2024 : 4 : 29
Ehsan Jahani

Ehsan Jahani

Academic rank: Associate Professor
ORCID:
Education: PhD.
ScopusId:
Faculty: Faculty of Technology and Engineering
Address: university of mazandaran
Phone: 09358549107

Research

Title
Evaluation of cracks in polymer samples with dual key hole notch using Artificial Neural Network (ANN) and Long Short-Term Memory (LSTM)
Type
Thesis
Keywords
(loads , matlab, angles , Artificial Neural network (ANN) , Long Short-Term Memory (LSTM) ,ABAQUS, Graphite,Keyhole, Notches)
Year
2023
Researchers Ehsan Jahani(PrimaryAdvisor)، Naser Kordani(PrimaryAdvisor)، Luma Ali Shahid Ali Almusawi(Student)

Abstract

Fracture assessment for serrated components based on the stress energy density criterion using different angles and loads, as it is used in several fields, including structural and mechanical, and to show the fracture behavior [1] . Taking into account the chance of cracking in the structural design, there are two suggested methods. One of the methods is universal and involves crack preface and crack spread , the other is more conservative and deal only with crack preface. Where several researches used slit shapes, including circular or v-shaped in terms of the effects on it. Several data were provided to be analyzed using the MTLAB program , using the artificial neural network(ANN), and (LSTM) [2]. When joining, with circumferentially cracked round bar piece, the fracture is a mixed pattern fracture response to a wide range of brittle materials. When dealing with circumferentially cracked round bar samples and mixed-pattern fracture response to a wide range of brittle materials and the effects on them, such as in the case of severe climatic conditions at the site such as high wind speed and in extremely hot or dry conditions, good workmanship , equipment and materials must be ensured at all times and the handling must be done properly[3]. Cracks in the component structure are very weak points, high risk of causing a brittle fracture; have been born cracks or lead to crack fatigue start. When the fissure is not apparent and a brittle fracture is expected, a conventional stress, and stress factors cannot be used, and a method for evaluating the max loads represented by a fissure component can persist is needed. The fracture begins systematically or unsystematically from the tip of the notch on the starting line , These are the points where the main and main stressors are at their maximum and as well as maximum shear stress and stress energy density[4,5]. On the contrary, the point at which the fracture begins in the mixed mode different from case to case because it depends on the type of the crack, the loads mode and the tension angle. Where different methods were used to verify this hypothesis the coherent area model and the samples based on type the stress (energy density) on the control (volume and the angle) of influence of the loads and the amount of loads on each slit. And all the methods allow us to evaluate cracks in different mix mode, when the material behavior can be assumed to be linear elastic. this is Wide degrees of experiential support with the use of simulations with finite element software to test theoretical predictions, the experiential program was made by changing the geometry of the sample, thus ensuring a wide range of loading modes. Theoretical predictions were compared with more than one model from data from multiple static tests performed on gear samples taken from a recent contribution from the author [6], the results will are evaluated twice on normal artificial neural networks and long-term memory (LSTM).