# NASA Gemini 9 Tragedy - How NASA Lost Its First Crew

## Метаданные

- **Канал:** Scott Manley
- **YouTube:** https://www.youtube.com/watch?v=FxphOosK08I
- **Дата:** 07.05.2026
- **Длительность:** 26:28
- **Просмотры:** 180,486

## Описание

In early 1966 the primary crew of Gemini 9 died in a plane crash when they tried to land a T-38 by circling to a runway under low clouds and poor visibility. Elliot See and Charles Basset were killed instantly when the aircraft collided with a building on the ground, McDonnell’s  building 101, which by coincidence was the place they were supposed to be visiting for training. 
The backup crew of Tom Stafford and Gene Cernan would take their place and both would fly to the moon during the Apollo program.
I recently re-read this story and with my recent training in instrument flying all the information now made sense, so I wanted to tell a story about the crash, and explain instrument approaches to non-pilots.

A lot of the best information comes from the book "Fallen Astronauts: Heroes Who Died Reaching For The Moon" by Colin Burgess
https://amzn.to/49fH5e0


Follow me on Twitter for more updates:
https://twitter.com/DJSnM

I have a discord server where I regularly turn up:
https://discord.gg/zStmKbM

If you really like what I do you can support me directly through Patreon
https://www.patreon.com/scottmanley

## Содержание

### [0:00](https://www.youtube.com/watch?v=FxphOosK08I) Segment 1 (00:00 - 05:00)

Hello, it's Scott Manley here. If you ask the average space fan what was the first accident in NASA's history that killed a crew, then Apollo 1 is the answer that most people will think of. But a year before NASA lost another crew to an accident during training. Elliot See and Charles Bassett were the primary crew assigned to Gemini 9. And on February 28th, 1966, they would be killed when the T-38 jet they were flying crashed while trying to land. And their deaths meant that the backup crew, Thomas Stafford and Gene Cernan, would be promoted to the primary crew. And the new backup crew was James Lovell and Buzz Aldrin. Now, the assignment of a new backup crew would have a large impact on the astronauts who would ultimately fly to the moon on the Apollo missions. Buzz Aldrin almost certainly would not have been on Apollo 11 if he hadn't flown on Gemini. So, this crash that most people don't remember actually set the stage for the Apollo crews that we do remember. But I want to look at this plane crash because while I read this story many years ago, I recently encountered it again. And as with my recent flight training, many parts of the story now take on new meanings that I think that I kind of want to start talking about and share. Specifically, if you read accounts of the crash, you'll see that the crash happened during a circling approach in low IFR conditions, which 5 years ago would just be words to me. But these days it triggers a whole bunch of words and thoughts, some of which are very much red flags. So, let me explain how two astronauts ended up in this deadly situation. And why a test pilot with thousands of hours of flight time wasn't able to handle landing an aircraft that is used to train new pilots for even faster and more complex aircraft. Northrop's T-38 Talon was first flown in 1959 and development completed in 1961. It was the first supersonic trainer aircraft. It was powered by a pair of General Electric J85-5A turbojet engines, which combined generated about 1. 9 tons of thrust, or 2. 6 tons with afterburners. It'll cruise at just over Mach 1, and thanks to the aircraft's lightweight, no-frills design, it was able to set a number of time to climb records in 1962. Records that it held for about a month before the F-4 Phantom came and took them away. When it was introduced, it was considered a lot easier to fly than the combat jets of the era. But as the aircraft with more modern flight control systems have been introduced, the situation has been reversed, and many people believe the F-35 is easier than the T-38. So, over a thousand of these were produced from 1961 to 1972. And incidentally, by the way, they were manufactured in the Northrop factory on the south side of Hawthorne Airport. The factory is now owned by SpaceX. So, that site was building vehicles flown by astronauts long before SpaceX moved in. NASA astronauts have access to a pool of T-38 twin-engine jet trainers. These were originally selected for NASA in 1964, and NASA astronauts will still fly them today. If you look at when the crew for Artemis 2 arrived for their final press conference on the ground, they arrived in the T-38s. NASA also used them as chase planes for the space shuttle and other aircraft development programs. You can see them in early space shuttle landings escorting the shuttle down and making sure the gear deploys. But, most of NASA's T-38 hours are clocked up by astronauts traveling. With multiple NASA centers spread around the country, NASA astronauts need a way to get across the country quickly, and many of them are also jet pilots. So, the T-38s also provide an opportunity for these pilots to continue to maintain the skills that they need in high-performance jets. Like the other astronauts in his group, Elliot See was a test pilot, but unlike most of the other pilots, he wasn't active military. In fact, the only other civilian in his group was Neil Armstrong. Neil Armstrong and See had actually been assigned together as a backup crew on Gemini 5. So, when that spacecraft had a problem with its fuel cell, See had taken lead in developing a fix and communicating it to the crew. Without his work, the planned long-duration mission would have been cut short. He'd saved the mission. Normally, the backup crew would then have been promoted together to primary crew status on Gemini 8, but instead, they'd been split up with Armstrong flying with Dave Scott on Gemini 8 and See becoming commander of Gemini 9 with Charles Bassett. Both missions had complicated EVA plans, and Deke Slayton thought that See wasn't physical enough for this task and would be better suited to the commander role. In his memoir, Deke Slayton wrote that he thought that Bassett would be strong enough to carry both of them. Gemini 9's plans involved a rendezvous

### [5:00](https://www.youtube.com/watch?v=FxphOosK08I&t=300s) Segment 2 (05:00 - 10:00)

and docking with the Agena target vehicle, followed by an EVA to test the Air Force's astronaut maneuvering unit. This was a jetpack that would let the astronaut fly free. It was more ambitious than the MMU that was demonstrated on the space shuttle decades later. The shuttle version used compressed nitrogen cold gas thrusters, but the Gemini era design used hydrogen peroxide monopropellant. The exhaust was a lot hotter, and the spacesuit for the mission added a protective chromalloy fabric layer to the legs to protect the astronaut. These astronauts literally wore stainless steel pants. So, in early 1966, they're training, and that takes them on visits to various NASA centers and aerospace companies who are involved in the mission. For the fateful flight, they were supposed to visit McDonnel, who were building the Gemini spacecraft in their facility next to Lambert Field in St. Louis. This would give them an opportunity to check out the spacecraft, which was nearing completion, and it would also allow the crews to train on the Gemini simulators for the mission ahead. They'd visited this facility about seven times in the last six months. The plan would be to fly from Ellington Air Force Base, which is next to the Johnson Space Center in Texas, making this flight of just under 700 mi. It's a relatively short hop for something as fast as the T-38. They would leave around 7:30 a. m. and they expected to be on the ground before 9:00 a. m. Both the primary and the backup crew would make the flight, with the primary crew of C and Bassett in the lead jet, and the backup crew of Stafford and Cernan flying in formation. But before they took off, they'd phoned ahead to get an idea of the weather situation, and it was bad news. There were clouds at about 800 ft with poor visibility below those clouds due to rain, snow, and fog. This would require an instrument approach, something of course that they trained for and flown in many, many times. The T-38 as a training aircraft included nav radios that could be operated with things like ILS, VOR, and TACAN, which is the military radio navigation system. With flying on instruments, the idea is that you use radio navigation to guide yourself to the destination and keep your aircraft away from the terrain. Nowadays, we have GPS everywhere, but in the 1960s, you'd be using radio beacons. And for the 700-mi trip from Houston to St. Louis, they'd be flying a route from one beacon to another, along what we call airways, imaginary lines in the sky that connect these beacons. You might have ATC give you specific altitudes or change headings to avoid traffic or weather, but generally you're following the beacons and reporting in with ATC. For the T-38, it would typically fly at altitudes far above the clouds. And while you would be in beautiful clear weather, you wouldn't be able to see the ground where you are. Once you get close to the destination, then you'll descend out of the on-route environment and into the approach onto the airfield. In the approach environment, you'll be closer to terrain and things will need to happen much more quickly. You'll need a plan, a way to use the navigational tools you have to descend through the clouds and emerge with the airfield in sight in front of you so you can then proceed to landing using your old good old-fashioned eyes as guidance. And so we have what are called approaches, which are standard ways of doing this. And these are being designed by specialists at the FAA who follow the terminal instrument procedures rules or TERPS. For a given airport, there'll be different approaches to each runway and there might be different approaches using different technologies. For example, there might be an approach that uses a VOR beacon at the field, another one that uses ILS next to the runway. And these days you might have GPS. That now those are actually found on most runways these days. All of these will be different, but they give aircraft with differing navigation systems a way to navigate to the runway safely. And to make it easier for pilots who have to fly this, all the information needed for an approach is condensed into a single page, a cheat sheet or in official terminology, an approach plate. The approach plate is the quick reference guide for how you will get through the last 20 miles or so and onto the ground safely. And so for this flight to St. Louis, Elliot had figured out the best way to get down through the clouds was to approach via runway 12 right. That had a new instrument landing system or ILS and would have them approach from the northwest, which was more desirable than the approach to runway 24. Runways are numbered according to their compass headings. So the landing approach direction on runway 12 was 120° to the southeast, whereas Runway 24, which had a an ILS, too, that was less desirable because they would actually have to fly over the field because they were coming from Houston, and then they would have to turn back and land going to the southwest. And so, this would take a whole lot longer. Runways are usually bidirectional, so Runway 24 is also Runway 06. And

### [10:00](https://www.youtube.com/watch?v=FxphOosK08I&t=600s) Segment 3 (10:00 - 15:00)

where you have parallel runways, you'll also have runways that can be like right, left, or center designation. And at this point, I should highlight the layout of the field because in 1966, there are runways running three directions. There's 24 and 06, 35 and 17, and 12 and 30. And that last direction has two parallel runways, but the smaller runway to the north. The jets want the wide, fast Runway 12 right, while the plebs in the piston engines can play on the northern Runway 12 left. The McDonnell factory was on the north side of the field near the approach end of Runway 24. This is a USGS image of the field from 1968 and showing the layout. These days, the airport has changed quite a bit to accommodate more airliners. That small parallel runway got widened and extended to accommodate jets. Runway 35 or 17 was removed, and a new runway was added to the northwest. Now, this is parallel to the existing Runways 12 left and 12 right. And when it was opened, they could have renamed 12 right to 12 center and made the new Runway 12 right. But to avoid confusing pilots and changing all the existing documentation, it's much easier and safer to call it Runway 11, even although it is actually parallel. So, the runway bearings don't always exactly correspond to the naming. I hope you're still interested in this story but after my excursion into runway naming, by the way. And this is before I even got into the fact that these are named based on magnetic north. And so, the bearings actually drift over time leading to the runway numbers having to be changed. And if you think this is a bunch of interesting but not 100% relevant information, then this makes it a perfect metaphor for calling up the aviation weather briefing phone line and getting an info dump of no times that probably won't matter to you. Anyway, Elliot did his homework. He came up with a flight plan and he filed it and there was one crucial thing that was missing for this plan. He didn't have the official approach plate for the ILS on runway 12 right. The essential quick reference or the cheat sheet for this landing. Instead, he took the information from the phone call and with Gene Cernan, he wrote up their own version of the plate with the frequencies, navigation aids, and the sequence all handwritten and copied to share with Stafford on the other aircraft in case they get split up. So, having flown hundreds of miles, the two aircraft are approaching the field and they start getting radar vectors from ATC. Essentially, headings and altitudes that they can fly so they can intercept the approach in towards the runway. As they descend through 12,000 ft, they get solidly into the clouds and they stay there until they land. The planes had to remain in close formation to keep sight of each other in formation. And apparently, the controllers were not aware that this was a flight of two aircraft in formation which led to some confusion later on. And staying in formation also meant they got dubious vectors to make their final intercept and that meant that they actually overshot their initial intercept of the approach and you had to come back from the north and get lined up and this delayed them getting set up for their final approach and landing. So anyway, the way the ILS works is that you tune the radio to the correct frequency and the radio signal contains two different tones. There's a 90 Hz tone on one side and a 150 Hz tone on the other side. And if you're lined up exactly in the middle, these will be equally strong. But if you're off-center, the navigation instrument can detect which side you're on depending upon which signal is stronger. Now, a runway might only provide horizontal guidance, and these days we would call this a localizer approach. The horizontal signal will use frequencies from 108 MHz to 112 MHz. But, to get a full instrument landing system, you also need vertical guidance, and that's known as the glide slope. There'll be a second set of antennas that operate at about 330 MHz to 335. The frequencies for the horizontal and vertical guidance are paired up, so you only tune the low-frequency one, and the other one is comes through automatically. So, anyway, they were on the center line in the clouds 2,400 ft moving about 220 kn, and ATC told them to switch over to the St. Louis tower frequency. And now, the question was when to start descending. For this approach, there was an outer marker beacon, a small installation that transmits a single a simple signal straight up, and this gets picked up as you pass over it, and it tells you you're at the point where you're supposed to descend. Now, Tom Stafford, he noted the passing of the outer marker. He heard it. And he heard C order the flaps down and the speed reduction to 200 kn, but C in the lead plane did not initiate a descent that was needed to reach the runway. A few moments later, Stafford radioed the lead aircraft saying, "Hey, the glide slope is out. You need to descend. " And moments later, C began to descend towards the runway. And so, here's the thing. There was no glide slope on that approach in February of 1966. I looked at the Federal Register, which contains

### [15:00](https://www.youtube.com/watch?v=FxphOosK08I&t=900s) Segment 4 (15:00 - 20:00)

this kind of information, and explicitly states there is no glide slope onto runway 12 right. As I understand, at the time the controllers didn't actually make a distinction between an ILS with only horizontal guidance and those which also had a glide slope. Now, now we'd call the approach without the glide slope a localizer to distinguish. And in general, an approach with a vertical guidance like a glide slope is called a precision approach. If you don't have the glide slope defined by navigational equipment, it's a non-precision approach. Both types of approach will give you a set of altitude restrictions for a series of points along the approach. Usually, a requirement to stay above a certain altitude before you reach a certain point. But when you hit the final section of a precision approach, you will then follow a smooth slope down until you reach the runway or give up. While on a non-precision approach, you descend down to a minimum descent altitude and then level off. And then you keep flying until you either see the runway or you reach the missed approach point and then give up and go missed. So anyway, in this case, without the vertical guidance, the rules allowed them to descend to 400 ft above the surface. So anyway, either C had been looking for a glide slope that wasn't there or he just missed passing off the outer marker, possibly because he wasn't reading well. We can't know exactly what happened, what led to this being missed. But the result was that while they began to descend and eventually emerged from the clouds, at that point they delayed too long and they were too close to the runway and too high above it to make a straight-in landing. And faced with the prospect of going missed and trying again, C decided that instead he was going to stay below the 800 ft cloud layer and circle to another runway to complete the landing. While this is technically legal, it's where things do start to go wrong. So C flying along runway 12 right, he began turning the aircraft to the left and Stafford of course followed at a safe distance. But he would lose sight of the lead plane and instead Stafford elected to head back up into the clouds and declare a missed approach. As C was maneuvering his aircraft at low speeds close to the ground and trying to keep sight of the runways out the left side of the aircraft while the weather included snow, rain, fog, and low clouds. And because the T-38 is a fast aircraft, he would have had a significant bank to turn it quickly enough while staying close enough to the field so that he could see the runway. I mean, so he could clearly see that runway 24 was there and it was better positioned. And he radioed to the ground that he was going to plan on landing on runway 24. He just needed to turn left a bit more and get lined up. And somewhere in there, he lost situational awareness. Or perhaps he lost too much airspeed. It's known that a few seconds before the crash, he reduced the bank angle, pitched up, and applied full power, spooling up the engines and lighting the afterburners. But only one of the engines would reach full power when the aircraft's left wing hit the roof of McDonnell's building 101. The wing cut up in the roof. Landing gear was stripped off, fuel tanks ruptured, and created a fireball. Somehow the ejection seat pyrotechnics fired on their own, but not enough to save the crew. They were thrown clear and received lethal injuries. The aircraft cartwheeled off the roof and landed in a parking lot just in front of McDonnell's Space Center. This was the building that they were supposed to visit. It was the place where the Gemini spacecraft was being built. If the aircraft had hit in a different way and caused more damage, it's possible the Gemini program would have lost spacecraft or the ability to build more hardware needed for the last few missions. As it was, only a small number of people in the building were injured. They were working on an assembly line for the Phantom aircraft. Also, the building was large enough that people working far away from the impact site didn't even know that an aircraft had hit the building. Elliott See had over 3,000 hours as a pilot. He'd probably made hundreds or even thousands of landings like this. Flying around an airfield in the pattern making turns from base to final and then landing is the basic way they teach landings from day one. How did he get this wrong? Well, I've learned that circling to a runway at low altitudes can be a trap. So much of what you do when trying to line up for a landing is looking out the window and trying to judge where you are in relation to the runway. So that when you make your turns, you end up lined up. But normally a jet aircraft like this would be circling at about 1,500 ft. And one of the ways that we judge the distance to an object on the ground while flying is by looking at the angle between that object and the horizon. This angle gets larger as the viewer gets higher. So when flying low you and staying trying to stay below the clouds, the runways can seem further away than they are. And so you end up moving closer in. And as you realize this, you realize it's moving a bit faster, you need to bank a little more to turn a bit harder to not overshoot

### [20:00](https://www.youtube.com/watch?v=FxphOosK08I&t=1200s) Segment 5 (20:00 - 25:00)

the runway. And then the T-38, this makes it worse. If you look at it, it's a skinny little thing. It has tiny wings. It is designed to fly fast and that makes it bad at flying slow. One of the things that's outlined on an approach plate is the minimum altitudes and visibility required for an approach. Faster planes generally require higher minimums to account for the shorter time that they have to make decisions and the wider turns that they will have to make. Aircraft are split up by categories based upon their stall speed. Like so your classic Cessna 172 is a category A. This Cirrus I fly is faster. It's a category B. Cat- category C covers small airliners and category D is for the larger faster airliners and maybe some of the slower military aircraft. The T-38 is a category E and this is a very exclusive club. It's basically the high performance military jets that get this designation. The T-38 needs to stay fast or fall out the sky. And when it's flying at approach speeds, it's already in a regime where the nose is pitched up and the aircraft is on the edge of the stall buffet with the air not running smoothly over the wings. And so you can see how this turn at low altitudes would be closer to the runway than expected and the aircraft is already close to the limits of stalling running with a small fraction of the lift available at higher speeds while trying to make a tight turn and so it might simply got too slow. It was losing altitude too quickly or it might have stalled as the angle of attack increased to try to make the turn. Either way, lighting the afterburners, rolling level, regaining speed and climbing would be a natural solution but the by the time this was happening he'd already run out of time. Now Stafford and Cernan in the other aircraft had climbed back into the clouds but initially there was some confusion. Again, the tower again didn't know that there were two NASA aircraft doing different things but another important thing that they needed had been left off the approach plate they'd made. There were no missed approach instructions. These are the instructions that tell you what to do if you can't land and it usually requires you climbing out in a straight line possibly with a turn to avoid terrain and then you head off to a place on the map where you can hold and figure out what to do next. Now eventually they did get helped by the controllers and they made another approach this time onto runway 24 which had a proper instrument landing system with a glide slope. It was only after they landed that they found out what had happened to the other aircraft. NASA would begin its own internal investigation led by Alan Shepard and staffed by a number of other experts. While this investigation would take weeks, the Gemini 9 plans continued with Stafford and Cernan now the primary crew. Just 2 days after the crash, the spacecraft was ready for shipping to the Cape in preparation for a launch on May 17th. However, the failure of an Atlas carrying the Agena target vehicle led to a postponement until a replacement could be launched. And on June 3rd, the mission launched with a new designation, Gemini 9A. When they rendezvoused with the Agena, they discovered that the payload fairing was still attached, so the docking had to be canceled. Then, Gene Cernan's spacewalk to the AMU was an unmitigated disaster. He overexerted himself, the suit overheated, his visor fogged up, and NASA learned a lot about the difficulties of operating on EVAs, and they realized more training was needed. Fortunately, a team at Langley had been experimenting with EVA training in a specially designed swimming pool, and soon the new neutral buoyancy simulator at Marshall Space Flight Center would be an essential facility to the space program. The AMU was planned to fly again on Gemini 12, but was pulled before the flight, and it never flew in space again. The report on the crash was completed and the findings announced on May 27th. While the report has never been published, we know a lot of the important things, like the general sequence of events and the runways used. It was published in our news about the Space News Roundup, which was a like an internal newspaper at NASA's Johnson Space Center. The June 10th, 1966 issue contained a lot of detail on the accident, and it's also in the same issue where they have the first post-flight report from Gemini 9A. Another source with much more detail is Colin Burgess' book Fallen Astronauts: Heroes Who Did Not Die Reaching for the Moon. He acquired a copy of the report from Bassett's widow, and he shared it with a few experienced pilots who offered their opinions, and that helped me understand what went wrong on the approach. Deke Slayton was later critical on his Cernan's piloting abilities in his memoir. He described Cernan's flying as old womanish. But Neil Armstrong didn't see anything problem anything wrong with uh Cernan's flying, and personally, I'm going to say my check ride was with a 65-year-old woman who had a bunch of different ratings and also went wing walking on biplanes. So, I have a

### [25:00](https://www.youtube.com/watch?v=FxphOosK08I&t=1500s) Segment 6 (25:00 - 26:00)

different perception of how an old woman flies. But, the report made it clear that there was more to this than one pilot flying beyond his skills. The handwritten approach plate, which all four pilots agreed on, was missing key information. The ATC weren't aware that this was a two-ship formation and they gave instructions that make it hard for the aircraft to get established on the approach, leaving them behind the planes. Uh the one thing that you can place squarely on C's shoulders is the poorly thought out decision to press on and attempt to circle to land with low clouds and poor visibility. Some people believe that he wanted to get on the ground before the backup crew so he could show the backup crew who's boss. Years later, Gene Cernan would say that if it wasn't for the crash, Charles Bassett would have walked on the moon. And again, we probably wouldn't have seen Buzz Aldrin on Apollo 11. The history that we know in this case was brought to us in part by questionable decision-making on the part of one pilot. And that sticks in my mind every time I'm asked to fly a circling approach with low minimums. I'm Scott Manley. Fly safe. —

---
*Источник: https://ekstraktznaniy.ru/video/51372*