Wright doesn’t post to discuss, he posts only to confront and cajole. Evidently it’s how he gets his jollies.

Wright wrote:
“No, they are to enable the pilot to take corrective action. That’s what pilots DO.”

I should have clarified a bit more. Some of the warning lights in military aircraft enable the pilot to take no corrective action. They only provide an indication of how crippled his aircraft is. In other words, the aircraft is equipped with sensors and indicators to enable him to decide when he must eject.

Unfortunately, I detect too much confrontational attitude (and just plain anger) in many of Mr. Wright’s posts. So I do not see any point in engaging him further. Life is too short to argue with people who take their frustrations out on the internet.

Acceptance testing is not just kicking the tires on an aircraft. Every plane that comes out of the factory completes a series of acceptance test flights. Such tests are never done on ELVs, for obvious reasons. Each ELV flies only once; it cannot do acceptance test flights before delivery to the customer.

The process is very well documented.

> It borrows heavily from the FAA certification process for aircraft

It has nothing to do with aircraft certification, which is based on hundreds of hours of flight testing, with hundreds of takeoffs and landings. Such a process has never been applied to ELVs and, for obvious reasons, it never will.

> why put an avionics system on a rocket that has a safety failure rate of 1 x 10-9, when the
> rocket fails catastrophically at a rate of 1 x 10-2?

For one thing, component failure rates are cumulative. To make a rocket 99% reliable, you may need parts that are more than 99% reliable.

More to the point, why design a rocket to fail catastrophically at a rate of 1 x 10-2, when it would probably be cheaper (and certainly more useful) to design a rocket that’s reliable?

> For instance, Griffin has stated his goal for the next vehicle is a failure rate of 1 x 10-3. Make that
> a requirement for the next vehicle

In 40 years, no ELV has come close to meeting that requirement. What’s the point of specifying an ELV then specifying a “requirement” that no ELV can possibly meet? Even if an ELV could meet that goal, how would you prove it, with a rocket that might fly a few dozen times in a decade, at best?

The problem is, the process doesn’t make sense. I ask again, why put an avionics system on a rocket that has a safety failure rate of 1 x 10-9, when the rocket fails catastrophically at a rate of 1 x 10-2? Here again, NASA weighs the contractor down with massive numbers of requirements, hires boat loads of people to monitor their performance with respect to the requirements, has hundreds, if not thousands of safety reviews to make sure all of the requirements are met, and yet ends up with a vehicle that still has a 1 x 10-2 failure rate.

Wouldn’t it be better if they provided the contractor with a top level failure rate they thought acceptable. For instance, Griffin has stated his goal for the next vehicle is a failure rate of 1 x 10-3. Make that a requirement for the next vehicle, provide a financial incentive to meet it (as well as a disincentive not to), and let the contractor allocate it to the various subsystems in a way that makes cost effective sense.

And while they were at doing things that make sense (for a change) they could also stop mandating that reliability and safety engineers work for the program office, and provide them with the autonomy they once had to make a difference on these vehicles. Everyone always has all of these theories on how to make space travel safer and more reliable, but they never seem to include pulling the disciplines that obviously should be working these issues back from the brink of total irrelevance.

This seems like some pretty obvious stuff, doesn’t it? Remember when the term “rocket scientist” implied someone had a bit firmer grasp of the obvious?

You expect lobbyists to refrain from lobbying?

One rather cynical reading of this agreement is that the Defense Dept. and the EELV venders are expecting Dr. Griffin’s Thoikol-centric designs to fizzle out, and then DOD/EELV can step in and pick up the funding …

… without being blamed for aborting the beautiful new Solid Rocket Booster baby.

Don’t get me wrong, this is a fairly minor bit of bitching on my part. I’m still very much on board with what Brant and Chris said, but the time period between ‘idea/team/funding’ and ‘first NASA contract’ has all been just hand waiving. Is anyone at NASA working on making that process transparent and predictable?

Unless the second and third stages of this proposed lunar Crew Exploration Vehicle stack have surprisingly high performance, I think it can be rigorously shown that the existing four segment Solid Rocket Booster will not do the job.

Build a five segment version of the SRB? A five segment SRB is neither off the self, nor “proven,” nor “man rated.”