arXiv subjects

David Hahn

Publications and source records attributed to David Hahn.

6 recordsLinked to original sources

arXiv2024

Inverse designing metamaterials with programmable nonlinear functional responses in graph space

Material responses to static and dynamic stimuli, represented as nonlinear curves, are design targets for engineering functionalities like structural support, impact protection, and acoustic and photonic bandgaps. Three-dimensional metamaterials offer significant tunability due to their internal structure, yet existing methods struggle to capture their complex behavior-to-structure relationships. We present GraphMetaMat, a graph-based framework capable of designing three-dimensional metamaterials with programmable responses and arbitrary manufacturing constraints. Integrating graph networks, physics biases, reinforcement learning, and tree search, GraphMetaMat can target stress-strain curves spanning four orders of magnitude and complex behaviors, as well as viscoelastic transmission responses with varying attenuation gaps. GraphMetaMat can create cushioning materials for protective equipment and vibration-damping panels for electric vehicles, outperforming commercial materials, and enabling the automatic design of materials with on-demand functionalities.

cond-mat.mtrl-sci↗

arXiv2023

View-Independent Adjoint Light Tracing for Lighting Design Optimization

Differentiable rendering methods promise the ability to optimize various parameters of 3d scenes to achieve a desired result. However, lighting design has so far received little attention in this field. In this paper, we introduce a method that enables continuous optimization of the arrangement of luminaires in a 3d scene via differentiable light tracing. Our experiments show two major issues when attempting to apply existing methods from differentiable path tracing to this problem: first, many rendering methods produce images, which restricts the ability of a designer to define lighting objectives to image space. Second, most previous methods are designed for scene geometry or material optimization and have not been extensively tested for the case of optimizing light sources. Currently available differentiable ray-tracing methods do not provide satisfactory performance, even on fairly basic test cases in our experience. In this paper, we propose a novel adjoint light tracing method that overcomes these challenges and enables gradient-based lighting design optimization in a view-independent (camera-free) way. Thus, we allow the user to paint illumination targets directly onto the 3d scene or use existing baked illumination data (e.g., light maps). Using modern ray-tracing hardware, we achieve interactive performance. We find light tracing advantageous over path tracing in this setting, as it naturally handles irregular geometry, resulting in less noise and improved optimization convergence. We compare our adjoint gradients to state-of-the-art image-based differentiable rendering methods. We also demonstrate that our gradient data works with various common optimization algorithms, providing good convergence behaviour. Qualitative comparisons with real-world scenes underline the practical applicability of our method.

cs.GR↗

arXiv2020

ADD: Analytically Differentiable Dynamics for Multi-Body Systems with Frictional Contact

We present a differentiable dynamics solver that is able to handle frictional contact for rigid and deformable objects within a unified framework. Through a principled mollification of normal and tangential contact forces, our method circumvents the main difficulties inherent to the non-smooth nature of frictional contact. We combine this new contact model with fully-implicit time integration to obtain a robust and efficient dynamics solver that is analytically differentiable. In conjunction with adjoint sensitivity analysis, our formulation enables gradient-based optimization with adaptive trade-offs between simulation accuracy and smoothness of objective function landscapes. We thoroughly analyse our approach on a set of simulation examples involving rigid bodies, visco-elastic materials, and coupled multi-body systems. We furthermore showcase applications of our differentiable simulator to parameter estimation for deformable objects, motion planning for robotic manipulation, trajectory optimization for compliant walking robots, as well as efficient self-supervised learning of control policies.

cs.GR↗

arXiv2018

A polynomial automorphism with a wandering Fatou component

We construct polynomial automorphisms with wandering Fatou components. The four-dimensional automorphisms $H$ lie in a one-parameter family, depending on the parameter $\delta \in \mathbb C \setminus \{0\}$, and as $\delta \rightarrow 0$ the automorphisms degenerate to the two-dimensional polynomial map $P$ constructed in Astorg et al (Ann. of Math., 2016). Our main result states that if $P$ has a wandering domain, then $H$ does too for $\delta$ sufficiently small.

math.DS↗

arXiv2017

Luminous and Variable Stars in M31 and M33 V. The Upper HR Diagram

We present HR Diagrams for the massive star populations in M31 and M33 including several different types of emission-line stars: the confirmed Luminous Blue Variables (LBVs), candidate LBVs, B[e] supergiants and the warm hypergiants. We estimate their apparent temperatures and luminosities for comparison with their respective massive star populations and to evaluate the possible relationships of these different classes of evolved, massive stars, and their evolutionary state. Several of the LBV candidates lie near the LBV/S Dor instability strip which supports their classification. Most of the B[e] supergiants, however, are less luminous than the LBVs. Many are very dusty with the infrared flux contributing one-third or more to their total flux. They are also relatively isolated from other luminous OB stars. Overall, their spatial distribution suggests a more evolved state. Some may be post-RSGs like the warm hypergiants, and there may be more than one path to becoming a B[e] star. There are sufficient differences in the spectra, luminosities, spatial distribution, and the presence or lack of dust between the LBVs and B[e] supergiants to conclude that one group does not evolve into the other.

astro-ph.SR↗

arXiv2016

Luminous and Variable Stars in M31 and M33. IV. Luminous Blue Variables, Candidate LBVs, and the B[e] Supergiants; How to Tell Them Apart

In this series of papers we have presented the results of a spectroscopic survey of luminous and variable stars in the nearby spirals M31 and M33. In this paper, we present spectroscopy of 132 additional luminous stars, variables, and emission line objects. Most of the stars have emission line spectra, including LBVs and candidate LBVs, Fe II emission line stars and the B[e] supergiants, and the warm hypergiants. Many of these objects are spectroscopically similar and are often confused with each other. With this large spectroscopic data set including various types of emission line stars, we examine their similarities and differences and propose the following criteria that can be used to help distinguish these stars in future work: 1. The B[e] supergiants have emission lines of [O I] and [Fe II] in their spectra. Most of the spectroscopically confirmed sgB[e] stars also have warm circumstellar dust in their SEDs. 2. Confirmed LBVs do not have the [O I] emission lines in their spectra. Some LBVs have [Fe II] emission lines, but not all. Their SEDS shows free-free emission in the near-infrared but no evidence for warm dust. Their most important and defining characteristic is the S Dor-type variability. 3. The warm hypergiants spectroscopically resemble both the LBVs in their eruption or dense wind state and the B[e] supergiants. However, they are very dusty. Some have [Fe II] and [O I] emission in their spectra like the sgB[e] stars, but can be distinguished by their absorption line spectra characteristic of A and F-type supergiants. In contrast, the B[e] supergiant spectra have strong continua and few if any apparent absorption lines.

astro-ph.SR↗