The 'maintenance' was maintained assiduously by AO staff. The dish panels--visually-- succumb to jungle rot in a very short time, and that did not affect the performance of the dish. The problem was the mean time between failures for the cables, and the imprudent desire of the NSF not to do 'ahead' replacement because of the cost. They justified this by saying that AO was not on the priorities outlined by the astronomical community. Of course, the outcome of that idiocy is they now spend $100M+ for a salvage operation and returning the karsk ridden former sink hole into yet another piece of Puerto Rican jungle.
Arecibo isn't the first radio telescope to unexpectedly fail. On Nov. 15, 1988, Green Bank Observatory's 300-foot telescope unexpectedly collapsed, leaving a tangle of twisted metal. A drone view of damage to a cable at the Arecibo Observatory in Puerto Rico captured after a second cable failed on Nov. 6, 2020. I'm no expert but it sure looks to me like like that dangling cable was at its end of life. I doubt that went unnoticed. Like so much other infrastructure, standard economic practice is to act blind of potential problems, wait until the catastrophe, and then react. (I-35 bridge colapse in Minneapolis) The Golden Gate suspender rope replacement occured after < 40 years, not a maintenance expense but a logical preemptive capital cost decision to extend its lifespan.
Many radio dishes have collapsed. AO is not your typical dish. Its basically a bridge suspension above a bowl. In most dishes, the dish itself is slewed and the focus point is locked and supported by the dish architecture. When cables snap they look 'aged' . A few hours earlier it probably looked just fine, except for strands that snapped but may not have been visible. Both the NRAO 300 foot and AO exceeded their MTBF life expectancy. In AO's case, much like your bridge reference, the cables should have been replaced more frequently rather than playing the dice on MTBF's. That would have further increased the AO lifetime. But hey---why spend an additional $15M in preventive maintenance when you can spend $100M+ to return an excavated karsk sink hole to a schistosomiasis ridden rivulet? Can't ya just hear the birds tweet already?
For radio amateurs who never visited the site like me, a fond memory is the EME special event conducted by their club KP4AO ten years ago. The ARRL appropriately called it "Moonbounce for Everyone" achieved through contributions from several well known professionals with roots in ham radio. From the article linked above - "What a great QSO party it was! The telescope’s huge forward gain, about 61 dBi at432 MHz, guaranteed that even small stations could get into the game. Many hundreds of stations copied the KP4AO signal after its half-million mile round trip to the Moon and back — some using small handheld Yagis or even a dipole, and in at least one case a rubber flex antenna. The wall of stations responding to our CQs sounded like 20 meters during a DX contest! Even with skillful, well-behaved operators spread out over 15 kHz and more,we had to work hard to pick call signs out of the din. Wideband real-time recordings have enabled usto copy hundreds of additional call signs, after the fact. A total of 242 lucky ones made it into the log with completed two-way QSOs, in about 8 hours of actual operation." First and probably the last time I'll ever hear an EME signal using only a FT-817 and a hand held Arrow beam, similar to this vid,
Yes. The cables. Anything structural, under stress, has the potential to fail eventually. With stress-corrosion and probably some fatigue damage as well, and the likelihood of an unknown defect or unplanned-for stress riser, it's nearly a sure thing - the only question is "when". One can design for a certain maximum lifetime assuming specific fixed environmental conditions and zero maintenance, but there's probabilities involved because everything is variable. That's why inspection and maintenance is so important, because (if you are modeling the lifecycle appropriately) you basically can reset the lifecycle management (assuming you're doing any of that at all). So, while the panels are an important functional component, if the structural components fail, it doesn't matter if the panels are cruddy-looking, or bright and shiny. They're useless, either way, if they're lying in a pile on the ground because a cable snapped and tore them loose. Oh, and of course if you're operating outside the safe-life design envelope anyway, all bets are off. I just saw one of Chip's posts talking about all the extra weight they've added over they years. Of course it was destined to fail if that's the case.
From the photo - besides the cable - there appears to be missing paint on the solid bars and other structural parts, and what looks like rust on them. It doesn’t give the impression of a well maintained facility. I’m not suggesting the on-site staff did anything wrong. No doubt they did what they could within their budget allocation.
Should comment that it may not be the cable itself that failed, but rather the anchoring of the cable either the cable at the anchoring point, or components of the anchoring system. Doesn't matter how strong the cable is, if one end comes loose! Apparently, that is what happened in the August cable fail. According to this article, that cable "slipped out of it's socket". The cable that failed Nov 6, the article says " ... Observers had noted and were tracking broken wires on the main cable that failed Friday. ...
The first cable that failed, in August, is one of the "auxiliary" cables added to compenstate for the extra weight added in '97. When that popped it understandably kicked off a less predictable "domino effect". Speculating, adding suspender cables of different capacities in parallel seems like it would greatly complicate stress modeling, especially in regard to peak effects shared between cables during randomly dynamic (storm) conditions. Being so much younger then the "main" cables, ie, less far into its MTBF curve, makes failure of an "auxiliary" cable even more premature and curious. Perhaps a manufacturing defect or construction botch during installation in the socket.
In a snarky comment about another radio observatory closure, I said that "Arecibo would make a good swimming pool too." Call me Nostradamus...
From Wikipedia regarding Joe Taylor K1JT: "In 1974, Hulse and Taylor discovered the first pulsar in a binary system, named PSR B1913+16 after its position in the sky, during a survey for pulsars at the Arecibo Observatory in Puerto Rico. Although it was not understood at the time, this was also the first of what are now called recycled pulsars: neutron stars that have been spun-up to fast spin rates by the transfer of mass onto their surfaces from a companion star." This is a real loss to the scientific community. I suspect Joe could talk more about the importance of this site.