What ultimately destroyed the Columbia was an uncontrolled reentry which caused severe overheating, burning and disintegration of the heat resistant materials, (the protective TPS tiles and RCC panels),  covering the orbiter.  This eventually allowed for the burnthrough of the aluminum skin and then the thermal attack of critical structures within the shuttle's airframe.  This weakening of the Columbia's structure was coupled with an aerodynamic moment applied to the structural frame of the orbiter in a manor which was inconsistent with its design.  As the Columbia began to tumble and turn sideways against its intended flight path, the forces applied to the airframe were simply far greater than the shuttle was designed to handle.  One general specification for the shuttle states that it should be able to withstand a 2.5 G load anywhere along the longitudinal X axis of the orbiter.  Although the space shuttle is by far one of the sturdiest aircraft ever built, it was never expected that the orbiters would see extreme long duration lateral aerodynamic forces against the fuselage.  This type of phenomenon is not unheard of, during the early years of X-Plane research one of the X-15 aircraft reentered the atmosphere sideways and broke up in the upper atmosphere due primarily to the aerodynamic forces.  The question is what caused the loss of control of the space shuttle Columbia?

Effects of a Wing Breach

Observational Analysis F1 11/18/2003 The Case for the Final Report was Made Using OEX Data It seems that the C.A.I.B. was unable to make the case for a wing breach using only the sensor data that was downloaded during reentry via the TDRS system.  Perhaps this data was too random in both location and sensor readings to be of much use.  Fortunately the OEX data recorder found in the debris field had enough data on the large magnetic tapes to help put together what happened after LOS as well as fill in some of the holes where data was lost during reentry due to the many intermittent communication blackouts with the S-Band antennas.  However, the existence of the OEX recorder itself is questionable. The sensors that Mission Control didn't see: Sensors for the MADS/OEX were originally located near RCC panel #8/9 location because that is the 55% point on the wing.  Sensors were also located at other strategic points along the length of the wing. Fig. OA-F1-1 V09T9910A - Wing Leading Edge Clevis Temperature V09T9895A - Wing Leading Edge Spar Temperature V12G9921A - Wing Leading Edge Spar Strain V07T9666A - Aft Panel Lower Surface Temperature Fig. OA-F1-2 There are three temperature sensors and one strain gauge that are cited as being the source of the data that made the strongest case for the RCC panel #8/9 breach scenario.  This side view, (above Fig. OA-F1-2), and the plan view, (near left Fig. OA-F1-1), shows the location of the sensors as stated in the STS-107 Accident Investigation Final Report Vol. I.   Fig. OA-F1-3 The above photo, Fig. OA-F1-3, is used in both the STS-107 Accident Investigation Working Scenario and Final Report Volume 1.  It is a closeout photo that should have been taken after repair work on Columbia's left wing or after an overhaul.  It is unknown where or when the closeout photo was taken. The sensor labels have been altered from the original documents to show the correct MSID's.   Because these four particular sensors are listed as sending data to the OEX recorder, they and their associated wiring are the type of items that should have been removed during Columbia's last overhaul in Palmdale, see Columbia's most recent overhaul.  They were installed when Columbia was new as part of a suite of sensors designed to aid in post flight engineering analysis.  As can be seen in the cutaway view of the wing there are many sensors for taking temperature, pressure and structural measurements located throughout the wing.  Those feed real-time data to the crew of the shuttle as well as mission control about the health of the shuttle. Of the temperature sensors listed above there are serious discrepancies as to what the MSID's actually are within both the Working Scenario document and the Final Report.  Some pages show the sensors as having MSID's beginning with V09T while other pages show the same sensor as being V07T.  After carefully reading through technical reports from the Dryden Technical Report Server it has been determined that sensors with MSID's that are V07T####A usually take temperature readings near the surface of the TPS material while those that are V09T####A are generally located on the inside of the shuttles skin. Further research indicates that the current MSID's applied to the sensors here are correct and the sensors will be referred to as such from this point on within this site.  The MSID's may still appear incorrect in the Working Scenario and Final Report documents. All the graphs below are from STS-107 Accident Investigation Final Report Volume I. Fig. OA-F1-4 Fig. OA-F1-4  is reported to be data from the strain gauge MSID V12G9921A that is shown in Fig. OA-F1-1, OA-F1-2 and OA-F1-3.  What the graph appears to show is a slight increase in strain during the period from approximately the 250 second mark to the 400 second mark.  After that there is another small increase followed by a sharp downward movement which indicates negative strain or that there is compression on this member.   There is strange data at the 500 second point that seems to show reading that are nearly off scale hi and low.   After that the strain goes back to zero until sometime after the 900 second point where another hi/low reading occurs.  The reading then stop at LOS. Fig. OA-F1-5 Fig. OA-F1-5  is temperature data from the sensor with MSID V09T9910A which is located on the outside of the wing spar and inside the cavity created by the RCC panel.  It also shows a slight increase in its reading just as the strain gauge did with the same hi/low occurrence at the 500 second mark. This sensor indicates a 50°F temperature increase over 200 seconds with a momentary temperature increase to 650°F at EI+487.  The temperature then drops to -200°F which is effectively off scale.   Fig. OA-F1-6 Fig. OA-F1-6  is data from a temperature sensor, MSID V09T9895A, on the inside of the wing spar.  The data from this sensor, inside the wing cavity, is similar to that shown in Fig. OA-F1-5. This graph shows a similar temperature increase as the one above, about 125°F increase over 100 seconds or less with a very small amount of data at 450°F for only a brief moment at EI+522.  Then a constant reading of -200°F which is also the off scale low for this sensor. For this sensor a heat transfer analysis should be done to check the period of time required for the inside edge of the wing spar to reach the stated temperatures, (a transient conduction analysis). Fig. OA-F1-7 Fig. OA-F1-7  is TPS temperature data from a sensor MSID, V07T9666A, on the underside of the wing near the leading edge.  Again anomalous readings occur at the same times as the other sensors.  The erratic data from the sensor shown as the sharp peaks and valleys of the graph along with an odd spattering of data points in those areas would tend to indicate something other than overheating. Something like this might be expected in the event of an explosion that destroyed the sensor.  It's also possible that the particular tile this sensor was attached to was destroyed or fell off the wing at this point. It is unknown what relationship this sensor data has to a breach in the WLE RCC panel. The effects of the strain reading, Fig. OA-F1-4, of -100 micro in./in., a negative strain says that a member is in compression, on the shuttle wing spar is unknown but based on earlier flights is appears the member is easily capable of handling that much strain in the positive direction.  The temperatures in the other graphs are not capable of harming the shuttles components. The temperature data shown in Fig. OA-F1-5 and OA-F1-6 would be an example of steady state thermal conduction through a flat plate which is the wing spar.  The temperature on the outside of the spar that is exposed to the super heated plasma Fig. OA-F1-5 should start rising earlier and rise faster and higher than the temperature taken from the inside of the flat plate or wing spar.  The outside temperature does start earlier but, when compared to the inside temperature, does not rise high enough or fast enough to be part of a steady state thermal conduction circuit.  Therefore it is unlikely that this temperature data has anything to do with a wing breach and thermal heating of the wing spar. Summary / Conclusions: If the data from the graphs in Fig. OA-F1-4, OA-F1-5, OA-F1-6, OA-F1-7 is actual data from from the reentry of STS-107 and not fabricated, it definitely indicates that something significant happened between EI+250 and EI+500 but exactly what is unclear.  Because these sensors reported directly to the OEX recorder, Mission Control would not have seen any of this data during reentry.  This site has already reasonably shown that the OEX recorder was not on Columbia during STS-107 and was planted in the debris field later to guide the investigation to a predetermined conclusion.  Because the discovery of the OEX was a reasonable explanation for the introduction of new data, it makes it extremely convenient that nobody else besides C.A.I.B. members would have seen it and would be able to question the data's authenticity.  Whether the data actually came from anytime during reentry of STS-107 or if it was simply fabricated is unknown. It has also been noted previously that areas of the shuttle left unprotected due to the loss of a thermal tile or in this case a large section of RCC material have remained undamaged.  From the document, "Risk Management for the Tiles of the Space Shuttle";1994; Interfaces 24:1; Pg. 72, the following observation was made, It is interesting to note, that in the two cases in which tiles have been lost in the past, burn-through did not occur (in one case, the tile was lost over a service hatch and the extra structure in the shuttle's frame was able to distribute the increased heating). Another document assessing the risk probability to the shuttle in the event that the TPS is damaged, "Safety of the Thermal Protection System of the Space Shuttle Orbiter", gives the statistical probabilities that the aluminum skin of the orbiter will burn-through if the TPS is damaged as well as the probability that the shuttle will be lost if that happens.  These values are based on both program flight history and engineering analysis of different locations on the shuttle.  These values may be added to get the total probability for loss of the Space Shuttle from a foam strike but the individual probabilities are 0.25 for a burn through and 0.05 that the shuttle will be lost due to a burn through at the leading edge of the left wing.  Here is a breakdown of some of the information from that document, Probabilities for Loss of the Space Shuttle. Copyright © 2003 - 2007 ColumbiasSacrifice.com

History of foam damage:

The current official theories imply that a damaged left wing caused the loss of control which is certainly possible.  Even though the shuttle had a computer controlled avionics system to compensate for flight errors during reentry, the damage may have become so great that the system lost its ability to compensate resulting in the loss of the vehicle and crew.  The Soviet space shuttle, "Buran", suffered significant thermal damage as it reentered the atmosphere  on its one and only orbital flight.  These pictures, Buran Pics, show the damage done to Buran, and some of it seems pertinent to possible damage incurred by the Columbia during STS-107.  Although the Soviets deemed the Buran too costly to repair after the flight, the shuttle was not destroyed and was by all accounts repairable.  (Note the missing section of leading edge RCC material on the wing).  

The document, STS-107 Master Timeline Rev15.xls, seems to show an odd sequence of events where communication losses occur during OMS/RCS jet operation.  This would imply that the five General Purpose Computers (GPC) that were in charge of running the flight software, controlling the OMS/RCS jets, communications and many other miscellaneous tasks were continuously going off line at almost regular intervals.  Each time one GPC went off line its backup would come on line and seek to reorient the wandering shuttle as it was also trying to restore ground communications via the S-Band antennas.  This appears in the timeline as the statement, "OMS/RCS jet firing bounded by data loss".  These events were caused by damage to the main GPC and all four backup computers in successive order.  These events began before any of the off nominal sensor readings and the GPC's are not located anywhere near the damaged area of the left wing.

From virtually the beginning of the shuttle program there has been a tremendous amount of interest in the effects of debris damage to the shuttles TPS.  Many technical papers have been written on the subject with the left wing, wheel well and wing leading edge being the main areas of concern.  Most of the analysis are done with the TPS at 100% thickness and also with some assumed damage leaving the TPS at perhaps 80% thickness.  Any technical studies showing the effect of missing tiles on the space shuttles reentry have not yet been located.  Many shuttle missions have collected specific data on the temperature variations at various critical locations on the orbiter during reentry.  These temperature measurements have been taken from both the surface of the TPS itself, the aluminum skin and internal structures of the shuttle.

For design purposes of the TPS for the space shuttle, engineers came up with a maximum allowable temperature that the skin of the orbiter could be exposed to without incurring serious damage.  Even though aluminum doesn't melt until it reaches approximately 1200°F, the critical temperature was set at 350°F.  This temperature was probably determined as the maximum allowable for critical systems under the orbiters skin that could be harmed by excessive heat.  These systems could probably take several times that amount of heat but standard engineering practices dictate putting in a Safety Factor (S.F.) of (2) or (3) to help insure that the systems remain safe even under extraordinary circumstances.  None of the temperature sensors ever registered anywhere near the critical temperature of 350°F. The documents, STS-107 Ground Track,  STS-107 Sensor Failure, and Event Sequence give some idea of what those sensors were showing during reentry.

Columbia's pre-reentry condition: (Based on official assessment)
The suggested post impact condition of these areas are damaged, (chipped), tiles and gouged RCC material on the leading edge of the wing.  There may also be some areas with a small number of missing tiles.  Fig F1 indicates the damaged areas of STS-107.

Fig. F1

TPS Damage Underside of Left Wing and Landing Gear Door

Damaged tiles:

The Columbia lost thermal tiles even before its first space flight when many of them fell off while the shuttle was riding on the back of its 747 transport aircraft on the trip from Palmdale to KSC for its first launch.  From that point on tremendous improvements were made in the fabrication of the tiles and their attachment to the orbiters skin.  After STS-1 the Columbia required the replacement of over 300 thermal tiles while after STS-4 the Columbia required only 40 tiles to be replaced.  After STS-5 tile damage was virtually negligible.  The most severe damage previously done to a shuttle TPS was on STS-87 when the Columbia sustained extensive tile damage.  The source of the damage inflected on Columbia during STS-87 was ET insulation foam which separated from the external tank during launch and ascent.  A new type of foam and the process for applying it turned out to be the culprit in that case.  A post flight inspection report counted over 300 hits with some of the damage penetrating 75% of the tiles thickness.  The damaged areas included the underside of the left wing and the wheel well door.  While no special precautions were taken for reentry, the orbiter did not sustain any structural damage nor was there damage to any of the shuttles systems.  At no time during reentry was the shuttle or her crew in jeopardy.

Fig. F2 shows the basic substructure in the effected area from Fig. F1.

Fig. F2

The document, NASA 2657 (Finite Element Reentry Heat-Transfer Analysis of Space Shuttle Orbiter), is a very detailed analysis of heat transfer to the orbiters skin in the same areas that are in question with STS-107.  The heat transfer analysis done in this document shows that with the TPS at 80% of its effective thickness there is only a small increase in heating to the internal structure of the orbiter.  Another document makes a similar analysis, NASA 85907 (Thermal Response of Space Shuttle Wing During Reentry Heating).  Fig. F3 from NASA 2657 (Finite Element Reentry Heat-Transfer Analysis of Space Shuttle Orbiter) shows the effects of moderately damaged TPS material, (80% of its effective thickness), on the shuttles aluminum skin temperature.  At no time does it approach the 350°F limit set by NASA and in fact at the time that LOS occurred the skin temperature is still well below 0°C.  If we go by the STS-5 data the temperature for undamaged TPS would have been about 23°F and if we double that to simulate damage the maximum skin temperature is still only 46°F.

Fig. F3

The temperature of the shuttles skin at the point where LOS occurred should have been well below the 350°F limit set by NASA irregardless of minor to moderate damage done to the TPS.

Missing tiles:
The method of bonding tiles to the surface of the orbiter has improved dramatically since 1981 based on experience and the introduction of new adhesives.  The are almost no instances of a shuttle missing tiles without some sort of cause such as debris impact since the first two or three launches.  The temperatures seen in these areas do get hot enough to melt the 6061-T6 aluminum skin of the orbiter, (the melting point 6061-T6 aluminum is 1250°F), or damage sensitive components under the skin, see shuttle temperature variations.  However, it has to be determined how long it would take to add enough heat to these areas to damage the orbiter.  Remember the orbiter had just left the Earths shadow when it began reentry so its temperature was 0 K or -273°F.  Fig. F4, from NASA document "Safety of the Thermal Protection System of the Space Shuttle Orbiter" December 1990, shows the stack up of materials that incorporate the TPS.

Fig. F4

There are two reasons why there has never been a burn through in areas of the shuttle with missing tiles.

1. The nature of the flow over the surface of the orbiter is such that the super heated plasma does not make constant direct contact with the TPS material.  Because the flow is laminar much of it passes over the small breach in the tiled surface and then trips to turbulent flow at the aft edge of where the missing tile should be.

2. The heat that does reach the aluminum surface of the orbiter is immediately conducted throughout the aluminum structure around the breach in the TPS.  In this case because the initial temperature was -273°F the structure had a long way to go before reaching a critical temperature.

Fig. F5 gives the heat flux, (heat that can transfer to the shuttle), during reentry.

Fig. F5

On the windward wheel well surface, an estimate of heat transfer to the shuttle on a missing tile patch that is 4.00 in. x 4.00 in. over a period of 900 sec. at 0.04 BTU/in.² - sec. would be,

For comparison a high end Weber Bar-B-Que running liquid propane or natural gas puts out 0.018 BTU/in.² - sec.

This is definitely enough heat input to melt through a thermally insulated portion of the wing.  However, the rate at which heat was being conducted away from the effected area to other parts of the wing should have been great enough that in the short time of exposure, (approximately 10 min.), there should not have been enough damage done to destroy the shuttle.  Remember, the Columbia was at or near absolute 0 when it began reentry.

Leading Edge of Left Wing

Damaged RCC Material:
The Reinforced Carbon/Carbon (RCC) material on the leading edge of the space shuttles wings cools the shuttle through ablation.  Ablation cooling means that the material loses heat by shedding small particles of carbon.  Over the course of time the RCC material loses mass and at some point must be replaced.  The document, "Orbiter Reinforced Carbon/Carbon Design and Flight Experience", Indicates the kind of damage incurred by the RCC material over the years and what has been done to improve its design.  Aside from pinholes which have been a common occurrence on all of the orbiters the RCC material has never been seriously breached, and there has never been a case of a missing section of the RCC material.  The inconel and cerachrome are extremely high temperature alloys that have to be burned through before any of the high temperature plasma can enter the wing.  Fig. F7, F8, F9 and F10 show the basic design of the wing leading edge section of the Columbia orbiter.  Even if the RCC material is missing, you would have to burn through nearly 0.500" of aluminum before exposing the interior of the wing.

Fig. F7	Fig. F8

Note that OV-102 Columbia has a different design of the wing leading edge structure than the other three orbiters.

Melting point Inconel 718 = 2420 °F
Melting Point Inconel 601/Cerachrome = 2460 °F

Fig. F9	Fig. F10

The following images are purported to be of Columbia in orbit during STS-107.  In any event, only the upper portion of the RCC material on the left wing leading edge is visible.  It is unclear whether these images are intended to show that there was damage done to the TPS during launch and ascent, or to show that the TPS was intact before reentry.  Any feedback as to the validity and significance of these images would be greatly appreciated.

webmaster@columbiassacrifice.com

Fig. F11	Fig. F12
Fig. F14	Fig. F15