Advanced examples¶
This section provides a very brief overview of the more advanced examples
currently included in the toolbox. These examples are all contained
in the examples/ directory.
Previous versions of the toolbox included other examples, most functionality is still present in this version of the toolbox but examples have not all been included.
Contents
Landscape¶
Calculates trapping landscapes. These show how optical trap properties,
such as trap stiffness or maximum optical force, vary as a function of
particle properties (such as radius and refractive index).
The ottLandscape.m example calculates how maximum trap depth varies
with different size and refractive index spherical particles, example
output is shown in [TODO].
These types of calculations often involve looping over parameters.
In some cases it is more optimal to pre-compute translations or rotations,
such as (in the example) translations along the beam axis for different
sized particles.
This version of the toolbox also adds support for re-using VSWF data
(for faster T-matrix calculation and field calculation), this is a feature
which may be useful for certain types of trapping landscapes.
See ott.utils.VswfData and its usage in the various beam and
T-matrix classes.
Dynamics¶
The ottDynamics.m example shows how the ott.tools.Dynamics
class can be used to simulate particle dynamics.
This requires a beam, T-matrix and drag tensor (which can be provided
as a ott.particle.Particle instance); and solves the Langevin
equation using a fixed time step method.
The example creates a spherical particle and generates a trace of the position and a 2-D histogram of the XY position, example output is shown in [TODO].
Wall Effects¶
This ottWallEffects.m example builds on the ottDynamics.m example
by using a Sphere-Wall drag tensor.
Currently there is not optical interaction between the particle/beam/wall,
this may be added in a future release.
The example generates visualisations of [TODO], example output is shown
in [TODO].
Non-spherical particles¶
This version of the toolbox is focussed on T-matrix methods for scattering calculations. Future versions of the toolbox aim to include DDA, shape surface approximation, and geometric optics. For now, the main methods for modelling almost-arbitrary shaped particles are point-matching, extended boundary conditions method (EBCM), and discrete dipole approximation.
The ottNonSpherical.m example shows how EBCM, DDA and PM can be used
to model a cylindrical particle. For certain sizes, all three methods
agree well, other sizes/aspect ratios, the methods start to disagree,
as shown in [TODO].
Calculating T-matrices using different methods is a good method for
validating the generated T-matrix/predicted force.
DDA Vaterite¶
The ottDdaVaterite.m example shows some of the more advanced features
of the DDA implementation: the implementation can be used to model a
inhomogeneous birefringent particle.
The example calculates T-matrices for vaterites with a sheaf-of-wheat
structure and calculates the torque about the z axis. Example output
is shown in [TODO].
The DDA implementation also supports calculating only specific columns of the T-matrix. This can be useful when the T-matrix is illuminated by only a beam with particular orders; or when calculating T-matrices in parallel.
Neural Networks for fast simulation¶
The ottNeuralNetwork.m example shows how a neural network can be
trained to rapidly predict optical force data. This requires first
generating a large training data set which is used to train the network.
Once trained, the network can be used for rapid force calculations
(for inputs within the bounds of the training data) or easily shared
with other researchers.
For this particular problem, interpolation could also be used, however the stored data sets are often much larger than the resulting neural network. Additionally, for problems with additional inputs/outputs (such as predicting force and torque from orientation and position), direct interpolation becomes a much more computationally expensive task.
Further reading¶
These examples cover most of the advanced functionality in the toolbox. For information about how different methods work, the reference section provides some insight. Otherwise, take a look at the Further Reading section or the previous version of the toolbox for additional examples. Experimental (viz., incomplete) new features can also be found in version 2 of the toolbox (available on GitHub).