1.

REVERSE ENGINEERING: THE THEORY

1.1. WHAT IS REVERSE ENGINEERING?

Reverse engineering is the process that identifies an object, a device, or a system technological properties
by performing a comprehensive analysis of its structure, functions and operations. In mechanical
engineering, this process aims to create a virtual 3D model from an existing physical object to duplicate or to
enhance it.

They are many reasons to use reverse engineering of physical objects. For example, the reverse engineering
process will be used if the original design is not supported by sufficient or adequate documentation or if the
original CAD model is not appropriate to support modifications and/or standard production methods. In some
cases, original manufacturer no longer exists nor manufactures a product, but some requirements remain for
that product. Reverse engineering would then help. In the case of worn or damaged components, for which
no source of supply is available, reverse engineering building features would also to analyze competitors'
products features, for performance improvement and/or product features, for lack of additional parts (spare
parts) or to update obsolete materials or antiquated manufacturing process.

In order to create a 3D model of the object, the object must be measured using 3D scanning technologies
(CMM, laser scanners, structured light digitizers…). Once the scanning operation is done, it is possible to
rebuild the 3D model using 3D CAD, CAM, CAE or other software.

1.2. WHAT IS 3D SCANNING TECHNOLOGY?

The 3D scanning technology is the application of scientific knowledge for analyzing an object (or
environment) by collecting data on its shape and, sometimes, on its appearance (color, texture) to build a 3D
digital model. The device used to perform this kind of analysis is a 3D scanner.

1.3. WHAT ARE 3D SCANNERS?

3D scanners devices are very similar to cameras: they have a conical visual field and can collect information
on visible surfaces. A detail makes them different: the camera collects color and surface information within
its field of view (creating images) while the 3D scanner uses the images acquired to extract 3D data
(collecting information on the distance and the surface within its field of view). It generally creates a geometric
point cloud of the object surface. The image produced by the 3D scanner indicates, for each point, the
distance to the surface. This provides the three-dimensional position for each point of the surface.
1.4. 3D SCANNERS: DIFFERENT TYPES FOR DIFFERENT USE

There are many types of equipment used for measuring 3D objects. 3D scanners, dedicated for digital form
acquisition, can be classified in two types of equipment: contact and non-contact 3D scanners. However,
non-contact 3D scanners are also divided into two categories: passive and active.

There are a broad variety of technologies for digital acquisition of 3D object shapes. Each device has its own
specifications and works differently. The following section describes each technology and will help you
understand them.

1.4.1. CONTACT

Contact 3D scanners explore (probe) physical objects while in contact with, or holding on to, a precision
plane surface, ground and polished to a specific maximum of surface roughness. When the object to be
scanned is not flat or cannot be stabilized on a plane surface, it has to be supported and held in place
using a fixture.

The scanner mechanism may either be a carriage system with fixed arms tightly clasped in perpendicular
with each axis sliding along a track, or an articulated rigid arm provided with high precision angular
sensors (Figure 1: Articulated probing arm equipped with a ruby touch probe.).

Figure 1: Articulated probing arm equipped with a ruby touch probe.

Resolution is one of the mesh file main properties, defining the level of details visible on scan data. It can
be compared to screen resolution, which is defined by the amount of pixels. A combination of both
methods can be used, such as an articulated arm suspended from a moving carriage, for mapping large
objects with interior cavities or overlapping surfaces.

A CMM (Coordinate measuring machine) is an example of contact measurement system (Figure 2). It is
mostly used in manufacturing and can be very precise. The disadvantage of CMMs is that they require
contact with the object being scanned. Scanning the object might modify or damage it, and it is very
significant when scanning delicate or valuable objects such as historical artifacts. The other disadvantage
of CMMs is that they are relatively slow compared to other scanning methods. The arm on which the
probe is mounted on can move very slowly, and the fastest CMMs can only run at a few hundred hertz.
In contrast, optical systems such as laser scanners can run from 10 to 500 kHz.

Contact measurements systems are also highly sensitive to environment changes (vibrations, part
displacement, temperature variations). In order for them to perform as indicated, these machines need
to be isolated from these changes (installed on an independent concrete slab or used in a metrology lab).

Figure 2: CMM (coordinate measuring machine) with rigid perpendicular arms.

1.4.2. NON-CONTACT PASSIVE

Passive scanners do not emit any kind of radiation; they rely on detecting reflected ambient radiation.
Most of these type of scanners detect visible light as it is found in the ambient radiation. Other types of
infrared radiation are also used. In most cases, passive methods are generally inexpensive because they
do not require particular equipment except for digital cameras. They provide low accuracy.
1.4.3. NON-CONTACT ACTIVE

In order to probe and object or environment, active scanners emit some kinds of radiation (X-ray,
ultrasound) or light and detect the reflection or radiation passing through object. The light or radiation
emitted by the scanner is reflected by the object scanned and sends back the information to the scanner,
recording the surface distance. However, this type of scanners has limitations: it has difficulty scanning
reflective or transparent objects.

1.4.3.1. TIME-OF-FLIGHT

The time-of-flight 3D laser scanner (Figure 3) uses a laser light to probe the subject. This type of
scanner, equipped with a time-of-flight laser rangefinder, finds the distance of a surface by timing
the round-trip time of a pulse of light. Moreover, the laser emits a pulse of light and the amount of
time is measured before the detector sees the reflected light.

Since the speed of light c is known, the round-trip time determines distance travelled by the light,
which is twice the distance between the scanner and the surface. If t is the round-trip time, then
distance is equal to c ∙ t/2. The accuracy of a time-of-flight 3D laser scanner depends on how
precisely the t time can be measured: 3.3 picoseconds is approximately the time taken for light to
travel 1 millimeter.

The laser rangefinder only detects the distance of one point in its direction of view. The device scans
its entire field of view, one point at a time, only by changing the range finder's direction of view to
scan different points. The direction view of the laser rangefinder can be changed either by rotating
the rangefinder itself, or by using a system of rotating mirrors. This method is commonly used
because mirrors are much lighter and can thus be rotated faster with a higher level of accuracy.

The typical time-of-flight 3D laser scanners can measure the distance of 10,000~100,000 points
every second.
 Figure 3: Lidar (light detection and ranging) 3D scanner used to scan buildings, rock formations, etc.

1.4.3.2. TRIANGULATION

Triangulation based 3D laser scanners are also active scanners. They use laser light to probe the
environment. Respecting time-of-flight 3D laser scanner, the triangulation laser shines a laser on the
subject and uses a camera to look for the location of the laser dot which appears at different places
in the camera field of view, depending on how far away the laser strikes a surface.

This technique is called triangulation because the laser dot, the camera and the laser emitter form a
triangle (Figure 4). Three pieces of information determine the triangle shape and size. First, the length
of one of the triangle sides the distance between the camera and the laser emitter is known. Second,
the angle of the laser emitter corner is also known. Third, the angle of the camera corner is
determined by looking at the laser dot position in the camera's field of view. In most cases a laser
stripe, instead of a single laser dot, is swept across the object to speed up the acquisition process.

The National Research Council of Canada (NRC) was among the first institutes to develop the
triangulation-based laser scanning technology in 1978.
 Figure 4: Principle of laser triangulation sensor.

1.4.3.3. STRUCTURED LIGHT

Structured-light 3D scanners project a light pattern on the subject. They look at the deformed pattern,
which is projected on the subject using either an LCD projector or other stable light source. The
camera, slightly offset from the projector pattern, looks at the shape and calculates every point’s
distance in the field of view. Some of those existing systems can also scan moving objects in real-
time.

Structured-light 3D scanners have two main benefits: to capture speed and precision. In fact, instead
of scanning one point at a time, they scan multiple points or the entire field of view at once. This
means they scan the entire field of view and generate exponentially more precise profiles than laser
triangulation, in a fraction of a second. This approach reduces (even eliminates) the problem of
distortion from motion.

1.4.3.4. MODULATED LIGHT

The modulated light 3D scanners project a continual changing light on the subject. This creates a set
of linear patterns on the object (sinusoidal pattern). A camera detects the reflected light and
determines the distance the light traveled, shifting each line of the pattern.

The modulated light also allows the scanner to ignore the light from sources other than laser, so
there is no interference.
 1.4.3.5. VOLUMETRIC TECHNIQUES

Industrial
Computed tomography (CT), microtomography and magnetic resonance imaging (MRI) techniques
are used to acquire an object digital image including its interior, such as non-destructive material
testing, reverse engineering or the study of biological and paleontological specimens. Although used
in industrial domain, these techniques are more common in the medical domain.

Medical
Computed tomography (CT) is a medical imaging method, which generates a three-dimensional
image of the inside of an object from a large series of two-dimensional X-ray images. Similarly,
magnetic resonance imaging is another medical imaging technique that provides much greater
contrast than computed tomography (CT), between the body’s different soft tissues, making it
especially useful in neurological (brain), musculoskeletal, cardiovascular, and oncological (cancer)
imaging. These techniques produce a discrete 3D volumetric image that can be visualized,
manipulated or converted to traditional 3D surface by means of isosurface extraction algorithms.1

Figure 5: X-ray computed tomography. The 3D model (top left corner) is

computed using the 2D images captures; the 3D model is reconstructed slice by slice.

1.4.3.6. STEREO VISION

Computer stereo vision is the result obtained from the extraction of 3D information of digital images
(such as CCD camera). By comparing two vantage points, 3D information can be extracted from
objects relative positions in the two panels. This is a very similar process than the stereoscopic vision.

1 URL: http://en.wikipedia.org/wiki/3d_laser_scanning (consulted on February 11, 2013)

Active vs. Passive Stereo
Active stereo vision uses a light, such as laser or a structured light pattern, to simplify stereo matching
challenges. The opposite of the active stereo vision is the passive stereo vision.

1.4.3.7. HANDHELD LASER SCANNER

The handheld laser scanners create a 3D image using the triangulation mechanism (ref. section
1.4.3.2). A laser dot (or line) is projected on the object from a handheld device (sensor) and the
distance to the measurement surface. An internal coordinated system collects the data. Therefore,
the scanning device position must be determined in order to collect data while the player is moving.
The scanner determines the object position by using the reference features of the scanned surface
(usually adhesive reflective targets). These scanners can also use the actual shape (object texture
or color) using a process of external monitoring.

A laser tracker providing the sensor position often represents external tracking. An integrated
camera determines the orientation of the scanner or a photogrammetric solution, using three (or
more) cameras, providing a six degrees freedom for the scanner.

Both techniques use infrared light-emitting diodes attached to the scanner. Infrared light-emitting
diodes are seen by the camera(s) through filters providing resilience to ambient lighting.

A computer collects the data (collected points), which is then recorded within a three-dimensional
space. That can be converted into a triangulated mesh with processing and then, to the CAD model.

Figure 6: HandySCAN 3D scanner used to digitize a metal casting.

The system uses stereo principle to simultaneously observe reflective
markers (for positioning) and laser projection (for geometry detection).
 1.4.3.8. HANDYSCAN 3D BY CREAFORM

HandySCAN 3D is a hand-held scanner manufactured by Creaform. It is a very portable and versatile

self-positioning 3D scanner. It uses Creaform positioning targets to reference its position from the
object to scan. This device’s main characteristic is its portability; it refers directly to the object with
targets. It can easy be carried to the object instead of bringing the object to the scanner.

Once the object and the scanner’s position have been located with targets, the surface acquisition is
completed through the camera. The camera sees the two laser lines, crossing each other, projected
on the surface. As the surface is swept over by the laser, data is recorded based on the triangulated
position. The output file format is a STL file.

1.5. LIVE SURFACE GENERATION

New 3D sensor technologies can capture 3D points at a very high rate and quickly capture very dense sets
of points on the object surface (live surface generation). This large quantity of 3D points observations allows
analyzing their local distribution on a section before validating their consistency with the expected error model
developed at the calibration stage. High quality surface point measurements can be produced in metrology
from these point observations. A smart and real-time measurement process is integrated to these steps
during capture.2

1.6. 3D SCANNING APPLICATIONS

NPD (New product development) is the process of bringing a new product to the market. There are two
parallel paths involved in that process: the first one involves the idea generation, product design and detail
engineering, while the second involves market research and marketing analysis.

One of the NPD concepts is product lifecycle management (PLM) in which 3D scanning plays a real key role.

2
http://www.creaform3d.com/en/resource-center/technological-fundamentals/intelligent-measurement/documentation
 1.6.1. 3D SCANNING: PART OF THE PRODUCT LIFECYCLE MANAGEMENT (PLM)

Figure 7: Simplified diagram of a typical PLM.

Conceive
The first step is to define product requirements based on the customer, the business, the market and
regulatory agencies’ perspective. The second step is to define the product main technical parameters. The
first design work is then performed and the product aesthetics and main functional aspects are developed
in parallel. Different media are used for these processes: from pencil and paper to clay models in CAID
software (Computer-aided industrial design).

Design
a. Detailed Design

The first stage is product design and development. Then come the prototype and production steps. Finally,
the product is launched. That can involve redesigning or improving existing products. The main tool used for
the design and development is CAD software is a simple 2D drawing/drafting or a 3D parametric feature
based on solid/surface modeling. CAD software includes technologies such as hybrid modeling, reverse
engineering, KBE (Knowledge-base engineering), NDT (Nondestructive testing), and the assembly model.
b. Prototyping

Prototypes are the first samples, models or versions of a product built to test the concept, process or act as
a thing to replicate. It is a term used in many contexts such as semantics, electronics, software programming
and product design. A prototype is primarily designed to test a new design (by analysts and/or users) and to
improve a product accuracy and quality.
Formerly known as rapid prototyping, the additive manufacturing or 3D printing is the process of making a
three-dimensional solid object of any virtual shape from a digital model. It is used to provide specifications
for a real object (instead of a theoretical system).

c. Validation, Testing and Analysis

Simulation, validation and optimization tasks are carried out using CAE (Computer-aided engineering)
software either integrated in the CAD package or stand-alone. These are used to perform tasks such as
stress analysis, FEA (Finite element analysis), kinematics, CFD (Computational fluid dynamics) and MES
(Mechanical event simulation).

Manufacture
a. Tool Design

Once the design is validated, manufacturing needs to be foreseen, including tool design. It is based on
planning how to mass-produce the project and which tools to use for the object manufacturing. At this step,
the tasks include selecting production processes with the material, determining the sequence of operations,
and choosing tools (such as jigs, fixtures and tooling). This task also involves testing a working prototype to
make sure the object created meets qualification standards.

b. Build/Assemble/Produce

Mass production is implemented following steps of design and manufacturing processes. This is standardized
products manufacturing of large quantities including (and most importantly) assembly lines.

Standardized production concepts are applied to different types of products such as fluids and particles
processed in bulk (food, fuel, chemicals and mineral extracts) or discrete solid parts (i.e. fasteners) to
assemble such components (e.g.: appliances and automobiles).

Mass production is a distinct field and can usually be contrasted with craft production or manufacture
distributed. This production process has been widespread and is at the center of the economy since the late
19th century.

c. Quality Control

Once the manufacturing process is defined and production has started, it is time for the products to under go
quality checks.
 Quality checks include CAD tasks such as tool design: creation of CNC (Computer numerical control)
machining instructions for the product’s parts as well as tools to manufacture those parts, using integrated
or separate CAM (Computer-aided manufacturing) software. This also involves using analysis tools for
operations process simulation such as casting, molding, and die press forming.

Service
a. Maintenance, Repair and Operations (MRO)

Once the product is in service, it will eventually suffer from wear and tear. A continuous quality control on the
product offers better visibility and can prevent heavy damage to occur. Therefore, surveillance and
maintenance operations are decisive. For instance, part of an object or assembly may have to be replaced
or repaired if it is damaged.

b. Dispose/Replace/Recycle

In the same way, there is an end-of-life for every product that needs to be planned. The disposal or recycling
of an object has a cost and that is part of its life cycle. The dismantling of sensitive parts, such as in nuclear
power plant, sometimes requires high caution and precautionary measures.

The 3D digitizing can be involved in all manufacturing industries and in every step of product lifecycle
management (PLM).

1.6.2. OTHER APPLICATIONS OF 3D SCANNING

Museology, Heritage Preservation, Restoration and Digital Archiving

3D scanning technology is used for archiving, restoration or virtual museum. For example, Creaform has
reproduced the monument of Francis de Gaston, knight of Lévis (Figure 8Figure 8). This gave the
Canadian city of Lévis, Quebec, the opportunity to own a bronze statue of the historical figure after whom
the city was named3.

3 URL : http://blog.creaform3d.com/2012/03/the-knight-of-levis-an-historic-figure-goes-digital/
 Figure 8: Reproduction of the monument of Francis de Gaston (knight of Lévis).

Another example of this application are the scans of ancient monuments realize by Ben Kacyra. Ancient
monuments give us clues of the astonishing past civilizations -- but are threatened by pollution, war and
neglect. Ben Kacyra invented a groundbreaking 3D scanning system (state-of-the-art technology) and is
using his invention to scan and preserve the world's heritage in archival detail4. To view his work, go to:
http://www.ted.com/talks/ben_kacyra_ancient_wonders_captured_in_3d.htm

Multimedia and Entertainment

Multimedia and entertainment industries can use 3D scanning tools in various applications, mainly to
reduce modeling time, to speed up process and to make 3D graphics more realistic.
For an example of multimedia application: http://blog.creaform3d.com/2013/12/when-3d-scanning-goes-
glamour-–-the-jenn-barlow’s-bubble-head-project/
Virtual Reality/Augmented Reality (Serious Gaming, 3D Training Systems)
3D scanning is not only for small/medium objects. Some devices are able to capture and reconstruct an
entire room or environment. The result may be reworked and transformed into a 3D virtual environment.
1.7. DEFINITIONS

1.7.1. POINT CLOUD

Figure 9 : Representation of point cloud shape.

4 URL : http://www.ted.com/talks/ben_kacyra_ancient_wonders_captured_in_3d.html
In a three-dimensional coordinate system, a point cloud is set of data point. These points are usually
defined by XYZ format and represent a set of 3D measurements. The cloud shape is due to the amount
of measured points. A single piece of data can be measured several times from different points of view.
For each of these measurements, uncertainty causes small agglomerates of points for the same data
and then creates the thickness of the model.

Point clouds may be created by 3D scanners, which measure a large number of points of an object
surface, and often output a point cloud as a data file. The point cloud represents the set of points the
device has measured. As the result of 3D scanning process, point clouds are used for many purposes
including creating 3D CAD models for manufactured parts, metrology, quality inspection and for a
multitude of visualization, animation, rendering and mass customization applications.

While point clouds are directly rendered and inspected, they are generally not directly usable in most 3D
applications. They are usually converted to triangle mesh models (STL file) or to NURBS (Non-uniform
rational basis spline) surface models through a process commonly called surface reconstruction.

1.7.2. STL/MESH FILE

Figure 10: Representation of STL shape.

STL (stereo-lithography) is a file format native to the stereo-lithography CAD software, widely used in 3D
printing and CAD industries. This file format is supported by many other software packages.

STL files only contains the surface’s geometry of a three-dimensional object without any color, texture or
other common CAD model attributes. They also contain a raw unstructured triangulated surface of the
unit normal and vertices (ordered by the right-hand rule) using a three-dimensional Cartesian coordinate
system.
The STL format specifies both ASCII and binary representations. Moreover, binary files are more frequen
since they are more compact. STL coordinates must be positive numbers. There is no scale information
and units are arbitrary.

1.7.3. RESOLUTION

Resolution is one of the mesh file main properties, defining the level of visible details on scan data. It can
be compared to screen resolution, which is defined by the number of pixels. The size of the triangles side
on the mesh file is called the resolution (measured in millimeters). Higher resolution increases the number
of triangles of a mesh file. For an equivalent object, the size of the mesh file is consequently proportional
to the resolution.

Important to remember: resolution and accuracy are two different factors. The resolution only defines the
level of details. A mesh can be equally accurate to another with a lower resolution. The difference is
significant in highly curved areas (Figure 8).

Figure 11: Difference between Resolution and Accuracy.

1.7.4. ACCURACY

In metrology, precision refers to measurement dispersion. The measurement error (the mean) can be
close to zero even if the system is not very precise (but it needs a good trueness). In other words, the
less scattered the measurement data is, the more precise is the equipment. A formal definition of
precision is: closeness of agreement between indications or measured quantity values obtained by
replicate measurements on the same or similar objects under specified conditions.
 The word trueness rather gives information on the difference between the mean of measurements and
the real dimension regardless dispersion. In other words, the closest the mean of measurements is to the
nominal value, the more trueness the equipment has. A formal definition of trueness is: closeness of
agreement between the average of an infinite number of replicate measured quantity values and
reference quantity value.

Accuracy is the conformity between scan data and reality. To evaluate the accuracy of a measuring
device, like a laser scanner, the data acquired with this device should be compared to the data acquired
with a more accurate measurement tool (e.g.: a coordinate measuring machine (CMM)) and the
measured item must be normalized.

Figure 12: Representation of accuracy trueness.

1.7.5. RAW/CLEAN

A raw mesh is a file directly issued from the scanning device, before any post-treatment operation, while
a file mentioned as clean, is a file that has been post-treated and ready for reverse engineering
operations.

1.8. DIRECT USE OF STL MODEL

Some applications do not need a CAD file, even reverse engineering process, when the STL file is
compatible and directly applies.

FEA (Finite Element Analysis)

 Figure 13 : Representation of finite element analysis (FEA).

In mathematics, the finite element method (FEM) is a numerical technique for finding approximate solutions
to boundary value problems for differential equations. It uses variation methods (the calculus of variations)
to minimize an error function and produce a stable solution. Similar to the idea that connecting many tiny
straight lines can approximate a larger circle, the finite element method (FEM) encompasses all the methods
for connecting many simple element equations over many small subdomains, called finite elements, to
approximate a more complex equation over a larger domain.
FEA, as applied in engineering, is a computational tool for performing engineering analysis. It includes the
use of mesh generation techniques for dividing a complex problem into small elements, as well as the use
of software program coded with FEM algorithm. In applying FEA, the complex problem is usually a physical
system with the underlying physics such as the Euler-Bernoulli beam equation, the heat equation or the
Navier-Stokes equations. The divided small elements of the complex problem represent different areas in
the physical system.

FEA is a good choice for analyzing problems over complicated domains (as in the case of cars and oil
pipelines), when the domain changes (as during a solid state reaction with a moving boundary), when the
desired precision varies over the entire domain or when the solution lacks smoothness.

Rapid Prototyping
Rapid Prototyping (RP) or more generally 3D printing can be described as a group of techniques used to
fabricate a scale model quickly of a part (object) or assembly using three-dimensional computer aided design
(CAD) data. What is commonly considered being the first RP technique is stereolithography. Therefore, the
STL file format has naturally become a standard in data transmission format for rapid prototyping industry.
The mesh file can either be created from an existing CAD file or come from a 3D scanning device.
Rapid Prototyping has also been referred to solid free-form manufacturing; computer automated
manufacturing, and layered manufacturing. RP has obvious use as a vehicle for visualization. In addition, RP
models can be used for testing or creating male models for tooling, such as molds and investment casting.
In some cases, the RP part can be the final one, but typically the RP material is not strong or accurate
enough. When the RP material is suitable, highly convoluted shapes (including parts nested within parts) can
be produced because of the nature of RP.

CNC Milling (Computer Numerical Control)

The key advantage of using an STL file format is that it communicates only the data that the NC programmer
needs -- the geometry of the object. There is none of the extra information in the file to indicate the structure
of the original CAD file. The process of transferring work piece geometry from one system to another is
simpler and saves work and errors since the STL file is by definition a closed, unambiguous surface.

Not having access to all of the original surface data can at times prove to be a limitation, but that's certainly
not a given. Indeed, if the STL file is properly constructed, it should preclude the NC programmer's need to
get at the surface data in the first place. In some cases, the structure of the CAD file is proven to be a
handicap, especially for small surfaces. They may need to be merged to be machined as one, when this kind
of modeling operations is not necessary with STL files as the tool path only relies on the shape of the object,
and not on the arrangement of component surfaces.