Wolfram Computation Meets Knowledge

Wolfram Summer School

Alumni

Vallorie Peridier

Summer School

Class of 2005

Bio

Vallorie Peridier is an associate professor of mechanical engineering at Temple University in Philadelphia, Pennsylvania. She is interested in computational methods of engineering mathematics for diverse applications: prior studies include Lagrangian methods for unsteady separation (fluid mechanics) and quantum filtering of electron emission, both at nanoscale point and at contact junctions.

Vallorie’s training includes a B.S. in physics (Bryn Mawr College) and a Ph.D. in engineering mathematics (Lehigh University). She worked in industry before joining academia, first for AT&T (information systems) and then for PP&L (nuclear plant engineering).

Vallorie’s favorite activity is dancing (English country dancing, Scottish country dancing, waltz, hambo, and Argentine tango).

Project: Calculating the 1D Inverse Problem for Surface Heating

This project considered a classic 1D inverse thermal-diffusion problem: can the (unknown) unsteady surface heating q“(t) be inferred from a timeseries of temperature measurements T(t) taken a with a thermocouple buried a distance, say r, from the surface?

A principal objective of the study was to see how successfully one could represent dimensioned, continuum physical systems using the Boolean strategies employed in NKS modeling. Using an NKS modeling strategy, is it then possible to make predictions about physically measurable quantities?

The results of this study are preliminary but very encouraging. Using a basic NKS-style energy propagation model based on rule 226, and a few scaling (Buckingham-Pi) concepts, agreement within 30% of the “exact” (PDE-diffusion equation) solution was achieved for several sample temperature time-series data, each corresponding to a continuum surface heating source. Furthermore, surprisingly good estimates of the surface heating were obtained from noisy data signals as well.

Favorite Four-Color, Nearest-Neighbor, Totalistic Rule

Rule chosen: 570696

I like rules that start off with something scrambled and evolve to pure propagating phenomena.