Hazim El-Mounayri

NC end milling optimization using evolutionary computation

Abstract Typically, NC programmers generate tool paths for end milling using a computer-aided process planner but manually schedule “conservative” cutting conditions. In this paper, a new evolutionary computation technique, particle swarm optimization (PSO), is proposed and implemented to efficiently and robustly optimize multiple machining parameters simultaneously for the case of milling. An artificial neural networks (ANN) predictive model for critical process parameters is used to predict the cutting forces which in turn are used by the PSO developed algorithm to optimize the cutting conditions subject to a comprehensive set of constraints. Next, the algorithm is used to optimize both feed and speed for a typical case found in industry, namely, pocket-milling. Machining time reductions of up to 35% are observed. In addition, the new technique is found to be efficient and robust.

Selecting an artificial neural network for efficient modeling and accurate simulation of the milling process

In this paper, two supervised neural networks are used to estimate the forces developed during milling. These two Artificial Neural Networks (ANNs) are compared based on a cost function that relates the size of the training data to the accuracy of the model. Training experiments are screened based on design of experiments. Verification experiments are conducted to evaluate these two models. It is shown that the Radial Basis Network model is superior in this particular case. Orthogonal design and specifically equally spaced dimensioning showed to be a good way to select the training experiments.

Prediction of surface roughness in end milling using swarm intelligence

A new technique from EC (evolutionary computation), PSO (particle swarm optimization), is implemented to model the end milling process and predict the resulting surface roughness. Data is collected from CNC cutting experiments using DOE approach. The data is used for model calibration and validation. The inputs to the model consist of feed, speed and depth of cut while the output from the model is surface roughness. The model is validated through a comparison of the experimental values with their predicted counterparts. A good agreement is found. The proved technique opens the door for a new, simple and efficient approach that could be applied to the calibration of other empirical models of machining.

Optimized CNC end-milling: A practical approach

This work proposes and implements a solution to achieve integrated product development that is seamlessly interfaced with optimized CNC end milling. The proposed solution is accurate and general and uses innovative techniques that are new to the field of metal-cutting modelling, simulation and optimization. In addition, the proposed solution is practical and its implementation can be easily and rapidly transferred to industry. First, the current work uses a novel and generic methodology for modelling the geometric aspect of the milling process. Second, techniques from artificial intelligence are introduced to model and optimize the milling process. Next, CAD/CAM and standard engineering development tools are adopted as the implementation tools and platform, and industry and CAD/CAM standards are used as the representations and standards in the development. Finally, the CAD/CAM platform is seamlessly interfaced with CNC production through the development of post-processors for typical CNC controllers found in industry. The system is demonstrated and verified for a typical industrial machining scenario, namely Pocket milling. The results demonstrate the system validity, applicability and accuracy.

VIRTUAL TRAINING ENVIRONMENT FOR A 3-AXIS CNC MILLING MACHINE

A virtual training environment for a 3-axis CNC milling machine is presented. The key elements of the environment are: (a) textured 3D photo-realistic virtual models of the machines and lab; (b) machine simulator for the machines’ controls and moving parts; (c) semi-empirical model of the machining operation; (d) hierarchical knowledge-base for process training; (e) unstructured knowledge-base for lecture delivery; (f) natural-language human-like intelligent virtual tutors. Applications of the AVML include: training students to operate manufacturing machines in a safe environment, allowing students and researchers to view and interact with highly accurate physical simulation of manufacturing machines, and optimization of the manufacturing process plan by testing various plans on the virtual machine before actual machining. The virtual training environment will significantly reduce the cost and increase the accessibility and safety of advanced manufacturing training.Copyright

MOLECULAR DYNAMICS SIMULATION OF NANOMETRIC CUTTING

Molecular Dynamics (MD) simulations of nanometric cutting of single-crystal copper were conducted to predict cutting forces and investigate the mechanism of chip formation at the nano-level. The MD simulations were conducted at a conventional cutting speed of 5 m/s and different depths of cut (0.724–2.172 nm), and cutting forces and shear angle were predicted. The effect of tool rake angles and depths of cut on the mechanism of chip formation was investigated. Tools with different rake angles, namely 0°, 5°, 10°, 15°, 30°, and 45°, were used. It was found that the cutting force, thrust force, and the ratio of the thrust force to cutting force decrease with increasing rake angle. However, the ratio of the thrust force to the cutting force is found to be independent of the depth of cut. In addition, the chip thickness was found to decrease with an increase in rake angle. As a consequence, the cutting ratio and the shear angle increase as the rake angle increases. The dislocation and subsurface deformation in the workpiece material were observed in the cutting region near the tool rake face. The adhesion of copper atoms to the diamond tool was clearly seen. The same approach can be used to simulate micromachining by significantly increasing the number of atoms in the MD model to represent cutting depths in the order of microns.

A generic and innovative approach for integrated simulation and optimisation of end milling using solid modelling and neural network

Accurate and integrated modelling and optimisation of 2½ - axis end milling using standard tools and novel approaches is presented here. The resulting system provides a practical bridge between current state-of-the-art CAD/CAM and optimised CNC production. First, a novel and generic approach for extracting the in-cut geometry is implemented using ACIS® open architecture solid modeller. Next, an ANN (artificial neural network) model is designed and implemented for force prediction. Finally, a new technique for optimising the cutting parameters is applied and verified. The technique is based on reverse mapping of the ANN model for cutting force estimation. As such, the machining process is optimised directly using the learned neural network model by adjusting the net inputs to optimal values that minimise a specified objective function subject to cutting constraints. The optimisation results are used to update the initial CL (cutter location) data file to produce an optimum NC (numerically controlled) code. The proposed approach is demonstrated and verified through a case study to show its validity, practicality, and applicability.

Design simulation of twisted cord-rubber structure using ProE/ANSYS

Abstract A three-dimensional model of a twisted cord embedded in rubber matrix was investigated to estimate the interface stresses. CAD model of the twisted cord geometry through blended feature was carried out using Pro/ENGINEER software. Twisted cord model representative of a typical cord was developed and analyzed for a realistic analysis of cords used in cord–rubber structure. Finite element analysis was performed on the models under axial, and combined axial and lateral loading using the ANSYS software. A circular cord with surrounding rubber model was also analyzed and the results obtained were validated with other studies that exist in the literature. The stresses and deformations obtained from the finite element analysis are shown for various cases illustrating the effects of twist, rubber modulus, and non-circular cord.

Molecular Dynamics Simulation of Nanometric Machining Under Realistic Cutting Conditions

Molecular Dynamics (MD) simulations of nanometric machining of single-crystal copper were conducted at a conventional cutting speed (5m/s) and different depths of cut (0.724 – 2.172 nm). The simulations were carried out to predict cutting forces and investigate the mechanism of chip formation at the nano level. The effect of tool rake angles and depths of cut on the mechanism of chip formation were also investigated. Tools with different rake angles, namely 0°, 5°, 10°, 15°, 30°, and 45°, were used. It was found that the cutting force, thrust force, and the ratio of the thrust force to cutting force decrease with increasing rake angle. However, the ratio of the thrust force to the cutting force is found to be independent of the depth of cut.Copyright

WEB-BASED MULTIMEDIA LECTURE DELIVERY SYSTEM WITH TEXT-TO-SPEECH AND VIRTUAL INSTRUCTORS

A web-based multimedia lecture delivery system is presented. The system provides natural-language instruction with synchronized: naturally sounding text-to-speech, written highlighted text, and animated 2D and 3D graphics. A nearphotorealistic animated human avatar can present the lecture with synchronized gestures and lip-synching. The course is presented using a hierarchical structured outline. The learner can ask the virtual instructor questions using natural-language speech or typed text. The instructor first tries to answer the question from the course content. If no information is found then a web search is performed.