Combined Airframe and Subsystems Evolvability Exploration During Conceptual Design
Total Page:16
File Type:pdf, Size:1020Kb
COMBINED AIRFRAME AND SUBSYSTEMS EVOLVABILITY EXPLORATION DURING CONCEPTUAL DESIGN Albert S.J. van Heerden, Marin D. Guenov, Arturo Molina-Cristóbal, Atif Riaz, Yogesh Bile Cranfield University, Cranfield, MK43 0AL, United Kingdom Keywords: Evolvability, commonality assessment, airframe subsystems, conceptual design. Abstract Several systematic methods for designing for and exploring evolvability during conceptual de- Evolvable designs allow development costs of fu- sign have emerged over the past few years, such ture products to be lowered significantly. This as in Lim [2]. However, these usually require work contributes to aircraft conceptual design ‘change rules’ (or commonality) to be specified space exploration by enabling simultaneous com- a priory and either focus exclusively on the air- bined exploration of both airframe and subsystem frame (i.e. the wing, fuselage, empennage, and evolvability. This is made possible by combin- undercarriage) or the subsystems (such as the hy- ing existing subsystems-architecting and design draulic, pneumatic, and electrical power systems, space exploration techniques with novel com- the environmental control system, and so forth). monality assessment algorithms. The proposed In this paper, a framework is proposed that techniques are demonstrated via a simple single- promotes evolvability exploration of the com- aisle passenger aircraft evolvability study. bined airframe-subsystems unit. Particularly, it enables searching for commonality across the air- 1 Introduction frames and subsystems of pairs of input aircraft designs, rather than requiring this commonality Evolvability is a vital consideration during the to be ‘pre-specified’. An important contribution design of a new aircraft. It refers to the extent is that the framework allows input airframes and to which components, subsystems, and associ- subsystems pairs that have dissimilar architec- ated processes (e.g. manufacturing) of the design tures (i.e. having possible different constituent can be reused or changed with ‘little effort’ to be components and/or layouts of these components). incorporated into possible future products. That This makes it useful for exploring evolvability is, aircraft manufacturers often attempt to retain across conventional and novel configurations. as much commonality between previous, current, The paper constitutes an extension of the and future designs, while trying to prevent com- research presented in the doctoral dissertation promising performance unduly. The following of A.S.J van Heerden [3] and is organised as definition for evolvability is adopted here: follows: This introduction is followed in Sec- tion2 by a discussion on designing for evolv- Evolvability: The ability of a system de- ability in conceptual design. In Section3, the sign to be inherited and changed across combined airframe-subsystems evolvability ex- generations (over time) [1]. ploration framework is introduced, followed by The purpose of designing for evolvability is to a demonstration of how it could be used (Section (hopefully) foster shorter development times and 4). Conclusions are presented in Section5. lower the development cost of subsequent gener- ations of the design, amongst others. 1 VAN HEERDEN, GUENOV, MOLINA-CRISTÓBAL, RIAZ, BILE 2 Background as some components/parts would be positioned further along the production ‘learning curve’ if In this section, a brief overview of important con- those components/parts are already being manu- cepts regarding designing for evolvability is pro- factured for the baseline [4]. vided. First, the concepts of evolvability and Commonality is therefore related to the re- commonality are described in further detail, fol- duction of redesign effort for a new descendant. lowed by a discussion on the design activity of To quantify this benefit, either a commonality evolvability exploration. score could be employed (which is only valid for comparison when the same baseline is used) or 2.1 Evolvability and Commonality the development and manufacturing cost reduc- tion could be estimated. A commonality metric Whether it was planned or not, many aircraft de- (also called commonality ‘score’ or ‘index’) is a signs (both military and civil) have undergone quantitative measure that aids in product evolv- significant evolution. For selected case studies ability related decision-making [6]. Many such regarding the evolvability of military aircraft, the metrics have been proposed and, for an overview, reader is referred to Lim [2]. Many civil pas- the reader is directed to Pirmoradi et al. [7]. Sim- senger aircraft have also been subject to substan- ple mass and cost weighted metrics are used in tial redesign to meet new requirements and re- this paper (see Section3). main competitive. An exemplary case is the Boe- ing 737 design, which has twice been upgraded 2.2 Evolvability Exploration with new engines; undergone substantial changes to increase wing area (through chord increases, The term ‘evolvability exploration’ is used here larger control surfaces, and tip extensions); had to refer to the activity of searching for aircraft de- ‘plugs’ inserted to increase fuselage length; and signs that appear promising for the ‘near-future’ received extensions to the empennage surfaces to entry-into-service (EIS) timeframe and could be increase surface area, amongst many others. changed with relatively little redesign effort to Despite these changes across the generations future designs that provide value in the possible of 737s, much of the original airframe design ‘far-future’ timeframes. has been re-used. This reuse of design features One of the pre-eminent methods for de- (as well as the associated processes) on later sign space exploration (particularly for evolv- generations of a design implies ‘commonality’ ability exploration) is ‘multi-attribute tradespace with the baseline design. According to Boas exploration’ (MATE). MATE is an approach [4], commonality is the “sharing of components, in which formal decision theory (particularly processes, technologies, interfaces and/or infras- multi-attribute utility theory) is incorporated into tructure across a product family” . It provides model- and simulation-based design [8]. The many life cycle benefits [5], but the focus here is multi-attribute tradespace is a two-dimensional on its potential to reduce development time and tradespace in which the ordinate represents cost (often referred to as Research, Development, multi-attribute utility and the abscissa cost. This Test, and Evaluation (RDT & E) costs, as well as plot is populated by all the enumerated designs, manufacturing cost and time. Development time such that each design is represented by a util- and cost are reduced, because the total ‘develop- ity/cost point. This enables the decision-makers ment scope’ is reduced [4]. Furthermore, devel- to view the value and cost of all the designs under opment of the first product is generally expected consideration in a single plot, regardless of the to cost more than if no commonality was planned, different subsystems architectures and configura- whereas subsequent variants are expected to cost tions the designs may embody. less [4]. Manufacturing cost is also expected to The temporal nature of the environment in be lower for the first unit of a subsequent variant, which complex systems operate can be accounted 2 COMBINED AIRFRAME AND SUBSYSTEMS EVOLVABILITY EXPLORATION for by combining MATE with a process called niques that could automatically: ‘Epoch-Era Analysis’ (EEA) [9]. In EEA, the • determine which major airframe compo- full life cycle is referred to as the ‘era’, whereas nents across two different aircraft are of the epochs refer to periods of time within that life same type and connected to other compo- cycle in which the context (factors exogenous to nents in the same way, such that they could the system) remain ‘constant’ and the system pro- be send for more detailed similarity assess- vides fixed value [9]. Each epoch is subsequently ment; and characterised by “static constraints, available de- • identify similar segments (based on user sign concepts, available technology, and articu- criteria) across pairs of components that lated attributes” [9]. Each epoch can therefore have complex geometries, complex mass be represented as a separate MAT. As a new distributions, several attachments to other epoch dawns, the changes in requirements, avail- components, and so forth. able technologies, regulatory environments, and Such techniques were developed in Ref [3]. so forth, could have an effect on the value of the However, it would also be beneficial if the system and its utility may increase or decrease. techniques could assess commonality across air- Several authors have applied a combination of frames and subsystems in a combined fashion. MATE and EEA specifically to the exploration of This is because subsystems are being increas- evolvability in complex systems (see for example ingly studied earlier in the overall aircraft design Refs [10], [11], and [12]. process (see for example Bile et al. [15]). Ex- Most evolvability exploration techniques tending the techniques from Ref [3] to provide usually assume that the change rules (change a combined airframe-subsystems is the subject of mechanisms) between different designs are al- this paper and the combined approach is provided ready specified. However, it would be beneficial next. if techniques were available that could automati- cally determine which components/parts could be 3 Framework common or