IN DEGREE PROJECT TECHNOLOGY,FIRST CYCLE, 15 CREDITS

, STOCKHOLM SWEDEN 2019

W.A.N.TWeightlifting Ant

FAIZA ALI

MARTIN SCHRÖDER

KTH ROYAL INSTITUTE OF TECHNOLOGYSCHOOL OF INDUSTRIAL ENGINEERING AND MANAGEMENT

W.A.N.T

Weightlifting Ant

FAIZA ALI, MARTIN SCHRODER

Bachelor’s Thesis at ITMSupervisor: Nihad SubasicExaminer: Nihad Subasic

TRITA-ITM-EX 2019:29

AbstractThe purpose of this project is to create a light weight roboticversion of an ant that can withstand great forces, trying tocome close to the ant’s lifting technique as much as pos-sible. This idea was chosen with inspiration from nature,especially from the obscure forces of an ant. These insectsare proven to be able to lift and carry heavy loads, up toa thousand times their body weight. Various lifting tech-nologies are used by several facilities today and thereforethere is a need for improvements in this field.

By trying to come close to an ant’s appearance andmimic certain ant movements, a hexapod was designed overa period of four months. The tests made in this project weredivided into three categories; stability, lifting and grippingability. The best balance was achieved by placing the legs’contact points on the ground as far away from each other aspossible. In total the robot ant could lift about 1.02 timesits own weight and bear 3.01 times its own weight on thethorax.

Keywords mechatronics, hexapod, lifting, weight, sta-bility, ant, gripping

ReferatTyngdbärande myra

Malet med projektet ar att konstruera en lagvikts robotmy-ra som kan utharda stora krafter och harma myrors rorelsevid lyft sa mycket som mojligt. Projektiden valdes med in-spiration fran naturen, speciellt fran de otroliga krafter hosen myra. Myror har bevisats kunna lyfta och bara tungalaster eller mer exakt tusen ganger sin egen vikt. Olika lyft-tekniker anvands av flera faciliteter idag och darmed finnsdet behov av forbattringar i detta omrade.

Genom att efterlikna en myras utseende och harma dessrorelser designades en sexfoting under en period pa fyramanader. Testerna delades in i tre olika kategorier; stabili-tet, lyft- och greppformaga. Den basta balansen uppnaddesda benens kontaktpunkter med marken placerades sa langtifran varandra som mojligt. Totalt klarade robotmyran attlyfta 1.02 ganger sin egen vikt och bara 3.01 ganger egnavikten pa ryggen.

Nyckelord mekatronik, sexfoting, lyfta, vikt, stabilitet,myra, greppa

Acknowledgements

We would like to thank Nihad Subasic and Staffan Qvarnstrom for always showingsupport and providing assistance. Especially thanks to Robert Romejko at theRoyal Institute of Technology’s department of automatisation for helping us withhydraulic solutions and supplying us with equipment. Without the help from theassistents, Sresht Lyer and Seshagopalan T. Muralidharanin, the project would nothave been as successful. We are also grateful for the help we have been provided byassistents at KTH Prototype Center for guiding us with the usage of 3D-printersand laser cutters.

Martin Schroder and Faiza Ali, Stockholm 28-05-2019.

v

Contents

Contents viList of abbreviations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii

1 Introduction 11.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.2 Purpose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.3 Scope . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11.4 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2

2 Theory 32.1 Past research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.2 Anatomy of hexapods . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.2.1 Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32.3 Slider Crank Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . 42.4 Mathematical model of the system . . . . . . . . . . . . . . . . . . . 5

2.4.1 End Effector . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.4.2 Leverage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62.4.3 Hydraulics and Pneumatics . . . . . . . . . . . . . . . . . . . 7

2.5 Microcontroller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82.6 Servo Motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

3 Demonstrator 93.1 Problem statement . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.2 Electronics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.2.1 Arduino . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93.2.2 Servo motor . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

3.3 3D-printed parts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.3.1 End effector . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.3.2 Thorax and legs . . . . . . . . . . . . . . . . . . . . . . . . . 10

3.4 Hydraulics and Pneumatics . . . . . . . . . . . . . . . . . . . . . . . 113.5 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11

4 Results 134.1 Testing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

vi

4.1.1 Hydraulics vs Pneumatics . . . . . . . . . . . . . . . . . . . . 134.1.2 Gripping procedure . . . . . . . . . . . . . . . . . . . . . . . . 144.1.3 Stability and movement . . . . . . . . . . . . . . . . . . . . . 144.1.4 Final Results . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

5 Discussion and Conclusion 175.1 Result expectation and analysis . . . . . . . . . . . . . . . . . . . . . 175.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

Bibliography 19

Appendices 20

A Flow Chart and Arduino Code 21

B CAD construction 27

C Electric Circuit 29

D Servo Motor Data Sheets 31

List of Figures

2.1 Ant Anatomy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.2 Triangle of stability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42.3 Slider Crank Mechanism . . . . . . . . . . . . . . . . . . . . . . . . . . . 52.4 Force Analysis of Object on End Effector . . . . . . . . . . . . . . . . . 62.5 Forces acting on lever . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

3.1 Mandible design in CAD . . . . . . . . . . . . . . . . . . . . . . . . . . . 103.2 Leg design in CAD . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113.3 The hydraulic system . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123.4 A part of the hydraulic system . . . . . . . . . . . . . . . . . . . . . . . 12

4.1 Parallel inner legs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 154.2 Angled inner legs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16

viii

List of abbreviations

CPU Central Processing UnitIDE Integrated Development EnvironmentMCU MicrocontrollerPLA Polylactic AcidRAM Random Access MemoryROM Read Only Memory

Chapter 1

Introduction

1.1 BackgroundAnts are well known for being able to withstand great forces and carry heavy loadswith respect to their own body weight. This is due to their biological appearancewith an exoskeleton and compact muscle textures. Studies have also shown thatants tend to use special techniques to facilitate the lifting procedure which willbe described more thoroughly later [1]. Weight lifting technologies are importantfor industry, for example the lifting of gigantic beams or other heavy equipmentwhen constructing buildings. Most heavy lifting cranes make use of hydraulics togenerate enough force. Therefore, it is a subject that will always have room forimprovements.

1.2 PurposeThe goal with this project is to research different lifting mechanisms and implementthem on a light weighted hexapod controlled by an Arduino microcontroller. Thefollowing research questions will be answered.

• How possible is it to mimic an ant’s lifting technique and movements withrobotics?

• How much can the robot lift in comparison to its own weight?

• How can this technology be implemented with the Arduino microcontroller?

1.3 ScopeThe lifting mechanisms considered in this project will be hydraulics, pneumaticsand leverage. The equipment that is available to use are 3D-printed plastic andlaser cut materials. Also, due to the budget provided by the KTH Royal Instituteof Technology, the number of servos per leg is limited to two.

1

CHAPTER 1. INTRODUCTION

1.4 MethodArduino is the microcontroller chosen for this project which will control all the servomotors and hence the hexapod’s movements in the construction. Even thoughelectronics will be used to make movements possible, the mechanical part of theconstruction plays a significant role. Thus, the main focus is to find the rightpart designs needed for accomplishing the project goals. The robot consists of sixmoveable legs, a head, thorax and abdomen. The mandibles is designed to grabobjects with and will be attached to the head. When it comes to the lifting part,this project tests both hydraulics and pneumatics in combination with leverageto generate enough force. The leg design and walking pattern is researched andtested. All parts are 3D-printed plastic which has adjustable durability and weight.Therefore, it is possible to experiment with different designs and infill densities.Furthermore, all sections of the hexapod were repeatedly redesigned and 3D-printeduntil the overall construction was steady enough to be tested.

2

Chapter 2

Theory

2.1 Past research

For robots, lifting heavy objects can be challenging. Not only does the constructionneed stability but the material must be strong enough to withstand heavy impactswithout breaking.

The end effector of a lifting hexapod plays a significant part on the lifting proce-dure and must therefore be designed precisely [2]. This device is the mechanical partof a robot that interacts with the surrounding [3]. For a weight lifting ant robot,the end effector has been chosen to be a gripper representing the ant’s mandibles.

2.2 Anatomy of hexapods

Since ants are the source of inspiration for this project, it is important to know theseinsects’ physiques in order to mimic their appearance and movements. Somethingall species of ants have in common is that the body consists of a head, a thorax andan abdomen, see Figure 2.1. The previously mentioned mandibles are only usedfor grasping and cutting while the actual lifting procedure is done by tilting thehead. Ants being hexapods, have all their six legs attached to the thorax. Duringmovements, these insects tend to move half of their legs at a time, using a tripodgait [4].

2.2.1 Stability

Whenever ants transport heavy weights, for example a leaf, they tend to tilt theirheads upwards not only for lifting but also to remain stable. The scientific expla-nation behind this action is that tilting the head causes the total center of mass tostay within the so-called stability triangle, also known as the support polygon. Antsare known for being both statically and dynamically stable because of the techniqueof having at least three contact points on the ground at a time when walking. Thisis called a tripod gait, which in turn creates a triangle of stability, see Figure 2.2.

3

CHAPTER 2. THEORY

Figure 2.1. The different segments, head, thorax and abdomen of the ant body ispresented above [14].

If the center of mass is not maintained within the support polygon, the whole bodycan risk tipping over. Furthermore, a good balance can also be obtained by keepingthe center of gravity of the body low or near ground level [5].

Figure 2.2. The three legs or contact points on the ground, create a blue arearepresenting the triangle of stability [15]. Edited in Adobe Illustrator.

2.3 Slider Crank MechanismA slider crank mechanism is used when there is a need for conversion between rotaryand linear motion. The system consists of a crank, a slider and a linkage betweenthem called a rod (see Figure 2.3). Moreover, the rod, being attached to both thecrank and the slider, could rotate freely around its joint with respective part. Whenthe crank is rotating, the rod transfers motion to the slider, making it slide between

4

2.4. MATHEMATICAL MODEL OF THE SYSTEM

a fixed plane. The maximum distance that the slider could move is equal to thecrank-disk diameter [6].

Figure 2.3. The principle of slider crank mechanism. [7].

2.4 Mathematical model of the systemBehind most mechanisms, there exists mathematical models which are often sim-plified but can be used in practice. The lifting and gripping procedures make useof different principles presented below.

2.4.1 End Effector

Using force analysis and Newton’s law of motion (see Equation 2.1), the frictionforce between the mandibles’ and the lifted object’s surfaces can be derived [8].

FR = ma (2.1)

• FR is the accelerated force.

• m is the mass of the object.

• a is the acceleration of the object.

The forces acting on an object that is being raised by an end effector, are shownin Figure 2.4. If the system is in equilibrium, a force analysis can be undertakenaccording to the calculations below. As seen in Equation 2.3, the total reactionforce, Ntotal, from the end effector is directly proportional to the friction forcebetween the surface of the weight and the end effector. Also, the normal force,

5

CHAPTER 2. THEORY

Figure 2.4. The red cylinder, representing a weight, gets exposed to a gravitationalforce mg, the acceleration force FR and reaction forces from the end effector’s surfaceNtotal. [9] Edited in Microsoft Paint.

Ntotal, and the acceleration force, FR, must be greater than the mass of the objectits lifting (see Equation 2.2).

FR +Ntotal > mg (2.2)

Ff = nµNtotal (2.3)

• Ntotal is the lifted object’s total normal force.

• g is the gravitational constant.

• Ff is the friction force for gripping and holding the weight.

• n is the amount of friction surfaces.

• µ is the friction coefficient between the gripper’s and the object’s surfaces.

By combining the expressions from Equation 2.1, 2.2, and 2.3 a formula of thefriction force required to grip and hold an object can be written as

Ntotal >m(g − a)

µn(2.4)

2.4.2 LeverageUsing a lever to facilitate lifting has been used throughout history. According tothe principle of leverage, the ratio between the piston force and the lifting force can

6

2.4. MATHEMATICAL MODEL OF THE SYSTEM

Figure 2.5. The forces F1 and F2 create an equilibrium state with the distance d1and d2 from the fulcrum. Drawn in Microsoft Paint

.

be controlled depending on where the fulcrum, the hinge of the beam, is placed (seeFigure 2.5). To maintain balance of the lever system, the torque that the pistonforce produces has to be equal to the torque that is generated from the lifting force.This statement is presented mathematically in Equation 2.5 [10].

M1 = F1d1 = F2d2 = M2 (2.5)

• M is the torque produced.

• F is the force applied.

• d is the distance between the application points.

2.4.3 Hydraulics and Pneumatics

Hydraulics and pneumatics both use the same principle. They consist of a closedcircuit, containing either liquid or air, and have two movable pistons with differentsurface areas. Applying a force to the smaller piston will generate a greater forcefrom the larger piston. This is possible because there is a constant pressure inhydraulic or pneumatic systems, see the equations below [11].

p1 = F1A1

= F2A2

= p2 (2.6)

A = V

h(2.7)

• F is the force on the piston.

7

CHAPTER 2. THEORY

• A is the area of the piston.

• p is the constant pressure in the circuit.

• V is the transported volume in the system.

• h is the compression height.

2.5 MicrocontrollerMicrocontroller Units (MCUs) are small scale computers, mostly used for control-ling different electronical devices. MCUs consists of a Random Access Memory(RAM), a Read Only Memory (ROM) and a Central Processing Unit (CPU) unlikemicroprocessors which only consists of a CPU. Depending on field of usage, MCUsvary in size and have different peripherals. The way they are used is by connectingcords between a source of power, like a battery, and itself in order to control exter-nal devices, for example a servo motor. To program them, code is first written ona computer and then transferred to the MCU [12].

2.6 Servo MotorA servo motor is a rotary actuator which can control its angle with high precision.When a servo motor is connected to an MCU it will give continuous feedback aboutits position. Controlling leg movement requires high accuracy of the angle whichmakes the servo motor a perfect choice for this application [13].

8

Chapter 3

Demonstrator

This chapter covers the electronic components, the final design of the most impor-tant parts and the software used in this project.

3.1 Problem statementThe following problem statements must be regarded.

• Can the micro servo motors produce enough torque?

• Will the hydraulic system stay leak proof?

3.2 Electronics

3.2.1 ArduinoThe chosen MCU for this project is the Arduino Mega. It is chosen due to the needof 15 output pins to control all servo motors.

3.2.2 Servo motorThe chosen servo motor to control the leg and mandible movements is the microservo SG90. This servo can rotate 180 degrees and is both light weight and compact.

There will be a total usage of 14 micro servos for the robot legs and mandibles.When it comes to the hydraulic system of the hexapod, a servo with a larger torquewas required. Thus, the FEETECH High Torque Servo FS5115M was chosen tocontrol the hydraulic forces. Both types of servo motors described above, comewith different sets of arms and screws for attachment. For further specifications onthe SG90 and FS5115M servo motors, their respective data sheets can be found inAppendix C.

To power the servo motors, the MCU and servo motors are connected to anadjustable voltage outlet. This makes it possible to tweak the voltage to maximizethe torque from the servos without overloading the whole system.

9

CHAPTER 3. DEMONSTRATOR

Figure 3.1. Mandibles and head seen from above in Solid Edge.

3.3 3D-printed parts

The 3D-printed parts are made of PLA plastic, mostly using low infill density at20 percent to make all parts as light weight as possible while still having enoughdurability. All parts were designed and modelled using Solid Edge software. Tomake sure all parts fit practically, imperfections from the 3D printers were correctedmanually, for example drilling holes and trimming the edges of parts to a moresuitable size. Glue were used to secure and stabilize the connections.

3.3.1 End effector

The end effector consists of a pair of mandibles (see Figure 3.1) controlled separatelyby two SG90 servo motors. The rotation causes a gripping motion, making grabbingand dropping loads possible.

3.3.2 Thorax and legs

The leg has been divided into three different parts, inner leg, outer leg, and a servoholder (see Figure 3.2). In addition, these parts are connected to each other withmicro servos, making the legs have two servo motor each. Moreover, the servoholders are designed to easily attach to the thorax of the hexapod. The thorax hasa top and a bottom plate which are connected through four bolts. In between theseplates, cables and the MCU is placed. On top of the thorax, the lever is placedwith its fulcrum in the middle, while the hydraulic system is attached underneaththe bottom plate.

10

3.4. HYDRAULICS AND PNEUMATICS

Figure 3.2. The leg design with two servo motors as joints. Rendered in SolidEdge

3.4 Hydraulics and PneumaticsThe hydraulic and pneumatic system consists of two identical 5 ml syringes con-nected to one 20 ml syringe. The surface area of the two different pistons is 227mm2 and 530 mm2 respectively.

When it comes to connections, three tubes were used together with a pipe di-vider. To avoid leakage and displacement of parts, all the connections and the twosmaller syringes were also hot glued in place. This whole system was mainly fixedunderneath the thorax except for the larger syringe, which was placed at the backof the body, acting visually as an ant’s abdomen. Moreover, the larger hydraulicpiston was connected to the lever to transfer forces to it (see Figure 3.3). Thesystem was also filled with water after the tests with air had been made.

When it comes to controlling the smaller pistons’ motion, the FS5115M servomotor was used due to its high torque. Three crank gears together with two rodsmade it possible to move both syringes at the same time. Figure 3.4 shows how theconnections were made between the smaller pistons and the rotating gear that wasattached to the high torque servo.

3.5 SoftwareThe software used in this project is the Arduino IDE and hence the programmingmake use of the Arduino IDE library servo.h. All servo motors are attached to theircorresponding pin on the Arduino board. Also, divided in pairs of six, all the micro

11

CHAPTER 3. DEMONSTRATOR

Figure 3.3. The smaller syringes (purple ones) are connected with the larger onethrough tubes. The large syringe is in turn connected to the lever. Rendered in SolidEdge.

Figure 3.4. The slider crank mechanism made it possible for the servo motor (at-tached to the middle gear) to control both syringes in the hydraulic system at thesame time. Rendered in Solid Edge.

servos’ movements during the walking process are synchronized. In Appendix A,the complete code of the project can be found.

12

Chapter 4

Results

4.1 Testing

The testing were made on a flat non slippery surface. There was no other envi-ronmental disturbances. The lifted load will consist of small weights of 100 gramseach, using a holder to stack them easily. These weights were one by one put ontop of the hexapod until tilting occurred or until the legs caved in. Each area, thelifting with hydraulics, gripping with mandibles and lastly the static and dynamicstability of the body, was tested five times. It is important to note that the walkingand lifting speed were not regarded in the testing process.

4.1.1 Hydraulics vs Pneumatics

The syringes and tubes were first tested with air, hence making it into a pneumaticsystem. The piston in the larger syringe would make a full movement of 30 mm whenthe smaller pistons were pushed 50 mm each which was made manually. However,when a complete retraction of the smaller plungers was made, there was no outcomeon the larger syringe at all, making it descend 0 mm. This made the head, thatis attached to the larger syringe through the lever, become stationary after onedownward motion.

The same tests were made after filling the tubes with water, making it intoa hydraulic system. For water, the set volume of 10 ml was transported everymovement. This made the smaller plungers move 50 mm each which in turn causedthe larger one to rise and descend 30 mm every time. Therefore, the head ofthe robot could successfully make an up and down movement. During the wholeexperimental process, tubes were continuously checked for leakage and after thetesting period of two weeks, the whole hydraulic system managed to successfullystay leak proof.

According to formula 2.6 the amplified force from the large piston is 1.16 timesthe force applied to the small pistons. Considering that the beam was 250 mm long,its fulcrum point was put 125 mm away from the larger syringe’s connection. This

13

CHAPTER 4. RESULTS

in theory produces a force gain of 1.16. The total load the hexapod could carry byfirst using pneumatics and then hydraulics is presented under Final Results.

4.1.2 Gripping procedureThe mandibles needed to rotate to a certain degree to grab the object. The frictionforce from the mandibles was enough to grip the object, however due to poor con-nection between the micro servos and the servo arms, the mandibles could not holdthe added load properly. Therefore, testing and calculating the friction force onthe mandibles were not possible. During weight lifting experiments, the loads werestacked on top of the hexapod’s head instead of letting the mandibles grip them.

4.1.3 Stability and movementAt first the ability to balance the whole body on three legs was tested. A few testswere made to find the right servo holder placement. There was only one placementof the servo holder that could bear the weight of the hexapod while standing onthree legs and that was then they were placed symmetrically on either side of thethorax with the same distance between each other. In Final Results the numbersof the total weight that the robot could bear by using this servo holder placing, ispresented.

When it came to walking, all legs were able to move independently in the ex-pected pattern without touching any surface. However, as soon as the robot wasput down, the actual walking motion failed since the micro servos began behavingincorrectly by making noises and bearly being able to move the whole body at all.

4.1.4 Final ResultsThe final weight of the ant robot, excluding the power source, ended up being 972g when the tubes were filled with air and 980 g with water. By making the headlift weights on top of its head, the strength of the pneumatics and hydraulics couldbe tested. When the construction used a pneumatic system to raise the head, it didnot move upwards at all. Therefore, when air was used as a medium in the tubes,the hexapod could lift 0 g weight while standing on all six legs. The outcome ofusing water instead of air gave different results. A maximum of 1000 g were able tobe lifted with the lever and hydraulic system combined.

By adding load on top of the thorax, the strength of the legs was tested. Thetilting point was also experimented with by loading the head with weights to pushthe total center of mass more to the front of the construction. Furthermore, onlywater was used in the tubes while testing the robot balance and leg strength to keepconsistency.

Firstly, the robot was put on all six legs with parallel inner legs like Figure 4.1.The average weight that the construction could hold was then calculated to be 2250g when load was put on the main body. When it came to only having three contactpoints, the ant robot could not balance properly and therefore the carried weight

14

4.1. TESTING

Figure 4.1. The inner legs placed horizontally seen from above. Rendered in Solid-Edge.

resulted in 0 grams on the thorax. Afterwards the tilting point was tested by addingweights on top of the head without moving it. Tipping started to happen when thehead bore about 700 g while standing on all six legs. Since the whole constructioncould not even remain stable on three legs while having the inner legs placed likein Figure 4.1, the tipping point for this leg set-up was unable to be tested.

Secondly, by rotating the inner front and back legs slightly outwards (see Figure4.2), the ant robot could bear about 2950 g on the thorax while standing on six legsand 2200 g while balancing on three legs. With loads stacked on top of the ant’shead, 1200 g were instead added before the robot started to tilt. Unlike the firstcase, the hexapod with angled legs held up to 500 g on the head while balancing onthree legs.

Lastly, the hexapod could lift 1.02 times its own weight thanks to the hydraulicsystem and hold up to 3.01 times the hexapod weight when the loads were stackedon top of the thorax.

15

CHAPTER 4. RESULTS

Figure 4.2. The front and the back inner legs slightly angled seen from above.Rendered in SolidEdge.

16

Chapter 5

Discussion and Conclusion

5.1 Result expectation and analysis

What was expected? Theoretically the forces should have gotten amplified 1.11times in the lifting process. Since pushing the syringes was made manually, it ishard to draw a conclusion whether this part of the theory was confirmed or not.Also, using air as a medium instead of water ended up being unsuccessful which isprobably because of compression. As seen in the results from the testing, retractionof the smaller plungers did not give any effect on the larger piston when the hexapodused a pneumatic system. Taking this into consideration, water was more suitableas a medium. Nevertheless, the disadvantage of using water in this project is thatwater weighs more than air, making the overall construction heavier which wasundesired considering the project goals. Other than these problems, the hydraulicsystem worked as expected with no leakage.

When it came to designing the parts, the time used for the 3D-printing tooklonger than expected. Some parts had to be reprinted at multiple occasions whichtook long time. Moreover, many printed parts did not appear to have the exactdimensions as the sketches made in Solid Edge. Almost all parts were optimizedwith the lowest and still durable infill density at 20 percent but some parts stillended up being less steadfast than expected. Parts like the beam was bending inthe testing process while lifting heavy weight. Other printed parts, such as the legsand the main body did not deform at all.

The best balance was achieved with the legs partly angled both when it came tostacking the weights on top of the thorax or putting them on the head. The theoryof the stability triangle suggests better balance if the contact points on the groundare far from each other and therefore the outcome of placing the servo holders andthe inner legs more wide apart was expected.

What went wrong? The linkage between the crank disk and the syringe pledgewas not durable and thus broke during multiple occasions while testing. Therefore,to keep consistency, all hydraulic and pneumatic tests that were analyzed in this

17

CHAPTER 5. DISCUSSION AND CONCLUSION

report, were made manually without any influence from servos.There were similar weak points for the construction independent of where the

legs were placed. When the maximum weight was almost reached, the connectionbetween the servo motor and outer leg was poor. Considering this, the results ofthe stability tests, especially the tests of leg strength and tilting point, could haveended up differently if all parts of the hexapod were error free.

When it came to the mandibles, there was an issue with weak connection betweenthe servo and servo arms. Using both screws and glue to keep them connected wasnot enough and the mandibles started bending as soon as any load was added. Thatis why the mandibles were almost impossible to test.

What can be improved? As mentioned above, the weakest points were themicro servo’s connection to its arms. A possible improvement is to either change toa more durable servo motor or using a different design for the connection. Havingmost of the parts glued together is not optimal, removing the possibility to changeand upgrade different parts. The connection could instead use only screws butwould in turn add more weight to the construction.

Instead of having two small syringes and one large it would be much easier tohave one medium sized and one large. This would not only remove the need fora pipe divider, but it would also remove the need to have two extra crank disksthat control one syringe each (see Figure 3.4). This would have made the overallhexapod weigh lighter. Since the slider crank system was placed on the front ofthe construction, the center of mass was closer to the head than to the middle ofthe body. In addition, the hexapod tended to tilt more easily forward while beingexposed to extra weight on the head. To avoid this problem, the whole slider cranksystem could be placed further back of the construction.

5.2 Future WorkEven though this project did not go as expected in many areas, the flaws whichcontributed to failed testing could easily be avoided. With enough testing and theright equipment, this hexapod could work properly and resemble the physiques andnature of ants even more than what was achieved in this project. The ArduinoCode is simple but yet there is room for adjustments, for example by adding morecomplex motions or walking patterns.

Taking some inspiration from this project, a lot of the theories presented herecan be tested on bigger machines that require stability and the ability to lift.

18

Bibliography

[1] Nguyen, Vienny and Lilly, Blaine and Castro, Carlos. The exoskeletal structureand tensile loading behavior of an ant neck joint. Journal of Biomechanics 47(2):p.497-504, 2014.

[2] Ueno, T and Sunaga, N and Asada, H. Mechanism and control of a dynamiclifting robot. Proceedings of the 13th ISARC, Tokyo, Japan p.95-102, 1996.

[3] Monkman, G. J.; Hesse, S.; Steinmann, R.; Schunk, H. Robot Grippers. Wiley-VCH p. 62. 2007.

[4] Moll, Karin and Roces, Flavio and Federle, Walter. How load-carrying antsavoid falling over: mechanical stability during foraging in Atta vollenweiderigrass-cutting ants. 8(1), 2013.

[5] Emam, Mohamed Ali and Marzouk, Mostafa and Shaaban, Sayed. A MONI-TORING DEVICE OF FORKLIFT’S STABILITY TRIANGLE. 43(2), 2017.

[6] [Online]. Encyclopaedia Britannica, Slider-crank mechanism, April 11th 2016.Available: https://www.britannica.com/technology/slider-crank-mechanism[Accessed: 31-05-2019].

[7] [Online]. Available: http://www.technologystudent.com/cams/crkslid1.htm[Accessed: 31-05-2019].

[8] Fantoni, G., Santochi, M., Dini, G., Tracht, K., Scholz-Reiter, B., Fleischer, J.,Lien, T.K., Seliger, G., Reinhart, G.,Franke, J., Hansen, H.N., Verl, A. Graspingdevices and methods in automated production processes. 63(2): p.679-701, 2014.

[9] [Online]. Available: http://www.honeybeerobotics.com/portfolio/door-opening-gripper/ [Accessed: 28-05-2019].

[10] Uicker J. Joseph, Pennock Gordon R., Shigley Joseph E., Theory of Machinesand Mechanisms, Oxford University Press, USA, 4th edition, 2010.

[11] Parr, Andrew. Hydraulics and pneumatics: a technician’s and engineer’s guide.UK: Elsevier,3rd edt, 2011, p.2.

19

BIBLIOGRAPHY

[12] Arduino. Arduino Mega. [Online]. Available: https://store.arduino.cc/mega-2560-r3. [Accessed: 29-04-2019].

[13] Sawicz, Darren Hobby servo fundamentals, 2001.

[14] [Online]. Available: https://wehavekids.com/education/Learn-About-Ants-and-Ant-Colonies-for-Kids [Accessed: 28-05-2019].

[15] [Online]. Available: http://jeb.biologists.org/content/jexbio/212/18/2893/F2.large.jpg[Accessed: 28-05-2019].

20

Appendix A

Flow Chart and Arduino Code

21

/*//////////////// INTRODUCTION /////////////////////////////// University - Royal Institute of Technology, KTH Course - MF133X ,Degree Project in Mechatronics TRITA ITM-EX 2019:29 Authors - Martin Schröder and Faiza Ali Name of the project - W.A.N.T Last edit: 29-04-2019//////////////////////////////////////////////////////////////*/

//Include the Servo library#include <Servo.h>

//Declaring servos for outer legs (odd numbers) Servo servo1; Servo servo3; Servo servo5; Servo servo7; Servo servo9; Servo servo11;

//Declaring servos for inner legs (even numbers) Servo servo2; Servo servo4; Servo servo6; Servo servo8; Servo servo10; Servo servo12;

//Declaring servos for mandibles Servo servo40; Servo servo41;

//Declaring servo for hydraulic motion// Servo servo30; void setup() {

//Attach each outer servo to their corresponding pin servo1.attach(13); servo3.attach(3); servo5.attach(5); servo7.attach(7); servo9.attach(9); servo11.attach(11);

//Attach each inner servo to their corresponding pin servo2.attach(2); servo4.attach(4); servo6.attach(6); servo8.attach(8); servo10.attach(10);

//Attach the mandibles- and hydraulic- servos to their corresponding pin servo40.attach(40); servo41.attach(14);// servo30.attach(30);}////////////////////////////////////////////////////////////////////////////////////////void loop() {

//Outer legs' initial position servo1.write(75); servo3.write(95); servo5.write(45); servo7.write(10); servo9.write(45); servo11.write(75);

//Inner legs' initial position servo2.write(80); servo4.write(110); servo6.write(90); servo8.write(90); servo10.write(90); servo12.write(70);//Mandibles' initial position servo40.write(80); servo41.write(80); delay(2000);

//Hydraulic servos initial position// servo30.write(0);

///////////////////////////////////////////Gripping motion servo40.write(120); servo41.write(30);

delay(1500);

//////////////////////////////////////////Lifting motion// servo30.write(180);

// delay(1500); ///////////////////////////////////////////Walking motion//Raise the first set of three outer legs servo1.write(20);

servo9.write(95); servo5.write(95);

delay(1500);

//Turn the first set of three inner legs forward servo2.write(60); servo6.write(70); servo10.write(110); delay(1500);

//Lower the first set of three outer legs servo1.write(75); servo5.write(45); servo9.write(45);

delay(1500);

//Raise the second set of three outer legs

servo3.write(40); servo7.write(65); servo11.write(20);

delay(1500);

//Return the first set of three inner legs to initial position servo2.write(80); servo6.write(90); servo10.write(90);

delay(1500);

//Turn the second set of three inner legs servo4.write(90); servo8.write(110); servo12.write(115);

delay(1500);

//Lower the second set of three outer legs servo3.write(95); servo7.write(10); servo11.write(75);

delay(1500);

//Raise the first set of three outer legs servo1.write(20); servo9.write(95); servo5.write(95);

delay(1500);

//Return the second set of three inner legs servo4.write(110); servo8.write(90); servo12.write(70);

delay(1500);

//Lower the first set of three outer legs servo1.write(85); servo5.write(35); servo9.write(35);

delay(1500);

//////////////////////////////////////////Descending motion of lever// servo30.write(0); // delay(1000);

///////////////////////////////////////////Opening of mandibles servo40.write(80); servo41.write(80);

delay(1500); }

APPENDIX A. FLOW CHART AND ARDUINO CODE

26

Appendix B

CAD construction

27

APPENDIX B. CAD CONSTRUCTION

28

Appendix C

Electric Circuit

29

Appendix D

Servo Motor Data Sheets

31

SERVO MOTOR SG90 DATA SHEET

Tiny and lightweight with high output power. Servo can rotate approximately 180 degrees (90 in each direction), and works just like the standard kinds but smaller. You can use any servo code, hardware or library to control these servos. Good for beginners who want to make stuff move without building a motor controller with feedback & gear box, especially since it will fit in small places. It comes with a 3 horns (arms) and hardware.

Position "0" (1.5 ms pulse) is middle, "90" (~2ms pulse) is middle, is all the way to the right, "-90" (~1ms pulse) is all the way to the left.

FEETECH RC Model Co.,Ltd. 产品规格书

Specification of Product V1.0 Page 1/2

产品名称：6V 15公斤模拟舵机 Product Name：6V 15kg.cm Analog Servo

产品型号 Model No.FS5115M

1.使用环境条件 Apply Environmental Condition

No. Item Specification

1-1 存储温度 Storage Temperature Range -30℃～80℃

1-2 运行温度 Operating Temperature Range -15℃～70℃

2.测试环境 Standard Test Environment

No. Item Specification

2-1 温度 Temperature range 25℃ ±5℃

2-2 湿度 Humidity range 65%±10%

3.机械特性 Mechanical Specification

No. Item Specification

3-1 尺寸 Size A：40.8mm B：20.1mm C：38mm D：49.5mm

3-2 重量 Weight 58.5g

3-3 齿轮类型 Gear type Metal Gear

3-6 机构极限角度 Limit angle 200°±5°

3-7 轴承 Bearing 2 Ball bearings

3-8 出力轴 Horn gear spline 25T

3-9 摆臂 Horn type Plastic,POM

3-10 外壳 Case Nylon & Fiberglass

3-11 舵机线 Connector wire 300mm ±5 mm

3-12 马达 Motor Metal brush motor

3-13 防水性能 Splash water resistance NO

4.电气特性 Electrical Specification (Function of the Performance)

No. 工作电压 Operating Voltage Range 4.8V 6V

4-1* 静态电流 Idle current(at stopped) 5mA 7mA

4-2* 空载速度 No load speed 0.18sec/60° 0.16 sec/60°

4-3* 空载电流 Runnig current(at no load) 160 mA 190 mA

4-4 堵转扭矩 Peak stall torque14kg.cm 15.5kg.cm

194.8oz.in 215.6oz.in

4-5 堵转电流 Stall current 1200 mA 1500mA

Note: "*"definition is average value when the servo runing with no load

TRITA ITM-EX 2019:29

www.kth.se