SBIR Phase II: Creating Functionally Decomposed Surface Models from Measured Data

这个小企业创新研究第二阶段项目涉及从测量数据重建复杂的自由形状的问题。 Raindrop Magic的主要兴趣是生成结构良好的高质量CAD模型。 有几种技术可以实现这一目标;不幸的是，自动表面处理系统只提供粗略的近似值，不能捕获原始设计意图，而手动分割方法不是很稳定，需要繁琐的工作。 使用功能分解，对象被构建为由较小的从属特征曲面（如圆角或扫掠曲面）连接的大型独立主曲面的集合。 在第一阶段，半自动方法被详细阐述，以创建良好的分段曲线网。 利用不同特征类型的特定属性，研究小组提出了计算每种特征的最佳表面表示的算法。 在第二阶段，该团队设想将他们的理论结果转换和扩展为强大而有效的计算算法。 提出了五个子系统：曲面指示器，约束拟合，曲线跟踪，光顺和几何拟合。 开发了新的核心技术来创建不同的几何实体，最终将其集成以获得高质量的表面模型。 这项技术将大大缩短相关工业设计和制造过程的交付时间，并生产出美观的产品，对整个社会产生积极影响。这项技术在两个关键市场领域产生了广泛的影响：逆向工程和先进的表面处理。 在研究前沿，该项目加深了对计算机辅助几何建模与扫描数据的理解，这一领域尚未受到大型CAD公司的关注，但却是一个活跃的研究领域。 它结合了离散和连续数学的知识，并利用这两种方法的优势。在技术方面，它引入了一种新的模式，通过更简单的操作，以更好的工程功能和先进的表面处理，显着改善当前的逆向工程商业系统。 主要应用将是产品设计，包括汽车，航空航天，消费品和医疗器械。 改进后的产品将有助于美国制造业在世界市场上更具竞争力，为引入按需设计和按需工程服务提供了一种途径。 拟议中的项目将帮助美国公司增加以客户为中心的生产，并减少产品迭代之间的时间。

This Small Business Innovation Research Phase II project deals with the problems of reconstructing complex free-form shapes from measured data. Raindrop Magic's primary interest is to produce well-structured, high-quality CAD models. Several techniques exist to reach this goal; unfortunately, automatic surfacing systems provide only rough approximations and do not capture the original design intent, while manual segmentation methods are not very stable and require tedious work. Using functional decomposition, objects are built up as a collection of large, independent primary surfaces being connected by smaller, dependent feature surfaces, such as fillets or swept surfaces. In Phase I, semi-automatic methods were elaborated to create good segmenting curve nets. Exploiting the specific properties of different feature types, the research team proposed algorithms to compute optimal surface representations for each. In Phase II, the team envisions transforming and extending their theoretical results into robust and efficient computational algorithms. Five subsystems are proposed: Surface-Indicators, Constrained-Fitting, Curve-Tracing, Fairing, and Feature-Fitting. New core technologies are developed for creating different geometric entities, which are eventually integrated to obtain high-quality surface models. This technology should significantly shorten lead-time in related industrial design and manufacturing processes and produce aesthetic objects, having a positive impact on the whole society.The proffered technology has broader impacts in two key market sectors: reverse engineering and advanced surfacing. At the research front, the proposed project deepens the understanding of computer-aided geometric modeling working with scan data, a field that has not received much attention from the large CAD companies, but is an active area of research. It combines the knowledge of both discrete and continuous mathematics and takes advantage of the strength of both approaches. On the technology front, it introduces a new paradigm that will significantly improve the current commercial systems of reverse engineering with better engineering features and advanced surfacing through simpler operations. The main applications will be product design, including automotive, aerospace, consumer products, and medical devices. The improved product will help the US manufacturing industry to be more competitive in the world market, providing a way to introduce design on demand and engineering on demand services. The proposed project will help US companies to increase customer-focused production and reduce the time between product iterations.