A historical survey with success and maturity estimates of launch systems with RL10 upper stage engines

Abstract

Pratt & Whitney's RL10 engine line has a long and rich history, beginning in 1958 and continuing today. This paper provides a historical summary of launch vehicles using RL10 engine derivatives dating from 1962-2005. The historical launch data is used to derive baseline launch success rates and growth curves for vehicles configured with RL10 engines in the upper stage. Because it was the first liquid hydrogen fueled rocket engine, the RL10 engine launch history provides a unique opportunity to investigate the maturity trends for revolutionary new complex systems. All of the data used in this survey was acquired through publicly-available sources [1-21]. In all, 190 vehicles configured with RL10 upper stage engines were launched between 1962 and 2005. There were 12 upper stage failures that either failed to reach orbit, or reached a lower, unintended orbit. The early failures were dominated by knowledge gaps in system interactions and operational flight conditions. There is a clear trend of early development growth with an eventual plateau as system knowledge improved as a result of flight experience and more thorough test programs. Failures due to process-based issues (fabrication techniques, quality control, etc.), however, do not appear to exhibit maturity growth. Eventually, as the knowledge-based failures are removed, these process-based failures become the dominant risk driver. Vehicles that use mature, highly-reliable components are still vulnerable to process or functional changes, and failures of this type occur fairly uniformly with flight experience. In order to improve future reliability estimates for such systems, it is important to understand the trends and relationship between the knowledge-based and process-based issues, and determine which class of issues currently dominates. It should be noted that of the 12 upper stage failures, only one was caused by a defective part.

How safe must a potential crewed launcher be demonstrated to be before it is crewed?

“…When a more specific subset of the data is investigated (Go, 2008) the situation is a bit better. As can be seen in Fig.…”

“…Most recently the classic problem of reliability growth has been addressed by a number of authors. These investigations have addressed the general problem and the specific problems as they were observed in the aerospace industry both in aircraft (Walls, Quigley, & Marshall, 2006) and in space launcher stages (Go, 2008). Immaturity becomes a particular problem as systems become more complex.…”

“…Fig. 1 (Go, 2008) shows a plot of the historical instantaneous failure rates of a set of launch vehicles and stages fit to a power law. A power law is traditionally the term used for any function that represents ''diminishing returns'' where the function diminishes as a power function of the independent variable and thus would plot as a straight line on log-log paper with a negative slope.…”

Physical simulation in space launcher engine risk assessment

“…Very often, it is observed that the most critical failures in a liquid propulsion system are manifested at the interfaces between elements of the integrated vehicle [15]. Understanding the physics of the rocket engine allows one to investigate the integrated propulsion system to identify system level synergistic characteristics.…”

Reliability growth and the caveats of averaging: A Centaur case study

“…All Centaur historical data in this paper results from a combined analysis of the data provided by [8], [9], [10], [11] and [12].…”