Repository hosted by TU Delft Library

Home · Contact · About · Disclaimer ·
 

Design and analyses of porous concrete for safety applications

Author: Agar Ozbek, A.S.
Type:dissertation
Date:2016
Publisher: TU Delft
Place: Delft
Identifier: 747141
doi: doi:10.4233/uuid:7c67133f-6c68-4385-a7e6-0b43fd5e2045
Keywords: Observation, Weapon & Protection Systems · EBP - Explosions, Ballistics & Protection · TS - Technical Sciences

Abstract

In an explosion taking place close to or inside a concrete structure, apart from the dangers of the explosive itself, the hazard due to the large debris originating from the concrete structure is an important threat. Protective structures with high probability to experience such extreme loadings during their service lives, such as protection walls outside important buildings or munition magazines for storing explosives, have to be designed to mitigate the effects of a probable explosion. With this goal, a PhD project was undertaken on designing and analyzing a cementitious material with sufficient static strength to carry the service loads and fractures into small-size fragments when subjected to an explosion. This thesis presents the design procedure of such a material, the experimental and numerical investigations and the results of the project. In the last decades, extensive research has been focused on understanding the behavior of concrete structures under extreme dynamic loadings. While most studies aim at designing materials that resist impact loading, the objective of this project was developing a cementitious material that is expected to fracture and disintegrate under impact. During the research, a sensitivity study was conducted on various forms of cementitious materials (Chapter 2). Porous concretes with enhanced strength, fracturing into small fragments when exposed to impact loading, were obtained and analyzed. In the process of modifying the porous concrete properties the main goal was to enhance the static strength while maintaining a high porosity. By modifying the mixture composition as well as the method of compaction, porous concretes with improved static strengths (at the range of 30-50 MPa) were produced as presented in detail in Chapter 3. In order to better understand the properties of the porous concretes produced, mechanical tests at macro and meso-scales were performed. Computer tomography, electron microscopy and X-ray diffraction analyses were also conducted to better explain the effective parameters which were summarized in Chapter 3. Among all the parameters that have been investigated, aggregate properties have the most significant effect on the mechanical properties of porous concrete. On the other hand, intense compaction in thin layers (to ensure good control of the production process), which directly affects the packing of the particles and the porosity, is essential to attain porous concretes with substantial strengths. The contribution of microsilica to the mechanical properties of porous concrete was not significant, even though the tests on the interfacial transition zone (ITZ) clearly showed that microsilica improved the strength of the ITZ phase. The reason for this „contradiction‟ was better explained by the CT scans of partially fractured samples. These scans revealed that in porous concretes crack patterns are very much influenced by the distinct porous structure and the aggregate skeleton. The cracks are forced to propagate into locations guided by the geometry of the present phases and the path does not always go through the weakest phase i.e. the ITZ. In the process of designing a material with desired dynamic performance, efficient dynamic testing techniques were highly required. In Chapter 4, dealing with the dynamic testing and evaluation of the experimental research, various experimental configurations were adapted or introduced for determining the dynamic response of porous concretes in a drop weight impact test. Different types of porous concretes and a mixture of normal concrete were subjected to drop weight impact tests. When the experimental configurations that have been used are compared, it can be concluded that stress gauge measurements have the advantage of direct measurement of the transmitted stress, while no analyses are needed afterwards to obtain the impact strength data. Among the other two monitoring and measuring techniques (LDV and high speed photography) that have been applied, LDV proved to be more accurate due to its higher sampling rate compared to high speed photography. It also has the advantage of being independent of factors such as the intensity of the light source. High speed photography, however, has its own advantage of also facilitating the visualization of the fracture process, which provides important qualitative information on the crack patterns. On the other hand, both techniques have the advantage of being non-contact methods. In the analyses of the drop weight impact test results, obtained using LDV and high-speed photography, the reverberation application of the impedance mismatch method was used. The analysis method proved to be very suitable to investigate the drop weight impact behavior of porous concretes. It also had the advantage of involving only the well-known dynamic impedance properties and the velocity measurements of the impactor, while the properties of the tested target specimen are not directly involved in the measurements and the analysis. The focus of the numerical part of the research was simulating and assessing the dynamic behavior of different porous concretes under impact loading. To achieve this goal, analyses were conducted both on real porous concretes and on fictitious (model) porous materials. The latter models were used to demonstrate better the individual factors affecting the porous concrete properties. In the research, explicit time integration was selected as the analysis method. For the realistic representation of the real concretes that were produced and tested, the aggregates present in the mixtures were directly determined through 3D computed tomography. A mesh generation program was developed for generating realistic finite element meshes. In the analyses, porous concrete was considered as a four-phase material consisting of aggregates, bulk cement paste, interfacial transition zones (ITZ) and meso-size air pores. When analyzing real porous concretes, the influences of the different parameters such as pore and aggregate size distributions and total porosity, are usually coupled. For distinguishing the sole effect of each parameter, virtual or model porous concretes were analyzed. The primary objectives of these analyses were computing the impact strength and examining the fragmentation behavior. The pore structures parameters are the most dominant factors affecting the mechanical properties of porous concretes. Therefore, model porous concretes having regularly and irregularly distributed circular pores of different radii and total porosity were analyzed. From the comparison of the impact responses of model porous concretes with different size pores, it could be concluded that for all porosities analyzed the impact strength of the concretes increased with decreasing pore size (at constant total porosity). It could also be concluded that for constant total porosity the sizes of the fragments that were formed drastically decreased when the pore size decreased. The results obtained from the analyses on the concretes with randomly distributed pores showed that for mixtures with larger pore sizes the scatter in impact strengths was larger. To understand the effect of aggregate size on the porous concrete properties, circular aggregate model porous concretes were analyzed. The results showed that the impact strengths of porous concretes with 8 mm and 4 mm particles (same porosity) were approximately the same, while the impact strength of the model porous concrete with 2 mm particles was lower. Reasons for this were explained in detail in Chapter 5: in different mixtures several factors influenced the results in opposite ways. In the numerical analyses of real porous concretes, the simulations were quite realistic. This result was achieved by applying proper finite element meshes resembling the actual porous concretes and accurate input data for the cementitious phases of the materials being quantified in the experiments at the meso-scale. Comparison of the peak values of the impact stress time histories showed that impact strengths of the samples obtained numerically (presented in Chapter 5) were in a good agreement with the experimental results. Pore and fragment size distributions of real porous concretes were also computed using the numerical models. This provided valuable information for better understanding of the fragmentation of the different mixtures. This thesis comprises experimental and numerical investigations performed on porous concretes meant to be used in safety applications. The results obtained from the enhanced strength porous concretes that have been produced were promising. The impact test monitoring methods proposed as well as the numerical analyses can be used in experimental and numerical investigation of the dynamic behavior of porous concretes.