The pogo bounce had been observed (although to a much smaller degree) on Apollo 4, so its appearance during Apollo 6 did not come as a complete surprise. Also, five years earlier, in 1963, pogo had threatened to end the Gemini program when the Titan II suffered this phenomenon on launch after launch. Its apparent cause was a partial vacuum created in the fuel and oxidizer suction lines by the pumping rocket engines. This condition produced a hydraulic resonance - more simply, the engine skipped when the bubbles caused by the partial vacuum reached the firing chamber. Sheldon Rubin of the Aerospace Corporation had finally suggested installing fuel accumulators and oxidizer standpipes, to ensure a steady flow of propellants through the lines. This had solved the Gemini launch vehicle problems, and NASA had this background experience to draw on when the Saturn V began having pogo troubles.* 39
Pogo on Apollo 4 had been measured at one-tenth g, much less than the one-fourth g set as the upper limit in Gemini. The lower oscillation was probably the result of carrying just "a hunk of junk," to simulate lunar module weight, on the earlier flight. But a test article flown on Apollo 6 had the shape and weight of a real lander in the adapter. This change in mass distribution coupled back into the fuel system problem and increased the pogo oscillations. The mission analysts later discovered that two of the Saturn engines had been inadvertently tuned to the same frequency, probably aggravating the problem. (Engines in the Saturn V cluster were to be tuned to different frequencies to prevent any two or more of them from pulling the booster off balance and changing its trajectory during powered flight.)
The rocketeers at Huntsville first wanted to know from Houston whether a crew could have withstood the vibration levels on Apollo 6. If so, the next Saturn V flight could be manned, even without a pogo cure. Low informed Saturn V Program Manager Arthur Rudolph that these levels could not be tolerated. Marshall also asked whether the emergency detection system could be used to abort the mission automatically if such high vibrations again occurred. During Apollo 6, the system had cast one vote for ending the mission. Had it cast a second vote, abort would have been mandatory. Low and chief astronaut Donald Slayton did not want to use the system in an automatic pogo abort mode. Low met with George H. Hage, Phillips' deputy, and they decided on the immediate development of a "pogo abort sensor," a self-contained unit that would monitor and display spacecraft oscillations. From what the sensor told him, a spacecraft commander could decide whether to continue or stop the mission.40
Marshall Space Flight Center pulled an S-IC stage out of Michoud Assembly Facility, brought it to Huntsville, and erected it in a test stand. By May, Huntsville, Houston, and Washington Apollo officials were ready to attack the pogo problem. Hage agreed to head the activity until Eberhard Rees could finish his task on the command module at Downey and take over. At one time during the pogo studies, Lee B. James (who had replaced Rudolph as the Huntsville Saturn V manager) said, 1,000 engineers from government and industry were working on the problem.41
Out on the West Coast, at the rocket engine test site at Edwards Air Force Base, Rocketdyne started testing its F-1 engine in late May. In the first six tests, helium was injected into the liquid-oxygen feed lines in an attempt to interrupt the resonating frequencies that had caused the unacceptable vibration levels. In four of the six tests, the cure was worse than the disease, producing even more pronounced oscillations. The Saturn V people at Marshall also tried helium injection, but their results were decidedly different. No oscillations whatsover were observed. Tests using the S-IC stage's prevalves as helium accumulators were then conducted at both Edwards and Marshall. The prevalves were in the liquid-oxygen ducts just above the firing chambers of the five engines and were used to hold up the flow of oxygen in the fuel lines until late in the countdown, when the fluid was admitted to the main liquid-oxygen valves in preparation for engine ignition. The prevalves were modified to allow the injection of helium into the cavity about 10 minutes before liftoff; the helium would then serve as a shock absorber against any liquid-oxygen pressure surges.
What had happened to the S-II and S-IVB stages, with two of the five J-2 engines shutting down in one case and the single J-2 engine refusing to start in the other, was more of a mystery than pogo. During tests at Arnold Engineering Development Center, at Tullahoma, Tennessee, engineers discovered that frost forming on propellant lines when the engines were fired at ground temperatures served as an extra protection against lines burning through. But frosting did not take place in the vacuum of space; the lines could have failed because of this. Also, in the line leading to each of the engines was an augmented spark igniter. Next to the igniter was a bellows. During ground tests, liquid air, sprayed over the exterior to cool it, damped out any vibrations. Vacuum testing revealed that the bellows vibrated furiously and failed immediately after peak-fuel-flow rates began. These lines were strengthened and modified to eliminate the bellows.42
Another item noticed by the flight control monitors during the boosted flight of Apollo 6 (and later confirmed by photographs) was that a panel section of the adapter that housed the lander had fallen away just after the Saturn V started bouncing. The controllers had been amazed that the structural integrity was sufficient to carry the payload into orbit. James Chamberlin in Houston discovered that thermal pressure (and therefore moisture) had built up in the honeycomb panels during launch; with no venting to allow the extra pressure to escape, the panel had blown out. A layer of cork was applied to the exterior of the adapter to keep it cooler and to absorb the moisture, and holes were drilled in the adapter panels to relieve the internal pressure if heat did build up inside on future launches.43
Although Marshall was responsible for stability and dynamic structural integrity throughout the boost phase, the Manned Spacecraft Center could not afford to sit on the sidelines and watch while its sister center wrestled with these problems. Houston had to get an Apollo payload stack together for structural testing. On 16 May 1968, Low and James decided to use a "short stack" (the S-IC stage would be left out at this time but could be incorporated later).** Astronaut Charles Duke was sent to Huntsville to keep information flowing between the centers, and Rolf Lanzkron was assigned by Low to manage the spacecraft dynamic integrity testing, which was satisfactorily completed on 27 August with no major hardware changes found necessary.44
* The Gemini launch vehicle engines were hypergolic, that is, its oxidizer and fuel burned on contact to produce thrust. Since the Saturn first stage (S-IC) engines were cryogenic, the propellant and oxidizer needed an igniter to produce burning - and no one expected a similar pogo problem with the larger booster.
** The stack comprised an S-IVB forward skirt, launch vehicle instrument unit, spacecraft-lunar module adapter, LM-2, a service module, a Block I command module, and the launch escape system from boilerplate 30.
39. MSC, "Apollo 6 Mission Report," p. 1-3; MSC, "Apollo 4 Mission Report," MSC-PA-R-68-1, January 1968, p. 5.1-8; Scott H. Simpkinson, telephone interview, 28 Aug. 1975; Barton C. Hacker and James I. Grimwood, On the Shoulder of Titans: A History of Project Gemini, NASA SP-4203 (Washington, 1977), p. 136.
40. Simpkinson interview; Boone to MSC Mechanical Panel Cochm., Attn.: Jenkins, "Spacecraft (structure/man) limits for longitudinal oscillations on the manned AS-503," n.d. [ca. April 1968]; Low to Rudolph, KSC, 22 April 1968; Hodge, minutes of Fifth Apollo Crew Safety Review Board Meeting, 16-18 April 1968; Low to Hodge, "Apollo Crew Safety Review Board activities," 23 April 1968; Low to Maynard, "POGO abort sensor," 22 May 1968; Armistead Dennett, minutes of Pogo Sensor Planning Meeting, 3 June 1968; minutes of 24th Crew Safety Panel Meeting, 19 June 1968.
41. Mueller Report, 29 April 1968; Low memo for record, "Management meeting on POGO," 18 May 1968; Abbey, ASPO Staff Meeting, 20 May 1968; MSC news release 68-50, [ca. 18 July 1968].
42. Mueller Reports, 15 April, 31 May, 7, 14, and 21 June, 8, 15, and 22 July 1968; 24th Crew Safety Panel Meeting; MSC release 68-50; Charles I. Duke, Jr., to Mgr., ASPO, "POGO activities," 12 July 1968; Astronautics and Aeronautics, 1968: Chronology on Science, Technology, and Policy, NASA SP-4010 (Washington, 1969), p. 135; NASA Nineteenth Semiannual Report to Congress, January 1-June 30, 1968 (Washington, 1969), pp. 18-19.
43. "Apollo 6 Review," 21 April 1968; Low presentation, "Manned Space Flight Management Council Review, Apollo Spacecraft Program," 7 May 1968; Low to Seymour C. Himmel, 14 Nov. 1968; MSC, "Review of AS-502 Structural Anomaly Activities," 27 June 1968; McClellan to Low, 1 July 1968; Donald D. Arabian to Low, "Summary of SLA Anomaly," 8 Oct. 1968; MSC, "Apollo Anomaly Status," PT-ASR-6, 1 Oct. 1968; JSC, "Apollo Program Summary Report," p. F-4.
44. Low to Phillips, 25 May 1968, with encs., "Apollo Space Vehicle Dynamic Integrity," MSC Announcement 68-67, 21 May 1968, and "Management of MSC Space Vehicle Dynamic Integrity efforts," MSC Announcement 68-69, 24 May 1968; Quarterly Status Rept. no. 25, 30 Sept. 1968, pp. 6, 7, 9; Low to Bolender and Kleinknecht, "Anticipation of pogo fixes," 3 May 1968; Low TWX to NASA Hq., Attn.: Phillips, "POGO Dynamic, Static and Other Structural Tests," 15 June 1968.