Riding Rockets - Mike Mullane [61]
But Young and Crippen would be taking an enormous risk and I feared for their lives. The only thing that had been positively demonstrated about the shuttle design was that it would glide from 25,000 feet to a landing. Four flight tests off the back of a 747 carrier aircraft had proven that. The solid rocket boosters and SSMEs had been ground-tested multiple times but had never actually flown in space. In fact, the SRBs had never even been tested vertically. In each of their firings, the rocket had been in a horizontal position, a fact that made many of us doubt the tests were really duplicating the stresses and strains of a vertical launch. The massive gas tank had never experienced the shake, rattle, and roll of a launch. There had been no full-scale flight tests of the 24,000-heat-tile mosaic that was glued toColumbia ’s belly. How would it do in the 17,000-mile-per-hour, 3,000-degree wind of reentry? And no spacecraft had ever glided 12,000 miles to a “one-chance” landing—but that’s exactly whatColumbia was going to have to do. And the unknowns weren’t just in the STS hardware. The shuttle’s computer system contained hundreds of thousands of lines of code. Billions of dollars and years of labor had been spent to validate that software but there were still thousands of permutations that had not been tested and that could contain fatal flaws. Would an engine failure at precisely T+1:13 in conjunction with an unexpected wind sheer at 65,000-feet altitude cause a software switch in some black box to flip to a different polarity and sendColumbia out of control? To an extent never before seen in spaceflight, the space shuttle was certified to carry astronauts based upon the wizardry of computer modeling. For a decade, engineers conducted thousands of ground tests in every imaginable engineering specialty: aeronautical, electrical, chemical, mechanical, hypersonic flight dynamics, cryogenic fluid dynamics, propulsion, flutter dynamics, aeroelasticity, and a hundred others. They digitized data gleaned from wind tunnel tests, engine tests, hydraulic tests, heat tile tests, and flight control tests and dumped the results into computers humming with the equations of Max Planck, Bernoulli, and Fourier. When the thousands of answers were finally assembled and examined, the engineers cheered. Computer models said their new shuttle system would work, that its twin SRBs and three SSMEs burning 4 million pounds of propellant in81 /2minutes would propel a quarter-million-pound winged orbiter to a speed of nearly 5 miles per second and an altitude of 200 miles. These same models also assured their brainy authors the orbiter would be able to make a powerless hemispheric-long glide to a 15,000-foot-long strip of runway. Of course, many of these same engineers had done the same thing in the development of the Redstone, Atlas, Titan, and Saturn rockets of the past manned programs, but in those programs, after all their testing and modeling were complete and the answer was “This rocket will fly,” they had still walked cautiously. “We could be wrong in this model, or maybe in this model, or in this one,” they had said. “We better test this puppy unmanned a couple times before we strap astronauts to it. And when we do, we better give the crew a way of surviving a booster failure throughall of