5th International & 26th All India Manufacturing Technology, Design and Research Conference (AIMTDR 2014) December

12th–14th, 2014, IIT

Guwahati, Assam, India

CONTROL OF OVALITY IN PIPE BENDING: A NEW APPROACH

A. V. Kale1* , H. T. Thorat2

1. Department of Mechanical Engineering, Yeshwantrao Chavan College of

Engineering, Hingna road, Nagpur, Maharashtra, India, Pin. 441110,
[email protected]
2..Director, NIT Srinagar, Uttarakhand,India.Pin 246174, [email protected]
Abstract
This paper formulates and analyses results of experimentation of pipe bending process on a new modified pipe
bending mechanism. Ovality or flattening of pipe in the bent portion is a major defect and contributes to poor quality
of the bend, deforms the cross section and reduces the strength of the bend to cause easy buckling of the pipe.
Several other defects in addition to this drawback of the bending process necessitated corrective action and
modification in the existing equipment to bend pipe with minimum defects. Hence, this work targets at design,
development, fabrication and use of equipment, which will produce pipe bends with better quality without
sacrificing simplicity of existing equipment. Analysis of experimental data is presented here
Key Words: Tube/Pipe Bending, Rotary Compression Bending, Ovality

equipment to bend pipe with minimum defects. Hence,

1. Introduction this work targets at design, development, fabrication
Considerable progress is made in all aspects of and use of equipment, which will produce pipe bends
equipment involved in pipe bending processes. The with better quality without sacrificing simplicity of
machines, dies, mandrels, tooling and automatic existing equipment.
controls have undergone substantial development in Existing compression bending mechanism is
recent years. Existing tools and equipments do not give modified and used for the experimentation. Pipes of
the output to the required standards. Bending process of constant diameter but different thicknesses are bent for
the pipe is accompanied by various defects such as various positions of top support plate i.e. with or
wrinkles, bumps, scratches and excessive ovality. These without pre-compression of pipe. Experimentation is
defects contribute to poor quality of the bend[1]. carried out using 80 pipes. All pipes are bent for total
Ovality or flattening of pipe in the bent portion is a bend angle of 850. Dimensions of each pipe are
major defect. Ovality in the bent portion deforms the measured before and after bending. Experimental results
cross section and reduces the strength of the bend, so it are tabulated. Tables contain diameters before and after
buckles easily[2]. The flattening of elastic tubes under bending at thirteen discrete positions along the length of
pure bending has been a focus of much attention over the pipe. From this data, ovality before and after
many decades since Brazier first studied this problem in bending are calculated.
1927[3]. Buckling of pipes during bending is studied by
Belke L.[4]. Gerber[5] studied flattening effect by 2. Proposed Concept for Control of Ovality
taking the plastic properties into consideration, although
In the existing bending mechanisms, a circular pipe
his investigation was relatively elementary. An oval
is loaded in the fixed form die as shown in Fig. 1 in
cross-section is assumed for the deformed pipe
which points A, B, C, and D are equidistant from the
after bending. Equation for radius of a point at any
centre. Arrows in the figure indicate the direction of
circumferential angle is established for this cross-
reaction by fixed forming die against the force applied
section by Pan K. and Stelson K.A.[6]. Numerous
for bending.
contributors have reported their experiments on this
phenomenon and provided some elementary analyses to
fit the experimental results. These results form the basis
for the further theoretical studies of the problem[7-12].

Drawbacks of the bending process necessitated

corrective action and modification in the existing

192-1
 CONTROL OF OVALITY IN PIPE BENDING: A NEW APPROACH

3. Operation of the Mechanism

Reaction of
A schematic of the new rotary compression bending
Bending A B Fixed
mechanism is as shown in Fig. 4. The mechanism is
Force Form Die
manually operated and similar to the compression
bender, which consists of two rollers. One roller is large
in size and having the radius equal to the bend radius.
D Angular movement is locked by using dowel pins, but it
can move up and down due to springs, which are placed
Figure 1 Pipe Fixed in Form Die Before Bending beneath this roller. Another roller, which is smaller in
size, is mounted on the separate bolt and is in the same
Pipe bends when the force is applied through horizontal plane with the big roller. The small roller is
wiping die. In this process, the point B remains at the fastened to the link by a nut through an oval hole. This
same position but point A is shifted towards the centre oval hole permits the horizontal adjustment of the small
due to compression of the pipe. As there is no support roller, for bending the pipes of different diameters.
at points C and D, therefore, these points are shifted
vertically outwards, as shown in Fig ig. 2, to form an
elliptical cross section having major and minor axes
resulting in oval cross section at the bend.
From the observations, it is concluded that the pipe
must be supported for maximum of its periphery and it
is also necessary to provide the support to the pipe
Location
during bending to avoid the deformation of the pipe Pin
vertically.

Original Section C Deformed Cross Fig. 4 Bending action in Modified Mechanism

4. Experimentation
Extrados
Intrados
The aim of experimentation is to use modified
A B
bending mechanism to bend the pipes. Experimentation
is planned for the pipes with constant diameter of 22
mm but of varying thicknesses. Tubular blanks of length
350 mm are used for experimentation.
ation. Pipes used are of
thickness 1.2, 1.25, 1.5 and 2.0 mm. Pipes are procured
Figure 2 Deformation of the Pipe without Vertical accordingly. These are grouped according to thickness
Support and colour coated. Pipes are marked with the grid
pattern as shown in figure 5 and made ready for
In the proposed method, an attempt is made to experimentation. Dimensionsions are measured before
restrict outward movement of points C and D by using bending. Based on these dimensions, the pipes are
top and bottom support as shown in Figure 3. If classified into sub groups having approximately same
increase in diameter while bending is restricted, dimensions. Five pipes are used from each group and
deformed cross-section
section cannot flatten easily. The roller bent on the mechanism.
support should not overlap the pipe for larger area. Each
roller envelops for 130º on the pipe diameter.
1
Top 8 2

Support Support
7 3
from Big Support from
6 4
Roller Small Roller 5

Bottom
Support
Figure 5 Grid Pattern on the Pipe
Figure 3 Proposed Concept

192-2
5th International & 26th All India Manufacturing Technology, Design and Research Conference (AIMTDR 2014) December 12th–14th, 2014, IIT
Guwahati, Assam, India

5. Description of Pipe Bending Process on 6. Measurement of Dimensions before and

Modified Bending Mechanism after Bending
Bending operation of the pipe on the mechanism is done As the circular cross section is divided into eight
in the following sequence: parts as shown in figure 5, therefore, outside diameter is
i. A tubular blank of the pipe is located in the pin measured before bending of the pipe at four positions
provided at rear end. It is located in such a way that i.e. between points 1-5, 2-6, 3-7 and 4-8 with the help of
the starting point of the bend goes towards the pin digital Vernier Calliper. The data is recorded in the
side and the seam is at the top. As there is an sheets similar to table1. Thickness is measured at each
excessive clearance between inside diameter of the point on the orthogonal grid along 13 discrete places,
pipe and outside diameter of the pin, therefore, it is corresponding to eight divisions, with the help of
easy to locate the pipes of various inside diameters. ultrasonic thickness tester. The instrument is calibrated
ii. Initially, when the pipe is inserted on the location pin, for Mild Steel specimen of thickness 1.2, 1.25, 1.5 and
outside diameter of the pipe is not guided in the 2.0 mm. After the measurements of diameter and
groove of floating big roller. Central roller is spring- thickness, mean diameter is calculated for each group of
loaded. Rotation of this roller is restricted by the pipe separately and recorded in the same table.
providing three dowel pins of diameter 12 mm. It is There are 80 such data recording sheets for 80 pipes.
always touching to the bottom of the top support From each group, twenty pipes are selected. Five
plate due to the spring force. Top support plate, pipes are bent for each condition of top support plate.
which has threads on its inside diameter, can be Diameters at four positions and thickness at each point
moved up and down. Central roller moves with the of the grid along the length of the pipe are measured and
top support plate to adjust the required bending recorded after bending. Moreover, change in length
condition. Thus, pipe is located exactly at the centre between the sectors as marked on the pipe along its axis
and at high points of the diameter. is also measured for each pipe for 12 sectors and
iii. Bending is performed in four conditions of the top recorded in the same table. Values of maximum ovality
support plate as mentioned below: increase in thickness on tension and compression side,
• Keeping the gap between the pipe and bottom of increase in length between the markings along the
the top support plate. This is called as open length of the pipe before and after bending is computed
condition. in the same table.
• Keeping the bottom surface of the top support plate
touching to extreme periphery of the pipe. This is 7. Analysis of Experimental Data
called as touching condition.
• Compressing the pipe by 0.25 mm by lowering the From the values of diameters before bending
bottom surface of the top support plate by 0.25 mm from table 1, percentage ovality is calculated as:
compared to the touching condition. This is called % Ovality before bending =
as load 1 condition.
• Compressing the pipe by 0.5 mm by lowering the D1B −D 3 B
×100 ( 1)
bottom surface of the top support plate by 0.5 mm Do
compared to the touching condition. This is called
as load 2 condition. From the values of diameters after bending, ovality is
iv.When the force is applied at the end of lever, the calculated as:
small roller advances and locates itself on to the
diameter of the pipe, gripping it against the central % Ovality after bending =
roller. Provision is made in the design for the Dmax − Dmin
×100 (2)
vertical adjustment of the small roller also, if Do
required.
Values of percentage ovality of the pipes before
v. The bending arm rotates about the fulcrum point of
bending are as shown in Ttable 1, for all thicknesses.
central bolt and the small roller wipes the pipe while
These pipes are bent for different positions of top
rotating about itself.
support plate during experimentation.
vi.During the rotation of the bending arm, compressive
force exerted by small roller is used to advance the
pipe and bend to conform to the shape of the central
roller.
vii.After bending, bending arm is brought back to its
original position, the top support plate is moved up
and pipe is removed.

192-3
CONTROL OF OVALITY IN PIPE BENDING: A NEW APPROACH

Table 1 Ovality before Bending at the Centre of the that, as the thickness increases, the value of ovality
Bend decreases at the center of the bend.

Open Touching Load 1 Load 2 The basic objective is to reduce the ovality in pipe
Group t ψ after bending with the help of new concept. This
% Ovality BB
reduction in ovality is achieved by bending the pipes
Group B 1.20 42.62 0.540 0.56 0.86 0.48 with different conditions of top support plate i.e. with
pre-compression. Values of ovality after bending at
Group A 1.25 42.62 0.144 0.19 0.14 0.16 various cross sections of pipe are as shown in table 3.
For all thicknesses, ovality is large for open condition
Group C 1.50 42.62 0.270 0.19 0.25 0.19 i.e. without the constraint on deformation in vertical
direction. If the top compression plate is kept touching
Group D 2.00 42.62 0.250 0.33 0.28 0.35 with the pipe diameter and the pipe is bent, ovality
reduces to half of its values in open condition. If the
pipe is pre-compressed by 0.25 mm, ovality further
Values of percentage ovality after bending, at the center reduces to half of its values than in touching condition
of the bend (at 42.620 bend angle), are as shown in for smaller thickness i.e. 1.2 and 1.25 mm. It is
Table 2 for all thicknesses and different positions of top marginally reduced in higher thicknesses. If the pipe is
support plate. pre-compressed by 0.5 mm, ovality reduces
considerably irrespective of the thickness and value of
In the Table 2, percentage ovality for thickness 1.2
ovality remains more or less same for all thicknesses.
mm, ranges from 46% to 6.93 % for open to load 2
condition. For 1.25 mm, it is 42% to 5 % depending The comparison of experimental percentage ovality
upon pre-compression of the pipe. Increase in the after bending for different positions of top support plate
percentage ovality after bending is very high for bend is as shown in Figure 6.
angles ranging from 14.210 - 71.030. It is also observed

Table 2 Ovality at the Centre of the Bend after Bending

Maximum Ovality after Bending at the Centre of the Bend
Open Touching Load 1 Load 2
Group t ψ Ovality % Ovality % Ovality % Ovality %
mm Ovality mm Ovality mm Ovality mm Ovality
Group B 1.20 42.62 10.86 46.43 4.81 22.32 2.31 10.4 1.54 6.93

Group A 1.25 42.62 10.77 42.03 4.47 20.12 2.11 9.5 1.13 5.55

Group C 1.50 42.62 4.19 18.86 1.76 7.91 1.67 7.52 1.24 5.51

Group D 2.00 42.62 2.7 11.81 2.46 6.87 1.89 6.41 1.05 4.88

192-4
5th International & 26th All India Manufacturing Technology, Design and Research Conference (AIMTDR 2014) December 12th–14th, 2014, IIT
Guwahati, Assam, India

(For Group A, 1.25 mm Thickness) buckling during bending and moreover, neutral axis is
shifted to tension side avoiding flattening of the pipe
50.00 Open thus reducing ovality.
40.00 Condition
% Ovality

30.00 Touching Percentage ovality range for open to fully pre-

Condition
20.00
Load 1 compressed condition of the pipe is as follows.
10.00 Condition 46.43% - 6.93 % for 1.2 mm thickness
0.00 Load 2 42.03% - 5.55 % for 1.25 mm thickness
Condition
0 50 100 18.96% - 5.10 % for 1.5 mm thickness
Bend Angle
11.81% - 4.88 % for 2.0 mm thickness
(For Group B, 1.2 mm Thickness)
References:
50.00
Open
Condition 1. Alexander Fred D. and Shaw R.M.,1985,Tool
40.00
Touching
% Ovality

30.00 Condition And Manufacturing Engineers Handbook of

20.00 Load 1 Metal Forming, 1985, Mc Graw Hill Book
Condition Company, New York, pp. 10-43. Chap.10.
10.00
Load 2
0.00 Condition 2. Alexander Fred D. and Shaw R.M., 1985, Tool
0 50 100 and Manufacturing Engineers Handbook of
Bend Angle
Metal Forming, 1985, McGraw Hill Book
(For Group C, 1.5 mm Thickness) Company, New York, pp. 661-673.
3. Brazier L.G., 1927, On Flexure of Thin
20.00 Open Cylindrical Shells and Other Thin Sections,
15.00 Condition Proc.R.Soc, London.,Ser.A, 116, pp. 104-114.
% Ovality

Touching
10.00 4. Belke L., 1984, A Simple Approach for Failure
Condition
5.00 Load 1 Bending Moments of Straight Pipes, Nucl.
0.00 Condition Engng. Des., 77, pp. 1-5.
Load 2
0 50 100 5. Gerber T.L.,Plastic Deformation of Piping Due
Condition
Bend Angle to Pipe Whip Loading, ASME Paper 74-NE-1.
6. Pan K. and Stelson K.A., Nov 1995, On The
(For Group D, 2.0 mm Thickness)
14.00
Plastic Deformation of Tube During Bending,
12.00 Journal of Engineering for Industry,
% Ovality

10.00
8.00
Open Condition Transactions of the ASME, 117, pp. 494-500.
Touching Condition
6.00
Load 1Condition
7. Prinja N.K and Chitkara N.R., 1984, Post
4.00
2.00
Load 2 Condition Collapse Cross Sectional Flattening of Thick
0.00 Pipes in Plastic Bending, Nucl. Engng. Des.,
0 50 100
83, pp. 113-21.
Bend Angle
8. Prija N.K., Parkar J.V. and Day B.V., 1985,
Figure 6 Variation of Ovality for Four Positions of The Large Strain-Large Displacement
Top Support Plate. Behaviour of A Whipping Pipe, Proc.
I.Mech.E.E., pp. 225-34.
In figure 6, highest value of maximum ovality is for 9. Reissner E., 1961, On Finite Pure Bending of
Cylindrical Tube, Osterr.Ing.Arch, 15, pp.165-
open condition and it decreases as pre-compression is
72.
increased by lowering the position of top support plate. 10. Seide P. And Weingarten V.J., 1961, On The
Its value is lowest for load 2 condition i.e. when the Buckling of Circular Cylindrical Shells Under
pipe is pre-compressed to 0.5 mm. It is clear from figure Pure Bending, J. Appl. Mech., Transactions of
that ovality decreases as the thickness increases. the ASME, 28,pp.112-16.
11. Ueda S., 1985, Moment-Rotation Relationship
8. Conclusion Considering Flattening of Pipe Due to Pipe
From the experimental data, percentage ovality is Whip Loading, Nucl. Engng. Des., 85, pp. 251-
computed at thirteen discrete places along the length of 259.
the pipe. Graphs are plotted to check its variation with 12. Yu T.X and Hua Y.L., 1986, The Whipping
respect to bend angle. It is observed that percentage During Pipe Rupture Accident and its Defence
ovality at the centre of the bent pipe decreases with in Nuclesr Power Plant, Journal of Pressure
increase in wall thickness and pre-compression of the Vessel Technology, Transactions of the ASME,
pipe. This is because the larger thickness pipe resists 3, pp.70-76. (In Chinese).

192-5