KAYE D. LATHROP (1932-2021)

BY WARREN F. MILLER

KAYE DON LATHROP, a pioneer in developing computational methods for solving the linear Boltzmann equation – used to model neutron behavior in complex media – died on Oct. 23, 2021, at the age of 89.

Kaye was born Oct 8, 1932, in Bryan, Ohio, and grew up in Macomb, Illinois, a farming community in the western part of the state. After graduating from high school, he entered the United States Military Academy at West Point in 1951 and graduated in 1955. While a cadet, he met Judith (Judy) Green, a student at nearby Vassar College. The two married in June 1955. After serving three years in the U.S. Army, Kaye pursued graduate studies at the California Institute of Technology, earning a master's and a doctorate degree. His dissertation, “Neutron Thermalization in Solids,” completed in 1962, signaled a lifelong focus on neutron transport theory.

That same year he joined Los Alamos Scientific Laboratory (renamed Los Alamos National Laboratory in 1981), where he would remain until his retirement in 1984. Over his 22-year career, Kaye emerged as a scientific leader and innovator in the modeling of neutron behavior. He played a formative role in developing computational approaches to solving the neutron transport, or linear Boltzmann, equation. This equation is used to model how neutrons move and interact with matter, providing critical information for understanding the behavior of both nuclear weapons and nuclear reactors. After World War II, the rapid advancement of scientific computing made it plausible to more accurately simulate the dynamic behavior of nuclear weapons. Among many challenges, this required significant progress in methods for discretizing the Boltzmann equation to model neutron transport.

Kaye provided leadership and creativity in advancing early neutron transport discretization methods, particularly by building on Bengt Carlson’s SN (discrete ordinates) method. This method addressed the angular variable by dividing it into a set number, N, of discrete segments. Kaye expanded this concept into a more general discrete ordinates approach to angular discretization. Working often with collaborators, he continued to refine the methodology – improving its ability to handle complex geometries, improving computational efficiency, increasing flexibility for modeling physical systems, and boosting accuracy by employing conservation rules. For instance, a straightforward approach to solving the complex, five-dimensional Boltzmann equation is to discretize the variables, form a matrix equation, and then solve it using matrix inversion. However, this is computationally intensive. Kaye was the first to formulate an accelerated source-iteration approach, which significantly reduced both computer memory requirements and runtime. His systematic development and adaptation of computational techniques – particularly as more advanced mainframe computers became available – helped establish a new research area that continues today. More efficient and sophisticated neutron transport codes are still being developed to operate on increasingly powerful hardware and system software.

An alternative to the discrete ordinates approach is the Monte Carlo method, a stochastic technique that solves the Boltzmann equation by sampling neutron transport events. This approach was also pursued vigorously at Los Alamos. Although some viewed Monte Carlo as a competing methodology, Kaye saw it as complementary. He welcomed collaborations with Monte Carlo developers, shared ideas to enhance Monte Carlo performance, and encouraged his colleagues in their work. In return, he sought insights from the Monte Carlo community to further improve the performance of discrete ordinates codes.

While serving as the T-1 Group Leader in the late 1970s, Kaye initiated a new computational program. Alongside W.H. Reed and others, he developed expertise in modeling the complex solid and fluid dynamics that occur under off-normal conditions in liquid-sodium cooled reactors. Coupling this phenomenon with neutron transport modeling led to the creation of computational tools critical for understanding reactor safety. As a result, Kaye is regarded as a pioneer in the development of computational tools for analyzing hypothetical accidents in fast-neutron reactors.

In recognition of his seminal work in computational methods, he received the Los Alamos National Laboratory Distinguished Service Award and, in 1976, the prestigious E.O. Lawrence Memorial Award from the U.S. Energy Research and Development Administration. He was elected to the National Academy of Engineering in 1986 – engineering’s highest professional honor.

Kaye was also highly active in the American Nuclear Society (ANS). In addition to being named an ANS Fellow, he held multiple leadership positions within the Society, including chairman of the Mathematics and Computation Division. He was recognized with the ANS Outstanding Performance Award in 1980. His leadership within ANS extended to serving as national treasurer and as a member of the board of directors – roles central to shaping the Society’s strategic direction and governance, reflecting his deep commitment to the nuclear science community.

At Los Alamos, Kaye was known not only as an outstanding research scientist, who made solutions to the Boltzmann equation feasible on advancing computer platforms, but also as a great leader. Colleagues and subordinates recall him as a disciplined professional who demanded excellence from himself and those around him. Those working under him thrived in the organized, focused, and friendly environment he cultivated.

I joined Kaye’s T-1 Group in 1974, working alongside B.A. Carlson, W.H. Reed, R.E. Alcouffe, and T.R. Hill, each widely recognized for advancing computation methods for solving the Boltzmann equation. In the years that followed, the group expanded to include other gifted contributors, notably E.W. Larsen and J. Morel. As the group's scope broadened (before and after 1974) to include more nuclear reactor core physics analysis, it welcomed reactor physics experts such as D.J. Dudziak and D. Odell to address new challenges. I witnessed firsthand Kaye’s remarkable talent for fostering teamwork, confidence, comradery, and a strong work ethic. It was an honor to be part of such a dynamic and powerful team. Recognizing Kaye’s leadership abilities early on, P.D. Carruthers, leader of the Theoretical (T) Division, appointed him Assistant T Division Leader in 1973, a role he held concurrent with his T-1 Group leadership.

Beyond his accomplishments in the T Division, Kaye’s exceptional leadership abilities were quickly recognized across the laboratory. Laboratory Director, H.M. Agnew (NAE 1976) reassigned him from T Division, and both Agnew and his successor, D. Kerr, promoted him to roles of increasing responsibility. These included associate division leader of the Reactor Division (1975-77); alternate division leader of the Energy Division (1977-78); division leader of the Computer Science and Services (C) Division (1978-79); and associate laboratory director for engineering sciences (1979-84). His appointment as C Division leader was particularly notable. Kaye was not only a master of transforming complex equations into forms solvable by computers but also became an expert in the most advanced high-performance computer platforms of the time.

After retiring from Los Alamos, Kaye joined the Stanford Linear Accelerator Laboratory (SLAC) in 1984. He served as associate director of the Technical Division and concurrently as a professor of applied physics from 1981 to 1994. He retired in 1994 with the title emeritus professor of applied research. At SLAC, he made noteworthy contributions to accelerator physics and was once again recognized as an outstanding technical leader. He was admired for his dedication to advancing scientific knowledge and for his collaborative spirit, which left a lasting impact on colleagues and the broader accelerator physics community.

In retirement, Kaye and his wife, Judy, moved to Ridgway, Colorado, where they spent time hiking, skiing, and biking in the San Juan Mountains. Their Greater Swiss Mountain dogs – Max, Mello, and Maisie – were their constant companions on these outdoor adventures. Even in retirement, he remained deeply engaged in science and public services. From 2006 to 2012, he served as an administrative law judge for the Atomic Safety and Licensing Board Panel of the U.S. Nuclear Regulatory Commission. He also regularly reached out to active computational scientists at Los Alamos and universities to discuss his latest ideas in computational neutron transport, nuclear reactor physics, and reactor safety.

Kaye was a pioneering giant in his field and a friend to many. He is deeply missed!