Advanced Materials Research Vols 194-196 (2011) pp 2216-2219 Online: 2011-02-21

© (2011) Trans Tech Publications, Switzerland

doi:10.4028/www.scientific.net/AMR.194-196.2216

Reverse Engineering Application in Automobile Parts Design

Xiaoling Yanga, Tianrui Zhoub and Limin Wangc

School of Mechanical and Electrical Engineering, Nanchang University,
Nanchang 330031, China
a
[email protected], [email protected], [email protected]

Keywords: Reverse Engineering; Point Cloud; Surface Reconstruction; CATIA

Abstract. Along with the development of the measuring technique and computer technology, market
competition is aggravating, the modern product of the design cycle demands shorter especially in
automobile field. In order to meet the development requirements of automobile, we must reduce the
design cycle and cost, so the Reverse Engineering (RE) technique has applications widely. This paper
describes some concepts about reverse engineering. It points out the key technologies in reverse
engineering like cloud data processing technology and surface reconstruction technology. Finally this
paper uses the reverse engineering software CATIA to recover one inner automobile panel and
describes the process of reverse engineering in details to build 3D model.

Introduction
Reverse Engineering (RE) is using the physical model to become the digital technology of CAD
model, geometric model reconstruction technology and the product manufacturing technology, also is
turning the product or physical existing model into engineering design model and the concept model,
and then based on this to anatomy, deepening and the process of recreation [1,2].
RE is widely used in automotive, aerospace, medical fields and etc. In contrast to this conventional
manufacturing sequence, reverse engineering represents an approach for the new design of a product
that lacks an existing CAD model. In the process of the product design and research, the use of RE
will largely reduced the production period and costs [3].
RE has key techniques including digital technology, data preprocessing, surface reconstruction
technique, 3D model reconstruction technique etc.
There are many software about Reverse engineering. Like Geomagic, Surface, Imageware, and
like normal 3D modeling software also including the function of RE. This paper uses CATIA because
of its great combination function about surface design, modeling design, reverse design and etc. In
RE design there are three main modules: Digitized Shape Editor (DSE), Quick Surface
Reconstruction (QSR), and Generative Shape Design (GSD). The typical process of RE in CATIA is
shown in Fig. 1 [4].

Fig.1. The typical process of RE in CATIA

Digitization of the inner automobile panel

Point digital technology has two methods: contact data collection and non-contact data collection [5].
Contact data collection machine typically is Coordinate Measuring Machine (CMM). In this paper,
we use the LSH800 three-dimensional laser scanning machine (non-contact data collection) to collect
data of the inner automobile panel. The scan precision is ±0.05 mm. This machine has two scan

All rights reserved. No part of contents of this paper may be reproduced or transmitted in any form or by any means without the written permission of Trans
Tech Publications, www.ttp.net. (ID: 152.14.136.77, NCSU North Carolina State University, Raleigh, USA-24/04/15,18:29:02)
 Advanced Materials Research Vols. 194-196 2217

methods including Planar scanning and Rotating scanning. Rotating scanning is applicable to scan
cylinder parts, and Planar scanning is applicable to single feature surface or complex
three-dimensional model, so we use planar scanning for the inner automobile panel. The part can’t get
complete point cloud data through one planar scanning; The result of one planar scanning is shown in
Fig. 2, and we can see the point data by scanning is not complete, so we should through many times
Planar scanning to get complete date at different positions through rotate rotary tables at
00.900,1800,2700 and so on, but the coordinate may be not the same so during the afterwards process
we should process the point cloud registration of multiple views.

Fig. 2. The result of one planar scanning

Data preprocessing of the point cloud

The data preprocessing is very important during the RE process, because its results will effect the
quality of the reconstruction model. In order to improve the rationality and integrity of the data, the
point cloud generally need some preprocessing such as registration of point clouds into one
coordinate system, and removes the outliers [6]. In DSE we use the registration tools to make multiple
views of the point cloud to the one coordinate system. The outliers we use Remove tool to operate.
And the point cloud data contains millions of points so that the computer runs slowly. In DSE Filter
has two methods including Homogeneous and Adaptive to reduce the point numbers. Homogeneous
is average algorithm to reduce, Adaptive is Based on the curvature to reduce the number of points in
order to make characteristics obviously. We use the Flitting by adaptive 0.04mm to reduce the point
number from 245583 to 111778.

Point cloud data mesh

From the scanning cloud point data, we are difficultly to discriminate the geometric configuration. In
order to identify each feature about point cloud and build model conveniently, we should need to
mesh the cloud point. In CATIA using 3D method to mesh, then to mesh smoothing, filling hole
mesh , decimate the mesh and so on in order to have good mesh for the behind operation.

Point cloud segmentation

Since the inner automobile panel has complex contour surfaces, and the whole inner automobile panel
can’t be approximated by a single mathematical surface, so the point clouds generally need to be
subdivided into many sections, each section can be fitted to a surface such as plane, cylinder, sphere,
cone, free form surface and so on. We divided the point clouds by segmentation by Curvature
Criterion in QSR. The point cloud segmentation by Curvature Criterion in curvature values by 0.015
is shown with blue curves in Fig. 3; also we can see the curvature of the position at any point with the
mouse, like the 0.0458603. Based on this we can operate the every single point cloud by segmentation
to construct contour lines to make surface.
2218 Advanced Engineering Materials

Fig. 3. The point cloud segmentation by Curvature Criterion

Curve processing
Curve processing is also very important during the RE process, and contour feature line is for
preparation to subsequent surface fitting. Only getting high degree fitting and smooth curve from the
point cloud, then it can get high quality of smooth surface. In CATIA QSR there are some kinds of
creating contour feature lines such as Project Curve, Planar Sections, Create Scans on Clouds, Create
Free Edges, 3D Curve, Curve From Scan and etc, also it can according to Point cloud segmentation by
Curvature Criterion to create contour feature lines. Using the operation of Project Curve, Planar
Sections, Create Scans on Clouds, Create Free Edges only to get the contour lines need to do some
dispose like merge, slice and smooth, some contour feature lines for power fit should operate curves
network. With the operation of the line intersection is obtained, and finally to convert curve.

Surface reconstruction
At present the free surface model of surface reconstruction generally has two methods based on
triangle Bezier surfaces and the rectangle domain parameters fitting methods on the basis of NURBS
(non-uniform rational B-spline curve and surface )[7, 8]. The method of triangle Bezier surfaces
fitting is adaptive to no rule and complex surface parts such as human face. This method has some
difficulties with the normal CAD/CAM system data switching because of the surface mode not match
with the product description standards, so finally the surface with this method should change into
NURBS surface. During this transition process it will lose some data inevitably, so the method of
NURBS surface has wide application.
There are two surface construction methods on the basis of NURBS: one is using the fitting curve
from the point cloud, and then constructing surface by surface modeling with the fitting curve; The
second is directly fitting surface with the scan data; Finally using surface edit operation such as
transition surface, surface joining, surface trimming and etc.
In order to ensure the surface precision reconstruction during the RE process, we should compare
the error between the reconstruction surface with the point cloud. In CATIA there are some
operations: Distance Analysis, Cutting Planes and etc. The Distance Analysis operation is to compare
the distance between the point clouds with the surface reconstruction. The Cutting Planes operation is
to observe the smooth of surface. The Distance Analysis about one surface with the original
equivalent point cloud is shown in Fig. 4. Every time we should analysis the error to guarantee the
surface reconstruction whether available. To this part we want to control the error in ±0.5 mm.
In this paper we use surface fitting in QSR, and surface modeling in FS and GSD. Finally using
surface edit operation such as transition surface, surface joining, surface trimming and etc to
complete surface reconstruction. After the surface reconstruction, we use the operation Thicken 1mm
to get the inner automobile panel 3D model. The final 3D model of the inner automobile panel by
reconstruction is shown in Fig. 5.
 Advanced Materials Research Vols. 194-196 2219

Fig. 4. The Distance Analysis about one Fig. 5. The final 3D model of the inner
reconstruction surface with the original point cloud automobile panel by reconstruction

Conclusion
Based on the theoretical knowledge in reverse engineering, we successfully use the reverse
engineering software CATIA to recover the inner automobile panel of 3D model, and also the model
meet the product precision. Through reverse engineering we can greatly reduce the design cycle and
cost. For now, the auto industry is fierce competition, so winning the first time is winning the market.
The RE is widely used in automotive, aerospace, medical fields and etc, so to improve the key
techniques is very necessary and important. Only the key techniques in the RE process become much
more perfectly, then we can get much higher precision. The precision improving I think mainly from two
sides, first we should improve the precision about point digital technology. The second is to uses a
better algorithm to point cloud data processing and some new methods about the surface
reconstruction to improve the reconstruction precision.
Acknowledgements
This work was financially supported by the Jiangxi Province science and technology plan supported
projects（2007BG096050）
References
[1] RAN Xian-sheng, HUANG Ze-ha, LIN Li: Journal of Chongqing University of Technology In
Chinese .Vol. 24 (2010), p. 53.
[2] Yan Shan, Binfei Xie, Editor: The basic reverse modeling technology on Imageware (Tsinghua
University Publishers, Beijing 2006)In Chinese.
[3] Yan-Ping Lin, Cheng-Tao Wang, Ke-Rong Dai: Medical Engineering & Physics. Vol.27 (2005),
p. 189.
[4] Longhan Xie, Yan Shan , Editor: CATIA V5 Reverse modeling design ( Tsinghua University
Publishers, Beijing2004)In Chinese.
[5] Zhiqin Xu, Changku Sun, Editor:3D Reverse engineering technology (China Metrology
Publishing, Beijing 2002)In Chinese.
[6] Woo H, Kang E, Wang S, et al: Int J Machine Tools Manuf. Vol. 42 (2002), p. 167.
[7] Zhao Liu, Ji Liting, Wang Lijian, Huang Fu: ELECTRONIC TEST In Chinese . Vol .2 (2010),
p. 19.
[8] WU Xiuchun, HU Lijuan, ZHAO Hongbo: Tractor & Farm Transporter In Chinese. Vol.36
(2009), p.24.
Advanced Engineering Materials
10.4028/www.scientific.net/AMR.194-196

Reverse Engineering Application in Automobile Parts Design

10.4028/www.scientific.net/AMR.194-196.2216

DOI References
[3] Yan-Ping Lin, Cheng-Tao Wang, Ke-Rong Dai: Medical Engineering & Physics. Vol.27 (2005), . 189.
doi:10.1016/j.medengphy.2004.09.006
[7] Zhao Liu, Ji Liting, Wang Lijian, Huang Fu: ELECTRONIC TEST In Chinese . Vol .2 (2010), . 19.
doi:10.1088/1674-1056/19/12/124211
[3] Yan-Ping Lin, Cheng-Tao Wang, Ke-Rong Dai: Medical Engineering & Physics. Vol.27 (2005), p. 189.
doi:10.1016/j.medengphy.2004.09.006
[5] Zhiqin Xu, Changku Sun, Editor:3D Reverse engineering technology (China Metrology Publishing,
Beijing 2002)In Chinese.
doi:10.1016/S0924-0136(02)00623-4
[7] Zhao Liu, Ji Liting, Wang Lijian, Huang Fu: ELECTRONIC TEST In Chinese . Vol .2 (2010), p. 19.
doi:10.1088/1674-1056/19/12/124211