Development of damage prediction models for plain concrete and FRP strengthened RC plates under near-field blast and projectile impact effects

• Main Author: Tu, Huan
• Other Authors: Fung Tat Ching
• Format: Thesis-Doctor of Philosophy
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Engineering::Civil engineering::Structures and design
• Online Access: https://hdl.handle.net/10356/152471
• Institution: Nanyang Technological University

With increasing terrorist attacks in the global scene, civilian concrete structures have been targets of weapon threats, especially suffering from explosive and ballistic threats. As concrete structures are vulnerable to these extreme effects, the research on the behavior of concrete structures under blast and impact effects and the corresponding protective measures have gained greater and greater momentum. Fiber Reinforced Polymer (FRP), a composite material, has been used to strengthen concrete structures to improve their blast and impact resistance recently. Since the protection of concrete structures is associated with the extent of caused damage, proper prediction tools are required to provide reasonable assessments on the vulnerability of structures to specific threats. Therefore, the research focuses in this thesis are to develop practical approaches to quantitatively predict local damage induced by near-field blast and projectile impact effects. The author firstly investigates local responses on plain concrete plates and reinforced concrete (RC) plates with externally bonded FRP layers and then proposes damage prediction models for each scenario. The study reviews extensively a number of research works conducted by previous researchers from literature. A near-field blast always produces transient non-uniform pressure with very high magnitude on concrete structures and the damage on targets is localized. To investigate the behavior of concrete plates under near-field blast effects, carrying out experiments and using numerical models are two common approaches. However, there are few of available analytical methods to assess the extent of damage on concrete elements. Hence, the author proposes a new analytical model to predict the size of local damage on concrete plates under contact detonation, which represents the most extreme condition of near-field blasts. The accuracy of the analytical model is verified by test results from references. Through the proposed approach, the local responses including crater, spalling and breaching on concrete plates can be well predicted without need of high economic and computational resource. In consideration of the vulnerability of concrete material to explosion loading, adopting externally bonded FRP layers is applicable in improving the blast resistance of RC structures. To quantitively determine the blast resistance of FRP strengthened RC plates under near-field blast effects, Artificial Neural Network (ANN) techniques are introduced and applied to estimate the size of local damage on targets. ANN methods have been successfully applied in some blast incidents, e.g. prediction of blast wave characteristics and estimation of influence of barrier walls on blast wave propagation. Furthermore, some works by using ANN technique to study RC panels/walls under close-in blasts are enlightening. In this thesis, the author constructs an ANN model based on the data from high-fidelity numerical simulations and this well-trained model can provide reasonable predictions on the local damage of RC plates and FRP layers. To the author’s knowledge, due to the limited number of relevant studies, there is no model available to quantitatively predict the damage on an FRP strengthened RC plate after a near-field blast. Using the results from the ANN model, a damage prediction chart is proposed for fast assessment. Based on the literature review, researchers have used experimental, numerical and semi-analytical approaches to investigate the behavior of plain concrete under projectile impacts. Different empirical models have been proposed to evaluate local effects on concrete targets. Besides, several semi-analytical methods are available to estimate the ballistic behavior of projectiles into concrete plates. However, the emphasis of most studies is on the penetration depth but not the surface damage (quantified by the spalling crater area). Ballistic attack constitutes a potent threat to nearby personnel and equipment and proper protection against projectile and fragments is of great importance. Therefore, to investigate the total quantity of concrete debris produced in one projectile impact, not only the penetration depth but also the surface damage needs to be well considered. To fill the gap, the author conducts a series of projectile impact tests and then develops an empirical model to predict the spalling damage on the impacted face of concrete targets. The RC elements exhibit limited resistance to ballistic effects by virtue of the poor energy absorption capacity of concrete material. It is widely accepted that externally bonded FRP layers is useful to protect targets from striking projectiles. However, there are only a small number of relevant studies on FRP strengthened RC plates. In the thesis, the author focuses on the response at the impacted face and conducts tests on RC plates with front strengthening FRP layers. Based on obtained results, the author modifies the proposed empirical model for plain concrete targets to quantitatively predict the spalling damage on FRP strengthened RC plates. The damage prediction models developed in this thesis can provide proper assessments on local damage size, which possess practicability in protective design.