节点文献

光固化三维打印快速成形关键技术研究

Research on the Key Techniques of UV-curing Three Dimensional Printing

【作者】 刘厚才

【导师】 莫健华;

【作者基本信息】 华中科技大学 , 材料加工工程, 2009, 博士

【摘要】 随着经济的全球化,市场竞争将越来越激烈,产品能否快速响应市场已经成为企业竞争胜败的关键,因此,对能够提高新产品开发成功率、降低产品开发时间和开发成本的快速成形设备的需求将越来越大。同其它主流快速成形技术相比,三维打印快速成形技术在成形速度、可成形材料种类、设备价格、运行和维护成本等方面具有明显优势,具有良好的发展潜力和广阔的应用前景,代表着快速成形技术的发展方向。根据三维打印快速成形技术发展的趋势,结合我国工业生产的需要,本文主要开展了光固化三维打印快速成形技术的研究,致力于开发具有自主知识产权的经济型光固化三维打印快速成形设备。论文的主要研究内容如下:对喷射液滴在成形表面上的扩展和再融合过程进行了理论分析,经计算确定了在作者设计的光固化三维打印快速成形系统中,XY平面上的实际打印分辨率要求应不低于300dpi,Z轴方向上的最小成形层厚为0.011mm。对光固化三维打印快速成形工艺中材料固化过程进行了理论分析,得到了最大固化深度与紫外灯的运动速度之间的关系,以及打印头沿X轴运动速度与切片厚度应满足的关系。基于保证使用性能和降低设备硬件成本的思想,进行了光固化三维打印快速成形原型机的机械系统、控制硬件系统和控制软件系统的总体规划和结构设计,为开发出具有自主知识产权的光固化三维快速成形设备奠定了基础。根据光固化三维打印快速成形技术中单层成形加工所需要的控制状态数,采用了2张单色位图来表示光固化三维打印快速成形系统中的层面数据,一张表示实体数据信息,另一张表示支撑数据信息。位图数据的生成采用扫描线填充方式,结合快速成形系统中CAD模型经切片后得到轮廓环数据的特点,提出了一种基于轮廓边的位图生成算法。该位图生成算法很好地解决了用常规的活性边表法难以判断和处理的当扫描线通过水平轮廓边所带来的奇点问题。实现了基于该方法的光固化三维打印快速成形层面实体位图生成算法和支撑位图的快速求解算法。根据光固化三维打印快速成形系统中层面位图数据的特点和常用无损位图数据压缩方法的不足,设计了一种基于字节的位图数据无损压缩方法。该位图压缩方法充分利用了光固化三维打印快速成形系统中层面位图数据的特点,压缩比高,经过对多个CAD模型的位图数据的压缩比较,最小压缩比为12.1:1,最大压缩比达53.9:1。该位图压缩方法在对位图数据进行压缩和解压的过程中,不需要进行复杂的计算,速度快,耗时少,对光固化三维打印快速成形系统这种有成百上千层位图数据需要处理的场合不会构成瓶颈,较好地满足了光固化三维打印快速成形系统中对数据压缩的要求。分析了多目标优化基准算法——第二代非支配排序遗传算法的不足,提出了相应的改进措施,在此基础上提出了一种新的多目标优化算法——非支配排序均匀遗传算法。通过仿真测试表明非支配排序均匀遗传算法在求解精度、计算效率和避免算法陷于局部最优解方面具有明显优势。研究了光固化三维打印快速成形工艺中零件制作方向对制件表面质量、所需支撑和制作时间的影响。提出了将基于Pareto最优原理的多目标优化方法用于零件制作方向优化求解的新思路。用本文提出的非支配排序均匀遗传算法进行零件制作方向的优化计算,表明其不仅可以求出比单目标优化方法更优的零件制作方向,而且在一次优化计算中就可得到多个在一定权重分配下零件的最优制作方向。

【Abstract】 With the economic globalization, the market’s competition will be more and more ruthless and time-to-market of the products is a key success factor of the corporations. As a result, the enterprise will need more and more Rapid Prototyping (RP) equipments, which can increase the successful ratio of developing a new product, decrease the time and cost required for a new product development. Compared with other RP technology, Three Dimensional Printing (3DP) has advantages in many aspects such as processing velocity, variety of processing materials, price of equipment, cost of running and maintenance of equipment. Consequently, 3DP has better developing potentialities and expansive applying foreground, and represents the developing direction of RP technologies. According to the developing trending of 3DP and the need of corporations in China, the UV-curing 3DP was researched in this paper, and the object was to develop 3DP system with our own knowledge right. The main content of research in the paper is as following:The spreading and reconsolidating of jetting droplets on the processing surface were studied in theory. It was calculated that the minimum resolution factor of printing in X-Y surface should be 300dpi in the developed the rapid prototyping system of UV-curing 3DP, and the minimum processing thickness of Z axle was 0.011mm. The solidified process of materials of photosensitive resin in UV-curing 3DP was studied in theory, the function of the maximum solidified depth with the moving velocity of the UV-lamp and the required relationship of slicing thickness with the moving velocity of printhead along X axis were educedBased on the idea of decreasing the cost of the hardware when the using performance was ensured, the project and structure of the mechanical system, controlling hardware system and controlling software system in the original UV-curing 3DP equipment were designed. It was established the groundwork of the first UV-curing 3DP equipment with our own knowledge right to be developed.According the number of controlling states needed in the single layer’s processing of UV-curing 3DP, two pieces of bitmap of black and white were adopted to express the layered data, one expressing the solid bitmap and the other expressing the support bitmap. The scanning beam way was used to creating the bitmap data. Combined with the character of the layered contour of slicing the CAD model in rapid prototyping system, a new algorithm of creating bitmap based on the contour edge was proposed. The new algorithm could dispose expediently all the odd points, especially the odd points produced when the horizontal contour edge was on the scanning beam, which were difficult to be distinguished and disposed by the classical algorithms. Based on the new algorithm, the algorithm of creating layered solid bitmap in UV-curing 3DP system and the algorithm of fast seeking the layered support bitmap were achieved.According the character of the layered bitmap in the UV-curing 3DP and the shortage of the lossless compressing methods of bitmap data in common use, one new compressing method of bitmap data based on byte was designed. The compressing way made the best of the character of bitmap data in UV-curing 3DP, the compressing rate of the method is high. It was showed that the minimum compressing rate was 12.1:1 and the maximum compressing rate was 53.9:1 when many CAD models’ bitmap data were compressed compared with the noncompressed bitmap data. The compressing way did not need complicated computing in the compressing and decompressing process, its compressing and decompressing velocity was fast, which did not form bottleneck of the condition of having thousands pieces of bitmap data to be disposed and which was well suitable for the data compressing in the UV-curing 3DP systemThe shortages of the classical multi-objective optimizating algorithm (MOGA) of NSGA II were analyzed and the improving methods were put forward. One new MOGA named nondominated sorting uniformly generic algorithm (NSUGA) was proposed, which was better in the precision of seeking pareto solutions, the efficiency of computing and the avoidance of gaining local optimal solutions than the MOGA of NSGA II.The effect of part building orientation on the part’s surface quality, supporting area and building time were studied in UV-curing 3DP. One new idea of optimizating part building orientation using multi-objective optimizating way based on the pareto optimal principle was proposed. Using NSUGA to optimize the part building orientation, not only the better solutions were found than the single-objective optimizating way, but also many pareto optimal solutions could be gained once.

节点文献中: