Path: utzoo!attcan!uunet!lll-winken!lll-tis!mordor!joyce!ames!think!craig
From: craig@think.COM (Craig Stanfill)
Newsgroups: sci.space.shuttle
Subject: Re: "Walking"
Message-ID: <29792@think.UUCP>
Date: 27 Oct 88 21:56:20 GMT
References: <665@sas.UUCP> <4964@hplabsb.HP.COM>
Sender: news@think.UUCP
Reply-To: craig@mneme.think.com.UUCP (Craig Stanfill)
Organization: Thinking Machines Corporation, Cambridge MA, USA
Lines: 59

In article <4964@hplabsb.HP.COM> dsmith@hplabsb.UUCP (David Smith) writes:
>In article <665@sas.UUCP> sasbrb@sas.UUCP (Brendan Bailey) writes:
>>do something they called "walking" which took place when the
>>main engines ignited

>When the boosters ignited, the shuttle lifted off and became free to
>"walk" under the side force of the liquid engines.  If you review a
>tape of the STS-26 liftoff, you can see the lateral movement.
>I seem to recall talk after STS-1 on the order of "Now that we know
>how much it walks, we can compensate for it in the future."  But I don't
>know why that aspect of the flight should have been an unknown to
>the engineers, nor do I know how it is compensated.  

I'm guessing on this, but the dynamics of the shuttle-EFT-SRB system
when the solids ignite and the holddowns release are probably complex,
and could not be exactly determined without a launch.  Remember that
the spacecraft is an elastic system, and at the moment of launch,
between SRB ignition and holddown release, the forces acting on the
system and the point at which they are acting suddenly change by several
million pounds.  The result will be short-lived system oscilation.
So what you've got is a system with 3 SME'e and two SRB's flexing around.
The result is that the axis of thrust is going to change, and the space
craft is going to ``walk''.

The crux of the problem: how much and how fast?  If you know the
dynamics of the system, you can use active control methods (gimboling
the engines) to compensate.  I'm sure the engineers had a pretty good
idea of system dynamics before STS-1 based on computer simulation, but
these simulations are never exactly right. If your simulation is off
by, say, .1%, you're going to walk a little between release and the
time you can gimbol the engines to stop the walk.  The only way
to be exactly right is to light the candle and find out.  Then you
know how much to compensate for the walk before it happens.

An additional source of walk could be interference between the SRB
plume and the launch pad, but I'm less sure of this; most of the plume
is deflected sideways by the launch pad, and the space craft doesn't
care what happens to the plume after it leaves the nozzle.  However,
if SRB ignition changes atmospheric pressure at the pad noticably, it
would affect the thrust of the various engines.  You might also get
aerodynamic forces acting on the wings as air displaced by the SRB
plume rushes forward.  Again, you don't know accurately enough to
compensate until you launch it.

A final possible source of walk could be if you don't know the thrust
profile of the SRB.  I suppose it's possible that there is a differnce
in the thrust profile between a horizontal SRB at a test facility and
a vertical SRB on a launch pad, but this seems unlikely.

By the way, to the extent that these effects vary from launch to
launch, every launch will see some walk.  If you change the
distribution of weight in the cargo bay, you change the system
dynamics, and you walk a little more or less.  Each SRB probably
has a subtly different burn rate at ignition, and again this changes
the system dynamics and you walk a little.

As I say, I'm guessing on this, but it seems reasonable based on first
principles.  My guess is that uncertainty as to system dynamics is the
major cause of the large walk on STS-1, but I wouldn't completely rule
out the other causes.