Wyniki wyszukiwania - BazTech

Ograniczanie wyników

Powiadomienia systemowe

Sesja wygasła!

Znaleziono wyników: 2

Liczba wyników na stronie

Wyniki wyszukiwania

Sortuj według:

Ogranicz wyniki do:

Investigation into the influence of vinyl polyester fibers on the basic properties of cement composites

Moj Mateusz, Walerus Tomasz

Materiały Budowlane

2025

nr 9

92--99

The purpose of this study was to evaluate the effect of vinyl-polyester waste fibers on the properties of fresh and hardened cement mortars. Mortars containing 1%, 2% and 4% fibers of different fractions were analyzed. The results showed that the addition of up to 1% does not significantly affect mortar parameters, while higher dosage causes deterioration. These fibers may be a useful component of mortars in the context of recycling waste materials and reducing the environmental impact of construction.

W artykule zaprezentowano wpływ włókien winylowo-poliestrowych pochodzących z odpadów na właściwości świeżych i stwardniałych zapraw cementowych. Analizie poddano zaprawy zawierające 1, 2 i 4% włókien o różnej frakcji. Wyniki wykazały, że dodatek do 1% nie wpływa istotnie na parametry zaprawy, natomiast większe dozowanie powoduje ich pogorszenie. Badane włókna mogą stanowić użyteczny składnik zapraw w kontekście recyklingu materiałów odpadowych i ograniczania wpływu budownictwa na środowisko.

Comparison of machine learning models predicting the pull-off strength of modified epoxy resin floors

Moj Mateusz, Kampa Łukasz, Czarnecki Sławomir

Studia Geotechnica et Mechanica

2024

Vol. 46, nr s1

377--388

It is becoming popular to replace destructive laboratory testing with related nondestructive testing (NDT) and/or machine learning (ML) techniques. Such an approach is becoming particularly desirable in operating facilities, where failing components result not only in the need for repair but also in the suspension of facility use for up to several months. Supporting construction work with artificial intelligence (AI) offers the potential for breakthroughs in this area. Commonly, this approach is already being used in the construction industry to determine compressive strength using, for example, information about the composition of a composite. Determination of pull-off strength can be approached in a similar way. In this paper, the ML model presented can be used to predict the pull-off strength of resin coatings containing granite powder and linen fibers. To obtain satisfactory results, the selected ML algorithms were analyzed on a database consisting of 140 sets of parameter values containing information about the composition of the resin coating. Indices indicating high performance (R = 0.885; RMSE = 0.138; MAPE = 3.72%) were obtained by a model based on the random forest (RF) algorithm containing 160 trees with a depth of 10 nodes. A comparison of the predicted fb pull-off strength with the strength determined by in-situ tests was developed. The results suggest that using artificial intelligence to determine the fb of resin coatings is a promising alternative.