Back to 2019 MSM Agenda
Session Description: The design of advanced materials in engineering has always proceeded hand in hand with advances in our ability to understand mechanisms of deformation, resilience, and failure. The tools for modeling tissue-engineered systems have exploded in the past few years, and the integration of these tools with experiment is now maturing to the point that protein-to-tissue level modeling may soon become routine. This session will bring together leaders in the development and application of hierarchical modeling tools to discuss the state of the art and to identify challenges and opportunities. The goal of the session is educate the basic science community on modeling approaches across multiple length scales.
Speaker Bios and Abstracts: Markus Buehler (MIT), David Kaplan (Tufts), and Zhao Qin (MIT)
Format: Brief didactic lectures combined with demosntrations of molecular design
Interactive Discussion (please put you name before your comments):
To what degree are effective
To what degree are effective motifs for toughening, identified at the macroscale, also effective at the nanoscale?
- Pheonemona at the macro and nano scale do not normally agree, and some adjustment will need to be made. Some patterns may translate. This could be a separate talk.
MD-to-3D synthesis requires high fidelity in synthesis. How well can you predict the reliability of materials?
- When we think about reliability, we want to think about the worst case scenario when there is a microcrack. That is why we add defects in the models. It is key to this concept -- how will adding defects perturb the components. We are also looking at this at the level of proteins.
The natural materials that inspire your techniques are all hierarchical. What is the state of the art in modeling and printing these hierarchies?
- A single technique for manufacturing is not enough - a hierachy of manufacturing techniques is also required. You design and manufacture, but a small change in boundary conditions will have a tremendous effect. Ideally, we would like to have a material where the manufacturing is built in -- where we can specify the function, and the material can design itself.
You have shown several
You have shown several techniques for handing very large simulations. How neccessary is this? Are the primary structures that you arise as being effective in your simulations typically simple, repetititve, and generalizeable?
- Many things in nature use universal building blocks (e.g. DNA). It is an advantage for the simulation if there are things that are repetitive -- unit cells can be used, and the predictions can be scaled up from there. There is a question - is this the best way to build things, or is the best that nature can do given what is available?
There are symmetries and regularities that show up that may be represented by a unit cell, but then reandomness shows up that is important for toughness. Why are these simulations so simple and regular?
- The building blocks have to be symmetric in these simulations to make the computations tractable. There may be gradients that will drive the system away from symmetrical responses.
If you break symmetry ahead of the crack, it will dramatically increase toughness. So, if there are symmetries around the crack plane, is this the result of the input parameters or something else?
- We are currently looking at crack initiation. The issue you raise is related to crack propogation. It is also an important question to look at. But it is not tractable to do that solution at this time due to the computational power needed.
Can you talk a little more about correlating the effects of mutation and the changes in protein structure to the mechanics of the material?
- Trying to do this with some simple materials like biofilms.
Markus Buehler (MIT)
David Kaplan (Tufts)
Zhao Qin (MIT)
Presentations to follow.