Codes

Taichi

Taichi is a FMM based collisional dynamics N-body code that is optimized to handle million and post-million body problems. The code is built on top of exaFMM with changes made to improve the force accuracy and optimizations for collisional dynamics. For accurate few-body dynamics, regularization is included. Work is in progress to implement post-Newtonian effects and gravitational wave recoil. Taichi includes time-symmetrized hamiltonian splitting integrators of second and fourth orders. For the latter, gradient extrapolation based forward symplectic methods are used to avoid negative timesteps and improve accuracy. Please contact me at dipto@cmu.edu if you would like to obtain a version of Taichi. Currently, I am working on creating an API that allows a user to use the gravity solver in Taichi easily in an N-body code. The API will be made publicly available.

Falcon

Falcon is a custom N-body code that is tuned to simulate intermediate mass ratio inspirals embedded in dark matter spikes. Since the self-gravity of the spike in minimal, Falcon neglects DM-DM interactions which speeds up the simulation. Post-Newtonian effects upto 2.5 order are included and work is in progress to implement post-Newtonian effects upto 3.5 order. Integration is performed using a second order auxiliary variable based method since the post-Newtonian Hamiltonian is non-separable. It includes adaptive timestepping via the usage of time-symmetrized methods to improve accuracy.

Rapid evolution of interstellar object capture and evolution

I used a combination of a time-symmetrized Hermite P(EC)^3 scheme + universal variables based Kepler solver to rapidly evolve a population of captured interstellar objects. The integrator is written in C and then interfaced using python. A user can run a number of simulations in parallel using multiprocessing. The above code resulted in a publication that investigated the capture of objects in near Earth orbit. Please note that there are some differences in the version that was used for the paper.

Evolution of Primordial Solar System

Using two coupled astrophysical codes, rebound and NBODY6++, my collaborators at Leiden University and I explored the effects of the Sun’s birth cluster on the dynamics of the giant planets and the secular evolution of the debris disk surrounding the planets.

PyStarsmasher

For my bachelor’s thesis, I created a new Python based interface for smoothed particle hydrodynamics (SPH) code Starsmasher which is used for studying stellar mergers. The new interface/code, named PyStarsmasher, can be coupled with any other astrophysical code using the Python interface. One can imagine coupling it to gas dynamics or gravitational dynamics codes using AMUSE. PyStarsmasher is an open-source code and can be downloaded here.

Formation of Runaway Collision Objects

For my bachelor’s thesis, I devised a novel method to study stellar collisions in NBODY6++ using Starsmasher using PyStarsmasher. This method of coupling codes had not been used before and provided better results while treating stellar mergers in star clusters. This coupled code was then used to study the formation of runaway collision objects in star clusters.

NBODY6++

For my bachelor’s thesis, I had to work with NBODY6++. I worked to modify the original code to allow for collisions to take place even when stellar evolution is switched off. In addition to that, the new code outputs details of all stars in the cluster for multiple timesteps before a collision has taken place so that it can be tracked. These files can be read using PyStarsmasher to set up a sequence of collisions. The modified code can be found here.