Industrial Dynamic Simulator. North Carolina State University, Box 5353, Raleigh, NC. 1965. 150 p. Marvel, Mason E., Extension Vegetable Specialist, University of Florida, Gainesville, FL, Personal correspondence, November 30, 1977. Moon, W. F., Personal correspondence, Zellwin Farms Co., Zellwood, FL, November 15, 1978. Naylor, T. H., Brlintfy, J. L., Burdick, D. S. and Chu, K. Computer Simulation Techniques. John Wiley & Sons, Inc., New York, 1968. Porter-Way Harvester Mfg. Co. Inc., RD 2, Waterloo, NY 31165. Rohrbach, Roger P., Brazee, Ross D., and Barre, Henry J., Evaluating precision planting based on Monte Carlo planter model, Transactions of ASAE, Vol. 14, No. 6, 1971, pp. 1146-1149. Roth, Karl Friedrick, Apparatus for harvesting and bundling plants, U.S. Patent #4037666, July 26, 1977. 70 14. 15. 16. 17. 18. 19. 20. 71 Sokal, Robert R. and Rohlf, James F. Biometry. W. H. Freeman and Company, San Francisco, CA, 1969, 776 p. Steel, Robert G. D. and Torrie, James H. Principles and Procedures of Statistics. McGraw-Hill Book Co., Inc., New York, 1960, 481 p. Siegel, Sidney. Nonparametric Statistics for Behavioral Sciences. McGraw—Hill Book Co., Inc., New York, 1956, pp. 52-60. Tawco Products, Inc., 1224 Chesapeake Ave., Columbus, OH. Watson, Geoffrey S., Selected Topics in Statistical Theory, Notes on lectures given at 1971 MAA Summer Seminar. Williams College, Williamston, MA, 1971, 167 p. Whitely, W. N., and Bayleg, W., Harvester and binder, U.S. Patent #357,971, January 1888. Wilde Manufacturing Inc., Rhubarb harvester, Bailey, MI. APPENDICES APPENDIX A DATA 72 Table A1. Hand Harvested Turnip Greens (6-20—78). Mass/Bunch (g) Box 1 Box 2 Box 3 Box 4 597 413 329 420 472 343 327 370 409 250 307 545 463 500 360 435 563 502 325 267 540 616 264 489 740 333 396 274 645 466 366 372 682 580 335 455 724 340 279 318 500 489 419 334 658 810 308 318 463 510 327 397 542 339 346 310 362 356 330 371 522 336 400 312 307 300 351 393 303 353 378 330 372 275 290 285 275 222 229 335 307 73 Table A2. Leaf Mass per Consecutive 2.54 cm of Row. Field Sampled Leafy Green Vegetables. (6/20-21/1978)1/. Row 1 Turnip Greens (g) Row 2 Turnip Greens (g) 07 08 27 48 43 43 ll 00 09 20 41 61 13 23 17 19 27 20 17 51 14 11 10 00 06 08 51 00 18 ll 19 00 03 12 17 28 31 13 18 17 41 39 35 16 66 14 20 30 36 00 11 10 20 34 05 14 05 l3 17 04 22 33 14 12 29 17 09 00 00 17 07 18 32 20 29 17 33 27 19 25 59 22 03 06 25 63 10 13 26 30 39 11 10 06 08 19 42 l3 17 17 00 14 12 10 21 19 41 26 10 00 16 28 23 00 24 18 43 27 ll 23 10 34 23 13 30 29 14 26 35 15 28 06 19 15 31 18 ll 18 49 31 31 25 22 34 51 13 00 01 15 18 38 03 09 23 47 03 42 20 16 03 21 12 37 00 31 15 21 09 19 28 35 26 19 31 21 06 33 18 22 ll 02 16 08 16 08 37 21 50 22 04 17 00 44 10 09 02 02 00 13 09 62 15 16 38 38 42 24 00 12 09 00 00 00 00 13 17 21 40 07 24 31 19 26 00 03 09 21 00 31 16 20 18 48 27 12 24 11 21 09 00 39 53 22 08 27 05 19 22 25 14 29 20 06 32 10 05 30 20 13 00 43 09 22 00 69 14 07 62 40 51 14 24 00 09 30 00 14 00 06 24 ll 18 17 12 19 46 16 13 39 14 28 08 26 06 08 19 32 19 20 17 26 ll 08 25 42 30 37 14 44 22 02 18 47 12 50 27 10 22 13 03 10 31 33 30 22 03 00 23 28 32 48 35 58 09 24 23 18 08 30 32 62 25 00 10 23 01 47 16 14 10 14 37 35 22 37 16 16 24 00 20 22 16 30 36 42 07 15 l4 18 47 06 27 54 16 O7 03 24 23 35 40 27 17 00 21 17 33 00 00 41 44 15 00 15 37 40 10 14 15 16 36 ll 10 08 08 26 08 17 00 31 11 37 38 38 14 30 28 33 50 00 19 08 34 34 00 17 42 07 06 39 17 14 03 34 37 04 38 00 21 28 15 06 20 03 14 76 60 31 15 00 18 00 34 22 28 14 15 13 22 26 20 08 07 15 21 40 22 06 15 07 26 15 17 12 12 20 09 46 14 34 25 41 18 27 12 42 15 35 08 17 20 33 23 15 23 16 29 30 04 59 15 25 21 18 25 04 00 35 14 ll 00 29 07 21 03 65 06 23 22 32 15 14 ll 05 15 02 20 13 19 24 16 19 05 00 13 13 36 20 46 43 23 14 15 21 06 00 39 36 03 ll 03 20 09 29 11 31 19 14 18 17 24 00 25 26 00 09 03 05 09 32 22 33 39 26 21 21 00 46 13 28 00 14 00 09 42 29 26 26 22 32 19 ll 00 14 14 22 00 19 24 00 39 17 08 39 25 20 14 19 05 06 18 04 06 46 28 04 13 32 00 15 37 07 27 12 00 00 20 14 15 30 12 14 16 46 39 36 34 44 29 19 00 00 09 15 07 31 21 03 07 41 31 33 39 23 17 ll 10 27 24 08 08 53 ll 12 25 30 23 31 16 22 17 20 47 22 10 02 15 19 07 05 05 45 09 20 00 30 14 19 19 20 00 00 10 11 15 09 00 42 09 30 32 20 15 18 12 00 ll 24 08 27 07 17 00 25 38 05 13 32 12 22 45 00 02 36 02 19 15 27 06 19 39 24 08 50 23 33 42 00 12 00 00 00 25 13 00 27 l5 14 25 31 29 14 35 00 10 00 10 16 00 00 10 13 12 39 25 19 08 20 08 Read data vertically. 74 Table A2. (Continued) Row 3 Turnip Greens (g) Row 4 Mustard Greens (g) 00 11 03 27 08 13 15 00 07 00 17 05 26 ll 16 14 08 25 19 38 16 ll 46 18 18 07 05 07 35 10 34 28 37 26 ll 35 31 14 05 13 12 07 07 23 30 17 23 28 08 18 46 03 06 06 00 05 31 16 34 33 ll 09 06 07 43 00 08 09 01 00 49 ll 23 31 28 16 05 23 17 00 22 22 16 00 37 34 34 16 32 07 05 08 24 00 10 12 02 18 26 45 l7 19 07 31 30 01 07 08 08 00 07 32 41 09 08 16 36 27 06 12 02 21 l4 13 09 20 48 23 00 ll 19 ll 14 07 08 ll 07 20 28 05 14 16 l4 14 10 37 08 08 10 00 00 10 21 03 ll 19 12 47 21 00 27 00 00 56 00 00 05 00 18 30 30 ll 16 34 12 00 02 12 00 31 10 ll 19 15 30 09 26 24 13 00 10 00 00 26 00 16 21 21 25 l7 17 41 14 00 10 12 06 47 00 21 28 36 24 08 18 02 13 07 21 31 13 11 13 22 O9 22 12 00 24 31 09 10 17 22 06 12 04 28 19 35 17 02 16 27 03 07 03 06 22 13 10 21 06 38 00 13 26 08 02 12 34 00 05 00 06 04 15 35 18 20 31 22 12 10 22 20 ll 00 03 00 18 50 10 08 23 30 16 31 19 31 06 00 04 00 16 24 03 08 24 15 25 21 37 22 15 00 05 00 41 21 00 16 10 17 31 23 00 30 22 00 02 06 17 ll 13 00 23 15 17 38 10 20 38 22 26 07 00 06 08 40 ll 17 11 22 00 ll 20 07 16 27 35 08 16 00 ll 25 12 30 00 00 05 08 05 32 29 26 13 28 42 21 14 ll 00 00 19 22 14 09 13 30 07 24 23 23 07 34 21 50 30 08 05 02 37 28 14 20 26 41 26 20 38 10 24 27 31 ll 25 25 33 45 29 05 12 00 17 06 10 12 09 ll 08 14 44 21 10 18 22 00 34 00 16 00 00 00 02 35 30 09 12 30 43 00 22 07 22 09 05 00 00 15 12 03 36 37 47 12 25 12 16 07 04 00 11 14 20 15 46 07 00 07 00 13 07 06 37 22 08 ll 36 37 13 23 09 18 10 26 00 10 30 25 32 18 27 21 16 24 00 00 00 18 ll 07 00 16 15 17 O6 23 21 15 00 04 16 10 04 08 00 07 15 07 18 43 14 21 12 16 14 06 03 00 08 10 14 00 20 23 32 l3 l6 14 09 37 00 00 05 05 22 52 24 33 18 08 52 18 06 20 00 48 00 00 32 ll 16 18 21 15 14 34 47 00 09 07 03 09 36 31 34 15 18 31 06 20 ll 00 25 21 58 07 33 25 33 46 09 18 19 13 00 00 43 09 48 16 00 11 30 17 25 45 30 09 08 00 00 00 28 33 20 01 24 12 14 21 15 03 00 18 19 06 00 22 13 10 52 02 22 32 05 14 34 06 38 08 00 07 30 27 36 ll 14 24 00 09 10 17 08 00 00 00 38 13 09 15 07 4O 00 12 26 12 10 OO 05 03 75 Table A3. Machine Bunched Turnip Greens, Design Two. Mass/Bunch (g) Row 1 Row 2 Row 3 Row 4 Row 5 350 545 344 375 328 353 432 370 355 407 324 589 222 205 269 458 313 224 450 301 228 241 287 315 699 228 350 317 185 339 544 338 197 355 350 268 321 219 310 331 359 539 428 340 384 311 388 166 275 299 286 226 274 340 167 273 158 215 531 246 286 400 262 215 307 311 406 448 269 425 437 665 425 337 627 331 381 354 Table A4. 76 Machine Bunched Turnip Greens, Design Three. Mass/Bunch (g) Replication No. 1 408 420 465 315 354 390 470 342 333 440 543 318 392 416 313 334 274 362 228 260 315 302 266 290 312 2 250 345 328 312 375 342 296 323 282 283 252 324 256 340 245 307 270 233 362 318 263 200 345 245 207 3 536 332 334 292 348 310 327 364 260 286 456 286 260 285 240 272 206 365 505 357 266 348 344 342 333 4 320 330 330 423 242 270 320 387 288 336 330 441 395 310 367 326 258 218 290 210 308 318 316 363 380 Treatment No. 2 Replication No. 1 2 3 4 303 232 308 267 354 420 414 283 400 275 256 436 385 297 303 326 411 293 360 377 427 306 343 280 204 Replication No. 1 2 3 4 424 480 365 333 360 272 387 384 190 378 453 412 460 397 277 597 306 305 437 383 310 375 593 293 317 340 337 437 384 425 418 231 333 379 297 337 353 443 352 353 282 297 321 308 367 348 420 345 414 260 306 284 384 362 235 326 421 317 353 297 374 332 276 266 332 268 352 273 395 277 520 393 407 338 366 278 227 287 308 170 338 332 203 400 396 265 256 374 394 277 234 251 328 475 350 267 353 255 294 348 280 178 267 433 405 213 457 458 370 318 307 307 318 283 310 185 298 304 277 287 327 279 266 377 277 304 402 374 682 285 380 305 307 302 200 350 548 216 343 410 321 246 285 636 234 319 298 263 456 228 249 277 247 249 440 173 260 322 283 286 226 288 226 232 347 275 289 354 413 323 341 363 367 300 471 77 Table A5. Machine Bunched Mustard Greens, Design Three. Mass/Bunch (g) Treatment No. l 2 Replication No. Replication No. 1 2 3 l 2 3 275 338 373 280 324 303 411 253 366 308 227 305 213 333 333 340 335 361 280 232 400 344 197 228 327 231 176 437 333 312 330 222 453 340 382 460 183 472 263 333 378 347 327 279 247 413 296 287 350 353 371 211 337 540 191 320 291 292 288 357 261 282 267 370 346 257 397 363 347 338 293 334 217 240 350 436 315 366 347 377 365 330 218 360 398 434 393 446 194 315 392 290 312 182 268 404 407 396 286 284 380 275 309 327 280 222 310 295 377 405 352 312 255 300 302 373 266 367 336 272 278 303 271 232 263 283 311 316 243 515 317 203 268 265 241 435 262 324 308 396 381 307 443 380 333 293 182 370 291 196 78 Table A6. Machine Bunched Turnip Greens, Final Design. Mass/Bunch (g) Run 1 Run 2 Run 3 Run 4 Run 5 388 316 279 398 318 271 355 273 310 316 316 327 363 330 276 307 338 264 355 279 332 319 279 342 338 324 283 293 395 414 352 264 288 389 392 356 283 268 290 280 282 287 480 366 307 327 370 305 397 372 333 353 382 359 430 311 370 442 338 320 315 365 351 371 309 366 365 433 271 375 452 364 385 395 393 342 275 310 APPENDIX B COMPUTER MODELS 79 Simulation Program of Machine Concept One. Table Bl. E C’NCEFT ONE USING MACHIN HERED Uri; Tor. rsccr. or] xv? IN. (G \I I rh F) \l 7. .Cfln I 21.x H2 3.21. C(\ 7.1: ND '74..” U. I I5 c N)vr1.) ILA ,- ..?s ,1“; (G I antenazlls r 3.1.... (x rcpt... V. C2D I..." D 9 ob RIVA“: .E FDGFDR 1 1.4.: r 3. q 1 UV 7. I .t I Pi. Pal . . .L w! In 3 a... \I I : )C r) 2 CH. 5 to rut IF S C. E 3 = I.” 7. 7. .h = 6 .16 I I C C7. U. 71. .. 5 fl. .lu 9.. I U her 5 n. C, U Fat 5 Irv. o 7. U. 2 F. r: .. 1+5 ) G ) GS 7.. :55. r = K .7 II P. 16L. 0 4| 0 \I . II o D U .13.”... .517. o .E A\C saucy G.P.<.I\sr.\4 0” . (~56 N 7 9. :CV? :u.IY. .1. .( :11; o O :66 p.1§?..—.GQ ‘v.r..»..>.+ 9.1. I 4 hit} D.(Iv.(:P..uthl.t\ 7| rL.L:M .. =rtwnf: :17.- ".337 A FDIGVV..DGIIN.I..GGINI M R I 2 Infarct-1.5.3. 11.12.10 I 7;: .(229 13531755“qu rr Cr. AN! STIVDARD DEVIATIGK OF BLNCV MASS EAR (' F 7. t i \l \I \I G 1 V - A 2 E N. P 1| \I "I. I I 5 Id (at. .‘Vr 9N \IQIU «(P/2 ILS ~M7.K (EI\ IGVH I EDT 41¢ Uni. Pénh =2.'\s :n r100 1" :1 oGMS UG P(U= 5<.VKJ = 2 (v2 5. .. AU AA :V 5&7..P9.1Tl~‘. .Prr ILLND .bUHrEEHT O‘CGDFDSC. CJI up ..:1 flap-Jo tarwutI? ”to.“ s 5.1;...- ..AL§..7. G§(5)=SEGG(0)=SEGG(7)=SEG¢(9) C.) (32 = a. )( 6 C It 1. .OF. GS C... c.) t) a . t) a . r, a t) q r) r. :4 . 1 3 L C 7 O A. 2 7. (a )4! In L. In In I! I! 5 rd 5 r; 7.‘ I I I I I I I I I I (G r; r p: ' . r; . p: r S ..... r06 . . 2 7a (a O C. 9 4| 2 In GE In I. In In In Id In 5 ca 51 ES ‘1 \l \I \I \l \I \I \l \I \l O S : S1 51 54! ’5‘- 51. 5.. Sal 51 54! S) 8) a. 9 29 30 I49 59 6+ 7.9 to 9+ fr. )0 3) 3) 3) 1..) 3) 3) 9:) 3) 3) Idol .34. SI .2 .3 .1. era 3H. .7. or». .0, .I\ ((2[( E‘ E‘ El\ El\ E‘ El\ E‘ E‘ E8 GGlnLru LG luau Lav [:0 LG LG LG Lao LG CG I .G .3 .G .G .G .5 .6 .G .5 .F. .CEAI).C )6 )C. )6. )2; )3. )5 )5. \IE )5 (:3 21:35.193. 189.18. 0.33. .169: ISRsIsc\ISA:I-. : ..1.I\:S..I\ 8.ILI\:P.(\ :5 It :i (351‘ =9.(=th\8 .P‘) xi) c.)r)P)SE)SF)5E)5£2)F\I5F)SD.)SBD 10191 P2 P3 94 Ct; to P7. PG. P9 F1 ((S w (Curlxncu (Pin (guy (Cu (a .v (G... (nus ((Wl‘ (abtvruGTGBTGGTGGIGGTruruvirvfcleGvIGGTcGG Gru (G (c (G (G (G ((0 (G (G (G (S riffyrricrUrrEfiirEr FEr...Vr:. FF: FEerCLLuF.tnUFE $96....15615515GISGI.DGISGIC.GIS!¢ISGIS 1 . ...t105..rar.rzr.5(51 rah.5n..5n..t/auso.. 5a 505.3 Q a ....4. ca?.27«34.b556£778.50 96. .11253336 7 .«luln‘lu‘lh‘lniubLh.“r~l~lfl.~l~‘5ta55t§<.ris C PISTTGRAV POINTS res auwcu rsss r 80 (Continued) Table Bl. I 9 3 Cu 9 4F F G i \I \I I i 9 ol 0.. S o c u. 6 or If i F I I G 4. «I. F. o G L .— r I T. I 2 N 0 o 74 = 8 C F; F r vb I D D i \l I O z 1.. V VI 4 e. H 8 H c F 7.. G N I I N U c .w I F. = F. F C .h .II C A E n. N P. p 7. U c, T P .C :- A ) I) p: 1 VII? I F 5 .rt (0 t. 5 pr. r 5 .L L C 5' (.4. 4| EGM 7.67: )9 Q va 20 KEV... S) \I oIH‘.CVnK IQSL 44. .C DRFGEA \1 (I. :41 4| oTFIDD 17.8 ol\ (\ ID "FIN (YA Ere G ITGEruA 2 (XE LG 6 )L)V\I1| 4| G)U or: E 7.E~.LoF.CS I FOLK \ICJ C. .3ch .. 4. $4IL. (1..., = S .C .C =4 :6 C ()5.).[ It I. Ii 10) I. .C‘ .L Shoot/PH. L(b.(.t(.|). 210‘ UUU P1. 4|.fu(ql(v-(vln 9V.I\I\T Nun.“ I U (r ('4'..4~ It It‘loon 4| ilk" Ii \III'J 7,601.67 TN TM Tr : 6:074 371.7: (C. GquFKIRvARZ Irv": orv~u3 C (LFC; .tfl .nrtrherFrLV pl YC.P nthturt' 1. Ni (3130.3! 4FkaFXh FPJLFFCCFCJL. 4. (fat). r7. :3. .C:~.C . «(Li (C P .5. ... 3.3.3.51»! S:>.r-. 71.) a T1131 4.27 4... (Jr)? star ..t 3. :25. A306 ..C ICC/Cit .C/CIth 1.. 8]. Simulation Program of Machine Concept Two. Table B2. IKG F‘CHINE CONCEPT 745 4- \ ~44 C 4. PEEK agTUAL FIELD 0.1. U h Tuaazr BUHCPIVS C C C I. c = : It In \.I 1 7 T N 3 .b : F It :7. M. Y. 4 = R. v.» a. It : C .. I» 3 b 1 C F. .Ku V r = = h 2. C 4| 4... S V.» 1. I = C 9' ./L "u .U 4| = F .N. r\ N : :Q I 4.. 7.. = 1 r... f. R M J C. = —. = F C 9....” 4H 4.. 33 T N 7.4 8 F... I 9 PM” 4| N. = = U : ....4 F in 7:: Y. K1 31M .0 = FD. as 9. Fr" 4\ = 2 I : ...\ r. c. 1. T. I. 11 4- 4.. C, 9.. 7 4| 5 9. 7. 7. U) N 7.4 C 7. 7 AJ 4... .1. 4| 1. 7. 5. 4| 1 1. or?» = FD. 5 .4. O C 4| I 4| 4. 4| 4| 4| I I 1 N.) \I 5 MV ...c I I I I 3 I I I I I 7 4| I 1.1) G N ) cc: 2 I 7 1. 5 O C 7 1 5 c. I. 7.. 3 5 (1. 412 a G 3 34.. 7.. 9. n7 0. O. 1 D. 4! 1 1 7. 1 4| 7. I\ 4 (s I k I (C 8 \I \l \I \l \l 1 4.. 4| 4| 4| \1 \l 1 R 2 IO .N ( 47....» ) C C C 0 O \I ) ) \4 ) C r... \1 E1 . LP1 3 I n. F PSC K. S c 5 5 54.. SC so .30 .3: 5 5 so .3 HrFr.MI3 F 5"“? O. 4.. 14 7.4 3. L4 54 .04 74 9.4 0,4 4.4 4... 2 C». 164.. v... N M. 4,... =7. . 9:... 1 E 7. E 3.4. 1...: 1...: 7 Fr. 1...». 3.... 7...: 3:... 1»... LE 5 NG‘- 1 8 I5 .u.n.E .L .L .L .l. .L .L .L .L .l. .L .L .L .L . U75: 2 1.4 )909... 94 F4 94 P4 9.4 94 0.4 84 54 P4 9.4 54 P4 9. 94.41.411.92 .l\ ..B. K274? 4 n F . D_ I D ...P r.D. .. F . C. {F 4 C. 1.).P 1. p 4... P 4... p 14 D. n}... f z er1 4.50.". I." 4 6V. . 4.0. 60... to". .04... 4...! éMr 6F N6».-46M {46M .46F. 46!. w I G... 4. 1M 4 EV u.62.1169116H.115111GZ11G7... 1G:«.4c1GC).4I1G.O4u4. 64.16941C9461.GOI.1G1515 AST.JI\.C 4.4/.5 Gc.ca»u:8:?_4 :9.4 :34 31.4 =7a4 :tJ4 H.434 =1.4 :34 :12! :12! 82.1. :‘1 8144! .- 3|... f. :43} 3447.4 . r..- .f4!r\. C!..[ f—flufvj.rfi... IDA... . rq(.... 42.3... can... Pu.“ r4 |.. r~f. Dufr ‘Nfi. 654 5...... D 147 F4 4 IA 4 e VITO u 0 .T. w r T1M \T.. w |.T3V.NT&U hT.....PNT6r. N17? :1... w =TC w. :T. w. :14." :T? WUNQ In 1‘14. P411.I|.(7 all... 7.4... 7‘: 7.1.: 7.(-. 1.(: 1.1.3111: 1.1-: 3‘. .. «((1 ?.I\.C ‘(93 on“ 1“ FIL: 5.. : .P : :. w- = : F FFF‘... r F.( . P73r1FFLrLPr5Cc F65: F7[CF.6[FFOCFF1CFF1I.$F10.» $10P¥1OP D. r.....r 00.5.3.6? Murray." 119M ”1.0. r. v 4... .UN. 4......Gv.14owru.w. In... G" 1.\C..M IVGMIWGNIV.GKIRGM.IV.(UN 1L.GM I...GVHIO\G~ 7.7.4.).r5 . 730116.380 1.21.13. I7 4. C r 1. 7.4.5.0799: 17.7715??? PC 0.1.57.4..107fizv r.15.7.L(.L\7 9.91.1234ra67 4 5: 5777- 777....» .CI 4. .r .54 r90 CCCQOO 0.93. f p. . . . 4...1114-11114..12222222222337.3314; 144....I4.,44.4..4414.444-I 4:4.14444. 14144401144 82 (Continued) Table B2. LUNCH fr. ) ) \I \I ) \I \l ) \l ‘1 \I ) ) u) \I \l E .C ,C F 6 t ,6 t .C .C O I... .0 .5 O C Lu P I I Y. 1 7; I Y. I. I I I I I 1 I I I I I I I I I I I I I I I I I I r. I I I I I I I I I I I I I I I I r. c. S C. S 9.. S Q. ea 5 S S S S S S S .9. I I I I I v. I I I I I I I I I I of.” O O O 0 O O O I O O I O O O O O S S 5 Cu 5 S S to S e. S S S S S S S I N. F In 0 M M. M. V w. M VI. 9. w, M- V Mn 0 G G G G r... G G G C. .u G G to G G G C a. .v L 5 r. F k ., .. . P. r. f. p D. ,.I . cc 9 n. a! 7. 3 Id 5 5. 7 ,3 AV 0 cl 2 3 S 7. 2 3 3 1. 3 95 3 95 7a 3 3 I“ In 1% k w = : t = : = = = = : = : = : : z A. T. T II II T 7. .I II II T T T II T. .I II F N N N N N N N N N Mn N N K In N N G Y. 1 I .I. I I I .1. I I I I I T. T. I «L C P. U f n‘. . a... r n. F _. a... rt nu... nt 0 I F P F F. F P F p. F D. or F or P .r P I D D D D h. n... D. D D D D D D D D D .1. I I 1.. I I I I I I I I I I I I I o w: II. N. V. M1 I.“ v... v r. N. M V. H U: I» v: A». .5. P. .r, H H H. H H N h h .H H H “r H H I T T In T T. T- I. T I! 7| .l T T T .l T I I I I Ya I V. v . T; I I I Y. 7‘ T. I 7.4 k k H uh N .u h .h h H U up H .n U h S . I C. .5 S S I“ Q. (a C. 0. S f. Cu 0. C. C. t. C. E E E E r: E at C. E r: E r: p: E I, SF P H H. H .h H H P H H H F H H. H H .C C c C C C 'L C 1.. C C C C C C C C 1 .CN N N N N u.‘ N N N N N N N N N N. N re I H... U H» U U U U U ”L U U U U U U U U 1 PF. .c E F P. 'b D. E D... P B B P. .H. a P. B I my C CR F F or F F. t. .7 tr or F. .0. rt F t. t. F 1.. It. a... r; C 0 n. rL r.. n. r. AL r0 0 h. n... r: 0 a. Ev?) .C \I 3 I 5 F F. R R R D c. on FOF1RERBFLRSR6F C M. SM (1.5/5.3:)» ESEfE? E: E9 F1E4.E1E1F1E1E1E . GSC ...NHCN9N9NENENDDN:.h.~.ND..NENa.NR-.NENPNENR, r. )AP. K..\l.l..\IM )M)M.\IM \l.\ \l" )¥)'\I.\‘.\IM )N‘I" )Mn)M)V L (MI 2CU305U3UA. UCU...U..U....USUQUTUOUOUCUGU . .c I )6N3NFN2N6N5N9..wCK../.NLfibre.» N..N7.NLN¢N Inap12vn9,23337.2.lulnlu6‘5555 2.054.- II o a II II II II II II It (I II II II It II II It It... 2.5.4 C4 ¢(.5 C.l.l\6( (( C(I‘wl‘lule,‘ C(.¢I\ C(6l‘ b(6(‘vl\ 6(61‘ U .Mll 4|<~!\7.Fv.G‘T‘T(TI\TI\TI\TI\TI\1I(TI\TI\TI\T(T(TI\TI\T~ r. o..(..~1 L... I: of a. Ill—21.52: In...3.:6...£r paIQu... QIAKrILCII: 1451.11 3.3!: ._T.. IL..ITNT.$T~,T» T» T» I.» 7n 7r Tu. .Iw Tu TV TV TNTPTF Ix...) CFVAP..\.rrllvlplrkY.PIPIFTID.1FIPIRI9MIO IO.IP.J;EIPIO‘N¢ h; F1F1FF. u...” .9? P..R.UD\eURrU§ iv: CFO‘DC n.2rn.r-.nv€.u.t (firtufiruruflzCTw. 1. CV». C nrr GIuIrruuF bop. IFIF,vrifnvpkurrovrdrt.b'rvrnvrreoFUrrr...\.t_ I ;.C.,..\1. .o a.2fi\ .bs\ cairn. n“ \fl.. -.,.,~p.nv%.o.r.. rv.;.fl.\. A..I1ap‘.ua « (5 «A- a 31.761. 679 C 15,07 F0 .1??Jlu.)t 757.7. 4.21.‘.) C79 0 . s12?.l~SI.. 0 Yet. 1.. 4 a. a a 1- . 2a..,.C3L27=:7..7d1..5‘.7.7 7 $5.15!»44LL4“L.>StJrJStaK/r;tzt 83 Simulation Program of In-Row Plant Mass. Table BS. I’LLATE (' u D STFIEUTIOE AND Twit A F GREEPS I s: 1 FOP FUh(dING TL AT? IK-FUV FLPNT-‘5SS 09¢! T ‘C I \I by .0 . .; 4| I r. ( a, N \I‘. 9.. to ulru o 6 U5 V try». a on TC. I/ I L U I C 7 L r...\) In 0 I v =6 .7) I p hal n3 I F C. ...(s o 9‘ W T I» v [V r9 .PA ‘6 6 o r3 I III. on! I S r I C. p I I .5 I C. In) I o \I 0 UC 1.1 G F1 7 I I Q. IuI\ .1 r“ s IJP c I n I. :G .1 r. w :J; I I o 5 ES 574 1. f S P I 5.! ( rU on ‘\J o o I! In TC :3 R N 4| . I G n. ..r t 6 S I C Info rat) I II N luv: u ) tr .. .J c S 0A 9!. I O L T I Al I )F. V «v c U) .c: r.» o r: : a; . .9 H; (r) (9 PF... PU V .- : ru .\ a. o be Ire = .r. O . 5. 7. ll 2.. . In .7 V) E D \I 1 V 7. Du. Ill! 9 I V IIPI F T G \I I F. I 'U Ul\ 4'14 - .I n l\ .H r. V \I .. fl» .. I DY. .7. F F); D : A 4.. at T h T NF I. I 7.1: N G I. o 3 S \l A 7." lac E 0.". A V or N I : )C I (G 8 I L (16 T A \l H I.\ .. 2 rs? V I of. L‘) (I Y. I r0 I S n.» P1 I10. 8 5. E F) 10 an LIC F l\ .AZ F. .5 .. (M- Z 2 D Fly - o 0 In AD I .I.. an IN )1" H. 3 5 I16 I I S Ill. 6,: R I "pp: N G" Ply ILU c C7. M. C I r.. B a. c Cl\ C I 47.. . 4 P §.GL A :U v.1“... (C...) N‘ C. I .U "v r) B (a R ND 57 . . N = U I‘ . 2N 5w. N ID( U P3. 5 Ire . 2 M 2 A IL 9.. 07. C. I PN)DI) V .3. I IU I FIT E w... : :IE ) G \I D 9....)7. ow..14un.ca .1 I IN‘GI A U2 gisfifixrqla rss .. :56 K = K N r.L .. 5C..: a. :4- r? .L‘T...‘ C. S: v : : .ruv Q r! 7.7. r 15... o 1 o )r. 4. o A v 10’ . 514}? . )n k. 1:FI).I. w... :1 U511 (US 0 Uni-1n. “t . E12 0 .F. T- a..a.(h. ILFF =. . 1?. I AuIJNLVAnr 4 SUI : :S = I .. no. (v 62.5(354 NA.SSG S r..( .....wW.L: (a I 3:..(w K at.. ..l7&:..u\ I 7... :ul :wlz ofi.(:I.L . 21(9):] BUG Lrv. 1: ar 1CnKlIG I UnlIalIlalleoL/Lt r... ‘7... .~.H.¢7:.—.GI Ilu‘c.bo II D hitv ., GGAFBPF L ..TP( v. c. NP: LLPD I .5 NT F(1.I(:FEII( N Pink .=V}.l,v-.:VoU U .L-N=wF Q =(=UN:EEPT v. .._r-:.\ ::r».F=:F./.Nr.=xf I Fulrvfu . Grrrur..rlrru I. Hi DELI» ‘51 f t Dr‘sehufflpspx I. hlvuv Aruf M huanvJUuINGGIIV I a! 1 Y; T- V r. S a. c1r.7.5 h p: 1231a.) #7:: p r 7. L 5:). 73.345; 4 ".3556 A D «In... 14 Ivauau P 5;: 7.3175474334441415]. t. F v V.C.r\ ((IL r.cc '9. .IA. -GMPEIVG)**2 PCIHTS FOR PUNCH vags SSVS .. 852 .3: hC fifi : V .1‘7vE2TTbc- I P LLN. D r, UVIHICEUYO "(inhuohss r. a 5. .I ‘dfd‘ao 55677- n C; y . .1 _. .. 2:: fiISTUGPAW c C IL SEGGK8)= (5)=SEGG(6)=SEGG(7) 84 (Continued) Table BB. I I = 8 \I av I: l\ V, Gfs A E o I! Q). 6 z: )) I PI... ) \I ' (III 9 i 4| 6‘ v. S o ECL C v. I. SE ‘ II. 3.3 O R I \I: I G I 7) 2 F (S o ru 61 I“ O EI‘ F I T $5 I 2 N =E I 0 Y. )5 .. 9v «L 6: E pf P (\I K I D 5;“ C I \I 1 E1 R .- 1. F 5( 7| 0 :6 t. H F. H )F. .1. tr T I (as G N. I I F Is: N U t V In G) I B e 6 E3 K ..r 54. C F F. C rr 2( or E rL tn C )G p F .1 U Int. 9, a 5 . c; .. '3 ~ 5 .. .3 T F K (S a. S .. 7. L A. 7 C . .4 7. c. .C F tr ) 5 G: .1 7. . r.. (. S I .u I. 1h In In In. 5 5 I) t. 5 E I 7...? R E) 7 7 I.) Iv .P C, 1 I I I I I I I I I I I F S V (C to $2 I I 7. 7 7 7 E ta r. u (a .\ 5 I I: IJ S S E F. 0 :1 I\ r. I I I I I . 3. 7. S. 6 8 9 AI .2. In S l. A 0 GV. cl )I\ 4| at rv fa he (I c L In 1h I! :5 In In (J I. 54: .31 4| from .CGT s 936 7 9| 4. S 7 8 o \I \I \I \I \l \l \l \l \l )4 \l§ 9 SVV 2t NE] 1 (E \l \I 7 7 7| 7 Q. :20. ran. '31 51 S1 t..1 S1 Sal ()1 S) (I) \I oAAEVp [.31. H GS r6 .. \l \I \I \I \l 1.9 2+ 1:. In+ 5+ (I 7+ 9+ 90 IL; 11 2 ZRPGEA )II E: o a C S C .3 C 3) I.) 1.) 3) 3) I.) 3) 3) 3) the! ‘1. 4| QTPADD 12.8 h 5) GI! .34. 11 1.1 21 21 31 oo| 0...; .3 a)“ ’3 .6 .7 :- 09 0‘ oI.\ l\ 5 "RIM. (VIA o :1“ 0.9 .9 0‘ 0* 0+ 0+ 0* .t( E( E( El. E‘ Elx El! [1‘ E( EG ES 5 ITGESA 2 678 )1UG) E) E) f) E) t.) E) .Lrn LIV Liv Irv L.“ LIV LIv Law Law Lru LG 6 )L)V\Ivl 4| G)rU Q. ~I.I\~rt4| Lfl. L74 L1. L5 L6 L? .G .G .G .5 .5 .6 .6 .G .6 2t 3t E 7E2A6$ I Etna 7| (G I.‘ .l\ -( o‘ ol\ .1\ alt \IE )5. \IE \IE )E )5. \IE \IE )E )S )S S 880 C II. 51:». MN GE2)G )G )6 )G )G )6 )6 :5. 15. 1.5, 15:15. 1.5515 IS. IS??... 1:.\: 5 5 6 :4. : t 1.. 55:11:: If IF(I[;IE..IE.,Y.E (:c.(:...(=:.(:P(:.€(=S(=: (=:.( :c ()0 ()9 )E II I. It 1+) It Eris a S=1(c.t7\chr.(Sr)(<.ra(¢.c.(5:..l\$ .r)5c.)r.B)SP)SE)5F)r)..:)ta:):,E)Sc.r.55155sU((C(6(A\ 21.91.. UUU F :) I:9]=9]=91:O 1:91:91: P1 f2 F3 F5 9:) F6 7? Par P0. Pal F1 1N(T(T(TQ9Y((T FNN )Ca F) F) F) 9) D) I.) F) "(P .o (its (1 V(. .v (7v (raw (1 v 1" uric” (.|M(fl_(1p..L.r.L..‘.. 4v ,Ft.l)1.;1 C. 4.1.. V, a. DVq.n.v1ntw.lufikh.Car\M 60.V9. CGTr:CT Cinleru?(JGTGGYCGTGGTGGTGGIGGTGT?VT» Tu ..Iu..ruvlw 71-1.7 N ((OG(TG(T GI\.|G(TG(T.GI\TG( (G (G (F0 (6 (G (G (G (G (G (G (G Gv.InNVARIINz ZGIQ QKNN I 6.“ (G (G (G (G (G (G (G FECF.tr FEr FE..I.E.UF..F.£E r.€_r.FEr.‘E., FE! t...\Fr.LF.UC3YOYEnK3:c.-CC v 55. FECFEL 7.51.}??? FETFEFLFE ISFISF1.096136JIJC.ISI..ISCISI,.V;SCISGI<.GSC.IFH‘VFYDXSHFFCCC R SSDv‘porquSGISrJIIsGISGISGVJS .1 E 4| rat/.45.. (a. S. £72535. .5. :5... can...) ”t, .2! (fur)... SHEIJrafi. L246 32L65 C .3C Tu 3.37515..S,L5.\‘S..S..S.\5.rcans .. n.11?.¢147..4~1~55€b~r7 3.?908...1.1 2.L7.7.6LSPI6¢ECL)LV. h...‘ 14 111 13:14 E . 014 «£21.3Inlqr.ta66.l78£.nw0\.u LL L LLIH£1~8L.~INI~L .hihlu.~.«l.L:P. > n C (.0 U C C 9C C h. CC 7. . $~ F \I ' 7. t . : ‘. \I I. p... i I o S 9 «L V 8 on .F C L f. G A V T Dr V t F. I I T nu 2 N P o 1 h. 8 I ..r p. M- I E i \I F. .M = 3 To L o s I A .r‘ V k v F or F I p L I. O G A T .. V “I r . p I D. p. E 0 CL T 9 I Z N E. T I F A I. I r! F S V n... to E V! A D 1| V. GM )1 )7. V P. a! .L 21 ‘:.VP (1 (L IGCI-ALCSNPVA F‘DDo/GA/GG. VRTN):.B)EA rnptSAAl cur- «(Curt \IV)TI\\/R()U 2&1561FG1R S C E? ‘.u.rr S D C 50. SC EG(16)*1 ( .....(=rn ...». I? It : IO = I9 ...)!» v4)o )Uél.‘ F (\I «(1‘).0.‘ I.N(.I(Yl1l\olflsl\vltatb61{vl w III V. .‘1IF. I .fb('. ..(c.‘| OOY‘...O-\lo Af‘vpl C(Trur‘qll TTN 7.3 f T? t TN 20 XFTI 7... vov T r! 0N]? VHF“: roN 10,1? 61.9 6.19 : ..f. ‘- '4‘ =. £2.55» C7 3. 8R 0 Q00 60C C75 '1. «L Q Q 23,1b C. ) VJ ( Pr GI .LL ‘8'. ..I (H. \Ivlnh TCAB (\I.U G7 on rich? c. O ." RTFKVFL kITH PIDFGIKT OF',“.7 0 J F D Y 'I' o: It 5) h.‘ : GI? I. C, c (4 C o A; 0 Q 097 009 “I , rt ‘1 3‘ run EN 5 NUV5F9*,12) r I.‘ r. Uf!.L “(TN 0 o ... A . .v I ‘P.. .....4. I . I PU"EY-SV&LL,‘.C),FL05T(K)*DIFF) -VAL(I))‘(DU“'FLOAT(I‘1)* vaL,s~ALL,r1rr,x,ouvrv) (.)I\ K)TA 5:. = .. .LFU Par r. 1(EFTD Uvot .. F‘T..LN Frftlvtvlrffi... a!