Until relatively recently, this has involved traditional methods relying on engineering experience and progressive evolution. For a long time, this produced satisfactory solutions for frequently manufactured components. Periodic challenges such as revised crash test requirements would cause some increase in the composite design department’s grey hair content, but experiments and experience produced solutions that fit the brief.
However, the quest for ever smaller performance gains inevitably pushes engineers to develop improved techniques to include maximum stiffness uniquely and deal with multiple load cases. So, can Formula 1 front wings have adequate beam stiffness to meet the FIA’s static vertical load test applied to the endplates’ top edge and retain aerodynamically beneficial flexibility in different directions or modes?
Design and optimisation software company GRM used their OptiAssist software with the GENESIS-based composite optimisation solution to investigate how Formula 1 engineers might be able to design a flexing wing that still meets the FIA deflection criteria. The first step was to generate a basic CAD model of a Formula 1-type front wing.
Next, the initial candidate ply lay-up was stipulated. This would typically involve outer and inner plies of woven carbon fabric orientated at 0 degrees, encapsulating a variable number of uni-directional fabric plies orientated at 0 and +/- 45 degrees. However, before GRM could do any analysis and optimisation, it was necessary to define the analytical constraints.
In this instance, the FIA’s static load test requirement (less than 20mm deflection with a 100kg load applied vertically to one or both ends of the wing). The wing could twist back at the outer ends under aerodynamic load to reduce the wing’s angle of attack at speed. Indeed, the objective of this exercise was to maximise the displacement of the wing’s twist under aerodynamic load.
Next, GRM ran automatic optimisation to generate multiple variations in the lay-up and considered the orientation of the plies, balancing absolute performance versus manufacturing complexity. The finite element technique used here is an element-by-element approach where the thickness of each individual ply of each wing element is optimised.
The downside of this technique in a composite application is that there is only one design variable for each ply per element. Solutions found here would be impractical in manufacturability as composite laminators would not appreciate cutting and laying up vast numbers of small squares of carbon or highly complex-shaped plies. Also, composite structures benefit from comprising long uninterrupted lengths of fibres.
Consequently, a ‘lay-up variation constraint’ was applied, which produced successively fewer plies to assist the manufacturing process. GRM calculated the resulting properties in terms of the aerodynamic deflection for each lay-up variation. This generates an informed selection of the optimal solutions that met the deflection criteria and were also practical to
manufacture. GRM notes that the optimisation also enabled an aerodynamic load failure index to be applied, which was essentially a strength constraint to ensure the wing would not fail when subject to a specified amount of rotational flex.
OptiAssist automatically generates ply patterns assets of data that GRM could import into the lay-up definition packages popular in the industry, such as Anaglyph’s Laminate Tools, Simulia Abaqus or BETA CAE Systems’ ANSA. At this stage in the sophistication of optimisation software, it is still necessary to manually edit ply boundaries to generate a fully manufacturable lay-up. GRM admits that this can be a time-consuming step and states that generating ply patterns from scratch can take a similarly long time. This is a part of the optimisation process that is under continual development.
The next step was to use OptiAssist to ‘drape’ (in the virtual environment) the plies onto the mould. This step considers how the complex curvature of an object such as today’s Formula 1 front wings affects how the fibres will actually lie in mould and hence in the finished product. While a skilled laminator can account for this during the physical lay-up process and minimise distortions, there can be deviations from the supposed fibre orientations, which could affect the end product’s performance.
See results in Table 1.
Table 1 – aerodynamic and ‘FIA’ displacement optimisation results
The optimisation results show the aerodynamic displacement measured at the rear of the front wing (the blue section being the fixed datum point on the nose turning to red outside the structural limit). It represents the outer sections twisting back under aerodynamic load a significant amount, which could yield substantial performance gains while having the beam stiffness required to pass the FIA static load test.
This test shows that a computational optimisation tool like GRM’s OptiAssist highlights significant scope in the Formula 1 rulebook for flexible wings without a great deal of time, cost, or guesswork required for carrying out digital trials. Undoubtedly, Formula 1 teams carry out physical lay-up and verification testing, but far fewer trials would be required to achieve the desired result using such a piece of software. This process is equally applicable to other primary aerodynamic structures such as floors, barge boards and rear wing.