Damage, deformation and residual burst strength of filament-wound composite tubes subjected to impact or quasi-static indentation

https://doi.org/10.1016/S1359-8368(00)00014-7Get rights and content

Abstract

Thin-walled filament-wound E-glass fibre-reinforced epoxy tubes were subjected to lateral indentation in quasi-static and low speed impact tests and were then tested under internal pressure to determine their residual burst strength. The behaviour of tubes subjected to quasi-static and low velocity impact loading tests was found to be the same. Experimental strain measurements in quasi-static indentation tests showed a large degree of redistribution of strain with increasing deflection which resulted in local buckling failure away from the indentation point. Damage in the form of matrix cracking resulted from low energy indentation but did not reduce the residual burst strength of the tube. Higher energy indentation, which produced buckling failure, reduced the tubes’ burst strength by 60%.

Introduction

Filament-wound E-glass fibre-reinforced epoxy resin tubes are known to have extremely high strength-to-weight ratio but may be susceptible to damage by rough handling. In order to evaluate the susceptibility of such tubes to impact damage the Defence Research Agency (DRA) Fort Halstead, devised a test in which tubes were subjected to controlled impacts by a dropped object and then tested under internal pressure loading to determine their residual burst strength [1], [2]. In the impact tests thin-walled tubes were found to fail at sites away from the point of impact [1].

In order to investigate this behaviour low velocity impact indentation of filament-wound tubes was simulated by quasi-static finite element analysis assuming linear elastic material properties [3], [4]. Related work on the indentation of composite tubes was reviewed previously [3]. The predicted initial strain distributions were confirmed by quasi-static experimental strain measurements at low strains [3]. The theory predicted a redistribution of stresses at high strains and that failure should occur under the indenter. In the quasi-static experiments there was evidence of damage due to microcracking under the indenter but final failure again occurred some distance away from the indenter. This final failure was predicted to be due to local axial shell buckling. There were no experimental strain measurements available to confirm the theoretical redistributions at high strains, the only available measurements being for a single point immediately under the indenter [4].

The objectives of the current work were to measure the strain distributions in quasi-static indentation tests up to final failure and to compare the deformation, damage and residual burst strengths of specimens subjected to lateral impacts with those of specimens subjected to quasi-static loading.

Section snippets

Theoretical analysis of tubes subjected to static indentation

It has been shown that the problem of filament-wound GRP tubes subjected to quasi-static lateral indentation can be modelled using the finite element code Abaqus [3], [4]. Abaqus is able to cope with the geometric non-linearities resulting from the significant change of shape of the tube during the deformation and the changing contact conditions between the indenter and the tube and the tube and the supporting plate (see Fig. 1). In the finite element model linear elastic material properties

Specimens

The specimens were filament-wound E-glass fibre epoxy tubes manufactured at DRA Fort Halstead. The E-glass fibre reinforcement was Silenka 051L, 1200 tex and the epoxy resin system was Ciba-Geigy MY750/HY917/DY063. A simple helical winding pattern was employed with all fibres at an angle of ±55° to the tube axis. The tubes were manufactured with two covers giving a fibre arrangement approximately equivalent to a +55°/−55°/+55°/−55° laminate and a tube wall thickness of approximately 1 mm. The

Burst testing

A schematic diagram of the burst test apparatus is shown in Fig. 12. In this type of burst test the tube specimen is subjected to hoop loading only, the axial loads being carried by the tie-bars. The tube was filled with water and pressure was applied using a two-stage hydraulic pressure intensifier. Each of the impacted specimens was mounted in the pressure test rig and loaded by internal pressurisation until failure occurred due to bursting (rupture) of the tube wall.

Fig. 13 shows the burst

Discussion

The experimental load versus indenter displacement results from both laboratories and from the interrupted tests are all very similar (see Fig. 2). The theoretical load versus indenter displacement curve (Fig. 2) is similar to the experimental curves but the theoretical curve is slightly stiffer. This is thought to be mainly due to the assumption of linear elastic material properties and ignoring the effects of the local resin cracking which occurs under the indenter and increases in crack

Conclusions

The experimental strain measurements confirm that there is a significant redistribution of strains during loading. Damage (microcracking) only reduced overall stiffness slightly but did result in a change in strain distribution. Measured redistributions of strains and variation in strain with indenter displacement (Fig. 4, Fig. 5) showed a similar pattern to that predicted by a large displacement finite element model but the magnitude of changes was even more pronounced than predicted by

References (11)

There are more references available in the full text version of this article.

Cited by (69)

View all citing articles on Scopus
View full text