Sei sulla pagina 1di 6

Imperial College London

Department of Aeronautics
3rd Year H401 / 4th Year H410

L3 Final Report
Group 1

March 13, 2015

Thomas REES 00642878


Yong Qian YONG 00636496
Joseph GIBBS 00642368
Mark SONES 00647820
Jialin LI 00636125
Robin SPINKS 00644225
Rikesh POONJA 00647775

Contents
1 Introduction

2 Design and Analysis

3 Materials, Manufacturing and Assembly

4 Test Results and Analysis

5 Figures and Tables

Introduction

The design brief for our project was to conceive, design, build, and test a long distance jumping robot
subject to constraints[3]. As part of the preliminary work on this project, the current literature was
reviewed in order to identify possible design paradigms (Table 1). The vast majority of these designs
use linear or torsional springs as energy storage. This provided a good point to start the design
process.
Robot
Hopper [2]
Mini-Whegs [7]
Jollbot [1]
Grillo [10]
Grillo [8]
7g Robot [4]
Closed Elastica [12]

Energy Storage
Linear Spring
Linear Spring
Steel Wire
Linear Spring
Linear Spring
Torsional Spring
Elastic Strip

Storing Mechanism
Lead Screw and Gear
Partially toothed pinion and gear
Face cam and slider roller
Lead screw, gear, and spiral cam
Partially toothed pinion and gear
Spiral Cam and gear
Motor bends elastic strip

Release Mechanism
Latch release
Escapement
Escapement
Escapement
Escapement
Escapement
Buckling

Table 1: Summary of existing robot designs. Adapted from [9]

Design and Analysis

During the design process it was noticed that the EPFL 7g robot [4] uses a four-bar linkage as the leg
mechanism. By changing the lengths of the bars in the mechanism, the force profile applied to the
ground could be designed to an optimum. It was reasoned that by replacing the four-bar mechanism
with a set of curved ski-like legs, where the curvature is chosen to give the optimum ground force
profile, it could be possible to design an optimum ground impulse with a lighter, simpler system. The
final design configuration chosen is presented in Figure 1. Two arms containing a gear and cam system
are connected to two skis via a set of torsional springs. By compressing the springs, the centre of mass
of the system is moved back and the robot rocks back on the skis. When released, the legs push off
the ground as the robot rocks forward giving a forward impulse.
A preliminary analysis of jumping performance was performed using an energy analysis. The ballistic
equations of motion were solved using a 4th order Runge-Kutta solver for various take-off velocities
and masses (Table 2). Using the most conservative estimate of kinetic energy, the estimated jumping
distance of the robot is 0.23 m.
Mass [g]
30

T/O Velocity [m/s]


5

Jump Length [m]


0.23

T/O Angle [deg]


45

Table 2: Estimated jump length


Due to the small form factor of the robot, detailed aerodynamic analysis is considered less important
at this stage of the design. For the jumping performance prediction, a CD of 1.3 was chosen, which is
in line with values used in the literature. Furthermore, very little structural analysis was performed.
The most structurally strained components were the arms, which were manufactured from carbon
fibre tubes and furthermore were reinforced by the arms of the torsional springs. Considering this
stiff construction and the small energies involved, the design and improvements of the mechanism was
decided to be of a higher priority than a detailed structural analysis.

Materials, Manufacturing and Assembly

During the initial fabrication procress, it was decided that the cam and gears would be laser-cut,
while the leg-assembly would be made of ABS plastic and 3D-printed. The other parts were off-theshelf products. These decisions were made based on considerations of structural strength, ease of
customisation and ease of manufacturing. Please refer to Section 5 for the drawing and full list of
parts, materials and fabrication processes of each robot part.
The robot design has 3 distinct sub-systems: motor and electronics, gear and cam assembly, and legs.
Therefore, after the fabrication process of the cam, gears and legs, each subsystem was assembled and
tested separately before being integrated into the final robot.
During this process, the fragile spindle joints were reinforced and thickened (see Figure 1). It was also
decided that the gears would be 3D-printed instead of laser-cut due to its ability to provide greater
resolution to the teeth.

Test Results and Analysis

For the first jump test, it was found that the motor lacked torque. Therefore, the first jump test was
performed with manual actuation. The robot performed a flipping motion, pivoting about the front
edge of the leg assembly. The jumping distance was 1cm, or 4.35% of predicted jumping performance.
After the first jump, an analysis was performed, the issues identified and solutions taken to mitigate
the situation as explained in Table 3 below. These updates and modifications were performed on the
robot prior to the testing day.

1
2
3

Problem
Heavy Robot

Action Taken
Leg assembly length halved to 5cm,

Leg assembly too stable inability to initalize high impulse jump


Lack of torque in motor

legs and steel bar hollowed out


Leg assembly length and
radius of curvature halved
New motor - rated torque of 10mNm

Table 3: Problems with initial jump test and mitigative actions taken
On the final test day, the robot did not jump, achieving 0% of predicted jumping performance.
Nevertheless, there are several key takeaways identified for improvement of the robot:
1. Increasing moment arm about pivot point
The cam and gear assembly, as well as actuation point of springs, should be reconfigured to the
top of the chassis as far away from the pivot of the leg-assembly as possible. This is to allow
for a higher moment arm during the actuation of the jump, compressing the spring further and
allowing for a high-impulse jump.
2. Reduction of friction on contact point
There was an extremely high coefficient of friction between the cam and metal primary axel of
the robot. This, combined with the fact that the springs tended to pinch the cam meant that
unnecessarily high torque was required to turn the cam against the axle. This could be done by
introducing a ball bearing at the point of contact.
3. Increasing efficient transmission of torque
This could be attributed to 2 factors: the low cam differential and the fact that the motor was
insufficiently constrained by the mounting on the CFRP arm. As a result, torque transmission
from the motor to the springs was inefficient and resulted in losses in torque.

Figures and Tables


Number
1,2,3

Part
Gears, Cam, Legs

Material
ABS Plastic

4
5
6
7,8,9

Arms
Gear Spindles and Joints
Springs
Motor, 3V Battery, Receiver

CFRP
Mild Steel
Music Wire
-

Fabrication Process
3D Printed
Off-the-shelf

Figure 1: Robot Parts, Materials and Fabrication Process

References
[1] R. Armour, K. Paskins, A. Bowyer, J. Vincent, and W. Megill. Jumping robots: A biomimetic
solution to locomotion across rough terrain. Bioinspiration Biomimetics, 2(3):S65S82, 2007.
[2] P. Fiorini and J. Burdick. The development of hopping capabilities for small robots. Autonomous
Robots, 14(2-3):239254, 2003.
[3] M. Kovac. L3 introduction presentation slides. Imperial College London Dept. of Aeronautics.
[4] M. Kovac, M. Fuchs, A. Guignard, J.-C. Zufferey, and D. Floreano. A miniature 7 g jumping
robot. In Proc. IEEE Int. Conf. Robot. Autom, pages 373378, 2008.
[5] M. Kovac, M. Schlegel, J.-C. Zufferey, and D. Floreano. A miniature jumping robot with self
recovery capabilities. In Proc. IEEE/RSJ Int. Conf. Intell. Robots Syst., pages 583588, 2009.
[6] M. Kovac, M. Schlegel, J.-C. Zufferey, and D. Floreano. Steerable miniature jumping robot.
Autonomous Robots, 28(3):295306, 2010.
[7] B.G.A Lambrecht, A.D. Horchler, and R.D. Quinn. A small, insect inspired robot that runs and
jumps. In Proc. IEEE Int. Conf. RObot. Autom, pages 12401245, 2005.
[8] F. Li, G. Bonsignori, D. Chen, C. Stefanini, W. Liu, P. Dario, and X. Fu. Jumping mini-robot
with bio-inspired legs. In Proc. IEEE Int. Conf. Robot. Biomimetics, pages 933938, 2008.
[9] Minkyun Noh, Seung-Win Kim, Sungmin An, and Kyu-Jin Cho. Flea-inspired catapult mechanism for miniature jumping robots. IEEE Transactions on Robotics, 2012.
[10] U. Scarfogliero, C. Stefanini, and P. Dario. Design and development of the long-jumping grillo
mini robo. In Proc. IEEE Int. Conf. Robot. Autom, pages 467472, 2007.
[11] U. Scarfogliero, C. Stefanini, and P. Dario. The use of compliant joints and elastic energy storage
in bio-inspired robots. Mech. Mach. Theory, 44(3):580590, 2009.
4

[12] A. Yamada, M. Watari, H. Mochiyama, and H. Fujimoto. A jumping robot based on the closed
elastica. In Proc. Int. Symp. Micro-NanoMechatron. Human Sci., pages 604609, 2007.
[13] A. Yamada, M. Watari, H. Mochiyama, and H. Fujimoto. An asymmetric robotic catapult based
on the closed elastica for jumping robot. In Proc. IEEE Int. Conf. Robot. Autom, pages 232237,
2008.

Potrebbero piacerti anche