Multi-criteria Evaluation Of Aerospace Vehicle Conceptual Design

Modern aerospace system conceptual design aims at choosing the most promising designconcept at first time and avoiding the costly design changes at a later stage. This requires system’stechnical performance parameters as well the overall program related issues such as cost, schedule,reliability and manufacturability etc. to be taken into account as early as at the conceptual designphase. In view of the numerous design and evaluation criteria which are often conflicting in nature,the final design concept is the result of a compromise solution. This is in contrast to the traditionalapproach of single objective of maximizing the performance or minimizing the cost. The designersdealing with conceptual design and evaluation studies are thus involved in balancing the multipleand potentially conflicting criteria involving diverse disciplines. Thus, one could state withconfidence that modern aerospace conceptual design is a multi-criteria decision making activity.The objective of the present thesis is to improve the process of aerospace vehicle conceptualdesign and decision making. To achieve this objective, this thesis presents a framework forconceptual design and multi-criteria evaluation of space launch vehicle. The proposed framework iscomprised of morphological matrix method to expedite the brain storming process by decomposingthe overall system function into sub-functions at lower levels of abstraction. Solutions to these sub-functions are then sought and synthesized together thus resulting in numerous candidate designconcepts. These concepts are then screened for compatibility in an automated way. The Tsiolkovskyideal velocity rocket equation is adopted to size the candidate space launch vehicle design conceptsfor required mission objectives. This has avoided the use of trajectory simulations based sizingapproaches which require precise information and data usually not available at the conceptual designphase. The use of Tsiolkovsky ideal velocity rocket equation supported by the proposed mass andpropulsion modeling approaches has improved the capability of the conceptual design process toestablish the performance trade space for a number of design concepts in an efficient manner.Reasonable estimates for the associated losses in velocity are introduced to result in more realisticcalculations. In addition to the technical performance parameters, certain program related issues suchas operability, flexibility, manufacturability and the cost related parameters are also calculated toserve as evaluation criteria. The qualitative judgment of program related issues is converted intonumerical values using fuzzy set theory. This is to address the prevailing subjectiveness inqualitative criteria rating and allowing natural language rules to be included in mathematicalcomputations. The cost modeling approach is based on the parametric cost estimating relationships which aredeveloped from the statistical data of similar systems completed in the past by correlating it with thebest fit curves. The cost items of space launch vehicle are categorized in design, development,testing and engineering cost and the theoretical first unit cost. This offers the flexibility ofcalculating the relevant cost items depending upon the scope of the study.From the relevant criteria values, it is observed that certain candidate design concepts performwell on certain criteria but performs bad on certain other criteria and there is no single designconcept which can be selected as best while simultaneously considering all criteria. This exhibits thetypical nature of multi-criteria decision analysis problem. An exhaustive review of the relevantliterature illustrated that multi-criteria decision analysis methods are pertinent to analyze thetradeoffs between various evaluation criteria and thus is a useful tool to support the space launchvehicle concept selection decisions. Hence, based on the performance parameters, program relatedissues and cost data, the candidate design concepts are subjected to multi-criteria decision analysis.This has ranked the design concepts and identified the most promising ones for furtherconsiderations. The robustness of the multi-criteria decision analysis is demonstrated with thesensitivity analysis. A comparative study is also conducted to observe the effect on the final rankingdue to the changing weight allocation and multi-criteria decision analysis technique. It is concludedthat for the space launch vehicle concept selection problem of present thesis, the subjective-objectiveweight allocation methods are more suitable than the purely objective weight method. This isbecause the objective weight method only solicits criteria weight from the decision matrix where theweight is assigned as a reference to the scatter in input data which may not be a true representationof the actual problem and the preferences of the decision maker. While the subjective-objectiveweight methods offer better chances of reflecting the criteria importance according to the nature ofthe decision problem and preferences of the stake holders by taking input from the decision maker.The innovation of the present thesis is the proposed framework which can be used for candidatedown-selection with an automated compatibility check when the morphological analysis results inhundreds of system concepts. Its successful application in conceptual design and evaluation of spacelaunch vehicle has demonstrated its utility in the early phase of the aerospace system design anddecision making.