Conference 7.286::space

    I hardly have a definitive stance on this, but I don't believe that
    the M-T engineers can be held responsible for the explosion.
    
    From all I've read and heard, I doubt that anybody KNEW that there
    would be a problem - there certainly was a higher probablibility
    of failure than ever before, which raised some questions regarding
    launch.
    
    
    I feel that the launch management team at NASA committed and error,
    and I would not be pleased to see any of these people's careers
    skyrocket over the next few years.  [Offhand, I'm not sure how they
    sleep at nights sometimes]
    
    People aren't perfect.  I'm willing to accept that.  And launching
    rockets is a risky business - practically by definition (take a
    few thousand pounds of high explosives, contain them, direct the
    blast, and aim up...), and I'm anxious to see what changes will
    be made in the future to make it safer.
    
    
    - dave

	I've been thinking about this a lot.   I'm sure that any heads-up
    engineer on the program would have made some personal contacts with people
    in the Astronaut corps.  The UseNet comment that the commander of the
    flight would be the person to get to rings best among the possibilities
    that I've considered.  
	After following the joint discussions in AW&ST, I believe that anyone 
    close to the design HAD to know that they were skating on thin ice.  
    Management had evidently closed off the 'thru channels' routes for getting 
    concerns of this sort aired, so I think a good engineer would have gone 
    outside of the normal channels and made darn sure the crew was appraised 
    of the M-T engineers' concern, especially since it seemed to be unanimous 
    on the technical level.
	I am disturbed about the level of competence demonstrated by the M-T
    engineers in the joint design.  Anyone designing things that spend time
    outdoors (houses, cars, etc) knows that you use gravity to shed water by
    shingling and draining.  Pointing a cup upwards is an invitation to 
    collect and maybe freeze water.  Even with insulation outside.  The new
    carbon wound casings have twice as many joints.  Not from my drawing
    board, they wouldn't (insert CAD terminal if you wish).   And the faith
    they put on that putty, with the few tests under environmental conditions,
    was not good engineering in my book. 
				- N. Chris Paulhus, P.E.

    I doubt they had the TIME to do any of this. The cold snap was a rather
    sudden event, and I would guess any engineer had less than 24 hours
    warning prior to the launch that it would be near freezing at the time
    of the launch. Could any of YOU come up with [pick a shuttle commanders
    name] phone number and give him a call within 24 hours? Could you get
    someone important enough at a major media interested enough to go
    national with this information in that much time? I don't think so. 

    re .3:  Yes, I agree with the same qualifications I made in .0.
    If I **KNEW** that it was a life or death matter, I feel that I
    could probably have done something.  (Surely all astronauts don't
    have unlisted numbers!)  However, it would take a LOT of work and
    I would take a LOT of shit for it.  Therefore, if I was worried,
    but not certain, I would not do it.
    
    Burns
    

    re: .*	I think some definition is in order here. If we
    		are debating a *design*, then there was plenty of
    		time to call of the shoot. If we are talking about
    		misuse, that falls squarely in the lap of NASA
    		flight directors.
    
    		I believe that the M-T tech's wrote the specification
    		for the SRBs plainly enough as related to launch
    		conditions (ie: temperature). I also believe (from
    		what I have read/heard) that they made a best-effort
    		to get the high-muck-a-mucks in NASA to postpone
    		the launch, because of the overnight temps being well
    		below the specification.
    
    		An analogy: some prescription drugs are lethal when
    		not used per directions. I don't expect ROHR or 
    		any other drug company to be held to blame because
    		someone swallowed a whole bottle of "reds". Misuse
    		is misuse, not bad design........
    
    		Summary: M-T engineers built the best bird that they
    		could, given the requirements.....NASA blew it!
                                        
    
    			DATman (...but they will redesign the SRB anyway)
    

	Re. .3   If I were a M-T engineer working on the SRBs, I most surely 
	could reach for my Roledex, or whatever, and within 10 seconds be
	dialing a number connecting me with John Young's office.  I may not
	be able to talk to Young, but I'd nail Crip or someone who I had
	met and who knew me.  To me that is a part of being a responsible
	engineer.  
	  I see too many engineers whose world is bounded by their workplace.
	They are not active in professional societies, standards writing
	organizations, or attending conferences of their peers.  They think
	of only the technical solutions/approaches to problems while a lot
	of what goes on in this world is operational/administrative, like it
	or not.  Having contacts with people throughout the industry, from
	competitors to suppliers to customers is needed in order for a good
	engineer to function well.  I'll admit that at DEC we engineers do
	not have as good a link to customers as we should, but we should keep
	trying.  At M-T with one customer for the SRB's, and a close working
	relationship with that customer (I'm familiar with this from my years
	at Electric Boat, working with the Navy), any good engineer would have
	made many personal acquaintances with customer type people  and should 
	be able	to get hold of them pronto! Leaving all of these contact 
	responsibilities to Field Service, Component Engineering, whatever
	decreases your leverage, your ability to affect what happens.  
	   So, in summary, 24 hours, hell. Give me 24 seconds! - Chris

    A recent issue of SUCCESS magazine had an article on networking,
    not E-NET, but similar. The proposition was made that (if it were
    important) anyone in the USA could be talking to Ronald Reagan in
    seven phone calls or less. That got me to thinking - my personal
    number is three. Think about it. Now all of you can do it in (at
    most) four!
    
    I still don't know what my views are on what the M-T engineers should
    or should not have done.

re: .5

You're right, NASA blew it! It was a spectacular explosion!


:-)
MartyJ

    re: .7	What does that mean "my personal number is 3" ?
    		And, what can we "do it in 4"? 
    
    		I guess I couldn't get a subscription to SUCCESS
    		magazine, as personally I believe that I couldn't
    		wake Ronny up with 700 phone calls, never mind 7.
    		
    
    			DATman (mystified_by_all_of_these_numbers)
    

    RE: .9
    
    Yes you could. In four calls:
    
    "my personal number is three" => I have a friend, who has a friend
    who knows Ronnie. So I call my friend, who calls his friend, who
    calls the White House. Three calls.
    
    "your number is at most four" => Call me - continue with scenario
    above. It is just a matter of full utilization of resources.

   re .10 and .7:  Yes, I believe my number is 3 also, but that does
    not mean I can reach the Pres in 3 calls.  I still have to know
    the right words to say.  (Would acquaintence number 2 believe that
    acquaintance number 1 believed that I believe that I have a sufficient
    reason for him to bother the next person in the chain?)
    
    That does not invalidate the reasoning regarding this topic, however.
    "The damn shuttle is going to blow up!" is pretty persuasive.  To
    me the question is still whether the engineers were sufficiently
    sure to go ahead with this strategy.
    
    Note that it is being said more and more recently that the engineers
    may have been right, but not necessarily for the right reason. 
    There were a number of problems with the joint which were not
    temp-related.  From anecdotal evidence, it looks like the engineers
    had only a .04 (1 in 25) chance of being vindicated (as opposed
    to being right) had they gone to extraordinary measures.
    
    Burns
    

    Actually, I suppose most people have a chain of 2 if they are willing
    to count their congressperson.
    
    Burns
    

It seems that the responsibility of the M-T engineers divides into two
categories:

	1. Design a joint which is reliable at extremes of ranges.
	2. Inform management of failures to do (1).

I am more concerned about item 1 than item 2.

I am concerned that there was not, apparently, adequate quality control
and adequate review of the joint design at Thiokol. This may have been
due to lack of time or money or training, or poor management structure.
I think it is unrealistic to expect engineers to do well at item 2,
particularly when management, as it certainly did in this case, turns a
deaf ear.

I do, however, believe it is realistic to expect that systems will be
designed to tolerate, with some considerable margin of safety, the
extreme ranges of conditions they will be subjected to. It is also
reasonable to expect that systems will be tested to the point of failure
on all ranges of conditions.

Does it freeze in Florida? Ask any citrus grower. Does it rain in
Florida? Lots. Should M-T design safety-critical joints to withstand
these conditions? Did they? Should we blame them for not convincing
management that the joints might fail, or is the problem really that the
joints should have been better designed, and more thoroughly tested?

A more general question: do we place enough emphasis, and enough value,
on quality engineering? That is undeniably an engineer's responsibility.

John O.

    I can see another way of looking at this.  I have seen this type
    of thing happen many times in the past as well.
    
    A M-T engineer manages to make the proper contacts by using enough
    pressure with enough highlevel people.  The word is the 'the shuttle
    will blow if launched'.  Very persuasive.  So, with the proper ammount
    of noisemaking, this engineer get the launched scrubed at a cost
    of many millions of dollars.  An insction is done to verify what
    the engineer as said.  Most likely, the only thing that would be
    found would be all of the 'known' problems with the launch vehicle.
    The engineer finds himself in a very 'bad' posistion and the shuttle
    blows up durring some other launch.
    
    This is one of many possibilities.  I can see it as a strong one.
    I can not fault the engineers at M-T or NASA.  This seems like a
    'system' problem.  Lets hope that THIS get corrected along with
    the launch vehicle.
    
    jim
    
    P.S. I also understand that NASA has been awarded 7.8 billion dollars
    for a shuttle replacement and recovery expenses.

    re .14
    
    Yes, I agree...most people would only go through the "call up the
    President or John Young" scenario if they were VERY close to certain
    that they were proveably right.
    
    Re the budget:  I believe that the $7.8B is their
    FYwhatever-the-next-one-is budget. I believe that it explicitly
    did NOT include funds for replacement with the idea that it would
    be put through as a supplemental when the powers-that-be decide
    who-what-when-where-how much, etc.
    
    The report of some inter-agency task force to do just that (figure
    out how to replace the shuttle) was due a few weeks ago and is expected
    any time as I recall.
    Burns
    

	  The 4/21 issue of AW&ST has some information on the new joint
	design. I'll try to draw it below. Basicly, they are going from
	one tang in a clevis to two mated cleviss.  They are canning the
	putty (never was properly tested and now found to act in ways
	they didn't want it to), and putting joint heaters on the outside
	that will keep rain out of the joint [ but the double clevis
	would also do that, except maybe wind driven rain that could
	swirl upward on the casing].   There's a couple of shots of the
	hole the fire was comming out of at the joint.  I'd like to know
	more about the insulation between the propellent and the casing -
	it must have contained the size of the opening quite a bit.


               | |                          | |
               | |                         /  |
               | |                        /   | 
               | |                       /    |
               | |                      |     |		
               | |                      |  ^  |
               | |                      | | | |
                -                        -   - 
                -                        -   -
               | |                      | | | |
                U                        U   U
  

              ^   ^                        ^   ^
             | | | |                      | | | |
              -   -                        -   -   <-- mating pin
              -   -                        -   -         hole
             | | | |                      | | | |
             |  U  |                      |  U  |
             |     |                      |     |
              \   /                        \   /
               | |                          | |

	Old Design		New Design

    re .16   insulation between the propellant and the casing
    
    
    In a typical solid engine design, like the SRBs, the combustion
    proceeds from a core down the middle of the propellant outwards toward
    the casing. That means for most of the burn time, you can rely upon
    the propellant itself to act as the insulator. The binders used
    in solid propellants are usually quite good at this.
    
    I do not know if you would need any insulation for the casing points
    near the joins between the propellant segments (i.e. at the casing
    joints). It may be that none is required, relying upon the pressure in
    the combustion chamber to force the propellant segments together. The
    puff of smoke at ignition that then stopped would support this guess. 
    
    Also, the geometry of the propellant grain is not a simple cylinder.
    The SRBs reduce thrust at about the same time the SSMEs are throttled
    back, before max Q. At about the same time the SSMEs are throttled
    up, the SRBs are increasing thrust again. This may mean that different
    amounts of the casing are exposed to combustion at different phases.
    It certainly means that combustion pressure varies which may have
    been a factor if the design relies upon pressure to force the
    propellant segments together.
    
    gary

    re: .17	I understand the reasons why the main engines are
    		throttled back prior to/during max Q...on a liquid
    		fueled engine, cutting back on the fuel volume/rate
    		would do that. But how do they control the "burn rate"
    		on the SRBs? 
    
    		BTW: I am sitting here at home and the 11PM news is
    		     showing the (finally) released footage of the
    		     orbiter free-falling after the "explosion"...
    		     crew compartment may have been intact......
    
    		It seems that our worst fears may have been real :^(
                         
    
    			DATman (let's get on with the program!)
    

    
    re -.1   The thrust of a solid rocket is proportional to the area
    of propellent being burned.  If you start out with a star shaped
    hole in the propellent, a lot of area will be exposed. As the fingers
    of the star burn off and the hole becomes circular, the area is
    reduced and the thrust goes down.  The natural tendency of a solid
    rocket is to have the thrust increase because the hole becomes larger
    as it's diameter increases. You can play these two effects against
    each other to get simple thrust profiles.  By mixing inhibitors
    in the propellent and having them in cylindrical layers that are
    exposed at various times, you can quickly reduce the thrust at those
    times.  Not quite as straightforward as it appears, eh? - Chris
    ps.  My question on insulation has to do with the section I saw
    that showed insulation over all of the casing EXCEPT just above
    and below the joint.  I'd have designed the insulation to dovetail
    over the joint, and keep it from the fire. I wonder if the insulation
    above and below the joint kept the hole in the casing as small as
    it was.  (I remember a quote shortly after the accident, that if
    fire breeched the casing, you would quickly have a hole "you could
    ride a horse through". The 15" X 22" hole shown in AW&ST would require
    a very small horse, and rider :-). )
    various 
    

    The grain geometry of the SRBs is a mixture of cylindrical
    (progressive, or thust increasing with time), star shaped core (neutral
    or constant thrust) and slotted (which can increase, decrease or
    remain constant over time).
    
    I don't think they vary propellant chemistry. From memory, the segments
    are loaded in a single pour (the propellant is mixed somewhat like
    an epoxy resin, loaded while liquid and allowed to cure in the casing).
    
    The star shaped core was originally design to give a constant thrust
    over time although it can be varied to be progressive or regressive.
    
    gary

    
    (From recollection of a recent AvWeek article:)
    
    There is inhibitor place over most of the "joining surfaces" of
    the propellant grain (i.e. where the meet surfaces of adjoining
    segments).  However, the bottom surface of the second segment (just
    above the joint that failed) and the bottom surface of the third
    segment are NOT inhibited.  Thus after a certain amount of time,
    the second and third segments are burning both on the end and in
    the center.
    
    I believe that at least part of the thrust increase after max-q
    comes from the fact that the ends of these segments have started
    to burn.  Note that this would imply that the flame surface would
    reach the outside casing near the joint that failed at just about
    exactly the time of the failure.
    
    Burns