MANUFACTURING

DESIGN
MAE 8510
Part I

Design for Manufacturing

and Assembly
Instructor: Professor A. Sherif El-Gizawy

Mechanical and Aerospace Engineering

University of Missouri-Columbia
Fall 2010
1

AssemblyProcess:AnalysisandDesign
IntroductiontoConcurrentEngineering
The concern of industrial productivity and manufacturing efficiency is at all time high.
The recent direction -in manufacturing development involves utilization of the Unified
Life Cycle Engineering (ULCE) approach to improve the competitiveness of
manufacturing industry through optimization of performance and minimization of total
life cycle cost of the product. The new approach adds several constrains to the product
delivery process. These include: design for manufacturability, reliability, and
maintainability and utilization of cost effective manufacturing techniques for producing
defect-free products.
Companies cannot meet quality and cost objectives with isolated design and
manufacturing engineering operations. The essence of design for manufacture (DFM)
and design for assembly (DFA) approaches is the integration of product and process
design in one common activity to ensure the best matching of needs and requirements.
This will help in identifying the product concepts that are inherently easy to manufacture.

GroupTechnology(GT)inConcurrentEngineering
Grouptechnologyisatoolthatutilizesclassificationandcodingsystemstoidentify
andunderstandpartsimilaritiesandtoestablishparametersforaction.Similarityofparts
couldbebasedontheirgeometricalshapeand/orsimilaritiesintheirmethodof
manufacturing.Figure1showsdesignfamily(commonshape).Figure2shows
productionfamily(commonmethods).Grouptechnologycanresultinsignificant
reductionindesigntimeandeffort.Fromacodethatdescribesthenewpart,itcanbe
determinedwhetherornot,thereisanexistingpartthatcanbemodifiedtogeneratethe
newdesign.Moreover,bynotingsimilaritiesbetweenparts,itisoftenpossibletocreate
standardizedpartsthatcanbeusedinterchangeablyinavarietyofapplicationsand
products.Similarityinmanufacturingmethodscanresultinsignificantcostreduction

duetoreductioninnumberofnewtoolsrequired(fixtures,jigs,gages,molds,etc.)and
machinessetuptime.

Fig.1DesignFamily
Fig.2ProductionFamily

DesignforManufacturingandAssembly
Concurrentengineeringrequiresanalysismethodstoevaluateconceptualdesigns
formanufacturingandassemblydifficultiesandcosteffectiveness.

Design for Assembly

The most effective way to improve a product design for assembly is the reduction
of the number of components involved in the assembly. Reducing part count will reduce
the number of assembly operations and their associated tools. By doing this you reduce
potential process failures. The result will be shorter cycle time for assembly, higher
quality products, and lower production costs. The designer can use the following rules as
guidance reduces the number of separate components in a conceptual design. Each part
will be examined against these rules. Satisfying any of them means that the component
has to stay as to stay as a separate one and can not be combined with others.

Rules for Maintaining Separate Components in the Design

1. If the part has to more relative to other parts in the assembly for functional reasons.
2. If the part has to be made from different materials or to be isolated from other
components in the assembly for functional reasons.
3. If the part is required for disassembly of the product to gain access for maintenance
and repair.
Figure 3 illustrates the idea of improving the design of assembly of a clipper throw
reducing the number of parts.

Fig. 3
Case Study
A conceptual design for a motor drive assembly is required to sense and control
the position of the drive assembly along two steel guided rails. The motor must
be fully enclosed with a removable cover for access to adjustment of the position
sensor. The basic requirements in this design are a rigid base designed to allow
sliding up and down the quid rails and to support the motor and house the sensor.
An exploded view of the initial design is shown in Fig. 4.
Applications of the design rules for reducing the number of components in the
assembly would proceed following the steps in table 1.

Figure 4. Design Concept for Motor Drive Assembly

Table 1: Application of DFA Rules for Part Reduction

(#1 indicates part required and # 0 indicates a separate part is not required.)

Part (quantity) Part no. Decision Justification

Base(1)
1
1 First component to be assembled
Bushing(2)

Motor(1)

Motor screw(2)

Sensor sub.(1)

Set screw(1)

Stand off(2)

End plate(1)

End plate
screw(2)
Plastic
bushing(1)
Cover(1)

9
10

Cover screw(4)

12

11

0 They can be combined with the base if made

from the same material
1 This sub assembly represents the basic
function of the design and can not be
combined with other components
0 Separate fasteners should be always the last
method for integrating different components
(they take long time for assembly
1 Basic standard component for the design
function
0 Separate fasteners should be always the last
method for integrating different components
0 They can be combined with the base if made
from the same material
1 It is required for purpose of disassembly and
maintenance
0 Separate fasteners should be always the last
method for integrating different components
0 It can be combined with the end plate using
same material, plastic in this case
0 It can be combined with the end plate using
same material, plastic in this case
0 Separate fasteners should be always the last
method for integrating different components

From the analysis shown in Table 1, the ideal design will consist of four
components instead of 19 components in the original design. The motor, sensor, base,
and a plastic cover that can snap on without need for separate fasteners. These four
components represent the theoretical minimum number of components needed the
product function. However, there will be always some practical and economical
constraints causing the actual number to be higher than the theoretical one. For example,
the two screws, which support the motor inside the base, and the set screw which secures

the sensor in its place, have to be kept in the new design. This is because the alternative
design solution will be more costly.
The DFA analysis can be used to estimate cycle time and cost needed for
assembling the proposed design. In this analysis, the design efficiency that relates the
theoretical cycle time of ideal design to the actual one, can also be determined. Table 2
shows the analysis results of the original design. The design efficiency is only 7.5%.
The new design is shown in figure 5. All fasteners are eliminated except those
required to secure the motor and the sensor in place. The bushings (#2) are combined
with the base, which will be made out of nylon, with low friction coefficient. The end
plate (#8), cover (#11), plastic bushing (#10), and six screws (#9, #12) replaced with one
snap on plastic cover. The results of DFA analysis for the modified design are shown in
Table 3. The efficiency of the new design is 26% (up about 350% of the initial one).
Table2
ResultsofDesignforAssembly(DFA)AnalysisfortheMotorDriveAssembly
ProposedDesign
Number Theoretical part count Assembly time(s) Assembly cost (cents)
Base
bushing
Motor sub.
Motor screw
Sensor sub.
Set screw
Stand-off
End plate
End plate
screw
Plastic
bushing
Thread leads
Reorient
Cover
Cover screw
Total

1
2
1
2
1
1
2
1
2

1
0
1
0
1
0
0
1
0

3.5
12.3
9.5
21
8.5
10.6
16
8.4
16.6

2.9
10.2
7.9
17.5
7.1
8.8
13.3
7
13.8

3.5

2.9

1
4

0
0

5
4.5
9.4
31.2

4.2
3.8
7.9
26

19

160

133

Design efficiency = 4x3/160 = 7.5 percent

10

Figure5.ModifiedDesignforAssembly
Table3
ResultsofDesignforAssembly(DFA)AnalysisfortheMotorDriveAssembly
Redesign
Part

Base
Motor sub
Motor screw
Sensor sub.
Set screw
Thread leads
Plastic Cover

1
1
2
1
1

Theoretical part
count
1
1
0
1
0

Total

Design efficiency = 4x3 /46.0 = 0.26 = 26%

11

Assembly
Time(s)
3.5
4.5
12.0
8.5
8.5
5.0
4.0

Assembly Cost
(cents)
2.9
3.8
10.0
7.1
7.1
4.2
3.3

46.0

38.4

Design concept

Suggestions
for
simplification
of products
structure

Design for
assembly (DFA)

Selection of
materials and
processes and
early cost
estimates

Suggestions
for more
economic
materials and
processes

Best design
Detail design
for minimum
manufacturing
costs

Design for
manufacturing
(DFM)

Prototype

Production

Fig. 6 Typical steps taken in a simultaneous engineering study using

DFA, and DFM techniques

12

Dimensional Accuracy in Assembly Design

Tolerancing methods have been incorporated in design and the manufacturing industry for
hundred of years. It was introduced because machines are incapable of fabricating parts to
exact size. There will always be deviations from ideal dimensions. Tolerancing methods
provide and analyze an acceptable range for part or assembly dimensions to deviate. Thus,
the purpose of this report is to evaluate efficient techniques that are viable in solving the
tolerance stack up problem for most common engineering applications. This chapter presents
an introduction to the type and scenarios of tolerancing. The objectives and sources of
variations are discussed.

Tolerancing
In general, three categories of tolerancing methods have been developed in the industry:
parametric tolerancing, geometric tolerancing, and operational tolerancing. Operational
tolerancing focuses on process design while the parametric and geometric ones are used in
product design.
Parametric Tolerancing
Parametric tolerancing is also referred as dimensional tolerancing. It is based on ordinary
dimensions (or vector proxies) such as length and width of a part but not the characteristics
of part features. Worst-case limit tolerancing, statistical tolerancing, and vectorial tolerancing
are versions of parametric tolerancing. The first two can be calculated using dimension chain.
All dimensions of a part or assembly represent a chain. For vectorial tolerancing models, the
dimensions of part or assembly are based on its manufacturing processes. It emphasizes on
orientation by means of vectors. This method can be used to account for small kinematic
adjustments during assembly.

13

Geometric Tolerancing
Geometric tolerancing applies tolerances directly to attributes of features. These
include concentricity of a hole, flatness or parallelism of surfacesetc. It is added to limit
the form, location, and orientation of parts. Geometric tolerancing is based on three central
notions:
1. Conformance to a geometric tolerance requires that a surface feature, or an attribute of a
feature (e.g., the axis of a hole), lie within a prescribed spatial zone. Note that this is a
true geometric criterion, whereas conformance to a parametric tolerance is inherently
numeric.
2. A geometric tolerance usually controls explicitly only one specified property of a feature,
such as form (flatness, cylindricity) or position. However, subtle interactions between
different tolerances on the same feature can complicate matters considerably.
3. Some containment zones (e.g., for form) can be positioned freely in space, whereas
others (e.g., for position) are located on parts through reference features called datums.
The use of containment zones deals directly with imperfect form and is the hallmark of
geometric tolerancing.
The semantics of geometric tolerancing are established primarily by a set of rules for
implementing datum systems from physical part features and another set of rules for
constructing spatial zones. The datum system is essential as a setup position for machine
parts. The recent introduction of Computer Aided Drawing (CAD) system based on datums
has pioneered the way for complete data transformation to manufacturing engineering.

14

Figure 1: Geometric variations

Operational Tolerancing
Operational tolerancing is used for process design using tolerance charting. Tolerance
charting is a precise mathematical technique for planning a manufacturing process for
mechanical piece parts in process having successive cuts and processes (etching,
electroplating, heat treating, etc.). Operational tolerancing is often used together with
parametric and geometric tolerancing to guarantee reasonable cost by means of
manufacturing processes.

Worse -case Methods

Usually, the assignment of parametric and geometric tolerance values utilizes a worst-case
scenario. The cumulative variability grows linearly with the length of the dimension links.
The zones or intervals of mating features in an assembly can be disjoint to enable a part
feature to lie in a preferred spatial zone. Generally, tolerances calculated would be tighter
than those obtain using statistical methods. Dimensions will be assigned minimum and
maximum values and the entire part takes a worst possible value.

15

Statistical Methods
Statistical tolerancing is merely an extension of classical parametric tolerancing that
provides an alternative to worst- case design. Currently, the statistical methods are preferred
for higher precision cases. It satisfies the Normal (Gaussian) Distribution. Basic dimensions
of a part are assumed as random variables.

Figure 2: Gaussian distribution in statistical method

1.2 Sources of Variations in Assemblies
In assemblies, different components are put together. With certain magnitude of variations in
every part, the total variation of an assembly would be the sum of its parts. This is called
tolerance stackup. This shows that although every component may satisfy the tolerance
allocated, the total assembly tolerance may be out of design specifications. By knowing the
sources of these variations, one can understand better how components of an assembly
interact between each other. There are three main sources of variation that must be accounted
for in mechanical assemblies:
1.

Dimensional variations (lengths and angles)

2.

Form and feature variations (flatness, roundness, angularity, etc.)

3.

Kinematic variations (small adjustments between mating parts)

16

Limits and Fits

Definitions

Figure 3: Illustration of Definition

Figure 4: Tolerance Type

17

Tolerances can be expressed with respect to the basic size as deviation in both upper and
lower directions (bilateral tolerancing), or only in one direction, if the consequences of
inaccuracy in that direction are less dangerous (unilateral tolerancing). Limit dimensioning is
just a different way of expressing tolerances based on bilateral or unilateral tolerancing.
Below are definitions of the terms in Figure 5:

Basic size: the size to which limits or deviations are assigned. The basic size is the
same for both members of a fit. It is designated by the number 40 in the example:
40H7. Sometimes it is also called Nominal size.

Deviation: the algebraic difference between a size and the corresponding basic size.
Algebraic means it can have positive and negative values.

Upper Deviation: the algebraic difference between the maximum limit size and the
corresponding basic size.

Lower Deviation: the algebraic difference between the minimum limit of size and the
corresponding basic size.

Fundamental Deviation: that one of the two deviations closest to the basic size. It is
designated by the letter H in the example: 40H7.

Tolerance: the difference between the maximum and minimum size limits on a part.

Tolerance Zone: a zone representing the tolerance and its position I relation to the
basic size.

International Tolerance Grade (IT): a group of tolerances which vary depending on

the basic size, but which provide the same relative level of accuracy within a given
grade. It is designated by the number 7 in the example: 40H7(IT7).

Hole Basis: the system of fits where the minimum hole size is basic. The fundamental
deviation for a hole basis system is H.

Shaft Basis: the system of fits where the maximum shaft size is basic. The
fundamental deviation for a shaft basis system is h.

Clearance Fit: the relationship between assembled parts when clearance occurs
under all tolerance conditions.

18

Interference Fit: the relationship between assembled parts when interference occurs
under all tolerance conditions.

Transition Fit: the relationship between assembled parts when either a clearance or
interference fit can result depending on the tolerance conditions of the mating parts

Preferred Fits
Tolerance zones are used to establish preferred fits as shown in Fig. for hole basis.
Normally, the hole basis system is preferred in mechanical design. Hole basis fits have
fundamental deviation of H on the hole. Full description of both systems is given in
the table of preferred fits. The hole basis and shaft basis fits are combined with the
preferred sizes to form the limits and allowances Tables. The tables for hole basis system
are given.

19

126
1

Design for Parts Handling and Mating.

General Design Rules
The design rules for assembly can be classified into two parts: Design guidelines for
handling (recognizing, orienting and moving the part); and Design guidelines for mating
operation (insertion, and fastening of part with another part).
Design Guidelines for Part Handling
1. Design parts that have end to end symmetry and rotational symmetry about the axis of
insertion,(Fig.1a).
2. For non-symmetric parts, make them more pronounced asymmetric,(Fig.1-b).
3. Provide features that prevent jamming and nesting,(Fig. 1-c).
4. Avoid features that will allow tangling of parts,(Fig. 1-d).
5. Avoid parts that are too small, slippery, sticky, or hazardous to handle (Fig2.).
Design Guidelines for Part Mating Operations
1. Design for smaller resistance to insertion following the general design features for
compliant parts discussed in section 3 (see Figures 4-5) .
2. Use unidirectional assembly about one axis if possible.
3. Avoid parts that require holding down for further assembly of other parts.
4. Provide features that can help in locating and orienting the part during insertion
(Fig3.).
5. Select the least expensive mechanical fastening method in your design starting with
this order: snap-fitting, plastic bending, riveting, and screwing (see Fig. 6-7..).
6. Avoid the need of repositioning the product in the fixture during assembly.
The above guidelines are simply a set of rules that provide the designer with a qualitative
background to help in generating a design that is suitable for assembly. However, they do
not provide any quantitative mean to evaluate a design for assembly and they do not help

226
2

in ranking different design solutions in terms of their effectiveness for assembly. In recent
years, a knowledge based system has been developed to systematically evaluate each
design for assembly in terms of cycle time, cost, and the overall design efficiency. This
quantitative approach for DFA will be discussed in details in the next few sections.

Fig. 1 Geometric Features Affecting Part Handling

326
3

Fig. 2 Special Features That Affect the Handling Difficulty

Fig.3 Design to Aid Insertion

426
4

Fig. 4 Use of funnel-shaped openings and Tapered ends to facilitate

insertion

Fig. 5 Use of Snap Fit Principal Instead of Screw Fasteners To Reduce

Assembly Cycle Time

526
5

Fig. 6 Common Fastening Methods

Fig. 7 Design Parts with Self Alignment

626
6

Geometric Attributes of Assembled Parts

A suitable classification system of different geometric attributes of parts to be assembled
provides a great assistance for economical assembly process and system design. The
main objective of this classification system is to allow for parts to be divided into groups
having similar handling and insertion problems. One design for assembly solution will fit
all parts in the same group. One tooling system will probably be sufficient for all the
family members. Thus cost and development time will be reduced. More over, with the
aid of motion and time study the handling and insertion time needed for assembly of each
class can be estimated.

Fig. 3 Envelopes

726
7

Envelopes.
Envelope of a part represents the smallest cylinder or regular prism, or rectangular prism
that can enclose the body of the part (Fig. ).
Degenerated Envelope.
Degenerated envelop is the cylinder or regular prism, or rectangular prism obtained by
eliminating all small projections of the original part (Fig. ). The definition of D.E helps
in determining the basic shape of a component (rotational, rectangular etc.).

Fig. 4 Definition of Envelop and Degenerated Envelop

Features
A feature can be defined as any of the followings:

The Space between the envelop and D.E. that is not occupied by the part ( step or
chamfer).

Space enclosed by the D.E. but not filled with material (hole and groove).

Space not enclosed by the D.E. but filled with part material (projection).

826
8

Fig. 5 Methods of defining the degenerated envelops

Rotational Symmetry
The part has a rotational symmetry when its orientation can be repeated with rotation
through an angle about an axis passing through the centroid of its D.E.
Principle Axis
The principle axis is the one that is perpendicular to the cross section of D.E. and passing
through its centroid.
Transverse Axis
The transverse axis is the one that is perpendicular to principle axis of D.E. and passing
through its centroid.

926
9

Alpha Symmetric
The part that repeat its orientation end to end when rotated about its transverse axis 180,
is alpha symmetric (Fig. ).
Beta Symmetric
The part that repeat its orientation when rotated about its principle axis, is beta symmetric
(Fig. ).
Figure

shows parts with different rotational symmetries

Fig.

Examples of Alpha and Beta Symmetries

Fig. Alpha and Beta Rotational Symmetries for various Geometry

1026
10

Part Thickness
The thickness of a cylinder is defined as its radius. For non cylindrical part, the thickness
is defined as the maximum height of the part with its smallest dimension extending
from a flat surface (Fig. ). Cylindrical part having diameter greater than or equal
to the its length, will be treated as non cylindrical part.

Fig.4 Definition of Part Thickness

1126
11

Part Size
The size of the part (the major dimension) is defined as the largest non diagonal
dimension of the parts projection on a flat surface (Fig. ).

Fig.5 Definition of Part Size

1226
12

Effects of Operational Constraints on Mating

Operational constraints are concerned with the interactions between mating parts.
Insertion and fastening consist of variety of basic operations. They include: peg-in-hole,
screw, weld, rivet, force fitetc. The common operational constraints that affect the
design for assembly are:

Accessibility for assembly location.

Ease of operation of assembly tools including operators hands.

Visibility of assembly location.

Ease of alignment and positioning during assembly.

Depth of insertion.

If the assembly associate cannot see the operation and when different obstruction are
present, the cycle time and associated cost for insertion will be large. Figure..shows
examples of restricted access and restricted vision with different degrees of difficulty.
Holding down is another operational constraint that reduces the design efficiency for
assembly. It is required to maintain position and orientation of parts during subsequent
insertion operations.

Fig. ---Operational Constraints

1326
13

Special Mating Operations (Mechanical and Metallurgical Bonding)

1426
14

Knowledge Base System for DFA

Using a classification schemes and time and motion studies, the following two data bases
for handling and insertion were established. Figure---displays the average time in seconds
required for handling a part. Both geometric attributes and operational constraint effects
are considered in the estimated values. Similar data base for estimating insertion times is
displayed in Fig--. A form for DEA work sheet is shown to help in organizing the
analysis. A quantitative index to compare different designs for assembly is presnted in
this sheet as Design Efficiency .

= Nmin ta / tma
Where Nmin is the theoritical minimum number of parts, ta is the assembly time for one
ideal part (3 seconds), and tma is the estimated time based on the data base for assembly.

1526
15

1626
16

1726
17

1826
18

Theoretical Min. No. of Parts

Operation Costs, Cents

Seconds
Total Operation Time

Insertion Time/Part

Insertion Code

Handling Time /Part

Handling Code

Number of Operation

Part ID #

Name of Assembly

tma

Cost

DFA WORKSHEET

1926
19

Nmin

=3 x Nmin/ tma =

Example:
The diaphrag,m assembly given in Fig.consists of 5 components. Analyze the Design
for assembly and recommend a new one with improved efficiency for assembly.
Solution:
1. Use the exploded 3D View and give identification number to each component.
2. Use DFA worksheet for evaluating the given design starting with the first part to be
added to the fixture and the proceed until the assembly is complete.
3. Calculate the total cycle time, cost, minimum theoretical number of parts, and
efficiency.

Redesign for Assembly

1. Reduce the number of parts.
2. Improve geometric features for easy handling and insertion.
3. Repeat DFA analysis to evaluate the improvement.
Figure shows the improved design. The new design efficiency is 84% up from 15.3% of
the original design

2626
26

Dia 8x16

48x10x14

Dia 12x1.5
Dia 8x7

Fig.Original Design and Work Fixture

Fig.Modified Design and its Fixture

2726
27

Handling Code

Handling Time /Part

Insertion Code

Insertion Time/Part

Seconds
Total Operation Time

11

1.8

00

1.5

6.6

6.6

Nuts

13

206

00

1.5

6.6

6.6

Washers

03

1.69

12

6.69

6.69

Plate

30

1.95

02

2.5

4.45

4.45

Bearing Housing

10

1.13

38

6.0

14.26

14.26

Screws

39.12
tma

39.12
Cost

2
Nmin

Theoretical Min. No. of Parts

Number of Operation
2

Operation Costs, Cents

(1Cent/s)

Part ID #
1

Name of Assembly
Diaphragm Assembly
Initial Design

=3 x Nmin/ tma
=3x2/39.12=15.35%

Number of Operation
Handling Code

Handling Time /Part

Insertion Code

Insertion Time/Part

Seconds
Total Operation Time

1
1
03
1.69
00
1.5
3.19
3.19
1
Plate

2
1
30
1.95
30
2.0
3.95
3.95
1
Housing

2726
27

Theoretical Min. No. of Parts

Operation Costs, Cents

Part ID #

DFA WORKSHEET
Name of Assembly
Diaphragm Assembly
Modified Design

7.14
tma

7.14
Cost

DFA WORKSHEET

2826
28

2
Nmin

=3 x Nmin/ tma = 3x2/7.14=84%