Documents / Report

Department of the Air Force Report, 1996

U.S. Department of War · 1996-09-10 · 181 pages · text from the file's own layer

This final report, dated September 10, 1996, was prepared by the Research Triangle Institute for the Department of the Air Force's 45th and 30th Space Wing safety offices. Titled Modeling Unlikely Space-Booster Failures in Risk Calculations, it shows how rare Mode-5 failures, in which a rocket veers well off its intended flight line, are modeled in the DAMP risk-analysis program. An appendix lists Atlas, Delta, Titan and Thor launch and failure histories through August 1996.

From the source:Release of 2026-05-08 Incident: 9/10/96, N/A. This report describes the Modeling of Unlikely Space-Booster Failures in Risk Calculations, documenting historical launch failure modes and recommending corrective actions to address them using novel modelling techniques.

density function. Since this could not be done in general, impacts from only the two
types of malfunction turns were considered. Several factors affect the results of the
simulations:
a. Weighting of tum data: Both random-attitude and slow-tum. simulations were
made for Atlas HAS. In combining impacts from the two data sets, random-
attitude turns were assumed to be three times as likely to occur as slow turns. A
factor of three was selected· since, among the Mode-5 failure responses in the
performance summaries for Atlas, Delta, and Titan, random-attitude turns
appeared to occur about three times as often as slow turns. In many cases, lack of
detailed information made it difficult to· decide whether a Mode-5 response
should be considered as a random-attitude tum, a slow tum, or some other type
of failure. The relative weighting of turns makes little difference, however, since
the impact distribution for the two types of turns are similar (as shown later in
Figure 5), and since the weighted composite must lie between the two. It was
assumed that similar results would be obtained for Delta, Titan, and LCVl, so
slow-turn computations were not made for these vehicles, cutting the number of
time-consuming simulations in half.
b. Breakup qa: In the tum calculations, the assumption was made that vehicle
breakup would occur if a certain value of qa. was reached~ In addition to the no-
breakup case which is considered unrealistic, separate runs were made for three
constant values of qa: 5,000, 10,000, and 20,000 deg-lb/ft2. As stated previously,
the determination of vehicle breakup is, in reality, much more involved than this
simplistic approach would suggest. However, to add realism to the malfunction-
tum calculations, use of a simple approach seemed better than none at all. For
Titan IV, allowable (but not breakup) qa.'s were provided as functions of Mach
number. The maximum permissible value and corresponding Mach number for
Titan/Centaur, Titan/NUS~ and Titan/lUS were, respectively, 6819 deflb/ft2 at
Mach No. 0.77, 5332 deg-lb/ft2 at Mach No. 0.815, and 17,000 deg-lb/ft at Mach
No. 0.325. For Atlas, Delta, and LLVl vehicles, no breakup qa. data were
available. The breakup qa.'s used in the calculations bracket the range of
permissible qa.'s for the Titan vehicles.
c. End time T5: The simulated impact distributions from random-attitude failures
and slow turns were compared with impact distributions computed from the
Mode-5 theoretical impact-density function. For the comparisons to be
meaningful, the value selected for T5 in the Mode-5 impact-density equation and
the stop time for thrusting-turn simulations must be the same. To some extent,
the shaping constants A and B derived by fitting the theoretical and simulated
impact data depend on TJY since the percentage of impacts in each 5° sector
depends on TB. However, after A and B have been established for a particular TJY
using a different TB in the DAMP calculations has no effect on computed risks
provided an adjustment is made in the probability of occurrence of a Mode-5
9/10/96 34 RTI

Not linked to a story yet.

About this file

Report, from the pursue collection. The PDF is mirrored here; the original link is above. 181 pages are in the text index: search them above, or from the library's search.