Olivier Guillermin
Director, Product and Market Strategy
Vistagy Inc.
Waltham, Mass.
www.vistagy.com
Circle 621

Acknowledgements:
Larry Pierce, senior applications engineer at MSC.Software, and Johan Grape, director of technology at Vistagy, contributed to this article.

Large and complex composite structures such as the Boeing 787, Airbus A350XWB, and Lockheed Orion CEV demand the design of lighter, less-costly, and better-performing assemblies. Unfortunately, analysts and designers usually work somewhat separately, focusing on their own domains rather than seeing the project from a more-comprehensive viewpoint. This approach makes it slow, difficult, and expensive to make changes after the first iteration. Product development in this case is largely a serial process, which doesn’t support shorter product cycles, less rework, or tight budgets.

The design of composite parts involves more unknowns and interdependencies than does that of metallic–part design. Serial product-development processes eliminate many chances to make complex adjustments. This negatively affects design benefits that are specific to composites, such as material directionality. Serial processes routinely inflate design allowances and safety factors, effectively treating composites as black aluminum and forgoing the benefits to be gained by the design of the material itself.

To help eliminate these problems, Vistagy’s FiberSIM composite-design software works directly with a variety of CAE packages, including SimXpert FEA software from MSC.Software, Santa Ana, Calif. This integration ties together the disciplines of composite-part design (CAD) and analysis (CAE), facilitating concurrent engineering from preliminary sizing through validation of final models with detailed ply-based parts definitions.

Structural composites have been used in aerospace for well over three decades but the volume production of large and complex assemblies is a recent development. Today’s focus is on overall structural and design optimization instead of traditional manufacturing concerns such as drapability, fiber deviation, residual stress and strain, and void formation.

There are many obstacles to effective collaboration between designers and analysts due to different domain knowledge, special techniques, and the use of different languages. Analysts think in terms of material properties, load cases, stress and strains. Designers work with ply coverage, nonstructural details, and design rules.

The common framework between FiberSIM and CAE software includes certain “touch points.” In composites, these comprise “zones,” which are built from loft surfaces provided by the team systems group and from material specifications and sizing data from the analysis group. Touch points are unlikely to change frequently or drastically and they represent logical information that can be used as the basis for shared concepts.

In the current development process, the designer provides the analyst with a definition based on the initial laminate specifications. The analyst maps this data onto the initial FE mesh of the part. The designer moves on to designing nonstructural wing elements, laying out transitions, detailing the design of drop-off areas, and preparing fasteners and inserts. The analyst applies physical properties to the meshed geometry (a critical consideration with composites) as well as loads and boundary conditions. Iterations that take place now involve concurrent data exchange between CAD and CAE.

The interface between CAD and CAE lets analysts directly use composite-design features such as system lines and zone partitioning to create and control a mesh for a composite skin. The interface also let analysts use lines of beams for stiffening elements such as stringers or frames in a fuselage section. In addition, the common access to native geometry exposes named attributes from CAD, which supports automated responses to design changes.

Another touch point for collaboration is the assignment of physical properties. The capability to seamlessly share detailed layup specifications helps the analyst’s efficiency and productivity and has a significant impact on a design’s accuracy and fidelity. The two programs working together supports everything from simple linear static to nonlinear buckling and progressive-failure analyses.

Optimization, too, should be multidisciplinary to account for all the composite structure’s performance requirements. There is more to designing a layup than simply cutting weight. The design must also meet multiple manufacturing constraints, design rules, and other requirements.

Here, the sharing of composite concepts across disciplines allows the seamless exchange of optimized layup data. Using a common geometry slashes the number of complicated dependency failures, while the logical relationship that persists between CAD and CAE removes the need for frequent, complicated refreshes.

In the new scenario, all changes flow from a constrained set of sources and allow for easy, automated remeshing in analysis as well as the automated translation and updating of designs. For example, the software automatically assigns new specifications to zones. Thus, increasing ply count or altering zone thickness triggers an automatic update that adds new ply drop-offs, while maintaining transition definitions, material choices, and detailed geometry.

With the detailed design almost finalized, the shared geometry further supports collaboration in the area of design validation. The analyst can access mesh control curves, which let him include the effects of ply drop-offs for precision meshing. Meanwhile, the framework maps the correct layup onto each finite element for detailed analysis.

Simulating the thermal loads the part will see during curing gives information on possible laminate spring-back “potato-chip” effects. This data can impact the final part shape and require the use of different tooling for the manufacturability of the part design. These functions close the feedback loop among design, analysis, and manufacturing.

It is obvious that bringing together design and analysis for composite structures has many benefits. First, the team can make modifications earlier in the development process, and it can accommodate changes late in the process to enhance optimization. Second, analysts can perform more-accurate analyses on the as-designed part definition using the true material properties.

Such concurrent engineering means shorter lead times and a parallel workflow that supports more and faster iterations. Both designers and analysts can continue working while synchronizing on significant changes. What’s more, the technique cuts the risks, program costs, and potential liabilities associated with the use of new materials and novel technologies.

Edited by Leslie Gordon