Transcript MS PowerPoint template

Haptic Manipulation of Microspheres Using
Optical Tweezers (Graphics replaced to keep file size small)
ICON
Ibrahim Bukusoglu, Cagatay Basdogan, Alper Kiraz, Adnan Kurt
Koc University, Istanbul, Turkey
ABSTRACT
Optical Tweezers
We report the manipulation of glass micro spheres having a diameter of 3-10 μm
using optical tweezers and with haptic feedback. We detect the position of a
micro sphere manipulated in a fluid bed using a CCD camera and calculate the
forces acting on it due to the optical trap and viscous drag. We estimate the
optical forces between the laser beam and the manipulated particle using a
mass-spring-damper model. For this purpose, we calibrated the optical trap and
calculated the coefficients of the mass-spring-damper model using image
processing and curve fitting techniques. The drag force is calculated using the
velocity of the sphere and the viscous damping coefficient of the fluid. We also
use a potential field approach to generate a collision-free path for the
manipulated micro sphere and display the optical trapping and drag forces and
the forces due the artificial potential field to the user through the haptic device
for achieving better results in manipulation and control.
Modeling
• Optical trapping is a non-contact
manipulation method.
• Object sizes ranging from single
atoms to microscopic particles can be
manipulated.
• Forces range from pN to nN.
Set-up
Set-up
XP
XL
Xs
m
b
U
: particle position
: laser position
: stage position
: mass of the particle
: viscous damping of the fluid
: potential field of the laser
Characterization
Haptic Feedback
We have implemented a path planning
algorithm based on a potential field approach
to compute the collision-free path of a trapped
particle. In this approach, obstacles (other
particles) and the goal are represented by
repulsive and attractive potential fields
respectively. The negative gradient of the
potential function gives the attractive force
applied on the particle.
We calculated the coefficients b, m, and spring
constant of laser potential. We applied a
sinusoidal displacement input to the scanner at a
frequency of 1 Hz and captured the motion of a
trapped microsphere using a video imaging
system at 25 Hz. We then calculated its position
from the captured video frames using the Image
Processing Toolbox of Matlab. We fitted
sinusoidal waves to the recorded scanner and
sphere positions to compute the velocity of the
scanner as well as the velocity and acceleration
of the trapped particle. The position, velocity, and
accelaration data were then inserted into the
dynamical equation and the unknown coefficients
were calculated using the least squares curve
fitting technique.
Manipulation Experiments
Modes of manipulation:
• WF : without haptic feedback
• DF : drag force as haptic feedback
• DPF : drag force and force due to potential field as haptic feedback