Coursera Flash Sale
40% Off Coursera Plus for 3 Months!
Grab it
Watch a technical presentation from the 2024 MFEM Workshop where Dohyun Kim from Brown University introduces SiMPL (Sigmoidal Mirror descent with Projected Lagrangian), a novel first-order method for density-based topology optimization. Explore how this innovative approach maintains point-wise bound preserving density fields throughout iterations while relying solely on first-order derivative information for design updates. Learn about the method's acceleration through adaptive step size and back-tracking line search, its mesh-independent behavior, and superior convergence rates compared to existing first-order optimization algorithms. Examine practical applications including compliance minimization and compliant mechanism problems that demonstrate the technique's versatility. The presentation is part of the fourth annual MFEM community workshop, which brings together users and developers of the Modular Finite Element Methods project to discuss software features, development roadmap, and showcase technical applications.
