You are on page 1of 7

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/221254289

Programming for evaluating strip layout of progressive dies

Conference Paper · September 2011


DOI: 10.1109/ICDIM.2011.6093334 · Source: DBLP

CITATION READS

1 604

3 authors, including:

Alan C. Lin Minh Tuan Ho


National Taiwan University of Science and Technology University of Strathclyde
56 PUBLICATIONS   599 CITATIONS    26 PUBLICATIONS   113 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

MIGRATE - MIniaturized Gas flow foR Applications with enhanced Thermal Effects View project

All content following this page was uploaded by Minh Tuan Ho on 30 May 2014.

The user has requested enhancement of the downloaded file.


Programming for Evaluating Strip Layout of
Progressive Dies
Alan C. Lin, Ho Minh Tuan and Dean K. Sheu
Department of Mechanical Engineering
National Taiwan University of Science and Technology
Taipei, TAIWAN
alin@mail.ntust.edu.tw

Abstract—A progressive die is an effective tool for efficient and algorithms to screen the solution space and reused previously
economical production of sheet metal parts in large quantities. calculated results for speeding up the bending planning. This
Nowadays, progressive die designers still spend much of their time study may successfully reduce the number of possible
on choosing better layouts among feasible ones. This study alternatives; however, may also eliminate better solutions.
employs Pro/Web.Link, Hyper Text Markup Language (HTML) Kannan and Shunmugam [5] used exhaustive searching to
and JavaScript to develop an application which helps evaluate produce bending layout. Nevertheless, they assume that sheet
automatically strip layouts in Pro/Engineer software environment. metal parts mostly have no more than 8 bends. Unfortunately,
This paper proposes solutions for calculating total evaluation there are, in general, more than 8 punches in practical cases.
score of the strip layout based on four factors: station number
Lin and Dean [6] suggested a modified scheme of exhaustive
factor, moment balancing factor, strip stability factor and feed
search and punch grouping rules so as to find all the feasible
height factor.
layouts with eliminating unwanted searching paths. They also
Keywords-CAD; Manufacturing; Programming proposed an evaluation function for ranking feasible strip
layouts in term of station number, moment balancing and strip
I. INTRODUCTION stability. Their solution, nonetheless, only dues with shearing
In today’s practical and cost-conscious marketplace, sheet operations. The research presented in is paper is the
metal parts have probably became the most versatile products succeeding project of [6] and it proposes a proper solution for
of modern technology. There are a variety of sheet metal dies, automatically evaluating feasible strip layouts in progressive
but progressive die is a good choice for mass producing small dies from their 3D models. The solution helps designers
and delicate parts where tolerances and specifications are quickly fine tune and choose better results among many
relatively severe. A progressive die combines two or more feasible layouts.
stations to perform simultaneous operations as the sheet is II. CRITERIA FOR STRIP LAYOUT EVALUATION
transported incrementally through the die. The design of
progressive die starts with the decision on which operations are In order to choose better strip layouts from many feasible
to be performed at each station, which is called “strip layout.” layouts, each feasible layout must be rank based on a total
In the design process, designers consider the sheet metal part score relevant for that layout. Among many factors in planning
(Fig. 1(a)) to understand its characteristics and specifications. strip layouts that affect the cost and quality of progressive die,
The appropriate shearing and bending punches (Fig. 1(b)) are four factors had proposed by Lin and Sheu [6] for layout
then determined depending on the internal holes, external evaluation function; they are: station number factor ‫ܨ‬௡ ,
outlines and bending features. Next, superimposing the moment balancing factor ‫ܨ‬௕ , strip stability factor ‫ܨ‬௦ and feed
designed punches on the strip (Fig. 2(a)) is a common method height factor. The evaluation score is computed based on these
to achieve strip layout (Fig. 2(b)). It is quite simple because four factors:
the designer merely tries to palace each designed punch on an
appropriate station of the die [1]. In addition to simplicity,
superimposition is also easy to spot any overlapping of the
superimposed punches as well as easy to exchange idea
between designers. However, superimposing of many punches
results in a huge number of possible strip layouts because the
number of possible solutions will grow exponentially with
number of features [2]. Schaffer [3] were probably the
explorer in researching strip layouts for progressive dies.
Computer-aided design techniques are employed for the
automation of die design. However, punch design and layout
are still relied on designer’s wisdom. In order to shrink the
enormous search space, artificial intelligent techniques and
heuristic searching schemes are suggested for overcoming Fig. 1 Solid models: (a) sheet metal part, (b) shearing punches
strip layout issue. Thanapandi et al. [4] had employed genetic
Pi and bending punches Bi

978-1-4577-1539-6/11/$26.00 ©2011 IEEE 229


is known. Since the limitation of calculation in JavaScript, the
Pro/Web.Link application can calculate exactly the centroid
for line segment and arc of circle (as shown in Table 1). Other
types of curve such as Spline, B-Spline, arc of ellipse, etc. are
not considered in this research. (4) Finally, the acting point of
shearing force is calculated as the centroid of its component
edges by the following equations:
σ௡௜ୀଵ ‫ܮ‬௜ ൈ ‫ݔ‬௜
‫ݔ‬௔௖௧௜௡௚̴௣௢௜௡௧ ൌ (4)
σ௡௜ୀଵ ‫ܮ‬௜ 

σ௡௜ୀଵ ‫ܮ‬௜ ൈ ‫ݕ‬௜ (5)


‫ݕ‬௔௖௧௜௡௚̴௣௢௜௡௧ ൌ
σ௡௜ୀଵ ‫ܮ‬௜ 
where ‫ݔ‬௔௖௧௜௡௚̴௣௢௜௡௧ ǡ ‫ݕ‬௔௖௧௜௡௚̴௣௢௜௡௧ : coordinate of acting point of
Fig. 2 An 8-station design: (a) superimposing punches, (b) shearing force, n: total number of sheared edges of that
resultant strip layout shearing punch, ‫ݔ‬௜ ǡ ‫ݕ‬௜ : coordinate of centroid of sheared edge
i, ‫ܮ‬௜ : length of sheared edge i.
‫ܧ‬௏ ൌ ‫ݓ‬௡ ൈ ‫ܨ‬௡ ൅ ‫ݓ‬௕ ൈ ‫ܨ‬௕ ൅ ‫ݓ‬௦ ൈ ‫ܨ‬௦ ൅ ‫ݓ‬௟ ൈ ‫ܨ‬௟ (1)
where ™୬ ǡ ™ୠ ǡ ™ୱ ǡ ™୪ are weighting factors for ୬ ǡ ୠ ǡ ୱ ǡ ୪ ,
correspondingly, Ͳ ൑ ™୬ ǡ ™ୠ ǡ ™ୱ ǡ ™୪ ൑ ͳ and ™୬ ൅ ™ୠ ൅
™ୱ ൅ ™୪ ൌ ͳ.
These four weighting factors are chosen by the designers
who determine how much important each factor contributes to
the strip layout evaluation. All these four evaluation factors are
formulated to range from 10 to 100, which higher score
indicates better efficiency in cost and production. The
evaluation score ୚ itself has relative meaning within feasible
layouts for one part; therefore, it can be used to rank these
feasible layouts so as to find out better solutions for fabricate
that product. In the other hand, the evaluation score ୚ cannot
be used to compare layouts of different parts.
III. FORMULAE OF FORCE CALCULATION
A. Magnitude of shearing and bending forces
If the friction between the punch and the sheet of metal is Fig. 3 Calculating the acting point of shearing force for a
neglected, the shearing force ‫ܨ‬௦ can be estimated form the sample punch
following equation:
Table 1 Centroid of various types of edge
‫ܨ‬௦ ൌ ‫ ܣ‬ൈ ߬௦ ൌ ‫ܮ‬௦ ൈ ܶ ൈ ߬௦ (2)
Type of
Image Centroid
where ‫ܣ‬: sheared area, ‫ܮ‬௦ : total length of sheared edges, ܶ: edge
strip thickness, ߬௦ : shearing strength.
Bending force ‫ܨ‬஻ can be determined by:
߬௦ ‫ݔ‬஺ ൅ ‫ݔ‬஻
‫ܨ‬஻ ൌ ‫ܮ‬஻ ൈ ܶ ൈ (3) Line ‫ݔ‬௖ ൌ
͸ ʹ
‫ݕ‬஺ ൅ ‫ݕ‬஻
segment
‫ݕ‬௖ ൌ
where ‫ܮ‬஻ : total length of bending edges ʹ

B. Acting points of shearing and bending forces


For shearing force, including punching and notching
forces, the method of calculating the acting point is ሬሬሬሬሬԦ ൌ ห‫ܥܩ‬
‫ܥܩ‬ ሬሬሬሬሬԦ ห ‫ כ‬ሬሬሬሬሬሬԦ
݊ீ஼
demonstrated on the sample shown in Fig. 3. (1) First, the ‫ݔ‬஽ ൌ ‫ݔ‬ሺͲǤͷሻ
sheared chain must be broken down into component sheared ‫ݕ‬஽ ൌ ‫ݕ‬ሺͲǤͷሻ
edges without considering the geometrical entities they form. ሬሬሬሬሬԦ
‫ܦܩ‬
Arc of ݊ீ஼ ൌ
ሬሬሬሬሬሬԦ
(2) The length of each edge is then calculated and it should be circle ‫ݎ‬
called ‫ܮ‬௜ for sheared edge number i. (3) Next, the distance of ሬሬሬሬሬԦ ห
ห‫ܥܩ‬
Ƚ െ Ƚ୉
edge’s center off the zero point in both x and y dimension is ‫ ݎ‬ൈ ‫ ݊݅ݏ‬ቀ ୗ ቁ
ൌ ʹ
determined. The center, or centroid, of an arbitrary edge can be ቀ
Ƚୗ െ Ƚ୉

obtained exactly by integral operation if equation of this edge ʹ

230
For bending force, the acting point is assumed as the
centroid of moved region (Fig. 4). Consequently, calculating
acting point of bending force is similar to calculating acting
point of shearing force in punching case. However, outer
contour of the moved surface is used instead of contour of the
punch.

Fig. 4 Acting point of bending force Fig. 5 An 8-station layout: (a) strip layout with superimposed
C. Center of Equivalent Reaction Force punches, (b) acting points of component forces
The centre of equivalent reaction force ( ‫ݔ‬ҧ ǡ ‫ݕ‬ത ) can be IV. PROGRAMMING FOR GEOMETRY TRAVERSAL
calculated as the centroid of its components forces by the
following equations: Geometry traverse means passing from handling one
geometry item (solid, surface, contour or edge) to handling
ቀσ௡௜ୀଵ ‫ܨ‬ௌ೔ ൈ ‫ݔ‬௜  ൅  σ௠
௝ୀଵ ‫ܨ‬஻ೕ ൈ ‫ݔ‬௝ ቁ
another geometry item. Following methods are introduced by
‫ݔ‬ҧ ൌ Pro/Web.Link in order to traverse the geometry of a solid
ቀσ௡௜ୀଵ ‫ܨ‬ௌ೔  ൅ σ௠
௝ୀଵ ‫ܨ‬஻ೕ ቁ block:
σ௠௝ୀଵ ‫ܮ‬஻ೕ ൈ ‫ݔ‬௝
ܶ ൈ ߬௦ ൈ ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔ ൈ ‫ݔ‬௜ ൅ ቇ To traverse the geometry, follow the following steps:
͸
ൌ • Starting from the current working directory of
σ௠ ௝ୀଵ ‫ܮ‬஻ೕ Pro/Engineer, use pfcBaseSession.CurrentModel () to get
ܶ ൈ ߬௦ ൈ ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔  ൅  ቇ
͸ handle to solid model.

σ௝ୀଵ ‫ܮ‬஻ೕ ൈ ‫ݔ‬௝ • From handle of a specific solid model, use
ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔ ൈ ‫ݔ‬௜ ൅ ቇ
͸ pfcModelItemOwner.ListItems() with an argument of

σ௠
௝ୀଵ ‫ܮ‬஻ೕ (6) ModelItemType.ITEM_SURFACE to list all the surfaces
ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔  ൅  ቇ contained in the current solid model. Then use
͸
pfcSurfaces.Item(i) to get handle to surface i.
σ௠ ௝ୀଵ ‫ܮ‬஻ೕ ൈ ‫ݕ‬௝
ሺσ௡௜ୀଵ ‫ܮ‬ௌ೔ ൈ ‫ݕ‬௜ ൅ ሻ • From handle of a specific surface, use
‫ݕ‬ത ൌ ͸ pfcSurface.ListContours() to list all the contours contained

σ௝ୀଵ ‫ܮ‬஻ೕ (7)
ሺσ௡௜ୀଵ ‫ܮ‬ௌ೔  ൅  ሻ in that surface. Then use pfcContours.Item(j) to get handle
͸
to contour j.
where ‫ݔ‬ҧ ǡ ‫ݕ‬ത: coordinate of center of the resultant reaction force, • From handle of a specific contour, use
n, m: total number of shearing forces, bending forces, ‫ܨ‬ௌ೔ ǡ ‫ܨ‬஻ೕ : pfcContour.ListElements() to list all the edges contained
shearing force i, bending force j, š୧ ǡ ›୧ : coordinate of acting in a specified contour. Then use pfcEdges.Item(k) to get
point of shearing force i, ‫ݔ‬௝ ǡ ‫ݕ‬௝ : coordinate of acting point of handle to edge k.
bending force j, ‫ܮ‬ௌ೔ ǡ ‫ܮ‬஻ೕ : shearing length of shearing force i, • From handle of a specific edge, use
bending length of bending force j. pfcEdge.Eval3DData() with an argument of parameter t in
order to get handle to specific 3D point of that edge.
If bending is upward, the direction of bending force is
bottom-up, opposite to the direction of shearing force (top- Once handle of a geometry item, an edge for instance, is
down); therefore, that bending length is assigned minus value, get, we can extract geometry information of that item by using
or ‫ܮ‬஻ೕ ൌ െ‫ܮ‬஻ೕ methods and properties introduced by Pro/Web.Link.

In Fig. 5, there is an example of eight-station layout, and V. PROGRAMMING FOR DETERMINING LIFT HEIGHTS
the zero point O is at the middle of the left-hand side of station Lift height at each station ୧ need to be calculated in order
1. The strip is punched simultaneously by the 8 shearing to determine the lift height needed for the whole strip layout.
punches and 4 bending punches (Fig. 5(a)). Centre of
equivalent reaction force is calculated based on the magnitudes This research concerns about wipe bending type. Wipe
and acting points (Fig. 5(b)) of these component forces by Eqs. bending can be divided into two groups: upward bending and
(6) and (7). Actually, magnitudes of component forces are downward bending.
estimated by Eqs. (2) and (3), and acting points of component For upward bending operation (Fig. 6), the lowest point of
forces are estimated by Eqs. (4) and (5) at earlier stage. the blank need to be higher than the highest point of the die

231
with an amount of S (safety lift height) in order to avoid
collision between the blank and the die. The reference plane of
the die and the reference plane of the strip are chosen for
calculating lift height (shown in Fig. 6(a)), so the lift height for
upward bending (shown in Fig. 6(c)) is estimated by:
‫ܪ‬௜ ൌ ܵ ൅ ‫ܤ‬௜ ൅ ‫ܦ‬௜ (8)
where ‫ܪ‬௜ : lift height need for station i, ܵ: safety lift height,
usually choose S = 2mm, ‫ܤ‬௜ : height of the bending die
(distance between the highest point of the die and the reference
plane of the die) at station i, ‫ܦ‬௜ : height of downward portion of
the blank (distance between the lowest point of the blank and
the reference plane of the strip) at station i.
Fig. 8 The Pro/Web.Link application for evaluating sheet
metal strip layout
Using the sheet metal part in Fig. 9(a) as an example, the 7
shearing punches and 2 bending punches employed to fabricate
the part are shown in Fig. 9(b)

Fig. 6 Upward bending: (a) before bend, (b) bend finish,


(c) lift after bend
In this case, the bending die used to create the upward
bending portion of the blank; therefore, the height of the
bending die ‫ܤ‬௜ and the height of upward portion of the blank ܷ௜
are proportional. If we assume that ܷ௜ is equal to ‫ܤ‬௜ , the lift
height for upward bending can be calculated as:
‫ܪ‬௜ ൌ ܵ ൅ ܷ௜ ൅ ‫ܦ‬௜ ൌ ܵ ൅ ݄௜ (9)
where ݄௜ : total height of the blank at station i.
For downward bending (Fig. 7), the lift height is estimated
by:
‫ܪ‬௜ ൌ ܵ ൅ ‫ܦ‬௜
(10)

Fig. 9 Solid models: (a) sheet metal part, (b) shearing punches
Pi and bending punches Bi
First, the feasible layouts, which are satisfied die design
Fig. 7 Downward bending: (a) before bend, (b) bend finish, rules, for this part are determined manually based on the
(c) lift after bend scheme suggested by [6]. Next, the Pro/Web.Link application
is utilized as an automatic tool for evaluating 12 feasible
For shearing operations, the lift height is similar to layouts of the sample part. Their calculation time, which is
downward bending. However, for the last stations, the blank is needed for the application evaluating each layout, evaluation
cut out of the scrap strip; so, the feed height is obviously as score (with wn/wb/ws/wl = 0.4/0.15/0.2/0.25) and four factors
follow: are shown in Table 2. Notice that Pi, Bj and i symbolize
shearing punch number i, bending punch number j and idle
‫ܪ‬ே ൌ Ͳ (11)
station, correspondingly.
where ܰ: the number of stations.
It is easy to find that different layouts will have various
VI. RESULTS punching areas, bending areas, connecting lengths, and center
of equivalent reaction force and lift height needed; as a result,
A Pro/Web.Link application (Fig. 8) has been developed the evaluation scores will be changed accordingly.
based on the proposed approach and calculation techniques in
order to compute automatically the evaluation score for each In order to demonstrate the calculation process, layout #2
given 3D model of strip layout. The application had been (Fig. 10(a)) is manually evaluated.
executed in Pro/Engineer Wildfire 5.0 environment on a PC.

232
Table 2 Evaluation results of 12 feasible ‫ܮ‬஻మ ൈ ‫ݔ‬ଶ ൅ ‫ܮ‬஻భ ൈ ‫ݔ‬ଵ
൬‫ܮ‬ௌభǡమ ൈ ‫ݔ‬ଵǡଶ ൅ ‫ܮ‬ௌయǡరǡఱ ൈ ‫ݔ‬ଷǡସǡହ ൅ ‫ܮ‬ௌల ൈ ‫ ଺ݔ‬൅ ൅ ‫ܮ‬ௌళ ൈ ‫ ଻ݔ‬൰
͸
layouts for sample part ൌ
‫ܮ‬஻మ ൅ ‫ܮ‬஻భ
൬‫ܮ‬ௌభǡమ ൅ ‫ܮ‬ௌయǡరǡఱ ൅ ‫ܮ‬ௌల ൅ ൅ ‫ܮ‬ௌళ ൰
# Layout
Time
Fn Fb Fs Fl EV ͸
(s) ሺെͷሻ ൈ Ͷͻ ൅ ͷ ൈ ͸͵
1 [P1,P2]+[P3,P4,P5]+P6+B1+B2+P7 8.33 48.57 68.69 51.75 10.00 42.58 ൬ͻǤͶʹ ൈ ͳʹǤͷ ൅ ͳͲʹǤͷʹ ൈ ʹͷǤʹ͵ ൅ ͳͲǤͲͺ ൈ ͶͳǤ͵͹ ൅ ൅ ʹͶǤ͹͸ ൈ ͹͸Ǥ͹͹൰
͸
2 [P1,P2]+[P3,P4,P5]+P6+B2+B1+P7 10.69 48.57 69.28 51.75 40.35 50.26 ൌ
ሺെͷሻ ൅ ͷ
3 [P1,P2]+[P3,P4]+P5+P6+B1+B2+P7 12.77 35.71 81.64 51.70 10.00 39.37 ൬ͻǤͶʹ ൅ ͳͲʹǤͷʹ ൅ ͳͲǤͲͺ ൅ ൅ ʹͶǤ͹͸൰
4 [P1,P2]+[P3,P4]+P5+P6+B2+B1+P7 12.55 35.71 82.13 51.70 40.35 47.03 ͸
5 [P1,P2]+[P3,P4,P6]+P5+I+B1+B2+P7 10.66 35.71 72.13 69.57 10.00 41.52 ൌ ͵ͶǤʹͻͷ
6 [P1,P2]+[P3,P4,P6]+P5+I+B2+B1+P7 10.59 35.71 72.66 69.57 40.35 49.19 σ௠ ௝ୀଵ ‫ܮ‬஻ೕ ൈ ‫ݕ‬௝
7 [P1,P2]+[P3,P5]+P6+B1+B2+P4+P7 12.42 35.71 71.43 51.75 10.00 37.85 ሺσ௡௜ୀଵ ‫ܮ‬ௌ೔ ൈ ‫ݕ‬௜ ൅ ሻ
8 [P1,P2]+[P3,P5]+P6+B2+B1+P4+P7 12.33 35.71 71.96 51.75 40.35 45.52
‫ݕ‬ത ൌ ͸
σ ௠
9 [P1,P2]+P3+P5+P6+B1+B2+P4+P7 14.06 22.86 85.99 51.70 10.00 34.88 ‫ܮ‬
௝ୀଵ ஻ೕ
10 [P1,P2]+P3+P5+P6+B2+B1+P4+P7 13.91 22.86 86.44 51.70 40.35 42.54
ሺσ௡௜ୀଵ ‫ܮ‬ௌ೔  ൅  ሻ
11 [P1,P2]+[P3,P6]+P5+I+B1+B2+P4+P7 11.69 22.86 77.29 69.57 10.00 37.15 ͸
12 [P1,P2]+[P3,P6]+P5+I+B2+B1+P4+P7 11.86 22.86 77.77 69.57 40.35 44.81 ‫ܮ‬஻మ ൈ ‫ݕ‬ଶ ൅ ‫ܮ‬஻భ ൈ ‫ݕ‬ଵ
൬‫ܮ‬ௌభǡమ ൈ ‫ݕ‬ଵǡଶ ൅ ‫ܮ‬ௌయǡరǡఱ ൈ ‫ݕ‬ଷǡସǡହ ൅ ‫ܮ‬ௌల ൈ ‫ ଺ݕ‬൅ ൅ ‫ܮ‬ௌళ ൈ ‫ ଻ݕ‬൰
͸

‫ܮ‬஻మ ൅ ‫ܮ‬஻భ
൬‫ܮ‬ௌభǡమ ൅ ‫ܮ‬ௌయǡరǡఱ ൅ ‫ܮ‬ௌల ൅ ൅ ‫ܮ‬ௌళ ൰
͸
ሺെͷሻ ൈ ሺെͳʹǤ͸Ͷሻ ൅ ͷ ൈ ሺെ͸ǤͲ͹ሻ
൬ͻǤͶʹ ൈ Ͳ ൅ ͳͲʹǤͷʹ ൈ ሺെͳǤʹʹሻ ൅ ͳͲǤͲͺ ൈ ሺെͳͶǤ͹Ͷሻ ൅ ൅ ʹͶǤ͹͸ ൈ ͳͶǤͳ͸൰
͸

ሺെͷሻ ൅ ͷ
൬ͻǤͶʹ ൅ ͳͲʹǤͷʹ ൅ ͳͲǤͲͺ ൅ ͸
൅ ʹͶǤ͹͸൰
ൌ ͲǤͷ͸ʹ

One thing should be notice from the previous equation is


that we group punches which perform at the same station. For
example, shearing punch P1 and P2 perform at station number 1,
so ‫ܮ‬ௌభǡమ and ሺ‫ݔ‬ଵǡଶ Ǣ ‫ݕ‬ଵǡଶ ሻ are the shearing length and acting point
of grouped punch (1, 2).
Next, the centre of the die for a six-station design is 42
௅ ௉௜௧௖௛ൈே௨௠௕௘௥ை௙ௌ௧௔௧௜௢௡௦ ଵସൈ଺
( ൌ ൌ ൌ Ͷʹ) units away from the
ଶ ଶ ଶ
zero point on the left-hand side (Fig. 11(b)).
After obtaining center of the equivalent reaction force (‫ݔ‬ҧ ǡ ‫ݕ‬ത)

and center of the die ( ǡ Ͳ); the following equations [6] are

employed to calculate maximum deviation ‫ܦ‬௠௔௫ and the real
deviation ݀ of the center of equivalent reaction force with
respect to the center of the die (as shown in Fig. 11(b)):
Fig. 10 Feasible layout #2:
[P1,P2]+[P3,P4,P5]+P6+B2+B1+P7 of the sample part: ͳͶ ൈ ͸ ଶ ͵Ͷ ଶ
(a) strip layout, (b) punch superimposing ‫ܦ‬௠௔௫ ൌ ඥሺ‫ܮ‬ȀͶሻଶ ൅ ሺܹȀͶሻଶ ൌ ඨ൬ ൰ ൅ ൬ ൰ ൌ ʹʹǤ͸ͷͷ
Ͷ Ͷ
(1) Stage number factor Fn ݀ ൌ ඥሺ‫ݔ‬ҧ െ ‫ܮ‬Ȁʹሻଶ ൅ ‫ݕ‬ത ଶ ൌ ඥሺ͵ͶǤʹͻͷ െ Ͷʹሻଶ ൅ ሺͲǤͷ͸ʹሻଶ ൌ ͹Ǥ͹ʹͷ
The value of nine punches for a six-station layout (Fig.
10(b)) is calculated by the following equation:
ܰ െ ܰ௠௜௡ ͸െʹ
‫ܨ‬௡ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͶͺǤͷ͹ͳ
ܰ௠௔௫ െ ܰ௠௜௡ ͻെʹ
(2) Moment balancing factor Fb
First, the shearing length ‫ܮ‬ௌ೔ and acting point (‫ݔ‬௜ ǡ ‫ݕ‬௜ ) of
each shearing punch as well as the bending length ‫ܮ‬஻ೕ and
acting point (‫ݔ‬௝ ǡ ‫ݕ‬௝ ) of each bending punch (illustrated in Fig.
11(a)) are investigated from the solid model in
Pro/ENGINEER. So we obtain the magnitude and acting point
of all component forces (shearing and bending forces)
Then, Eqs. (6) and (7) are employed to find the center of
the equivalent reaction force (‫ݔ‬ҧ ǡ ‫ݕ‬ത) for these component forces:
σ௠
௝ୀଵ ‫ܮ‬஻ೕ ൈ ‫ݔ‬௝
ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔ ൈ ‫ݔ‬௜ ൅ ቇ
͸
‫ݔ‬ҧ ൌ
σ௠
௝ୀଵ ‫ܤ‬௝
ቆσ௡௜ୀଵ ‫ܮ‬ௌ೔  ൅  ቇ
͸
Fig. 11 Moment balancing calculation: (a) acting points of
component forces, (b) deviation of center of equivalent
reaction force

233
Finally, the following equation [6] is used to determine ‫ ܪ‬െ ‫ܪ‬௠௜௡ ͳͲ െ ʹ
‫ܨ‬௟ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͶͲǤ͵ͷʹ
moment balancing factor Fb : ‫ܪ‬௠௔௫ െ ‫ܪ‬௠௜௡ ͳͶǤͲ͹Ͳͺ െ ʹ
In summary of these four factors, the evaluation score Ev is
݀ ͹Ǥ͹ʹͷ obtained by Eq. (1):
‫ܨ‬௕ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͳͲͲ െ ͻͲ ൈ ൌ ͸ͻǤ͵ͳͳ
‫ܦ‬௠௔௫ ʹʹǤ͸ͷͷ
‫ܧ‬௏ ൌ ‫ݓ‬௡ ൈ ‫ܨ‬௡ ൅ ‫ݓ‬௕ ൈ ‫ܨ‬௕ ൅ ‫ݓ‬௦ ൈ ‫ܨ‬௦ ൅ ‫ݓ‬௟ ൈ ‫ܨ‬௟
(3) Strip stability factor Fs ൌ ͲǤͶ ൈ ͶͺǤͷ͹ͳ ൅ ͲǤͳͷ ൈ ͸ͻǤ͵ͳͳ ൅ ͲǤʹ ൈ ͷͳǤ͹͸ͳ ൅ ͲǤʹͷ ൈ ͶͲǤ͵ͷʹ
ൌ ͷͲǤʹ͸ͷ
The total connecting length of the sheet metal part in Fig.
12 is 81.490, and it will be reduced gradually by punching VII. CONCLUSIONS
operations performed at every station (illustrated in Fig. 12).
The detail information of connecting length Lk at each station Process planning, which is the core issue in progressive
for layout 2 is investigated from the solid model in dies, is a challenging and time-consuming task which requires
Pro/ENGINEER (shown in Table 3). There is no reduction of experienced designers and their great deal of trial and error.
connecting length at stations 1 and 5; so, connecting length for After utilizing designers’ expertise and design rules, there are
the strip stability factor would consider only stations 2, 3 and 4. still so many feasible layouts for designers to select from. The
subjects for this research are feasible layouts, which are
manually determined using the scheme proposed by Lin and
Dean [6]. In this study, we have suggested an approach for
recognizing punching, notching and bending operations from
given 3D model of strip layout. In addition, we have also
proposed a method for automating layout evaluation and
developed a Pro/Web.Link application which helps designers
quickly determine the most appropriate layouts among many
feasible ones. The followings are the main features of this
application:

Fig. 12 Connecting length Lk at each station of layout #2 of the • 3D models of strip layouts are used as input data for the
sample part process of operation recognitions and calculations.
• The sheet metalworking range varies from punching,
Using the following equation [6] to compute strip stability notching and bending operations.
factor Fs for layout 2 as follow:
• The four factors, or criteria, used for layout evaluation are
‫ܮ‬௞ ͷͲǤ͸ʹ ʹͳǤ͵ʹ ͳ͹ǤͶ͸ combined with their weighting factors. Hence, the
σேିଵ
௞ୀଵ ݇ ൈ ʹൈ ൅͵ൈ ൅Ͷൈ
‫ܮ‬௅௞ ͸ͳǤͳʹ ͶͲǤ͹Ͷ ʹͲǤ͵͹
‫ܨ‬௦ ൌ ͹Ͳ ൈ ൌ ͹Ͳ ൈ evaluation function can be flexibly adjusted to satisfy
σேିଵ
௞ୀଵ ݇ ʹ൅͵൅Ͷ
ൌ ͷͳǤ͹͸ͳ various designers’ preferences.
• The evaluation score can be obtained very quickly; for
Table 3 Strip stability factor for layout #2 of the sample part example, approximately sixteen seconds for an eight-
station layout in comparison with one hour of manual
Reduced
Station
LLk 1.43LLk Punches connecting
Connecting extraction and calculation.
number length (Lk)
length
ACKNOWLEDGMENTS
0 81.490 --- --- --- 81.490
This project was supported in part by the National Science
1 NA NA [P1,P2] 0 81.490 Council, Taiwan, under project number 97-2628-E-011-002-
MY3.
2 61.12 87.40 [P3,P4,P5] 30.87 50.62
REFERENCES
3 40.74 58.26 P6 29.21 21.32 [1] Schubert P. B., Die Methods: Design, Fabrication, Maintenance, and
Application, Industrial Press Co., 1967.
4 20.37 29.13 B2 3.86 17.46 [2] Gupta S. K., “Sheet metal bending operation planning: using virtual
node generation to improve search efficiency,” Journal of
Manufacturing Systems, Vol. 18, No. 22, 1999, pp. 127-139.
5 NA NA B1 0 17.46
[3] Schaffer G., “Computing design of progressive die”, American
Machinist, Vol. 22, 1971, pp. 73-75.
6 0 0 P7 17.46 0 [4] Thanapandi C. M., Walairacht A., and Ohara S., “Genetic algorithm for
bending process in sheet metal industry”, IEEE Electrical and Computer
(4) Feed height factor Fl Conference, Toronto, Vol. 2, 2001, pp. 957-962.
[5] Kannan T. R., and Shunmugam M. S., “Processing of 3D sheet metal
The feed height for the whole strip of that six-station layout components in STEP AP-203 format. Part II: feature reasoning system”,
is determined by Eqs. (9) and (10): International Journal of Production Research, Vol. 47, No. 5, 2009, pp.
1287-1308.
H = max (H1, H2, H3, H4, H5, H6) = max (2, 2, 2, 10, 8, 0) = 10 [6] Alan C. Lin and Dean K. Sheu, “Knowledge-based sequence planning of
‫ܪ‬௠௜௡ ൌ ܵ ൌ ʹ (safe lift height) shearing operations in progressive dies,” International Journal of
‫ܪ‬௠௔௫ ൌ ܵ ൅ ͳʹǤͲ͹Ͳͺ ൌ ͳͶǤͲ͹Ͳͺ Production Research, 2010 (accepted for publication).

234

View publication stats

You might also like