believed strongly in the commercialization of space, a policy that he tried with Landsat and foisted on the sts and nas Since nASa wanted the Sts to be primary US launch vehicle and wanted to justify the projected high flight rate it had to capture most of the launch market. Thus it got the Air Force to agree that all future military missiles would fly on the shuttle. The Air Force also agreed to refurbish the old MOL Space Launch Complex at Vandenburg to have a site to launch into polar orbit from military missions. It of course required that all NASa payloads went on the shuttle Thus the Hubble and Galileo were designed to go up on the Shuttle. It enticed the commercial customers in two ways. It offered very attractive prices for the first three years of Shuttle operations. Thus a PAM-d class satellite launch could be had for $15 million whereas to get the same launch in an Ariane was $30 million and $25 million on Delta. It also pulled it's payloads from Delta and Atlas. Since there were now being used less and less but they needed to sustain their infrastructure their launch costs rose Thus delta cost rose from $5 million a launch in 1970 to $26 million a launch by 1980. NASA also terminated the Delta and Atlas production lines in 1985. The Air Force did buy some Titan 34D's and contracted to buy only a few Titan 4s but did so over the objections of NASa and agreed to stop doing this. Thus NASa and the government moved to a one launcher policy driven by the desire for cost effectiveness. By January 1986, the STS had only flown twenty four times and had proven to be neither cheap nor reliable. However, so committed was NASa to the thesis that this was an operational vehicle that after only four test flights they had declared it an operational vehicle and on the 25 flight they were going to fly a teacher into space, an event to be watched by millions of schoolchildren. Instead of quick turnaround what they had found with this "operational " vehicle was that every one of the 17,000 tiles on it needed to be inspected after every flight and every SSme needed to be replaced every time. They had also noticed some worrisome erosion in the solid rocket joints where the segments ere put together. Thus each Shuttle, instead of a turnaround of days, took months to prepare and required a large standing army of people to maintain it at human flight safety levels(0.99999) How could the 1977 estimates have been so wrong? In retrospect, there were a number of factors. There was a deliberate naSa strategy of getting support for large programs with optimistic operational estimates and low cost estimates. This is the well known Camel's nose under the test strategy which basically relies on getting things going and building supporters who would sustain the program as the costs mounted This strategy w be very clear on Station. In addition, the designers were overly optimistic about the technical process of NASA. Perhaps they were still living in the glory days of Apollo. In any case they clearly underestimated the Ssme difficulty. Still they seemed to have taken leave of commor sense. The SSME is operated at 109% of total rated thrust. This is at the red line. Any mechanic will tell you that an engine routinely operated at the "red line"will break down frequently. Truck engines(the model for the STS)work so reliably because they operate far from the maximum capabilities of the engine. The STS was certified as operational after only 4 flights with the really flight critical part the ascent, being only 8 minutes each. Thus it was certified after 32 minutes of critical flight. In contrast the F-22 is required to be tested for a minimum of 183 hours of flight time before Congress authorizes buying the aircraft. Finally, the historical probability, based or many launchs, of solid rocket failure has been 1 out of 25. How the NASA engineers managed to convince themselves that the catastrophic failure rate would be 1 in 10,000 when the sts hadbelieved strongly in the commercialization of space, a policy that he tried with Landsat and foisted on the STS and NASA. Since NASA wanted the STS to be primary US launch vehicle and wanted to justify the projected high flight rate it had to capture most of the launch market. Thus it got the Air Force to agree that all future military missiles would fly on the shuttle. The Air Force also agreed to refurbish the old MOL Space Launch Complex at Vandenburg to have a site to launch into polar orbit from military missions. It of course required that all NASA payloads went on the shuttle. Thus the Hubble and Galileo were designed to go up on the Shuttle. It enticed the commercial customers in two ways. It offered very attractive prices for the first three years of Shuttle operations. Thus a PAM-D class satellite launch could be had for $15 million whereas to get the same launch in an Ariane was $30 million and $25 million on Delta. It also pulled it’s payloads from Delta and Atlas. Since there were now being used less and less but they needed to sustain their infrastructure, their launch costs rose. Thus Delta cost rose from $5 million a launch in 1970 to $26 million a launch by 1980. NASA also terminated the Delta and Atlas production lines in 1985. The Air Force did buy some Titan 34D’s and contracted to buy only a few Titan 4’s but did so over the objections of NASA and agreed to stop doing this. Thus NASA and the government moved to a one launcher policy driven by the desire for cost effectiveness. By January 1986, the STS had only flown twenty four times and had proven to be neither cheap nor reliable. However, so committed was NASA to the thesis that this was an operational vehicle that after only four test flights they had declared it an operational vehicle and on the 25th flight they were going to fly a teacher into space, an event to be watched by millions of schoolchildren. Instead of quick turnaround what they had found with this “operational” vehicle was that every one of the 17,000 tiles on it needed to be inspected after every flight and every SSME needed to be replaced every time. They had also noticed some worrisome erosion in the solid rocket joints where the segments were put together. Thus each Shuttle, instead of a turnaround of days, took months to prepare and required a large standing army of people to maintain it at human flight safety levels (0.99999). How could the 1977 estimates have been so wrong? In retrospect, there were a number of factors. There was a deliberate NASA strategy of getting support for large programs with optimistic operational estimates and low cost estimates. This is the well known Camel’s nose under the test strategy which basically relies on getting things going and building supporters who would sustain the program as the costs mounted. This strategy would be very clear on Station. In addition, the designers were overly optimistic about the technical process of NASA. Perhaps they were still living in the glory days of Apollo. In any case they clearly underestimated the SSME difficulty. Still they seemed to have taken leave of common sense. The SSME is operated at 109% of total rated thrust. This is at the “red line”. Any mechanic will tell you that an engine routinely operated at the “red line” will break down frequently. Truck engines (the model for the STS) work so reliably because they operate far from the maximum capabilities of the engine. The STS was certified as operational after only 4 flights with the really flight critical part the ascent, being only 8 minutes each. Thus it was certified after 32 minutes of critical flight. In contrast the F-22 is required to be tested for a minimum of 183 hours of flight time before Congress authorizes buying the aircraft. Finally, the historical probability, based on many launchs, of solid rocket failure has been 1 out of 25. How the NASA engineers managed to convince themselves that the catastrophic failure rate would be 1 in 10,000 when the STS had 2