RP 3rd Internal

Chapter 9

Software for RP

Learning objectives:

After studying this chapter you should be able to do the following:

1. What is STL file?

2. List out the common errors during conversion of STL file

3. Effect of file resolution on the final part

4. List out a few software used in RP

9.1 STL File:

STL (Standard Tessellation Language) is a file format native to the

stereolithography CAD software created by 3D Systems. This file format is supported

by many other software packages; it is widely used for rapid prototyping and

computer-aided manufacturing. STL files describe only the surface geometry of a

three dimensional object without any representation of colour, texture or other

common CAD model attributes. The STL format specifies both ASCII and binary

representations. Binary files are more common, since they are more compact.

An STL file describes a raw (not much changes) unstructured triangulated

surface by the unit normal and vertices (ordered by the right-hand rule) of the

triangles using a three-dimensional Cartesian coordinate system. This format has long

been the industry standard in rapid prototyping. Let us look into the process of

approximating surfaces with triangles: Each 3D form is made out of polygons. A

polygon is defined as a flat shape which is constrained or confined by a closed circuit.

Each polygon with n sides can be represented using n-2 triangles.

If we look at a box, for example: It is easy to see the box has 6 sides- each one is a

polygon. Each one of those sides is a square, meaning it can be represented using 12

triangles.

Since we are dealing with 3-dimensional shapes, each triangle has a direction.

This direction is expressed by the normal of the triangle. (The outward direction is

represented by the normal)

The two triangles above, though they are identical, are facing in opposite directions.

STL manipulation software

While all CAD software allows the creation of STL files, not always this process

ends with a printable file. STL manipulation solutions allow:

1. Fixing those models in order to produce watertight models.

2. Performing several simple actions to change the model (such as cutting and

labelling).

Magics RP is the most extensive solution available

Meshlab is satisfactory as a free tool, although not very easy to use and lacking

some functionality

Pricing is not specified, as this greatly varies from region to region.

Additionally, most vendors offer several packages. Please consult your local

distributor for pricing.

Online fixing Services

As an alternative to purchasing an STL fixing software, where the user can

upload the file and receive a fixed file, usually within a few minutes. Some vendors

offer free automatic service, while some will have a specialist look into more complex

issues. Pricing (if applicable) is typically per use, greatly minimizing the initial cost.

Common errors in CAD to STL conversion

Inverted normal:

The meaning of an inverted normal is that one surface does not have a

consistent direction. Occasionally, the interpretation or reading of the surface between

CAD and STL results in inverted normal lets look at an example:

The box above has one face with an inverted normal. This means we will need

to fix the file in order to print it using an Objet 3D printer. (You will realise it, in the

approaching topics)

Zero thickness:

Since files printed on Objet printers have to be fabricated in real world. The

files have to have a volume which is larger than zero. Sometimes, a model is

represented on the CAD software using just a 2D model, which has no volume:

The part shown above is a sheet of material which has no volume, though it is three dimensional.

In order for the file to be printable, we must give it some volume:

The second part is printable, since it has a positive volume.

Bad edges:

If we look into the definition of watertight, we realise that all triangles must

be connected: we cannot allow gaps between triangles. Gaps like this are commonly

referred to as bad edges. Let us look at an example: The two hemispheres of the

shape above are no connected, and are marked with a thick yellow line to indicate this

(indication varies with the software suite) in order to close those gaps we would need

to fill those gaps- an action referred to as stitching.

We distinct between two cases of bad edges:

Near bad edges:

Near bad edges are defined as edges which have a neighbour triangle which is

closer than a set threshold. Those are generally closed automatically using your

software of choice.

Real bad edges/Planar holes

Real bad edges typically enclose a hole in the shape. This is slightly more complex, as

it might cause a case of zero thickness.

The part above has a real bad edge in one of its faces. As a result, the boundaries of

the box have zero thickness.

Near bad Edges

Since Objet printers require a positive thickness in order for a file to be printable, this

will require one of two solutions:

a. Close the hole

Once the user closes (adds triangles) to the hole, the model is once again watertight

and has a positive volume. Once again, it is printable.

b. Create thickness

If the design intent was to create a box with one missing face, the user would need to

create thickness, making the part printable. This is typically achieved using the

offset command.

9.2 STL File Resolution

Irrespective of the CAD programming used to design a part or product

assembly, the ultimate input file for stereolithography and most other RP processes is

called the STL file. By chance, most major CAD programming offers the capability to

output an STL file suitable for the process. The catch is that the output process involves

a configuration step, and the quality of the final RP part is dependent upon a proper

configuration. To understand what is involved with proper STL output file

configuration requires a look at the STL file itself.

The STL file is intended to simplify the complex mathematical descriptions of

surface and solid geometry into a form that can be readily used to drive the imaging

systems of RP machines. To do this, the architects of the STL file type chose to use the

most basic form of surface that can be described from point data only the triangle.

If any three points are chosen from a 3-dimensional surface then a triangle can be

described by those points to approximate a portion of that surface. Of course, a

triangle is by definition flat-lacking any curvature whatsoever-so if the surface in

question (problem) contains any curvature then there is some error, or deviation, in

the approximation. Though, if the triangle size is reduced to the point where it is much

smaller than any curvature in the surface, then the deviation can be brought down to

a level where it is negligible. There is a practical limit-as the triangles decrease to an

extremely small size, the number of them required to complete the surface becomes

infinitely large, as does the size of the final STL file.

Most configuration settings for outputting STL files are aimed at this issue- how

small to make the triangles to best approximate the surface geometry while not

making enormous file sizes. Let us take a look at a perfect CAD shape and a rather

coarse STL file approximation of it (see the below figure). We would not suggest using

a file as coarse as this-we purely include it for the sake of illustration: Notice how the

areas of curvature are approximated in the STL file with large flats-deviating from the

proper geometry. This effect is called faceting and it would be evident in the final

RP part if built with this coarse file.

Lets take a closer look at the small hole in the centre, and highlight the edges of the

triangles used to create it:

Note that the true shape of the hole is drawn here, and an indication as to the deviation

from true, due to the use of large triangles. Now lets reduce the size of the triangles

to get a better approximation and compare:

Clearly, this does a much better job of approximating the curvature, and the

triangles are not so numerous as to make a huge file size. The coarse STL file size was

77KB while the fine STL file is 308KB. Equipped with this information, you can use

the help function for the CAD software that you use, and develop an approximation

that works for your needs. The major CAD software programming such as

ProEngineer and SolidWorks actually show you an image of the faceted STL file so

that you can see for yourself how closely the triangles approximate your geometry.

The proof is in the pudding, so to speak, so we recommend installing a file viewer for

STL files just to be sure. Doing so, you can rotate, zoom, pan and cross-section the file

to be sure that it is true to form. Here are some good STL viewers that are free to install

and use.

3D Tool

Minimagics

9.3 Solid view

SolidView software allows non-CAD users to easily view, measure, translate

and mark-up CAD data, opening up communication to all who need to be involved

in the design process. SolidView is used across the world by those needing access to

CAD data but not trained in using CAD systems. It is a low price solution to access

CAD data for manufacturing engineers, scientists, structural engineers, technical

illustrators, managers, product managers and sales people.

9.4 Magics

Magics is rapid prototyping software and is a key element of the Magics e-

Solution Suite, a full range of market-leading software products that will stream,

automate and boost almost every step in your rapid prototyping and manufacturing

(RP&M) process. Magics rapid prototyping software enables you to import a wide

variety of CAD formats and to export STL files ready for rapid prototyping, tooling

and manufacturing. Its applications include repairing and optimizing 3D models;

analysing parts; making process-related design changes on your STL files; designing

fixtures; documenting your projects; production planning and much more.

9.5 Mimics Z

For the first time, doctors, nurses, and technicians who have no previous

experience with 3D modelling or 3D printing can create 3D structural (anatomical)

models from MRI and CT scan images quickly and easily. Developed by Materialise

(creator of Mimics, the leading medical imaging software for the rapid prototyping

industry), Mimics Z is optimized for output on 3D Systems high-definition, 3D

printers that produce full-colour models in only hours. Healthcare organizations

worldwide increasingly rely on 3D anatomical models for pre-operative planning,

specialist consultation, implant fit and design, patient counselling and medical

education.

Key Features

Software wizards guide users with minimal training through the entire process.

Includes an extensive help function and templates.

Automates all steps needed to import MRI and CT scan files, select structures,

conduct editing and masking, and export data.

Processes industry standard DICOM data and outputs ZPR files optimized for

printing on 3D Systems ZPrinters.

Tight integration with ZEdit software, enabling users to add colour to highlight

areas of interest (such as a bone tumour), and annotation (explanation) features

that make it possible to label models with critical patient data and doctor

comments.

Bridge the gap between 2D image data & 3D engineering applications

Mimics software allows you to process and edit your 2D image data (CT, CT,

MRI, etc.) to construct 3D models with the utmost accuracy, flexibility and user-

friendliness. The powerful segmentation (subdivision) tools allow you to segment

your medical CT/MRI images, take measurements and engineer directly on your 3D

model. From there you can export your 3D data to a wide range of output formats

and engineering applications; such as FEA, design, surgical simulation, Additive

Manufacturing and more. Mimics is part of Materialise total solution for advancing

biomedical R&D

9.6 Magics communicator

Visualise

View STL, IGES*, VDA* and DXF 3d faces*, with fast rotation, zooming and

cross sectioning.

Annotate

Add 2D and 3D annotations, shapes, text and bitmaps.

Measure

Easily create 2D drawings from 3D files. Extensive feature recognition allows

measuring of distances, radii and angles in 3D. Add tolerances and additional info.

Present

Make a 3D slide show with adjustable colours, shading and transparency.

Ease of use

Communicator's straightforward interface ensures that even non-CAD users will be

comfortable with the programming in no time.

9.7 Other important software tools for RP

3D3 Solutions Makes FlexScan3D, a relatively cheap (inexpensive) programming that

allows digitizing objects using your own digital SLR camera and TV projector.

3D-SHAPE GmbH (Germany) SLIM makes it possible to create a model from a set of

different single 3D views. Registration, mesh reconstruction and visualization

processing steps can also be run almost entirely automatically via Visual Basic Script.

Alias Wavefront Spider, is a module for the Studio Tools suite for editing, translating

and operating on point cloud data in a variety of formats.

Anatomics Pty. Ltd. BioBuild converts volumetric imaging data to rapid prototyping

file formats. While designed for medical bio-modelling, it can also be used for reverse

engineering applications.

Geomagic, Inc. Geomagic enables the creation of STL files and 3D models from point

clouds of data from any source.

InnovMetric Software Inc. (Canada) PolyWorks/Modeller 5.0 3D modelling software

for milling, rapid prototyping, reverse engineering, and finite-element analysis

applications.

Rapidform, Inc. Produces RapidForm XO Verifier for comparing scan and CAD data,

and RapidForm XOR Redesign for generating parametric CAD models from scan

point clouds.

Chapter 10

Process optimization

Learning objectives:

After studying this chapter you should be able to do (answer) the following:

1. What is the need of process optimization in RP?

2. List out the factors that influence the accuracy of the model or part

3. Describe the surface errors

4. The importance of build orientation

Process optimization

The parameters of rapid prototyping can be classified as nuisance parameters,

constant and control parameters. Nuisance parameters include age of the laser, beam

position accuracy, humidity and temperature, which are not controlled in the

experimental analysis but may have some effect on a part. Constant parameters

include beam diameter, laser focus and material properties, etc. the constant

parameters will affect the output of the process and are controllable in a run. These

include layer thickness, hatch space, scan pattern, part orientation, shrinkage of the

material and beam width compensation, etc. Layer thickness, hatch space, part

orientation and depth of cure are the most vital among the control parameters.

Identification of requirements and key manufacturing parameters

The functional requirements of a manufacturing process include accuracy,

strength, build-time and efficiency of the process. All the manufacturing requirements

are also applicable to RP. Surface accuracy is gaining a greater significance as more

parts are used as master patterns for secondary manufacturing process. Build time is

very significant in the general context of manufacturing for scheduling and cost

estimation. Layer thickness, hatch space and orientation are the key control

parameters for SLS and SLA. These are required indeed process-independent

parameters, and can be applied to other processes, such as LOM, FDM, etc. Support

structures are essential for SLA and FDM, but they are not needed for LOM and SLS

processes.

10.1 Factors influencing accuracy

Accuracy of a model is influenced by the errors caused during tessellation and

slicing at data preparation stage. Decision of the designer about part deposition

orientation also affects accuracy of the model.

Errors due to tessellation: In tessellation surfaces of a CAD model are estimated

piecewise by using triangles. It is true that by reducing the size of the triangles, the

deviation between the actual surfaces and estimated (approximated) triangles can be

reduced. In practice, resolution of the STL file is controlled by a parameter namely

chordal error or facet deviation as shown in figure 10.1a

Fig 10.1a Tessellation of a typical surface of CAD model

It has also been suggested that a curve with small radius (r) should be

tessellated if its radius is below a threshold radius (ro) which can be considered as one

tenth of the part size, to achieve a maximum chordal error of (r/ro). Value of can be set

equal to 0 for no improvement and 1 for maximum improvement.

Errors due to slicing: Real error on slice plane is much more than that is felt, as shown

in figure 10.1.b

Fig 10.1.b Real error slice plane Fig 10.1.c Error due to replacement of

arcs with stair-steps cusp height

For a spherical model, error due to the replacement of a circular arc with stair-

steps can be defined as radius of the arc minus length up to the corresponding corner

of the staircase, i.e., cusp height. Thus maximum error (cusp height) results along z

direction and is equal to slice thickness. Therefore, cusp height approaches to

maximum for surfaces, which are almost parallel with the x-y plane. Maximum value

of cusp height is equal to slice thickness and can be reduced by reducing it; however

this results in drastic and great improvement in part building time. Therefore, by

using slices of variable thicknesses (widely known as adaptive slicing, as shown in

figure 10.1.d), cusp height can be controlled below a certain value. Except this,

mismatching of height and missing features are two other problems resulting from

the slicing. Although most of the RP systems have facility of slicing with uniform

thickness only, adaptive slicing scheme, which can slice a model with better accuracy

and surface finish without losing important features must be selected.

Figure 10.1.d: Slicing of a ball, (a) No slicing (b) Thick slicing (c) Thin slicing

(d) Adaptive slicing

10.1.1 Surface errors:

The common surface errors in Rapid prototyping errors are:

1. Representational errors

2. Process errors

Different orientations have different errors and number of layers requirement,

the preferred orientation is based on the users selection of primary measure. The

effect of laminar (consisting of laminae) building style results in stair-stepping. It is

possible to reduce or completely eliminate the stair-stepping effect by changing the

direction of build. For example, for a rectangular block, the stair-stepping effect will

not be noticed if it is built on one of the flat faces, as in Figure-A, but will be prominent

in any other orientation. The effect increases as slope of the face increases with respect

to the vertical axis (build direction), as shown in Figures-B and C.

A B C

Part building

In the course of part deposition generally two types of errors are observed and

are namely curing errors and control errors. Curing errors are due to over or under

curing with respect to curing line and control errors are caused due to difference

(variation) in layer thickness or scan position control. Figure 10.1.1 illustrate effect of

over curing on part geometry and accuracy. Adjustment of chamber temperature and

laser power is necessary for proper curing. Calibration of the system becomes

compulsory to minimize control errors. Shrinkage also causes dimensional inaccuracy

and is taken care by choosing proper scaling in x, y and z directions. Polymers are also

designed to have nearly (almost) negligible shrinkage factors. In SLA and SLS

processes problem arises with downward facing layers as these layers do not have a

layer underneath and are slightly thicker, which generate dimensional error. If proper

care is not taken in setting temperatures, curling is frequently observed.

(a) Thicker bottom layer (b) Deformed whole boundary

Figure 10.1.1 Over-curing effects on accuracy in Stereolithography

Part finishing

Poor surface quality of RP parts is a major constraint and is primarily due to

staircase (stair-stepping) effect. Surface roughness can be controlled below a

predefined threshold value by using an adaptive slicing. Further, the situation can be

improved by finding out a part deposition orientation that gives minimum overall

average part surface roughness. Yet, some RP applications like exhibition models,

tooling or master pattern for indirect tool production etc. require additional finishing

to improve the surface appearance of the part. This is generally carried out by sanding

and polishing RP models which leads to change in the mathematical definitions of the

various features of the model. The model accuracy is mainly influenced by two factors

namely the varying amount of material removed by the finishing process and the

finishing technique adopted. An expert operator is required as the amount of material

to be removed from different surfaces may be different and inaccuracies caused due

to deposition can be brought down. A finishing technique selection is important

because different processes have different degrees of dimensional control. For

example models finished by employing milling will have less influence on accuracy

than those using manual wet sanding or sand blasting.

Selection of part deposition orientation

This is one of the crucial decisions taken before slicing the part and initiating

the process of deposition for a particular RP process. This decision is important

because it has potential to reduce part building time, amount of supports required,

part quality in terms of surface finish or accuracy and cost as well. Selection of part

deposition orientation is process specific where in designer and RP machine operators

should consider number of different process specific constraints. This may be a

challenging (difficult) and time consuming task as designer has to trade-off among

various conflicting objectives or process outcomes. For example better part surface

quality can be obtained but it will lead to increase in the building time.

The importance of build orientation:

As more choices of materials and build processes become accessible in layer

based manufacturing, it is increasingly important to identify fundamental problems

that underlie (hidden cause) the entire field. Determination of best build orientation

is one important issue that must be considered in any layer based manufacturing

process. It is crucial to choose a proper build orientation for the rapid prototyping. In

the fabrication of RP parts, the quality of the prototyping, build time, the support

structure, the build accuracy, part cost and load capacity, among others, can be

seriously affected the selection of part build orientation.

A better surface quality yields higher precision of dimensions, which can

facilitate a more accurate test on assembly or functionality of the prototype parts. And

surface quality of rapid prototype parts depends on the stair-stepping effect on

inclined planes and the curved surfaces. Build time is an important factor in building

a part. If the same model takes a different orientation, the build time may vary since

it is largely dependent on the height of a part. Also, in the case of SLA, SLS and FDM,

build time can differ greatly in terms of the amount of support structure. And build

time generally comprises of three elements: data preparation time, part build time,

and post-processing time. The support structure costs money and affects the post-

processing time; we must choose the orientation to handle this aspect. The part cost is

a sensitive factor to the user or designer when determining the part orientation in RP.

The cost includes the material cost (parts and supports), additional equipment cost,

post-processing cost, and labour cost etc. The build accuracy depends on the staircase

(stair-stepping) effect and volumetric error. And the designers should minimise the

volumetric error to get higher accuracy of RP model. At last, for the load capacity

aspect, the same model build in different orientations can have great differences of the

load capacity.