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.

Verified Facts - Post 17 Stanley Frankel

Status: COMPLETE

Sources Consulted

  • Wikipedia: Stan Frankel
  • Atomic Heritage Foundation: Stan Frankel profile
  • Wikipedia: LGP-30
  • Caltech Archives: Feynman “Los Alamos From Below”
  • Haigh, Priestley, Rope arXiv 2103.05705 (PDF, authoritative on Los Alamos computing)
  • Masswerk: LGP-30 technical details
  • Various secondary sources on ENIAC H-bomb calculation

Verified Facts (post is correct on these)

  • Stanley Phillips Frankel born 1919 Los Angeles (post says June 6 1919, sources confirm 1919)
  • Died May 1978 at age 58 (sources confirm)
  • PhD from UC Berkeley
  • Post-doc under Oppenheimer starting 1942 (sources confirm)
  • Worked with Eldred Nelson on neutron diffusion / critical mass
  • Arrived Los Alamos early 1943 (post says 1943 - confirmed)
  • Wife Mary Frankel involved in computing pool
  • Marchant and Friden desk calculators used
  • Recognized IBM tabulators could be programmed (PCAM machines)
  • Caught “computer disease” - arctangent story confirmed by Feynman
  • Feynman took over IBM operation
  • August 1945 traveled with Nick Metropolis to Moore School (ENIAC) - confirmed
  • ENIAC: 30 tons, 40 panels, 17468 tubes (the figures are standard)
  • Half a million punch cards - confirmed
  • ENIAC calculation revealed flaws in Teller’s Super; Teller’s spring 1946 report
  • 1947 paper with Metropolis on computational methods
  • Berni Alder collaboration at Caltech (1949-50)
  • 1950 Manchester Mark 1 trip, paper published 1955 in J Chem Phys
  • Lost security clearance early 1950s red scare (sources say 1949 plausible)
  • Designed MINAC at Caltech (1954) - 113 vacuum tubes + diodes
  • LGP-30 first manufactured 1956 by Librascope
  • Royal McBee/Royal Precision joint venture for sales
  • Designer with James Cass (Caltech grad student)
  • ~500 units sold (320-493 actually manufactured)
  • $47,000 price
  • ~800 pounds
  • Drum 6.5 inches diameter, 7 inches long
  • 4096 word memory, 64 tracks of 64 words
  • Flexowriter for I/O
  • LGP-21 transistorized successor
  • Oppenheimer security clearance: suspended Dec 21 1953, hearings April 1954, revoked June 29 1954 (well-established history)

DISPUTED / INCORRECT CLAIMS in the post

CRITICAL ERROR 1: Group T-5 vs T-6

The post says: “Frankel created Group T-5: teams of women…”

This is WRONG on two counts:

  • T-5 was the desk-calculator HUMAN computing pool, led by Donald (Moll) Flanders, formed summer 1943
  • Frankel and Nelson led T-6 (IBM Computations), formed September 1944 when IBM Computations were removed from T-2 (Serber’s Diffusion Theory group)
  • Mary Frankel set up problems for the T-5 pool to calculate but she was not just a “wife”; she was a key staff member
  • The “teams of women” calculating with Marchants/Fridens were T-5 under Flanders

Source: Haigh, Priestley, Rope arXiv 2103.05705 (peer-reviewed authoritative source on Los Alamos computing infrastructure).

CRITICAL ERROR 2: LGP-30 word length

The post says: “4096 words of 32-bit memory” Correct: 31-bit (the 32nd bit was a spacer on drum, always zero in memory) Sources: Wikipedia LGP-30, Masswerk

CRITICAL ERROR 3: ENIAC calculation date

The post says: “in November 1945 ran the calculation… results were available by December” Correct: The calculation was RUN in December 1945 and January 1946 (not November-December) ENIAC was first put to work on December 10, 1945 on the Los Alamos problem. Programming was designed in fall 1945 after August arrival. Sources: Multiple, including APS, vanemden, Computing and Manhattan Project (AHF).

CRITICAL ERROR 4: Frankel moved aside timing

The post says: “Frankel was moved aside” after Feynman took over. Correct: Frankel was moved to Teller’s F-Division in January 1945 (because of “computer disease”) He was NOT moved aside vaguely after Feynman - the specific date is January 1945. This actually makes the chronology cleaner: Frankel moved to F-Division Jan 1945, then was already with Teller’s group when he and Metropolis went to ENIAC in August 1945 to do the Super calculation. The post buries this connection. Source: Haigh et al.

MODERATE ERROR 5: Adjacent address time

The post says: “The time between adjacent addresses was approximately 2.34 milliseconds” The LGP-30 drum rotated at 3700 rpm (~16.2 ms/revolution); maximum access time per revolution was 17 ms; quick access registers were 0.26 ms. The “2.34 ms” figure is not in standard sources I checked. Possibly the average access time using interlace addressing, but should be verified or removed.

MINOR ISSUE 6: PhD year

The post does not give a year but implies Frankel had PhD by 1942 (when he started post-doc with Oppenheimer). Wikipedia and AHF confirm postdoc began 1942 but neither gives PhD year explicitly. The user-prompt watch-list says “UC Berkeley PhD 1942 under Oppenheimer” - but sources don’t explicitly state 1942, only that postdoc began 1942.

MINOR ISSUE 7: Frankel “supervised” the calculation

The post says: “supervised by a 26-year-old physicist from Los Angeles” 26 fits: born June 6 1919, by November-December 1945 he was 26. But “supervised” oversells - he and Metropolis jointly designed and ran it.

POSSIBLE ERROR 8: 1949 clearance loss date

The post says: “In early 1949, Stanley Frankel lost his security clearance” Sources only confirm “early 1950s red scare.” The 1949 date is plausible but I cannot find a primary-source confirmation in the time available. Should be presented less definitively or sourced.

Source URLs (for Wayback posting)

  • https://en.wikipedia.org/wiki/Stan_Frankel
  • https://ahf.nuclearmuseum.org/ahf/profile/stan-frankel/
  • https://en.wikipedia.org/wiki/LGP-30
  • https://calteches.library.caltech.edu/34/3/FeynmanLosAlamos.htm
  • https://arxiv.org/pdf/2103.05705
  • https://www.masswerk.at/nowgobang/2019/lgp-30
  • https://ahf.nuclearmuseum.org/ahf/history/computing-and-manhattan-project/