Project Title:

Ice-Accretion Prediction on Massively Parallel Computers

03.01-9457A
910458
Ice-Accretion Prediction on Massively Parallel Computers
Nielsen Engineering & Research, Inc.
510 Clyde Avenue
Mountain View
CA
94043-2287
Steven C.
Caruso
415-968-9457
LeRC
NAS3-26321
028

03.01-9457A
910458

Abstract:

Ice-Accretion Prediction on Massively Parallel Computers
This project addresses the use of massively parallel computers for the prediction
of time-dependent ice accretion on two- and three-dimensional aerodynamic bodies.
Currently, the LEWICE computer code is being developed at NASA's Lewis Research Center
for the prediction of aircraft aerodynamic performance under icing conditions. There
are several distinct components to the LEWICE program, including water droplet trajectory
calculations and airfoil aerodynamics predictions. These components can require large
amounts of CPU time. The extension of this computer code to three-dimensional geometries
will be severely restricted by the computational power of present-day single- or
serial-processor computers. The goal of this project is to demonstrate the feasibility
of using massively parallel processing techniques to gain significant computational
efficiencies for typical calculations performed in time-dependent icing analyses.
In Phase II, the complete LEWICE code will be ported to a parallel computer.
An accurate and efficient tool that can perform three-dimensional aircraft icing
analyses could be used by both government and industry to further understand aircraft
icing problems, decrease development time and costs of ice protection systems, and
aid in the qualification and certification of aircraft to operate under icing conditions.
icing analysis, parallel processing, computational fluid dynamics