Vous êtes sur la page 1sur 5

I MPERIAL C OLLEGE L ONDON

A ERONAUTICAL E NGINEERING - 3rd Y EAR


AE3-418 A PPLICATIONS E XERCISE

L3 Final Report
G ROUP 7
www.imperialjumpingrobot.weebly.com

Minglun DAI
CID: 00731629

Marc E MMANUELI
CID: 00732254

Jan F IGALA
CID: 00731892

Tereze G AILE
CID: 00731562

Mehmood H ASSAN
CID: 00731501

Juan L ARREA
CID: 00732526

Marcus R OSE
CID: 00730955

Supervisor:
Mirko KOVAC

March 13, 2015

Design process and performance modelling


The objective of this L3 Application Exercise is to design, manufacture and test a jumping robot ensuring
it meets all the criteria outlined in the design envelope. Multiple jumping mechanisms were considered
such as the use of torsional and linear springs, pulleys and elastic materials. The final design is inspired
by an airsoft gun due to its simplicity and adaptability to design modifications. It consists of a linear spring
that is compressed by an electric motor through the use of gears. Electrical components are distributed
on both sides for stability and to prevent spinning. Moreover a worm drive was adopted for its large
reduction ratio and a 14 tooth gear is chosen as a robust release mechanism.

(b) Original rough sketch of model

(a) Airsoft pistol - design inspiration

Matlab trajectory
During modelling, a point mass assumption was adopted coupled with an estimated weight of 30 grams.
The equations used for theoretical calculations are as follows taken from the applications handout:



1
y(t)

x
(t) =
ACd cos arctan
(x(t)
2 + y(t)
2)
(1)
2m
x(t)






y(t)

1
2mg + ACd sin arctan
(x(t)
2 + y(t)
2)
(2)
x
(t) =
2m
x(t)

These equations were solved with a 4th order Runge Kutta method. The theoretical take off velocity
was calculated to be 8ms1 - this is twice as much as required. However, this is not possible since
friction losses and inefficiencies of the motor impact the performance. The maximum distanced jumped
is predicted to be 1.75m. The graphs obtained for the trajectory are shown below.

Spring calculations
For the spring, we used the equation 12 mv 2 = 12 kx2 . We had already calculated the robot size and
approximated that required spring compression is 3cm. Subsequently, the equations were arranged to
find out the value of the spring stiffness knowing m, v and x from previous calculations. The value of K
was determined as 0.9N mm1 . This is a fairly high spring compression stiffness and after conducting
tests we decided to go for a spring with a lower K value so that a small motor could compress the spring.

Description of mechanical implementation


Mechanical and electrical components were chosen to meet the requirements based on initial calculations and a compromise between price and market availability. The chosen components include: a
micro-planetary geared motor with gear ratio 1 : 136, 2.4GHz micro receiver, 3.7V River cell at 20mAh
battery, 7.62cm steel spring, MOD0.5 50 tooth worm drive, MOD0.5 rack, carbon fibre rods, 3D printed
motor housing, tube and a syringe plunger.
The first prototype is shown below, weighing 42g.

(b) Original CAD design

(a) Prototype 1

List of prototypes
Prototype 1: Spring was too stiff for motor and manufacturing wasnt accurate due to many separate
glued parts. The inner piston was made from a syringe plunger
Prototype 2: Inner piston designed and printed to hold rack and to reduce friction
Prototype 3: Gear, feet and motor holders printed as one part with the main tube. The spring was
made smaller and less stiff. These modifications reduced the weight from 42g to 26g.
Prototype 4: A worm drive guard was designed to stop the worm drive from moving off axis when
turning the spur gear.

Prototype 5: Feet repositioned (to reduce moment on takeoff)


Prototype 6: 2.4Ghz micro receiver went unserviceable. Replaced with a heavier RF receiver from
a micro RC car.
Prototype 7: Silicone putty was added to feet to remove angular momentum. Improved jump
distance by 300%.
Multiple iterations were made for the external tube, internal piston, electronics housing and motor housing
and due to their complex geometry the only feasible option left was to 3D print these parts. This resulted
in significant weight reduction and better fitting parts. Moreover feet location was optimised for better
grip. The final weight of the manufactured robot was 29.98g which satisfies the design weight constraint.

(a) Prototype 7

(b) CAD design of Prototype 4

Description of robot performance


Our first jump test resulted in a distance travelled of 100mm. Upon carrying out analysis with a 240fps
camera we discovered that upon takeoff the positioning of the foot as shown in the CAD design above
resulted in a moment to be produced. This angular momentum was effectively wasted energy. Upon
repositioning of the feet as shown in prototype 7, the robot jumped 150mm. After carrying out slow
motion analysis it was still discovered to rotate on takeoff: this was due to the lack of friction on the feet.
Once silicon putty was added to the feet the jump distance increased by 300% with no rotation at takeoff
and with a trajectory of 45 .

Best Jump (mm)


295

Robot Size (mm)


90

Robot mass (g)


29.98

Merit Function
4.39

Year Group Positioning


2nd

Table 1: Final Test Results

Futher improvements
The robots performance could have been improved in numerous ways outlined below.
More budget for buying 2 motors: test one motor to its limit for better understanding of maximum
compressible length of the spring.
Remove leg holders from the tube to save weight.
Replace RF receiver with a working 2.4Ghz receiver, which will reduce the weight by 2g.
Reprint tube with hole cutouts for weight saving and use finite element analysis to ensure the
material is not subjected to critical stresses.

Appendix
Costing
Our overall expenditure for the project was 129.32. The project went over budget for a couple of reasons:
There was a minimum order charge and postage cost for the springs. They were ordered before
the workshop announced that they could order in springs for us.
In order to keep the weight down micro digital receivers were bought (28 each) and the cheapest
suitable transmitter was bought (17).
At all points in the project the cheapest parts were obtained. The worm drive and all gears along
with the motor were sourced from Hong Kong.