FIFTH INTERIM TOPICAL REPORT

on

A STUDY OFrTHE MECHANICS OF

CLOSED-DIE FORGING

1CALCULATION OF GEOMETRICAL PARAMETERS IN

DESIGNING THE FORGING PROCESS FOR
AXISYMMETRIC SHAPES

to

ARMY MATERIAL AND MECHANICS RESEARCH CENTER

October 15, 1969

oo&j~ 0 119/131r
4b by

T D. E. Nichols, H. J. Henning,
and A. M . Sabroff •XO1111O.

Contract DAAG46 -68-C-0 111,, . .

BATTELLE MEMORIAL INSTITUTE

Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201

1D>&2
01
 FIFTH INTERIM TOPICAL REPORT

on

A STUDY OF THE MECHANICS OF

CLOSED-DIE FORGING

CALCULATION OF GEOMETRICAL PARAMETERS IN

DESIGNING THE FORGING PROCESS FOR
AXISYMMETRIC SHAPES

to

ARMY MATERIAL AND MECHANICS RESEARCH CENTER

October 15, 1969

by

T. Altan, D. E. Nichols, H. J. Henning,

and A. M. Sabroff

Contract DAAG46-68-C-0111

BATTELLE MEMORIAL INSTITUTE

Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
 FOREWORD

This topical report, "Calculation of Geometrical Parameters in Designing the

Forging Process for Axisymmetric Shapes", covers the analytical work performed under
Contract No. DAAG46-68-C-0111 by Battelle Memorial Institute, Columbus, Ohio, from
April 30, 1969, to July 15, 1969.

This work was administered under the technical direction of Mr. Dennis Green of
the Army Materials and Mechanics Research Center, Watertown, Massachusetts 02172.

This program was carried out under the supervision of Mr. A. M. Sabroff, Chief
of the Metalworking Division, and Mr. H. J. Henning, Associate Chief of the Metal-
working Division. Dr. T. Altan, Senior Scientist, is the principal investigator.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 FOREWORD

This topical report, "Calculation of Geometrical Parameters in Designing the

Forging Process for Axisymmetric Shapes", covers the analytical work performed under
Contract No. DAAG46-68-C-0111 by Battelle Memorial Institute, Columbus, Ohio, from
April 30, 1969, to July 15, 1969.

This work was administered under the technical direction of Mr. Dennis Green of
the Army Materials and Mechanics Research Center, Watertown, Massachusetts 02172.

This program was carried out under the supervision of Mr. A. M. Sabroff, Chief
of the Metalworking Division, and Mr. H. J. Henning, Associate Chief of the Metal-
working Division. Dr. T. Altan, Senior Scientist, is the principal investigator.

BATTELLE MEMORIAL INSTITUTE -- COLUMBUS LABORATORIES

 ABSTRACT

A computer program in FORTRAN language is developed for calculating geometri-

cal parameters for axisymmetric forgings. Using the dimensions and the angles given
in the engineering drawing of the forging, the program calculates the forging volume and
weight. The shape-difficulty factor may prove useful in process design, for cost esti-
mating, and for assisting engineers to make decisions about feasibility and difficulty in
manufacturing. Empirical equations suggested by German and Soviet workers are used
for calculating the flash dimensions and the flash weight. The computer program is
applied to various forgings, and theoretical results are found to agree with actual experi-
mental data. The developed computer program is now available to interested forging
companies. It could be used for a systematic analysis of existing forging data, for
making cost estimates after incorporating in-house cost factors, or for designing axi-
symmetric forgings with minimum additional effort.

BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

 TABLE OF CONTENTS

Page

INTRODUCTION. ....... . . .......................... 1

BASIC VOLUME COMPONENTS OF AN AXISYMMETRIC FORGING . . . . 2

Cylinder ......................... ... . 2

Truncated Cone..................... 5
Truncated Cone With Fillet................. 6
Truncated Cone With Corner............... .. 9

CALCULATION OF GEOMETRICAL QUANTITIES FOR THE ENTIRE FORGING. 11

Perimeter of the Axial Cross-Sectional Surface. ...... . . . ..

Surface Area of the Axial Cross-Sectional Surface .............. 13
Volume of the Forging.............. .. . .... 13
Radius of Center of Gravity of the Half Cross Section. ... . . . .. 14

COMPUTER PROGRAM TO CALCULATE DESIGN VARIABLES FOR

AXISYMMETRIC FORGINGS ................ ..... ........ 14

Program Input . . .................................... 15

Program Calculations and Output.. . ...................... 16

APPLICATION TO PRACTICAL FORGINGS ........................ 18

DISCUSSION.. ........ . .......................... 24

REFERENCES . .......................................... 24

APPENDIX A

THE SHAPE-DIFFICULTY FACTOR FOR AXISYMMETRIC FORGINGS . A-1

APPENDIX B

EMPIRICAL FORMULAS FOR PREDICTING FLASH DIMENSIONS AND

FLASH WEIGHT...................... B-1

APPENDIX C

SYMBOLS USED IN THE FORTRAN PROGRAM ...................... C-1

APPENDIX D

FORTRAN COMPUTER PROGRAM FOR CALCULATING DESIGN PARAMETERS

IN AXISYMMETRIC FORGINGS ................................ D-1

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 CALCULATION OF GEOMETRICAL PARAMETERS IN
DESIGNING THE FORGING PROCESS FOR
AXISYMMETRIC SHAPES

by

T. Altan, D. E. Nichols, H. J. Henning,

and A. M. Sabroff

INTRODUCTION

The design of a forging process involves the calculation or the prediction of

(1) Forging volume and weight

(2) Number of blocker shapes to be used, if any
(3) Flash geometry, flash width, and flash thickness
(4) Flash weight and scale losses
(5) Total forging weight, including scale and flash losses.

The second interim topical report, "Methods for Estimating Flash-Gap Dimensions
,and Flash Weight in Closed-Die Steel Forging", described a shape-difficulty factor for
axisymmetric forgings first introduced by Soviet workers. (1)* At this time, this factor
includes only the geometrical shape of the forging but excludes the effect of forging ma-
terial, die wear, and forging tolerances. The shape-difficulty factor appears to be use-
ful in predicting the difficulty involved in forging a part and in calculating the, flash di-
mensions and flash weight by using statistically derived empirical formulas. These
formulas and the derivation of the shape-difficulty factor were given in the second topical
report, and they are summarized in Appendix A. In the present work, the derivation
and practical use of a FORTRAN computer program is described. This program calcu-
lates, from the dimensions given in the engineering drawing of a particular forging, the
following variables:

(1) Volume of actual forging, excluding flash and scale losses

(2) Shape-difficulty factor
(3) Flash width and flash thickness
(4) Weight of the flash and scale losses
(5) Total forging weight, including scale and flash losses
(6) Plane area of forging, including flash.

The computer program is written for axisymmetric steel forgings forged from a cylin-
drical billet. However, it can be extended easily to analyze intermediate blocker forg-
ings necessary for forging complicated axisymmetric shapes. Practical shop experience
will be necessary to develop the shape-difficulty factor into a characteristic quantity
that indicates, for a given material, the number of necessary blocking operations and
the fabrication cost associated with a specific forging shape.

*References to work described in three papers by Russian authors have been discussed in the second interim report.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 2

In order to calculate the volume of a forging using the dimensions in the engineer-
ing drawing of the forging, the forging is divided into various small components for which
the volumes can be calculated by means of well-established formulas. This procedure is
carried out in practice in most companies by using formulas, charts, or nomograms.
Complete computerization of this procedure is proposed, at least for the axisymmetric
forgings. With the time-sharing computer terminals available today, even a small com-
pany that does not have a computer can acquire a computer terminal at minimal expense
to carry out the calculations described in this report.

In order to use this computer program, the user is not required to know the details
of the mathematics or of the programming involved. He must, however, know how to
prepare the input data for the program. However, the details of the geometrical deriva-
tions and of the computer program are given in this report so that those interested in
research and development can use this information for future expansion of the present
approach.

BASIC VOLUME COMPONENTS OF AN

AXISYMMETRIC FORGING

The practical method for determining the total volume of a forging consists in
dividing the forging into simple basic geometrical shapes and then calculating and adding
the volumes of all components. Figure I illustrates the volume components of a simple
axisymmetric forging. Using well-known formulas, the volumes are easily calculated.
Tables are available for determining the volumes of the small-volume components de-
scribed by the arc and the edges of the corners or fillets of the forging. This procedure
can be computerized along with the determination of the shape-difficulty factor. The
shape-difficulty factor described in the second interim topical report and in Appendix A
is evaluated by determining the following variables for an axisymmetric forging:(I)

(1) The perimeter of the axial cross-sectional surface, P

(2) The surface area of the axial cross-sectional surface, S

(3) The radial distance from the axis to the center of gravity of half of the
cross section, RG.

For determining the forging and the flash weights, it is also necessary to calculate the
volume of the forging.

For this purpose, the forging is divided into basic volume components that are
analyzed below. Throughout this analysis, the computerization of all calculations is
emphasized.

Cylinder

The half cross section of a cylinder is shown in Figure 2a. The radii Ri and Ri_1
describing the cylinder are subscripted in order to facilitate the adaptation of derived
formulas for computer programming. These dimensions are obtained from the engineer-
ing drawings of the forging.
BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES
 3

_ R4

i• R5

/7r4a
r6 0H

<
R6

RII
FIGURE 1. DIVISION OF A FORGING INTO SIMPLE GEOMETRICAL
VOLUME COMPONENTS

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 4

P1 R1

Center of gravity R

Pi-I RI-,I H

a. Cylinder Half Cross Section

a1 (positive)

PP Ri C

Pi- I
A Ri-I B

b. Truncated Cone Above the Parting Line

FIGURE Z. SYMBOLS USED IN DESCRIBING THE GEOMETRY OF

THE HALF CROSS SECTION OF A CYLINDER AND OF
A TRUNCATED CONE

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 5

Using the symbols of Figure 2a, the following variables are easily calculated:

The segment of the perimeter, i.e., the distance between the points Pi and Pi-l1:

Pi Ni-i = H (1)

The surface area of the half cross section:

Si = Ri H (2)

The volume of the volume component:

Vi = ?TRi2 H (3)

The radial distance from the center of gravity of the half cross section:

RGi = Ri/2 (4)

For a half cross section of a cylinder as shown in Figure 2a, Equations (1), (2), (3), and
(4) give the segment of the perimeter, the surface area, the volume, and the radius of
center of gravity, respectively. These four quantities are necessary to calculate the
shape-difficulty factor and the volume of the entire forging.

Truncated Cone

The half cross section of a truncated cone is shown in Figure 2b. The angle of
taper, ci, measured with respect to the centerline, is considered positive if the taper
is above the parting line of the forging.

As in case of the cylinder, the geometrical variables for the half cross section of
the truncated cone seen in Figure 2b, are calculated as follows:

The segment of the perimeter:

Pi Pi-l Hj/cos aji (Pi-Ri-l)/sin aix (5)

The surface area:

Si = (Ri + Ri_1) Hi/2 (6)

The volume:
2
V 7T(Ri2 + R + RR /3.(7)
Si ii) H/31(-)

The radial distance from the center of gravity of the half cross section seen in Fig-
ure 2b is obtained by considering the triangle PiPi_lA and the rectangle PiABC as the
components of the half cross section. The following relation holds:

RGi x total surface area = surface of triangle x RGT of the triangle + surface
of rectangle x RGR of rectangle,
BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES
 6

whe re

RGi = center of gravity radius for the half cross section

RGT = center of gravity radius for the triangle
RGR = center of gravity radius for the rectangle.

Thus,
i (Ri_1 - Ri)] (Ri_ 1 - Ri) H RiRi H
RG Si = Ri + 3 J 2 + 2 (8)

Using the value of the surface Si from Equation (6), Equation (8) is solved to give:

(Ri2 + RiRi-i + Ri-12

RGi - 3(Ri + Ri-l)

For a half cross section of a truncated cone as seen in Figure 2b, Equations (5), (6),
(7), and (9) give the segment of the perimeter, the surface area, the volume, and the
radius of center of gravity, respectively.

Truncated Cone With Fillet

The half cross section of a volume component, described by two radii, Ri and Ri-1,
and by fillet radius ri, is illustrated in Figure 3a for positive draft angles (draft angles
are positive for the part of the forging above the parting line) and in Figure 3b for the
negative draft angles. Following Figure 3a, the values which are given in the engineer-
ing drawing are:

(1) Edge radius, R

(2) Fillet radius, ri
(3) Angles cmi and 0i_2.

In some cases, the value of the edge radius, R, is not explicitly given in the engi-
neering drawing but has to be calculated from other given dimensions. These secondary
calculations are incorporated in the final computer program.

From Figure 3a, the following quantities, which are also valid for the conditions
of Figure 3b, are calculated by using the geometrical relations:

Angle = i2 - ori f (10)

BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

 7

a i ( positive)

a1i2 (positive) ~RiIle Rs

rii
______
a. Positive Draft Angles ai and ai.2 (Above the Parting Line)
. Ri-I

(negative) p

2 ( negative) I
b. Negative Draft Angles ai and ai_2 (Below the Parting Line)

FIGURE 3. SYMBOLS USED IN DESCRIBING THE GEOMETRY

OF A
A TRUNCATED CONE WITH CONCAVE FILLET

BATTELLE MEMORIAL INSTITUTE COLUMBUS

-- LABORATORIES
 8

and, using

£= ri tan (/3/2) (11)

AKi = £ sin mi (12)

AHi = 2 cos ai (13)

ARi_1 = sin 0i-2 (14)

AzHi_1 = cos ai-2 (15)

Using Equation (12), the radius for the point Pi is

Ri = R - ARi . (16)

Using Equation (14), the radius for the point Pi-I is

Ri1--I =R + Ai_1 (17)

The height, H, of the volume component considered is

H = AHi + AHi_1 , (18)

where Hi and Hi_1 are given by Equations (13) and (15), respectively.

The surface area of the circular segment defined by the straight line PiPi_ I and by
the arc PiPi-l is
r2
A =--(13 - sinl3) ,(19)

where the angle, /3, is given by Equation (10).

The distance between the center of the arc and the center of gravity of the circular
segment is
3
2 r.3~ sin /3/2
Xs=
3 A (20)

The distance, Ax, as seen in Figure 3a, is

Ax = x( sin (21)

whe re

angle y = /3/2 + 7 - _2 for ai_ 2 > 0 (Figure 2a) (22a)

-/3/2-
P= ai_2 for ci_2 < 0 (Figure 2b) (22b)

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 9

The radius for the center of gravity of the circular segment defined by the arc
PiPiI and by the line PiPi_l is given by

Rs = R + Ax (23)
The quantities given by Equations (10) through (23) are calculated by using only the values
of the edge radii R, fillet radii ri and of the angles ci and ai-2, all of which are given in
engineering drawings.

Now, the important geometrical variables for the volume component under consid-
eration can be calculated:

The segment of perimeter, arc PiPi = 3 ri (24)

(Ri + Ri-1) H
The surface area, Si = 2 - A (25)

The volume, Vi = 'T(Ri 2 + Ri_ 2

+ RiRiI)H/3 - 2'RsA (26)

The radius of the center of gravity, obtained from RGi = (RGcSC - A Rs)/Si . (27)

In Equation (27), the radius RGC is the radius of the center of gravity of the half-
truncated cone described by Pi and Pi-l" RGC is calculated by applying Equation (9).
SC is the surface area of the same half cross section of the truncated cone and is calcu-
lated with Equation (6). All other quantities used in Equations (24) through (27) are cal-
culated by using one or more of Equations (10) through (23).

Truncated Cone With Corner

The half cross section of a volume component with a corner is shown in Figure 4 a
for positive draft angles and in Figure 4b for negative draft angles. Again, the edge
radius, R, the corner radius, ri, and the angles ai and ai_2 are given in the engineering
drawing of the forging.

Equations (10) through (21) are still valid for calculating various geometrical quan-
tities shown in Figure 4a.

The angle -y is given by

- - - /3/2 - ai_2 for ai_ 2 > 0 (Figure 3a) ( 2 8a)

' = f- I 0i-2 + /3/2 for ai_2 < 0 (Figure 3b) (28b)

The radius for the center of gravity of the circular segment is given by

Rs = R - Ax . (29)

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 10

AR.C1

RiIR
a1 2 i -.. .. '
R Pi-I
-2(ngtv )

i~

a. Positive Draft Angles

'~~~~~ ~~~ S R S

b. Negative Draft Angles

FIGURE 4. SYMBOLS USED IN DESCRIBING THE GEOMETRY OF A

TRUNCATED GONE WITH CONVEX FILLET

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 11

The following quantities can now be calculated:

The segment of perimeter, arc PiPi-l = /3 ri (30)

(Ri + Ri-l.) H
The surface area, Si = 2 +A (31)

2 (32)
The volume Vi = 7T(Riz + Ri- 1 + RiRiI)H/3 + 27f Rs A

The radius of the center gravity for the entire volume component is

RGi = (RGC SC + A Rs)/Si , (33)

where RGC, SC, RS, Si and A are obtained from Equations (9), (6), (29), (31), and (19),
respectively.

Thus, Equations (30), (31), (32), and (33) give the necessary geometrical quantities
for the half cross section of a truncated cone with a corner radius, as seen in Figure 4a
and b.

CALCULATION OF GEOMETRICAL QUANTITIES FOR

THE ENTIRE FORGING

The half cross section of an axisymmetric forging is illustrated in Figure 5. The

draft angles, ai, the fillet radii, ri, the heights between corners, HCi, and some of the
corner radii, Ri, i-1, are obtained directly from the engineering drawing. For instance,
for the example given in Figure 5, the corner radii R 2 3, R 6 , 7, and R 8 , 9 are obtained
from the drawing. Using the geometrical relationships, the corner radii R 4 5 and
RI0, IIare easily calculated by the computer.

The geometrical quantities necessary for calculating the shape-difficulty factor,

the flash dimensions, the forging weight, and the flash weight are calculated by adding or
subtracting the quantities determined for each volume component. In Figure 5, for in-
stance, the calculated quantities for volume components within the zone defined by the
radii R 8 9 and RI 0 11 must be subtracted from the total. In order to use the "sub-
scripted variables" of the FORTRAN computer language, the dimensions and the angles
are subscripted as shown in Figure 5.

Perimeter of the Axial Cross-Sectional Surface

The total perimeter of the axial cross-sectional surface, seen in Figure 5, is ob-
tained by adding all the segments of the perimeters calculated for various cross-
sectional components. Thus, considering the axial symmetry of the forgings, the total
perimeter P is
n

P =2 PiPil 1 (34)
i=l

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 12

"R6 17

R 8 ,9

7 8 _

9
6 a9

a 5 o
5
a5
12

I 5/

R3 HC4 5

2R2

R 2 ,3

R 4 15

FIGURE 5. DESCRIPTION OF THE FORGING GEOMETRY BY DEFINING

THE CONTOUR OF THE HALF CROSS SECTION WITH VARIOUS
RADII, Ri, FILLET RADII, ri, AND DRAFT ANGLES, ai

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 13

where

PiPi-l = segment of the perimeter obtained from Equations (1), (5), (24),
or (30) corresponding to the specific volume component
n = number of volume components in the forging.

Equation (34) gives the total perimeter of the axial cross-sectional surface for an axi-
symmetric forging.

Surface Area of the Axial Cross-Sectional Surface

The total surface area is given by

n
S=2 ZkSi (35)
i=l

where

Si = surface area of a volume component, from Equations (2), (6), (25) or (31)
n = number of volume components in the forging

k = + 1 or - 1.

The factor k has the value +1 if the surface area of that volume component must be added
to the total. If that surface area must be subtracted from the total, as, for instance, for
the volume components within the zone defined by R 8 , 9 and RIO, 11 of Figure 5, then
k= -1.

Equation (35) gives the total surface area of the axial cross-sectional surface for an
axisymmetric forging.

Volume of the Forging

The total volume of the forging is calculated by

n
V = k Vi , (36)
i=1

where

V. = volume of a volume component, from Equations (3), (7), (26), or (32).

For n and k, the same considerations are valid as in Equation (35), which was used for
calculating the area of the cross-sectional surface.

BATTELLE MEMORIAL INSTITUTE-COLUMBUS LABORATORIES

 14

Equation (36) gives the total volume of an axisymmetric forging calculated from the
volume components.

Radius of Center of Gravity of the

Half Cross Section

The radial distance from the axis to the center of gravity of half of the cross sec-
tion is calculated by

RG= k RGiS)/S , (37)

i--1

where

RGi = radius of the center of gravity for a volume component, from Equations
(4), (9), (27), or (37)
Si = surface area of a volume component, from Equations (2), (6), (25),
or (31)
n = number of volume components in the forging
k = +1 if the volume of that component is added to the total
= -1 if the volume of that component is subtracted from the total.

After the radii of centers of gravity are determined for the half cross sections of
all volume components, the radius of center of gravity for the half cross section of the
entire forging is obtained with Equation (37).

COMPUTER PROGRAM TO CALCULATE DESIGN

VARIABLES FOR AXISYMMETRIC FORGINGS

The geometrical relationships derived in this report are computerized in order to

calculate the necessary variables for obtaining the volume, the flash thickness, and the
flash width of the forging. The computer program is given in Appendix D. The
FORTRAN symbols used to designate various geometrical variables are listed in Appen-
dix C. The practical use of this computer program is described below by using the half
cross section of the example forging obtained from literature and shown in Figure 6. (2)

In order to use this computer program, the user is not required to know the details
of the mathematics or of the programming involved. He must, however, know how to
prepare the input data for the program. However, the details of the geometrical deriva-
tions and of the computer program are given in this report so that those interested in
research and development can use this information for future expansion of the present
approach.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 15

Program Input

The input data for the computer program are obtained from the engineering drawing
of the forging, in which all the draft angles and the fillet radii, but not necessarily all
the corner radii (or diameters), are given. The unspecified dimensions are calculated
by the computer from the input data. The input data, punched on IBM cards, are given
in Table 1 for the forging shown in Figure 6.

Following Table 1, the first line (or the first IBM card) contains (a) the number of
corners or fillets below the parting line starting with the point at center line and (b) the
number of corners above the parting line starting with the corner at parting line as seen
in Figure 6.

TABLE 1. INPUT DATA FOR THE EXAMPLE FORGING OF FIGURE 6

(EACH LINE CORRESPONDS TO ONE IBM CARD)

Computer Card Column Number

10 20 30 60 50 60
7 9 Corner or Fillet
k Ri Hci ri 0i f No. in Figure 6

1 0.0000 0.0000 0.0000 -90.0000 0.0000 1

1 0.7870 0.0000 0.2360 -6.0000 0.0000 2
1 0.0000 0.3940 0. 1180 -90. 0000 1.0000 3
1 1.3790 0.0000 0.3940 -6.0000 0.0000 4
1 0.0000 0.5900 0.1180 -90.0000 1.0000 5
1 0.0000 0.0000 0.3940 -6.0000 -1.0000 6
1 2.7570 0.3940 0.0000 0.0000 0.0000 7
1 2. 7570 0.0000 0.0000 6.0000 0.0000 1
1 0.0000 0. 1970 0. 1575 90.0000 1.0000 2
1 2. 1650 0.0000 0. 1970 5. 7000 0.0000 3
1 0.0000 1.5750 0. 1970 90.0000 1.0000 4
-1 0.0000 0.0000 0. 1970 9.0000 -1.0000 5
-1 1. 1820 0.9850 0. 1575 90.0000 0.0000 6
-1 0.0000 0.0000 0. 1970 9.0000 -1.0000 7
-1 0.5910 0.7870 0. 1970 90.0000 0.0000 8
-1 0. 0000 0.0000 0.0000 90.0000 0.0000 9
0.2830 5.5140 5.5140 0.3935 0. 1967 3. 1450
P Do Dc ho hA Hs

Each line in the main body of Table 1 represents the data for a corner point. All
dimensions are given in inches. As seen in Table 1, the following quantities are
punched on one card and given in one line:

k = +1 if the volume component is to be added to the total volume; otherwise,

k = -1. The role of k is already discussed in Equations (35), (36), and
(37).

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 16

Ri = corner radius; Ri = 0 if that corner radius is not available in the engi-

neering drawing, as seen for the third corner, Table 1 and Figure 6.
Hci = axial distance between two consecutive corners.
ri = fillet radius.
(Xi = angle between the i'th and (i + 1)th corner, as discussed in deriving the
geometrical relations for basic volume components.
f = 0, if the corner radius Ri is obtained from the engineering drawing.
f = + 1, if the corner radius must be calculated from the radius of (i-1)th
corner, as for the third corner in Figure 6.
f = - 1, if the corner radius Ri must be calculated using the (i+l)th corner,
as for the sixth corner in Figure 6.

The last card of the input data contains the following information:

p = density of the workpiece material in lb/in. 3

Do = diameter of the initial round stock

Dc = maximum diameter of the forging, i. e. , diameter of the circumscribing

cylinder
ho = minimum distance between flat surfaces upon which stock was resting
when dies were closed
hA = distance between internal and external parting lines
Hs = final height of the forging.

The input data given in Table 1 are obtained from the drawing of the example forg-
ing shown in Figure 6 by following the conventions described above.

Program Calculations and Output

Using the input data, the program first calculates all the edge radii for the half
cross section. It then divides the forging into volume components and determines for the
cross section of every volume component (1) the segment of the perimeter, (2) the area
of the half cross section, (3) the volume, and (4) the center of gravity.

With these values already calculated, the perimeter, the surface area of the half
cross section, the center of gravity of the half cross section, and the volume are calcu-
lated for the entire forging. These calculations are carried out by using the formulas
developed earlier in this report, and the results are printed out.

At this point, all the variables necessary for calculating the shape-difficulty factor
are known. Using the formulas given in the second interim topical report( 1 ) and in Ap-
pendix A, the value of the shape-difficulty factor is calculated and printed out.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 17

2.757

'- 2.165

1.182 -

0 •0.591

0.O197 R g
0.985
5.70 --

1,575

0.157R 0.787

0.3940

--O.394R () 0.0.197R 0. 1 97RI 05

0.0359
0.394

-0.78 7

1.379-

FIGURE 6. AXIAL HALF CROSS SECTION OF AN AXISYMMETRIC

FORGING ILLUSTRATING THE INPUT DATA TO COM-
2
PUTER PROGRAM( )
All dimensions are given in inches.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 18

The flash thickness is calculated by using the empirical formula suggested by

Wolf. (1) For the computation of the flash ratio, the formula suggested by Wolf is used
if the forging weighs less than I pound. Otherwise, Teterin and Tornovskij's formula
is applied for both in calculating the flash ratio and the flash weight. All these formulas
were discussed in detail in the second interim topical report and are given in Appendix B.

The output of the computer program for the example forging shown in Figure 6 is
given below:

PERIMETER, SURFACE, VOLUME, R OF C. GRAVITY.

PERIMETER = 18. 08550 SURFACE = 8.47459.
VOLUME = 33.25583 RADIUS OF C. OF GRAVITY = 1.24910.
SHAPE DIFFICULTY FACTOR IS 2.02221.
FORGING WEIGHT WITHOUT FLASH IS 9.41140.
FLASH THICKNESS, FLASH WIDTH, FLASH RATIO ARE 0. 11403 0.46062 4.03952.
FORGING WEIGHT, FLASH WEIGHT, TOTAL WEIGHT 9.41140 1.04123 10.45263.
THE PROJECTED AREA INCLUDING FLASH IS 32. 52509.

The calculated quantities useful for the design of the forging shape are

(1) Shape -difficulty factor

(2) Forging weight with and without flash
(3) Flash thickness
(4) Flash width
(5) The projected area, including flash.

Thus, this computer program calculates in very short time some of the useful
variables for the overall design of the forging process.

APPLICATION TO PRACTICAL FORGINGS

The computer program developed in the present study is applied to various axi-
symmetric steel forgings. Dimensions for most forgings were obtained from the liter-
ature and some were supplied by the Steel Improvement and Forge Company.

Guha( 2 ) in his dissertation analyzed the forging shown in Figure 6. He studied the
effect of draft angles and fillet radii upon die fill, forging load, and forging energy in
forging 1043 and 1045 steels. He carried out experiments under a mechanical press and
a counterblow hammer. The actual design values used by Guha for the forging seen in
Figure 6 are compared below with the calculated values:

Calculated With the

Actual Computer Program

Shape-difficulty factor Not calculated 2. 022

Forging volume Not calculated 33. 26 in. 3
Forging weight (without flash) Not calculated 9. 41 lb
Forging weight (with flash) 12. 29 lb (calculated 10. 5 lb
from stock
dimensions)

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 19

Calculated With the

Actual Computer Program
Flash thickness in die 0. 12 in. 0. 114 in.
Flash width in die 0. 59 in. 0.46 in.
Flash ratio in die 5 4.04

The above comparison clearly shows that the flash dimensions determined by the com-
puter program are close to the actual values used in Guha's experiments. The flash and
scale losses used by Guha are about 30 percent of the actual forging weight, while the
flash losses predicted by the computer program are 12. 0 percent of the forging weight.

The forging analyzed in the present study on "Mechanics of Closed-Die Forging" is

illustrated in Figure 7. (1) This forging was designed before the present computer pro-
gram was developed. Selection of the flash dimensions of the forging are based solely on
experience, while the values for the total volume and weight were obtained by using the
same empirical formulas used in the computer program. Forging experiments on this
shape with lead at room temperature, with 6061 aluminum at 800 F, and with 1020 steel
at 2100 F are now completed. During these trials, the forging shown in Figure 7 was
completely filled for all three materials. The actual and the calculated design values
for this forging are given below:

Calculated With the

Actual Computer Program

Shape-difficulty factor -- 1.316

Total forging weight 3.42 lb 3.42 lb
Flash thickness in die 0. 1 in. 0. 085 in.
Flash width in die 0.5 in. 0.41 in.
Flash ratio in die 5 4.8

The above comparison shows that the theoretical values are within the useful range of the
empirically estimated values used in experiments.

Brill, in his study on modeling of the hot forging process, used the forging shape
seen in Figure 8. (3) He forged model materials such as wax, lead, plasticine and plas-
tics, and 1043 steel in the same die. The actual and the calculated significant parameters
are given below:

Calculated With the

Actual Computer Program

Shape -difficulty factor -- 1. 134

Forging weight without flash 0. 336 lb 0. 336 lb
Forging weight with flash 0. 36 lb (calculated from 0. 41 lb
stock dimensions)
Flash thickness in die 0. 079 in. 0. 058 in.
Flash width in die 0. 394 in. 0. 236 in.
Flash ratio in die 5 4.06

The comparison indicates that the flash dimensions predicted by the computer program
are close to the actual values. The flash weight predicted by the computer program is
2Z percent of the forging weight, while the actual flash weight is only 7. 2 percent of the
forging weight. A careful examination of the photographs of forgings given in Brill's

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 20

0.09 R

3 0
1.2
L
0.3R
0.09 R

0.35 _3:
0.5

0.35 0.09 R

2.0

FIGURE 7. AXISYMMETRIC FORGING USED IN THE PRESENT

PROGRAM FOR CLOSED-DIE FORGING STUDIES

All dimensions are given in inches.

0.197/
--- ....___ 0.2755
0.0787-
0.197

0.985

0.06 R

FIGURE 8. AXISYMMETRIC FORGING USED IN BRILL'S

3
EXPERIMENTS( )
BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES
 21

dissertation indicates, however, that the die cavity was not always completely filled.
The top of the spikes on the top and bottom of some forgings do not appear to have the
exact contour of the die cavity at the spike. (3) Consequently, the theoretically predicted
values for flash weight and flash dimensions seem to be more realistic design values than
the ones used by Brill. (3)

Tolkien made an extensive experimental study on various lubricants in closed-die

forging. (4) He evaluated the friction and the sticking forces of lubricants by using the
closed-die forging seen in Figure 9. The actual and the calculated design parameters
for this forging are given below:

Calculated With the

Actual Computer Program

Shape -difficulty factor -- 1.014

Forging weight without flash 0. 260 lb 0. 260 lb
Forging weight with flash 0. 336 lb 0. 317 lb
Flash thickness in die 0. 0552 in. 0. 0564 in.
Flash width in die 0. 138 in. 0.23 in.
Flash ratio in die 2.5 4. 1

The agreement between theoretical and actual values is good. The flash thick-
nesses are nearly identical, while the theoretical flash width is slightly larger. The
flash weight predicted by computer calculations is less than the actual weight used by
Tolkien (22 percent versus 29. 2 percent).

One of the participating companies in this program, Steel Improvement and Forge
Company, supplied the die and forging drawings for the axisymmetric forging seen in
Figure 10. Using the dimensions given in Figure 10, the computer program was used to
calculate the following parameters:

Shape -difficulty factor 1. 8

Forging weight without flash 5. 74 lb
Forging weight with flash 6. 33 lb
Flash thickness 0. 100 in.
Flash width 0.431 in.
Flash ratio 4.35

These values appear to be reasonable for forging the shape seen in Figure 10 from
low-alloy steel. In SIFCO's actual forging, the material was 347 stainless steel, with
one blocker die and one finishing die used. It is interesting to observe that SIFCO's forg-
ing (Figure 10) has a shape-difficulty factor of 1. 8, while Guha's forging (Figure 6) has a
shape-difficulty factor of 2. 02. Guha could forge his shape without using a blocker die
because the workpiece material was low-alloy steel. SIFCO's example illustrates the
well-known fact that, in addition to the geometrical shape-difficulty factor, factors de-
scribing the influence of forging material, die wear, and forging tolerances must also be
considered in designing a forging process. Those latter parameters are not included in
the computer program described in this report.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 22

0.788

0.1692
0.05521 ..

0.1692
1004
0.158 R 0. 158

0.158 R
I 295
0.. 06

FIGURE 9. AXISYMMETRIC FORGING USED BY TOLKIEN IN

STUDYING LUBRICATION EFFECTS IN CLOSED-
4
DIE FORGING( )

All dimensions are given in inches.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 23

"2.12
"A--1.685
""0.955"-"
0.38

A0.46

0.45
0.775

0.785 •03
S0.51 03

0.37
1.68

2.14

FIGURE 10. AXISYMMETRIC FORGING DIMENSIONS SUPPLIED BY

SIFCO FOR THE PRESENT PROGRAM

All draft angles = 70; all fillet radii = 0. 25 in. ; all corner
radii = 0. 12; other dimensions shown in inches.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 24

DISCUSSION

The determination of the number of preforming operations, the design of flash, and
the prediction of flash and scale losses are significant steps in designing a forging pro-
cess. The flash geometry has a marked influence upon die loads and stresses and upon
die filling. The prediction of flash weight, including scale losses, is a very important
factor in the overall cost of large-quantity-production forgings. The blocker shapes are
necessary in forging complex shapes of difficult-to-forge materials, and they add to
forging costs. It normally takes extensive experience to reasonably estimate these vari-
ables and to design the forging process, and they are not included in the current com-
puter program. It may be possible to handle such decisions later as experience and data
accumulate and are introduced into the program.

The difficulty in forging a part depends upon various factors, such as the geomet-
rical shape of the forging, the forging material, the forging tolerances, and the expected
die wear. The "shape-difficulty factor" has been suggested for axisymmetric shapes by
Soviet workers. (1) Using the shape-difficulty factor described in Appendix A and statis-
tically obtained empirical equations given in Appendix B, it appears to be possible to
predict the flash dimensions and the flash weight for axisymmetric low-alloy steel forg-
ings. This procedure is extended to calculate the volume of a forging, also, and it is
programmed for a computer in FORTRAN language. The program uses only information
given in the engineering drawing and calculates the total weight of the forging (forging
weight and flash weight). In order to use this computer program, the user is not re-
quired to know the details of the mathematics or the programming involved. He must
only know how to prepare the input data, as shown in Table 1.

The computer program is applied to various forgings, and results have been com-
pared with actual experimental data. The agreement between theoretical-empirical pre-
dictions and with design data developed experimentally (or suggested by extensive forging
experience) is very good. The developed computer program is now available to inter-
ested forging companies. It could be used for a systematic analysis of existing forging
data, for making cost estimates after incorporating in-house cost factors, or for design-
ing axisnymmetric forgings with a minimum amount of additional effort.

The information presented in this report is not intended to be a ready answer for
design problems of individual forgers. In order to establish a more exact analysis of
the forging process, the concept of shape-difficulty factor must be complemented by
experimental-empirical factors describing the effect of forging material, forging toler-
ances, and die wear. Every company could find and develop this information to suit its
own needs and operations. The concept of a shape-difficulty factor could be further ex-
tended to nonaxisyrmmetric forgings, if there is interest within the industry for this kind
of developmental work.

REFERENCES

(1) Altan, T., Henning, H. J., and Sabroff, A. M. , "Methods for Estimating Flash-Gap
Dimensions and Flash Weight in Closed-Die Steel Forging", Second Interim Topical
Report to AMMRC, Watertown, Massachusetts, January 30, 1969.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 25 and 26

(2) Guha, S. P. , "Einfluss der Abrundungen an den Werkzeugen beim Gesenkschmieden

auf Kraft-und Arbeitsbedarf sowie den Steigvorgang" (Influence of Fillet Radii on
Load and Energy Requirements and on Die Fill in Closed-Die Forgings). Disserta-
tion, Aachen, 1964.

(3) Brill, K. , "Modellwerkstoffe fdr die Massivumformung von Metallen" (Model Mate-
rials for Forging of Metals), Dissertation, Hannover, 1963.

(4) Tolkien, H. , "Schmierwirkungen in Schmiedegesenken" (Lubrication Effects in Forg-

ing Dies), Dissertation, Hannover, 1961.

BATTELLE
MCT3,ORIES

- STU•PTh,

BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

 APPENDIX A

THE SHAPE-DIFFICULTY FACTOR FOR

AXISYMMETRIC FORGINGS

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 A-1

APPENDIX A

THE SHAPE-DIFFICULTY FACTOR FOR

AXISYMME TRIC FORGINGS

The shape-difficulty factor for axisymmetric forgings was originally suggested by

Teterin, et al., (1) and was discussed in the second interim topical report. The deri-
vations are summarized in this Appendix.

A shape factor, xf, is defined as:

f =-(A-i)

where

P = perimeter of the axial cross-sectional surface

F = surface area of the axial cross-sectional surface.

Since the shape factor xf is dimensionless, any unit of length can be used for its calcu-
lation. A cross section through a 10-in. -diam x 10-in. -high cylinder, for example,
has a perimeter of 40 inches and a cross-sectional surface area of 100 inchesZ. The
shape factor, xf, would thus be
402
f 0? 16 (dimensionless)
xf-100

To arrive at a factor for expressing complexity of shape, Teterin, et al, considered

the cylinder that circumscribes the shape of the forging. Thus, the height, Hc, of the
cylinder is equal to the maximum height of the forged shape, and the diameter, Dc, is
equal to the largest diameter of the forged shape. The cylinder is considered to repre-
sent the simplest shape and any modification of the cylinder to introduce more com-
plexity. Mathematically, the shape factor of any cylinder, xc, can be expressed as

2 4
Pc (Hc + Dc)2
xc = F7 H D (A-2)

The authors proposed that a "longitudinal shape" factor, a, be designated to com-

pare any shape with its circumscribing cylinder as follows:
Xf
S=--(A-3)
xc

The shape factors xf and xc in Equation (A-3) are calculated from Equations (A-l)
and (A-2), respectively. If xf = xc, the value for a•c would be 1, expressing the longi-
tudinal shape factor for a simple cylinder.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 A-2

Teterin, et al, recognized that projections on circular forgings (e. g. , shaft or

rims) become progressively more difficult to forge as the projections are located at
progressively greater distances from the axis of the forging. To accommodate such
variations in forging difficulty, the authors proposed a "lateral" shape factor, 3, which
is used to express the degree of forging intricacy in the lateral direction. Thus,
2 Rg
=3 , (A-4)
Rc

where
Rg = radial distance from the axis to the center of gravity of half of the
cross section
Rc = maximum radius of the forged piece which equals the radius of the
circumscribing cylinder.

For a cylinder, Equation (A-4) gives Pc = 1.

In summary, the parameters a and /3 define the relative difficulty of forging a

given workpiece in two different planes. A shape-difficulty factor, Sf, incorporating
both of these parameters, a and P, is defined as

Sf . (A-5)

Substituting Equations (A-3) and (A-4),

=xf ZRg (A-6)

xc Rc

Continuing the substitution for these components, using Equations (A-i) and (A-2),

2
P

Sf=p2 "RcRc (A-7)

c
Fc

The authors use the subscript 1 to indicate the shape of a forging under study (same as
the shape of the forging-die cavity). Thus,

Si = _ _ . 2Rgl (A-8)
PC 1

If the part is being forged in more than one step, a shape-difficulty factor must
also be defined for the starting shape of the workpiece, i. e.,

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 A-3 and A-4

(Z 2Rg
so = mogo = " (A-9)

kF

where the subscript o indicates the shape factor for the workpiece before the forging
step under consideration.

If the starting shape is a cylinder, the overall shape-difficulty factor, S, is ex-

pressed as

S i (A-10)
Sc

Since the shape-difficulty factor for a cylinder Sc = -c = I • 1 = I,

then S = S1 (A-Il)

However, if SO represents the shape-difficulty factor of a preform and S1 represents

that for the final forging shape, the shape-difficulty factor is expressed as

S =S .
so (A-12)

In this case, the values for both S1 and So must be calculated using Equations (A-8) and
(A -9).

Teterin, et al, conducted a statistical-regression analysis of shape factors deter-

mined with these procedures on several shapes of various difficulty. They found that
the shape-difficulty factor, S, forging weight, Q, and forging diameter, Do, represent
the most important factors in determining the flash ratio (w/t) for the forgings studied.

In this regard, the determination of shape-difficulty factor S permits the selection

of flash dimensions with greater confidence.

BATTELLE MEMORIAL INSTITUTE -COLUMBUS LABORATORIES

 APPENDIX B

EMPIRICAL FORMULAS FOR PREDICTING FLASH

DIMENSIONS AND FLASH WEIGHT

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 B-1

APPENDIX B

EMPIRICAL FORMULAS FOR PREDICTING

FLASH DIMENSIONS AND FLASH WEIGHT

Flash Dimensions

Empirical formulas for predicting flash dimensions and flash weight were given in
the second topical interim report and are summarized in this Appendix.
1
The flash thickness is obtained with the equation suggested by Wolf( ) and is
given below:

t=(1.13 + 0.89/Q/2.2 -0.017Q/2.2) (B-i)

25.4

where t is in inches and Q is the forging weight in pounds.

more than 1 pound is given

The flash ratio (width/thickness) for forgings weighing
1
by Teterin and Tornovskij's ( ) equation:

w- = -0.02 + 0.0038S(DO) + 4.93 (B-2)

t \t (Q/2. 2)0. 2

1
weighing less than 1 pound, the equation established by Wolf( ) is
For forgings
suggested:
S= 1. 25 exp (-1. 09 Q/2. 2) + 3 . (B-3)
t

In both formulas,

W = flash-land width, in.

Do = diameter of the initial round stock, in.

S = dimensionless shape-difficulty factor (S designates the difficulty

involved in forging a shape, increasing with increasing shape
complexity).

Q = forging weight, lb,

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 B-2

Flash Weight

The following variables usually are considered in estimatimating the amount of

flash in filling a forging-die cavity:

(1) Weight of the forging

(2) Dimensions of the flash gap

(3) The location of the internal and external parting lines

(4) Shape difficulty.

The experienced die engineer usually considers a base-line flash-metal loss of

about 5 percent of the forging weight and adjusts this value according to respective geom-
etries of the flash and any internal punch-outs. Extra allowances for shape complexity
are generally arbitrary, and, depending on the individual's experienced judgment, the
allowances may be as much as an additional 5 percent of the forging weight. For ex-
ample, a 100-lb axisymmetric forging having a 5-lb punch-out and a complex configura-
tion might be estimated to require

5 lb for flash

5 lb for a punch-out

5 lb extra for shape complexity

15lb total allowance for flash losses.

Such estimates usually are checked out by forging a few billets representing a range of
weights that bracket the calculated weight. The results of these "tryout" runs then are
used as a basis for finalizing the cutting weight for the billets.

Teterin and Tornovskij( 1 ) conducted a statistical analysis of data on numerous

axisymmetric forgings aimed at developing mathematical procedures for estimating
flash-metal losses. The investigators derived a dimensionless parameter, 'Y, which
takes into account the original stock dimensions, the dimensions of the final forging, and
the vertical distance between internal and external parting lines (Figure B-1). The for-
mula for -yis

Hs(B-4)
"ho + hA

whe re

Hs = final height of the forging

ho = minimum distance between flat surfaces upon which stock

was resting when dies were closed

hA = distance between internal and external parting lines.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 B-3

4 -- s

\\ \0\\

"hA= D, \ DI'
hA 0

I - sh ho i

FIGURE B-I. VARIABLES INDICATING THE SHIFT OF THE EXTERNAL

PARTING LINE AND THE INITIAL POSITION OF THE
STOCK IN THE DIES (1)

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 B-4

Statistically analyzing data from a sufficiently large number of forgings, the authors
obtained the following expression for calculating the weight of flash:

Qf( 10 2) , (B-5)

where

Qf = weight of the flash

Q = weight of the forging

and the parameters K 1 and K 2 are determined from

K 1 = 0.54 + 15.44 (Q /2. 2)-0 2 (1 + 0. 00757-0), Q in lb, (B-6)

and

K 2 = 0. 7026 (1 + 0. 01969r)) w/t , (B-7)

where w/t is obtained from Equations (B-2) or (B-3).

The parameter,?), represents a dimensionless expression for shape that incorporates

the shape factor, S, the shape-change parameter, y, and the diameters of the stock
(DO) and the final forging (Dl) thus:

D i 2 (B-8)

whe re

S is the shape-difficulty factor

and

y is obtained from Equation (B-4).

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 APPENDIX C

SYMBOLS USED IN THE FORTRAN PROGRAM

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 C-1
APPENDIX C

SYMBOLS USED IN THE FORTRAN PROGRAM

NRl = number of corners in a half cross section below the parting line, in-
cluding the corners on the parting line and on the axis.

NR2 = number of corners in a half cross section above the parting line, in-
cluding the corners on the parting line and on the axis.

MARK(I) = +1 if the volume component must be added to total

-1 if the volume component must be subtracted from total.

R(I) = radius of i'th corner, R in Figures 2 and 3 or Ri, i_ in Figure 5.

GH(I) = axial distance between two consecutive corners, Figure 5.

FR(I) = fillet radius at the i'th corner, Figure 5.

ALFA(I) = draft angle between the i'th and (i+l)th corners, Figure 5.

PER(I) = 0, if the radius of the i'th corner is given

+1 if the radius of the i'th corner must be calculated from the

radius of (i-l)th corner

-1 if the radius of the i'th corner must be calculated from the

radius of (i+l)th corner.
segment of the perimeter PiPi- 1 from Equations (1), (5), (24) or (30).

DENS = density of the forging material, lb/in. 3.

DIAM = initial stock diameter.

DCYL = diameter of the cylinder circumscribing the forging.

HCYL = height of the cylinder circumscribing the forging.

TAHZ = minimum distance between flat surfaces upon which stock was resting
when dies are closed.

SURF(I) = surface area of the half cross section of a volume component, from
Equations (2), (6), (25), or (31).

VOL(I) = volume of a volume component, from Equations (3), (7), (26), or (32).

RG(I) = radius of center of gravity for a half cross section of a volume

component.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 C-2
TPER = perimeter of the axial cross section of an axisymmetric forging.

TSURF = area of the cross-sectional surface.

TVOL = volume of the forging.

RGT = radius of center of gravity of the half cross section.

SHAPE = shape-difficulty factor, as described in the Second Interim Topical

Report. (1)

WEIGH = weight of forging without flash.

FTHICK = flash thickness.

RATIO = flash-width-to-thickness ratio.

FWIDTH = flash width.

WTOT = total forging weight, including flash losses.

AREA = total plan area of the forging, including flash.

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 APPENDIX D

FORTRAN COMPUTER PROGRAM FOR CALCULATING DESIGN

PARAMETERS IN AXISYMMETRIC FORGINGS

BATTELLE MEMORIAL INSTITUTE- COLUMBUS LABORATORIES

 D-1

0 0 00000000cl00
aC
0C0000 00 C)Ca 0 c0

4444V44v4444
444444444444
6p bFt. w
aaa6pW, W

Ai
ip.~~~~a 4At A . W.
ta-

V. w tp
ipatp ta t.
ep4 a*
V4~~a t aA
wt
$A tAa~ 4Awwtp* A w

w w1A.W [Link]-
~
AtfA
*A~~~~ ~ ~ 0w6AV#At ORt

a~~to (f wa
wta
[Link]. i 4 I 4

alAal aatAl a
if, 4 ba
aaFaw
Y.

Wa
4F if o... V W 6F -
*~a
aa. Vff ata6a
w 6F tp 0. A 6Fba6a
Ir- I ba EpIdA
EA 6 #A UP to
tp

wa #A w ak aA6 ,. aa
V~a ~
V *FkFVaa a t Va
a .
6b vt v w w to a b,
a.
v ip w 6 * EaE
Vý~~~~ v aab .'

LF9 VF.W c a aaFb.

t. Yaa v WtaWW -
a. a a ap p p ta 'r-
a- ap- F Or
A V4
aP1bp- aL a-,A P
h.

t. W, t a1 ai
tr U' tP- b- aýW

a-a v _w cv aa a

aea. f. a.
ba,a'I
a V.

a.F a Wa v
a F v--v
v-

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D- 2

0 00 000 0pi *0
44A444444o4A4
6A

im6AfA 4 4A p Vbp 4pp

ip ip4A 4 6A 4 to ip
to~~ ~ ~ ~ qp# ppAI^W4

p op 6p p
p p p p 4A W #4AWA p p pAf
p" p pA pp k"Vp pP
4
fAtA#A# #AP W, WAR
pp #A Aw w 6A
P FAA4A#

fp w I 6
P *A IAP*
(APP

fp wp tp pA fp pA
ppF p p 4P

ip~~~ ~ ~~ ot i pA
A4 p" pA6

bp~ pF p p
ip ~~
wp ~ f pp#

p p p iP
P 6p 6p
PA t.
PA P PP pP

pp ppFW
Wpv pp pp
if t w w b wp *p
ip vpp
V pwVV ~
~ ~ ~ ~ ~ cp ppt
ppp v
pp to .bp p pF
P V. 46
[Link] v w Ib
Pp

pp ppp
F A ýp pF pF pFq P V
w~~~ ~ ~ v p4po.iV4p
te V4 bg ýp p pA pA%
p t&bVp
ppp pp p p. pF

PPpp vVP tp tp
6p p PP pppwp

P. PP P p P W_ pbp
IF~~ #p pf pA
v ývv ~ ~~ ~ ~ ~ ~ ~ vv4~ p .ppv1-9

BATTELL CLMBSLBRTRE ppORA ppTTT

 D-3

PROGRAm SHAPF([Link]~=INPUTTAPE-6 OUTPUT)

C**DETERMI!NATION OF SHAPEFfACTOR IN FORGING
C**MOOTFCATION IF BEGINNTNeA AND END CORNERS ARE
C** NOT FLAT
(COMMON R(100),NPCH(100),ALFA(100),PER(100),SURF(100),VOL(100),
iPGUOnO) FP (100') MARK (100) ,DFLH( 10.0).
pl=,3.1419q?693A
C*4s ~R,(J)m RADIUS O)F .TI+E [Link]. AT VARIOUS
C ** LOCATIONS
C** ~NP= NU.M8ER OF CORNERS. ON ONE HALF..OF THE
C** FORGING
C** ~~CH (I)![Link]-TH_.AND
C** (!-H)TH CORNER
C** ~~ALFA (1).M ANGLE .[Link]- TAPER .AT .THEI-T
C** CORNER WITH THE AXIS
.c*4 PER.(1)= ~PFRTMETER OF A..SLTCE
C** qliRF(I)=SURFACE AREA OF A qLICE
cl* v ~~~OL(I v 0L IME -F -.
A - [Link]-....
C ** FRUT)= FILLET RADIUS BETWEEN THE I-TH AND
C ** (I-I)[Link] =F(I0.9
FOR. SHARP CORNER
MAR'K(I):1 FOR INCREASING VOLUME
C ~~MARK,(1,Y-1 .[Link]!ASING .VOLUME
C** r\PI=N'JMRFR OF CORNERS FROM REGTN TO PARTING LINE
C** ~ NP?=NIHIMRR OF COPNFPS FROM PARTING-LINE, TP-END
C** TN PFP(T) WF STORE SIGNS TO CALCULATE SOME R(I)S

PEA!. 21,NIP1,NR;,
PRINT 2?,NR1,Np?
NPO(NQ1-1)*?+Npl*
NPDM (NPI.-1)
PRINT86
no 8111,$Npnm,p
PEAD 8?,gMARK(T),P (I) eCH (1) 9FP.(.).,ALFA (1),9PER(I)
PRINTF8i? MARK (I) R I),CHCI) ,FP( I) eALFA(T') ,PER (I)
TFCIE(O.1)GO TO 8_]
pERK(1-1)= APRK( T)
p I-i) =R (I)
rH H1-1) =CH,( )
FR(I-I)=FR (I)
Al A ýFA (1+1_)=ALFA (T) __. . . . .. ..

PE.A!) 8?.-M-A'R-K(-N'n),R(NRD),CH(NRD)qFR(NRfl)+,ALFA(NRD)q0ER(NRD)
pRINIT82,MARK.(NRO)oR(NRD) ,CH(NRD) ,FR(NRO)),_ALFA.(NRO) ,PER (NRD)
NR(pm=Npr3,
nO 8j1=N~r)PqN!Rmq
[Link] .(I.R(I0)__jCH( I) FR( I) ,ALFA(I) ,PER (I)
TFI(T9 [Link]) SO _TO 83
M4RK (j-_])'=MAPK (1)
PfR(T-1 )=PFER(T)

r H (I
1-1) =CH. (J)

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D -4

FR (II) FR T)
A3 11LF A i+! i=nLF A(T)
READ) 8?,MAPK(NP),q(NR.),CHCNP),FRCN.R).,ALF.A(NR)
PRINT82,oMAPK(NP))R(NR)t
, CH (NQ) FR NR) .ALF4CNR)
NRM.N=NP_-I
n084T=1 ,NRMTN,?
QFCE.1)G-0 Tn, !44
yF(PCI)eNF.A.)C,) TO R4
.ANG3LF=ALFAC,T-P)
ANGLF=ANGLF *PT/iRno
TF(.PER(I). L.T.O.'CO To 85
R CI)=R I-?) -PFP (11*TAN (ANGLE) *CH CI)
p R CJ)
=11
SO To 84
85..ANGLE:ALFA(T)
ANGLE=ANGLF*0T /TR A
(CI)=RC(.T+I)-.PER
(1)*TAN (ANGLE) *CHCI1#1)
p CI-])=R T)
A4 CONTINUE
PRINT229NR
PRI NT24
pPINT239(RCT),y=19NR)
PPRINT 2
PR1NT23, CCH(t) ,T1,NR)
.PRINT?9
PRINT23, (FR(Tl oT=l ,NR)
PRINT27
PRINT23, (ALFA( T)% =1-NR)
PRINT23, (PERCI) ,T~[Link].)
PRINT?], (MARK(T),T=1 NR)
READ ?3,OENS~flTAM,DCYLTAHZTAHAHCYL
PR4IN-T?.2,nENSD)[Link] 'TAHZTAHAHCYL
C** 0 F N5= DEN SITT.Y
t** DIAM=STOCK DIAMETER
C** DCYL=LARGEST DIAMETER OF-FORGING
C** TAHZ=MINIMUM DISTANCE BETWEEN FLATS
C ** TAH=ISAC Y~TEEN JNJERN AL- AJND
C* EXTERNAL PARTING LINES
C** ~~~HCYL=,TOTALHEI~GHT [Link]
TVOL=O.
TPER=Oe
TRGS=O.

3?' ALFA(I)=ALFA (T)*PT/IRO.

KIL IM=M'R-1
C** START CALCULATIONS

TFC(ALFA(T-1).Eo.o.)Go TO 41
C** THIS* CHECKS THE PARTING LINE WHICH IS
C** ~~CONSIDERED ..
AS_.CYLINQER
nIVF=AlFA CI-?) -ALFA CT)
RFTA=AHS (01FF)
EP CT) =RFTA*FR(T)
R2=HFTA/?o

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D-5

T?=TAN CR?)

FR1FPPCT) *T?_
n)FLRM:FRl*STN([Link] (I-2))
r)ELRP =FRj *tT I AI_FACTI))
C** CORNIER RADIUS WAS ORIGINALLY STORED AT
C ** ROTH R(I) AND RUT-I)
rFLHdI)=F~R1*CO5(ALFA(I).)
0FLH(I)=AR9(CDEIH(I))
nELHCT-I) =FRl*CnS (ALFA (1-?)_)
nFLHCI-1=ARSCDFIH(T-1))
[Link]( T) +r)FIH ( T1-i
sHETA=SIN CAFTA)
TAS=FR (1)*42* (RTA-SRETA) /2.
TAX(S=2**FR CT) **3*5?**3/ (3.*TAS)
PALF?=ARS(ALFA(T-2))
A~AM M rA=82+*0,5~* PT_-PAL F2
c* * mfo IFY H FRPE
rF(ALFA(T-2).LrT.0.)C,AMllMA=O.5*PI-82-PALF2
rsAtMA=ARq (GAMMA) 5
TF(ALFA(T).,GTAtFAC1-?_))GO To 5
r)FLX= (FR(I) /C;?-TAXS) *STN ([Link]).
PS=R (I) *F[_LX
5? CONTTNIIFJ

PC I-ti R CT-i) +nlLPM

11=R()
+P CT-I)

V/OL Cl) =PT* (R!RMT*TAH/3.-TTAS*2.,*RS)

PGCON=RTPMI/ (l,*s1)
p G I) =(R GCO N*S *O,5* TAH- T AS*PS),/SUR-F T-)
42 cONTTIN(E
miOW CALCULATF. F~oPTHE TRUNCATDCN
C~i AS DEFINED BY POTNTS 1-1 AND 1-2
TAHl=CH( T)-OELH CT-I )DELHC I-?)_,_.
I C* CHCT)=HEiGHT BETWEEN CORNERS STORED AT
C*BOTH CHCT_ AN -±4T1)
43 CON7T NUF
50RFCI-I)=PTl *n,c;*TAHI

VOLCT-)=T*TAH14$RIRM1/3*
nEN~cIN(ALFAC!-'>))
[Link]-;?),Eno..)tFN1l.
PERCT-l)=ARS((P(T-1)-P(I-2))/flEN)
PG I-I.)=PTRM1 /(l.*R~l)
TABS=A6S CTAR)_l
7F([Link].0.0OnO0l)Cl) TO 31

PGl(-1) =O.
\/OL Cr-i) =0
GO To-3.

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D-6

41 rONTTNUE
C** CALCULATIONS FOR CYLTNDER AT PARTING LINE
C** AKII) FOR THE [Link] CONE
PG (1=R (T)/?.
cvOLR (1),IOR (I)*CHCH
(1)
PEIR (.Tf=C-H.(T)
TAH1=CH (I-1 -fFLH(CI-?)
oELH( I)=o.
60 To 43
51 [Link]
PALF2=AI.4(ALFA (I-2))
GFAMM 'A (T~*-2)LT?-P4L AF?* I- AL P+
sklA=~A3ARS (GAMMA)
0*ELx(=FR(
( T)/C?-TAXS)*SIN(GAMmA)
TAS=.TAS
n'O To 52
31 CONT INUJE
TK=I.1
n013pK=1 .3
KK=IK-K
13?, [Link](KK),SURF(KK),VOL(KK),RG(KK)
ORINTl33,IRFTAI)FLH(I),DELH(T..1),TAHR(T).RCI1m1),TAHlGAM
qFT8=8ETA*1R0.,p
PRINT133, I ,IETgRRT2,S2,C2,FR1,DELRDELRM
131 CONTINUE
.SU#F(NR)=0.
VOL (NR)=0,
PG (NR)=0.
PER (NR) =A8 5ý((R (NR.) -.
[Link]-11
TPER=O.
7 SURF = 0
TVOL=Oo
C******CALCUJLATE PEPIMETFRPVOLUMEtSURFACE AND RADe OF CENT. OF
C*****GPAVITvT
n,06l T=2,N
.FMARKI.
TPER=TPER+PER CT)
IF CMAR((T) _,_!I) FMARKZ-1
T SURF TSkJPF +FMAP4K*SUPF (1)
TVOL=TVOL+FMRKVLC)
PGS~pGS+FMARK*PG (I)*SURF (I)
61 CONTINUE
PG T=PG / TStJR F
TSUPF=2.*TSUPP
TPE.4=2.*TPFR
PRINT 911
PRINT67,TPE-Rpe,TSIRF TVOLtRGT
C** STAIRT CALCULATION OF SMAKE FACT-OR
FXF=TPER**2-/TSIJPF

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D-7

ALFS=EXF/EXC
PCYL=DCYL/2.
PETS =2,*RGT/Pc*YL
SHAPE=ALFS*8ETe,
PRINT 62,,SHAPRE_,[Link]
WEIGH=TvOL*DENS
PRINT 63,[Link].L.
C** FOR FLASH THICKNESS WE USE WOLF-S FORMULA

FTATUO0.89*Tl**6*5uu.O.017*Tl
[Link] c(.1.1,tF. 'lEA)15
C*41 FOR FLASH RATIO WE USE wOLF-S FORmULA IF WEIGHT IS
C** ~~SNIALLE-P.-HAN 1 LJR$.QI.HERAI5_E_.[Link] S FORMULA
[Link].3*)tGO TO 71

rO To 72
71 PARTx-1. 094TI
pATIO=1 .?5*F2XP (PART) .3.
72 .[Link]
FWIDTH*RATTO4&FTHICK
PP I .T. ~
_4,tk FTH KFWDJ A1.
oAMZHCYL/ ([Link])
FTAý=SHAPF*O I A.Mi4*?GAM**R/ Y**2

FLAS`HW=(AK) -AKp2)*WETGH/1OO.

PRINT 65vWFIGH9FLASHW9WTOT
[Link] CYL,2q.*FWjRTH
AREA=PI*fTOT**P/4,
PRINT 66oAREA
9.11. FORMAT (40H PFR.IMFTERSU5RF-A-C-EVLUMER OF C,AR-A-ITY)
62 FORMAT(27H SHAPE DIFFICULTY FACTOR IS93F1505)
63 FORMAT(32H FORGINC9, WFIGHT WTTHOUT FLASH 4_qSgIji5 A 5 Is
64 FORMAT (44H FLASH THICKNESSFL-ASH WIDTHFLASH RATIO ARE,3F15#5)
65 FORMATC4I.HF0L~eTjNG WFIGHTFLASH WEIGHT9+oTAL WEIGHT,3FI5S.),
66 FORMAT (38H THF -PROJECTED AREA INCLUDINr, FLASH ISfF15.5)
67 FORMATC1IH PERTM4ET.ERzFj5,_5,9H SURFACEl!UFA5ff5//,SH
1 VOLUME.,F1595,2.
19H RADIu)5 OF C. OF CRAVITYtF1S,5)
?A FORMAT(19H MATFRIAL OENSITYx AJ.A±.SOKDAEE
,1.,1
IH LARGEST DIAMETER =qF1O.4,/lOH4 HEIGHTS .,oF1Oo4)
27 FORMAT(;?7H ANGLESBETWEEN CORNERS ARE)
29 FORMAT(lRH FILLET RADII ARE
25FORMAT(36H AXTA DSTANCES BETWEEN CORNERS ARE)
?4 FORMAT(25H RAnTT OF THE CORNERS ARE)
22 .FORNIAT(IPH NUMRF-R-OF CORNERSo2l1 s)
86 FORMAT(9?H INPUT DATA92N40 NUMRER CORNER OADIIAXIAL DISTANCE BETWE
lFN.CnRNEPS9F,Flj FT RAnIIDRAFT ANGLES)
23 FORMAT(8FI0,4)
1-33 FORMAT(/T5,RF)P.3)
21 FORMAT(8TIO)
82 FORMAT(II0,[Link],4)
END

BATTELLE MEMORIAL INSTITUTE - COLUMBUS LABORATORIES

 D- 8

'C 0,

CD 0
o C>

o C

o 01

C>C
Q 00

11,
%frr-- P.-% p.-
0
im IL P- u)tr
*m f-. C, m,

a CD 0 0 0:

*MM c-
W cMin P- F- - i . CN
ýcm 0
r-I nM-
COOmmm4w4f-WL o NNM 00 M Q f-Ne N
00 0 c 0z c0= 0 Q0>C 0 000
000
CýCQC'C 'L C C00C ýC "CCC 0aQCC'C ' Da0 0 000C -

z zlL, -
M i I jVIFUN4---WCL
-C 1 Ca ZIC4(I-p.
.- C -. rp- Ctý-*
- r-m ?- I C%,C

Ul .fti C 0 CO 0 000c C~0

. =c Z
C ý0 c C>C0 00000c C 41 0 C0'C0 000
I= 0 LjC C(C0
-

cl a, Z C
v- cr
A- IL m ICr , A c ,4 4 m m C " 'N t - C c

Cr L
.cI
c z
IaD _jccN>C J---.-m
mc
4(cc C
c~ (
.
Ct C'- 0ý 0 [Link]. CL Cf- 'rC--.
CC 00.-..-'44cD0
~
-0c0 r(ý0=c V'-4= -C o
f_- Q .0
> _ C--- ILJ t jc c

ui i
2O
LLJ i0 0
C
0
0
C
O 0
0
0
0
0
.
0
0
0
0
o
0 00
o o
0
~ 0 0
00
C O O
0
-r ft
fr. LL
z~L.
W
~M U ~ L) U J.....J U U )...'~L .~ ' U

UL c W. op W,~ >

BATEL MEORA INTTT -CLUBU LABRATRIE