scholarly journals METHOD TO DEFINE MEASUREMENT UNCERTAINTY FOR DESIGN SPACE EXPLORATION IN ADDITIVE MANUFACTURING

2021 ◽  
Vol 1 ◽  
pp. 2067-2076
Author(s):  
Valentine Cazaubon ◽  
Audrey Abi Akle ◽  
Xavier Fischer
Keyword(s):  
Additive Manufacturing ◽  
Design Space Exploration ◽  
Design Space ◽  
Design Method ◽  
Optimal Solution ◽  
Feed Speed ◽  
Parameters Selection ◽  
Wire Feed Speed ◽  
Geometrical Defects ◽  
The Given

AbstractAdditive manufacturing is a process used for quick prototyping in industries. Geometrical defects are observed on printed parts. The aim of the paper is to propose a design method to implement measurements uncertainties into a Design Space for Additive Manufacturing parameters selection. To do so, two tests have been realized. The first test consists in determining the instrument’s uncertainty by measuring a standard length several times by an operator. The second test aim to determine the uncertainty within operators mesurement of geometric outputs (clad’s height, clad’s width, dilution’s height, dilution’s width and contact angle). Based on the results of our tests, uncertainties have been applied in our Design Space populated with 31 real printed clads. The uncertainties display with error bars on scatterplots allow to capitalize the knowledge for his/her exploration of the Design Space for future prints. The given information provides to ease the engineer to select the optimal solution (laser power, tool speed and wire feed speed) for his/her given Additive Manufacturing problematic among candidate points

scholarly journals Design Space Exploration for YOLO Neural Network Accelerator

Electronics ◽  
2020 ◽  
Vol 9 (11) ◽  
pp. 1921
Author(s):  
Hongmin Huang ◽  
Zihao Liu ◽  
Taosheng Chen ◽  
Xianghong Hu ◽  
Qiming Zhang ◽  
...  
Keyword(s):  
Neural Network ◽  
Design Space Exploration ◽  
Design Space ◽  
Space Exploration ◽  
Computation Time ◽  
Memory Bandwidth ◽  
Block Transmission ◽  
Transmission Strategy ◽  
The Neural Network ◽  
The Given

The You Only Look Once (YOLO) neural network has great advantages and extensive applications in computer vision. The convolutional layers are the most important part of the neural network and take up most of the computation time. Improving the efficiency of the convolution operations can greatly increase the speed of the neural network. Field programmable gate arrays (FPGAs) have been widely used in accelerators for convolutional neural networks (CNNs) thanks to their configurability and parallel computing. This paper proposes a design space exploration for the YOLO neural network based on FPGA. A data block transmission strategy is proposed and a multiply and accumulate (MAC) design, which consists of two 14 × 14 processing element (PE) matrices, is designed. The PE matrices are configurable for different CNNs according to the given required functions. In order to take full advantage of the limited logical resources and the memory bandwidth on the given FPGA device and to simultaneously achieve the best performance, an improved roofline model is used to evaluate the hardware design to balance the computing throughput and the memory bandwidth requirement. The accelerator achieves 41.99 giga operations per second (GOPS) and consumes 7.50 W running at the frequency of 100 MHz on the Xilinx ZC706 board.

scholarly journals Design space exploration of Convolutional Neural Networks based on Evolutionary Algorithms

2017 ◽  
Vol 3 (1) ◽  
Author(s):  
Abeer Al-Hyari ◽  
Shawki Areibi
Keyword(s):  
Neural Networks ◽  
Genetic Algorithms ◽  
Evolutionary Algorithms ◽  
Convolutional Neural Networks ◽  
Design Space Exploration ◽  
Design Space ◽  
Space Exploration ◽  
Handwritten Digit ◽  
Fully Connected ◽  
The Given

This paper proposes a framework for design space exploration ofConvolutional Neural Networks (CNNs) using Genetic Algorithms(GAs). CNNs have many hyperparameters that need to be tunedcarefully in order to achieve favorable results when used for imageclassification tasks or similar vision applications. Genetic Algorithmsare adopted to efficiently traverse the huge search spaceof CNNs hyperparameters, and generate the best architecture thatfits the given task. Some of the hyperparameters that were testedinclude the number of convolutional and fully connected layers, thenumber of filters for each convolutional layer, and the number ofnodes in the fully connected layers. The proposed approach wastested using MNIST dataset for handwritten digit classification andresults obtained indicate that the proposed approach is able to generatea CNN architecture with validation accuracy up to 96.66% onaverage.

scholarly journals A Parametric Study of Additive Manufacturing Process: TA6V Laser Wire Metal Deposition

2021 ◽  
pp. 15-20
Author(s):  
Valentine Cazaubon ◽  
Audrey Abi Akle ◽  
Xavier Fischer
Keyword(s):  
Additive Manufacturing ◽  
Laser Power ◽  
Pearson Correlation ◽  
Metal Deposition ◽  
Feed Speed ◽  
New Process ◽  
Wire Feed Speed ◽  
Subtractive Manufacturing ◽  
Working Principle ◽  
Wire Feed

AbstractAdditive Manufacturing has proven to be an economically and industrially attractive process in building or repairing parts. However, the major issue of this new process is to guarantee a mechanical behavior identical to the subtractive manufacturing methodologies. The work, presented in this paper, is centered on the Laser Wire Metal Deposition (LMD-w) method with the metallic alloy TA6V. Its working principle is to fuse a coaxial wire on a substrate with a laser as a heat source. To better understand the interaction between the input parameters (Laser Power, Wire Feed Speed and Tool Speed) and the clad geometry output variables (Height, Width and Contact Angle) and the substrate displacement, we have realized an experimentation. We printed 9 clads according Taguchi’s experimental design. Pearson correlation coefficient and Fisher test performed on the experimental measures showed as main result: Tool Speed is the parameter with the most significant influence on the output variables.

Design Space Exploration Using Cycle Accurate Simulator

2016 ◽  
pp. 66-79
Author(s):  
Arsalan Shahid ◽  
Bilal Khalid ◽  
Muhammad Yasir Qadri ◽  
Nadia N. Qadri ◽  
Jameel Ahmed
Keyword(s):  
Design Space Exploration ◽  
Design Space ◽  
Space Exploration ◽  
Optimal Solution ◽  
Access Time ◽  
Performance Improvements ◽  
Computing Platforms ◽  
On Chip ◽  
Number Of Cycles ◽  
Multilevel Cache

Multi-Processor System on Chip (MPSoC) architectures have become a mainstream technology for obtaining performance improvements in computing platforms. With the increase in the number of cores, the role of cache memory has become pivotal. An ideal memory configuration is always desired to be fast and large; but, in fact, striking to balance between the size and access time of the memory hierarchy is considered by processor architect. Design space exploration is used for performance analysis of systems and helps to find the optimal solution for obtaining the desired objectives. In this chapter, we explore two design space parameters, i.e., cache size and number of cores, for obtaining the desired energy consumption. Moreover, previously presented energy models for multilevel cache are evaluated by using cycle accurate full system simulator. Our results show that with the increase in cache sizes, the number of cycles required for application execution decreases, and by increasing number of cores, the throughput improve.

Process Development for a Robotized Laser Wire Additive Manufacturing

2017 ◽  
Author(s):  
Meysam Akbari ◽  
Yaoyu Ding ◽  
Radovan Kovacevic
Keyword(s):  
Additive Manufacturing ◽  
Process Parameters ◽  
Process Development ◽  
Laser Scanner ◽  
Travel Speed ◽  
Main Process ◽  
Feed Speed ◽  
Wire Feed Speed ◽  
Deposition Processes ◽  
Complicated Geometries

Additive manufacturing has attracted the attention of industries such as aerospace and automotive as well as the medical technology sectors in recent years. Among all metal-based additive techniques, laser metal wire deposition offers some advantages like shorter processing time, more efficient material usage, and a larger buildup envelop. It has been found that robotized laser/wire additive manufacturing (RLWAM) is a demanding process. A plethora of process parameters must be controlled compared to other laser-based metal deposition processes. The influence of main process parameters such as laser power, stepover increment, wire feed speed, travel speed and z-increment was investigated in this study to find the optimal values. Droplet formation, wire dripping, irregular deposition in the first layer, and deviation of the wire tip were also found to be the main obstacles throughout the process and practical solutions were proposed to deal with these issues. In this study, an 8-axis robot (6-axis arm robot with a 2-axis positioner) and a 4 kW fiber laser along with a wire feeder were integrated to print the different geometrical shapes in 3D. In order to verify the geometrical accuracy of the as-built part, the buildup was scanned using a portable 3D laser scanner. The 3D representation, the Standard Tessellation Language (STL) format obtained from the buildup, was compared with the original CAD model. The results show that RLWAM can be successfully applied in printing even complicated geometries.

scholarly journals Prediction and Optimization of Production Quantities in Innoson Manufacturing Extraction Plastic Product

2019 ◽  
pp. 1-11
Author(s):  
Ezeliora, Chukwuemeka Daniel ◽  
Okoye, Peter Chukwuma ◽  
U. Mbabuike, Ikenna
Keyword(s):  
Factorial Design ◽  
Design Space ◽  
Design Method ◽  
Signal To Noise Ratio ◽  
Optimal Solution ◽  
Coefficient Of Determination ◽  
Anambra State ◽  
Plastic Pipes ◽  
Surface Plot ◽  
Production Quantity

In this research, it focused on the prediction and optimization of the production quantity in Innoson Plastic Manufacturing Company, Nnewi, Anambra State, Nigeria. The research method used is the application of factorial design methods to model, to evaluate the best optimal solutions for the production quantity of extrusion plastic pipes in the aforementioned company. The analysis shows that the parameters used to model the production quantity are significant and the model produced is also significant with its coefficient of determination to be 0.9968 and the adjusted R-Squared is 0.9823. Adequate Precision measures the signal to noise ratio. A ratio greater than 4 is desirable. The ratio of 29.271 indicates an adequate signal. This model can be used to navigate the design space. The Model F-value of 68.99 implies the model is significant. There is only a 1.44% chance that an F-value this large could occur due to noise. Values of "Prob > F" less than 0.0500 indicate model terms are significant. The 3D surface plot shows the effect of the variables in production system. It describes the variations of the input and output parameters in production of plastic extrusion products. The factorial design method applied shows the optimal solution which revealed that the best quantity of the product that is necessary to produce in any given month is 14414.112 units of a 25mm diameter plastic pipes with the optimal desirability of 100%. The tool also shows that the pigment is almost not important in the optimization of the product production quantity due to its insignificant quantity. However, the results further revealed that the industry should be conscious of highly influence input variable during production.

Formation and improvement of surface waviness for additive manufacturing 5A06 aluminium alloy component with GTAW system

2018 ◽  
Vol 24 (2) ◽  
pp. 342-350 ◽  
Author(s):  
Haibin Geng ◽  
Jinglong Li ◽  
Jiangtao Xiong ◽  
Xin Lin ◽  
Dan Huang ◽  
...  
Keyword(s):  
Travel Speed ◽  
Feed Speed ◽  
Surface Waviness ◽  
Forming Process ◽  
Line Energy ◽  
Content Type ◽  
Wire Feed Speed ◽  
The Given ◽  
Wire Feed ◽  
Matching Relation

Purpose As known, the wire and arc additive manufacture technique can achieve stable process control, which is represented with periodic surface waviness, when using empirical methods or feedback control system. But it is usually a tedious work to further reduce it using trial and error method. The purpose of this paper is to unveil the formation mechanism of surface waviness and develop a method to diminish it. Design/methodology/approach Two forming mechanisms, wetting and spreading and remelting, are unveiled by cross-section observation. A discriminant is established to differentiate which mechanism is valid to dominate the forming process under the given process parameters. Findings Finally, a theoretical method is developed to optimize surface waviness, even forming a smooth surface by establishing a matching relation between heat input (line energy) and materials input (the ratio of wire feed speed to travel speed). Originality/value Formation mechanisms are revealed by observing cross-section morphology. A discriminant is established to differentiate which mechanism is valid to dominate the forming process under the given process parameters. A mathematical model is developed to optimize surface waviness, even forming a smooth surface through establishing a matching relation between heat input (line energy) and materials input (the ratio of wire feed speed to travel speed).

scholarly journals Study of Arc-wire and Laser-wire processes for the realization of Ti-6Al-4V alloy parts

2020 ◽  
Vol 321 ◽  
pp. 03002
Author(s):  
A. Ayed ◽  
G. Bras ◽  
H. Bernard ◽  
P. Michaud ◽  
Y. Balcaen ◽  
...  
Keyword(s):  
Additive Manufacturing ◽  
Travel Speed ◽  
Feed Speed ◽  
Laser Additive Manufacturing ◽  
Powder Bed ◽  
Wire Feed Speed ◽  
Manufacturing Techniques ◽  
Parameter Ranges ◽  
Wire Arc Additive Manufacturing ◽  
Wire Feed

Arc-wire or laser-wire additive manufacturing seems promising because it allows large parts to be produced with significant deposition rates (ten times higher than powder bed additive manufacturing), for a lower investment cost. These additive manufacturing techniques are also very interesting for the construction or the repair of parts. A versatile 3D printing device using a Wire Arc Additive Manufacturing (WAAM) station or laser device Wire Laser Additive Manufacturing (WLAM) for melting a filler wire is developed to repair and build large titanium parts. The final objectives of the study are to optimize the process parameters to control the dimensional stability, the metallurgical and mechanical properties of the produced parts. In this paper, an experimental study is carried out to determine the first order process parameter ranges (synergic law, laser power, wire feed speed, travel speed) appropriate for these two techniques, for repair or construction parts on Ti-6 Al-4V.

A Wire Feed Control System Based on Detecting Induced Electromotive Force of the Motor

2012 ◽  
Vol 569 ◽  
pp. 781-784
Author(s):  
Zhen Yang Lu ◽  
Peng Fei Huang ◽  
Yong Gang Zhang ◽  
Shao Jun Bai
Keyword(s):  
Electromotive Force ◽  
Feed Speed ◽  
Motor Speed ◽  
Feed System ◽  
Motor Current ◽  
Feed Control ◽  
Wire Feed Speed ◽  
The Given ◽  
The Relationship ◽  
Wire Feed

A stable and even wire feed speed is a guarantee of good quality welding. A new wire feed system, without any speed-measuring equipment, has been designed. And the motor current of this system works in a discontinuous state. When the motor current is zero, it detects the induced electromotive force; acquires the motor speed according to the relationship between the induced electromotive force and the motor speed; calculates and compares with the given speed then changes the duty cycle of the voltage added on the motor by PWM technology, that control the motor speed. Then a stable and even wire feed speed can be achieved. Verified by testing resistance, the change rate of the wire feed speed is slightly, which proves that the new system has good mechanical properties and a stable speed. Besides, it runs reliably.

Sign in / Sign up

Export Citation Format

Share Document