Go To Home Page

Observational Analysis E1A
12/27/2003

Examining The Official RCC Panel Debris Impact Test

 Go To Table Of Contents

For a number of months the public waited for the C.A.I.B. to provide an answer to one of the most burning questions which was,  where did the foam strike the left wing of the Columbia?  There was a rather complicated shell game of where the foam may have impacted the orbiter.  Possibly near one of the RCC panels or maybe it plowed into the main landing gear door.  There was also the possibility that a magic piece of foam similar to that magic bullet that entertained us so well in the Warren Commission Report ricocheted off different surfaces producing several different holes.  Of course the most likely candidate for a breach location needed to be near temperature sensors that trended upward rather than those that went low, stayed neutral or fell off scale during reentry.  The final location for the hole placed it somewhere on RCC panel 7, 8 or 9.  Conveniently there were several sensors near the area that collected data confirming the existence of a breach in that area.  There was a delay in getting that information because the sensors sent their data only to the OEX data recorder which was found later.  Miraculously once again the OEX recorder saves the day.

News story with comments from Admiral Gehman on impact testing

A few observations made by the C.A.I.B.:

Fig. OA-E1A-1

STS-107 Accident Investigation Final Report Vol. I, Pg. 60, Fig. 3.4-4 

The image to the left, Fig. OA-E1A-1, is directly from the STS-107 Accident Investigation Final Report Vol. I.  The graphic simply shows a couple of different scenarios of how foam traveling parallel to the orbiters body might impact the leading edge of a shuttle wing.  An impact near the apex of the RCC panel would result in a large angle of incidence, nearly 90°, and will impart a much greater percentage of its kinetic energy to the panel than a foam strike that occurs further back on the panel.

The report also explains that the strongest part of the panel tends to be at the apex near the tip where the dome like structure is the most pronounced.  The panel then becomes weaker as you move back from the apex towards the midpoint of the surface where the greatest damage is likely to occur.

 

 

Fig. OA-E1A-2

STS-107 Accident Investigation Final Report Vol. I, Pg. 55

Fig. OA-E1A-3

STS-107 Accident Investigation Final Report Vol. I, Pg. 55

This is just a brief description of how the RCC panels are fabricated.  It is only posted here for the interest of the reader.

Fig. OA-E1A-2 is a brief summary of design requirements for the RCC panels probably written near the beginning of the shuttle program.  Per the design requirements the RCC panels are to have the same resistance to kinetic energy impacts as do the TPS ceramic tiles.  The kinetic energy impact they must be able to survive has a value of 0.006 Ft.-Lbs.  One engineer studying this specification during the investigation discovered that the kinetic energy released by a regular wood No. 2 pencil dropped from a height of 6 in. would be equal to 0.006 Ft.-Lbs.  Because this value for the maximum allowable energy for impact is almost jokingly low, it might be deduced that the engineers responsible for it did not expect that an object of any significance would ever come near the leading edge of the shuttles wings.  The following statement was also part of the design requirements.

"...the wing leading edge would not need to withstand impact from debris or ice since these objects would not pose a threat during the launch phase."

When the Boeing engineers performing the 1ascent analysis on STS-107 needed a maximum allowable value for debris  impact. they referred to the statement that required the RCC panels to have the same value as the TPS ceramic tiles and the 2test report done on the ceramic tiles.  whether or not the new information on the RCC panels would have led to a different conclusion for those reports will require another complete analysis.

1. COL_DEBRIS_BOEING_030121, 030123, 030124

2. Orbiter Impact Tile Testing,
    SwRI Project No. 18-7503-005, March 5, 1999

 

RCC material properties:

If absolutely nothing else at all was accomplished in the creation of the STS-107 Accident Investigation Final Report Volumes I Through VI, the team testing the RCC panels established some significant guidelines that may provide a starting point for developing a true analytical method to predict impact damage to the panels.  This was quite an accomplishment given that they started from scratch with absolutely no material strength data available for the Reinforced Carbon Carbon and no loading data of any kind available for the panels themselves unless you count the specification for surviving a Kinetic Energy Impact of 0.006 Ft.-Lbs. presented above in Fig. OA-E1A-2.

The problems found in the impact testing that was done are detailed in the following sections, the areas of concern are the test methods used and the material properties.  Although the test methods as described in the STS-107 Accident Investigation Final Report Vol. II, D.12 are somewhat flawed, they provide a template that any follow up work may build on.  The allowable values for the material need to be independently verified and most likely determined using standard empirical methods for finding materials properties.

Figures E1A-4 through E1A-6 are from STS-107 Accident Investigation Final Report Vol. II, Appendix D.12, Pgs. 375 to 390

The strength values tabulated in Fig. OA-E1A-4 were created by the RCC panel testing team based on the strength properties of the raw materials used to makeup the RCC composite.

Fig. OA-E1A-4

The next step was attempting to develop an analytical solution that would predict the extent of damage to an RCC panel resulting from a debris impact.  The results of the analytical solution are represented in
Fig. OA-E1A-5 by equation 12 below.  Equation 12 calculates the maximum boundary stress seen by the RCC panel during impact.

Fig. OA-E1A-5

Fig. OA-E1A-6 is a graph of pressure Vs. Velocity for equation 12.  The units on either side are kPa on the left and psi on the right.

Fig. OA-E1A-6

 

The test procedures:

Fig. OA-E1A-7A

Fig. OA-E1A-7B

Fig. OA-E1A-7A shows the most likely scenario for a debris impact on RCC panel 8 during launch and ascent.  The final report states that the debris was traveling parallel to the orbiters body which makes the trajectory angle 0° as represented by the graphic.  The final report also indicates that the geometry or shape of the underside of an RCC panel would result in angles of incidence to the panel's surface that are at the most 20° and usually much less.

Fig. OA-E1A-7B shows the angles and impact locations that were used in the test that caused failure to a #8 RCC panel.  This is an extreme worst case scenario that probably would not and possibly could not ever happen during an actual launch ascent.

Structural Strength of the RCC Panel

As shown in Fig. OA-E1A-1 above the strength of an RCC panel changes as to move across it.  The strongest point is at the apex or the very front of the panel.  The weakest point occurs midway along the surface of the panel.  Both Fig. OA-E1A-7A and Fig. OA-E1A-7B upper and lower range of panel impact locations along with the point of greatest damage.

Changing the angle of the foams trajectory decreases its original kinetic energy.  If a 0° course change means the foam retains all of its original energy then a 90° course change would mean that all of the original kinetic energy has been lost.  Any values in between can be calculated using simple trigonometry.

 

Fig. OA-E1A-8

STS-107 Accident Investigation Final Report Vol. I, Pg. 79, Fig. 3.8-1

Fig. OA-E1A-8 is a photo of the Nitrogen gas gun at the Southwest Research Institute facility used to conduct the tests.  This gun is a much larger one than was used to do the original foam impact tests on tile samples for.

 

Fig. OA-E1A-9

STS-107 Accident Investigation Final Report Vol. I, Pg. 81

Fig. OA-E1A-10

STS-107 Accident Investigation Final Report Vol. I, Pg. 82

 Per the test procedures described in STS-107 Accident Investigation Final Report Vol. II, Appendix  D.12.  For the WLE panel impact tests both fiberglass panel sections from Enterprise were used along with real RCC panels from other orbiters that had a flight history similar to Columbia's.  The fiberglass panels were used to setup the test and determine exactly where the impact would occur.

Fig. OA-E1A-9 is the moment of impact on a section of fiberglass panels showing how the foam fragmented after impact.

Fig. OA-E1A-10 shows the hole made in an RCC panel 8 with a foam projectile at an angle of incidence of 25.1°

Higher resolution photos of RCC panel 8 after test.

STS-107 Accident Investigation Final Report Vol. I, Pg. 83

Reanalyze STS-107 using new data:

STS-107 can be reanalyzed using the analytical method, Equation 12 and graph developed in the final report with the added advantage of knowing that the official damage assessment placed the breach in RCC panel 8 of the left wing.  The final report uses the total velocity multiplied by the Sin of the angle of incidence to get the velocity component normal to the surface.  This analysis based on Equation 12 is stated to be more conservative than a computer modeled solution but might be used as a preliminary determination of whether more extensive analysis is required.

Allowable working stresses for the RCC panels are given in the final report as 29 ksi (200 MPa) for the panel rib and 27 ksi (185 MPa) for the panel face.  These values seem significantly lower than what would be expected but no other literature is currently available.

Stress values are computed for impact incidence angles ranging from 5 to 20° and a velocity of 774 Ft./ Sec.  The principle stresses found for angles of incidence 5, 10, and 20° are 34, 42 and 56 ksi respectively.  These values are conservative but fall short of the allowable working stress of 27 to 29 ksi.  Damage to the panel would have resulted in each case.

If  this calculation were done for the Columbia on STS-107 it would have shown that a more complete analysis should be done including computer modeling and visual inspection.

This Velocity Vs. Pressure graph along with Equation 12 may provide a quick and reliable method for determining the potential damage done to RCC panels after a debris strike.  It might be used to get a preliminary result before more time consuming and expensive methods are used.  As the equation and graph appear to work now with the allowable working stresses found in the final report, Almost every possible debris strike would be shown as significantly damaging a panel.  Even though the preliminary method needs to be, and has been shown to be, conservative, the current values will show damage for every event of a debris impact.  After 100+ shuttle missions this is known not to be the case.  Therefore the allowable values and or the results of the calculation need to be compared more closely to actual impact damage and modified to increase accuracy.

Summary / Conclusions:
The very nature of this test itself may be misleading.  What the test shows is the extent of damage done to an RCC panel when struck with foam debris at a given velocity and angle of incidence.  What the test does not show is the maximum possible impact velocity and angle of incidence of foam debris with some given initial aerodynamic properties on the leading edge of a wing.  The actual impact velocity on the RCC panel in question is then debatable based on how the aerodynamic properties of the debris change when it encounters the boundary layer of the wing shockwave.  Since there is no information on how the velocity and trajectory of foam debris may change when coming into contact with such a boundary layer this possible difference can only be considered for discussion purposes.

  1. It was stated in the description of the test procedures that both fiberglass WLE panels from Enterprise were used along with RCC panels from other shuttles that had similar flight histories to Columbia.  The report describes damage to both the fiberglass and RCC panels and It is often unclear which type of panel is being tested.  The fiberglass panels were designed and fabricated to look like the RCC panels for display purposes but were supposed to act only as a template for the actual test.  Anytime you have such an item in a test environment that does not provide any usable data for future design purposes it should be clearly marked as something other than a test specimen.  It does not appear anywhere in the test description that the fiberglass panels were marked in such a way as to separate them from the RCC panels and no different markings can be observed in any of the associated images.

  • Therefore damage done to any test article is questionable at best and data from these tests should not be used for any future purposes until more conclusive testing can be done.

  1. The "smoking gun" that was found during the impact tests was the large hole, 16" x 17", created on the lower side of an RCC panel 8, Fig. OA-E1A-10.  The test procedures stated that this test was performed with an angle of incidence of 25.1° to the lower surface of the panel.  One of the conclusions in the test report was that increasing the angle of incidence was one of the single greatest factors in increasing the amount of damage done to an RCC panel during an impact test.   Based on information within the official final report the trajectory angle of the foam during the actual event was most likely 0° and not more than 5° with the resulting angle of incidence being between 5° and 20°.  To achieve an angle of incidence of 25° requires that the trajectory angle of the foam debris to be between 15° and 20°.  The actual angle of incidence is dependent on exactly where along the lower surface of the RCC panel the impact occurs.  The impact tests may be conducted using any velocity and angle of incidence.  In a flight environment that represents the actual debris impact any change in the trajectory angle of the debris will result in a reduction of its kinetic energy and velocity.

  • The given parameters and test conditions for the result known as, "The Smoking Gun", were far too conservative to represent an event that occured durring the launch and ascent of Columbia on January 16, 2003.

  1. The debris strike video never clearly shows the piece of foam hitting the wing of the shuttle.  The piece of debris appears to disappear below the shuttle wing followed by the breakup and scattering of the foam debris.  Therefore the video evidence is at best inconclusive.

  1. Per STS-107 Accident Investigation Final Report Vol. I, Chap. 3, Pg. 60

[ Image analysis determined that the foam was moving almost parallel to the Orbiter's fuselage at impact, with about a five-degree angle upward toward the bottom of the wing and slight motion in the outboard direction. ]

Copyright © 2003, 2004 ColumbiasSacrifice.com

Official references for RCC panel testing