Re: The agency is recognized the world over as the most careful and risk-averse space agency.
The N1 blew up three times, killing nobody.
Nitpick: 4 times.
It killed nobody because
Correct.
they were sensibly doing unmanned test launches.
There was little sensible about the Russian N1 test program because the Russians skipped a lot of basic tests under limited budget, limited time, and enormous political pressure. The N-1 flights were analogous to jumping to Apollo 4 without doing Little Joe II, Saturn I, Saturn IB test flights, and skipping dynamic ground test modules. Details below.
NASA put the Saturn V into human-carrying service despite unresolved and potentially lethal issues in unmanned testing, for political reasons.
The N1 had vastly less testing than the Saturn V in every way.
1) Only one of every three N1 first stage engines were factory-tested before they were launched. Every Saturn V engine, F-1 to J-2 to RCS cluster, was test-fired before it was installed in a stage. Several N-1 flights were destroyed by engine or plumbing problems that would've been identified in individual engine testing.
2) None of the N1 stages were test-fired on the ground, either in the R&D phase of the N1 program or in pre-flight stage check-outs. In addition to dynamic test modules that the N-1 lacked, every Saturn V stage was test-fired (see figure 231) before it went to the launch pad.
3) The N-1 never had test flights of individual stages on supporting rockets. The Saturn V in the Apollo configuration had numerous test flights of individual stages including:
a) 4 Little Joe II flights (AS-001 to AS-002) to evaluate Apollo escape system
b) 5 Saturn I flights testing the Apollo CSM stack (flights AS-101 to AS-105); also evaluated the Saturn IV stage
c) 6 Saturn Ib flights testing the Apollo CSM, Apollo LEM, and first manned Apollo flights (AS-201 to AS-203; Apollo 1, Apollo 5; Apollo 7). Apollo 1, of course, killed three astronauts on the pad.
d) 5 Saturn V test flights, including the unmanned Apollo 4 and manned Apollo 6, 8, 9, and 10.
The Russian approach to the N-1 was, "The rocket's on the pad? Hold my vodka and watch this." Due to lack of budget and logistics issues (each first stage had to be disassembled at the factory and shipped in pieces to the launch site), they never performed the basic hardware testing that the Americans did from 1960 to 1967.
For example, the F-1 engine had been in testing for 5 years before the Russians even got the N-1 budget approved. Yet, the first N-1 launch was just before Apollo 9.
As a result, this is how the four N-1 launches played out:
1) Flight 1: though the N-1 had never been tested in whole, the Russians decided to use the N-1 to make a lunar flyby with the Soyuz 7K-L1 "Zond" module - basically skipping tests the US performed with Little Joe II, the Saturn I, Saturn IB, and some Saturn V. At T+25 seconds, vibrations start tearing apart rocket plumbing and started a fire in the engine bay. The flight control system ("Kord") responded by shutting down all engines at T+68. Kord also locked the second and third stages, preventing ground control from separating those and salvaging some upper stage test data. Problem: crappy construction. Solution: launch another rocket, this time with fire extinguishers on each engine.
2) Flight 2: Not 5 months later with no interim testing and an incomplete investigation of launch 1, the Russians tried again. Another Zond capsule would flyby the moon looking for landing sites. After engine start but before launch pad release, engine #8 blew up (debris was found on the pad). Kord methodically began shutting down damaged nearby engines from T+10 to T+12 as the rocket gained altitude, then decided "let's shut all the engines down!" Except Kord failed at that, too, and left #18 burning. The unbalanced thrust tilted the N-1, which then blew up close to the launch pad in one of the biggest non-nuclear explosions in history. The upper stage escape system worked perfectly, though. The root cause: bad welds shed metal into the operating engine. Solution: add filters to the fuel lines and reprogram Kord not to shut down engines in the first 50 seconds of flight.
3) Flight 3: 2 years later with only a dummy payload, the Russians discovered something that the US would've found in wind tunnel testing or individual stage test flights: the complicated aerodynamics of the semi-plug nozzle design of the N-1 first stage created weird eddies and counter-currents that induced an uncontrollable roll. In light of launches 1 and 2, Kord was now banned from shutting down engines in the first 50 seconds of flight to protect the launch pad. So, the third N-1 went through an accelerating roll until the third stage tore off and Kord was finally able to shut down first stage engines. Solution: extensively analyze the first stage aerodynamics, redesign the tail section, and...just kidding. The Russians added powerful roll-control rocket engines to the first stage.
4) Flight 4: 16 months later with a Zond payload, the Rooskies tried again. Because the Russians had never done individual stage ground testing, individual stage test flights, or operated an N-1 longer than 68 seconds, they had little knowledge of problems in the first stage engine plumbing. An attempt to reduce loads during the well-understood Max-Q period by shutting down 6 first stage engines led to the less-understood problem of water hammer blowing apart the fuel lines. The launch escape system worked again, and the upper stages separated from the disintegrating first stage correctly.
The N-1s were flown as all-up unmanned launches similar to Apollo 4. Those N-1 flights were wrecked by issues that the US a) avoided through quality control (e.g., good welding), or b) identified in ground testing, or c) on unmanned test flights that actually flew as planned. Apollo 4 and 7 did identify plumbing problems like pogo oscillations that were still problematic as late as Apollo 13.
By aerospace industry standards, the Saturn V and shuttle had very abbreviated flight testing before they were pushed into service. The crews of Apollo 1 and Challenger died because NASA got complacent. But to say the non-existent testing of the N-1 was "sensible" is not correct since it was even more abbreviated, budget-starved, and rushed than NASA's test plans.