t [It lillll llllllllllllllllllll 3' 3 1293 o 58 3118 , “BRA! Y Hickman 9 rate University This is to certify that the thesis entitled Impact Testing of Navy Bean Pods presented by Nor Mariah Adam has been accepted towards fulfillment of the requirements for M.S. degree in Agricultural Engr. W Major professor Thomas H. Burkhard 0-7639 MSUis an Affirmative Action/Equal Opportunity Institution MSU LIBRARIES m RETURNING MATERIALS: Place in book drop to remove this checkout from your record. FINES will be charged if book is returned after the date stamped below. “Jr-'2‘] K’M’I./'-\. “’5‘" .1. . A . ‘1 r v ‘ t H ‘ ‘ '5 E {in L; ‘ l . “"'l‘. 2:5“5325 IMPACT TESTING OF NAVY BEAN PODS by Nor Mariah Adam A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Agricultural Engineering 1985 iCDL/ «*9 do 333' Nor Mariah Adam Approved :%W4 é/W Major Professor Approved : 0 /(?((-<<./' Department Chairman ACKNOWLEDGEMENTS The author wishes to express sincere gratitude to the following: Dr Thomas H. Burkhardt (Agricultural Engineering) for his advise, encouragement and suggestions as major adviser; Dr. George Mase (Metallurgy, Mechanics and Materials Science) and Dr. James D. Kelly (Crop and Soil Science) for their time, constructive criticism and moral support; Mr. Gary Connor and Pedro Herrara for helping with the construction of equipment, and Abdul Razak Habib for helping with the experimental design, statistical analysis and computer work; Staff and students in the Department of Agricultural Engineering for their helpful support in the course of this study; Last but not least, her parents, Encik Adam Abdul Aziz and Puan Azizun Dahalan for their unfailing love, encouragement and faith in their daughter's pursuit for knowledge. ii TABLE OF CONTENTS LIST OF TABLES LIST OF FIGURES 1. INTRODUCTION .........................................1 1.1 Background .......................................1 1.2 Bean Plant .......................................3 1.3 Anatomy of Bean Pod ..............................4 1.4 Shatter ..........................................6 1.5. Objectives .......................................7 REVIEW OF LITERATURE .................................8 2.1 Some Parameters Related to Seed and Grain Damage .8 2.2 Previous Research Work on Navy Beans in Michigan.10 2.3 Measurement of Mechanical Properties of Biological Materials .......................................13 2.4 Instrumentation Methods .........................15 2.5 Summary .........................................16 THEORETICAL CONSIDERATIONS AND EXPERIMENTAL DESIGN ..18 3.1 Theoretical Considerations ......................18 3.1.2 Impact Energy ..............................18 3.1.2.1 Without Specimen .........................20 iii 3.1.2.2 With Specimen (With Impact) ..............21 3.2 Impact Action ...................................25 3.3 Experimental Design .............................26 EQUIPMENT ...........................................29 4.1 Pendulum ........................................29 4.2 Bean Pod Holders ................................29 4.3 Electrical Release Mechanism ....................32 4.4 Angular Transducer ..............................32 4.5 Force Transducer ................................34 .4.5.1 Strain Gauge ...............................34 4.5.2 Temperature Compensation ...................36 4.5.3 Wheatstone Bridge ..........................37 EXPERIMENTAL PROCEDURE ..............................42 5.1 Force Transducer ................................42 5.2' Angular Transducer ..............................50 5.3 Air Resistance ..................................51 5.4 Data Collection .................................55 5.4.1 Beans IOOOOOOOOOOOO0.0.0.00.0...OOOOOOOOOOOOSS 5.4.2 Pod Impaction ..............................56 5.5 Computation of Calculated Values ................61 5.5.1 Impulse ....................................61 5.5.2 Maximum Impact Force .......................64 5.5.3 Absorbed Energy ............................64 5.6 Accuracy of Measurements ........................65 RESULTS AND DISCUSSIONS OOOOOOOOOOOOOOOOOOOO0.00.00.070 6.1 Resu1ts 0......OOOOOOOOOOOOOO0.00.00.00.00000000072 iv 6.2 Impulse .........................................75 6.3 Maximum Impact Force ............................82 6.4 Absorbed Energy .................................88 6.5 Summary .........................................96 7. CONCLUSIONS AND SUGGESTIONS .........................98 7.1 Conclusions .....................................98 7.2 Suggestions .....................................99 BIBLIOGRAPHY O...0....O...OO...OOOOOOOOOOOOOOOOOOOOOOOO.101 6.8 6.9 6.10 LI ST OF TABLES A 2x3x3 Factorial Combination Table ..............28 Determination of Angular Speed Using Photography (32 frames/s) and Angular Transducer ..................52 Experimental Layout For A Single Year .............58 Moisture Content For Beans and Pods (wet basis) for All samples 0....O00.0.0.0...00.000.00.00000000000063 Results for 1983 ..................................73 Results for 1984 ..................................74 Table of Mean and Ranges of Computed Values .......74 Analysis of Variance for Impulse ..................76 Table of Means for Impulse With Pod Position ......78 Analysis of Variance for Maximum Impact Force .....83 Table of Means for Interaction Between Release Angle and Variety for Impact Force.......................85 Mean Maximum Impact Force .........................86 Analysis of Variance for Absorbed Energy ..........89 Tables of Means for Release Angle for Absorbed Energy...’0.0.00I.0.0.0.0...OOOOOOOOOOOOOOO0.0.0.0090 Table of Means of Absorbed Energy for Pod Position.91 Table of Means for Year-Variety Interaction for Absorbed Energy .00....OOOOOOOOOOOOOOOOOOOOOOO...0.94 vi LIST OF FIGURES WhOIe and Opened Navy Bean POdS ..OOOOOOOOOOOOOOOOOOS Transverse Section of a Phaseolus Pod ............5 Sketch of a Pendulum Without Input Power and SpeCimenOOOOOOOCOO...OOOOOOOOOOOOOOOO0.0.0.00.0000019 Sketch of Pendulum With Specimen But No Input Power 19 Graph of Angular Velocity versus Pendulum Release Height for a Pendulum Impacting a Specimen ........24 Typical Force-Time or Deceleration-Time Curve for a Fruit or Vegetable With a Non-Yielding Surface.....24 Equipment Assembly for Impact Testing of Navy Bean Pods...............................................30 Bean Pod Holders ..................................31 Close-up View of the Angular Transducer............33 Angular Transducer Fixed onto a Bracket ...........35 Angular Transducer Facing Away From the Observer ..35 Wheatstone Bridge Circuit .........................38 Location of the Strain Gauges .....................40 Force Transducer Circuit ..........................40 Various Ways of Taking the Shielded Cable From the PendUIum ..OOOOCOOOOOOOOOOOOOO00.0.000000000000000043 Calibrating Force Transducer With Loading Platform 46 Block Diagram of the HP-85 System for Impacting Navy Bean Pads CO...OOOOOOOOOOOOOOOOOOOOOOOI0.0.0.00000047 Calibration Curve for the Force Transducer.........48 Force-Time Output from the HP-85 System ...........49 vii 5.9 5.10 Comparisons Between the Photographic and Angular TranSducer ”ethOds 0.0.0...0.0.0.0....00.0.0000000053 Coding of the Pod Position ........................57 Typical Output on (A) Polaroid and (B) P1astic.....60 Output from an Undeformed Cylindrical Body ........62 Twisted Navy Bean Pods After Shatter...............62 Coding and Computation on the SPSS Program ........71 Plot of Average Impulse for Pod Positions .........79 Plot of Number of Pods Shattered After Impact at Different Release Angles and Pod Positions ........81 Interaction Between Variety and Release Angle for "aXimum Impact Force O0.0....OOOOOOOOOOOOOOOOOOOOOOBS Output from the Instron Machine ...................87 Plot of Mean Absorbed Value With Pod Position .....92 Plot of Mean Absorbed Value With Release Angle ....92 Interaction Between Year and Variety for Absorbed EneIQYOOOOOOOOOOOOOOOOOOOOOOOOOOOOO0.0.0.000000000094 viii ABSTRACT IMPACT TESTING OF NAVY BEAN PODS by Nor Mariah Adam A pendulum system was developed for impacting 1983 and 1984 Swan Valley and C-20 navy bean pods. The impulse, maximum impact force and energy absorbed to shatter were measured. A 2x3x3 factorial design at 2 levels of variety, 3 levels of pendulum release angle and 3 pod positions (side, top, base) was used. The mean absorbed energy was 10.16 mJ with a range of 1.40-31.50 mJ, the mean maximum impact force value was 24.82 N with a range of 11.32-36.97 N and the mean impulse value was 24.32 N-ms with a range of 6.04-71.67 N-ms. Impulse and absorbed energy to shatter depend on pod orientation. The bean pod is weakest when it is hit along the base. Impulse is the most suitable parameter to characterize impact resistance because it is dependent on pod orientation and independent of variety, time of harvest and pendulum release angle. CHAPTER I INTRODUCTION 1.1 Background Michigan has been a leading producer of navy beans. In 1983 over 80 percent of the navy beans produced in the United States were grown in Michigan, which was equivalent to a crop value of about 110 million dollars (Michigan Agricultural Statistics, 1984). The current method of harvesting navy beans in Michigan is by windrow harvesting (Kelly, Burkhardt, Varner, Adams and Srivastava, 1981). This method of harvest involves two operations. Early in the morning when the beans are damp, a bean puller pulls the plants from the ground and forms windrows. The windrows are then left to dry in the sun. Later in the day, a combine harvester collects the windrows to harvest the crop. Usually navy bean harvest starts in middle September and continues into early October. In Michigan this is the time of year when foggy nights and high humidity days are common. If the grower pulls more beans in the morning than can be harvested in the afternoon, some windrowed beans must be left in the field for harvest at a later time. If rains come before these windrowed beans can be harvested, they will be more severely damaged than bean plants which are still standing in the field. If the grower pulls too few plants, he reduces the length of his effective harvest day. Indeed direct combining of navy beans was a salvage operation for the wet seasons of 1975 and 1977 because fields were too wet for conventional harvest (Pickett, 1982). The advantages of direct combining navy beans have been realized by farmers as early as 1952 (Khan, 1952). In windrow harvesting losses from shelling, machine damage and degradation are common when windrowed beans are exposed in wet weather (Gunkel and Anstee, 1962). Direct harvesting on the other hand, is a one-step operation. Merits of direct harvesting mentioned by Kelly gt a1, (1981) are as follows: i. standing plants dry faster than windrowed plants, so combining can begin earlier in the day, ii. eliminates the need for guessing the hectares of crop to be pulled in the morning, iii. the grower can continue to harvest the crop until the beginning of a rainstorm, iv. direct harvesting requires less equipment, hence operation cost is lower, v. a combine harvester has better floatation on wet soil than a tractor-mounted bean puller, which tends to bury more beans during the pulling operation for wet soil conditions. 1.2 Bean Plant According to Kelly and Adams (1981), major navy bean plant types grown commercially in Michigan are as follows: i. bush type ii. short vine type iii. vine type Since 1973, the navy bean breeding program at Michigan State University has been working towards the development of the taller, narrow profile and more erect plant type called architypes (Adams, 1981). Architypes are modified bush type and short vine type navy bean plants. For example, C-15 is an upright bush type variety while Swan Valley is an upright short vine type variety (Kelly and Adams, 1981). Architypes have been developed using parental stocks which were more tolerant to air pollution and soil compaction as compared to the standard bean varieties. They can be grown in narrower rows than the usual 71 to 76 cm (28 to 30 inches), resulting in a higher plant density which leads to a higher yield per unit area (Adams, 1981). The architype bears bean pods which grow well above the ground and these pods mature uniformly. The stem is erect and has a high resistance to lodging. These characteristics make architypes ideal for direct harvest. In addition, the erect and narrow profile appears to create less favorable humidity conditions for white mold growth (Adams, 1981). 1.3 Anatomy of Bean Pod In North America, beans are classified under the genus Phaseolus. Most of the common dry edible beans like kidney beans, navy beans, pinto beans and snap beans belong to the varietal classification Phaseolus Vulgaris L. (Narayan, 1969). Figure 1.1 shows a picture of whole and opened navy bean pods. Usually there are six or seven seeds in a navy bean pod. Pod dimensions vary with different seed size and class. For example, navy bean mass is 20 g per 100 seeds while kidney bean mass is 60 g per 100 seeds (Kelly, 1984). Beans belong to the legume family and they possess dry dehiscent (or self-exploding) pods. The exocarp, which includes the epidermal layer and the subepidermal layer, is composed of thick walled cells (Esau, 1960). Figure 1.2 shows a transverse section of a Phaseolus pod. From the diagram, it can be seen that the seed is attached to the pod suture by the fine funicle. Differential shrinkage of different tissues in the pericarp of a legume, is assumed to be the primary force inducing split in longtitudinal dehiscence (Esau, 1953). When the pericarp splits, the two valves of the dried Figure 1.1 Picture of Whole and Opened Navy Bean Pods Pod Suture Funicle Testa or Seed Coat Cotyledons Pericarp or Region of . Dehiscence Fruit Wall F1'Qure 1.2 Transverse Section of A Phaseolus Pod (From An Atlas of Plant Structure Vol. 2, by Bracegirdle and Miles, 1973) legume twist, thereby expelling the seeds. 1.4 Shatter Shatter is the splitting of the bean pod. During harvesting, shatter can occur when the cutter bar cuts into the bean pod, causing the pod to pop open, thereby releasing the seeds. Kelly et al. (1981) defined shatter loss as "...loose beans and beans in pod not attached to plant.." which were on the soil surface, excluding the preharvest losses. In this study the focus is on loose beans and not loose pods and the criteria for pod shatter is when one or more seeds pop out of the bean pod after impact. Shatter is brittle failure of the bean pod (Hoag, 1975). In engineering, material properties related to brittle failure include ultimate strength and strain energy. It is then possible to know the parameters related to pod shatter when the pod's engineering properties are known. It has been observed that the amount of elapsed time between maturity and harvest affects shatter loss for both navy beans and soybeans (Singh, 1975 and Hoag, 1975). Hoag (1975) suggested that fatigue, caused by cyclic wetting and drying of the bean pods, may contribute to the increase in shatter loss. One of the ways to reduce field shatter loss is by developing suitable varieties. Indeed shatter resistance has been used as a criterion in the selection of soybean varieties (Hoag, 1975). 1.5 Objective The purpose of this work is to build a portable machine for finding physical properties of navy bean pods. Specific objectives are: i. To find pod orientation most susceptible to shatter, ii. To measure the impulse required to initiate shatter, iii. To measure the energy absorbed by the bean pod during impact, iv. To use the above information to find the parameter most suitable to characterize impact resistance for variety development. CHAPTER II REVIEW OF LITERATURE According to Mohsenin (1978), mechanical damage is a by-product of mechanizing harvesting and handling of agricultural products. Impact is a major cause of mechanical damage in harvesting and subsequent processing and handling operations (Fluck and Ahmed, 1973). Product damage like bruises, burst or split fruit, can lead to significant economic losses during storage, handling and marketing of agricultural products (Finney and Massie, 1975).. It has been found that some crop varieties are more susceptible to impact damage than others (Hoag, 1973 and Mohsenin, 1978). This offers an opportunity for plant breeders to develop new varieties which would withstand the mechanical force imposed during production. 2.1 Some Parameters Related to Seed and Grain Damage Common experimental variables used in the analyses of seed and grains subjected to impact are as follows: i. impact velocity ii. moisture content iii. product orientation. 8 The common measured quantities are i. imparted impulse ii. energy absorbed to initiate damage iii. peak resistive force. Bilanski (1966) analyzed damage resistance of various seed grains under gradually applied load, as well as under low and high impact velocities. He found damage resistance dependent on grain size, moisture content and grain orientation. He also observed that grains with high moisture content required more energy to initiate damage than low moisture grains. Perry and Hall (1966) also found impact velocity, moisture content and product size influenced damage resistance when they dropped navy beans at various heights in a silo. In addition to these factors, they found that temperature also influenced damage. Turner, Suggs and Dickens (1967) varied impact velocity, moisture content and specimen orientation in their impact experiment on peanuts. They found the coefficient of restitution depended on specimen orientation. When low moisture peanuts were subjected to high impact velocities, they observed that these peanuts were mostly damaged by the brittleness of the bulls. Hoag (1972) used experimental variables similar to Turner 33 213's (1967) for his impact experiment on soybean pods. Contrary to Turner et al.‘s finding, Hoag (1972) found that 10 specimen orientation did not give significant differences in energy adsorption to initiate pod shatter. Hoag (1975) extended his research and found that the maximum force to cause soybean pod damage dependent on the pod moisture content. He recommended imparted impulse to be used as an indicator of field shatter loss. Reported studies on corn breakage due to impact were mainly concerned with defining parameters that can characterize corn kernel resistance to shearing. Srivastava, Herum and Stevens (1976) impacted corn kernels at various orientations and moisture levels. They measured maximum resistive force, imparted impulse and energy absorbed to initiate failure. They found, generally, that these values increased with moisture content. They recommended energy absorbed per unit area in longtitudinal impact to characterize impact strength of corn kernels. Mensah, Herum, Blaisdell and Steven (1981) obtained similar measured quantities as Srivastava et a1. (1976) for their impact experiment on corn kernel. Although they obtained general observations similar to Srivastava et a1. (1976), they recommended the peak shear stress to characterize impact strength of corn kernels. 2.2 Previous Research Work on Navy Beans in Michigan Post war research on navy beans were mainly concerned with improving harvesting techniques (Khan, 1952). 11 Research on direct harvesting of the bush type navy beans in the early 1950's has been reported (Mc Colly, 1958). Mc Colly (1958) found that the finger type reel was more efficient than the standard bat type. This is because the tines on the finger type reel could be adjusted such that the bean plants were lifted towards the cutter bar, thereby reducing cutter bar losses and shatter losses. He also mentioned that the available plant variety at that time was not suitable for direct harvesting. Some research related to navy bean harvesting has dealt with the pod moisture content. Pickett (1973) found mechanical damage to navy beans during harvesting dependent on both pod and seed moisture content as well as on the cylinder speed. He also stated that pod moisture content was likely to affect threshability. For optimum harvest conditions, he recommended a moisture content of under 12 percent for the bean pod and between 17 and 20 percent for the bean seeds. Singh (1975) studied the Sanilac and Seafarer navy bean varieties to evaluate environmental effects on field drying and harvesting. He developed models for rate of change of moisture levels and overnight rise in moisture level for both bean pods and bean seeds separately. He also developed a model for unthreshed loss as a function of pod moisture content and cylinder speed. Since the models indicated varietal differences, be suggested that physical properties of newly introduced 12 varieties be found before using his prediction models. He agreed with Pickett (1973) that threshability of the crop was influenced by the bean pod moisture content. For maximum threshability with minimum bean damage, Singh (1975) recommended moisture levels of under 13 percent for the bean pod, and between 18 and 20 percent for the bean seeds. Knowledge about physical and mechanical properties of navy bean seeds is important for seed production and processing; and for reduction in seed damage during harvesting and handling operations. Reported research on the determination of these properties was done under quasi-static and/or impact loading. Perry and Hall (1965) used a wooden bar to strike individual navy bean seed at various moisture levels and various impact velocities. These velocities were similar to velocities that would have been attained by the navy beans after a free fall of 6.1 to 7.6 meters (20 to 25 feet). They used high speed photography to evaluate the impact force, impact duration and deformation of each seed. Narayan (1969) used the column stability theory to compute stability modulus and elastic modulus of navy bean seeds under quasi—static loading. He used a high velocity impact arm to measure impact forces required to cause seed checking. Checking is splitting of the seed coat. Hoki (1973) measured Young's modulus and ultimate strength of 13 navy bean seed coat and cotyledons separately under quasi-static tensile loading. He used the contact theory to predict deformation of the navy bean seed under compressive loading. The contact theory was incorporated with the impact theory to predict damage under impact loading. 2.3 Measurement of Mechanical Properties of Biological Materials. Several types of measuring techniques have been developed for specific agricultural products. Uniformity in testing techniques is difficult because of the complex structure and variations in size and shape of agricultural products. To date, only the Stein tester is commercially used to test mechanical strength of grains (Singh and Pinner, 1983). Measuring techniques can be divided into quasi-static and dynamic methods. Common quasi-static loading is either under compression loading or tension loading. Tension testing is less popular due to the difficulty in gripping the specimen without damaging the tissues. Dynamic testing includes simple drop tests, pendulum, pneumatic impact device, rotary arm, centrifugal impactor and vibration tests. Simple drop tests, either of a product upon a rigid surface, or of a mass upon the product, have been 14 extensively used (Perry and Hall, 1966; Hammerle and Mohsenin, 1966; Sharma and Bilanski, 1971 and Pluck and Ahmed, 1973). With this method, the velocity at impact is limited by the drop height and orientation of a dropped specimen is impossible to control. The pendulum is a popular impacting device (Perry and Hall, 1965; Bilanski, 1966; Turner et al., 1967; Hoag, 1972; Srivastava et al., 1976 and Mensah et al., 1981). The pendulum is versatile; the impact velocity can be varied either by changing the pendulum length (Lyon and Zable, 1973) or by changing the release angle (Bilanski, 1966 and Srivastava g£_gl:,l976). This method allows easy control of specimen orientation. High velocity impact arms driven by variable speed motors have been employed for high speed impact testing (Bilanski, 1966; Turner et al., 1967 and Burkhardt and Stout, 1971). This method is used to simulate free impacts during threshing or handling. Keller, Converse, Hodges, and Do Sup Chung (1972) evaluated corn kernel damage by pneumatically projecting the kernels against selected materials. Hoki and Pickett (1973) developed a high speed impact tester which consisted of a rotating impact disk and a vacuum bean holding disk. The centrifugal impactor has been used to provide random impacts under controlled speeds for corn kernels and soybean seeds (Cooke and Dickens, 1971; Paulsen, Nave and 15 Gray, 1981 and Singh and Finner, 1983). Vibrational characteristics of agricultural products have to be known before vibratory harvesting can be done. Research has been done on the mechanical impedance of blueberries (Rohrbach and Glass, 1980), the vibrational characteristics of blueberry canes (Ghate and Rohrbach, 1975) and the resonant frequencies of strawberries (Idell, Holmes and Humphries, 1975). 2.4 Instrumentation Methods High speed photography is a popular method of measuring the impact velocity of a pendulum (Perry and Hall, 1965; Turner et al., 1967; Hoag, 1972 and Burkhardt and Stout, 1974). This method gives a continuous documentation of impact and allows an accurate and reliable way to measure impact duration. Careful coordination of film exposure rate with the impacting arm movement is important. Precaution must be taken to prevent the specimen from drying under the heat of the camera lights. Perry and Hall (1965) used an asbestos shield to protect their specimens from the heat of the camera lights. The piezo—electric or quartz type accelerometer has been used for continuous measurement of acceleration (Hammerle and Mohsenin, 1966; Burkhardt and Stout, 1971; Hoag, 1972; Lyon and Zable, 1973; Pluck and Ahmed, 1973; Srivastava et al., 1976 and Mensah et al., 1981). The 16 acceleration signal is usually preamplified and is commonly displayed on a storage oscilloscope. An angular transducer, such as a tachometer transducer or a shaft encoder, may give a continuous measurement of the pendulum angular position (Bilanski, 1966 and Mensah gt 3}; 1981). It is convenient to have the angular transducer output displayed on a chart recorder. Strain gauges have many applications in experimental work. Bledsoe and Swingle (1972) measured detachment properties of snap beans using strain gauges attached on a cantilever. Goyal, Drew, Nelson and Logan (1980) used strain gauges to evaluate seed emergence forces. The strain gauges were placed on an aluminium ring which acted as a sensing element. The force transducer was sensitive enough to measure forces to the nearest 0.01 N. 2.5 Summary Product damage can be reduced in today's mechanized production of agricultural products, when the physical properties of the products are known. Several types of measuring techniques and instrumentation exist because of variations in product size, shape and complex structure. Common parameters to characterize impact strength are imparted impulse, peak resistive force and energy absorbed to initiate damage. Common measuring techniques are quasi-static loading, pendulum, simple drop tests, 17 pneumatic impact device, high velocity impact arm, centrifugal impactor and vibration tests. .Popular instrumentation techniques uses the accelerometer, shaft encoder and strain gauges. CHAPTER III THEORETICAL CONSIDERATIONS AND EXPERIMENTAL DESIGN 3.1 Theoretical Considerations Characteristics of an acceleration-time or force-time curve are important for analysis of impact. Hammerle and Mohsenin (1966) stated that absorbed energy and duration of impact are important factors to initiate impact damage. It is also possible to determine the type of damage from the force-time curve (Pluck and Ahmed, 1973). 3.1.2 Impact Energy For a pendulum with no input power, as in Figure 3.1, impact energy is provided by the gravity field to the pendulum mass, and is a function of the release height. By the Law of Conservation of Energy, the total energy, that is the sum of kinetic energy T, plus the potential energy V, remains constant up to the instant of impact. In equation form, T1 + Vi = T; + V2 where the subscripts refer to arbitrary positions of the pendulum prior to impact. 18 19 Rebound At C h1 = h2 ° J”//Down Bottom of Travel Swin At 8 9 Figure 3.1 Sketch of A Pendulum Without Input Power and Without Specimen (no impact) a Pivot , x.“ <\\ I 1 Datum I l \ \\\\ - h I I, “4' ‘\ \\ h]- 2 ,' ' - ' ‘ Q I. ; II ’I a“ ‘\\\\\\ ll 2 . g \\ \y / \ ,’/1 Release Il’Mg At A I I [III] A Rebound At c 43’ i, “2 ._ Specimen striked h > h (f f By Pendulum 2 1 E A 7"5 p Figure 3.2 Sketch of Pendulum Without Input Power With Specimen (with impact) 20 3.1.2.1 Without Specimen (No Impact) With no impact, the total energy at A equals the total energy at all positions of the down swing, including at the bottom of travel (See Figure 3.1). This can be expressed as: TA+VA=TAB +VAB =TB+VB where: subscript A refers to position at A, subscript AB refers to position between A and B, subscript B refers to position at B. Substitution in the above equation yields: 0 + Mg(L-h‘) = 1/2 Iou2+ Mg(L-h) . 1/2 Iou§+ Mg(L - L/2) which simplifies to: o - Mgh1 = 1/2 10.} - Mgh = 1/2 10.»: - Mg L/2 (3.1) where: Mg - weight of the pendulum, N, h - position of the pendulum center of gravity at A (at release),m, I - mass moment of inertia about pivot, kg-mz, w - angular velocity at an arbitrary position between A and B, rad/s, h - position of pendulum center of gravity at an arbitrary position between A and B, m, ab - angular velocity at bottom of travel, rad/s, L - length of pendulum arm, m. 21 From Equation 3.1 we get: Mg ( h - h1) = 1/2 Iowz (3.2) Mg (L/2 -h1) = 1/2 10...: (3.3) Additionally, with no specimen in the fixture, and hence no impact, energy will be conserved between position A and position C. Here T+V=T+V A A C C or 0 - Mgh1 = 0 - Mgh2 so that, for a completely frictionless pendulum, h1 = hz. However, experimentally h2 > h1 because of the presence of friction in the pivot bearing. Therefore, if h‘ and h2 can be measured, for a range of values of h1, the corresponding friction values can be calculated. 3.1.2.2 With Specimen (With Impact) When the pendulum strikes the specimen (bean pod), as in Figure 3.2, the motion of the pendulum can be divided into three parts, namely: i. down swing, ii. impact (negligible motion of the pendulum), iii. up swing. Down Swing Applying Equation 3.1, the total energy at pendulum release and just prior to impact can be written in the energy balance equation as: 22 o - Mgh = 1/2 10.»: - Mg L/2 where a” is the angular velocity just before impact in rad/s. The above equation can be simplified to give a general equation before impact as Mg (L/2 - h1) - 1/2 Io”: (3.4) Impact When the pendulum hits the specimen, some energy will be absorbed by the bean pod. The absorbed energy may appear as some form of damage to the bean pod. Equation 3.1 then becomes TC + Vb = TB + VB (just after impact) or o - M9112 = 1/2 10...: - Ep - Mg L/2 (3.5) where Ep is the energy absorbed by the bean pod during impact' and h2 is the pendulum center of gravity at C. Substitution of Equation 3.3 into Equation 3.5 yields Ep = Mg (h2 - h‘) (3.6) and h2 > h1. Up Swing. During impact some energy is transferred to the bean pod. Following impact, the Law of Conservation of Energy holds for the up swing motion. The total energy at the bottom of travel equals the total energy at the end of the up swing, C. The energy equation can be written as: 2 0 - Mgh2 = 1/2 low2 - Mg L/2 23 where “5 is the angular velocity after impact in rad/s. The above equation can be simplified to give a general equation after impact as: 2 Mg (L/2 - h2) = 1/2 Io.»2 (3.7) Equations 3.4 and 3.7 are the energy balance equatiOns for conditions before and after impact. Clearly, the difference in energy levels before and after impact is attributed to the energy transferred to the bean pod during impact. In equation form, 2 2 =Ep=Mg (h2 -h1) (3.8) The value of the energy absorbed by the bean pod during impact, Ep, can be computed when both at and wz, or both h1 1 and h2 are known. From trigonometry, h1 = L/2 cos 6‘ and h2 = L/2 cos 92, 9‘ )92 (3.9) where 91 and 92 are the release angle and the maximum rebound angle respectively. The impact motion can also be graphically described by the angular velocity, w , versus the position of the pendulum center of gravity, h, as in Figure 3.3. Between A and B is the region of downswing, and at impact BB', some energy is transferred to the bean pod during a very short period of time. This is characterized by the sudden decrease in w . Between B' and C is the region of upswing after impact. Clearly, due to the energy transfer to the 24 U1; —————————————————————— B “2. ................... At Impact 3 :3. i? Before : '8 Impact : ,2 u g I L l .2 I (:35 I C l < I I A I I h1 h2 L_ Pendulum Height, h 2 Figure 3.3 Graph of Angular Velocity id versus Pendulum Height h for A Pendulum Impacting A Specimen Force (or Deceleration) Time Figure 3.4 Typical Force-Time or Deceleration-Time Curve For A Fruit or Vegetable with A Non-Yielding Surface (From Fluck and Ahmed, 1973) 25 bean pod, the motion ends at hz' where h2 > h 1 The values of u! for conditions before impact can be computed using Equation 3.2, where the U2 term can be simplified to give 2 2 Mg/Io * (h - h1) = (v This can be written as w = fzngm - h‘)/Io (3.10) From Equation 3.10 it is possible to get a plot of w for a range of values of h. 3.2 Impact Action According to Mohsenin (1978), the concept of impulse-momentum forms the classical theory of impact. From Goldsmith (1960), the Linear Impulse-Momentum Law is given by: t JP dt 8 d(mv) (3.11) 0 where: '11 I impact force, N, t - duration of impact, 5, v - linear velocity, m/s, m - mass of object, kg. For a constant mass, from Turner et a1. (1967), Equation 3.11 becomes 26 t vp (I F dt = mp dv = mp vp (3.12) 0 0 where: mp'- mass of bean pod, kg, vp - linear velocity of bean pod just after impact, m/s. A typical curve such as in Figure 3.4 can be obtained by an instrumented target with which the specimen collides (Fluck and Ahmed, 1973). The peak force is the highest point on the curve. The impact duration, t time interval between the initiation and the termination of impact. The area under the force-time curve is the experimental impulse value. High speed photography can be used to verify velocity calculations and to determine specimen deformation. 3.3 Experimental Design In this study three factors of interest were variety, pod orientation and pendulum release angle. These variables are independent of each other. Variety and pod orientation are qualitative factors while the release angle is a quantitative factor. Two bean varieties were tested, which means that variety was a two-level factor. Both pod orientation and release angle were three-level factors. 27 The factorial treatment combination of 2 x 3 x 3 is shown in Table 3.1. This type of treatment combination allows one or both of the following analyses to be performed: i. Computation of main effects and interaction between factors, ii. Identification of factors that can lead to future work. A random complete block design was used with time as the blocking factor. This means that within a particular block of time, the bean pods were subjected to all 18 treatment combinations. Each experimental unit was subjected to .3 particular treatment combination, which was chosen at random. Moisture content was not included as a fourth factor so as. to increase the precision of the statistical tests. A 2x3x3x3 factorial treatment combination with moisture content at three levels would require 54 treatment combinations. This is not possible due to the limited number of bean pods available and the size of work anticipated. 28 Table 3.1 A 2x3x3 Factorial Combination Table Run Variety Release angle Pod orientation 1 C-20 40 top 2 C-20 40 bottom 3 C-20 40 lateral 4 C-20 50 top 5 C-20 50 bottom 6 C-20 50 lateral 7 C-20 60 top 8 C-20 60 bottom 9 C-20 60 lateral 10 Swan Valley 40 top 11 Swan Valley 40 bottom 12 Swan Valley 40 lateral 13 Swan Valley 50 top 14 Swan Valley 50 bottom 15 Swan Valley 50 lateral 16 Swan Valley 60 top 17 Swan Valley 60 bottom 18 Swan Valley 60 lateral CHAPTER IV EQUIPMENT A pendulum system was set up as shown in Figure 4.1, to impact navy bean pods. The system was comprised of: i. a pendulum ii. bean pod holders iii. electrical release mechanism iv. angular transducer and counter timer v. force transducer vi. storage oscilloscope with special camera. 4.1 Pendulum The pendulum, a modified steel bar, was supported on two ball bearings. A pointer was placed at the top of the pendulum to read off the angle 6n the graduated angle plate (protractor) for measurement of release angle, rebound angle and angular speed of the pendulum. 4.2 Bean Pod Holders The bean pod holders were screwed into supports in a stand (See Figure 4.2). The bean pod was simply supported on the holders such that the pod could be fixed at any 29 3O cowuwmoa a? nausea gar: onoumo—Pwomo mangoum .m amusementh muted .m Louzvmcmch ampsmc< .e Emwcmsomz mmmwpmm qu_cpowpm .m mtmupoz com :mmm .N e=_:u:ma .H muoa comm A>mz Go mcwumwh aomasH to» SFBEmmm< ucmsnwzcm Ho e mesmwm 31 Bean Pod Holders Figure 4.2 32 orientation to the impacting ring, and moved to any position along its length. The height of the holders and the distance of the support stand from the pendulum could be adjusted. 4.3 Electrical Release Mechanism The electrical release mechanism consisted of a solenoid, a 6 volt dry cell and switch. When the current was switched on, the magnetic force from the electromagnet was strong enough to hold the pendulum at its release angle. The pendulum was released when the current was switched off. The bracket which held the solenoid could be moved along a slot to vary the release angle. This enabled a consistent method of positioning and releasing the pendulum arm. 4.4 Angular Transducer The angular transducer was made up of two photosensors placed at 13 mm (0.5 inch) apart (See Figure 4.3). The photosensors functioned in a bright environment. When the pointer passed by the first photosensor at an arbitrary position A, the pulse triggered the counter timer (Fluke model 1953A counter-timer). When the pointer passed by the second photosensor at an arbitrary position B, the second pulse stopped the timer. The time scale on the timer ranges from 0.1 us to 10 8. Knowing the fixed angular 33 Figure 4.3 Close-up View of the Angular Transducer 34 distance between A and B, and the time taken to move from A to B, the angular velocity of the pointer can be calculated. To avoid interference from the shielded cable, the photosensors were placed away from the observer (Figures 4.4 and 4.5). This arrangement required an extra light source for the photosensors to perform consistently. 4.5 Force Transducer This section will describe some strain gauge properties and temperature compensation, and Wheatstone bridge circuit before describing properties and construction of the force transducer. 4.5.1 .Strain Gauge Strain, c , is a geometric property of a deformed body and, is defined as extension/original length. An electrical-resistance strain gauge will change in resistance when the wire is stretched and applied strain is developed, according to the following equation: .AR/R - G.F. t (4.1) where 1AR/R is the change in resistance (ohm)/original resistance (ohm). The gauge factor (G.F.) is a property of the strain gauge (a number) and c is strain (dimensionless). Also E = Force/Area/c = stress/c (4.2) 35 I Figure 4.4 Angular Transducer Fixed Onto A Bracket Figure 4.5 Angular Transducer Facing Away From the Observer 36 where E is the Young's Modulus (N/mz) and a force (N) is applied to a cross-sectioned area (m2). Equation 4.2 can be rewritten as: c = stress/E (4.3) A , Wheatstone bridge circuit is commonly used to convert the .AR/R value to a voltage signal which can be measured with a recording instrument. Since the strain gauge is small, light, precise and inexpensive, it is commonly used as the sensor in a variety of transducers (Dally and Riley, 1978). A transducer is a device which uses one property to measure another property. 4.5.2 Temperature Compensation In many test programs, the strain gauge installation is subjected to temperature changes during the test period, and careful consideration must be given to determine whether the change in resistance is due to applied strain or to a temperature change. Strain developed from a temperature-induced resistance change is referred to as apparent strain (Measurements Group Tech Note TN 128-3, 1976). The temperature of the strain gauge is influenced by ambient temperature variations and by the power dissipated in the gauge when it is connected into the Wheatstone bridge circuit. To reduce the power dissipated in the gauge, gauge current must be minimized. Steps taken to reduce the gauge current and hence, the power 37 dissipation in the force transducer circuit include: 1. using 350 ohm resistance strain gauges instead of the common 240 ohm gauges, 2. employing a 5 volt instead of the 10 volt excitation voltage, 3. a temperature compensated circuit design was also used via the active full bridge circuit where the (actual strain)/(apparent strain) equals unity (Measurements Group Tech Note TN 128-3, 1976). 4.5.3 Wheatstone Bridge The material for this section is obtained from Dally and Riley (1978). The Wheatstone bridge circuit may be used as a direct readout device where the output voltage is measured. The bridge may also be used as a null-balance system, where the output voltage is adjusted to zero by adjusting the resistive balance of the bridge. Consider the Wheatstone bridge circuit in Figure 4.6 where the excitation voltage is E volts. The voltage drop across R1 is denoted as VABand is given as VAB' E * R1/(R1 + R2) volts (4.4) Similarly, the voltage drop across R4 is denoted as VADand is given by VAD: E * R4/(R3 + R“) volts (4.5) The output from the bridge is VBDand VBDR VAB- VAD volts (4.6) 38 E volts Figure 4.6 Wheatstone Bridge Circuit With E Excitation Voltage 39 Substituting Equations 4.4 and 4.5 into Equation 4.6 and simplification yields VBD= R R3 - R2R4 E volts (R1 +122) (HT-3+ 4 V will go to zero and the bridge will be considered in BD balance when R1 R3 = R2R4 The bridge is initially in balance before strains are applied to the gauges in the bridge, thus the output voltage is initially at zero and the strain inducethV can be measured directly for both static and dynamic applications. Since the force is linearly related to strain (Equation 4.3), as long as the ring remains elastic, the force transducer can be calibrated (See Section 5.1) so that the output signal is interpreted as a force reading. Aluminium was chosen because it has a '1ow elasticity modulus and it is a good heat dissipator. Four 350 ohm strain gauges made for aluminium applications, model EA-l3-125AC-350, were glued onto the ring with M-bond 200 adhesive (Figure 4.7A) and covered with silicon rubber for protection from the environment. These gauges were connected in an active full bridge circuit with 5 volt excitation voltage (Figure 4.7B). The letters A, B, C and D were the corresponding pin connections on the shielded cable which connected the gauges to the signal conditioner. The signal conditioner is a black box with a high 40 Pendulum Arm Allen Screw Figure 4.7A Location of Strain Gauges on the Ring To Signal "A Conditioner "C and Amplifier Figure 4.73 Sketch of the Force Transducer Showing the Location of the Strain Gauges and Its Circuit 41 input impedance (2 MI) ). It is used for balancing a Wheatstone circuit (1/4, 1/2 or full bridge) and to supply a well regulated 5 or 10 volt excitation voltage to the Wheatstone circuit. It also has a built-in voltmeter for measuring the bridge output voltage. The power dissipation from each strain gauge is 1.14 W/in2 . This is far below the 5 W/in2 high precision rating for static and dynamic loading (Measurements Group Tech Note TN-502, 1979). Hence, thermal effects can be considered to be negligible. In an active bridge such as in Figure 4.78, temperature-induced resistance change in opposite pairs is cancelled out when the strain from the inside gauges is combined with the strain from the outside gauges. The important features of this circuit were i. it was inherently temperature compensated, hence, ii. true strains were equal to apparent strains, iii. effects were additive, resulting in a larger output. The output from the signal conditioner was put into a Tektronik 59 storage oscilloscope. The storage oscilloscope provides clear visual displays of the output, with a sweep rate ranging from 0.5 us to 5 8 per division. The display on the scope can be photographed with the special equipment camera which uses a high speed polaroid film (Type 47,speed 3000). CHAPTER V EXPERIMENTAL PROCEDURE This chapter is divided into two parts. The first part deals with the calibration techniques for the force transducer and the angular transducer. The latter part deals with impacting bean pods, the data collection procedures and the calculations involved to obtain impulse, maximum impact force and the energy absorbed to shatter the bean pod. 5.1 Force Transducer A 0.6 cm (1/4 inch) diameter shielded cable was placed along the center line of the pendulum which had dimensions of 27x3x0.9 cm. The cable was used to protect the strain gauge signals from stray voltage in the environment. Therefore, the cable had to be placed as close possible to the gauges. The size and inflexibility of the cable called for a careful method of taking the cable from the pendulum to the signal conditioner without affecting the motion of the pendulum. Different ways of taking the shielded cable, perpendicularly, vertically and parallel, from the pendulum were tried (See Figure 5.1). The most suitable cable 42 .43 sapzccma on» sot; mpnmo umcpmwsm as» acmeP we who: maowtm> H.m crammu 363 mdz< mm \ art :3 22:: mi . 4 d44 Ila-l. m<4=u~ozmmmwm Ital... 4 man: haghm) l:+..l m" mm. mm mm mm SQ mv am mm ( 58p) 313m UNTDSBN 44 arrangement yielded free swing rebound angles similar to those of a pendulum without the shielded cable. A photodiode, consisting of an infrared emitter and detector embedded in a U shaped resin fixture and connected to a black box, was made to measure the maximum free swing rebound angle. The photodiode was attached to a bracket which was placed near the angle plate. The bracket was made such that it could be moved along the circumference of the angle plate. When the pendulum pointer blocked the light path from the emitter to the detector, the green light on the black box would light up. The pendulum was released at a given angle as, if the green light on the black box did not light up, the photodiode was moved by 0.5 degree, and the pendulum was again released at 6. This procedure was repeated several times until there was green light, and the corresponding rebound angle was recorded. Each point in Figure 5.1 is an average of 3 readings. The results in Figure 5.1 suggest that taking the cable vertically was the most suitable method. The bridge was checked for balance with a strain indicator. A balanced bridge indicates that proper solder and gauge installations were made. To ensure reliability and sensitivity, the force transducer was first connected to the signal conditioner for 70 hours to check for any drifts, then it was put under 12 continuous cycles of loading and unloading using weights and a loading platform 45 (See Figure 5.2). The force transducer was loaded up to 10.7 N with 0.4 N increments (up to 2.4 lb with 0.1 lb increments). The signal conditioner was connected to a Hewlett-Packett (HP) 85 computer and 3 3497A data acquisition unit (See Figure 5.3). The calibration equations obtained were: Loading: Y = 0.4995 + 0.665 x; r = 0.999 Unloading: Y = 0.6060 + 0.662 x; r = 1.000 Average: Y = 0.5552 + 0.663 x; r = 0.999 where Y refers to the output voltage in volts and X refers to the load in newtons. To determine an average calibration value, a single curve was obtained by combining the loading and unloading conditions because both the loading and unloading curves have similar slopes (See Figure. 5.4). The sensitivity of the force transducer was 1.51 N/volt (.34 lb/volt). This means that if the output voltage from the circuit is 1 volt, then the corresponding load the force transducer registers is 1.51 N. The HP 85 computer was also used to check the impact duration with the set up shown in Figure 5.3. The pendulum was released at 40 degrees. The smallest time interval for the data acquisition unit is 1 ms. In Figure 5.5 the impact duration is 2 ms and the impulse curve is made up of only 3 data points. Even though a triangular wave can be thought of as a rough approximation of a sine wave, digitally characterizing a triangular wave requires a 46 Figure 5.2 Calibrating the Force Transducer Using A Loading Platform and Standard Weights 47 muoa :mmm z>mz mcwuumuEH tom Embmxm mmua: ecu Co Emcmm_a goopm m.m mgzmwm mam mowsnflnpmflm owmnopm mommmoch Hmccmno Memo oppommwo com mopsmsoo pass honoflpfl6200 enom IOLOHE soapflmflsdo< chmflm a mm m: spam Lzu :OSMLQPFQU mmmLm>< #6 95¢: m20h2mz zH o :mzm omuu zumwgm> m “mqum < mom h=o><4 4