BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:Asia/Tokyo
X-LIC-LOCATION:Asia/Tokyo
BEGIN:STANDARD
TZOFFSETFROM:+0900
TZOFFSETTO:+0900
TZNAME:JST
DTSTART:18871231T000000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20250110T023312Z
LOCATION:Hall B7 (1)\, B Block\, Level 7
DTSTART;TZID=Asia/Tokyo:20241205T092300
DTEND;TZID=Asia/Tokyo:20241205T093400
UID:siggraphasia_SIGGRAPH Asia 2024_sess126_papers_1048@linklings.com
SUMMARY:Barrier-Augmented Lagrangian for GPU-based Elastodynamic Contact
DESCRIPTION:Technical Papers\n\nDewen Guo (Peking University), Minchen Li 
 (Carnegie Mellon University), Yin Yang (University of Utah), and Sheng Li 
 and Guoping Wang (Peking University)\n\nWe propose a GPU-based iterative m
 ethod for accelerated elastodynamic simulation with the log-barrier-based 
 contact model. While Newton's method is a conventional choice for solving 
 the interior-point system, the presence of ill-conditioned log barriers of
 ten necessitates a direct solution at each linearized substep and costs su
 bstantial storage and computational overhead.\nMoreover, constraint sets t
 hat vary in each iteration present additional challenges in algorithm conv
 ergence. Our method employs a novel barrier-augmented Lagrangian method to
  improve system conditioning and solver efficiency by adaptively updating 
 an augmentation constraint sets. This enables the utilization of a scalabl
 e, inexact Newton-PCG solver with sparse GPU storage, eliminating the need
  for direct factorization. We further enhance PCG convergence speed with a
  domain-decomposed warm start strategy based on an eigenvalue spectrum app
 roximated through our in-time assembly. Demonstrating significant scalabil
 ity improvements, our method makes simulations previously impractical on 1
 28 GB of CPU memory feasible with only 8 GB of GPU memory and orders-of-ma
 gnitude faster. Additionally, our method adeptly handles stiff problems, s
 urpassing the capabilities of existing GPU-based interior-point methods. O
 ur results, validated across various complex collision scenarios involving
  intricate geometries and large deformations, highlight the exceptional pe
 rformance of our approach.\n\nRegistration Category: Full Access, Full Acc
 ess Supporter\n\nLanguage Format: English Language\n\nSession Chair: Paul 
 Kry (McGill University)
URL:https://asia.siggraph.org/2024/program/?id=papers_1048&sess=sess126
END:VEVENT
END:VCALENDAR