Previous abstract Next abstract
Session 20 - HAD II - New Telescopes and New Tools.
Division, Oral session, Monday, June 08
Vastly improved understanding of the internal structure and evolution of the stars is one of the extremely successful accomplishments of late 20th century physics and astrophysics. Electronic computers played an essential and pivotal role in the theoretical phases of these developments. Theoreticians had attempted to construct mathematical models of stars since the late nineteenth century, but growth of understanding was slow because: The requisite atomic and nuclear physics was then largely unknown; (2) The four simultaneous non-linear partial differential equations of stellar structure were impossible to solve analytically, thus requiring questionable approximations or extremely tedious and error-prone large-scale numerical integrations with hand-operated mechanical calculators. By 1940 some important basic properties of white dwarfs and main-sequence stars had been deduced but most other stellar types in the Hertzsprung-Russell diagram remained puzzles.
Urgent computational needs of the Manhattan Project and its immediate post-war extensions greatly spurred the invention and rapid development of modern, high-speed, stored-program electronic computers. Simultaneously, new numerical techniques were conceived to take full advantage of the unique capabilities of the new machines. John von Neumann and Los Alamos associates were highly influential in both of these revolutions. Many problems important for the nuclear laboratories required detailed solutions to complex equations similar to the basic equations of stellar structure. Thus, when Martin Schwarzschild, Louis Henyey, Marshal Wrubel and their students and collaborators began applying modern computers to problems in stellar physics, they often had to explore new numerical techniques to fit their own problems. This paper outlines how astrophysicists adapted their insights, thinking and working methods in beginning the transition to computer modeling which came to totally dominate the field by the late-1960s.
Program listing for Monday