Loading...

Reducing part Warpage by Optimizing Process Parameters in Selective Laser Sintering

Ahmadi Dastjerdi, Ali | 2015

724 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 48335 (08)
  4. University: Sharif University of Technology
  5. Department: Mechanical Engineering
  6. Advisor(s): Movahhedy, Mohammad Reza; Akbari, Javad
  7. Abstract:
  8. One of the newest methods of rapid prototyping is SLS . In this day, SLS is more common in compare others methods, because it uses little time and man does not nearly interfere in the producing stages of part which is made by SLS method. Also, sophisticated parts, which cannot be built by other methods, can be produced by SLS method. One of the disadvantages of SLS method is low dimensions accuracy of parts produced in this way. This problem has two major reasons: Shrinkage and Warpage. These two phenomena are created because of ununiformed temperature distribution. Others features which have effect on dimensions accuracy are residual stress, porosity and removing space of between powder particles, etc. As a result of that, many researches, included are conducted about improving dimension accuracy of parts produced by SLS. In this project, warpage of part produced by SLS is reduced by optimizing process parameters. At first, sintering process is modeled for a single layer by using finite element software ABAQUS. Then, experiment is done in order to validate the result of the model. After that, 40 tests with different parameters involved scanning speed, laser power, surrounding temperature of working, layer thickness and pattern scan, are modeled with ABAQUS. The results of these tests are used as training and test sets for the artificial neutral network. A back-propagation neural network is constructed as the surrogate model to formulate the mapping between the design variables and the objective function. Then, Genetic Algoritm is applied to optimize the neural network function. Finally, influences of each parameter on warpage are survived by help of neural network function. The results illustrate that the scanning pattern in which changes its scanning direction frequently by 908 at every turn reduced the warpage of layer. Although many parameters influence warpage, in constant state whose name is normal, more scanning speed and more surrounding temperature of working lead to less warpage of part. On the other hand, less layer thickness, less laser power depending on pattern scan may lead to less warpage of part
  9. Keywords:
  10. Waviness ; Sintering ; Neural Network ; Genetic Algorithm ; Process Parameters ; ABAQUS Software ; Rapid Prototyping ; Selective Laser Sintering (SLS)Method

 Digital Object List

 Bookmark

No TOC