Define the aerodynamic system
Begin by listing the surfaces exposed to the flow in the intended operating state. The shell, crew, runner assemblies, openings and component gaps may all contribute to the measured forces. A model that simplifies one of these elements can still be useful, but the simplification must be visible in the test record. Otherwise a result for a clean model may be misreported as a result for the complete competition sled.
The governing rules constrain the design space before optimisation begins. Check the applicable shell provisions in the current International Bobsleigh & Skeleton Federation documents. Treat a dimensional envelope and an aerodynamic objective as separate requirements. A geometrically attractive feature is not acceptable merely because a simulation suggests lower drag.
Published requirements: IBSF international rules and documents. Confirm the applicable discipline, edition and inspection procedure.
Choose a quantity that answers the question
Drag force is often the headline quantity, but the experiment needs a more precise definition. State the reference area, air density, flow speed and coordinate axes if reporting a coefficient. Without a common reference area, two drag coefficients cannot be compared directly. If the objective concerns stability, the relevant side forces and moments also need consideration rather than being hidden behind one drag number.
The familiar expression D = ½ρv²CDA is a way to organise a force measurement or model. It is not evidence that the coefficient remains constant across every speed, attitude and environmental condition. Reynolds number helps describe the relationship between inertial and viscous effects. Scale models therefore require a reasoned account of which similarities are preserved and which are not.
Make crew geometry repeatable
The crew is part of the external configuration. Head position, shoulder placement and the fit between bodies and cockpit openings change what the flow encounters. An aerodynamic comparison must document the representation used: a physical crew, a rigid surrogate or a digital approximation. Photographic references and defined positioning fixtures can reduce ambiguity, but they must not obstruct safe athlete movement.
A rigid surrogate improves repeatability without reproducing every aspect of a real crew. That trade-off should be explicit. Start with controlled geometry to isolate a shell change, then investigate whether the effect remains when realistic positioning variation is introduced. A design that is sensitive to small posture changes may require a different interpretation from one with a similar nominal result but a wider usable operating range.
Use simulation as a disciplined hypothesis
Computational fluid dynamics can expose pressure patterns and candidate flow structures that are difficult to observe directly. Its usefulness depends on the geometry, computational domain, mesh, boundary conditions, turbulence treatment and convergence checks. Save these settings with the result. A colourful surface plot without that record is an illustration, not an auditable design comparison.
Investigate mesh sensitivity and monitor the quantities that answer the engineering question. A residual history alone does not demonstrate that the predicted force is independent of the numerical setup. Consider yaw, nearby surfaces and the ground representation when relevant. The German Aerospace Center, DLR, publishes aerodynamic research; NASA’s Turbulence Modeling Resource provides model definitions and verification material. Neither source validates a particular bobsleigh design automatically.
Background and methods: German Aerospace Center (DLR); NASA Turbulence Modeling Resource.
Close the loop with physical evidence
A wind-tunnel programme needs a calibrated force balance, a documented mounting arrangement and an assessment of blockage and support interference. Establish the baseline repeatedly, including after removing and reinstalling the model. This exposes setup sensitivity that a single uninterrupted sequence can hide. Log air conditions and explain corrections rather than presenting corrected values without their derivation.
Compare simulation and experiment at matching conditions. Agreement in a single integrated force can conceal compensating errors, so use additional observations when practical. Surface pressure, flow visualisation or repeated configurations may help diagnose a discrepancy. Treat disagreements as information about the model or experiment, not as inconvenient data to omit.
The final shell dossier should connect each geometry revision to its manufacturing feasibility and inspection method. A feature that exists only in the digital surface cannot be credited to every moulded part. The next stage asks whether the workshop can reproduce the geometry, stiffness and interfaces assumed by the analysis.
