Thanks.
I can partially answer -- Kim has a lot more details on it, I'm sure.
The carrier plate is a piece of metal that has tiles bolted to it. It
fits between the RCC panels that make up the leading edge of the wing and
the tiles attached with adhesive to the underside of the wing, filling
the gap between the two. I'm not positive what the carrier plate is made
of, perhaps inconel, which is very heat-resistant.
One theory (one that Kim Keller has espoused) is that a carrier panel was
damaged by the foam strike and possibly came off the vehicle on flight
day two due to solar heating cycles. The resulting gap in TPS protection
caused the breach and, ultimately, the break-up of the vehicle.
Doug Van Dorn
dvan...@mn.rr.com
Actually, it is "carrier panel" (the guy on the board misspoke). There
are lots of them all over the vehicle. Basically, they are service
panels giving access to interior components. They are called carrier
panels because they "carry" or have attached to them thermal
protection material, either tiles or blankets.
> What is it made of?
Aluminum, about 1/8" thick.
> Any link of where there is a drawing of the assembled part in place.
Typically, they are rectangular panels with bolt holes that allow them
to be fastened to the ship. Shape may vary, depending on the shape of
the opening they cover.
-Kim-
former shuttle tech
Combining this information with the 0.3 m x 0.4 m areal dimensions of the
"mystery object" (source: CAIB), I estimate an average A/m (area to mass
ratio) of about 0.038 m^2/kg. That agrees well with the mystery object's
A/m, that I estimated from its apparent rate of orbital decay, about 0.040
m^2/kg.
The calculation of the average A/m takes into account that the objected
(source: CAIB), as described at the start of Section 4 of my analysis:
http://www.satobs.org/columbia/STS107mysteryobject.html
Given the uncertainties in my analysis, the closeness of the above
comparison probably is more apparent than real, but it does appear that a
carrier panel is in the ballpark, as are RCC and high density HRSI, as I
previously reported. Presumably radar signature will be helpful in
discerning among these and other possibilities.
Kim, what are typical areal dimensions of the carrier panels adjacent to the
wing leading edge RCC?
What is the density and typical thickness of the tiles attached to the those
panels?
Would those tiles likely have had remained attached to a carrier panel that
had broken free?
Ted Molczan
I don't have exact figures; maybe Steve W. will post them. They're something
like 4" wide by 22" long?
> What is the density and typical thickness of the tiles attached to the
those
> panels?
The tiles are going to be around 4" thick; again, I'd defer to Steve for a
better measurement.
> Would those tiles likely have had remained attached to a carrier panel
that
> had broken free?
Depends on how the foam hit them. The densified layer may have remained
after impact.
--
-Kim-
former shuttle tech
That's a good estimate. Their length varies depending on the RCC panel that
it's associated with. The RCC panels have varying widths.
> > What is the density and typical thickness of the tiles attached to the
> those
> > panels?
>
> The tiles are going to be around 4" thick; again, I'd defer to Steve for a
> better measurement.
Probably closer to 3" on the lower surface, 2" on the upper surface panels.
I don't have the density info handy.
> > Would those tiles likely have had remained attached to a carrier panel
> that
> > had broken free?
>
> Depends on how the foam hit them. The densified layer may have remained
> after impact.
Agreed.
--
Steve Wachowski
KSC OPF3 Technician
The opinions above are mine and mine alone,
I do not represent NASA or my employer.
> "Kim Keller" <keke...@cfl.rr.com> wrote in message
> news:nLxea.17057$o8.2...@twister.tampabay.rr.com...
>
>>I don't have exact figures; maybe Steve W. will post them. They're
>> something like 4" wide by 22" long?
>
> That's a good estimate. Their length varies depending on the RCC panel that
> it's associated with. The RCC panels have varying widths.
and
>
> Probably closer to 3" on the lower surface, 2" on the upper surface panels.
> I don't have the density info handy.
>
Thanks for the details...
In the mega-picture that was posted earlier, it looked as though the
carrier panel attachments were spaced around 12" apart, assuming that
the tiles immediately behind them are approximately 6" in width. That
would suggest that there are two bolts per panel?
--
Reed
I found that a carrier panel, with or without the densified layer, could
have matched the area to mass ratio of the mystery object. One that retained
3 inch thick, high density (22 lb/ft^3) HRSI, would not have been a good
match.
I guessed that the densified layer is 50 percent more dense than the bulk
density of a tile. Also, I guessed that strain isolation pads are not used
on carrier panels. If you have contrary information, I am willing to
reconsider the analysis. If pads are used, it would be helpful to know their
thickness and density.
Ted Molczan
"Steve Wachowski" <swach...@cfl.rr.com> wrote in message
news:DWZea.28335$o8.4...@twister.tampabay.rr.com...