Carlos Chastre

The bias factor of the Eurocode 2 [CEN (European Committee for Standardization) (2008). Eurocode 2: Design of ConcreteStructures–Part 1-1: General Rules and Rules for Buildings] and ACI 318 [ACI (American Concrete Institute) (2014). Building CodeRequirements for Structural Concrete and Commentary] flexural resistance models of reinforced concrete beams are compared withemphasis on the effect of the incorporation of coarse recycled aggregates sourced from concrete waste. The bias factor of the yielding momentcalculations according to both codes is also investigated, and the bias in the cracking moment when Eurocode 2 material clauses are used. Thedatabase was composed of 174 beams, and the criteria that led to its development are discussed. The effect of recycled aggregate incorporationon the statistical descriptors of the bias factor is evaluated and probabilistic modeling using lognormal distributions is argued for. Preliminarypartial safety factors for the bias factor of recycled aggregate concrete beams are proposed. No significant differences in the bias of theultimate moment were found between the two comparison vectors: Eurocode 2 versus ACI 318 specifications and recycled versus naturalcoarse aggregate. The bias of the cracking moment increased when coarse recycled aggregates were incorporated, most probably due to thehigher heterogeneity of recycled aggregates.

TThis chapter concerns the need to evaluate the reliability of structural elements produced with recycled aggregates concrete (RAC) in order to address some designers’ scepticism towards the use of this eco-friendly material. The current knowledge on RAC’s behaviour demonstrates its viability for structural purposes. However, of the investigations performed so far very few are related to a fundamental aspect towards RAC world-wide application as a structural material: structural codes have a probabilistic basis.After briefly presenting the state-of-the-art knowledge on the material properties and structural behaviour of RAC, the limitations of the current knowledge are debated. Afterwards, an introduction to structural codification is presented, as well as the fundamentals of reliability analysis. Examples of code verifications are contextualized with their underlying assumptions and the information necessary for code calibration is discussed. The role of reliability in the calibration of structural codes is shown, common techniques for reliability calculations are briefly explained, and relevant references in the area are cited for the readers’ perusal.Having established how structural codes are calibrated, the state-of-the-art on the probabilistic and statistical knowledge of RAC properties is reviewed. The implications of the very reduced number of studies on this area are discussed and the need to conduct further studies is emphasized.Afterwards, investigations that used reliability analysis to calibrate partial safety factors applicable to RAC are reviewed. The methodology of each investigation is presented, the experimental tests that led to the definition of the probabilistic information of the RAC’s parameters are described and the need to have a wide range of data coming from different RAC compositions and aggregate sources is debated.This chapter finishes by contextualizing the current knowledge on RAC properties with the necessary information for code calibration procedures. The relevance of a code proposal towards RAC affirmation as a structural material is highlighted, as well as the requirements of such code. Suggestions for future studies are made.

A produção de betão com substituição de agregados naturais por agregados reciclados minimiza os impactes da indústria da construção. Contudo, o recurso a este tipo de agregado é limitado devido a dúvidas dos diferentes agentes da indústria da construção e à ausência de regulamentação específica para o projecto de betão com agregados reciclados. A percepção de betões com agregados reciclados como um material heterogéneo com comportamento imprevisível é o principal entrave, quer para a aceitação da indústria, quer para o desenvolvimento de regulamentação. Diferentes aspectos relacionados com o efeitos da variabilidade das propriedades de betão com agregados reciclados na regulamentação de estruturas são abordados conceptualmente: a heterogeneidade dos agregados reciclados, o efeito desta heterogeneidade nas propriedades mecânicas de betão e os efeitos da variabilidade das propriedades de betão na sua fiabilidade estrutural. Os primeiros resultados de uma campanha experimental desenvolvida especificamente para abordar o efeito de agregados grossos reciclados na fiabilidade são apresentados e as suas implicações no projecto de estruturas discutidas. São apresentadas sugestões de desenvolvimentos futuros que visam a calibração de coeficientes parciais de segurança que possibilitem o projecto de estruturas de betão armado dimensionadas segundo o formato do Eurocódigo 2 e mantendo os mesmos níveis de segurança de betão convencional.

An investigation on the scatter of code-type constitutive models that relate compressive strength (fc) with tensile strength (fct) and Young’s modulus (Ec) of standard concrete specimens is presented. The influence of the mix design on the accuracy of the fc vs. fct and fc vs. Ecrelationships is discussed, with emphasis on the lithological type and morphology of the coarse aggregates. The uncertainty of the constitutive models is analysed in probabilistic terms and random variables that model the uncertainty of the fc vs. fct and fc vs. Ec relationships are proposed for reliability analyses of serviceability limit states. The suitability of the models proposed is assessed through preliminary conservative estimates of their design values.

Research on the variability of the properties of recycled aggregate concrete is lacking and is necessary for the development of reliability analyses and code calibration procedures. This paper presents an experimental programme on the within-batch variability of the compressive strength, Young’s modulus, and splitting tensile strength of several recycled and natural aggregate concrete mixes. The influence of the recycled concrete aggregates on the mechanical properties and variability of concrete is analysed and discussed and benchmarks with standard predictions for the variability of natural aggregate concrete are made. It was found that full recycled aggregate concrete incorporation did not increase the variability of any of the properties tested, but intermediate ratios of recycled aggregate incorporation did. The properties of high-strength concrete mixes were more variable than that of all other mixes, irrespective of recycled aggregate incorporation. All properties of all compositions were suitably modelled by normal distributions. The coarse recycled aggregates were sourced from concrete waste.

This paper investigates whether the model uncertainty of reinforced recycled aggregate concrete (RAC) beams subjected to bending differs from that of reinforced natural aggregate concrete (NAC) beams. An introductory remark concerning the importance of the codification of RAC structural design is made and notions concerning model uncertainties and their role on structural codification are given. Afterwards, the criteria used in the construction of a database of RAC and NAC beams are referred before presenting the key findings of an analysis on the model uncertainty of the cracking, yielding and ultimate moments of beams subjected to four-point bending tests. The analytical moments were calculated following Eurocode 2 provisions. Probabilistic models for model uncertainties are proposed. Negligible differences in the model uncertainty of NAC and RAC beams are reported.

This paper analyses the studies made so far on the variability and reliability of recycled aggregate concrete. Since recycled aggregate concrete is seen by different agents of the construction industry as a variable material and no structural code has specifically been calibrated to its use, its role as a structural material is limited. Such calibration is hindered since specific research on the statistical and probabilistic data of recycled aggregate concrete properties is lacking.Investigations on the probabilistic knowledge of recycled aggregates and recycled aggregate concrete properties are discussed, and the studies made so far on the reliability of recycled aggregate concrete elements are summarised. Final remarks regarding the future prospects towards the consensual acceptance of recycled aggregate concrete structures are provided.

This paper investigates the effect of recycled coarse aggregate incorporation on the relationship between 150 mm cubic and Փ 150 mm cylindrical compressive strength (the reference strength of standards) by comparing data from recycled and natural aggregate concrete compositions in which both cubes and cylinders were tested. A conversion factor from cubic to cylindrical strength is proposed in two versions: A deterministic and a probabilistic one. Such factor has not been studied before and researchers have been converting cubic data as if natural aggregate concrete were tested. The probabilistic factor is intended for reliability analyses on the structural behaviour of recycled aggregate concrete using data from laboratory cube tests. It was found that the incorporation of recycled coarse aggregates sourced from concrete waste significantly decreases the expected value of the factor but the factor’s scatter is relatively unaffected.

Este artigo avalia se, para especificações de betão correntes, a resistência à compressão de betão pronto é significativamente dependente da central que o produziu, bem como se a variabilidade entre amassaduras de uma determinada composição é dependente da classe de resistência especificada. É apresentado um parâmetro probabilístico que converte a resistência à compressão característica especificada para a resistência à compressão expectável em provetes de betão cúbicos em condições padronizadas.É feita uma análise estatística e probabilística dos resultados de ensaios de resistência à compressão aos 28 dias de betão pronto produzido em três centrais de betão portuguesas durante o ano de 2017. Após a avaliação da representatividade da base de dados analisada, é feito um estudo comparativo dos parâmetros estatísticos dos dados de produção das diferentes centrais e é avaliada a influência da classe de resistência à compressão do betão nestes parâmetros. Define-se e analisa-se um parâmetro probabilístico que converte a resistência à compressão especificada para a resistência potencial de betão, para cubos de 150 mm. Este parâmetro contribui para uma futura proposta de um modelo probabilístico, a ser utilizado em análises de fiabilidade, que converte a resistência à compressão especificada para a resistência à compressão de elementos de betão em obra.Quer a análise estatística dos ensaios de resistência à compressão, quer o parâmetro de conversão são comparados com os resultados de investigações internacionais na área, aferindo-se se a qualidade da produção de betão pronto nacional é semelhante à de outras regiões.

An analysis on the effect of the incorporation of coarse recycled concrete aggregates on the bond strength between concrete and embedded steel reinforcement is presented. The model’s uncertainty of the Level I provision of the anchorage length of fib Bulletin 72 on ribbed steel/recycled aggregate concrete bond is quantified. Afterwards, reliability analyses on the bond strength are made and a partial safety factor for the anchorage length of recycled concrete elements is proposed. The model’s uncertainty is evaluated through data from pullout tests, the only type of bond test that has been so far performed extensively on recycled aggregate concrete specimens. The limitations of this test in reproducing the bond of actual structural elements is discussed, and the model’s uncertainty is converted to that of lap splice tests. The bond strength of recycled aggregate concrete design was found to be less reliable than that of natural aggregate concrete, especially in the absence of confining reinforcement. For concrete with full recycled aggregate incorporation, a 25% increase in the anchorage length is proposed. Additional testing on the bond strength of lapped splices or beam-end specimens is recommended.

This paper evaluates and compares the statistics of compressive strength data from three Portuguese ready-mixed concrete plants. A hierarchical model showed that different groups of concrete strength records are not statistically equivalent, even if they were produced in the same plant and using the same concrete composition. This finding is related to autocorrelation. For the same specified strength class, compositions produced less often result in higher average compressive strength and variability. The statistics of one of the plants were quite different from those of the others, even though the concrete of this plant also complied with the specifications. It was found that the average compressive strength of a mix may be quite dependent on the plant that produced it, even if the compressive strength complies with quality control specifications. Conformity with the target slump and strength class was checked following the conformity criteria of EN 206-1 for continuous production. Nonconformity with slump is more frequent than failure to comply with the strength class. A bias factor for reliability analyses was proposed.