Elsevier

Composite Structures

Volume 131, 1 November 2015, Pages 90-98
Composite Structures

Crush responses of composite cylinder under quasi-static and dynamic loading

https://doi.org/10.1016/j.compstruct.2015.04.057Get rights and content

Abstract

Despite the abundance of studies investigating the performance of composite structures under crush loading, disagreement remains in the literature regarding the effect of increased strain rate on the crush response. This study reports an experimental investigation of the behaviour of a carbon–epoxy composite energy absorber under static and dynamic loading with a strain rate of up to 100 s-1. Consistent damage modes and measured force responses were obtained in samples tested under the same strain rate. The energy absorption was found to be independent of strain rate as the total energy absorption appeared to be largely associated with fibre-dominated fracture, which is independent of strain rate within the studied range. The results from this study are beneficial for the design of energy absorbing structures.

Introduction

Interest in energy absorbing structures for crashworthiness applications has been growing due to the increasingly safety-conscious environment in which the aviation industry operates. Design guidelines [1] and standards [2] impose a maximum allowable acceleration envelope experienced by the occupant in a crash. Composite materials have been gaining popularity in aircraft structures due to their superior specific strength and stiffness, corrosion and fatigue resistance. Their complex failure modes [3], [4], [5] facilitate a high level of energy dissipation making them suitable for use in energy absorbing structures to meet crash protection requirements. Jackson et al. [6] showed that composite energy absorbing structures can significantly reduce the acceleration experienced by the occupants in a crash environment which would reduce risks and severity of injuries. Hence, the performance of composite structures under crush loading is of great interest.

Performance of energy absorbing structures can be measured through their specific energy absorption (SEA), peak force (Fpeak), steady-state force (FSS) and crush efficiency (CE). The energy absorption is area under the force (F) – displacement (x) curve, and is directly related to protective capability. SEA is defined as the energy absorption of unit mass (m) of structure (Eq. (1)). Hence, for weight-conscious applications such as aviation, SEA is a critical measure of performance.SEA=energyabsorbedmassofstructureconsumed=xstartxendFdxmFpeak is the highest force (hence highest acceleration) experienced during the crush event and is directly related to the potential for injury suffered by the occupants in a crash situation. One of the purposes of the energy absorber is to keep acceleration levels within human tolerance limits. FSS is the mean force during steady-state crushing of the specimen after the consumption of the trigger and the passing of peak force, and is a good indicator of the overall energy absorption capability of the structure. CE [7] (Eq. (2)) is the ratio between FSS and Fpeak (Eq. (2)) and is indicative of the nature of the crush response.CE=FSSFpeakA catastrophic failure is characterised by a high peak force as well as a low steady-state crushing force, and hence a low crush efficiency.

A wide variety of composite energy absorber configurations have been reported in the literature since the pioneering work by Thornton [8] and Farley [9]. Simple geometries such as circular tubes [9], [10], [11], [12], rectangular tubes [13], [14], [15] and flat plates [13], [16] have been studied extensively. More complex geometries have also been investigated, including C sections [17], [18] and I sections [15]. Triggering has been shown to be an important aspect of energy absorbent structure design, with chamfering on the loading surface being most common [19]. Other trigger mechanisms studied include the tulip [12], [20] and ply drop [21]. Some authors have designed the structures so that they are self-triggering, for example, using corrugated [22] or hourglass profiles [20]. Others have evaluated different types of material systems and layup configuration [19], [23]. However, most of these experimental data were based on quasi-static testing, despite crushing being a dynamic event. Consequently, the effectiveness of energy absorbers can only be reliably determined once an assessment of possible rate dependence is made.

Currently, there is disagreement in the literature [24] over the effect of intermediate nominal loading rate between 0.1 and 100 s-1 on the response of composite structures. Initial tests completed by Thornton [8] suggested rate independence for glass and graphite–epoxy cylindrical tubes. Farley [10] noted that for chamfered cylinders, [0/±θ]2 graphite–epoxy specimens were rate insensitive, but Kevlar–epoxy and [±θ]3 graphite–epoxy specimens displayed increased specific energy absorption (SEA) as the testing speed increased. Palanivelu et al. [7] found the SEA of specimens with a circular cross-section were rate insensitive, whereas those with a square cross-section increased slightly with crush speed. Work done on rectangular tubes by Mamalis et al. [25] has found that both the SEA and peak force of square tubes increased with respect to increasing strain rate. On the other hand, crush testing conducted by Jackson et al. [17] on chamfered C-section found an approximately 10% reduction in SEA for specimens impacted at 8.5 m/s when compared, with those crushed at 20 mm/s. This was confirmed by David et al. [18] who also observed a reduced SEA on dynamically tested C-section specimens with a [(0/90)2/0/(90/0)2] layup. Brighton et al. [26] reported a decrease in SEA for chamfered carbon–epoxy tubes with a [0/90]4 layup when the test speed was increased whereas the rate effect for a chamfered 4 ply glass-polypropylene fabric tubes were inconclusive.

A sufficiently large sample size is often required to measure the scatter in the experimental data. Brighton et al. [26] noted that the lack of manufacturing control has a significant effect on the specimen response which leads to possible strain rate effect being hidden within the noise. Unstable collapse of specimens [25] also presented challenges in their measurement due to the presence of high force spikes in the resulting force response. Furthermore, unstable crushing response is also more dependent on any microscopic defects or weak points within the structure, which are random in nature. In order for a conclusion to be drawn with confidence, the observed trend must be compared with the size of the scatter inherent in the experimental results.

This study presents a comparison between the response and damage mechanisms of a tulip triggered composite cylinder subjected to quasi-static and dynamic crush loading with strain rates of up to 100 s-1. Qualitative analysis of the specimens was conducted to identify the damage modes and their propagation through the structure. Quantitative analysis of the force response for each test condition was also completed. The reliability of the results was assessed through analysing the scatter of the measured data.

Section snippets

Specimen design

The specimen is a cylindrical tube with a series of tulip triggers cut into the top surface. The tubes were manufactured using a unidirectional Hexcel HexPly T700/M21 carbon–epoxy prepreg with a [0/90/0/90]s layup. The tubes were autoclave cured as per manufacturer’s specifications. The effect of seams is minimised by placing seams of adjacent plies on opposing sides of the cylinder. After curing, the composite tubes were machined to the geometry specified in Fig. 1.

A tubular geometry was

Comparison of damage modes

The typical crush progression of a low rate test is shown in Fig. 2. The eight tulip peaks provided trigger points for damage initiation. The sharp tips of the tulip triggers were easily damaged by the loading surface, which created points of initiation where the damage then spread progressively throughout the entire structure. This process maximised the amount of material damaged and hence the total energy absorption whilst preventing high peak load at the onset of crushing. Significant

Data reliability

Composite crush structures exhibit complex damage mechanisms, which lead to variability in experimental results. To ensure the validity of the results of the present study, a sample size of between 4 to 7 was used so the scatter in the results of each test configuration can be quantified. In contrast, many authors [7], [8], [18], [25] tested only one specimen per test configuration. The observed damage modes in the present study were also compared against the literature.

The specimens in the

Conclusions

The response of a tulip-triggered cylindrical energy absorbing structure undergoing crushing at increasing strain rates was investigated. The damage mechanisms have been identified from post-test inspections. The recorded force response was used to calculate the specific energy absorption (SEA), peak and steady-state forces as well as the crush efficiency.

A high level of consistency was achieved in the present study, indicating good reliability of the results. The observed fracture and damage

Acknowledgement

This work was undertaken within the Systems for Crashworthiness project, part of a CRC-ACS research program, established and supported under the Australian Government’s Cooperative Research Centres Program. Prof Brian G. Falzon acknowledges the financial support of Bombardier and the Royal Academy of Engineering.

References (36)

Cited by (0)

View full text