Methodology for the Generation of Cabin Load Data Bases by Numerical Crash Simulation Approaches

Kurzfassung

The potential for passenger (Pax) survivability during crashes of transport aircraft is influenced by the integrity of the fuselage and the cabin environment, the occurrence of fire, access to emergency exits, and the loads which are transmitted to the occupants during a crash sequence (trauma injuries and fatalities). Loads or accelerations of the cabin environment at attachment points of cabin furniture, seats and on occupants are normally not well known from accident investigations and reconstruction. However, they are fundamental information for further Pax safety improvements based on new cabin safety features such as airbags, new concepts for seat-/restraint systems, overhead bins, galleys, cabin dividing bulkheads, emergency exits and evacuation procedures. A systematic full-scale experimental access to cabin load data and accelerations during crashes is not feasible and is too costly. Therefore, numerical crash simulation methodologies with full-scale aircraft models offer promising capabilities. The knowledge of crash simulation methodologies for metal and composite A/C sub-structures have been improved and validated through FW3 and FW4 research projects such as “Crashworthiness of Commercial Aircraft” and CRASURV. Within these projects, a high confidence in numerical crash simulation of A/C sub-structures and full-scale structures has been established, in particular with typical aluminium designs. However, further methodology and tool improvements are needed in particular in the area of material and joint failure models, and models for impacted surfaces such as water. As modelling approach a so called local/global approach seems to be an effective way to generate full-scale A/C crash models for hybrid (multi body) codes such as KRWIN. Hybrid models comprise far fewer elements (mass points, beams, external springs) compared to full FE models, and even complex models require just a couple of hours of computation time. In the local/global approach the non-linear behaviour of representative structural components are analysed in detail with FE crash codes. The detailed FE-models include both, stiff fuselage sections such as the pressure bulkhead area (hard section), as well as more deformable sections, representative for other large parts of the fuselage (soft section). Those “macro-elements” are used as input data in the hybrid crash code. This methodology will be used to generate models for a commuter A/C, a typical single aisle A/C, a typical wide body A/C, and a double deck configuration. Figure 1 demonstrates the local/global modelling approach. In full-scale FE technique a two level approach is defined to perform the modelling of a typical single aisle A/C considering the different behaviour in the different areas of the aircraft. An existing coarse mesh such as those used for static FE full aircraft modelling is taken first. This model includes information related to mass distribution and to aerodynamic behaviour of the full aircraft, including lift and ground effects. Further CAD information is provided to develop a fine model of the impact areas of the A/C, using an appropriate methodology to optimise the size of the corresponding models and to guarantee an acceptable time step. For ditching simulations, the “Smooth Particle Hydrodynamics”-Method (SPH) is used in combination with a full FE model of a single aisle type A/C, also, for other aircraft a hybrid modelling technique in combination with a hydrodynamic potential theory is applied. Other new aspects concerning the simulation methodology not considered in previous projects are related to full scale 3D-hybrid models for the A/C fuselage and simulations of benchmark crash scenarios, comprising various impact surfaces such as water and slopes, and flight into obstacles. A major outcome of the simulation studies is the exploitation of a load database of the cabin environment of the different sizes and categories of A/C considered. The expected data and information comprise acceleration components (x, y, z direction) on the cabin/cargo floor (seat/cargo attachment points), acceleration components (x, y, z) at the cabin furniture attachments points (e.g. overhead bins, galleys etc.), resulting velocity changes and displacements, global exploitation of the crash sequence and structural deformation, loads/moments of structural components and potential areas of structural failure. With the simulated data of the cabin load environment, prediction of cabin furniture behaviour under crash loads will be established and pre-normative design criteria for the cabin will be developed with the aim to improve Pax survivability and reduce fatality rates in potentially survivable accidents.