Blaise Bourdin

The journal Club theme of Aug. 15th on iMechanica is "Variational formulations in fracture mechanics", and features a discussion of our book "The variational Approach to Fracture".

Follow this link for the thread on iMechanica.

The following computations were performed a model for dynamic fracture mechanics derived by the variational approach for quasi-static fracture. It is an ongoing work in collaboration with Christopher J. Larsen (WPI) and Casey Richardson (UCLA). Click on the pictures for animations.

A low speed tearing experiment:


A faster one


Two cracks trying to avoid each other at high speed

About this book:
Presenting original results from both theoretical and numerical viewpoints, this text offers a detailed discussion of the variational approach to brittle fracture. The key to this approach is viewing crack growth as the result of a competition between bulk and surface energy. The variational approach that the authors champion provides an incisive picture of initiation and propagation whose features are detailed. It treats crack evolution from its initiation all the way to the failure of a sample. The authors set forth tested-and-proven models that you can use to gauge crack initiation, crack path, and crack extension for arbitrary geometries and loads. Although the material is mathematical in nature, the authors avoid unnecessary technicalities. They also connect the variational approach with more classical treatments of fracture, demonstrating the distinct results of each approach in simple test settings and via relevant numerical simulations. In short, this text offers a new and unified view of fracture evolution in a brittle solid.

Written for:
Mechanicians, especially those involved in materials science and/or computational mechanics; applied mathematicians involved in the mathematics of materials and/or the calculus of variations; structural, civil and mechanical engineers interested in fracture and its numerical simulation at a structural level.

[ISBN-13: 978-1402063947]

I am currently looking for one or several undergraduate assistants (math, science or engineering major) for the summer, and possibly continuing throughout the academic year.

Duties include the pre-processing, running and post-processing large scale fracture mechanics experiments on supercomputers.
The position requires some UNIX knowledge, some scripting or web programming. Experience with supercomputer is not necessary but welcome. Some knowledge of numerical analysis, optimization or pde is a plus. Additionally, I have several coding projects also related to fracture mechanics for a macOS developer.

Hours and salary will be based upon availability and experience.
The positions are funded through NSF REU funds, and US citizenship or permanent residency is required.

Contact me for more information.

3dconnexion just released a beta version of a mac SDK for some of their 6 degree of freedom input devices, including the really neat (and affordable) SpaceNavigator. unfortunately, the interface is incompatible with the old-style UNIX/X11 SDK. MagellanProxy (get it from the Download page) is a quick and dirty hack acting as a proxy between both interface. Not very sophisticated, not very well written, but challenging enough to write, considering that I know nothing about cocoa programming, X11 programming and can barely write C...

LSU research has a nice piece on my work on fracture mechanics here.

New 3D simulations similar to that of the Fracture page, but with fixed displacements boundary conditions...
Click on the pictures to see an animation

My proposal "A Free Discontinuity Approach to Brittle Fracture Mechanics: Analysis and Numerical Implementation" in the Division of Mathematics Sciences of the NSF has been funded!
More here!

My proposal to the Medium Allocation Review Committee of the Teragrid was just funded. I now have 175,000 System Units to use for Fracture computations.

References:
Allocations on the TeraGrid by the Cyberinitiative Partnership, a project of the Office of Cyberinitiative at the National Science Foundation

My fracture mechanics code scales linearly on up to 800 processors... (or perhaps more fairly, petsc's linear solvers scale on up to 800 processors.)

The First Louisiana Joint Workshop for Academia and Industry will take place at LSU on March 30 and April 1st. Check the web pages.

The Backtracking Algorithm, aimed at avoiding some local minimizers for the fracture problem.
More 3D simulations in the Fracture page.
More crack propagation under thermal loads (see also my page on the cluster schwarz)

Applied Analysis Seminar @ LSU
IGERT on Computational Fluid Dynamics
Schwarz cluster