Model-driven design and analysis for sustainable lightweight design

Biomimetic design has become an important approach in the quest for more sustainable composite products. However, the systematic transfer and application of materials, structures and principles from nature to products and industry requires a paradigm shift from constant efficiency optimization and maximum growth to sustainable strategies, long-term efficiency and robustness. In addition to new material and product properties, this also requires new production architectures and changes in product development, processes and planning. Model-driven design and analysis serves as a powerful tool to facilitate this transformation and enables controlled management of complex interdependencies along the life cycle of a product.

Achieving a balanced design

Lightweight construction aims to increase resource efficiency by reducing the mass of components. However, lightweight materials like fiber-reinforced polymers (FRPs) or high-performance metal alloys often have a relatively unfavorable ecological footprint in production and disposal. Aiming for a favorable overall balance, environmental accounting as part of design studies along the whole life cycle of a product is crucial, particularly in the manufacture, use and end-of-life (EOL) phases of a product (Fig. 1). However, because of complex dependencies between these phases, and the fact that each comes with a distinct set of parameters and optimization tasks for its evaluation, achieving a balanced design is not that simple.

Results of the manufacturing phase largely depend on an accurate prediction of material demand and the availability of valid manufacturing process models. For the use-phase model, contributions are determined by the mission scenario, service life and assumed overall system efficiency —consider, for example, an aircraft — as well as the allocation formula of the system’s energy consumption to its components. Contribution of the EOL phase hinges on the prospective disposal or recycling process and whether or not credits for recycled value streams are awarded.

To identify an optimal part design in terms of sustainability measures, all relevant design variants have to be evaluated. This becomes possible by the use of model-driven design synthesis of lightweight components in combination with life cycle assessment (LCA) and life cycle cost assessment (LCCA) tools. Specifically, the Fraunhofer Institute for Casting, Composite and Processing Technology (Fraunhofer IGCV, Augsburg, Germany) develops model-driven design workflows for lightweight design with the help of a graph-based design language implemented in the commercial software tool suite Design Cockpit 43, developed by the engineering firm IILS mbH (Leinfelden-Echterdingen Germany). The underlying design methodology is being developed at the University of Stuttgart (Stuttgart, Germany) and IILS.

Zeroing in on the best for each purpose

Considering an example in urban air mobility (UAM), a modular design program refines the component model step by step, and is driven by physical or heuristic design rules. A design task is specified through a set of functions and requirements, such as mechanical loads. A solution principle, such as a load-carrying beam, implements the specified functions. Thereafter, design steps like shape selection, material selection, sizing and detail design (production-specific) are executed automatically. Thanks to a modular architecture of the design sequence, many different design aspects can be implemented in each of these steps (Fig. 2). The design sequence is executed automatically, so that multiple variants can be generated without manual engineering efforts.

Each variant is analyzed with respect to three competing design goals of lightweight design:

• Minimization of structural mass to increase payload and energy storage capacity.
• Minimization of LCCs to ensure economic viability.
• Minimization of the product environmental footprint (PEF) to foster environmental sustainability.

While computing the structural mass from the CAD models’ volume and the presumed material density, LCC factors in material and manufacturing costs are calculated from different production models, as well as the cost of energy in the use phase. A more detailed model could also consider maintenance efforts and costs of disposal. The PEF is calculated with a simplified, forward-looking LCA using the software solutions GaBi (Sphera, Chicago, Ill., U.S.) and OpenLCA (GreenDelta, Berlin, Germany). Both modules — LCCA and LCA — rely on coordinated data with the same input parameter set containing design and process parameters.

Combining automated design and analysis, the characteristics of all generated design variants are compared in a Pareto analysis that estimates the benefit delivered by each action, then downselects the most effective actions. This helps engineers to identify variants that offer a favorable tradeoff between different criteria. This analysis is used to identify the most promising variants. For an in-depth analysis of a particular variant (Fig. 3), all design data — such as forces and moments, laminate stacking, CAD models, process chains, use models, etc. — are readily available.

 

Mastering the challenge

While the general approach is relatively straightforward and generates different designs and evaluates them automatically, there are some challenges. First and foremost, each model involved requires a set of parameters characterizing materials, processes, machines, etc., participating in the model. The discussed models currently process more than 50 independent design inputs and parameters, and the LCA and LCCA modules pull even more parameters from specific data sources. For every design variant computed, all of these parameters must be retrieved. However, not all of them are available at a sufficient level of confidence. Therefore, analysis of uncertainties can help to better understand the accuracy, precision and sensitivity of the computed product designs and evaluation results.

In order to gradually increase the accuracy and variability of the design and evaluation process, more and more parameters are needed, so that efficient data acquisition is key. This is where the presented method may profit from other digitalization and Industry 4.0 initiatives such as smart factories and digital twins that deliver large amounts of data and can be used to model the production and use phase more precisely.

 

Yielding the fruit

The design methodology presented here provides a framework for studying complex interdependencies in lightweight design with respect to the entire product life cycle. In this way, it is possible to evaluate various generated product variants with respect to the ecological and economic impact of applying bionic structures, bio-based materials or bio-inspired algorithms to a specific design problem. Thus, the presented approach and design tools make a significant contribution to biological transformation in manufacturing, which aims (by applying processes and principles from nature into technological systems) to generate sustainable added value and a resource-efficient economy, as well as minimizing waste production.

The resulting “biological” or “resource-saving” variants can be compared to conventional variants, in terms of sustainability and performance measures, without substantial manual engineering efforts, in order to verify whether they have a positive impact and meet the company’s requirements. This approach is transferrable to all kind of products and manufacturing processes in different sectors including commercial aviation, UAM and ground transportation. Furthermore, it is expandable for further optimization tasks, such as acoustic noise reduction and heat transfer optimization.

With this approach, one can identify numerous beneficial applications for sustainable lightweight structures — both as single components and assembled systems. This approach also enables engineering organizations to make informed decisions with respect to the myriad relevant dimensions of creating sustainable value in the designs they develop.