Aug 5, 2008

X-Ray Machine Zaps Itself

By John Fleck, Albuquerque Journal Staff Writer

A $350 million nuclear weapons X-ray machine at Los Alamos — already years overdue and millions of dollars over budget — inadvertently damaged one of its internal components last week, causing at least another three-month delay in the project.

A powerful beam of electrons used to create the machine's X-ray pulse vaporized a piece of graphite inside, fouling a vacuum chamber that needs to be ultraclean, according to lab spokesman Kevin Roark.

Seventeen of the X-ray machine's 74 massive doughnut-shaped power cells will need to be taken apart and cleaned, Roark said.

No one was injured and there was no damage other than the mess to the machine, known as the Dual-Axis Radiographic Hydrotest Facility, or DARHT, Roark said.

DARHT has been plagued by problems.

When it was proposed 21 years ago, the managers of the U.S. nuclear weapons program called it "essential." With a redesign midway through the project's development and repeated delays, its cost grew from an initial $30 million estimate to at least $260 million.

Then, when it was completed in 2003, lab scientists found that it did not work. They had to completely tear it apart and rebuild it, adding $90 million to the cost.

In May, the National Nuclear Security Administration declared the reconstruction a success, and lab officials said the machine would be ready for its first full-scale nuclear weapons test by September or October.

That test will now be delayed until early 2009, Roark said Monday.

Last week's accident, first reported on the "LANL: The Rest of the Story" blog, happened during tuning of the machine in preparation for the first weapons test, according to Roark.

DARHT generates two powerful X-ray beams. The first beam was completed in the 1990s and has been used successfully ever since. The problems, including the most recent incident, involve the second X-ray beam.

Fired at right angles into a target, they can create a three-dimensional image. The X-rays are powerful enough to penetrate metal, allowing nuclear weapons designers to study mock nuclear weapons as they are being detonated.

The pictures allow weapons designers to study the details of the early stages of a weapon's explosion without conducting a full nuclear test.

The machine works by firing a powerful electron beam, which in turn creates a burst of X-rays. Last week's accident happened when the electron was inadvertently focused on a piece of graphite.

Under normal conditions, the piece of graphite acts as a sort of window shade, allowing routine testing of the electron beam without actually using it to X-ray anything. But for reasons that are not yet clear, the electron beam was too tightly focused, burning off a piece of graphite, according to Roark.

"It was an unexpected result," he said.

Roark said cleaning up the mess can be done within the project's current budget.

27 comments:

Anonymous said...

GREAT!
Now another round of fixing the problem, declarations of success, promotions, distinguished performance awards, and bonuses for management.

DARHT is one of the genuine management successes at LANL.

Anonymous said...

What is left out is that LLNL designed and built the 2nd axis of DARHT. LLNL is responsible for the original failure to meet the specified accelerating gradient and is reponsible for delivering a product that did not have engineered control of the beam position.

But, when the problems arise, LANL gets the blame.

LLNL is a teflon laboratory.

Anonymous said...

From LLNL: The True Story

Saturday, August 2, 2008

A vote of no-confidence in "Livermites!"

What our friends at LANL are saying

Posted by scooby at 8:11 PM

3 comments:

Anonymous said...


"Don´t worry about comments from uninformed posters to the LANL blog - in this case they don´t even know the role LLNL played on the 2nd axis. Of course if something happened that involved the downstream transport and target, LLNL will be involved - it´s their hardware after all."

(August 2, 2008 8:41 PM)

(//llnlthetruestory.blogspot.com/2008/08/vote-of-no-confidence-in-livermites.html)

Anonymous said...

I think a more complete history of the 2nd axis would be in order. I think a 3rd national laboratory was involved...

Anonymous said...

Don't the designers at LLNL know about feedback control loops for high power equipment like DAHRT? It sounds to me like LLNL's bad design was at fault here.

And don't you know that LANS is beginning to really hate this blog...

"Last week's accident, first reported on the "LANL: The Rest of the Story" blog, happened during tuning of the machine in preparation for the first weapons test, according to Roark."

If LANS was more forthcoming communicating to both their employees and the press then this blog would have long since withered away.

Anonymous said...

Be credited with fixing this latest DARHT screw-up and maybe you, too, can make it to the inner sanctum of exalted TSM Level 6!

Anonymous said...

FTFA: "It was an unexpected result"

I would hope so!

Anonymous said...

On another topic. Staff are finally finding out what job box they're being stuffed into. Surprise! surprise! those who the lab screwed over and were partially compensated for by the Welch report are being screwed again and crammed into the lower boxes. Wonder how long before one of them flips out and and does a bit of box stuffing of their own.
How can the DOE be so f*** up?

Anonymous said...

"I think a more complete history of the 2nd axis would be in order. I think a 3rd national laboratory was involved..."

LBNL designed the injector, injector cells, and accelerator cells. LLNL designed the downstream transport and target.

Anonymous said...

For 7:46. DARHT is a pulsed accelerator. As such, it doesn't have the luxury of feedback stability. It also delivers more charge in a microsecond than the accelerators you are familiar with do in a week. BTW, the National Lab you are looking for in your design work finger pointing is LBL, not LLNL. Neither are culpable in this incident. Get your facts straight or get out of Dodge.

Anonymous said...

Gimme a break! anybody who's worked on accelerators knows the damage the beam can do. If they haven't designed systems to prevent it or the systems didn't work it can only be described as a really big screw-up!
Yeah, sh** happens but its still Sh**

Anonymous said...

All eyes on LANL. Any error or misstep means your (professional) life. The result is nobody wants to work there, just as planned by the "watchers." The bad guys are winning - and of course they've never, themselves, ever made a mistake. Especially not one they'd want to be kept secret...

Anonymous said...

Driving up past the big concrete barriers where the beam went under the road at LANSCE was always a good reminder of why you want to keep the beam where it belongs.

Anonymous said...

"It was an unexpected result"

You don't say.

Anonymous said...

"Unexpected" - Is that like 10e-1, 10e-2 or 10e-3 per shot?

Anonymous said...

"...maybe you, too, can make it to the inner sanctum of exalted TSM Level 6!"

Wow, and my GL told me there were no Level 6's at the Lab. Guess they just repeat what they're told. Of course, they also said there wouldn't be a TSM class system at LANL now as well.

Anonymous said...

Did anybody notice that there is no mention of the event in the News Bulletin?

Anonymous said...

From POGO:


DARHT Misses Target, Again, and Again

The Albuquerque Journal reported today that the Dual-Axis Radiographic Hydrotest Facility (DARHT), a $350 million nuclear weapons X-ray machine at the Los Alamos National Laboratory (LANL), vaporized a part of itself last week. According to lab spokesman Kevin Roark, "seventeen of the X-ray machine's 74 massive doughnut-shaped power cells will need to be taken apart and cleaned," causing a three-month delay in the project.

As the Journal pointed out, the National Nuclear Security Administration (NNSA) declared in May that the X-ray would be ready for its first full-scale nuclear weapons test this fall, following a $90 million reconstruction effort to repair the machine after it failed to work in 2003. Prior to the repair, the cost of the DAHRT program had already escalated from $30 million to $260 million due to repeated delays and a project redesign.

In addition to the issues raised in the Journal article, there is a major concern often voiced by LANL insiders and experts: the internal audit function at the Lab is failing to provide reasonable assurance to taxpayers that costs being incurred on projects like this are necessary and allowable. When there aren't enough project controls in place to gauge progress and identify problems early on, the cost associated with changing course or fixing problems becomes significantly higher.

The lack of a credible audit function is a serious problem given that LANL is embarking on major initiatives like plutonium pit production. Without better internal auditing, what's to stand in the way of tremendous cost overruns and endless delays such as those that have plagued DARHT?

-- Ingrid Drake

==================================

The article states that the cost "escalated from $30 million to $260 million." Now, that's what we at LANL call a "management success!"

Anonymous said...

He meant level six in the new job classifications formerly known as TSM. The new job classifications for the TSM series are too generic to even be worth bothering with, IMHO. For example we don't have electrical engineers, mechanical engineers, materials engineers, etc. Just R&D engineers.

And since I have a "computer science" degree, I am currently uncategorized because HR thinks I might belong in some "scientist" category instead of calling me an engineer, which is what I am. My GL has yet to iron that out with the bureaucrats.

Anonymous said...

Wasn't six sigma suppose to ensure these kinds of mishaps wouldn't occur?

Anonymous said...

8/6/08 5:41 AM said;
"The lack of a credible audit function is a serious problem given that LANL is embarking on major initiatives like plutonium pit production. Without better internal auditing, what's to stand in the way of tremendous cost overruns and endless delays such as those that have plagued DARHT?"

The Lab has auditors?

Naaah! You're kidding, right?

Anonymous said...

regarding auditors:

An auditor is just an accountant who is too dumb to steal!

Anonymous said...

The LANL newbulletin providng actual news?! Bwaahaaahaahaaaa....

Anonymous said...

As for DARHT's flameout...Thats a "feature".

Anonymous said...

" Roark said cleaning up the mess can be done within the project's current budget."

While that may be a true statement, the fact is that they were given those funds to run experiments, not to clean up messes.

Anonymous said...

"While that may be a true statement, the fact is that they were given those funds to run experiments, not to clean up messes." 8/7/08 8:23 AM

You are mistaken, 8:23 AM.

Cleaning up messes is EXACTLY what LANS is here for, as the science is slowly going away (or haven't you noticed?). Thus, the creation of a big new mess at DARHT should be seen as a great victory for LANS management.

Anonymous said...

Anonymous at 8/10/08 8:35 PM writes:
"..the creation of a big new mess at DARHT should be seen as a great victory for LANS management."

S/he is correct. AND, there will be bonuses for Mikey et al when the cleanup has been completed.