Computer Aided Design of a

Pattern and Risers for Casting

Jong Cheon Park Processes: Part 1
Kunwoo Lee
An interactive computer program has been developed to design a pattern and risers
Department of Mechanical Design and Pro-
for the production of castings of high quality. In our system, the user models the
duction Engineering,
shape of a final product by using the system's modeling capability, a pattern is
Seoul National University, generated in a three-dimensional model by eliminating the holes and adding
Seoul, Korea shrinkage allowances and drafts, the proper riser is created automatically, and they
are united together to yield a three-dimensional model of this portion of a mold
assembly. The mold can be completed after the runners and the gating systems are
designed, modeled, and united which will be described in Part 2 of this work. The
unique feature of this work is a realization of an automatic design of the pattern and
risers by integrating the modeling capabilities and the design equations used in com-
mercial practice.

Introduction
The use of computer-aided design and computer-aided linked together to form a casting CAD system. In this paper,
manufacturing in industry increases the efficiency of the product modeling capability, automatic design of a pattern,
design and manufacturing processes. The resulting high pro- and automatic design of risers will be described. The remain-
ductivity makes the use of CAD/CAM an attractive and prof- ing topics will be covered in Part 2.
itable goal for industry. Our work is unique in that it generates a mold in a complete
An ideal CAD system for the casting process would three-dimensional geometric model by uniting the three-
minimize the time delay between the design stage and the dimensional models of a pattern, risers, and gating systems, so
manufacturing stage of a product. In the CAD/CAM area, that a finite element analysis pre-processor, NC processor, etc.
there exist already the technologies which minimize the time can be linked directly.
delay between those two stages. To apply the existing The casting material to be handled in the system is primarily
technologies to the casting process, the three-dimensional ductile cast iron but the same method may be applied to gray
model of a casting mold is the essential part. To produce a iron.
three-dimensional model of a casting mold, it is necessary to
prepare the proper pattern, the riser, and the gating system.
These are designed based on the geometry of the final product Related Work
to a large extent. This fact indicates the possibility of creating The use of computer-aided design as a tool for the casting
a CAD system which generates the three-dimensional model of industry has been addressed by Graham [1] and Welbourn [2].
a casting mold once the geometry of the final product is Graham puts emphasis on the usage of existing CAD systems
specified. for the casting industry and Welbourn describes the
According to this point of view, the objective of this work is capabilities that the casting CAD system should have. Work
the development of a CAD system for casting processes by by Berry and Pehlke[3] has been directed at modeling a casting
which the complete three-dimensional model of a casting mold for thermal analysis; however, direct pattern and mold design
can be automatically generated. To develop this system, the has not been discussed. Heine and Uicker [4] have developed
following problems should be solved: the CAST METALS PROGRAM in which riser design and
• Product modeling capability freezing order calculations are performed. Gerhardt [5] also
• Automatic design of a pattern developed a program to compute the dimensions of a gating
• Automatic design of risers and a risering system by a systematic/logical approach. The
• Automatic design of gating systems Dynarad Corporation developed software for calculating
• Completion of a mold risers and gating systems, NO VAC AST [6], by using some ax-
In our system, each of these segments is developed and ioms and certain well-established empirical information. None
of the works above generates a mold in a complete three-
Contributed by the Production Engineering Division for publication in the
dimensional geometric model. Smith and Lee [7] developed a
JOURNAL OF ENGINEERING FOR INDUSTRY. Manuscript received May 1989. procedure that generates the proper patterns in complete

Journal of Engineering for Industry FEBRUARY 1991, Vol. 113/59

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Copyright © 1991 by ASME
 three-dimensional geometric models; however, the design of a
mold was not discussed.

Product Modeling
Since the design of a mold is heavily dependent on the shape o
H i_ |2J
N

r~7
N2 1
of a final product, there must be a way to provide the 1
geometric model of the final product which will act as an input L !
to the casting mold design system. To this end, the modeling
capability of a three-dimensional object was implemented by
incorporating the routines of' a constructive solid geometric
modeling system. The modeling capability of this system is
similar to that of commercially available constructive solid (a) (b)
geometric modeling systems [8, 9]. Hence, the Boolean opera- Fig. 1 The modification of a pattern based on shrinkage allowance
tions, such as union, subtraction, and intersection, can be per-
formed on the primitives. The primitives supported in this
product modeling system are box, cylinder, cone, prism,
sphere, and torus. As can be inferred from the supported
primitives, a product bounded by sculptured surfaces cannot
be handled in the system.

Automatic Design of a Pattern

Pattern Requirements. Since all cast metals in a foundry
will shrink in size when the metal changes from a liquid state
to a solid state, and since a further decrease in volume will also
occur as a result of cooling from a high temperature in the Fig. 2 Parameters to define a conical surface
solid state, the pattern must be designed to compensate for
shrinkage. Also, the pattern must be designed such that it can
be withdrawn easily from the mold without damaging the their surface type, e.g., planar, cylindrical, or conical.
mold itself, so a taper, so-called the pattern draft, is required In the current implementation, the shrinkage allowance
on all the faces parallel to the direction of withdrawal. Addi- can be handled for planar, cylindrical, and conical
tionally, the pattern has neither cavities nor holes because they faces only.
are taken care of by placing cores in the mold later. Therefore, For a planar face, calculate the increment (LI) by
all the holes in the final product model must be eliminated multiplying the normal distance of the face from the
before the pattern drafts and the shrinkage allowances are center (LO) by the factor of [shrinkage ratio/
considered. (1-shrinkage ratio)], and move the face in its outer
normal direction (Nl) by the increment LI as il-
Elimination of Holes. The first step taken to design the lustrated in Fig. 1(b). When a face is moved, the sur-
pattern is to eliminate all the holes in the final product. First, a face equation associated with the face is updated, and
hole is detected by the user's picking of all the edges contained the updated surface is intersected with the surfaces
in the hole loop on the outer face of the final product. Next, associated with the adjacent faces to the moved face to
our system identifies all the inner faces of the hole, which are yield the new edges. These new edges are intersected
the faces associated with the selected edges except the outer with each other to yield the new vertices.
face. Finally, all entities geometrically connected with the hole
faces, such as the points, vertices, edges, and dimensions, are • For a cylindrical face, the original cylindrical face is
deleted. Once the holes are eliminated, the pattern is ready to replaced by a cylindrical face with an enlarged diameter
be resized for the shrinkage allowance. scaled by a factor of [1/(1-shrinkage ratio)] if it is
convex. Reversely, in the case of a "concave" face, the
Shrinkage Allowance. Shrinkage allowance is the amount diameter is scaled down by a factor of [1/(1 + shrinkage
added to the pattern to compensate for the solidification and ratio)] as shown in Fig. 1(b). The new edges and ver-
cooling phenomena. Thus, the pattern must be made larger tices for the new face are generated in the same way as
than the product to provide the correction for all the contrac- for the planar face.
tions. It is known that each face of the pattern shrinks in a • For a conical face, a procedure similar to the method
direction such that it breaks away from the sand mold. Thus applied to the cylindrical face is followed. In our
the dimensions of "convex" faces have to be enlarged while system, a conical surface is defined by a base point
those of "concave" faces have to be reduced compared to (PI), an axis direction (Ul), radius in the base plane
those of the final product. (Rl), and semi-convergence angle (6) as shown in Fig.
For this development, our system has been implemented on 2. The system defines a new conical face by adjusting
the following assumptions: First, a linear correction for the only the base radius of the cone while the remaining
contraction is usually acceptable [7] even though the true con- parameters are invariant. Here, the scaling factors are
traction is volumetric. Second, the mass center of a product determined in the same way as for the cylindrical face
exists inside the product itself and approximately coincides depending on the convexity or concavity of the conical
with its thermal center. This assumption holds if the shape of face. Once a new conical face is defined, the original
the product is not excessively complex or thin. Thus, the pro- conical face is replaced by the new face and then the
cedure for generating the pattern with consideration of the new edges and vertices associated with the new face are
shrinkage allowance can be described in the following steps. obtained in the same way as for the planar face.
• Calculate the coordinates of the mass center of/he Pattern Draft. After the pattern is resized, the remaining
product and regard them as the center of shrinkage modification is adding the pattern drafts. To add the pattern
(PO), as shown in Fig. 1. drafts, a parting plane should be defined first. A parting plane
• Search all the faces bounding the pattern and find out can be defined by its location and orientation. For easy im-

60 / Vol. 113, FEBRUARY 1991

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org Transactions
on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use of the ASME
 ^>
O
Parting plane
Semi—convergerice Semi—convergence
angle = o a n g f e = +10°

Fig. 3 Separation of a pattern C a)

d i r e c t i o n of withdrawl
A

reference face

draft angle
O
O
Semi—convergence Semi—convergence
angle = o a ngl e = — 1 0°
Fig. A Draft operation on the pattern
( b )

plementation, the orientation of the parting plane has been Fig. 5 Draft operation on a cylindrical or a conical surface. Draft angles
of +10 degrees were provided in (a) and (b).
fixed to be the X-Y plane and thus only the location input is
required. Also, a stepped parting plane is not supported in the
current implementation. Once the parting plane is defined, the angle of a new surface is calculated as follows. If the
pattern is cut into two bodies as shown in Fig. 3. As in Fig. 3, original conical surface is "convex" as in Fig. 5(a), the
the bodies are separated for easy draft operation even though new semi-convergence angle becomes the original semi-
they will be glued together after the draft operation. convergence angle plus the draft angle. In the "con-
The principles used in the implementation of the draft cave" case, as in Fig. 5(b), the new semi-convergence
operation are described in the following steps. angle would be the original semi-convergence angle
• For each face parallel to the direction of withdrawal, minus the draft angle. Once the new parameters are
the user interactively assigns the reference face and the determined as discussed above, the new conical surface
draft angle shown in Fig. 4. In our system, the sign of is defined.
the draft angle is positive when the volume of the pat- By applying the draft operation as described above, the pat-
tern as a result of the draft operation increases. tern will be created with the proper draft. In the current im-
• The surface type of the face to be tilted is identified. plementation, the surface type of the face to be tilted is limited
For a planar face, the system searches one point on the to planar, cylindrical, and conical.
common edge between the reference face and the face
to be rotated and determines the rotational axis direc- Automatic Design of Risers
tion by taking the cross product between outer normal In this system, Karsay's applied risering method [10] has
vectors of the faces. Then the system computes the been employed and implemented. His method is based on a
rotational transformation matrix by means of the draft modulus which is the casting volume divided by its cooling
angle, the rotational axis, and the point on the rota- surface area, the metallurgical quality of the iron, and the
tional axis, and thus defines a new surface by applying mold rigidity. The family of applied risering methods is shown
the transformation matrix to the original surface. For a in Fig. 6[10], All the methods for riser design introduced in
cylindrical or a conical surface, the face is treated as a Fig. 6 were implemented in the system as discussed in the
conical face because a cylindrical face can be con- following sections.
sidered as a special case of a conical face for which the
semi-convergence angle is zero. Note that the semi- Selection of Risering Method. The first step to be per-
convergence angle is defined to be positive when the formed in designing a riser is to select the proper risering
axis direction of the cone points toward the apex. The method. The following explanation refers to Fig. 6. First, the
procedure for drafting a conical face is described as average modulus of a pattern is calculated and the proper
follows with reference to Fig. 5. First, the system iden- risering method based on the modulus is recommended by the
tifies the axis direction of the conical surface. If the system. The modulus of a pattern can be easily calculated
axis has a reverse direction with respect to the outer because the volume and the surface area of the pattern were
normal vector of the reference face, the axis direction already computed when it was modeled. Then, the proper
of the conical surface is reversed and thus the sign of risering method in consideration of the mold rigidity is
the semi-convergence angle is reversed too. The arrows selected interactively by the user as recommended by the
in Fig. 5 represent the axis direction and the outer nor- system even though the user can select a different risering
mal vector of the reference face after the adjustment method.
described above. Next, the base point and the base In this paper, our work is focused on the riser design based
radius of a new conical surface are obtained from the on pressure control risering which is the most popular in prac-
original conical surface. Finally, the semi-convergence tice. Even though theriser design based on the directly applied

Journal of Engineering for Industry FEBRUARY 1991, Vol. 113/61

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 11 1
Mold
1 1
Mt cm il.O

-0.9
1 1 N\Vi
2 . 0 - -0.8 M w >*>'
-0.7 o°"
Weak Mold Strong Mold 1 1
-0.6

-0.5 sM
1 1
1.0—1-0.4 6 •^
Modulus(cm) Modulus(cm) r,n u
-0.3 _,
1 1
-0.2
1 /*
> 0.4 < 0.4 <2.0 > 2.0 -ni"
0.2-03- 0.4-0.5 0.6 -0.7 J 0.8-0.9 - I.O-U-, -1.2- Min.
1 ,i _,l . it ll
"1
0 1.0 2.0 3.0 Mem
Fig. 7 Relationship of geometric modulus to transfer modulus

enced by metallurgical quality is shown in Fig. 7 [5,12]. In

order to calculate the transfer modulus of each segment, the
Pressure Riserless Design curves in Fig. 7 were fitted and stored in the form of B-spline
curve equations.
Control
Once the geometric modulus and the transfer modulus of
Risering each segment have been determined, the feeding units of the
casting should be established. The feeding unit is defined as
Fig. 6 The family of applied risering methods the group of neighboring heavy segments isolated by the
"thin" segments having geometric moduli smaller than the ad-
risering also has been implemented, this method is not de- jacent M,'&. The system investigates the geometric moduli and
scribed in this paper and can be found in the references transfer moduli in adjoining order and determines the ex-
[10, 11]. istence of any segment for which the geometric modulus is less
than the transfer moduli of the adjacent segments. If any, it
Modulus Analysis. The primary objective of the modulus groups the segments isolated by such segments into one
analysis is to determine the feeding units for a casting which is feeding unit. After the feeding units are identified, the system
the basis for determining the number and the locations of the also determines a specific segment to which the riser is at-
risers, and to calculate the proper size of each riser and riser tached for each feeding unit. By Karsay [11], the riser may be
neck. First, a casting with a complicated shape needs to be attached to any segment in one feeding unit, but our system
separated into simple shapes called segments. recommends the user to attach the riser to the significant seg-
In the system, the casting is separated into several segments ment, which is defined as the one with the largest geometric
interactively as described below. In the current implementa- modulus in each feeding unit.
tion, only planar and cylindrical surfaces are allowed as sec-
tioning surfaces. If a planar sectioning surface is needed, the Dimensions of Riser and Riser Neck. To calculate the
user identifies an existing planar face of a pattern, through dimensions of the riser and riser neck, the moduli of the riser
which the sectioning is to be performed, by picking any two and riser neck should be determined. In principle, the riser and
edges on the face. If a cylindrical sectioning is needed, the user riser neck must be permeable to liquid iron as long as the liq-
picks a circular edge of the pattern which will lie on the sec- uid transfer from a significant segment is necessary.
tioning surface. If more segments are required, the operations Therefore, the riser modulus, Mr, and the riser neck modulus,
above are performed repeatedly. M„, must equal the transfer modulus, M,, of the significant
Once the pattern is split into segments, the next step is to segment of each feeding unit.
calculate the geometric modulus and the transfer modulus of --M,
each segment. First, the system calculates the volume and the
total surface area of each segment. Also, the surface area of M„ -M,
the new surfaces which have been exposed by the sectioning is Due to the superheat around the riser neck, however, the riser
calculated. This surface area is subtracted from the total sur- neck may have a smaller modulus, i.e.,
face area to yield the effective cooling area. Now, the system
calculates the geometric modulus of each segment by dividing M„=0.67M,
the volume by the effective cooling area. The shapes of standard risers are shown in Fig. 8 [5,10,12].
Next, the system will calculate the transfer modulus of each As illustrated in Fig. 8, riser dimensions can be calculated easi-
segment. By Corlett and Anderson [12], the transfer modulus, ly from the riser modulus. The cross-sectional shapes of the
denoted by M„ is defined as follows: "the modulus required riser neck supported in this system are shown in Fig. 9 [11],
of a segment through which liquid transfer will take place i.e.; round, square, and rectangular. Similarly, the dimensions
from an adjacent larger segment for as long a time as of the riser neck are calculated from the neck modulus (M„) as
necessary. The necessary time is that required for the heavier illustrated in Fig. 9.
segment to dissipate its expansion pressure by transferring ex- The procedure for calculating the dimensions of the riser
panding liquid through an adjacent smaller segment." and the riser neck is as follows: First, all the segments belong-
Accordingly, it is rather obvious that the necessary liquid ing to each feeding unit are displayed on the graphics terminal
transfer through the "thin" segment having a geometrjc and the segment to which a riser should be attached is selected
modulus smaller than the transfer moduli of two adjacent by the user as recommended by the system. Next, the
segments is not possible. A proposed relationship between the allowable riser types shown in Fig. 8 are displayed to be
geometric modulus and the transfer modulus, M,, as influ- selected by the user with the number required. More than one

62 / Vol. 113, FEBRUARY 1991 Transactions of the ASME

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 Side Risers

Poor metallurgical quality

I
"i r
2 3 4 5 6 7
,X> I -/ Modulus
Fig. 10 Relationship of modulus to shrinkage
Topmost point of riser
Type I Type 2
(contact in cope) (contact in drag.)
Effective teed metal ( shaded volume)
Topmost point of casting

Riser Riser Diameter Riser Volume

Type (D)
1 D= 5.68 Mr V = I.06D 3
2 D = 4.9I Mr V = I.I6 D 3
3 D = 4.53Mr V = l.04 D 3
Fig. 11 Effective feed metal of a side riser

EFM= [(0.007*3) + (0.095x2D) + (OA36xD2)]

Type 3 In the case of a top riser, as well as a side riser, only the
(Top Riser) portion above the topmost point of the casting is effective.
Fig. 8 Courtesy of standard riser shapes The procedure for adjusting the dimensions of a riser is as
follows.
• First, the system calculates the difference between the
riser height and the cope height of the casting and
calculates the effective feed metal volume from this dif-
b=a
ference and the riser diameter. If two or more risers
should be constructed, the total effective feed metal
b=2ak volume by the risers is computed.

H
Next, it calculates the volume of feed metal required by
to
the casting from the percent feed metal as in Fig. 10
and the casting volume.
• Finally, the system compares the effective feed metal
volume with the volume of feed metal required.
a=Mnx3 a = Mn x 4 If the effective feed metal volume is less than the volume of
Fig. 9 Shapes of riser neck the feed metal required, the system iteratively increases each
riser height.
Once the feed metal requirement is satisfied, the system up-
riser for each feeding unit is allowed in our system. Also, the dates the heights of the lower risers so that all risers have their
shapes of the riser neck shown in Fig. 9 are displayed and top in the same horizontal plane. Otherwise, in the initial stage
selected. Finally, from the moduli of the riser and the riser of the liquid contraction the lower positioned risers will be fed
neck for the feeding unit, the dimensions of the riser and its from the higher risers, and consequently this makes the lower
neck are calculated. Even for the case of multiple risers for risers useless and will create a very large shrinkage void in the
each feeding unit, the dimensions of the risers and their riser topmost riser.
necks are determined as for a single riser.
Modeling of Riser and Riser Neck. By using the Boolean
Adjustment of Riser Size. Since the riser has been con- operations between the primitives, the system models risers
structed based on the modulus, the riser dimensions must be and riser necks so that they have the dimensions as previously
adjusted so that the effective riser volume exceeds the volume determined. To model the risers represented in Fig. 8, the
of initial liquid feed metal required by the casting. primitives used for type 1 are cone and torus, for type 2 are
Figure 10 [5,10,12] illustrates the percent feed metal re- cone and sphere, and for type 3 are cylinder and sphere. Also,
quired by the casting based on the metallurgical quality and the system uses a box or cylinder to model the riser neck. For a
the average modulus of the casting. For programming pur- side riser, the system models the riser neck along the + X axis
poses, the curves were fitted and stored in the form of B-spline from the datum point, denoted by R l , and then models the
curve equations. riser at the position apart from the datum point by the neck
A schematic diagram to indicate the effective portion for a length along the + X axis, denoted by position A in Fig. 12.
side riser is shown in Fig. 11. The volume of effective feed The datum point is the origin of the reference coordinate
metal (EFM) for side risers is expressed[5] as system used by the embedded geometric modeling system. The

Journal of Engineering for Industry FEBRUARY 1991, Vol. 113/63

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 datum point will be used later in the layout of risers. In the in Fig. 12) and it is rotated about +Z axis by an angle
meanwhile, a top riser is modeled so that the center point of 0 (position C in Fig. 12). For the circular edge type, a
the neck base is located at the datum point and the riser is similar procedure can be applied. The type of the edge
along + Z direction. After modeling the riser and the riser to which the riser is attached is limited to straight and
neck separately, the system unites them. circular in the current implementation.
Layout of risers. Once the risers are constructed in three- The procedure for the layout of a top riser is described as
dimensional models, they are moved and rotated interactively follows.
to the desirable locations and orientations with respect to the • The position of the riser is located from the top view of
pattern. the segment to which the riser is attached, i.e., X and Y
The procedure for the layout' of a side riser is described as coordinates of its position are interactively provided.
follows with reference to Fig. 12. • Then the position of the riser is located again from the
The system asks the user to pick an edge (El) on the front view of the segment, i.e., the Z coordinate is also
parting plane from the top view of the segment and to interactively provided.
locate the riser position (PI) on the edge. The vector from the datum point to the position
• If the edge type is straight, the outer normal vector specified above is computed, and then the riser is
(Nl) of the edge, the vector (VI) from Rl to PI, and translated in the direction of the vector by its
the angle (6) of the outer normal vector (Nl) from the magnitude.
+ X axis are computed. Then, the riser is translated in Once the risers are transformed to the positions as specified
the direction of VI by the magntidue of VI (position B interactively by the user, the system unites the pattern and the
risers.

Case Study
edge E1
By using the system developed in this work, the pattern and
the risers for a gear blank casting have been designed. Figure
13(a) shows an initial menu and a final product constructed by
the system's modeling capability. The pattern with drafts and
adjusted dimensions based on the shrinkage allowance after
the elimination of the holes is shown in Fig. 13(6). Once the
final pattern is created, the pattern is split into three segments
and the feeding units from the analysis of the geometric and
transfer moduli of each segment are identified as shown in
Fig. 13(c). Then the segments to which the risers are attached
are identified, and the proper risers are designed considering
the feed metal requirement. After the layout of the risers, the
risers and the final pattern are united as illustrated in Fig.
Fig. 12 Layout ot a side riser
13(d).

SELECT MENU

MODELING

PATTERN DESIGN

RISER DESIGN

GATING DESIGN

MOULD ASSEMBLY

EXIT

ENDPICKI CANCEL J

QADAL
PLOT SCREEN SCREEN TOGGLE

Fig. 13(a) Case study

64https://manufacturingscience.asmedigitalcollection.asme.org
Downloaded From: / Vol. 113, FEBRUARY 1991 Transactions
on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use of the ASME
 PATTERN DESIGN

HOLE ELIMINATE

SHRINKHGE

DRAFT

BLENDING

RETURN

BOOT Litrr

ENDPICK[CRNCCUI

CADAL
PLOT SCREEN SCREEN TOGGLE DISPLAY

Fig. 13(b) Case study

Ge J act. command PRESSURE CONTROL

PATTERN SECTION

zoom
ccroI I

MODULE BNRLYEIG

RISER LOCUTION

RISER CREATION

RETURN

•DOT LISt

ENDPlCKtCPNCELi
TWHWtK
SCO N D . HODULUS
fiopuuic
1 a.aa t.u
2 a. 1.0G

S 1.75 1.

0.5
CUMULATIVE VOLUME GHMEIM/10BI

PLDT SCREEN SCREEN TOGGLE

Fig. 13(c) Case study

Conclusion with their recommended locations through modulus analysis

and interactive layout. In order to generate a complete three-
An interactive computer program to design a pattern and dimensional mold, a program to generate the appropriate
the risers has been developed. It includes an automatic hole gating systems needs to be linked to this system.
elimination, an automatic scaling for shrinkage allowance, With the three-dimensional model of the mold, this system
and an automatic draft addition for the pattern design. Also, would allow an NC tool path generation for integration with
it includes an automatic generation of risers and riser necks computer aided manufacturing and a solidification simulation

Journal of Engineering for Industry FEBRUARY 1991, Vol. 113/65

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use
 SELECT MENU

MODEL INS

PATTERN DESIGN

RISER DESIGN

GATING DESIGN

MOULD ASSEMBLY

EXIT

Boo-r L i s t

C N P P I CK I C J 3 N C E L I

CABAL
PLOT SCREEN SCREEN TOGGLE

Fig. 13(d) Case study

for the integration with computer aided engineering. Further- 5 Gerhardt, Jr., P. C , "Computer Applications in Gating and Risering
more, this system would become a very powerful tool for prac- System Design for Ductile Iron Castings," AFS Transactions, 1983, Des
Plaines, IL.
tical use if it could be linked with a knowledge-based expert 6 Dynarad Corporation, Novacast Demonstration Manual, available
system to collect the proper knowledge of human experts. through 14733 Catalina Street, San Leandro, CA, 1984.
7 Smith, C. T., and Lee, K., "Computer Aided Pattern Design for Casting
References Processes," AFS Transactions, 1986, Des Plaines, IL.
8 Requicha, A. A. G., "Representations for Rigid Solids: Theory, Methods,
1 Graham, H., "Computer Graphics for Casting Design," AFS Transac- and Systems," ACM Computing Surveys, Vol. 12, No. 4, Dec. 1980.
tions, 1981, Des Plaines, IL. 9 Cary, Chris, ROMULUS User's Reference Manual, Shape Data Ltd.,
2 Welbourn, D. B., "The Foundry and CAD/CAM: What is Required?" Apr. 1983.
AFS Transactions, 1983, Des Plaines, IL. 10 Karsay, S. I., Ductile Iron III Gating and Risering, QIT-Fer et Titane, Inc.
3 Berry, J. T., and Pehlke, R. D., "Computer-Aided Design For Castings," 11 Karsay, S. I., Encyclopaedia of Design Logic, QIT-Fer et Titane, Inc.
AFS Transactions, 1984, Des Plaines, IL. 12 Corlett, G. A., and Anderson, J. V., "Experiences with an Applied Riser-
4 Heine, R. W., and Uicker, J. J., "Risering by Computer Assisted ing Technique for the Production of Ductile Iron Castings," AFS Transactions,
Geometric Modeling," AFS Transactions, 1983, Des Plaines, IL. 1983, Des Plaines, IL.

66 / Vol. 113, FEBRUARY 1991 Transactions of the ASME

Downloaded From: https://manufacturingscience.asmedigitalcollection.asme.org on 06/30/2019 Terms of Use: http://www.asme.org/about-asme/terms-of-use