"Why don't you just fix your little problem and light this candle?"
Words as significant now as they were 44 years ago. Al Shepard said these words, tired after more than 3 hours of holding inside his tiny Mercury capsule on the pad at
Today I feel like saying the same thing. Let's light this candle. Some of you already know this, but for those of you who don't: I'm a huge space buff. Space.com, NASA Watch, and Spaceflight Now are among my daily reads. I've watched Apollo 13 so many times I know not only the dialogue but the way the actors deliver the lines. HBO's miniseries "From the Earth to the Moon", hosted and produced by Tom Hanks, is highly cherished. I've read more books on the subject (Failure is Not an Option, by Apollo Flight Director Gene Krantz, remains my favorite,) than I can remember. To this day I wish Comcast would stop wasting bandwidth on stupid channels and just carry NASA TV. (They broadcast it, in digital, for free!! All you have to do it point a dish at the satellite!)
I've been waiting for the Return to Flight launch for months now. The current delays make me look back to the first Mercury flight and laugh. I could see the old curmudgeonly Al Shepard in the seat of the shuttle saying "You know what, I don't give a shit about some damned sensor, let's light this candle."
For those of you who don't understand exactly what is kept NASA management from letting Discovery fly on July 13th, I'll explain it to you. The big orange thing attached to the shuttle at the pad, that's the fuel tank, External Tank to be exact. During launch the main engines on the shuttle need so much fuel that it has to carry an ET with separate sections for the liquid hydrogen and liquid oxygen. The larger of the two sections inside that tank is the liquid hydrogen tank, and at the very bottom of that tank is a sensor called the Engine Cutoff Sensor. It's the one that is causing the problem...during the countdown on July 13th, readings intermittently read dry when it is obviously supposed to be reading wet while the tank is fueled. The launch was scrubbed because, according to post-Challenger rules, this sensor and 3 others exactly like it must read correctly. Yep, that's right, this sensor is one of 4 in the tank that senses how much fuel is still in tank.
So let's imagine, for a moment, if you had a big bottle of Coke (or Pepsi, or RC, or Barq's. Whatever you prefer.) The bottle is completely opaque except for four windows, equally spaced down the side of the bottle, so that the status of the bottle can be quickly determined just by looking at it (Full, 3/4, 1/2, 1/4.) You don't know the exact level, because looking through the window you can only see when the soda is at that particular level. Unlike a car gas tank, which is usually a strip-type or a float-type, that can show you the level of the gas on a gauge that is infinitely adjustable anywhere from full or empty, you only have a 'wet' or 'dry' status on 4 windows. Now imagine a 'soda' bottle 738,418 times the size of your regular 2 liter. That's
In my mind, it's like this. The shuttle burns those 390-some-odd-thousand gallons of hydrogen in less than 10 minutes. In fact, we already know it burns at 45,283 gallons per minute, (the math doesn't work out correctly because there's fuel stashed aboard the Discovery too, albeit it in much smaller amounts.) We can tell, by timing the time between the first sensor running dry, the second sensor running dry, and the third sensor running dry how long it would take before the balky fourth sensor ran dry and ergo calculate the rate at which the tank was running empty. Imagine that bottle. Cover up the bottom window, and introduce a leak in the bottle. The flow of the soda out of the bottle is going to happen at a pretty steady rate (well actually the rate reduces slightly over time as the weight of soda is reduced, thereby reducing the pressure on the point where the soda is leaking, but I'm confident NASA engineers could graph that, even with old-school slide-rules.) Even an elementary school student could guess how long it would take for the level to get from the third to the fourth window by timing the intervals between the first and second, then the second to the third. It's easy math. In reality, there is a system for monitoring the actual fuel consumption on the shuttle anyway, any abnormality would be instantly seen.
And let's not forget one thing. If the shuttle does use enough fuel to empty the tank, and make a fully functioning fourth sensor read dry, we're already screwed. An abnormal flow condition is an abort situation on its own, and they could tell by the flow rate whether or not they'd be able to abort-to-orbit, or abort to a landing using a variety of methods. Having a faulty sensor wouldn't affect the flow rate at all, in fact it's a completely redundant thing in my mind. They'd know long before that sensor went dry if they were out of gas.
So, why don't we light this candle?
Good luck to the crew of the Discovery.