White Papers
Probabilistic Performance and Life-Cycle Metric Analysis for Liquid Rocket Engines Using Redtop-2 and ProbWorks
J. E. Bradford, A. Charania, and J. R. Olds
2003
Abstract
At the early stages of any engine design, there is a high degree of uncertainty due to a lack of supporting detailed component analysis. It is generally prohibitive in terms of time and cost to resolve these uncertainties via detailed simulation and testing. One approach that is relatively insensitive to these uncertainties is probabilistic, instead of deterministic, analysis. The ever-present factor of uncertainty, due to both controllable and uncontrollable factors, can be simulated to create a robust engine design. The Rocket Engine Design Tool for Optimal Performance-2 (REDTOP-2) is a code developed by SpaceWorks Engineering, Inc. (SEI) for use in the conceptual and preliminary stages of liquid rocket engine design and analysis. REDTOP-2 provides engine performance, power balance, weight, cost, and reliability, given only a minimal set of user inputs. Numerous engine cycles and configurations such as single vs. dual preburners, staged combustion vs. gas generator vs. expander cycles, or series vs. parallel turbine flow can be selected.
The equilibrium sub-model allows for the analysis of many common engine propellants, as well as the ability to incorporate new species by the user. Additionally, the engine can be sized by specifying a throat area, total mass flowrate, or a required thrust at an ambient condition. A probabilistic assessment of a conceptual liquid rocket engine design is performed using REDTOP-2. Two specific probabilistic methods are utilized for comparison: a direct Monte Carlo Simulation (MCS) and potentially faster Discrete Probability Optimal Matching Distributions (DPOMD) method. Statistical metrics such as mean, standard deviation, and certainty level are compared for engine metrics that include specific impulse, weight, exit area, length, and cost. Both of these assessments are performed using REDTOP-2 in Phoenix Integration's ModelCenter© collaborative design environment utilizing the SEI developed ProbWorks© suite of uncertainty analysis components.
Introduction
The discipline of propulsion is one of the most important when it comes to space launch vehicle design. Yet there are often confusing issues related to the best computational code to use to model such problems. Many companies and governmental organizations have their own proprietary propulsion codes. Those codes/programs that are readily available are often not the most updated, relying on older methods and programming languages. The reality within the conceptual design community is that there is a lack of modern, easy-to-use, reliable, fast acting, commercially available computational programs that can model both the performance and life cycle metrics (i.e. cost and safety) of liquid rocket engines.
Another issue at these early stages of engine design includes the high degree of uncertainty due to a lack of supporting detailed component analysis. It is generally prohibitive in terms of time and cost to resolve these uncertainties via detailed simulation and testing. One approach that can provide a more confident answer is probabilistic, instead of deterministic, analysis. The ever-present factor of uncertainty, due to both controllable and uncontrollable factors, can be simulated to create a robust engine design.
The metrics used to evaluate the impact of technologies on a conceptual engine or even Reusable Launch Vehicle(RLV) concept can be composed from various disciplines (i.e. performance, operations, cost, economics, safety and reliability) representing both a system’s technical feasibility and economic viability. Uncertainty, an ever-present character in the design process, can also be embraced through a probabilistic design environment [1]. The objective is to probabilistically quantify the impact of applied technologies on the output metrics of interest from the full design process, notionally referred to here as Probabilistic Data Assessment (PDA). Robust design methods such as PDA allow quantitative assessment of risk. Monte Carlo simulation techniques can be used to place uncertainty distributions on internal design parameters. The resultant outputs are cumulative and frequency probability distributions rather than simple deterministic values. Confidence intervals can be placed upon output metrics of interest to determine the 80% or 95% likelihood of meeting a target (e.g. Thrust-to-Weight, payload capability, gross weight, launch price).
Download The Complete Paper
(A one-time registration will be requested)
