Graduating from Stanford Phi Beta Kappa in mechanical engineering in 1944, Harvard was soon in the U. S. Navy and assigned to Moffett Field as an engineer. The only course he had found difficult at Stanford was machine shop. Nevertheless, the Navy made him a research scientist in a 16-foot high-speed (for that time, but less than Mach 1) wind tunnel. His theoretical ability soon made it evident that Harvard should join the National Advisory Committee on Aeronautics Ames Theoretical Aerodynamics Branch under Dr. Max Heaslet. In this branch, Harvard soon made many important contributions, including a little-known but extraordinarily informative derivation of the supersonic area rule [NACA RM A55A18]. He understood the hyperbolic equivalent of the Green’s function for elliptic equations, i. e. , Riemann function, which he used to derive the supersonic area rule. This derivation made it clear how to separate wave drag due to lift from that due to volume, and thereby how to minimize the wave drag for a given lift and volume (or maximum cross-sectional area). Subsequently, Harvard’s results were widely used throughout the industry. This culminated almost twenty-five years of theoretical aerodynamics research, which for most would be a career in itself. But Harvard’s career was far from over. The age of the “numerical wind tunnel” was on the horizon, and Harvard was among the first to see its potential.

In 1958 the NACA became NASA with both aeronautics and space in its charter. In an effort to speed up the reduction of data from their wind tunnels, NASA Ames arranged for the purchase of its first computers in the late 1950s. This attracted Harvard’s attention, and he soon learned machine language to be able to evaluate the results from his theories. The Theoretical Aerodynamics Branch was subsequently used to educate others at Ames, by their rotation through this branch, in the use of computers and computational methods.

With the arrival of transistorized computers and FORTRAN, Harvard began a long career devoted to the development of computer methods for the solution of aerodynamic problems. Hans Mark, NASA Ames, and Dean Chapman, division chief for thermo- and gas-dynamics, wisely decided to form a Computational Fluid Dynamics (CFD) Branch with Harvard as its chief,