快速成型技術(shù)采用離散/堆積成型原理，根據(jù)三維CAD模型，對(duì)于不同的工藝要求，按一定厚度進(jìn)行分層，將三維數(shù)字模型變成厚度很薄的二維平面模型。再將數(shù)據(jù)進(jìn)行一定的處理，加入加工參數(shù)，產(chǎn)生數(shù)控代碼，在數(shù)控系統(tǒng)控制下以平面加工方式連續(xù)加工出每個(gè)薄層，并使之粘結(jié)而成形。實(shí)際上就是基于“生長(zhǎng)”或“添加”材料原理一層一層地離散疊加，從底頂完成零件的制作過程。快速成型有很多種工藝方法，但所有的快速成型工藝方法都是一層一層地制造零件，所不同的是每種方法所用的材料不同，制造每一層添加材料的方法不同。

Rapid prototyping technology adopts the principle of discrete/stacked molding. Based on the three-dimensional CAD model, for different process requirements, it is layered according to a certain thickness, transforming the three-dimensional digital model into a two-dimensional flat model with very thin thickness. Further process the data, add machining parameters, generate CNC codes, and continuously process each thin layer in a flat machining manner under the control of the CNC system, and make it bond and form. In fact, it is based on the principle of "growth" or "addition" of materials, layer by layer discrete stacking, completing the manufacturing process of parts from bottom to top. There are many process methods for rapid prototyping, but all of them involve manufacturing parts layer by layer. The difference is that each method uses different materials and the method of adding materials to each layer is different.

快速成型的工藝過程原理如下

The process principle of rapid prototyping is as follows

三維模型的構(gòu)造:在三維CAD設(shè)計(jì)軟件中獲得描述該零件的CAD文件。一般快速成型支持的文件輸出格式為STL模型，即對(duì)實(shí)體曲面做近似的所謂面型化(Tessellation)處理，是用平面三角形面片近似模型表面。以簡(jiǎn)化CAD模型的數(shù)據(jù)格式。便于后續(xù)的分層處理。由于它在數(shù)據(jù)處理上較簡(jiǎn)單，而且與CAD系統(tǒng)無(wú)關(guān)，所以很快發(fā)展為快速成型制造領(lǐng)域中CAD系統(tǒng)與快速成型機(jī)之間數(shù)據(jù)交換的標(biāo)準(zhǔn)，每個(gè)三角面片用四個(gè)數(shù)據(jù)項(xiàng)表示。即三個(gè)頂點(diǎn)坐標(biāo)和一個(gè)法向矢量，整個(gè)CAD模型就是這樣一個(gè)矢量的集合。在一般的軟件系統(tǒng)中可以通過調(diào)整輸出精度控制參數(shù)，減小曲面近似處理誤差。如Pre/1E軟件是通過選定弦高值(ch-chordheight)作為逼近的精度參數(shù)。

Construction of 3D model: Obtain the CAD file describing the part in the 3D CAD design software. The file output format supported by general rapid prototyping is STL model, which approximates the solid surface through the so-called Tessellation process, using planar triangular patches to approximate the model surface. To simplify the data format of CAD models. Facilitate subsequent hierarchical processing. Due to its simplicity in data processing and independence from CAD systems, it quickly developed into a standard for data exchange between CAD systems and rapid prototyping machines in the field of rapid prototyping manufacturing, with each triangular surface represented by four data items. The entire CAD model is a collection of three vertex coordinates and a normal vector. In general software systems, surface approximation processing errors can be reduced by adjusting the output accuracy control parameters. Pre/1E software uses the selected chord height as the accuracy parameter for approximation.

三維模型的離散處理:在選定了制作(堆積)方向后，通過專用的分層程序?qū)⑷S實(shí)體模型(一般為STL模型)進(jìn)行一維離散，即沿制作方向分層切片處理，獲取每一薄層片截面輪廓及實(shí)體信息。分層的厚度就是成型時(shí)堆積的單層厚度。由于分層破壞了切片方向CAD模型表面的連續(xù)性，不可避免地丟失了模型的一些信息，導(dǎo)致零件尺寸及形狀誤差的產(chǎn)生。切片層的厚度直接影響零件的表面粗糙度和整個(gè)零件的型面精度，每一層面的輪廓信息都是由一系列交點(diǎn)順序連成的折線段構(gòu)成。所以，分層后所得到的模型輪廓已經(jīng)是近似的，層與層之間的輪廓信息已經(jīng)丟失，層厚越大丟失的信息越多，導(dǎo)致在成型過程中產(chǎn)生了型面誤差。

Discretization processing of 3D models: After selecting the production (stacking) direction, the 3D solid model (usually STL model) is discretized in one dimension through a dedicated layering program, that is, slicing along the production direction to obtain the cross-sectional profile and solid information of each thin layer. The thickness of the layers is the thickness of the single layer stacked during molding. Due to layering, the continuity of the CAD model surface in the slicing direction is disrupted, inevitably losing some information of the model, resulting in errors in part size and shape. The thickness of the slicing layer directly affects the surface roughness of the part and the overall surface accuracy of the part. The contour information of each layer is composed of a series of intersection points connected by a sequence of folded line segments. So, the model contour obtained after layering is already approximate, and the contour information between layers has been lost. The thicker the layer, the more information is lost, resulting in surface errors during the forming process.