Hasil Pencarian  ::  Simpan CSV :: Kembali

"

---

ABSTRACTThe aim of this research is to reproduce the experimental result using finite element approach with damage model coupled to creep. The calibration of coupling between creep and local damage model associated with size effect law proposed by Bazant are presented in this thesis. Its behavior was investigated through two kinds of studies, the loading rate effect and residual capacity test. Three different sizes of beam, which are geometrically similar specimens, were simulated in three point bending test and creep test divided into load level test and residual capacity test. Numerical simulation was doneusing existing finite element code Aster_ Code developed by EDF (Eiectricite de France) for coupling between local damage based on bi-linear elasto-damage model and creep based on Benboudjema theory. Result show that the finite element code is capable to reproduce the experimental result qualitatively. The interaction between creep and local damage is shown through size effect plot by giving the behavior shift to the right, which means that the materials become more brittle when creep appears.
"

"Sebagai karakter yang mengakibatkan variabilitas properti mekanis, sifat heterogenitas beton perlu dipertimbangkan karena mempengaruhi perilaku material. Salah satu metode untuk memperhitungkan sifat tersebut adalah random field yang digenerasi dengan Turning Band Method (TBM) pada aplikasi berbasis finite element, Cast3M. Objek pemodelan pada penelitian ini adalah beton polos berukuran 15 x 15 x 15 cm3, beton polos 10 x 10 x 50 cm3, dan beton dengan tulangan tunggal 10 x 10 x 50 cm3. Studi numerik dilakukan dengan mengaplikasikan beban tekan pada setiap sampel kubus dan beban tarik pada sampel balok. Model kerusakan Mazars diimplementasikan untuk mempelajari perilaku keruntuhan pada masing-masing sampel terhadap variasi parameter ukuran mesh dan panjang korelasi. Hasil pemodelan menunjukkan bahwa penggunaan TBM sangat berpengaruh pada pola keruntuhan dari setiap sampel. Parameter ukuran mesh dan panjang korelasi memiliki pengaruh yang signifikan terhadap propagasi dan sensitivitas kerusakan serta estimasi properti mekanis. Pemodelan dengan ukuran mesh 1 cm dan 2,5 cm terhadap masing-masing panjang korelasi menghasilkan pola keruntuhan yang bervariasi. Perbandingan hasil pemodelan beton polos 15 x 15 x 15 cm3 yang diuji tekan dengan studi eksperimental menunjukkan bahwa pola keruntuhan dengan mesh 1 cm mendekati kerusakan pada sampel pengujian. Sedangkan, perbandingan hasil pemodelan beton tulangan tunggal 10 x 10 x 50 cm3 dengan sampel pengujian di laboratorium menunjukkan bahwa penggunaan panjang korelasi 1 cm dan 3 cm memiliki pola kerusakan yang memiliki kemiripan masing-masing terhadap sampel ER1 dan ER2.

---

As a character that results in the variability of mechanical properties, the heterogeneity of concrete needs to be considered because it affects the behavior of the material. One of the methods to take into account these properties is a random field that is generated by the Turning Band Method (TBM) on a finite element based application, Cast3M. The modeling objects in this study were plain concrete with dimension of 15 x 15 x 15 cm3, plain concrete 10 x 10 x 50 cm3, and concrete with single reinforcement 10 x 10 x 50 cm3. Numerical studies were carried out by applying compressive loads to the cube sample and tensile loads to the beam samples. The Mazars damage model is implemented to study the damage behavior of each sample with varying parameters of mesh dimension and correlation length. The modeling results show that the use of TBM greatly influences the failure pattern of each sample. Parameters of mesh size and correlation length have a significant effect on the propagation and damage sensitivity as well as the estimation of mechanical properties. Modelling with 1 cm and 2.5 cm size of mesh for each correlation length produces a variety of damage patterns. Comparison of modelling results of plain concrete 15 x 15 x 15 cm3 tested in compression with experimental studies reveals that the pattern of failure with a mesh of 1 cm is close to the damage in the laboratory test sample. Meanwhile, a comparison of the results of the 10 x 10 x 50 cm3 single-reinforced concrete modelling with samples tested in the laboratory indicates that the use of a correlation length of 1 cm and 3 cm has a pattern of damage that has similarities to the ER1 and ER2 samples, respectively."