Journal of Materials Processing Technology 162–163 (2005) 76–82 A tensile impact apparatus for characterization of fibrous composites at high strain rates G.H. Majzoobi ∗ , F. Fereshteh Saniee, M. Bahrami Mechanical Engineering Department, Faculty of Engineering, Bu-Ali Sina University, Hamadan, Iran Abstract A high rate tensile testing apparatus called “flying wedge” for testing fibrous composites at strain rates up to 10 3 s -1 has been described. R2000 glass/epoxy composite has been tested at a very low strain rate of 10 -3 s -1 using the universal Instron tensile testing machine, and at high strain rates up to nearly 850 s -1 using flying wedge. The results show significant increase in failure stress and reduction in failure strain for dynamic tests with respect to quasi-static experiments. While the reduction in failure strain fluctuates from 60 to 75%, the increase in failure strength varies from 300 to even 500% for different ply angles. Although, the results may be affected by the stress waves, which propagate through the parts of the testing machine on the impact, it is found that the stress wave’s effects do not play a determinant role in this regard. The results indicate that the rate of the increase of stress versus strain rate slows down as ply angle increases. Scanning electron microscopy of the fracture surface of the specimens did not reveal significant difference between the failure mechanisms at low and high strain rates. © 2005 Elsevier B.V. All rights reserved. Keywords: Glass/epoxy composite; Strain rate; Flying wedge; Impact; Strength 1. Introduction Because of their high strength to weight ratios and a high resistance to corrosion and abrasion, fiber-reinforced com- posites have gained wide application in the design of struc- tures exposed to explosion, shock loading and impact such as aircrafts, missiles, spaceships, high-speed boats and many other structures. Therefore, mechanical behavior of these sorts of materials must be fully characterized within the range of strain rates at which the structure is exposed. Over the past decades, many attempts have been devoted to the study of behavior of glass fiber-reinforced compos- ites at high strain rates. Most of the results indicate that the properties of this kind of composites such as elastic modu- lus, yield stress and ultimate stress increase with the increase of strain rate. However, there is not a consensus among the authors about the degree of this increase. The reason may be attributed to the fact that the experimental results are highly dependent on the test rig, specimen geometry, manufactur- ∗ Corresponding author. E-mailaddress: gh majzoobi@yahoo.co.uk (G.H. Majzoobi). ing process and so on. Most of the experimental works on the high strain rate response of composites involve the use of split Hopkinson bar (SHB) with different load applications including tension [1], compression [2] and shear [3]. Lifshitz and Leber [4] investigated the inter-laminar ten- sile strength and modulus of E-glass/epoxy at strain rates of 100–250 s -1 using SHB. Most of their results were higher than the quasi-static values by a factor of about 1.3. Tsai and Sun [5] examined the properties of S2/8552 glass/epoxy composites at strain rates 400–700 s -1 with different speci- men geometries using compressive SHB. They showed that the results for coupon specimens were different from those obtained with block specimens. Their results also suggest that the viscoplasticity model established with lower strain rates test data could be used for high strain rates up to700 s -1 . Staab and Gilat [6] investigated the effects of strain rate up to 10 3 s -1 on 1002 glass/epoxy laminates. Their results indi- cate that, depending on the value of strain rate, the maximum normal stress is 50–100% higher for dynamic tests than quasi- static loading conditions. They also showed that although the fibers and matrix are both sensitive to strain rate, the fibers influence laminate rate sensitivity more than the matrix. The 0924-0136/$ – see front matter © 2005 Elsevier B.V. All rights reserved. doi:10.1016/j.jmatprotec.2005.02.182