Niels Vonk

The WaterLily.jl article is now available on Computer Physics Communications! Together with Gabriel Weymouth, we review our CFD solver's design (runs on any CPU or GPU backend!), performance, accuracy, and also show some very cool examples (like the one below). Check it out! 🔗 https://lnkd.in/drfwRFmf #CFD #GPU #HPC #JuliaLang #Julia

208

8 commentaren

As silicon photonics continues to push the limits of high-throughput optical communication for space and near-Earth missions, understanding how radiation impacts device performance and reliability has become crucial. In our recent work called "RADPIC: A Radiation-Aware Design-Space Exploration and Simulation Framework for SOI-Based Photonic Integrated Circuits," to be presented at #IEEE #PAINE conference in CO next week, we bridge this gap by integrating optical and electrical simulation analyses that account for total ionizing dose (TID) effects in silicon photonics. We demonstrate RADPIC’s capabilities through a microring resonator modulator (MRM) case study, analyzing changes in frequency response, extinction ratio, modulation efficiency, and quality factor under radiation exposure. The results show strong agreement with experimental data, highlighting RADPIC’s potential as a foundation for radiation-aware digital twins in photonic circuits, and developing more robust and radiation-hardened PICs for future deep-space communication and sensing missions. This work was done by my MS student, Kevin M., and in collaboration with my colleague, Biswajit Ray, at CSU! #SiliconPhotonics #SpaceCommunications #RadiationHardening #PhotonicIntegratedCircuits #SOI #Research #OpticalEngineering #DeepSpace

178

Density Functional Tight Binding (DFTB) is a quantum chemistry technique used in the Amsterdam Modeling Suite (AMS) to conduct fast and precise simulations. It's a simplified version of density functional theory (DFT) that speeds up calculations by using a minimal basis set and pre-calculated parameters. This significant speed boost over standard DFT allows for the simulation of larger systems and longer time frames. DFTB is a popular choice for a variety of applications in fields like materials science, chemistry, and biology because it effectively balances computational speed with a good level of accuracy. The DFTB module within AMS is a powerful and flexible tool for a wide array of research challenges. It helps scientists study the molecular structures, reaction paths, and electronic and vibrational properties of systems with thousands of atoms. It's commonly used for initial material screening, mapping out the conformational changes of large biological molecules, or simulating dynamic processes where standard DFT would be too computationally intensive due to the system's size. AMS's integration of DFTB offers a user-friendly platform with robust analysis tools, creating a smooth workflow from setting up the system to visualizing the data. The DFTB method in AMS is constantly being enhanced and refined. It features various versions, including the common DFTB2 and the newer DFTB3, along with extensions for non-covalent interactions (DFTB-D3) and excited states (TD-DFTB). These improvements broaden its use for more intricate systems and phenomena. The method's accuracy depends heavily on its parameters, and AMS provides a variety of pre-set parameters for different elements, ensuring a dependable starting point for many applications. This blend of speed, accuracy, and adaptability makes the AMS DFTB module an invaluable resource for computational researchers. For videos, please check the links in comments Nyro Research India Sreejit Soman #QM #MD #compchem #DFT #Materials

67

3 commentaren

Are you working with hybrid systems models and curious about the saltation matrix and various tools we have built in recent years for it? If your system has changing contact conditions (or other hybrid events), you need to understand the saltation matrix and determine if should be accounted for in your work. We have put together a collection of example code & tutorials for calculating the saltation matrix and using it for state estimation and control. Specifically, this has code for our salted Kalman filter (SKF), hybrid iterative linear quadratic regulator (h-iLQR), and analysis of simple systems from our tutorial paper. The code is in Matlab and Python (with more python code coming soon!): https://lnkd.in/eZX5VgjW Our tutorial paper on saltation matrices is here (and you can now run all of the examples from this paper in the tutorial code): https://lnkd.in/e9xbKwA7 We had bits and pieces of code out before, but we've cleaned it up and brought it all together in this repository. Sorry we didn't post this earlier! Hope this helps, and let me know if you have any issues. Thanks to Diana Frias Franco, Joshua Ramos, Derek Fan, and Karla Soto Cuevas for getting this out, and to Nathan Kong, Joe Payne, and James Zhu for their earlier code. (My students actually finished this months ago and I forgot to post about it...)

89

1 commentaar

Are you a researcher in aerial robotics and also think it is time for a fresh way to perform system- and autonomy design? Have you also considered that perhaps using the same few design concepts (e.g., planar quadrotors, helicopters, or fixed-wings) for all diverse applications and environments in which robots are deployed may not be the best path forward? Have you too thought that we could develop algorithms that computationally design a robot’s airframe and its control such that they are *jointly* optimized for a certain task? If yes, then have a look at our latest work! In this Proof-of-Concept study, we show that we can create an end-to-end algorithm that designs a morphologically novel multirotor *and* its control policy such that it outperforms state-of-the-art designs in a given task both in simulation and the real world. Critically, computational design is driven by navigation performance on the task considered as opposed to simplified and partially task-agnostic airframe metrics like thrust-to-weight ratio. This is a first step towards morphological diversification in aerial robotics and a wave of new systems that have their airframes and navigation policies co-optimized to best match the needs of a mission and the pressures of an environment. Naturally, a lot of work lies ahead to make this deliver truly “unimaginable” results. We are committed to making it happen! This work was led by Etor A. as part of the #SPEAR Horizon Europe project funded by European Health and Digital Executive Agency (HaDEA), and also paves the way for a set of investigations to take place in the newly funded Norwegian Centre for Embodied AI. Further contributors: Welf Rehberg, @Phillip Weiss, Mihir Kulkarni * Video link: https://lnkd.in/d6gv8_d7 #ntnu #robotics #autonomy

185

New publication: Physics-Informed Neural Networks applied to adsorption processes. Happy to be part of this collaboration between Eindhoven University of Technology, University of Lille 1 Sciences and Technology and TNO, aiming at advancing physics-informed AI for climate solutions! Main highlights: - With abundant data, both traditional NNs and hybrid-PINNs perform well. - In data-scarce conditions, PINNs significantly improve generalization and accuracy. - A curriculum learning strategy keeps PINNs stable even under extreme low-data scenarios. - The trade-off: higher computational cost due to enforcing physics constraints. The paper can be accessed here: https://lnkd.in/emKM5fCu Mattia Galanti Jean-Yves Dieulot Jurriaan Boon Martin van Sint Annaland

59

1 commentaar