Chair Force Engineer

Wednesday, October 28, 2009

Mike Griffin, Saturn I, and the Potemkin Rocket

After three years of anticipation, the "Potemkin Rocket" known as Ares I-X has launched. The flight appears to have been nominal. While Ares I-X was a low-fidelity test of a bad rocket design, the test's fundamental flaws should not detract in any way from the Ares I-X program personnel who devoted the last three years of their life to making this test a success. While I strongly believe that Ares I-X should have waited until the 5-segment SRB was available, Ares I-X still taught NASA personnel much about ground handling operations and ocean recovery for the Ares rockets. Perhaps the thrust oscillation questions will look a little bit more clear after the test, in spite of the SRB that was used for the launch being so different from the final SRB.

Ares I-X also serves as a backhanded endorsement of Michael Griffin's approach to Project Constellation. Ares got a foot in the door before Norm Augustine's panel could have the final word. For members of Congress who don't comprehand how much more work beyond Ares I-X is necessary before the real Ares I is ready, the visual of Ares I-X lifting off is evidence that the program is on track. Indeed, mebers of Congress are already spinning the Augustine Report as evidence that Ares is being well-executed, even though the committee largely ignored that question (and largely endorsed the idea of commercial spacecraft for low earth orbit missions, with Ares V Lite for deep space exploration.) Somewhere, Mike Griffin is smiling with glee. Not just because Ares I-X succeeded, but because the political winds growing across the Potomac will keep his Shaft Rocket airborne for the forseeable future.

When I think of Ares I-X, I start to think of the original Block I Saturn I rockets. (This is as flattering to Ares I-X as it is insulting to the Saturn I.) The four Block I flights flew with no fins, a dummy upper stage, a shortened first stage and less powerful first stage engines than the later Block II Saturn I and Saturn IB. It would be possible to look at the original Saturn I's and say they contributed nothing to the final man-rated Saturn IB. But that misses the point entirely. After the successes of the first Saturn I's, it was fairly easy to tweak the H-1 engines and stretch the propellant tanks. Even the redesign of the tail fairing wasn't the biggest challenge in the Saturn I evolution. Block I retired the risk on the first stage; Block II tackled the challenge of the new liquid hydrogen upper stage. Saturn IB went a step further by testing the new upper stage engine that would also take men to the moon. But Ares I-X didn't accomplish any major risk reduction, since the only hardware commonality is the SRB case segment design.

The Ares I-X test doesn't tell us much about the ultimate success of the Ares I program. But because the test did succeed, we know that the riddle and challenge of Ares will carry on.

Friday, October 23, 2009

Controlling the Platform

For once in my life, I'm actually excited about upgrading to the new version of the Microsoft Windows operating system. After the major embarrassment of Windows Vista (which could never shake the somewhat-true perception that it was bloated and slow) comes a refined, streamlined Windows that breathes new life into computers from the 2003-2004 time period. I'll probably take advantage of the upgrade after I stick more RAM in my custom tower PC. I'll even forgive Microsoft for the fact that the new OS carries the confusing name "Windows 7" when it should be "Windows NT 6.1."

At the same time, the Mac fans are using the new Windows rollout to justify the fact that they've had a rock-solid OS for years. The Mac OS X family is based on BSD UNIX and benefits from the fact that few virus-writers want to spend their time attacking systems that only 11% of the population use.

If Mac OS is so great, why doesn't Apple release it for everybody to run on their PC's? It's theoretically possible, since the Mac platform migrated to Intel CPU's back in 2006. But Mac OS for Intel-based systems still requires Apple's firmware on your motherboard in order to run. Apple is primarily in the business of selling hardware, not software (Microsoft's business model is the exact opposite.) In Apple's view, having a good OS helps to sell hardware. (Intel's thinking isn't very different, as every Windows upgrade helps Intel to sell the more powerful hardware that the new Windows requires.)

With the brief exception of the mid-90's Mac clones, Apple has maintained exclusive control of the Macintosh platform. Is that a good business model for the PC industry? The conventional wisdom is that the openness of the Intel-Microsoft architecture has been good for the PC industry as a whole. But it may not be good for the individual PC manufacturers. Case in point is IBM, who pioneered the architecture we're familiar with today. "Big Blue" was beaten to the punch when Compaq introduced the first 386-based PC in 1986, and eventually quit the home PC business when it sold out to Lenovo.

It's debatable if IBM could have maintained its position as a market leader if it had adopted the 386 CPU earlier (which it didn't, because Intel refused to let IBM produce its own 386 chips under license.) But that point marked IBM's slide from being an innovator into being an also-ran. From that point on, it was clear that Intel, not IBM, would dictate the future of personal computing.

One of my favorite PC "what if" scenarios is the Commodore Amiga. When the Amiga platform was launched in 1985, it was easily the most advanced personal computer available, capable of pre-emptive multitasking and stunning color graphics with as little as 256 kilobytes of RAM. It was even priced competitively (nearly half the cost of the inferior Macintosh models of the day.) Yet Amiga never caught on like the PC or Mac. It was partly due to Commodore's reputation for producing low-cost computers aimed at the children's video game market, and partly due to an inept marketing department. But it's debatable whether Amiga clones could have saved Commodore from its 1994 bankruptcy. While Commodore could have made money selling its Amiga OS to owners of Amiga clones, Commodore was always in the business of selling computer hardware. Amiga clone computers would have likely taken away sales from Commodore.

It's been said that the Macintosh platform currently controls 11% of the personal computer market. Different flavors of Windows control nearly 90%, and Linux has around 1%. (I don't know if that figure comes from recent sales numbers, or surveys of individuals to see what computers they use at home.) While that might not sound good for Apple, it's pretty remarkable to think that the Apple hardware has an 11% market share. Just think of how many other PC makers dominate the market--Dell, HP, Gateway and Toshiba, just to name a few. I'd be interested to look at the total sales volume for the PC vendors to see where Apple stacks up. It's clear that the company is doing very well for itself.

All things considered, Apple's exclusive control over the Mac platform is good for the company. And regardless of whether your computer runs Mac OS or Windows or even Linux, it's likely that Intel is laughing all the way to the bank.

Tuesday, October 20, 2009

The World's Largest Stick of Dynamite

Just when it seemed like the history books had been closed on the Challenger disaster, I came across a review of Truth, Lies & O-Rings, an interesting look at the faulty decision-making leading up to launch. (hat tip to Clark Lindsey's Hobbyspace.) The reviewer makes an interesting point about the dangers inherent in ground handling of solid rockets. Many of the inherent disadvantages of SRBs have been long-discussed, such as the inability to shut them down during abort situations. But handling and storing the motors carries all the potential dangers of riding on them. For that reason, SRB stacking operations are classified as “hazardous operations,” and all non-essential personnel are banned from the Vehicle Assembly Building. The procedure is similar for stacking the stages of other solid-fuel launch vehicles. In spite of all the precautions and built-in safety mechanisms, the potential always exists for a catastrophic solid-fuel detonation, as occurred with Brazil’s orbital launch vehicle.

While I tend to think that the risk is overstated (the industry has been dealing with large solid rockets since the 1940’s,) it can never be entirely eliminated. For this reason, Jeff Bell predicted that the SRB would be deleted from the shuttle-derived launch vehicles under development by NASA. Many “space boosters” are dismissive of Jeff Bell, viewing him as a cynic whose arguments aren’t worth the paper they’re written on. I’ll concede that his predictions often come with fatal flaws, but he does make a lot of solid arguments and presents plenty of pertinent facts. In the case of the aforementioned prediction, Jeff Bell’s fatal flaw is assuming that NASA would choose a safe, clean-sheet launcher design over one that protects the shuttle’s entrenched workforce and contractors.

The ground-handling of large solid rockets (and even the individual segments) was an issue that should have been re-examined when Ares I was designed to be "safe, simple and soon." While NASA personnel have done an admirable job in handling the SRB's up to this point, it's sobering to know that just one mistake could cost a lot of lives and pull the plug on the nation's manned space program. The Ares 5-segment SRB will be the world's largest stick of dynamite, and that risk should never be lost on anybody who works in the space business.