The Society of Professionals, Scientists, and Engineers (union) at LLNL has a short six question on-line poll for LLNL and LANL employees on the future of management models at these labs. They plan to share these results with LLNS & LANS, with NNSA, DOE, and Congress.
SPSE Survey
Main website:
SPSE
12 comments:
Sharing the results with the very powers that created this mess. Seriously, how is this going to change how LLNL is run?
SPSE has no teeth, no clout, nothing. Like a large dumb animal, the only way to get the US goverment to pay attention to you is cause it a lot of pain. And then you'll get it's full undivided attention.
Showing congress a piece of paper claiming to be a survey is a bad joke at best. You want them to react, hold a work stoppage and get the media to cover the action. Tell the country that thanks to DoE and Bechtel our nuclear inventory is now uncertifiable and unable to be upgraded.
Please tell us something we don't already know.
Tell the country that thanks to DoE and Bechtel our nuclear inventory is now uncertifiable and unable to be upgraded.
How do you know this? Do you know or are you making an assumption? I think you are making an assumption which is never a good thing.
6:31 asks how..
Aside from the very public concerns in the news recently by SecDef Gates?
How about common sense? Take the fact that everyone at LLNL including the designers have taken a hit in bodycount. The people still left will be leaving soon, taking knowlege not in the notebooks with them.
Take a look at the public description of the RRW. Based on existing designs. Public admission that we no longer can innovate, only replicate.
But we have computer models to use for verification...
Yes, models. Of what we think is how things work. Not fact, a guess. Remember what happened last time a weapon was built using guessing ( during the first test ban )? A fizzle. Will a model reveal a manufacturing defect, out-of-spec part, or an unaccounted-for degeneration mode? Not a chance. Only real tests will reveal unforeseen problems. Which is why you'll never seen a car maker rely solely on computer models for product safety. NASA lost two shuttles because they trusted models more than engineers with hands-on experience.
As long as the designers were still around you could reverse the mistake of no testing. But DoE and Betchel have made sure via layoffs etc. that we could not field a weapon that is guaranteed to work if our lives depended on it. Which unfortunately they do.
SPSE is a joke. Always has been. Always will be.
More evidence of why replacing testing with just computer modeling is a bad idea.
Russian deputy PM urges Roscosmos to focus on Bulava ICBM -
RIA Novosti
Quoting part of the article....
"There are flaws in the testing of the [Bulava] components on the ground. Often, real tests are substituted by mathematical calculations to match technical requirements," Sergei Ivanov said at a meeting with Russia's space officials.
This coming from a nation that has a reputation for top-of-class mathematicians and computer programmers I just can't see how blandly saying we can manage to design without testing simply because we have the TSF and they don't either. Anyone who has had to struggle with a graphics driver update will tell you the hardware is irrelevant if the software can't do the job properly.
More evidence of why replacing testing with just computer modeling is a bad idea.
This was a test of the delivery vehicle, not the weapon. You are comparing apples and oranges.
January 11, 2009 5:47 AM
Just about everything you have represented as fact is pure fiction.
There was no "fizzle" and the space shuttle disasters had nothing to do with computer simulations gone awry.
January 31, 2009 9:49 AM and 9:52 AM...
Apples vs Oranges.
No it's not. It's whether you can model a system completely enough to avoid testing. One could argue that compared to a physics package a missile should be a piece of cake to model. The article supports the statement that modeling without supporting testing is inadequate.
Claims of fizzles being fiction and shuttle disasters having nothing to do with models.
Go look at the photo of Upshot-Ruth in the B111 lobby. That picture was put there as a reminder to all designers to not get cocky.
Count how many LRL tests failed in '53 and how many during the Nougat and Dominic series.
Anyone can read the official reports about the shuttle disasters and see the references to models.
Go look at the photo of Upshot-Ruth in the B111
That picture has nothing to do with a test ban. That was Lawrence Livermore's failed attempt at building it's first nuke long before there was supercomputing capability.
The space shuttle, the most complex machine ever built (orders of magnitude more complex than a nuke) was never tested. It was launched as a fully integrated system with out an all up test (successfully launched). The failures that resulted years later were due to human error (ego and arrogance), not a fundamental hardware failure.
I'm afraid you don't know what you are talking about.
February 2, 2009 4:37 PM.
First you denied ANY fizzles. I listed a bunch. Including the shots after the test moratorium. Of course Upshot-Ruth was earlier! It still was a huge failure of a theoretical model.
And you agree the shuttle was not tested properly and futhermore that management blindly trusted what they believed in ( models and statistics tables ) rather than line workers and testing.
It's not whether the numbers were crunched by a computer program or done on a napkin. It's whether you are so obsessed with your precious theory that you field a weapon without bothering to check if it really works.
Rather than wave your hands claiming I'm wrong, cite your evidence to prove your statements.
Better still - PROVE that the existing models capture ALL modes of failure for a design. If you can't, then we'd damn well better test.
And until then - the inventory is uncertifiable.
The fact that SPSE cannot get the professional workers to be represented by it, even in this time of disaster at the Lab, indicates what these people really think about SPSE. SPSE has a reputation for being a group of bottom quartilers who are not representative of the real Lab.
Post a Comment