Mechanical characterization of textile reinforced cementitious composites under impact tensile loading using the split Hopkinson tension bar

Abstract

Strain-hardening cement-based composites (SHCC) reinforced additionally with continuous textile reinforcement exhibit a high tensile strength, ductility and low crack width up to failure localization, being suitable as strengthening layers on structural elements subject to impact loading. The quasi-static tensile properties of such composites are usually derived on long, planar specimens, as dictated by the geometry of the textile reinforcement and by the necessity of sufficient yarn anchorage. However, with regard to high strain rate testing in split Hopkinson bar systems, both the length and the planar shape of the textile reinforced specimen represent major drawbacks. This explains the lack of comprehensive investigations on the mechanical performance of textile reinforced cement-based composites from the perspective of material characterization. This paper presents two new configurations of a gravity-driven split Hopkinson tension bar (SHTB) purposefully developed for investigating the tensile behavior of such composites under strain/displacement rates in the range of impact loading. The first configuration is designed for uniaxial tension tests on planar textile reinforced composites. The planer specimens are attached to the input and output bars using special aluminum adapters. The influence of the adapters and of specimen geometry on wave propagation and dynamic stress equilibrium is discussed in detail based on the results of experimental and numerical investigation. Corresponding amendments to the traditional wave analysis and suitable evaluation methods are proposed for an accurate assessment of the material response. Additionally, a novel testing configuration for single-yarn pullout experiments is presented. This setup allows for a detailed description of the rate effects on the bond between textile yarns and cementitious matrix.

Introduction

Textile reinforced concrete and fabric-reinforced mortars have emerged as a promising solution for lightweight construction elements and structural retrofit and strengthening [1,2]. These novel composites consist of fine-grained cementitious mortars and two- or three-dimensional textile meshes made of carbon, alkali-resistant glass or polymer multi-filament yarns [3,4]. Given the high tensile strength of the textile yarns and their durable nature under normal service conditions, construction elements and strengthening covers reinforced with textile meshes feature small thicknesses and offer higher flexibility in terms of fabrication and application technology as well as element shape.

Furthermore, by providing a volumetric micro-reinforcement of the cementitious matrix through the addition of short, randomly distributed fibers [[5], [6], [7]], enhanced crack control, damage tolerance, and energy dissipation capacity can be achieved. The latter is especially important in the case of structural strengthening against severe dynamic actions, such as impact or blast. However, for fully exploiting the features mentioned above at high loading rates and for enabling an efficient and reliable material design, appropriate material investigation techniques are needed both at composites' and constituents' scales [8].

Modern high-rate servo-hydraulic testing machines enable displacement speeds higher than 10 m/s, which apparently qualifies them as applicable for impact tension tests [9]. However, the sudden release of hydraulic pressure for loading the specimen and the loading process itself trigger vibrations in the entire testing frame or in individual parts of the setup, which distort the effective specimen response, especially when the oscillations are of high amplitudes due to resonance effects [10,11]. The oscillations can be filtered if the vibration modes of the testing setup are known [9]. However, this is problematic when the frequency content of the setup ringing is close to that of the specimen response, such as in the case of a short experiment duration, i.e., at high loading rates. For this reason, and because of other aspects presented hereafter, the servo-hydraulic testing setups mainly enable reliable measurements of the material response up to strain rates of 10 s⁻¹ only.

The development of the Kolsky bar [12], also known as split Hopkinson bar, formed the basis for most of the currently used material testing techniques at high loading rates [13]. Split Hopkinson tension bars (SHTB) [14,15] are modifications of the original Kolsky's configuration, and they consist of an input bar, in which the loading wave is induced, a transmitter bar, in which the wave transmitted by the specimen is recorded, and of a specimen fixed between the two bars. As opposed to the Kolsky bar configuration, the SHTBs are designed for directly generating a tensile loading pulse usually by using a striker system or a pre-tensioned bar. The split Hopkinson bar testing configuration relies on the one-dimensional wave propagation in elastic bars for assessing the loading process and the specimen response [16]. For avoiding pronounced wave dispersion effects, the bars usually have round cross-sections, while their length is considerably larger than their diameter. Moreover, the uniformity of stress distribution along the sample, known as dynamic stress equilibrium, is a prerequisite of reliable quantification of the material response in general and strength in particular. These boundary conditions define to a large extent the shape and size of the tested specimens commonly of cylindrical shape and relatively small length.

At the same time, it is known that textile reinforced concretes or mortars require a relatively large anchorage length of the textile mesh in the cementitious matrix, which usually implies large specimen lengths [17]. For dynamic material characterization, this condition might impose significant challenges related to dynamic stress equilibrium in the sample. Furthermore, given that the textile fabric imposes planar sample geometries, dedicated testing methods must be developed [[18], [19], [20]].

The paper at hand presents a split Hopkinson tension bar which was purposefully adapted for testing planar samples of textile reinforced composites under impact tension at displacement rates of up to 8 m/s. The composite specimens are attached to the input and output bars through specially designed adapters. Since these elements represent additional geometrical discontinuities in the testing setup, their effect on the wave propagation was analyzed experimentally and numerically, which facilitated the definition of appropriate evaluation techniques for differentiating the specimen response from various wave effects related to the setup itself.

Moreover, a novel testing configuration for single-yarn pullout tests at high displacement rates was proposed and analyzed. This configuration enables experimental investigation of the pullout behavior of textile yarns at displacement rates of up to 8 m/s for capturing the rate effects on the yarn-matrix bond. To demonstrate the suitability of the developed setups for the proposed task, experimental results on planar specimens consisting of high-strength strain-hardening cement-based composites (SHCC) with and without one layer of carbon textile were presented. The article focuses on the characterization of the testing configurations, while the detailed material characterization is a matter of other works performed by the authors [6].