Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Wyszukiwano:
w słowach kluczowych: ściskanie betonu

Sortuj według:

Ogranicz wyniki do:

Stochastic identity of test result series of the compressive strength of concrete in industrial production conditions

Jasiczak J., Kanoniczak M., Smaga Ł.

Archives of Civil and Mechanical Engineering

2015

Vol. 15, no. 2

584--592

The article provides the scientific basis of interval identity of the compressive strength of concrete, as well as includes calculation examples concerning the application of concrete apron paving with the use of concrete pavers. It was found that in order to control the compressive strength of concrete formed in a continuous, cyclically repetitive mode, standard statistical assessment of concrete strength is not sufficient since, in the case of deviations in the studied characteristics of concrete from the project assumptions, it is not possible either to implement corrections in the concrete manufacturing process or to change the location of the concrete structure with decreased strength parameters. Due to this, the analysis of the qualitative and quantitative production of concrete mix and its application at the construction site required the use of appropriate calculation procedures, allowing to extract subsets of the results of the compressive strength of concrete projecting technological disturbances and assign to them time intervals, in which the disturbances occurred. This is a completely different approach to the issue of determining the compressive strength properties of tested concrete volume than presented in the construction standards. For this reason, to present the real picture of the compressive strength variability, the division of tests results was made by grouping the data belonging to one population with probabilistically stabilized strength parameters forming a family of concrete. In order to do this it was necessary to use appropriate calculation procedures based on verification of statistical hypotheses. The authors proposed three independent methods based on statistical tests: Pearson χ2 test, t-Student's test and Mann–Whitney's U test. The results series were divided into families of concretes and were given different sensitivity of each method. However, the breakdown of the whole series achieved by the three methods is comparable and occurred in the areas of high variability of strength.

Nieniszcząca identyfikacja wytrzmałości na ściskanie betonu wysokowartościowego przy wykorzystaniu sieci neuronowych

Schabowicz K.

Zeszyty Naukowe. Budownictwo / Politechnika Śląska

2001

z. 93

417-426

W referacie przedstawiono rezultaty zastosowania sztucznych sieci neuronowych do identyfikacji wytrzymałości na ściskanie betonu wysokowartościowego, na podstawie parametrów określonych metodami nieniszczącymi. Bazę dla tego obliczeniowego eksperymentu stanowiły badania doświadczalne kilku betonów wysokowartościowych różniących się składem, w tym ilością domieszki superplastyfikatora i dodatku pyłu krzemionkowego. Wykorzystano parametry określone dla badanych betonów metodami ultradźwiękową, sklerometrycznymi i "pull-out". Uzyskano pozytywny rezultat predykcji neuronowej wytrzymałości na ściskanie badanych betonów, na podstawie parametrów określonych nieniszczącymi metodami. Na uwagę zasługują otrzymane bardzo wysokie wartości współczynników korelacji i niskie błędy względne. Wynik eksperymentu wskazuje na możliwość dokonywania neuronowej identyfikacji wytrzymałości betonu wbudowanego w konstrukcje budowlane.

Results of the application of artificial neural networks to the identification of the compression strength of high-quality concrete on the basis of parameters determined by non-destructive testing are presented. Tests were carried out on several high-strength concretes differing in their composition, also in their amounts of superplasticizer and silica fume additives. Their parameters were determined by respectively ultrasonics, sclerometry and "pull-out". The neural prediction of the compression strength of the concretes on the basis of the parameters determined by non-destructive testing proved to be accurate. Very high correlation coefficients and low relative errors were obtained. The experiment has shown that the neural identification of the strength of concrete built in constructions is viable.