Prediction of Marshall Parameters of Modified Bituminous Mixtures Using Artificial Intelligence Techniques

Khuntia, Sunil; Das, Aditya Kumar; Mohanty, Monika; Panda, Mahabir

doi:10.1260/2046-0430.3.3.211

Cited by 21 publications

(11 citation statements)

References 12 publications

(15 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Between the two different models developed within the scope of the study, the ANN model provides better results. However, the R 2 values obtained with the estimations made by this model gave lower predictive values for some parameters (MS, F) while providing better estimates for the remaining parameters compared to [60] on the estimation of the Marshall parameters of bituminous mixtures using artificial intelligence techniques. One of the reasons for this performance is the attempt to estimate seven different parameters using only two input values, AC and HL.…”

Section: Figure 18 -Effect Of Hl On Mq and Air Voidsmentioning

confidence: 82%

Modelling the Effects of Hydrated Lime Additives on Asphalt Mixtures by Fuzzy Logic and ANN

2019

View full text Add to dashboard Cite

In this study, Marshall design test parameters of hot mix asphalt samples with various rates of Hydrated Lime (HL) content were modelled using Fuzzy Logic (FL) and Artificial Neural Networks (ANN). HL was used as an additive material in asphalt mixtures and it affects the properties of the mixture. The effect of this material varies depending on the rate of use and the asphalt content of the mixtures. With the Marshall Stability test, optimal Asphalt Content (AC) ratios in the mixtures were obtained. The effect of the HL additive, which was introduced precisely in the mixtures in a wide range, on the Marshall parameters and depending also on the asphalt content was investigated. For this purpose, 15 Marshall design sets were prepared by decreasing the ratio of the mineral filler in the mixture starting from 6.8% by weight, by 0.5% intervals, and replacing it with the same ratio of HL. In addition, 45 control samples were produced for soft-computation. Marshall test results showed that the use of HL additive with lower amounts in the mixtures yields better results compared to higher rates in terms of the material properties. The Marshall test results were used to develop the FL and ANN models. The models which were developed produced acceptable estimations of the mixture parameters.

show abstract

Section: Figure 18 -Effect Of Hl On Mq and Air Voidsmentioning

confidence: 82%

Modelling the Effects of Hydrated Lime Additives on Asphalt Mixtures by Fuzzy Logic and ANN

2019

View full text Add to dashboard Cite

show abstract

“…Tablo 1 incelendiğinde akma için elde edilen regresyon katsayılarının (R 2 ) % 13,5 ile % 25,7 arasında değiştiği görülmektedir. Küçük veri gruplarıyla yapılan çalışmalarda akmayı tahmin etmek üzere kurulan modeller ile daha yüksek R 2 değerleri elde edilmiştir [13,15]. Regresyon katsayıları, bağımlı değişkenlerdeki varyansın en fazla yüzde kaçının kullanılan bağımsız değişkenlerce açıklanabileceğini göstermekte olup elde edilen regresyon katsayıları bağımsız değişkenlerin bağımlı değişken olarak seçilen akmayı açıklamada yetersiz kaldıklarını göstermektedir.…”

Section: Araştirma Sonuçlari Ve Tartişma (Results and Discussion)unclassified

“…Bu teknikler kullanılarak bilinen özellikler ile bilinmeyen değişken tahmin edilmeye çalışılır. Marshall Testi ile belirlenen stabilite ve akmanın tahminine yönelik olarak laboratuvar ortamında üretilmiş, sınırlı sayıda ve homojen nitelikte gözlemlerden oluşan veri gruplarına dayanan çalışmalarda başarı sağlanmış olsa da [13][14][15] veri sayısı arttığında ve gözlemler çeşitlilik gösterdiğinde oluşacak durum bilinmemektedir. Bu çalışmada, 2001 yılında Ankara Büyükşehir Belediyesi sınırları dahilindeki aşınma ve binder tabakalarının imalatıyla ilgili bitümlü sıcak karışım uygulamalarına ait veriler derlenerek uygulama temelinde 4680 adet gözlem içeren büyük bir veri grubu oluşturulmuştur.…”

Section: Gi̇ri̇ş (Introduction)unclassified

Bi̇tümlü Sicak Karişimlara Ai̇t Akma Değeri̇ni̇n Regresyon Modelleri̇ İle Tahmi̇ni̇

Arslan

Bayırtepe

2018

Gazi Üniversitesi Fen Bilimleri Dergisi Part C: Tasarım Ve Teknoloji

View full text Add to dashboard Cite

ÖzBu çalışmada, bitümlü sıcak karışım özelliklerinden olan ve Marshall Testi ile deneysel olarak belirlenebilen akma değerinin malzeme ve karışım özellikleri temelinde regresyon modelleri ile tahmin edilmesi amaçlanmıştır. Bu kapsamda Ankara Büyükşehir Belediyesi sınırları dahilindeki aşınma ve binder tabakalarının inşaatına yönelik bitümlü sıcak karışım uygulamalarına ait veriler derlenmiş ve uygulama temelinde 4680 adet gözlem içeren veri grubu oluşturulmuştur. Veri grubuna regresyon, temel bileşen, kümeleme ve diskriminant analizleri uygulanmıştır. Kümeleme analizi, veri grubunun 4 ana kümeye ayrılabileceğini göstermiştir. Belirlenen kümelere ve çeşitli kombinasyonlarına uygulanan standart aşamalı regresyon analizleri ile akma bağımlı değişkenini tahmin etmek üzere geliştirilen modellerde regresyon katsayısı % 73,9'a kadar ulaşmıştır. AbstractIn this study, flow value of the bituminous hot mixtures that can be determined by Marshall Test has been aimed to be predicted by regression models based on material and mixture properties. In this context, data on bituminous hot mix applications for the manufacturing of wearing and binder layers within the borders of Ankara Metropolitan Municipality have been compiled and a data group containing 4680 observations has been created on the application basis. Regression, principal component, clustering and discriminant analyses have been applied on the data group. The clustering analysis shows that the data group can be divided into 4 main clusters. With the stepwise regression analyses carried out on the determined clusters and their various combinations, the regression coefficient of models, which are developed to predict the flow variable, has reached to 73.9 %.

show abstract

“…Tapkın et al (2010) employed ANNs for estimating Marshalltest results for dense asphalt mixes modified with different types and percentages of an additive, polypropylene [9]. Khuntia et al (2014) used the ANNs and support vector machine for estimating the Marshall test results for the polyethylene modified mixes [25]. Ozturk et al (2016) employed ANNs for estimating the volumetric characteristics of the asphalt mixes prepared following Marshall procedures [26].…”

Section: Historical Background Of Neural Network Applications In Pavement Engineeringmentioning

confidence: 99%

Prediction of the optimum asphalt content using artificial neural networks

Othman¹,

Abdelwahab²

2021

Metall Mater Eng

View full text Add to dashboard Cite

The performance of the asphalt mix is significantly influenced by the optimum asphalt content (OAC). The asphalt content is responsible for coating the aggregate surface and filling the voids between the aggregate particles. Thus, the aggregate gradation has a significant influence on the required asphalt content. The Marshall design process is the most common method used for estimating the OAC, and this process is called the asphalt mix design. However, this method is time consuming, labor intensive, and its results are subjected to variations. Thus, this paper employs the artificial neural network (ANN) to estimate the OAC from the aggregate gradation for the two most common gradations used in asphalt mixes in Egypt (3D, 4C). Results show that the proposed ANN can predict the OAC with a coefficient of correlation of 0.98 and an average error of 0.026%. As a result, a new approach for the Marshall test can be adopted using results of the proposed ANN, and only three specimens, instead of fifteen, are prepared and tested for estimating the remaining parameters. This approach saves the time, effort, and resources required for estimating the OAC. Additionally, the ANN was validated with previously developed models, and the ANN shows promising results.

show abstract

Prediction of Marshall Parameters of Modified Bituminous Mixtures Using Artificial Intelligence Techniques

Cited by 21 publications

References 12 publications

Modelling the Effects of Hydrated Lime Additives on Asphalt Mixtures by Fuzzy Logic and ANN

Modelling the Effects of Hydrated Lime Additives on Asphalt Mixtures by Fuzzy Logic and ANN

Bi̇tümlü Sicak Karişimlara Ai̇t Akma Değeri̇ni̇n Regresyon Modelleri̇ İle Tahmi̇ni̇

Prediction of the optimum asphalt content using artificial neural networks

Contact Info

Product

Resources

About