Biomechatronic Hand
Thread Rating:
  • 0 Vote(s) - 0 Average
  • 1
  • 2
  • 3
  • 4
  • 5
electronics seminars
Active In SP
**

Posts: 694
Joined: Nov 2009
#1
28-12-2009, 04:01 PM



.doc   biomechatronic.DOC (Size: 674 KB / Downloads: 202)

.doc   biomechatronic report.DOC (Size: 823 KB / Downloads: 228)


ABSTRACT

An "ideal" artificial hand should match the requirements of prosthetics and humanoid robotics.

It can be wearable by the user which means that it can be perceived as part of the natural body and should replicate sensory-motor capabilities of the natural hand.However, such an ideal bionic prosthesis is still far from reality.

This paper describes the design and fabrication of a novel artificial hand based on a "biomechatronic" and cybernetic approach. The approach is aimed at providing "natural" sensory-motor co-ordination, biomimetic mechanisms, force and position sensors, actuators and control, and by interfacing the hand with the peripheral nervous system.


Chapter I
INTRODUCTION

The objective of the work describe in this paper is to develop an artificial hand aimed at replicating the appearance and performance of the natural hand the ultimate goal of this research is to obtain a complete functional substitution of the natural hand. This means that the artificial hand should be felt by the user as the part of his/her own body (extended physiological proprioception(EPP) ) and it should provide the user with the same functions of natural hand: tactile exploration, grasping , and manipulation (cybernetic prosthesis). Commercially available prosthetic devices, as well as multifunctional hand designs have good (sometimes excellent) reliability and robustness, but their grasping capabilities can be improved. It has been demonstrated the methodologies and knowledge developed for robotic hands can be apologies and knowledge developed for robotic hands can be applied to the domain of prosthetics to augment final performance. The first significant example of an artificial hand designed according to a robotic approach is the Belgrade/USC Hand. Afterwards, several robotic grippers and articulated hands have been developed, for example the Stanford/JPL hand and the Utah/MIT hand which have achieved excellent results. An accurate description and a comparative analysis of state of the art of artificial hands can be found in. These hands have achieved good performance in mimicking human capabilities, but they are complex devices requiring large controllers and their mass and size are not compatible with the strict requirements of prosthetic hands.

In fact, the artificial hands for prosthetics applications pose challenging specifications and problems, as is usually the case for devices to be used for functional replacement in clinical practice. These problems have forced the development of simple, robust, and reliable commercial prosthetic hands, as the Otto Brock Sensor Hand prostheses which is widely implanted and appreciated by users. The Otto Bock hand has only one degree of freedom(DOF), it can move the fingers at proportional speed from 15-130 mm/s and can generate grip force up to 100 N.

According to analysis of the state of art, the main problems to be solved in order to improve the performance of prosthetic hands are

1) lack of sensory information gives to the amputee;
2) lack of natural command interface;
3) limited grasping capabilities;
4) Unnatural movements of fingers during grasping.

In order to solve these problems, we are developing a biomechatronic hand, designed according to mechatronic concepts and intended to replicate as much as possible the architecture and the functional principles of the natural hand.

The first and second problems can be addressed by developing a natural interface between the peripheral nervous system (PNS) and the artificial device (i.e., a natural neural interface (NI) to record and stimulate the PNS in a selective way. The neural interface is the enabling technology for achieving ENG-based control of the prostheses, i.e., for providing the sensory connection between the artificial hand and the amputee. Sensory feedback can be restored by stimulating in an appropriate way userâ„¢s afferent nerves after characterization of afferent PNS signals in response to mechanical and proprioceptive stimuli. The biomechatronic design process described above is illustrated in the scheme depicted in Fig.1.



Fig. 1: The control scheme for wearable artificial hands


The research described in this paper is focused on the third and the fourth points. In general, cosmetics requirements force to incorporate the entire device in a glove and to keep size and mass of the entire device comparable to that of the human hand. It turns out that the combination of robust design goals, cosmetics, and limitation of available components, can be matched only with a drastic reduction of DOFâ„¢s, as compared to those of the natural hand. In fact, in prosthetic hands active bending of joints is restricted only to two or three joints (metacarpo-pha-langeal joints of the thumb, of the index and of the middle finger), while the other joints are fixed. Due to the lack of DOFâ„¢s prostheses are characterized by low grasping functionality and, thus they do not allow adequate encirclement of objects in comparison to the human hand ; low flexibility and low adaptivity of artificial fingers leads to instability of the grasp in presence of an external perturbation, as illustrated in. In conclusion, commercial prostheses have been designed to be simple, robust and low cost, at the expense of their grasping ability.

This paper presents a novel multi-DOF hand several active joints, which is designed to obtain better grasping performance and natural fingers movements. The hand is designed according to a biomechatronic approach: miniature actuators and Hall-effect position sensors are embedded in the hand structure in order to enable the control of available DOFâ„¢s. This paper describes a prototype of the artificial hand which has been designed, fabricated, and tested in vitro, in order to assess the feasibility of the proposed approach.





Chapter -II
DESIGN OF THE BIOMECHATRONIC HAND

2.1 BIO MECHATRONIC DESIGN

The main requirements to be considered since the very beginning of prosthetic hand design are the following: cosmetics, controllability, noiselessness, lightness, and low energy consumption. These requirements can be fulfilled by an integrated design approach aimed at embedding different functions within a housing closely replicating the shape, size and appearance of human hand. This approach can synthesized with the term: biomechatronic design.

2.2 ARCHITECTURE OF THE BIOMECHATRONIC HAND

The design goal of the biomechatronic hand is to improve to some extent one of the most important limitations of current prosthetic hands (no dexterity and no adaptability), while preserving the main advantages of such hands, that is lightness and simplicity. This objective has been pursued by using small actuators(two of each finger) instead of one single large actuator( as in most current prosthetic hands) And by designing a kinematics architecture able to provide better adaptation to object shape during grasping. It turns out that the use of micro motors allows to augment functionality in grasping objects by means of human-like compliant movements of fingers. This result addresses the very basic requirements of cosmetic appearance of the hand in static and dynamic conditions.


The biomechatronic hand has three fingers to provide a tripodgrasp: two identical fingers(index and middle fingers) and the thumb(see Fig.2)


Fig. 2. Architecture of the biomechatronic hand

In fact, as explained in, at least three fingers (non rolling and non sliding contact) are necessary to completely restrain an object.
The hand performs two grasping tasks:
1) Cylindrical grasp
2) Tripod grasp

The finger actuation system is based on two microactuators which drive the meta carpophalengal (MP) and the proximal interphalengal (PIP) joint. The thumb actuation system is based on microactuators and has two active DOFâ„¢s at the MP and the interphalengeal (IP) joint, respectively.

The grasping task performed by the hand compromises two subsequent phases:
¢ Reaching and shape-adapting phase
¢ Grasping phase with thumb opposition.

In phase one ,the first actuator system allows the finger to adapt to the morphological characteristics of the grasped object. In phase two, the second actuator system provides thumb opposition for grasping.

In section III, the basic criteria for designing the actuation system according to biomechatronic approach are described.

2.3 ACTUATION SYSTEM

The adoption of bulk and heavy actuators in the design of commercial upper limb prostheses, leads to an extreme reduction of DOFâ„¢s. The goal is to achieve stable grasp by means of high grip forces. This design philosophy can be represented as a loop (see Fig.3)



The above schematization shows how this approach leads to design hands with a maximum of two DOFâ„¢s and able to obtain stable grasps using high pinch force (about 100N). To summarize, mechanical grippers such as state of art prosthetic hands, can generate large grasping forces and are simple to implement and control, but they are not adaptable and may cause problems of low grasping stability.
The approach we propose (see Fig.4) to invert the previous loop by using microactuators and by exploiting the advantage of increasing DOFâ„¢s.

According to the design philosophy, an artificial hand actuated by a plurality of microdrives would have enhanced mobility and, thus, larger contact areas between phalanges and grasped object. Therefore, a reduction of power actuation could be accepted and compensated by increasing contact areas in order to augment grasp stability. In fact according to a hand with independently movable fingers and multiple phalanges can encircle the object much better than a hand with rigid fingers. In addition, the contact area between an object and the finger can be larger and, thus, grasping stability is enhanced.









Chapter -III
DESIGN OF HAND PROTOTYPE

In order to demonstrate the feasibility of the described biomechatronic approach, we have developed a three fingered hand prototype with two identical fingers (index and middle) and thumb. Actuators, position sensors and a 2-D force sensor are integrated in the hand structure.

The index/middle finger has been designed by reproducing, as closely as possible, the size and kinematics of a human finger. Each finger consists of three phalanges and a palm needed to house the proximal actuator (see fig 5).


Fig 5. Detail drawing of middle finger



3.1 ACTUATOR SYSTEM ARCHITECTURE

In order to match the size of a human finger, two micro motors have been integrated within the palm housing and the proximal phalange of each finger.

The selected micromotors are Smoovy (RMB, eckweg, CH) microdrivers (5mm diameter) high precision linear actuators, based on the dc brushless motors with planetary gears. The rotary motion of the shaft is converted to linear motion using lead screw transmission.

The main mechanical characteristics of the linear actuators are listed below (see table I).

TABLE I
Summary of the main characteristics of the SMOOVY (RMB, eckweg, CH) microdrivers (5mm diameter)

Nominal force 12 N
Maximum Speed 20 mm/s
Weight 3.2 g
Maximum load (axial) 40 N
Maximum load (radial) 25 N
Transmission rate 1:125
Gear stages 3



The selected actuator fulfills almost all the specifications for application in the prosthetic finger: small size and low weight. The main problem encountered is related to noise which turns out to be relatively high, at least in the current implementation. Despite of his limitation, we decided to proceed with the application of the linear actuator in order to investigate integration problems and global performance.

The shell housing provides mechanical resistance of the shaft to both axial and radial loads system. This is very important during grasping task, when the forces generated from the thumb opposition act on the whole finger structure

3.2 KINEMATICS ARCHITECTURE

The kinematics of each finger joint is described in the following subsections
1. MP joint: the proximal actuator is integrated in the palm and transmits the movement through a slider- crank mechanism to the proximal phalanx, thus , providing flexion/extension movement. The slider is driven by the lead screw transmission mounted directly on the motor shaft.
2. PIP joint: the same mechanism used for mp moves the pip joint. Only the geometrical features in order that the size of the mechanism fits within the space available according to the strict specification of the biomechatronic hand.
3. DIP joint: a four bars link has been adopted for the dip joint and its geometrical features have been designed in order to reproduce as closely as possible the natural dip joint flexion. The mechanism has been synthesized according to the three prescribed position method.

Due to the high transmission rate (planetary gears and lead screw transmission) friction is high and, thus, the joints are not back-drivable. This causes problem in controlling accurately in hand. However the positive side effect of the friction is that the grasping forces can be exerted even when power supply is off, a very important function for hand procestheses.

3.3 THUMB DESIGN

The thumb has been designed to perform grasping task by thumb opposition. The thumb has been obtained by simply removing the distal phalanx from the index/middle finger see fig 6 .





Chapter -IV
HAND FABRICATIONS

The hand protected (see fig 7) comprises the three fingers (index middle and thumb), each with two-DOFâ„¢s actuated by micro motors and sensorised by hall-effect position sensors and by strain gage-based force sensors. The characteristics of the position sensors and of the force sensors are illustrated in following sections.

The three fingers have been fabricated using the Fused Deposition Modeling (FDN) process. This process allows obtaining 3-E complex shapes from CAD models easily, quickly and cheaply. The main limitation of the FDM process resides in poor mechanical characteristics of the material that must be used, which is acrylonitrile/butadiene/styrene (ABS).how ever; this is acceptable for a prototype.

Fig 7. Hand prototype in grasping task


Chapter V
POSITION AND FORCE SENSORS

5.1 SENSORS

Sensors are used as peripheral devices in robotics include both simple types such as limit switches and sophisticated type such as machine vision systems. Of course sensors are also used as integral components of the robots position feed back control system. Their function in a robotic work cell is to permit the robotics activities to be co-ordinate with other activities the cell.

(1) TACTILE SENSOR: These are sensors, which respond to contact forces with another object; some of these devices are capable of measuring the level of force involved.

(2) PROXIMITY AND RANGE SENSOR: A proximity sensor that indicates when an object is close to another object but before contact has been made. When the distance between the objects can be sensed, the device is called a range sensor.

(3) MISCELLANEOUS TYPES: The miscellaneous category includes the remaining kinds of sensors that are used in robotics.

(4) MACHINE VISION: A machine vision systems is capable of viewing the workspace and interpreting what it sees. These are used in robotics to perform inspection, part recognition and other similar tasks.



5.2 USE OF SENSORS IN ROBOTICS

The major uses of sensors in robotics and other automated manufacturing systems can be divided into four basic categories.

(1) Safety monitoring.
(2) Inter locks in work cell control.
(3) Part inspection for quality control.
(4) Determining positions and related information about objects in the robot cell.

5.3 POSITION SENSORS

A position sensor, based on the Hall Effect sensor is mounted at each active joint of the hand. The main advantages the Hall Effect sensors are there small sizes and there contact less working principle. In each finger, the hall sensors are fixed, respectively, to the palm and to the proximal phalanxes, where as the magnets are mounted directly on the sliders of each active joint.

In this configuration the sensor measures the linear movement of the slider, which is related to the angular position of the joint. In each MP joint, the linear range of the sensor is 5.2mm, where as in the PIP joint the linear range is 8mm.

Using a micrometric translator stage we found to optimal configurations for the position sensors. In the first optimal configurations two magnets are used at a distance of 3.5mm.this configuration has a working range of 5.4mm with a linearity of5.34%.

The second optimal configuration (suitable for MP joints) has six magnets and a working range of 8.4mm with a linearity of 3.81%. A finger prototype with integrated position sensors is showed in fig 8.



Fig. 8 The first prototype with the 2 integrated sensors.

5.4 HALL EFFECT SENSORS

When a beam of charged particles passes through a magnetic field, forces act on the particles and the beam is deflected from its straight line path. A current flowing in a conductor is like a beam of moving charges and thus can be deflected by magnetic field. This effect is known as HALL EFFECT. Consider electrons moving in a conductive plate with a magnetic field placed at right angles to the plane of the plate. As a consequence of the magnetic field, the moving electrons are deflected to one side of the plate and thus that side becomes negatively charged while the opposite side becomes positively charged since the electrons are directed away from it.

This charge separation produces an electric field in the material. The separation continues until the forces on the charged particles from the electric field just balance the forces produced by the magnetic field. The result is a transverse potential difference given by

V=KhBI/t
Where,
B is the magnetic flux density at right angles to the plate, I current through the plate, t the plate thickness, K the constant called Hall Co-efficient. Thus if a constant current source is used with a particular sensor, the hall voltage is a measure of the magnetic flux density.

Hall Effect sensors are generally supplied as in integrated circuit with the necessary signal processing circuitry. There are two basic forms of such sensor, LINEAR where the output varies in a reasonably linear manner with the magnetic flux density and THERSHOLD where the output shows a sharp drop at particular flux density.









LINEAR THRESHOLD


5.5 2D FORCE SENSORS

A 2-D force sensor, based on strain gages technology, has been developed in order to sensorize the digital phalanx of index and middle fingers. The force sensor measures both normal and tangential forces. The sensor design has been optimized using the Pro/Mechanica structure software.

5.6 SENSORS CHARACTERIZATION

1. Characterization of position sensors: we found that the best simplest way to characterize these sensors is use an optical method. Used a Nikon Coolpix 950 digital camera mounted on a tripod in order to record the movement of the finger. The movement of each Smoovy actuator was driven by a CCS00001 controller (RMB, CH).each controller has a power supply of 11V,.while each sensor was supplied with 6V.

For each active joint 100 different frames, 50 for flexion and 50 for extension movements where acquired. For each frame the output value of the sensor was measured with a digital multimeter and recorded, where as the position of the joint was measured using the module measures to Adobe Photoshop 5.5 with a precision of 0.1o .

Results are presented in fig 9. For the sensor in MP joints and in the PIP joints, respectively. The flexion phase is indicated with a small dark circles, while the extension is indicated small light squares.

It is important to point out the both curves for both sensors generally present low hysterieses. The difference between the flexion and the extension curves is mainly due to the mechanical clearance of the sensorised slider.

2. Characterization of 2-D 4 sensor: the force sensor was characterize using an INSTRON 4464 testing machine.

A traction-compression loading cycle (0N-10N-0-N) was performed for each direction. Results are presented in fig 11, for the normal loading direction and the tangential loading direction respectively. Diagram show a linear behavior of the 2-D force sensor.


Chapter VI
FINGERED TRIP FORCE ANALYSIS

A first set of experimental test has been performed in order to evaluate the force that the index/middle finger is able to exert on an external object. To this aim we are measured the force resulting when the finger is pressing directly on the high accuracy piezoelectric load cell corresponding to different configuration of the joints.

To pressing task were identify in order to evaluate separately and independently the force generated by actuators of the fingers.

TASK1: the pushing action is exerted only by the distal actuator
TASK2: the pushing action is exerted only by the proximal actuator
The ten test where performed for each sub task. The result obtained is illustrated in fig 11.

Fig.11 Experimental results of tests aimed at evaluating force performance of the biomechatronic fingers

Chapter VII
FUTURE IMPROVEMENTS

The experimental tests showed promising results, but there is still room for improvement. First of all, natural fingers movements during grasping activities will be further investigated in order to achieve a truly human-like behaviour of the artificial finger. The force sensor measurements will be further investigated in order to sense incipient slippage and to obtain force sensing abilities. Finally, suitable control strategies will be investigated and applied in order to develop a natural control of the wearable hand.



Chapter VIII
CONCLUSIONS

In this paper, novel approach to the design and fabrication of prosthetic hands, called biomechatronic design, has been presented. The biomechatronic design consists of integrating multiple DOFâ„¢s finger mechanisms, multi sensing capabilities and distributed control in order to obtain human like appearance, simple and direct controllability and low mass. The biomechatronic design approach can lead to the development of hand and prostheses, when combined with other important factors, such as low energy consumption for adequate autonomy (at least eight hours between recharges), noiseless operation fore not disrupting social interactions, cost suitable for support by the health insurance system and above all sensory feedback to the amputee through interfaces. A biomechatronic hand prototype with three fingers and a total of six independent DOFâ„¢s has been designed and fabricated. This paper is focused particularly on the analysis of the actuation system, which is based on miniature electromagnetic motors.

Current work in our lab is directed to improve the limitations of the prosthesis is presented in this paper. First of all, a new design has been devised aimed at increasing the grasping the force of the hand while retaining the main positive characteristics of previous design. The new hand architecture is based on under actuated mechanisms, comprising a total of two dc motors. The hand has nine non independent DOFâ„¢s (so it can still grasp rather effectively objects of complex shape) and can generate grasping force of about 30 N.


A second important objective that we are pursing is to implement a neural of the hand by means of interfaces implanted at peripheral nerves of the amputee. This very challenging goal could ultimately lead to the development of a truly cybernetic hand, controlled and received by the amputee almost at his /her own lost hand and, therefore, a real potential alternative to hand transplantation.



Chapter IX
REFERENCES

1. CHAIARA@MAIL-ARTS.SSSUP.IT
2. SCHILLING, FUNDAMENTALS OF ROBOTICS, PRENTICE HALL INDIA, PAGE NO.: 43-50, 117-150.
3. W.BOLTON, MECHATRONICS, ADDISON WESLEY LONGMAN LIMITED, PAGE NO. 1-12, 33-35, 172, 176-178.
4. GROOVER, INDUSTRIAL ROBOTICS, MCGRAW-HILL INTERNAL EDITION INDUSTRIAL ENGINEERING SERIES, PAGE NO. 32-46, 60-83, 144.
5. S.R.DEB, ROBOTICS TECHNOLOGY AND FLEXIBLE AUTOMATION, TATA MCGRAKI HILL PUBLISHERS,
PAGE NO. 152.


ACKNOWLEDGEMENT

I am very grateful to almighty God who led me in the right way to attain this goal.

I express my sincere gratitude to Dr: T.N. SATHYANESAN,
Prof. & Head, Department of MECHANICAL ENGINEERING, MES College of Engineering, Kuttippuram, for his cooperation and encouragement.

I also at the outset thank and express my sincere gratitude to my seminar and presentation guide Mr.: ALEX BERNARD (Lecturer, Department of ME), and Mr. KRISHNAKUMAR (Lecturer, Department of ME), Asst. Prof.
Mrs. JUMAILTHU BEEVI D.(Staff in-charge, Department of ME) for their invaluable advice and wholehearted cooperation without which this seminar and presentation would not have seen the light of day.

I am also indebted to all the teaching and non- teaching staff of the department of mechanical engineering for their cooperation and suggestions, which is the spirit behind this report. Last but not the least, I wish to express my sincere thanks to all my friends for their goodwill and constructive ideas.


Saumya G




CONTENTS

1. INTRODUCTION 1

2. DESIGN OF THE BIOMECHATRONIC HAND 5
2.1 BIO MECHATRONIC DESIGN 5
2.2 ARCHITECTURE OF THE BIOMECHATRONIC HAND 5
2.3 ACTUATION SYSTEM 7

3. DESIGN OF HAND PROTOTYPE 9
3.1 ACTUATOR SYSTEM ARCHITECTURE 10
3.2 KINEMATICS ARCHITECTURE 11
3.3 THUMB DESIGN 12

4. HAND FABRICATIONS 13

5. POSITION AND FORCE SENSORS 14
5.1 SENSORS 14
5.2 USE OF SENSORS IN ROBOTICS 15
5.3 POSITION SENSORS 15
5.4 HALL EFFECT SENSORS 16
5.5 2D FORCE SENSORS 18
5.6 SENSORS CHARACTERIZATION 18
6. FINGERED TRIP FORCE ANALYSIS 20

7. FUTURE IMPROVEMENTS 21

8. CONCLUSIONS 22

9. REFERENCES 24

.pdf   biomechatronic help.pdf (Size: 99.94 KB / Downloads: 149)
Use Search at http://topicideas.net/search.php wisely To Get Information About Project Topic and Seminar ideas with report/source code along pdf and ppt presenaion
Reply
project report tiger
Active In SP
**

Posts: 1,062
Joined: Feb 2010
#2
16-02-2010, 08:36 PM


.ppt   MECHATRONIC HAND.ppt (Size: 1.66 MB / Downloads: 148)

Introduction

The objective is to develop an artificial hand aimed at replicating the appearance and performance of the natural hand or to obtain a complete functional substitution of the natural hand.
Main problems to be solved in order to improve the performance of prosthetic hands are:
1) lack of sensory information gives to the
amputee
2) lack of natural command interface.
3) limited grasping capabilities
4) unnatural movements of fingers during grasping.
Reply
seminar addict
Super Moderator
******

Posts: 6,592
Joined: Jul 2011
#3
23-01-2012, 12:28 PM

to get information about the topic Biomechatronic full report,ppt, related topic refer the link bellow

topicideashow-to-biomechatronics

topicideashow-to-biomechatronic-hand--5405

topicideashow-to-biomechatronic-hand-ppt-ppt

topicideashow-to-biomechatronic-hand-full-report
Reply

Important Note..!

If you are not satisfied with above reply ,..Please

ASK HERE

So that we will collect data for you and will made reply to the request....OR try below "QUICK REPLY" box to add a reply to this page

Quick Reply
Message
Type your reply to this message here.


Image Verification
Please enter the text contained within the image into the text box below it. This process is used to prevent automated spam bots.
Image Verification
(case insensitive)

Possibly Related Threads...
Thread Author Replies Views Last Post
  fabrication of automatic hand break release jaseelati 0 276 10-01-2015, 04:26 PM
Last Post: jaseelati
  electricity generated by hand wheel jaseelati 0 327 01-01-2015, 04:46 PM
Last Post: jaseelati
  HAND PORTABLE SCANNER PPT study tips 0 310 31-07-2013, 03:48 PM
Last Post: study tips
  HAND GESTURE : FOR HUMAN MACHINE INTERACTION science projects buddy 4 5,540 15-05-2013, 07:18 AM
Last Post: Guest
  A Study on Hand Gesture Recognition Technique pdf project girl 0 393 16-01-2013, 10:44 AM
Last Post: project girl
  biomechatronic -hand full report seminar tips 0 346 06-12-2012, 04:47 PM
Last Post: seminar tips
  HAND GESTURE RECOGNITION uploader 0 593 30-05-2012, 12:29 PM
Last Post: uploader
  DESIGN AND DEVELOPMENT OF A HAND-GLOVE CONTROLLED WHEEL CHAIR seminar ideas 0 694 19-05-2012, 11:20 AM
Last Post: seminar ideas
  Contactless and Pose Invariant Biometric Identification Using Hand Surface seminar paper 0 385 12-04-2012, 11:39 AM
Last Post: seminar paper
  Tele-Graffiti (A Camera-Projector Based RemoteSketching System with Hand-Based User I computer science crazy 6 5,463 29-03-2012, 10:58 AM
Last Post: seminar paper