Coupled crash mechanics and biomechanics of aircraft structures and passengers

Highlights

• Have you ever seen the eigenmodes of vibration of a seated human body? Here, we show how they look and the possible associated injuries from the viewpoints of biomechanics.

• Aircraft structural damage and passenger (dummy) injuries in several different crash scenarios can be clearly visualized in the video animations from our supercomputer simulations. They will add intuition and insights to researchers and engineers for future safety designs of aircraft.

• Our work has successfully validated the experimental drop test results as well as basic structural properties of a ring-stiffened thin-shelled structure.

Abstract

The DYCAST (Dynamic Crash Analysis of Structures) experiments that started at NASA Langley Research Center during the late 1970s have greatly influenced the methodology and thinking of aircraft crashworthiness and survivability studies, and was continued and refined at other aerospace establishments. Nevertheless, so far most of the existing work has emphasized the impact damage to the aircraft section. Issues related to potential passenger injuries have not been properly addressed in the literature, to the best of our knowledge. Here, we study the DYCAST problem integrally by treating and combining impact damage and passenger injuries altogether. We develop the biomechanics by way of modal analysis of passenger dummy motions coupled with the vibration of aircraft structures in order to understand their basic interactions. Two types of mechanical dummies are used in this study. Such a modal analysis can help identify basic injury types, but is valid only in the constructed models, linear regime. However, we are able to extend the linear elastic model to a nonlinear elastoplastic computational model by using the versatile software LS-DYNA as the platform. Computer simulations are carried out on the supercomputer clusters and the numerical results are rendered into video animations for visualization and analysis. One can see, for example, how the passenger-dummy interactive motions with the fuselage and fixtures and the potential injuries caused in the event of general aircraft crashes on a fractal domain.

Introduction

Aircraft crashworthiness and survivability are always top concerns for airplane designers, travelers and transportation regulation agencies. According to some estimates [1], at any given time, there are some 500,000 people in the air flying. Air travel safety has always been rated as the best among all modes of travel because of the long term investments made by the governments of the world and by the time and efforts of numerous researchers on aircraft safety and structural integrity issues.

Even though there is a long history of aircraft structural safety studies, systematic, integrated research on aircraft structural crashworthiness and integrity modeling and computation did not begin until the work of E.L. Fasanella and his collaborators at the IDRF (Impact Dynamics Research Facility) in the NASA Langley Research Center in Hampton, Virginia during the late 1970s [2], [3], [4]. The Project was designated DYCAST (dynamic crash analysis of structures). At IDRF, a section of a retired Boeing 707 aircraft was dropped to the ground from a control tower about $55ft$ high, with dummies inside. See Fig. 1. Structural damage was carefully documented, and a finite element model was built to explain the experimentally observed damage. In an early paper by Fasanella, et al. [2], only 300 pieces of finite elements were used.

The work by Fasanella and his collaborators (see [2], [3], [4]) established initial benchmarks and has guided the directions of future studies on crash damage modeling and computation for aircraft structures. This work has been continued and refined in academia and the aerospace industry in many countries. We may mention the various more recent papers by researchers in Japan [5], [6], [7], China [8], [9] and the European Union [10], [11], for example. In 2020 Paz et al. [12] proposed enhancement of crashworthiness of commercial aircraft structure by crushable energy absorbers for struts and studied them for a vertical drop scenario. See also the references [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28] for other designs, optimization and composite materials related to the drop test and crashworthiness studies. The single section of an aircraft in the DYCAST experiment consists of the following components: fuselage, ring stiffeners, stringers, seats, passengers/dummies, floor, struts, safety belts, overhead cabins, plus other possible minor items. These components interact with each other through coupling or interface conditions. Thus, there is a great deal of structural complexity in the work of model building. In turn, each component is made of different materials and has varying geometries, yielding diverse dynamic responses in a crash environment. Computationally, there is a large memory requirement in order to achieve the desired accuracy through increasingly fine resolutions. Such computations normally exceed the capacity of a desktop or small workstation and necessarily must be implemented on a supercomputer. These constitute major challenges for the present study.

In the earlier studies cited in the preceding subsection, the emphasis of the work was on aircraft structural integrity and damage after the crash tests. Little work has been addressed to the study of passenger safety. Passengers’ injuries and fatalities can happen due to fires, high-G forces, blunt force trauma and bodily penetration by flying projectiles. The types of injuries suffered by the victims depend on various factors of the air accident, such as the positions of the passengers, whether seat belts were on, the type of collision, the intensity of the impact, etc. As there are so many variables, there are many scenarios. There is obviously a large void to be filled in regarding such a passenger injury study. Yet one thing that is quite certain is that such a study largely depends on the interactions between the passenger and the aircraft and fixtures. Here, our treatment does not include the effects of fires.

We study this passenger-aircraft interaction, mathematically speaking, by the method of vibration modal analysis of an aircraft fuselage section furnished with passenger dummies. The aircraft section is designed in the same spirit as the DYCAST experiment. Such a modal analysis is commonly regarded as fundamental in the analysis of most mechanical vibrations. Modal analysis is the determination of natural frequencies and mode shapes of a structure. Usually, damping is neglected as a first step in performing the dynamic analysis. Using the natural frequencies and mode shapes, one can extract characteristic properties of the structural dynamics and understand the response of the system subject to dynamic loading and disturbance. In biomechanics/bioengineering, the use of modal analysis is not new. We mention several representative articles [29], [30], [31], [32], see also the references therein, as examples, where the biomechanics of human body movements is an important subject originating from the applications in ergonomics, kinesiology, sports medicine, aging, vehicle transport comfort, etc. In addition, in civil and mechanical engineering, a prominent application of modal analysis is the tuned mass damper (TMD) [33] that can help damp out vibrations in tall skyscrapers due to wind or seismic effects. TMD application can prevent discomfort, damage, or outright structural failure [33]. In aerospace engineering, for example, in Dimitrijević and Kovačević [34] Dimitrijević and Kovačević performed computational modal analysis on their LASTA aircraft in order to avoid wing flutter in the range of operating speeds of the LASTA. In the automotive industry, passenger safety is a critical design requirement that also depends on car-driver/passenger interactions. Nevertheless, we have not found any work on modal analysis related to automobile car crash tests. Thus, our work here may very well fill in portions of the similar needs for the automotive industry, as the crash conditions being treated here are actually more general. Overall, what we are doing may be best termed as developments in injury and forensic biomechanics.

The vibration modal analysis will be carried out computationally. To achieve this objective, we need to first build a mathematical and computer model for the ring-stiffened aircraft fuselage section in the DYCAST experiments, calibrate and then validate any computational results against experimental data. Structurally speaking, such a fuselage section is primarily a ring-stiffened aluminum thin-shell structure. However, as it turns out, there is a lack of commonly accepted benchmarks for such ring-stiffened structures with a scale in the range of an aircraft fuselage section, despite a broad online search by the authors. Therefore, computer modeling, validation and benchmarking are a crucial first study for the present paper.

Here, we use the versatile software LS-DYNA [35] as the platform for our structural modeling and design, while the experimental data are chosen from the Japanese work [5], [7] as it contains, relatively speaking, the most comprehensive data for our benchmarking purpose. However, the CAD work in Minegishi et al. [5], Kumakura [7] still does not satisfy 100% of our computer modeling needs as some of the seemingly minor design details were omitted in Minegishi et al. [5], Kumakura [7]. Thus, we don’t have a model’s precise data set to validate against - rather, what we are validating against is the data of an approximate model of a ring-stiffened structure.

The paper is organized as follows; cf. also the flowchart in Fig. 41.

In Section 2, we develop a basic DYCAST model by following the design in Minegishi et al. [5], Kumakura [7] as closely as possible, obtaining a ring-stiffened single-section fuselage structure equipped with seats and (passenger) dummies. We then build the finite element model to compute, analyze and compare predictions against those from [5], [7]. Our work here can also be viewed as a benchmark for a ring-stiffened thin shell model for other general purposes.

In Section 3, we first discuss the choices and construction of two types of passenger dummies. The first one is an LS-DYNA dummy [35] while the second is a “make-shift” rubber dummy made due to the limitation of available supercomputer memory storage. We then study their biomechanics by computing and presenting their modal analysis with natural vibration frequencies and mode shapes.

In Section 4, we present the modal analysis of an entire fuselage section containing fuselage, structural supports, seats and passenger dummies. We then again compute and present their modal analysis with natural vibration frequencies and mode shapes, so one can understand the basic types of structure/dummy interactions.

The work of modal analysis in Sections 3 and 4 is based on a linearized elastic model, as modal analysis can only be performed for linear systems or those in the linear regime. In Section 5, we return to the viscoplasticity model and numerically simulate general crash tests for an aircraft section in order to visualize the damage to the aircraft section and the interactive motion of the passenger dummies. The configurations and geometries of general crashes, including a fractal ground and a hill slope, go well beyond the vertical crash test in the literature in order to illustrate and understand the crash mechanisms of structural damages and personnel injuries.

Section 6 provides concluding remarks.

All the mechanical parameters for the various components in this study, such as fuselage, ring stiffeners, stringers, struts, floor, passenger dummies and seats are presented in Appendix A, while the LS-DYNA computer codes are given in Appendix B.