Special section: New Problem Formulations for Design Under Uncertainty

Design of Long-Endurance Systems With Inherent Robustness to Partial Failures During Operations

[+] Author and Article Information
Jeremy Agte1

 Department of Aeronautics and Astronautics, Air Force Institute of Technology, Wright-Patterson AFB, OH 45433jeremy.agte@afit.edu

Nicholas Borer

Systems Design Engineer  The Charles Stark Draper Laboratory, Cambridge, MA 02139nborer@draper.com

Olivier de Weck

 Department of Aeronautics and Astronautics, Engineering Systems Division, Massachussetts Institute of Technology, Cambridge, MA, 02139deweck@mit.edu

Innate here refers to the system’s characteristics from first “creation” or initial design, as opposed to modifications or improvements applied at some later stage of its development process.

If the Markov chain has been simplified due to failure symmetry (as in Fig. 3), the members excluded from Gs,t are those in which all of the arrival paths come from states that have already been classified as unacceptable.

Total thrust, T, is equal to drag in straight, level flight. Thus, for a given flight condition, the moment created by the differential thrust is the only thrust related unknown.


Corresponding author.

J. Mech. Des 134(10), 100903 (Sep 28, 2012) (15 pages) doi:10.1115/1.4007574 History: Received January 19, 2012; Revised June 18, 2012; Published September 21, 2012; Online September 28, 2012

This article presents an integrated multistate method for the early-phase design of inherently robust systems; namely, those capable, as a prima facie quality, of maintaining adequate performance in the face of probabilistic system events or failures. The methodology merges integrated multidisciplinary analysis techniques for system design with behavioral-Markov analysis methods used to define probabilistic metrics such as reliability and availability. The result is a multistate approach that concurrently manipulates design variables and component failure rates to better identify key features for an inherently robust system. This methodology is demonstrated on the design of a long-endurance unmanned aerial vehicle for a three-month ice surveillance mission over Antarctica. The vehicle is designed using the multistate methodology and then compared to a baseline design created for the best performance under nominal conditions. Results demonstrated an improvement of 10–11% in system availability over this period with minimal impacts on cost or performance.

Copyright © 2012 by American Society of Mechanical Engineers
Your Session has timed out. Please sign back in to continue.



Grahic Jump Location
Figure 1

Increasing probability of failure with longer mission durations; Technology Level (TL) 1 represents current civil aviation component failure rates (actuators—1/500,000 h, engines—1/125,000 h); TL 2 represents current UAV rates (actuators—1/10,000 h, engines—1/100,000 h); TL 3 represents past UAV rates (actuators—1/3000 h, engines—1/50,000 h); y-axis zoomed in on (a) for readability

Grahic Jump Location
Figure 2

Markov chain formulation for a nonsymmetric three-element system

Grahic Jump Location
Figure 3

Markov chain formulation for a symmetric three-element system

Grahic Jump Location
Figure 4

Two-engine failure scenario

Grahic Jump Location
Figure 5

Multistate analysis and design process. Solid arrows represent process flow. Dashed arrows represent internal iterations. Letters and symbols along arrows represent variable flow.

Grahic Jump Location
Figure 6

Iterative processes in UAV performance model

Grahic Jump Location
Figure 7

Design performance space for nominal endurance versus A/C flyaway cost; this shows the initial baseline design very near to the Pareto front for optimal cost and nominal endurance. Aircraft geometry is as one might expect for a design based on nominal performance, with high aspect ratio and low wing sweep.

Grahic Jump Location
Figure 8

Design performance space for availability versus cost—varying only λs; Here, the design space for availability and cost is constructed taking the initial baseline geometry and varying only the system’s component failure rates. Geometry shown is that of the initial baseline. Aircraft geometry is the same for each of the design points since only failure rates are varied.

Grahic Jump Location
Figure 9

Design performance space for availability versus cost—varying only x ; The design space for availability and cost is constructed by keeping failure rates constant and varying only the system’s static design variables. Maximum system availability is nearly the same as that accomplished by varying only failure rates. The resulting aircraft geometry is characterized by relatively high wing sweep, a larger vertical tail, and increased engine power.

Grahic Jump Location
Figure 10

Design performance space for availability versus cost—full multistate, λs and x ; This shows the design space for availability and cost that results from allowing variation of both failure rates and static design variables. The resulting design exploits both groups of variables to improve availability, replacing extra wing sweep and tail size with more cost effective improvements in component failure rates.

Grahic Jump Location
Figure 11

Design performance space overlays for availability versus cost. An overlay of three design spaces shows that full multistate design allows a region of availability and cost to be reached that is unobtainable via the independent variation of either static design variables or component failure rates.

Grahic Jump Location
Figure 12

Design comparison via Pareto nondominated points; Circled design points show the location of two optimal design geometries in three separate design spaces. Design A favors higher expected performance over system availability and lower cost. Design B favors higher availability and lower cost over expected performance.

Grahic Jump Location
Figure 13

UAV model stability and control sign convention

Grahic Jump Location
Figure 14

Variation of best cruise altitude (at end of cruise-climb) with throttle setting for aircraft with failed inboard engine and rudder. For straight and level flight, trim drag, Dtrim, must equal thrust, T. To the right of maximum, reduction in throttle of the operating engines is accompanied by a reduction in thrust lower in magnitude than the resulting decrease in trim drag, thus improvement is accomplished and the aircraft climbs. Once the maximum is passed, moving from right to left on the chart, any benefits in decreasing throttle are not realized as the loss in thrust is greater than the reduction in trim drag. Thus, the aircraft must descend until the two are again equal.

Grahic Jump Location
Figure 15

Comparison of reliability costs for various components, normalized by averaged cost and MTBF of medium unit



Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In