Key Takeaways

Learn about the benefits of the finitedifference timedomain method.

Gain a greater understanding of the application of the finitedifference timedomain.

Learn more about the formulations associated with the finitedifference timedomain.
Electrodynamics, quantum physics, electrostatics, thermodynamics, and Maxwell's equations.
In the fields of science and electronics alike, there are always problems or equations that require solving. Throughout the advancement of these and similar areas of study or focus, there are methods in use for this precise purpose. In smartphone marketing, the saying is, "There's an App for that."
However, in the areas of electronics and science, there is always a method for that. In this case, the technique or method in question is the finitedifference timedomain (FDTD). We utilize this method to model computational electrodynamics or find approximate solutions to the associated system of differential equations.
FiniteDifference TimeDomain (FDTD)
Kane S. Yee first introduced the numerical analysis technique we call the finitedifference timedomain method in 1966. To this day, many still refer to this method as Yee's method. It did not officially receive the FDTD method designation until 1980.
The FDTD method is a discrete approximation of James Clerk Maxwell's equations that numerically and simultaneously solve in both time and 3dimensional space. Throughout this process, the magnetic and electric fields are calculated everywhere within the computational domain and as a function of time beginning at t = 0.
Note: James Clerk Maxwell's equations symbolize one of the most refined and aphoristic ways to state the fundamentals of magnetism and electricity. These equations include:
 Gauss' law for electricity
 Gauss' law for magnetism
 Faraday's law of induction
 Ampere's law
Maxwell's equations are the basis for FDTD and describe the effect and behavior of electromagnetism. In summary, it is the technique for the simulation of computational electromagnetism. Those in the field consider it the easiest and most effective way to model the effects of electromagnetism on a specific object or material.
The FDTD Approach
Utilizing the FDTD method will divide both time and space into distinct segments. It provides the segmentation of space into boxshaped cells that are small in comparison to the overall wavelength. The precise location of the electric fields is on the edges of the cube. However, the position of the magnetic fields is on the faces of the boxshaped cells.
In summary, there are three steps in the FDTD computation:

The discretization of the whole area into small cells (defined as Δx, Δy, and Δz) in the Yee cell.

Defining the electrical properties (conductivity and permittivity) of the model in a mesh grid.

Substitution of partial derivatives by differential quotient in James Clerk Maxwell's equations.
Note: In the above summary, the uppercase Delta (Δ) represents a change.
If you obtain the parameters of the primary fields in time and space, you can also calculate other secondary measures. These quantities include frequencydomain characteristics such as impedance (input), radar crosssection, scattering parameters, and farfield radiation patterns, to name a few.
FiniteDifference TimeDomain Applications and Formulations
The FDTD method is a finite domain numerical method; therefore, we must truncate the computational domain of the problem. We must also enforce the proper boundary conditions at the boundaries of the computational domain. For example, in a structure (shielded), we enclose all objects within a perfect magnetic or electric conductor box.
With open boundary problems, such as an antenna, we utilize an absorbing boundary condition like a PML (perfectly matched layer) to mimic the free space. In this case, the absorbing boundaries allow the incident waves and fields to flow through them without back reflection. In terms of FDTD simulation time, it directly correlates to the following parameters:

The size of the computational domain.

The proximity of the PML walls to the enclosed objects.
We must approximate the computational domain of the FDTD by utilizing an appropriate meshing scheme, but the method itself provides a wideband simulation of a physical structure. To produce the necessary spectral information, we require a suitable wideband temporal waveform to excite the physical structure. Your choice of a waveform, time delay, and bandwidth will affect the merging behavior of the FDTD loop (time marching).
Using the FiniteDifference TimeDomain Method
One particular issue of concern is the numerical stability of the FDTD method’s time marching scheme. Therefore, to satisfy the CourantFriedrichsLewy (CFL) stability condition, your time step needs to be inversely proportional to the maximum grid cell size.
Note: The CourantFriedrichsLewy condition is a necessary condition for convergence while numerically solving certain PDEs. This emerges in the numerical analysis of distinct time integration schemes when we utilize these for a numerical solution.
Functionally, a highresolution mesh requires a smaller time step. Operationally, a smaller time step requires a more significant number of time steps to converge if letting the fields in the computational domain fully evolve.
Using the FiniteDifference TimeDomain Method, Continued
We must establish a computational domain when implementing an FDTD solution of James Clerk Maxwell's equations. We define a computational domain as the physical region in which we perform the simulation. In this case, we determine the H and E fields at each point in space within the computational domain. We specify the material of each cell within this computational domain as well. The material in question is generally either metal, dielectric, or freespace (air).
You can utilize any material, provided you specify its conductivity, permittivity, and permeability. Keep in mind that you cannot directly substitute the dielectric constant of dispersive materials in tabular form into the FDTD scheme. Therefore, you approximate by utilizing multiple Debye, Drude, Lorentz, modified Lorentz, QCRF (quadratic complex rational function), or CCPR (complexconjugate poleresidue) models.
Upon establishing both the grid materials and the computational domain, we then specify a source. The source can be either an applied electric field, impinging plane wave, or current on a wire. With regard to an impinging plane wave, we can use the FDTD method to simulate light scattering from planar periodic structures, the photonic band structure of infinite periodic structures, and arbitrarily shaped objects.
Since we directly determine the E and H fields, the simulation's output is usually the H or E field at a series of points or a point within the computational domain. During this process, the simulation will evolve the H and E fields forward in time. Generally, it does the processing on the H and E fields returned by the simulation. Data processing can occur during the ongoing simulation. Lastly, during the time the FDTD method calculates the electromagnetic fields inside a compact spatial region, you can obtain radiated and/or scattered farfields via neartofarfield transformations.
The finitedifference timedomain is a numerical analysis method utilized for modeling computational electrodynamics. Though it is still a timedomain technique, FDTD solutions are capable of covering a wide range of frequencies within a single simulation run.
Magnetic Flux and Faraday's Law of Induction.
Whether you are utilizing Gauss' law for electricity and magnetism, or Faraday's law of induction, you will need the right set of PCB layout and design software. Allegro PCB Designer and Cadence's full suite of design tools can help you create designs from verified component models and analyze all aspects of functionality. You'll also have access to a set of tools for MCAD design and preparing for manufacturing.
If you're looking to learn more about how Cadence has the solution for you, talk to us and our team of experts.
About the Author
Follow on Linkedin Visit Website More Content by Cadence PCB Solutions