___________________

¹ The Wiss, Janney, Elstner Associates (WJE) report on the initial auxiliary socket and cable (WJE, 2021, Auxiliary Main Cable Socket Failure Investigation, WJE No. 2020.5191, June 21); NASA Engineering and Safety Center (NESC) and Kennedy Space Center report (G.J. Harrigan, A. Valinia, N. Trepal, P. Babuska, and V. Goyal, 2021, Arecibo Observatory Auxiliary M4N Socket Termination Failure Investigation, NASA/TM−20210017934, NESC-RP-20-01585, NASA Engineering and Safety Center, Langley Research Center, June 15, https://ntrs.nasa.gov/api/citations/20210017934/downloads/20210017934%20FINAL.pdf), and forensic investigations performed by Thornton Tomassetti, Inc. (TT)—the “TT Interim Report” (TT, 2021, Arecibo Telescope Collapse: Forensic Investigation Interim Report, NN20209, prepared by J. Abruzzo and L. Cao, November 2) and the “TT Final Report” (TT, 2022, Arecibo Telescope Collapse: Forensic Investigation, NN20209, prepared by J. Abruzzo, L. Cao, and P.E. Pierre Ghisbain, July 25, https://www.thorntontomasetti.com/sites/default/files/2022-08/TT-Arecibo-Forensic-Investigation-Report.pdf).

² TT Final Report, p. 1.

³ TT Final Report, Appendix C, Table 4, p. 9.

⁴ TT Final Report, p. 188, Figure 25.

of 78⁵ zinc-filled sockets. In over a century of successful use prior to the Arecibo Telescope’s collapse, all the forensic investigations agreed that such a spelter socket failure had never been reported. The committee considered the following unanswered questions about the Arecibo Telescope’s collapse:

1. Why did the Arecibo Telescope sockets/cables fail despite their widely used safety factor above 2? (the ratio of cable strength to applied load)
2. Why did this unprecedented spelter socket accelerated zinc creep failure mode appear in the Arecibo Telescope and nowhere else in more than a century of successful spelter socket use?
3. In the Arecibo Telescope environment, why did a relatively young auxiliary socket, M4N-T, fail first, long before any main cable socket with more than twice as much age with three-plus decades of previous service (1963–1997) at an even lower safety factor?
4. Why did four of the six platform auxiliary cable sockets have no further pullout beyond the ⅜ inch they gained during installation (as measured in the Lehigh University socket test)⁶ and thus exhibited zero measured creep even after 23 years of service?
5. Why did all six auxiliary cable tower sockets exhibit cable pullout beyond their initial ⅜ inch, indicating some creep after installation, and why did all the auxiliary cables exhibit more socket pullout on the tower end than the platform end of the same cable?
6. Why did the rate of wire breakage in the main cables decrease significantly after 1974?
7. Why did the auxiliary cables have no recorded wire breaks even after 23 years of service?

The only hypothesis the committee could develop that provides a plausible but unprovable answer to all these questions and the observed socket failure pattern is that the socket zinc creep was unexpectedly accelerated in the Arecibo Telescope’s uniquely powerful electromagnetic radiation environment. The Arecibo Telescope cables were suspended across the beam of “the most powerful radio transmitter on Earth.”⁷ The other investigations failed to note several failure patterns and provided no plausible explanation for most of them. To answer these questions with empirical evidence instead of only the inferences that can be made from the existing data, a more comprehensive and widespread forensic analysis of “good” and “bad” socket workmanship and the low-current, long-term effect on zinc creep is required. This report makes the recommendation to provide this evidence using the remaining socket and cable sections recovered from the site.

Recommendation: While still available, the National Science Foundation should offer the remaining socket and cable sections to the research community for continued fundamental research on large-diameter wire connections, the long-term creep behavior of zinc spelter connections, and materials science.

In preparing this report, the committee first assessed “what happened” and then provided analysis as to the most probable explanation it could develop with the available evidence as to “why it happened.” The committee gathered important information from the Arecibo Telescope’s administrators, engineers, and technicians, as well as NSF, as well as expertise from bridge and structural engineers. The committee also reviewed the in-depth forensic reports performed after the failure and reviewed the structural analysis, engineering plans, inspection reports, and photographs of the Arecibo Telescope, as well as the various repair proposals. Finally, a literature review was performed of zinc spelter sockets, cable connection pullout, zinc creep, static and dynamic loading of zinc, and electric current effects on zinc creep. In addition, due to the technical complexity of this report, before the report was finalized, NSF was provided an opportunity to review portions of the text and make suggestions relating to any perceived technical inaccuracies or factual errors.

The committee concluded that a 39-month failure sequence (Figure 2-1), starting with Hurricane Maria (hereafter “Maria”) on September 20, 2017, then a Category 4 storm, led to the telescope’s collapse. Structural analysis

___________________

⁵ TT Final Report, p. 46.

⁶ TT Final Report, Appendix N.

⁷ A.P.V. Siemion, et al., 2011, “Developments in the Radio Search for Extraterrestrial Intelligence,” 2011 XXXth URSI General Assembly and Scientific Symposium, https://doi.org/10.1109/URSIGASS.2011.6051263.

performed as part of the forensic investigations established that the wind loading from the hurricane should not have damaged the telescope’s cable structure or caused any additional socket cable pullout. The structural analysis demonstrated that Maria’s wind loading did not increase any cable tension below a nominal safety factor of 1.8.⁸ However, the failure sequences occurred at cable loads well below the traditional factor of safety through some mechanism(s) not previously reported. The sequence likely began with Maria because inspections that were performed on the Arecibo Telescope sockets in 2003 and 2011 by Ammann & Whitney (A&W), a structural engineering firm, reported that “the end socket ½” cast zinc leading edge separation was observed at all cables but has not measurably increased since reported in the 2003 survey.”⁹ The best available information would indicate cable pullouts remained on the order of ½ inch until after Maria. As a result of sparse inspection documentation, there is uncertainty around how much damage was evident immediately post-Maria. Inspections made in late 2018 and early 2019 observed cable slippage greater than 1.5 inches on auxiliary sockets at the ground end of backstay B12W and 1.125 inches at the tower end of M4N. “The cable slips were evidence of structural distress in the sockets and should have raised a concern that cables may fail.”¹⁰ The question is why cable slip did not spark greater concern. It might be that the gravity of the situation was not recognized for the following two reasons: (1) The traditional safety factor based on cable strength divided by cable loads was still comfortably above 2 depending on the particular cable, and (2) in structures using spelter connections, the weak link has never been the connector itself.

The Arecibo Telescope gave fair warning post-Maria that it was in structural distress through increasing cable socket pullout. Upon reflection, the unusually large and progressive cable pullouts of key structural cables that could be seen during visual inspection several months and years before the M4N failure should have raised the highest alarm level, requiring urgent action. The lack of documented concern from the contracted engineers about the inconsequentiality of cable pullouts or the safety factors between Maria in 2017 and the failure is alarming. Given the observed excessive cable pullout, continued use of the factor of safety calculations based on the original design ignored the impact of any degradation mechanism. Safety factors based on cable strength are not pertinent to failure modes involving creep, stress corrosion cracking, hydrogen embrittlement, fatigue, or the integrity of the socket connections.

These mechanisms routinely occur at applied stresses below half of the yield strength in a material undergoing one of these processes. Unprecedented structural distress evidence in any of NSF’s massive structures should not be casually dismissed. The risks posed by the structural distress should have been alarming to inspectors, taking into consideration the safety and lives of the facility’s personnel. Despite the safety factors, these risks merited an urgent response.

Nothing in the ASCE 19¹¹ or AASHTO M 277¹² standards indicates that any movement (pullout) of the cables from their sockets after initial installation (approximately ⅜ inch) was acceptable or justifiable. The structural analysis demonstrated that even Maria wind loading should not have added significant incremental stress to affect the safety factor nor occurred over a time period great enough to affect PLC or cause additional cable pullout. The reliance by the consultants (before and after the first cable failure in 2020) on a perceived allowable pullout of one-sixth of the cable diameter, which should only be seen at loading at 80 percent of ultimate cable strength, does not align with the AASHTO M 277 standard guidance. The committee, therefore, disagrees with the suggestion made in the Thornton Tomasetti, Inc. (TT) 2022 report, Arecibo Telescope Collapse: Forensic Investigation,¹³ to use the D/6 limit as a threshold for slip monitoring.

Almost 3 years after Maria, on August 10, 2020, the tower end of auxiliary cable 4, labeled M4N-T, at less than half its design load, pulled out of its zinc-filled spelter socket and failed. The loose cable struck the Gregorian dome and crashed onto the dish below. At the time of failure, it had only been in service 23 years after the 1997

___________________

⁸ TT Final Report, Appendix J, Figure 34, p. 35.

⁹ Ammann & Whitney, 2011, “Arecibo Radio Telescope Structural Condition Survey,” Cornell University Archives, Arecibo Ionospheric Observatory Records #53-7-3581, Division of Rare and Manuscript Collections, Cornell University Library Box 37, Folder 8, March, p. 2.

¹⁰ TT Interim Report, p. 25.

¹¹ Refers to the American Society of Civil Engineers standard “Structural Applications of Steel Cables for Buildings,” ASCE 19.

¹² Refers to the American Association of State Highway and Transportation Officials standard “Standard Specification for Wire Rope and Sockets for Movable Bridges,” AASHTO M 277.

¹³ TT Final Report, p. 49.

upgrade. The main cables had seen roughly 57 years of service by August 2020 with no socket failures at that point in time, with at least 30 of those years (1963–1993) at a ~10 percent higher static dead-load safety factor (1.98 as opposed to 2.18) than the auxiliary cables saw before the M4N-T failure. The socket that failed first, M4N-T, was not the most heavily loaded, nor was the brooming (the physical spread of the cable wires before being cast in the molten zinc, Figure 1-6¹⁴) of this socket’s wires the worst found among the Arecibo Telescope’s socket connections.

After the M4N-T failure, the main cables to Tower 4, M4 (×4), were loaded to 646 kips,¹⁵ compared to 516 kips in the M12 (×4) mains and 497 kips in the M8 (×4) mains, as illustrated in Figure 2-7. This re-distribution of loads reduced the TT-computed (but incorrect) safety factor for the M4 (×4) mains to 1.6, making the M4 (×4) mains significantly more heavily loaded compared to their original strength than any other remaining cables on the telescope. Thus, while the cables should not have failed even at this load, it is not surprising that, given the evident socket degradation, the second and third failures were in the M4 (×4) mains.

On November 6, 2020, one of the M4 (×4) main cables on Tower 4, M4-4, failed again at the zinc-filled spelter socket (Figure 2-7). Interim repairs were set to begin on November 9. The material was already being staged when the second cable pulled out. After this second cable failure, the loads on the three remaining M4 (×3) cables increased to 800 kips, reducing the erroneously determined safety factor to 1.3¹⁶ for the remaining M4 cables On November 19, NSF announced that safe repairs would not be possible and that Arecibo Telescope would be closed in a controlled decommissioning. On December 1, M4-2, one of the three remaining zinc-filled sockets holding the M4 (×3) cables on Tower 4, failed, increasing the load in the two remaining M4 (×2) cables to 1,062 kips above their nominal strength. The 913-ton platform collapsed, swung across the dish, and smashed through the reflector.

Absent Maria, the committee believes the telescope would still be standing today but might have eventually collapsed if its unique, accelerated zinc creep had not been addressed before it was decommissioned. The long-term zinc creep failure of the Arecibo Telescope sockets and the subsequent cable pullout has never been documented elsewhere despite a century of zinc-filled cable spelter sockets use.¹⁷ The type, size, length, and fittings of the cables used in the Arecibo Telescope (whether the original cables constructed in the 1960s or the auxiliary cables installed in the 1990s) were catalog-selected items, not at all unusual, with decades of proven performance. Extensive structural modeling of the Arecibo Telescope, independently confirmed by laser cable sag surveys, validated that under all static and cyclic loading conditions, the cable loads barely exceed half the nominal cable strength. While PLC can occur at stresses well below yield, such a creep failure has never been reported in spelter socket zinc.

The Arecibo Telescope’s cable design adhered to the standards of practice during the original design and the subsequent addition of the Gregorian dome in 1997. Accounting for static load, cyclic dynamic loading from thermal effects, wind, and earthquakes, and operating in a corrosive tropical environment, the telescope’s engineering design and material specifications were reasonable for the original construction and subsequent upgrades. Construction procedures and workmanship were adequate to ensure long-term structural integrity. Extensive investigation and testing revealed no defects in the Arecibo Telescope’s design, materials, or workmanship that contributed materially to its collapse. The committee did not find sufficient evidence or analysis pointing to an unrecognized design fault, defects in the socket construction, other environmental effects such as hydrogen embrittlement, or some unobserved dynamic load condition based upon connection design and platform geometry. However, these factors cannot be fully discounted. As stated previously, a more comprehensive and widespread forensic analysis of both “good” and “bad” socket workmanship and mechanisms to accelerate zinc creep would be required.

A potential mechanism for spelter socket zinc creep acceleration not considered in other analyses was the effect of low-current electroplasticity (LEP). The cables whose sockets failed were suspended in a unique and powerful radio telescope environment, capable of inducing current in the cables at some level. Electric current flowing through zinc has been found to increase its creep rate but under laboratory conditions significantly different

___________________

¹⁴ Brooming refers to the physical spread of the cable wires before being cast in the molten zinc. Refer to Figure 1-6.

¹⁵ A kilopound (kip) is a non-metric unit of force equal to 1,000 pounds-force. 1 kip is equivalent to 4448.2216 Newton.

¹⁶ TT Final Report, Appendix G, Figure 22, p. 17.

¹⁷ “Wire Cables of Various Types and Materials Tested by U.S. Bureau of Standards,” 1915, Engineering Record 72(19):567–568.

than the spelter socket service in the Arecibo Telescope.¹⁸ While there is not enough data or empirical evidence to prove LEP as a causal mechanism for the acceleration of the socket zinc creep, no other mechanism has been found likely. The circumstantial evidence and the cable pullout patterns offer support for the role of LEP. LEP provides a physically plausible but unproven mechanism to answer the outstanding questions described above about why this spelter socket failure mode was seen in the Arecibo Telescope and nowhere else in history, the non-uniformity in the rate and pattern of the cable pullouts, the failure of a young auxiliary cable socket first, and the timing of the Arecibo Telescope’s cable wire breaks.

All the reported experimental zinc electroplasticity (EP) data were developed at current densities orders of magnitude higher than those possibly present in the Arecibo Telescope but measured in laboratory experimental periods that were orders of magnitude shorter than the telescope’s socket zinc service. There are no reported experimental data concerning low-current, long-term EP, which the committee has lumped together under the term “LEP,” affecting zinc’s creep mechanisms over decades. There was also no reported measuring capability or data from the Arecibo Telescope concerning induced currents or electromagnetic effects in the cables. Measurements of induced current in the Arecibo Telescope’s cables at peak broadcast power or the quality of the various tower and platform grounding paths were not recorded and thus could not empirically validate this explanation. The timing and patterns of the Arecibo Telescope’s socket failures make the LEP hypothesis the only one that the committee could find that could potentially explain the failure patterns observed. This mode of accelerated PLC would have also been operative at stresses below the strength of the cable.

By necessity, NSF’s cutting-edge research will occasionally require unique custom-designed facilities that place conventional structural designs and materials into new and unprecedented operating regimes where prior experience in conventional environments is not a reliable guide. Unprecedented failure modes in NSF’s other large facilities can never be fully anticipated. But to the extent reasonable, their onset and progression can and should be detected by careful condition monitoring of performance, which must be increased and not decreased with age. TT observed, “The available information [about the Arecibo Telescope’s condition] is generally less comprehensive and detailed after the second upgrade in 1997. The scope of the inspections performed by A&W was considerably reduced, and the quarterly maintenance reports were replaced with simple maintenance logs.”¹⁹ For aged structures such as the Arecibo Telescope, additional facility maintenance and monitoring (and their associated costs) may be warranted. The committee does not know how much of this monitoring and inspection reduction was caused directly or indirectly by NSF’s reduction in Arecibo funding over its final decade of service. The committee concluded that the safety consequences of a structural failure of the Arecibo Telescope were not considered in decision-making during its design and operation or in decisions about extending its life.

This report makes recommendations for critical facilities to have formal facility operation and inspection manuals and routine independent monitoring and assessment.

___________________

¹⁸ A. Lahiri, P. Shanthraj, and F. Roters, 2019, “Understanding the Mechanisms of Electroplasticity from a Crystal Plasticity Perspective,” Modelling and Simulation in Materials Science and Engineering 27(8), https://doi.org/ARTN08500610.1088/1361-651X/ab43fc.

¹⁹ TT Interim Report, p. 6.