A general thermal modeling and control methodology for thermal processing o
f layered materials for rapid prototyping technologies is established in th
is article. An analytical multivariable model of lumped temperature outputs
generated by heat inputs on a surface grid is developed, based on Green's
function and state-space descriptions. The few independent parameters neede
d in such a linearized formulation are experimentally identified and their
time-variability reflects the heat transfer nonlinearities and process dist
urbances. A robust controller with thermal feedback is designed by pole pla
cement methods, to obtain a specified dynamic temperature field yielding th
e desired material structure and properties. The regulated thermal processi
ng is optimized in real time by proper heat source power modulation and tor
ch guidance through a simulated annealing strategy. Its performance is rest
ed on both the computer model and a laboratory station, using robotically g
uided plasma-arc cutting and infrared thermal sensing, in regulating the se
nsitized zone during blanking of an elementary contour pattern on stainless
steel.