STS-93 flight plan
The primary objective of STS-93 was to deploy AXAF, which was officially named the Chandra X-Ray Observatory in late 1998. Chandra was the last Shuttle payload deployed with the upper stage, but thirteenth IUS mission in the program; IUS deploy missions had a familiar template, with the deploy typically planned within the first half-dozen orbits of the Earth on Flight Day 1.
“There is a pretty rigorous template of going through those [IUS pre-deploy] checkouts and that’s an ambitious template, too — you’ve got to get right on it. It’s kind of like when Shuttle is going through their de-orbit prep, that four-hour schedule or timeline for the crew is very busy and everything really builds on the next item, in order to get the vehicle to ‘go for deorbit.’
It’s the same for deploy, it’s a deploy checklist, we have a product that we build into the deploy checklist and it’s choreographed out, down to the minute. Some of these things are specifically constrained when the Shuttle is flying over a particular ground site where the IUS team is going to be sending some commands to the IUS, well we have to do a checkout over that site to make sure it’s active and in a good configuration.
So it’s a pretty busy time right after we get the vehicle up on orbit. Of course the Shuttle and the flight crew are going through their post-insertion checklist to really convert the vehicle from an ascent launch vehicle to an on orbit vehicle, getting the doors open, getting everything reconfigured and then they jump right into the payload checkout and deploy checklist stuff and…I think it was 6:15 somewhere in there is where deploy typically is.
Credit: NASA.
(Photo Caption: The summary timeline of flight activities for STS-93. The nominal deploy of Chandra and IUS was at the beginning of Orbit 6 on Flight Day 1. Landing was planned for just under five days after liftoff.)
Chandra and IUS took up so much of the vehicle performance for the mission and given Columbia’s weight, it could have been a very short flight.
“Actually, the only objective for this mission really was to deploy this satellite, so it could have been a three day mission if you really get down to it — launch on Day 1, maybe have a backup on Day 2, and then come home on Day 3,” Austin noted. “Well, our medical rules for the crew and the missions long ago had established five days as a minimum mission duration and that’s really to protect for the crew health, so five days was the baseline and so we filled the rest of the mission.”
“So Day 1 was the deploy and it was busy; then after that it became a mid-deck payload mission with the crew and they had probably ten or twelve different we called them secondary payloads that they were processing and doing experiments for the rest of the mission. So it was really a tale of two different missions, the Day 1 payload deploy, then after that it was a smaller internal science mission after that.”
Austin had worked on the Simulation team prior to entering the Flight Director Office; as a part of working out the detailed flight plan, a schedule of training would be developed for the flight.
“Something that I did before I was a flight director is I was a lead training instructor for crew training and then later a Simulation Supervisor in charge of the overall training for the crew and the flight control team in the mission,” he said. “I have that background so I joke with everybody that I always knew all the malfunctions that were going to happen on a flight because I put them in before.”
“So there would be a Lead Simulation Supervisor that would be assigned at about that time, too,” he said, referring to initial selection of other flight leads. “We would work together and by this time in the Shuttle’s history we had a pretty standard template of types of training we would do, the amount of ascent/entry training we do, and it’s just based on the complexity of the in-flight events.”
“This was a Day 1 payload deploy so we’ll exercise that several times, depending on the different possible contingencies that we know that we need to be prepared for, like a one rev wave off or possibly even waving off until the next day, so we’ll have that preliminary training template built out there in that first few months of getting together and then it gets refined a little bit.”
Credit: Philip Sloss for NSF
(Photo Caption: The Shuttle Flight Control Room (also known as the White FCR) in 2011. In addition to flight control itself, in between Shuttle missions individual flight control team members trained for their certifications and teams assigned to specific upcoming flights trained here.)
Austin provided a little history and overview of the training flow for the flight control teams: “Long ago after Challenger we rebuilt the whole training profile [for] the flight control team and the individuals,” he said.
“There would be a whole scheme called generic training and we would use different kinds of mission simulations that were around that represented the type of activities we would be doing like rendezvousing with a payload or deploying something and we would have those simulations scheduled for the flight team. That allows a particular individual, a flight controller, or even a flight director that has a predefined training flow [to get] to a level of being certified as a flight controller or a flight director to be assigned to any flight.”
“And so there’s always this generic training flow that’s going on, maybe two or three sims a day throughout the week and there’s always a team over there doing training and a sim team in the simulator and a pickup crew if you will that’s in the simulator facilitating the mission and the training,” he added. “When you’re doing generic training not everybody is being certified; you can fill the sim so I would go over there and participate in sims after I was certified and that was really just to facilitate the training for everybody else.”
“So that would go on pretty constantly, I guess I would probably go do a sim a week probably. Probably in that final six months or so before flight then the focus for me and my team is really more on the paperwork and the product and the reviews and then our specific training activities, so that’s where it kind of tailors itself down to just focusing on my mission and not just broader staying proficient as a flight director.”
“We trained for all aspects of the mission, mostly the deploy, then Ascent, then Entry,” Collins said in her email. “We spent less time training on experiments, but enough to insure we could operate them quickly and efficiently with no mistakes.”
“[The] Pilot and commander flew “Shuttle Training Aircraft” flights,” she added. NASA owned and operated training jets used by the astronaut office to maintain flying proficiency. The two pilots on the flight crew would use the modified Gulfstream jets at the Shuttle Program’s Continental United States (CONUS) landing sites (Kennedy Space Center, Edwards Air Force Base, and White Sands Missile Range) to practice the final landing approach they would fly in the orbiter for real at the end of the mission.
Credit: NASA.
(Photo Caption: One of NASA’s Shuttle Training Aircraft as seen flying a dive into Shuttle Landing Facility Runway 33 from the roof of the Vehicle Assembly Building. Flight crew pilots (Commander and Pilot) used the modified business jet to simulate the Shuttle’s approach and landing characteristics. The STA engines would have to be put in reverse thrust and the main landing gear deployed in order to realistically simulate the low lift over drag aerodynamics of a Shuttle orbiter.)
“We had extra training to allow practice for returning with the Chandra in the event of an ascent abort,” Collins noted. “Due to the heavy payload, we had a riskier launch: in the event of an engine failure, there were less options for aborting to orbit.”
“Also, a potential emergency landing with the IUS/Chandra combination onboard would have been tricky: the structures engineers were concerned about over-stressing the shuttle landing gear, so they analyzed the numbers closer than normal,” she added.
With respect to crew training on the primary payload, Collins said: “Our crew interacted mostly with the Inertial Upper Stage (IUS) which was the solid rocket motor attached to Chandra. This IUS took Chandra 1/3 of the way to the moon!”
“Most of our payload training was deploying the IUS, both nominal and malfunction training,” she explained. “Only a small portion of our training interacted with the Chandra itself.”
Getting ready, handling delays
Issues completing assembly and checkout of the X-ray telescope and later an in-flight anomaly with a separate IUS delayed the mission. Columbia started its processing flow for STS-93 at the conclusion of its previous mission, STS-90.
The orbiter completed the Neurolab mission, the last one flying a habitable Spacelab module, on May 3, 1998, landing in the afternoon at the Kennedy Space Center’s Shuttle Landing Facility. Columbia was towed into Bay 3 of the Orbiter Processing Facility (OPF) overnight to begin mission deservicing while NASA considered a reflight of the Neurolab mission.
Credit: NASA.
(Photo Caption: Columbia is flared for main landing gear touchdown just prior to the completion of the STS-90 mission on May 3, 1998. Several hours after wheels stop, Kennedy Space Center ground operations towed the orbiter in Orbiter Processing Facility Bay 3 for post-flight deservicing and to begin configuring the vehicle to fly STS-93.)
Assembly and testing of the unique spacecraft was behind schedule and launch readiness was already projected to be no-earlier-than (NET) late in 1998. That coupled with ongoing delays in starting the ISS assembly sequence led NASA and the Shuttle Program to consider a quick reflight of the Neurolab mission similarly to how the Microgravity Science Laborary-1 (MSL-1) / STS-83 was reflown.
In this case Columbia and crew had flown a successful, full-duration Neurolab mission; however, the first Shuttle assembly flight for ISS was likely to slip from September to December, 1998, leaving several months open to fly Columbia an additional time when no other Shuttle flights were scheduled. Under that scenario, Columbia would have reflown Neurolab in August, 1998, and STS-93 would slip into early 1999; as it turned out, NASA decided not to refly Neurolab and STS-93 was delayed to late January, 1999, anyway.
Following standard mission deservicing such as removal of the three Space Shuttle Main Engines (SSME), the Extended Duration Orbiter (EDO) pallet was removed from the orbiter in late May. By October, shipment of the telescope to the launch site wasn’t expected until early January and eventually the launch date was adjusted to March, 1999.
Columbia’s extra weight meant additional measures needed to be taken for the orbiter to fly AXAF and IUS. “I mentioned how heavy the payload was, well because it was so heavy we had to offload a lot of the other consumables that would allow us to fly a longer mission, the cryogenics, the oxygen, the hydrogen,” Austin said, referring to the cryogenic fuels used for the orbiter’s electricity-producing fuel cells.
The STS-93 mission patch. Credit: NASA.
“Everything that allows the mission to go the typical ten, twelve, fourteen days, we had to start skinny-ing that weight back down just so we could get this payload up there and off the ground.”
Columbia was the vehicle that was flying almost all the EDO missions for the program; for STS-93, instead of carrying extra cryogenic tank sets on the EDO pallet in the back of the orbiter payload bay, both the whole EDO system and two of the five standard cryogenic tank sets in the orbiter mid-body were removed from the orbiter to reduce weight.
Columbia typically carried 1000 pounds of ballast or more within boxes in the aft compartment, but during orbiter processing for STS-93 both the ballast and the ballast boxes were removed.
Announcement that the telescope was renamed to the Chandra X-Ray Observatory just before Christmas was accompanied by a delay in shipment to KSC until the end of January and a new targeted launch date of NET April 8, 1999. The telescope was shipped from its TRW final assembly facility in Redondo Beach, California, to KSC in early February; as Chandra was arriving in Florida, NASA decided to move STS-93 to early July behind the first station logistics mission in mid/late May.
All four orbiters were back at KSC in early 1999; with only three OPF bays to work with and Atlantis back from its recent OMDP in Palmdale, Columbia traded places with Atlantis. Columbia moved into the VAB in mid-February until there was a new OPF vacancy and Atlantis moved from storage in the VAB to OPF Bay 3.
On April 15, 1999, Discovery rolled out of OPF Bay 1 for the STS-96 mission in May and Columbia immediately moved into the processing bay for final orbiter preparations. By that time, a failure of the IUS on an April 9th Titan IV launch had suspended integration activities between Chandra and its upper stage unit while the Air Force investigated the failure.
Credits: NASA.
(Photo Caption: A rare dual orbiter appearance of Discovery and Columbia outside occurred on April 15, 1999. Discovery was rolled out of OPF Bay 1 and moved to the VAB transfer aisle to be stacked with the rest of the Shuttle vehicle for the STS-96 mission. Columbia had been stored elsewhere in the VAB when STS-93 was delayed, so it was pre-positioned adjacent to the processing bay as can be seen in the low-res, inset, video frame. Discovery was rolled over to the VAB (left) and Columbia was towed into OPF Bay 1 on her landing gear to complete STS-93 orbiter processing.)
Austin took the delays in stride: “Eventually the launch delays get smaller and smaller, right? I joke with people that we kind of sneak up on launch,” he said.
“The early delays were kind of significant and we would look at that and either know that we’re going to have this gap and we want to keep the training proficiency at a certain level with both the crew and the flight team so we would adjust the remaining training schedule to add another training sim. I don’t remember exactly with the different delays how we reacted but I know just traditionally that’s what you do.”
“We had plenty of training to do,” Collins said. “The extra time allowed us to add training we felt we needed: more malfunction procedures, more STA flights, and more payload training (we had “secondary” payloads: payloads in the mid-deck that we performed on Flight Days 2-5).”
Following the Flight Readiness Review (FRR), the Shuttle Program set the launch date for July 20, 1999. The final countdown and tanking would get underway the evening of the 19th, with liftoff planned about a day prior to the thirtieth anniversary of Apollo 11’s lunar landing.
Lead image credit: NASA.