IDEAS (Intelligent Design and Engineering of Aerospace Systems) is an open-source project being directed by NASA's Glenn Research Center, in collaboration with Penn State University, UC Irvine, and Arizona State University. IDEAS will become a central hub for a vast amount of discovery tools that enhance our experimentation selection and allows us to combine new radical ideas from a variety of different of different domains. It will create full prototypes of aircraft building from the bottom up.
To prototype this system, we boil down the application into a single, feasible benchmark problem in thermal management. The IDEAS prototype will feature an AI system that will create new, feasible heat exchanger designs:
Develop a lightweight, multi-functional heat exchanger for waste heat recovery that allows more than 150kW of heat to be extracted from the gas turbine engine core’s exhaust nozzle with the overall objective to use waste heat in a productive manner.
Brooke Weborg
Integration and Data Science Lead
brooke.r.weborg@nasa.gov
Paht Juangphanich
Automation Lead
paht.juangphanich@nasa.gov
Brandon Ruffridge
PeTaL Lead
bruffridge@nasa.gov
Dan Sutliff
Multifunctional Liner Integration Lead
daniel.l.sutliff@nasa.gov
David Ashpis
Integrated Waste Heat Recovery Lead
ashpis@nasa.gov
Ezra McNichols
Thermal Transport and Processes Lead
ezra.o.mcnichols@nasa.gov
Arman Mirhashemi
Heat Exchanger Testing Lead
arman.mirhashemi@nasa.gov
Vikram Shyam
Technical Lead
vikram.shyam-1@nasa.gov
The PeTaL (Periodic Tabel of Life) labeller is a project that seeks to categorize biomimicry papers for engineers. Engineers are interested in lifeforms that have great thermal resistance, structural rigidity, ability to stick to things under water, etc. Life has has millions of years to evolve into the best version of itself. We as humans are in the beginnings of all our designs. If we can based designs off nature's evolution then perhaps we can have highly efficient airplanes with flapping wings or drones that look like bugs. Nature is one of our most valuable resources. There are many researchers dedicated towards understanding nature. Engineers often do not look at biology papers or know their terminology but we would like to search them. The goal of this tool is to help enable that.
DESCRIPTION
PINNs stands for Physics Informed Neural Networks. This exploratory project on using PINNs to solve various partial differential equations (PDEs). Comparisons between this method and classical iterative solvers is also included.
*Available Summer 2022. This Repo is not currently available to the public.
