Short bolts on space shuttle Atlantis 3

[kenvlach]

After 26 flights over ~20 years, NASA has checked the paperwork on bolts holding the KU-band antenna on the space shuttle Atlantis, scheduled for launch Aug. 27.
(see antenna as stowed in cargo bay & as deployed in space:

http://spaceflightnow.com/shuttle/sts115/060813kubolts/)

"A recent engineering review indicates two of the four bolts holding the KU antenna support box in place are too short. Engineers cannot directly inspect the bolts at the launch pad, but a paperwork review shows the bolts in question may be engaged by less than 2.4 threads. The requirement is 8.4 threads engaged."

Should the box break free during ascent, it would fall the length of the shuttle's 60-foot-long cargo bay and could cause catastrophic damage."

http://www.cbsnews.com/network/news/space/current.html

"Incidentally, NASA has known of potential problems with the bolts for some time... in fact, the bolts were replaced onboard sister shuttles Discovery and Endeavour after it was found the bolts may have been manufactured too short to safely accomplish their task.
CBS News reports the bolts were not replaced in Atlantis, however.
A two-day flight readiness review is now underway... to determine, among other things, if... and how... the bolts should be replaced."
--

http://www.aero-news.net/index.cfm?ContentBlockID=8af65a8d-52d2-4dac-8347-bc27d7e78e53&

 

[NickE]

Good Grief!

 

[metengr]

Isn't NASA required to set the standard for safety? They need to replace current administrators.

 

[MikeHalloran]

The 8.4 threads is clearly a statutory requirement.

If they had any actual engineers there, they'd figure out if 2.4 threads is enough, based on analysis of the joint, not analysis of the paperwork.



Mike Halloran
Pembroke Pines, FL, USA

 

[israelkk]

This is a symptom and not just at NASA. Instead of hiring more professional engineers and actually let them engineer, the companies higher industrial engineers, QA engineers and managers. There is so much paper work that actual thinking, analysing, designing (I do not mean 3D CAD modeling) but stress, dynamic and kinematic analysis, tolerancing, assembly design, etc. Designing more then one prototype, build them and test them and then decide which is the best to fit the task.

Today the QA of the actual parts replaced by the QA of the paper work. No one care that the bolts are short and no one double check them to save money. Every QA person just check the paper work.

As one of my colleague used to define ISO: ISO rules and paper work will assure that all your products are consistent. Therefore, if you are lousy designer and manufacturer all your products will be lousy. Paper work will not assure good products only careful design and participant of every worker in the process. For example the worker who assembled the bolt had to see (check) that they are too short. However, he is a pin head and expected by his boss to stay a pin head.

 

[swertel]

That is sad but true. I have known QA engineers performing a source inspection and not once did they even look at the product, just all the paperwork and certifications.

--Scott

For some pleasure reading, try FAQ731-376

 

[NickE]

TS16949 is not much better... Even here where I work I've seen the: Fix the print to fit the part not the part to meet the print... Then fix the paperwork to match....

 

[unclesyd]

The right thing is being done,,,for a change

http://www.mercurynews.com/mld/mercurynews/living/health/15306692.htm

 

[alexit]

Though they would never tell you, the guy mounting the antenna 20 years ago probably told the shop foreman "Hey these two bolts are too short", then went and put in a couple longer ones he had in his toolbox left over from working on his neighbors' car.

The six month process with 30+ engineers tracking down the QA forms and reviewing the drawings was time wasted because they went out and found bolts of sufficient length (though undocumented quality) already installed.

 

[Kenneth]

Wayne Hale (

http://www.nasa.gov/centers/kennedy/about/biographies/hale.html),

excerpted from the Post Flight Readiness Review Briefing Transcript (

http://www.nasa.gov/pdf/154909main_postFRRtranscript _2.pdf);

"Some 25 or 30 years ago, a mistake was made in the design of this particular component, the way this antenna is bolted onto the orbiter, and for the last 25 or more years, we had been flying with these threaded fasteners and bolts that just barely have a thread or two engaged in to the nut that holds them on."

"We had some questions about threaded fasteners and went back through an exhaustive review, which is still ongoing, of all the threaded fasteners on board the orbiter, in particular, and we found that in this particular application that bolts weren't long enough and had been that way for a number of flights."

"So, on Discovery and Endeavour, the two orbiters that we have in the maintenance facility, we immediately went to those vehicles and changed those bolts out, and that problem is resolved."

“On Atlantis, unfortunately, the access is very difficult. So we are doing some more work to try to understand exactly how much risk is involved with, A, either changing those bolts out, because any time you go to do non-standard work at the launch pad -- and this is in a particularly difficult location -- you run some risk, or, B, what is the real risk for this particular flight and could we accept this less-than-perfect application of a screw-thread fastener for one more flight."

"I think in this particular application, it is like a minimum of six and preferably eight plus through threads engaged, and we pulled one of the bolts off, and it had two-thirds of a screw thread engaged. Some of the other orbiter bolts had one and a half screw threads engaged."


As noted above, ultimately the "short" bolts were replaced:

http://www.nasa.gov/mission_pages/shuttle/behindscenes/bolt_replacement.html

and

http://www.nasa.gov/pdf/154918main_kuband_antenna_access.pdf

Paperwork review does not replace good design. Bad mouth "paperwork" (design) review all you want, but realize that in this case it identified this 25+ year old design engineering error. Just too bad the proper design review didn't apparently happen 25 years ago.

Re the Installer/Foreman “Correction”:
"Tracy Young, spokesperson for the Kennedy Space Center, confirmed the bolts removed from Atlantis were the wrong size -- a mistake made during the original manufacture of Atlantis some 25 years ago." (Source:

http://www.aero-news.net/PlainPage.cfm.)

So the designer erred and the installer didn't catch/flag the error.


Re "actual engineers": Do you really believe that NASA didn't extensively analyze this joint before making the decision to replace the bolts on the pad, but instead relied on a statutory requirement?

 

[MikeHalloran]

Given the labor turnover, I wonder if they really could analyze the joint. Certainly not during a power outage.

Okay, they make an easy target. Like shooting fish in a ... hubcap.



Mike Halloran
Pembroke Pines, FL, USA

 

[corus]

After 25 flights and 20 years of service without failure it seems to me that the bolts were designed perfectly adequately for the job, or am I missing something here?

corus

 

[Kenneth]

Based on the available information, I wouldn’t personally characterize anything about this situation as "perfectly adequate" (an oxymoron?).

While it was apparently "adequate" for 25 flights (based solely on the absence of a catastrophic failure), what about the 26th? The "it worked last time" rationale is not completely unlike that advanced by my wife when something (the car, washing machine, etc.) breaks/quits: "I don't understand how it can be broken, it was working yesterday." Extending that line of reasoning, nothing would ever fail.

If not supported by engineering, that's not risk analysis, and you might as well be shooting craps. It's also part of the "culture" that both the Challenger and Columbia Accident Investigation Boards charged NASA with changing.

If you discovered that two of four lug nuts on your car were engaged by only two thirds to one and a half threads would you fix it, or continue to drive on it secure in the knowledge that it hadn't stripped or loosened yet? And that's for a vehicle that's not going to potentially fall out of the sky if you guessed wrong (or you hit a pothole in the road that you somehow managed to avoid for the last 25 trips).

Reading through the FRR Briefing Transcript (

http://www.nasa.gov/pdf/154909main_postFRRtranscript _2.pdf)

you'll see that NASA challenged the analysts to determine if those two bolts were really necessary, and if the remaining two bolts alone were adequate to react the loads on the antenna.

Based on the "last minute" on-the-pad bolt replacement, the answer those analysts apparently came back with didn't appear to leave the decision makers at NASA with a warm fuzzy feeling (otherwise described in the FRR Transcript as "a good positive engineering margin on the structure as it exists".)

 

[unclesyd]

From the Wayne Hale statement.

"Some of theother orbiter bolts had one and a half screw threads engaged."

This is a wide open statement depending in how you interpate the word "other".

 

[Kenneth]

From the FRR Briefing Transcript:

QUESTIONER: Traci Watson, USA Today, for Wayne Hale.

First, a clarification. When you talked about the bolts that had only two-thirds of the thread or one-and-a-half threads engaged, was that on the KU antenna, or was that somewhere else? I missed that.

MR. HALE: Tracy, I could just barely hear your question. The discussion of the bolt thread engagement and I think that is the gist of your question -- there four bolts that hold this antenna onto the vehicle. Two them are very well engaged, and two of them are not, and the two that aren't, when we took those bolts off the two orbiters, we saw as little as, say, two-thirds of thread engaged and as much as one-and-a-half threads engaged. Neither one of those are good numbers.