Meteorologist, oceanographer, Python solvem probler, experimenter, Raspberry Pi tinkerer, geek dad, manager, space enthusiast, ham radio lover, science columnist, ADHDer, synthetic intellect curious, linux admin cosplayer.

Stanley Phillips Frankel

Born: June 6, 1919, Los Angeles, California Died: May 1978 (aged 58)

Education

  • Graduate study at the University of Rochester
  • Ph.D. in Physics, University of California, Berkeley
  • Post-doctoral work under J. Robert Oppenheimer at UC Berkeley (1942)

Pre-Manhattan Project: Berkeley, 1942

In spring 1942, Oppenheimer recruited his post-doctoral fellow Robert Serber to work with two graduate students – Eldred C. Nelson and Stanley P. Frankel – on problems of neutron diffusion and bomb hydrodynamics. Frankel and Nelson derived a formula for estimating the critical mass of a uranium-235 assembly based on diffusion theory. Oppenheimer used this formula in his estimates presented at the crucial Berkeley Summer Study of June–July 1942 – the conference where Oppenheimer, Fermi, Bethe, Teller, Konopinski, Serber, Van Vleck, Frankel, and Nelson laid the theoretical foundations for both the fission and fusion bomb programs. Frankel was thus present at the very inception of both weapons programs – a 23-year-old graduate student sitting in the room where the atomic age was planned.

Manhattan Project (1943–1945)

Los Alamos T-Division

Frankel joined the Theoretical (T) Division of the Manhattan Project at Los Alamos in 1943, recruited through his connection to Oppenheimer. He was central to the computational infrastructure that made the bomb possible.

Organizing the Human Computers

Frankel, together with Eldred Nelson, made early calculations on neutron diffusion – the key physics problem for determining whether a chain reaction would work. But his most distinctive organizational contribution was creating Group T-5: he organized teams of scientists’ wives (and other women with mathematical training) into assembly-line workflows, using Marchant and Friden desk calculators to perform the complex, repetitive implosion hydrodynamics calculations essential for the plutonium bomb design. His own wife, Mary Frankel, worked as one of these human computers in the same division.

The calculations were organized so that each person performed one step, passed the result to the next person, and so on – a human pipeline that prefigured the architecture of electronic computers.

The Feynman Connection

Richard Feynman told the famous story in “Surely You’re Joking, Mr. Feynman!” and in his “Los Alamos From Below” talk. Frankel realized that the computation problems at Los Alamos could be solved more efficiently using IBM tabulating machines. He designed a system and ordered IBM machines from the company.

But then Frankel caught what Feynman called “the computer disease” – the irresistible delight in seeing how much you can make a machine do. While the IBM group was running far behind schedule on critical calculations, Frankel was sitting in a room programming one of the tabulators to automatically compute a table of arctangent values – “absolutely useless because they had tables of arc-tangents.” Feynman was pulled from his own work and asked to take over the IBM group, which he reorganized into a parallel processing pipeline that dramatically accelerated output.

The anecdote is affectionate rather than damning: Frankel was the first at Los Alamos to grasp the potential of programmed machines, even if his enthusiasm temporarily overwhelmed his priorities.

ENIAC and the H-Bomb Calculations (1945–1946)

Timeline

In March 1945, before the Trinity test, von Neumann brought Frankel and Metropolis to the Moore School of Electrical Engineering at the University of Pennsylvania for an advance look at ENIAC. Von Neumann had recommended to Teller that a thermonuclear problem be used to test ENIAC, because it would be far more demanding than the ballistic trajectories the Army had designed it to calculate.

In August 1945, immediately after the war ended, Frankel and Metropolis returned to the Moore School to set up and program the calculation. They were the first Los Alamos scientists to program ENIAC.

What the Calculation Computed

The problem was a one-dimensional simulation of thermonuclear burning: specifically, whether a cylinder of deuterium-tritium fuel, heated by a fission primary, would sustain a self-propagating fusion reaction. The calculation modeled the behavior of deuterium-tritium systems at various initial temperature distributions and tritium concentrations. It attempted to answer two questions: (1) Could the thermonuclear fuel be ignited? (2) Would the burning propagate through the fuel?

Scale and Duration

The calculation consumed half a million punch cards of data (some sources say “a million” – the discrepancy may reflect different stages of the multi-run computation). It used 95 percent of ENIAC’s control capacity – nearly every one of the machine’s 18000 vacuum tubes and 1000 bits of internal memory. The program ran for approximately six weeks beginning in late November or early December 1945. Results were available by early 1946, and subsequent calculations continued through February 1946 (the month of ENIAC’s formal public dedication).

The Results and the Flaws in Teller’s “Super”

The ENIAC computation was simplified – only the first part (ignition) of the Super problem could be modeled given ENIAC’s memory limits. Several physical effects had to be omitted. Even so, the results were deeply troubling for Teller’s “Classical Super” design:

  • Radiation losses were catastrophic. Energy emitted as bremsstrahlung (braking radiation from electrons) escaped the fuel mass rather than being reabsorbed – the mean free path for these high-temperature photons was measured in kilometers. The inverse Compton effect caused additional cooling of electrons by collisions with photons from the fission initiator. These radiation losses drained energy faster than fusion could produce it.
  • Ignition was uncertain. Even using 95% of ENIAC’s capacity, the calculation “did not truly answer the question of whether or not a Super could be ignited, much less propagate.”
  • Tritium requirements were enormous. The design demanded quantities of tritium that were impractical to produce.

In the spring of 1946, a conference was held at Los Alamos to review the ENIAC results. The principals included Edward Teller, Enrico Fermi, John von Neumann, and Director Norris Bradbury. Teller used these results in his 1946 report, but remained optimistic about the Super concept.

The final blow came in the summer of 1950, when Fermi and Ulam performed additional calculations showing that liquid deuterium “would probably not burn” – there would be no self-sustaining, self-propagating reaction. Every important assumption behind the Classical Super was wrong. This forced the radical redesign (Teller-Ulam configuration, using radiation implosion) that eventually produced the working hydrogen bomb.

Historical Significance

This was one of the first demonstrations that electronic computers could tackle problems impossible by hand or with desk calculators, and it set the stage for the entire field of computational physics. The ENIAC thermonuclear problem was, in a real sense, the birth of scientific computing.

Monte Carlo Work (1949–1955)

After the war, Frankel worked with physicist Berni Alder at Caltech to develop Monte Carlo analysis methods for molecular dynamics. In 1950, he traveled to Manchester, England, to run Alder’s calculations on the Manchester Mark 1 computer – one of the earliest stored-program computers. Their results were published in the Journal of Chemical Physics (1955).

In 1947, Frankel and Metropolis published a paper on computer integration techniques, further extending the computational methods pioneered on ENIAC.

Security Clearance and the Red Scare

The Revocation

Frankel lost his security clearance in early 1949 – five years before Oppenheimer’s more famous clearance revocation in 1954. The stated reason, according to multiple sources including Grokipedia and Alchetron (drawing on the same underlying material), was “familial political associations” – meaning the accusations were based on the political activities or associations of family members, not on Frankel’s own conduct. Some sources describe the questions raised about his clearance as having been considered “baseless” and related to “past acquaintances.”

Context: Oppenheimer’s Students

Frankel was not named among the Oppenheimer students who were known Communist Party members – that group included David Bohm, Giovanni Rossi Lomanitz, Joseph Weinberg, and Bernard Peters. Frankel appears to have been in a different category: not accused of personal CP membership, but caught by the expanding net of guilt-by-association that characterized the early Cold War security apparatus. The phrase “familial political associations” is tantalizingly vague – it could refer to a relative’s CP membership, to attendance at CP-adjacent gatherings, or to any number of associations that the FBI considered suspect.

Frankel’s clearance revocation (1949) preceded Oppenheimer’s (1954) by five years. Both were former Berkeley physicists, both were Jewish, both were associated with the left-liberal milieu of 1930s–1940s Berkeley. But there is no direct evidence that Frankel’s case was triggered by the investigation of Oppenheimer specifically. The timing – 1949 – corresponds to the broader tightening of security procedures following the Soviet atomic test (August 1949) and the Fuchs espionage case, and to the intensification of HUAC investigations targeting the Manhattan Project.

No Hearing Records Found

No transcript or record of a formal hearing for Frankel’s clearance revocation has been located in publicly accessible archives. He does not appear in the published transcripts of the Oppenheimer security hearing (1954), either as a witness or as a subject of discussion. His case appears to have been an administrative action rather than a formal adversarial proceeding.

The Effect

The loss was devastating: Frankel was cut off from the national laboratory system and the government-funded research that dominated Cold War physics. He reinvented himself as an independent computer consultant – a path that, ironically, led to his most consequential contribution (the LGP-30).

Caltech, MINAC, and the LGP-30 (1954–1956)

MINAC

Recruited by Caltech in the early 1950s to head its new digital computing unit, Frankel designed the MINAC (Minimal Computer) in 1954. Who specifically recruited him is not documented in accessible sources – Caltech’s president was Lee DuBridge and the physics division chair was Robert Bacher (both with Los Alamos connections), but no source names the individual who brought Frankel to Caltech. It is notable that Caltech hired him despite his loss of security clearance – the institute was not dependent on cleared work in the same way the national labs were.

The MINAC was a compact, general-purpose computer prototype that emphasized simplicity and reliability. It used:

  • 113 vacuum tubes for logic circuitry (most computers of the era used thousands)
  • Germanium diodes from Hughes Aircraft for switching (solid-state diode logic)
  • Magnetic drum memory as primary storage

Frankel described it as a “simple general purpose computer.” The design philosophy was radical for its time: instead of building the largest, fastest machine possible, Frankel optimized for small size, low cost, reliability, and single-user operation.

LGP-30: The Librascope Deal

Librascope was a Glendale, California division of General Precision Inc., a defense contractor. The company had been founded in 1937 by Lewis W. Imm to improve aircraft load balancing, and was acquired by General Precision in 1941. Librascope saw commercial potential in Frankel’s MINAC design and licensed it from Caltech.

Frankel and Caltech graduate student James Cass (B.S. ‘50, M.S. ‘53) were hired by Librascope to transform the prototype into a production-ready product. Cass served as the primary engineer at Librascope; his engineering group “productized” the MINAC – turning a laboratory prototype into a reliable commercial machine. The nature of Cass’s contribution appears to have been translating Frankel’s design-for-minimum-components philosophy into something that could be manufactured, serviced, and sold.

The result was the Librascope General Purpose 30 (LGP-30), first manufactured in 1956:

  • Size: Desk-sized (roughly the size of a large office desk)
  • Weight: approximately 800 pounds
  • Memory: Magnetic drum, 6.5 inches diameter x 7 inches long; 64 tracks of 64 words each (32 bits per word) = 4096 words total
  • Speed: Time between adjacent addresses approximately 2.34 ms
  • Interface: Flexowriter typewriter with punched paper tape
  • Price: $47,000 (approximately $500000 in 2020s dollars)
  • Sales: Over 500 units – extraordinary for that era

The LGP-30 has been called the first personal computer: a single-user machine that sat on a desk and plugged into a wall socket. It was quiet, reliable, and required no special cooling or power supply. Later marketed as the Royal McBee LGP-30 through a joint venture with the Royal McBee Corporation (typewriter manufacturer).

LGP-21

Frankel designed a successor, the LGP-21, using transistors instead of vacuum tubes, further reducing size and cost.

Thread from Nuclear Weapons to Chaos Theory

Lorenz and the LGP-30

The LGP-30’s most famous user was Edward Lorenz at MIT. Around 1960, Lorenz installed a Royal McBee LGP-30 in his office at MIT’s Department of Meteorology. His supervisor Robert White, in the General Circulation Project led by Victor Starr, had suggested he buy a computer. The LGP-30 was 1000 times faster than a desk calculator, and Lorenz used it to run a simple 12-variable weather model (the geostrophic form of a two-level baroclinic model with 12 spectral variables), producing a day’s worth of virtual weather every minute.

In 1961, with the assistance of programmer Margaret Hamilton (later famous for Apollo guidance software), Lorenz wanted to re-examine a particular sequence of output. To save time, he restarted the simulation from the middle rather than the beginning, typing in values from the printout. The LGP-30 worked internally with 6-digit precision, but the printout rounded to 3 digits – so a value like 0.506127 was entered as 0.506. The resulting forecast diverged completely from the original run. This was the discovery of sensitive dependence on initial conditions – the foundation of chaos theory and the “butterfly effect.”

Lorenz’s 1963 paper “Deterministic Nonperiodic Flow” in the Journal of the Atmospheric Sciences, computed on a machine Frankel designed, is one of the most consequential scientific papers of the twentieth century.

Did Frankel and Lorenz Ever Meet?

No evidence of any contact has been found. Frankel was a commercial computer consultant in the Los Angeles area; Lorenz was a meteorologist at MIT in Cambridge, Massachusetts. They occupied completely different professional worlds. Lorenz purchased the LGP-30 through Royal McBee’s commercial sales channel – there is no indication he knew or cared who designed the machine. Frankel, for his part, was designing calculators for SCM Marchant and Diehl by the early 1960s and appears to have had no connection to meteorology or atmospheric science.

The connection between the two men is extraordinary but almost certainly entirely coincidental – a machine designed by one man for one purpose was used by another man for an entirely unrelated purpose, and the result changed science. Frankel died in 1978; Lorenz died in 2008. Neither, as far as any accessible source records, ever acknowledged the other’s existence.

The Thread

The thread runs thus: Frankel organized human computers at Los Alamos for the atomic bomb (1943), learned to program ENIAC for the hydrogen bomb (1945), applied Monte Carlo methods to molecular dynamics (1949–55), designed a cheap personal computer at Caltech (1954), and that computer ended up in Lorenz’s hands at MIT, where it revealed that deterministic systems can behave unpredictably. From nuclear weapons through the first personal computer to the discovery of chaos – Frankel is the connecting tissue.

Later Career

After the LGP-30, Frankel continued as an independent computer consultant and designer:

  • CONAC: Designed a computer for Continental Oil Company (1954-1957)
  • Packard Bell PB-250: Served as consultant on this commercial computer
  • SCM Marchant Cogito calculators (1965): Developed programmable desk calculators
  • NIC-NAC: A microcoded calculator prototype
  • Diehl Combitron: Contracted by Diehl (West Germany) to develop a desktop calculator

Death

Stanley Phillips Frankel died in May 1978, at the age of 58 (some sources say 59). No public obituary has been located. The cause of death is not documented in any accessible source. Wikipedia’s “List of suicides” does not include him. No major scientific organization published a memorial or tribute – a telling indicator of how thoroughly the security clearance revocation erased him from the scientific establishment. He received no awards, no honorary degrees, no named lectureships. The man who helped design one of the first personal computers and whose ENIAC work launched computational physics died in near-total obscurity.

Personality and Appearance

The Los Alamos Badge Photo

A Los Alamos security badge photograph survives (Wikimedia Commons: Stanley_P._Frankel_Los_Alamos_ID.png). It shows a young man in his mid-twenties – dark hair, serious expression, the standard wartime ID portrait. No other photographs have been located in public sources.

Feynman’s Portrait

The most vivid character sketch comes from Richard Feynman’s “Los Alamos From Below” (1975 lecture, published in Engineering and Science, February 1976). Feynman calls him “a rather clever fellow” – high praise from Feynman, who was stingy with compliments. The “computer disease” passage paints Frankel as someone consumed by intellectual fascination to the point of practical dysfunction: the kind of mind that, given a machine, cannot stop until he has explored every capability, even when the urgent work of building a nuclear weapon sits waiting. The tone is affectionate and amused, not critical. Feynman is recognizing a kindred spirit – a person who gets lost in the joy of figuring things out.

What Can Be Inferred

No colleague memoirs, oral histories, or interviews with detailed personality descriptions of Frankel have been found. He left no autobiography, no published reminiscences, no recorded interviews. He is absent from the extensive oral history collections at both the Atomic Heritage Foundation / Nuclear Museum and the Niels Bohr Library. This silence is itself revealing: Frankel was evidently not a self-promoter, not a public figure, not someone who cultivated a reputation. He did the work and moved on. The contrast with contemporaries like Teller, Feynman, or even Metropolis – all of whom left rich personal records – is stark.

The “Computer Disease” Quote – Full Passage

From Feynman’s “Los Alamos From Below” (Caltech Archives):

“Well, Mr. Frankel, who started this program, began to suffer from the computer disease that anybody who works with computers now knows about. It’s a very serious disease and it interferes completely with the work. The trouble with computers is you play with them. They are so wonderful. You have these switches – if it’s an even number you do this, if it’s an odd number you do that – and pretty soon you can do more and more elaborate things if you are clever enough, on one machine.

“And so after a while the whole system broke down. Frankel wasn’t paying any attention; he wasn’t supervising anybody. The system was going very, very slowly – while he was sitting in a room figuring out how to make one tabulator automatically print arctangent X, and then it would start and it would print columns and then bitsi, bitsi, bitsi, and calculate the arc-tangent automatically by integrating as it went along and make a whole table in one operation.

“Absolutely useless.”

Feynman then explains that he was pulled from his own theoretical work to take over the IBM computing group, which he reorganized into a parallel pipeline – “and we got speed.” The key context: Feynman introduces Frankel earlier in the passage as “a rather clever fellow by the name of Stanley Frankel” who “realized that it could possibly be done on IBM machines.” So the arc of the story is: Frankel had the original insight (use IBM tabulators for bomb calculations), set up the system, then got lost in the machine’s possibilities. Feynman rescued the schedule but credits Frankel with the foundational idea.

Connections to Others

  • J. Robert Oppenheimer: Mentor at Berkeley; post-doctoral advisor; both lost security clearances in the Red Scare
  • Richard Feynman: Colleague at Los Alamos; the “computer disease” anecdote
  • Nicholas Metropolis: Traveled together to Moore School to learn ENIAC programming; co-designed H-bomb calculations; co-authored papers on computational methods
  • Edward Teller: The H-bomb calculations Frankel ran on ENIAC directly informed Teller’s thermonuclear weapons work
  • Berni Alder: Collaborator on Monte Carlo methods at Caltech
  • Edward Lorenz: Used Frankel’s LGP-30 to discover chaos theory (though they likely never met)
  • Mary Frankel: Wife, who worked as a human computer in Group T-5 at Los Alamos
  • Margaret Hamilton: Programmer who assisted Lorenz on the LGP-30 at MIT

Mary Frankel – Expanded Details

Mary P. Frankel arrived at Los Alamos with Stan in the spring of 1943. She was hired as a human computer in the T (Theoretical) Division. Her maiden name and educational background have not been found in any accessible source – she is consistently identified only as “Mary P. Frankel” and described as a “mathematician.”

Her role was more significant than a simple calculator operator. According to a direct quote from one of the women she supervised: “Our immediate supervisor was Mary Frankel, who set up the problems for us to run through the machines.” Mary designed the calculating sheets – detailed rows specifying each operation – that the other human computers followed step by step on the Marchant and Friden desk calculators. Complicated operations were handed off between different human computers in a pipeline arrangement. She was, in effect, a programmer before the word existed: translating physics problems into sequences of arithmetic operations that could be executed by a team of people with mechanical calculators.

The fact that she was supervising other women while her husband Stan was organizing the broader computational infrastructure of T-Division (and later the IBM tabulating machine operation) means the Frankels were a husband-and-wife team at the center of Los Alamos computational work. No information about Mary Frankel after Los Alamos has been found – whether she continued in computing, what happened to her after Stan’s death, or whether she is still living.

The phrase “familial political associations” that cost Stan his security clearance in 1949 is intriguing in this context – it could refer to Mary’s family, to Stan’s own family, or to other relatives. No further details have been found.

Sources