Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.14 Placement of transducer in a typical specimen
(Left: transducer on the top side of chord, Right: transducer on the rotating saddle)
The load-displacement relationship of the trunnion brace provides an indication of the
yield and ultimate load capacity of the trunnion. The load-displacement response of
the trunnion brace to shear indicated the elastic, inelastic and failure behaviour of the
plate trunnions under shear loading. The load-displacement relationship was obtained
by plotting the readings obtained from the transducers.

3.3

Governing failure mode of trunnion

In the following section, a description of the failure modes of trunnions is made. It is
observed that the ultimate failure tests produces a distinct load deformation path in all
the specimens and it is possible to ascertain the failure mode, ranging from chord
plastification effect to fracture of the shear plate.

3.3.1 Pure pipe trunnion
This set of specimens consists of trunnions with attached pipe only. It is observed that
there are two distinct failure mode of this type of trunnions, namely chord tension pull

90


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

out failure and brace shear failure through fracture and chord plastification of the
chord resulting in the indentation of the chord. Figures 3.15 to 3.19 show the effect of
direct shear failure of the brace. The deformation of the brace and fracture line on the

brace can be seen clearly from the diagrams shown.

Figure 3.15 Deformation and governing failure mode of specimen CT1

Figure 3.16 Deformation and governing failure mode of specimen CT2

91


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.17 Deformation and governing failure mode of specimen CT3

Figure 3.18 Deformation and governing failure mode of specimen CT4

Figure 2.9 below shows the deformation and fracture on the chord. Due to the thinner
chord wall used in this specimen, it is possible for the chord to fail by fracture on its
chord wall.

92


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.19 Deformation and governing failure mode of specimen CT5

Only specimen CT3 shows that there is shear brace failure, all the other four
specimens indicated chord tension pull out failure. The latter failure mode is usually
accompanied by large chord indentationd as seen in the diagram. The load
deformation curves for specimens CT1 through CT5 are shown in Figure 3.20. The

plots show a distinct elastic range, onset of strain hardening, ultimate load and the
final fracture point. The ultimate load reached for specimen CT1, CT2, CT3, CT4 and
CT5 are respectively 4,800kN at a maximum deflection of 27mm, 2,175kN at a
maximum deflection of 20mm, 5,417kN at a maximum deflection of 50mm, 2,940kN
at a maximum deflection of 17mm, and 4,568kN at a maximum deflection of 25mm.
All the specimens show high ductility prior to the ultimate failure load indicating that
is there is a lot of reserve strength in the trunnions. As the grommet are placed about
200mm away from the face of the chord wall, the load on the brace is predominantly
shear and it is beneficial in the design of trunnions that there is high ductility. Thus
there is a less likely chance for the trunnion to suffer premature failure due to sudden

93


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

fracture at the limit load. Table 3.20 gives a summary of the ultimate loads and
displacement for specimens CT1 to CT5
6000

Total Load, P (kN)

5000

4000

3000

P

2000

 

Specimen CT1
Specimen CT2
Specimen CT3
Specimen CT4
Specimen CT5

1000

0
0

10

20
30
40
Displacement, ' (mm)

50

60

Figure 3.20 Load deformation curves for specimens CT1 to C5

Table 3.2 Summary of the ultimate loads and displacement for specimens CT1 to CT5
d0

t0

d1

t1

Fu,test

'

mm

mm

mm

mm

kN

mm

CT1

508.0

20.5

324.0

17.6

4800

27

CT2

508.0

12.5

324.0

12.4

2175

20

CT3

508.0

20.5

406.4

12.5

5417

50

CT4

508.0

12.5

406.4

12.5

2940

17

CT5

508.0

15.2

406.4

17.0

4568

25

Specimen

94


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

3.3.2 Through pipe trunnion
This set of specimens consists of trunnions with through pipes only. In this case, the
pipe is slotted though the chord wall and welded together, thereby providing greater
strength in shear and prevents less chord tension pull-out failure where the chord wall
is not thick. It is observed that there is one distinct failure mode of the trunnion here,
namely brace shear failure through fracture of the brace.

Figures 3.21 and 3.22 show the effect of direct shear failure of the brace. The
deformation of the brace and fracture line on the brace can be seen clearly from the
diagram shown. The shear failure occurs just at the intersection between the weld and
the brace.

Figure 3.21 Deformation and governing failure mode of specimen CT6
The load deformation curves for specimens CT6 and CT7 are shown in Figure 3.23.
The plots show a distinct elastic range, onset of strain hardening, ultimate load and the
final fracture point. The ultimate load reached for specimen CT6, and CT7 are

95

Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

respectively 4,136kN at a maximum deflection of 34mm and 5,150kN at a maximum
deflection of 41mm. All the specimens show high ductility prior to the ultimate failure
load indicating that is there is a lot of reserve strength. Table 3.3 provides a summary
of the ultimate loads and displacement for specimens CT6 and CT7. Figure 3.23 is the
load deformation curves for specimens CT6 and CT7.

As the grommet are placed about 200mm away from the face of the chord wall, the
load on the brace is predominantly shear and it is beneficial in the design of trunnion
when there is high ductility. Thus there is a less likely chance for trunnions to suffer
premature failure due to sudden fracture at the limit load.

Figure 3.22 Deformation and governing failure mode of specimen CT7

Table 3.3 Summary of the ultimate loads and displacement for specimens CT6 and CT7
d0

t0

d1

t1

Fu,test

'

mm

mm

mm

mm

kN

mm

CT6

508.0

12.4

324.0

12.4

4136

34

CT7

508.0

12.5

406.4

12.5

5150

41

Specimen

96


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

6000

5000

Total Load (kN)

4000

3000

P

2000

✁

1000
Specimen CT6
Specimen CT7
0
0

10

20
30
Displacement (mm)

40

50

Figure 3.23 Load deformation curves for specimens CT6 and CT7

3.3.3 Combined shear plate and pipe trunnion
This set of specimens consists of trunnions with shear plates slotted through the chord
walls and welded together. The brace are welded to the shear plate inside the core to
provide the circumference for the grommet during loading and they are also welded
onto the chord wall. That is the static strength of these specimens utilises the strength
of the brace as well as the shear capacity, effectively increasing the shear and bending
carrying capacity for consideration in the trunnion design. This combined trunnion
provides clues to the effectiveness of the brace and shear plate when combined
together compared to the isolated cases as described in the earlier sections. It is

97

Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

observed that the governing failure mode for this specimen is through chord tension
pull out failure for all the four specimens. Figures 3.24 to 3.27 shows the deformation
and fracture of the brace and shear plate. The inserts in the pictures show the buckling
mode of the shear plate inside the chord wall. The buckling of the shear plates
effectively reduces the shear carrying capacity of the trunnion. Due to the width of the
outer diameter of the chord used, the shear plate used for such pipe trunnions must be
sufficiently thick to reduce the effect of buckling of the shear plate.

Figure 3.24 Deformation and governing failure mode of specimen CT8

Figure 3.25 Deformation and governing failure mode of specimen CT9

98


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.26 Deformation and governing failure mode of specimen CT10

Figure 3.27 Deformation and governing failure mode of specimen CT11

The load deformation curves for specimens CT8 to CT11 are shown in Figure 3.28.
The plots show a distinct elastic range, onset of strain hardening, ultimate load and the
final buckling of the shear plate. The ultimate loads reached for specimen CT8 to
CT11 are respectively 5,482kN at a maximum deflection of 14mm, 2,862kN at a
maximum deflection of 9mm, 7,596kN at a maximum deflection of 30mm and

3,500kN at a maximum deflection of 16mm. These combined pipe and shear plate
specimens also show high ductility prior to the ultimate failure load indicating that is
there is a lot of reserve strength in such trunnions. The grommets are placed about
200mm away from the face of the chord wall, thus the load on the trunnion is

99


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

predominantly shear. Table 3.4 provides a summary of the ultimate loads and
corresponding displacements for specimens CT8 to CT11
9000

8000

Total Load, P (kN)

7000
P

6000

✂

5000
4000

3000

2000

Specimen CT8
Specimen CT9

1000

Specimen CT10
Specimen CT11

0
0

10

20
30
40
Displacement, ' (mm)

50

60

Figure 3.28 Load deformation curves for specimens CT8 to CT11

Table 3.4 Summary of the ultimate loads and displacement
for specimens CT8 to CT11
d0

t0

d1

t1

Fu,test

'

mm

mm

mm

mm

kN

mm

CT8

508.0

20.5

324.0

17.5

5482

14

CT9

508.0

12.5

324.0

12.4

2862

9

CT10

508.0

20.7

406.4

12.5

7596

30

CT11

508.0

12.5

406.4

12.5

3500

16

Specimen

100


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

3.3.4 Tubular X-joints
This set of specimen consists of trunnions with only attached pipes but with the length
of brace extended. The length of brace used for specimens CT12 to CT17 are
respectively 400mm, 800mm, 1300mm 400mm, 800mm, 1600mm. It is observed that
there are two distinct failure modes of the trunnions here, namely chord tension pull

out failure and brace shear failure through fracture and chord plastification of the
chord resulting in the indentation of the chord. Figures 3.29 to 3.34 show the effect of
direct shear failure of the brace and chord indentation. The deformation of the brace
and fracture line on the brace can be seen clearly from the diagrams shown.

Figure 3.29 Deformation and governing failure mode of specimen CT12

Figure 3.30 Deformation and governing failure mode of specimen CT13

101


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.31 Deformation and governing failure mode of specimen CT14

Figure 3.32 Deformation and governing failure mode of specimen CT15

Figure 3.33 Deformation and governing failure mode of specimen CT16

102


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.34 Deformation and governing failure mode of specimen CT17
The load deformation curves for specimens CT12 through CT17 are shown in Figure
3.35. The plots show a distinct elastic range, onset of strain hardening, ultimate load
and the final fracture point. The ultimate load reached for specimen CT12 to CT17 are
respectively 2,780kN at a maximum deflection of 50mm, 1,512kN at a maximum

deflection of 90mm, 966kN at a maximum deflection of 180mm, 3,780kN at a
maximum deflection of 60mm, 2,126kN at a maximum deflection of 80mm, and
1,051kN at a maximum deflection of 240mm.

All the specimens show high ductility prior to the ultimate failure load indicating that
there is a lot of reserve strength for such trunnions. As the grommet is placed at
lengths ranging from 400mm through 1600mm away from the face of the chord wall,
the load on the brace is predominantly via bending moment. The results are later used
in the comparison of the amount of shear that gives rise to bending moment as the
loading arm of the brace becomes longer.

103


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

4000
Specimen CT12
Specimen CT13
Specimen CT14
Specimen CT15
Specimen CT16
Specimen CT17

3500

Total Load, P (kN)

3000

P
✄

2500

2000

1500

1000

500

0
0

50

100
Displacement, ' (mm)

150

200

Figure 3.35 Load deformation curves for specimens CT12 to CT17
Overall, it is observed from the experimental tests of all the four different types of
trunnion specimens and X-joints tested that there is high level of ductility, due to
strain hardening, in the trunnions regardless of the configuration type. There is also
distinct yielding behaviour under load and the load deformation plots show clearly

defined ultimate failure load. The failure modes of these specimens are predominantly
shear failure through fracture of the shear-loaded arm of the trunnion except for
tubular X-joints where the loads gives rise to bending moments. Table 3.5 shows a
summary of the ultimate loads and displacement for CT12 to CT17. The next section
will focus on the static strength and how they compare with the current practice.

104


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Table 3.5 Summary of the ultimate loads and displacement
for specimen CT12 to CT17
d0

t0

d1

t1

Fu,test

'

mm

mm

mm

mm

kN

mm

CT12

508.0

20.7

324.0

17.5

2780

50

CT13

508.0

20.7

324.0

17.6

1512

90

CT14

508.0

20.7

324.0

17.6

966

180

CT15

508.0

20.7

406.4

12.5

3780

60

CT16

508.0

20.7

406.4

12.5

2126

80

CT17

508.0

20.7

406.4

12.5

1051

240

Specimen

3.4

Discussions on the test results

The specimens in the tests were carefully selected so that different types of failure
mechanisms of the trunnions can be observed. The behaviour and strength of these
specimens are discussed below.

3.4.1 Design strength of pure pipe trunnions
Based on the joint shear resistance for the pipe trunnions as described in the previous
chapter, the same formulation is used in this case for the large pipe trunnions. This
series of tests have been calibrated and uses more advanced technology and resources
which reflects more accurately vis-a-vis the first set of tests that grommets are more
prone to unpredictable behaviour with better control of the experimental results
obtained. Hence this series of tests serves to enhance the previous series of tests and
also the larger dimensions used and the higher shear loads generated help to reduce
any error due to sizing effects.

105


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

The test results for specimens CT1 to CT5 are given in Figures 3.36 and 3.37, which
shows the various observed behaviour including first yield (indication of flaking of
white wash on the chord or brace), initial crack, ultimate load reached and the fracture
point. These points have been painstakingly recorded in order to fully understand the

actual behaviour of the pipe trunnion during the various loading stages. The first yield
has been observed through the white wash flaking, and this should not be
misconstrued as the elastic yield strength. However, it serves as an important
indication of the final failure mode of the pipe trunnion and where failure occurs.

106


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

2500

6000
Ultimate Load
@4830kN
Fracture
@4460kN

Initial Crack
@4200kN

Load (kN)

4000

2000

Initial Crack
@2000kN

Load (kN)

5000

Ultimate Load
@2300kN
Fracture
@1970kN

1500

3000

1000
First Yield (Chord )
@2000kN

2000

First Yield (Chord )
@730kN
500

1000

Specimen CT2

Specimen CT1
0

0
0

10

20
30
40
Displacement (mm)

0

50

20
40
Displacement (mm)

60

4000

6000
Ultimate Load @5420kN
5000

Fracture
@5350kN

Initial Crack

@4620kN

3000

Load (kN)

Load (kN)

4000

3000

2000

First Yield (Brace )
@1890kN

Ultimate Load
@2950kN
Initial Crack
@2470kN

2000

Fracture
@2460kN

First Yield (Chord )
@1080kN

1000

1000
Specimen CT3

Specimen CT4

0

0
0

20
40
Displacement (mm)

60

0

20
40
Displacement (mm)

60

Figure 3.36 Design load for specimens CT1 to CT4

107

Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

6000
Ultimate Load
@4580kN

5000

Load (kN)

4000

Initial Crack
@3780kN

Fracture
@4070kN

3000

2000
First Yield (Chord )
@1290kN

1000

Specimen CT5

0

0

20
40
Displacement (mm)

60

Figure 3.37 Design load for specimen CT5
The first yield occurred on the chord surface for specimens CT1, CT2, CT4 and CT5
with these specimens ultimately failing through chord tension pull out. First yield
indication for specimen CT3 started on the brace and the ultimate failure mode was
through shear fracture. Thus once the initial deformation starts, the load deformation
generally follows the weakest link and the ultimate failure path is thus determined.

Table 3.6 is a summary and comparison of the loads and formulations used to
calculate the pipe trunnion design strength. The following formulations, Equations 3.1
to 3.3, are used to compare with existing recommendations on shear and bending
moment effects on the specimens.

Mu
f y 0 ˜ t0 ˜ d1

5.1 ˜ J 1.04
☎

✆

0.43
☎

(1  0.4E )  (1  0.4E ) 2 

2

2  (0.4 E ) 2

(3.1)

J2

108


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

V1

M1

S ˜ d m1 ˜ t1 ˜ 0.4 ˜ f y1

(3.2)

V1 ˜ w1

(3.3)

For the failure mode caused by chord indentation, M1 is very close to Mu or even
higher. This is because the full effective shear capacity of the brace can be mobilised

before chord indentation sets in and this path is continued to the ultimate failure.
When Mu is low, the joint design resistance for bending moment is low and this
generally results in chord indentation when M1 increases. On the other hand, for
specimen CT3 where the failure mode is due to brace shear failure, the effective shear
area is mobilised and shows that the existing shear strength estimate is very
conservative. Here, it is noted that M1 is much lower, indicating the high capacity of
the joint to resist chord indentation. There is also a lot of reserve strength from the
ductility. Through determining the amount of joint design resistance against bending
moments, the trunnion can be designed against such failure modes, such that the full
effective shear capacity of the brace is mobilised.

Table 3.6 Summary of the ultimate and design strength for CT1 to CT5
Specimen

d0

t0

d1

t1

Fu,test
kN

V1
kN

M1
kNm

CT1
CT2
CT3
CT4
CT5

508.0
508.0
508.0
508.0
508.0

20.5
12.5
20.5
12.5
15.2

324.0
324.0
406.4
406.4
406.4

17.6
12.4
12.5
12.5
17.0

4830
2300
5420
2950
4580

3368
1879
2326
2326
3128

337
188
233
233
313

(All other units in mm)
Mu
Failure
kNm
330
132
539
217
393

chord

chord
shear
chord
chord

109


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Thus it is concluded that the existing design formulation under predicts the static yield
strength of the pipe trunnion by a wide margin. Further, it is observed that the high
reserve strength enjoyed by the pipe trunnion makes the configuration suitable as a
lifting point as the ductility is very high. Due to this under estimate of the static yield
strength of the pipe trunnion, there is an opportunity to use this as a basis to further
optimise the basis of recommendations later.

This phenomenon is further investigated in the numerical analysis where the tubular
X-joints were used to determine the design load where the level of chord plastification
occurs so that indentation of the chord will not reduce the shear capacity.

3.4.2

Design strength of through pipe trunnions

Since chord wall thickness is an important component in the design of pipe trunnions,
there is potential benefit when the brace, instead of being attached to the outside of
the chord, is slotted through the chord wall and welded, similar to the configuration
when the shear plate is slotted through the chord. The main aim of fabricating a pipe
trunnion with through pipe is to prevent chord indentation and these are suitable for

situations where the chord wall is thin and no replacement can be found. Since the
configuration is similar to shear plate pipe trunnions, this gives the designer greater
confidence in utilising the full shear capacity of the pipe, even though this may not be
necessary considering that a well-designed pipe trunnion can easily handle this task.
This method of pipe trunnion design is investigated here and tested experimentally to
determine whether there is any advantage in designing through pipe trunnions.

110


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

The test results for specimens CT6 and CT7 are given in Figure 3.38, which shows
the various observed behaviour including first yield (indication of flaking of white
wash on chord or brace), initial crack, ultimate load reached and the fracture point.
The first yield has been observed through the white wash flaking, and this should not
be misconstrued as the elastic yield strength. However, it serves as an important
indication of where the failure mode of the pipe trunnion is likely to.

5000

6000

Ultimate Load
@5160kN

Ultimate Load
@4140kN
5000

Initial Crack
@3560kN

3000

Initial Crack
@4520kN

4000

Fracture
@3510kN

Fracture
@4830kN

Load (kN)

Load (kN)

4000

3000

2000

1000

First Yield (Brace )
@1940kN

2000

First Yield (Brace )
@1540kN

1000
Specimen CT7

Specimen CT6
0

0
0

10

20
30
40
Displacement (mm)

50

0

20
40
Displacement (mm)

60

Figure 3.38 Design load for specimens CT6 and CT7

The first yield occurred on the brace surface for specimen CT6 and CT7 and these
specimens ultimately failed through shear fracture. Thus once the initial deformation
starts, the load deformation generally follows the weakest link and continues until
ultimate failure is reached.

111


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

Figure 3.39 is a comparison of the load deformation plots for CT2 & CT6 and CT4 &
CT7. It is noted from the earlier discussions that specimens CT2 and CT4 both fail
through chord indentation, that is the full effective shear area of the braces were not
fully mobilised before chord plastification sets in. This reduces considerably the joint
shear resistance of specimens CT2 and CT4. The figure shows the sharp contrast
between pipe trunnions that are designed with and without through pipe configuration.
Both specimens CT2 and CT4 have lower stiffness and the ultimate loads reached
were much lower than that of specimens CT6 and CT7 respectively although they
have identical dimensions. Thus it is possible for pipe trunnions, to have a lower shear
capacity due to the thin chord. These should be re-designed to carry much higher
shear capacity without changing any configuration, by merely inserting the brace
through the chord wall. The calculation for the design would be exactly the same as
that for normal pipe trunnions.

CT2

CT6

CT4

6000

CT7

6000

5000
4000
Load (kN)

Load (kN)

4000
3000
2000

2000

1000
0

0
0

20
40

Displacement (mm)

60

0

20
40
Displacement (mm)

60

Figure 3.39 Comparison load deformation plots for CT2 & CT6 and CT4 & CT7

112


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

This configuration is further investigated in the numerical analysis where a series of
through pipe were used to determine the applicable range of parameters that are
useable and effective.

3.4.3 Design strength of combined shear plate and pipe trunnions
Combining both the effective shear area of pipe and shear plate based on the
formulation as discussed above, the load deformation curves are plotted in Figure
3.40. The table presents the values calculated using Equation 3.4 to 3.6.

V1 = S dm1 t1 (0.4 fy1)

(3.4)

Vs = 2 ds ts (0.4fys)

(3.5)

V1 + Vs = 0.4 (2 ds ts fys + S dm1 t1 fys)

(3.6)

The plot shows the various observed behaviour including first yield (indication of
flaking of white wash on chord or brace), initial crack, ultimate load reached and the
fracture point. The first yield has been observed through the white wash flaking,
which should not be misconstrued as the elastic yield strength. However, it serves as
an important indication of where the failure mode of the pipe trunnion would be. The
first yield occurred on the chord surface of specimens CT8 and CT10 while in the
case of specimens CT9 and CT11 it occurred on the brace. Table 3.7 provides .a
summary of the design strength for CT8 to CT11.

113


Experimental Investigation of Large Pipe Trunnions and Tubular X-Joints

7000

3500
Ultimate Load
@5510kN

Load (kN)

5000

3000
2500

Initial Crack
@5060kN

4000

Fracture
@5100kN

Load (kN)

6000

Ultimate Load
@2870kN

3000
2000

Initial Crack
@2610kN

1500
First Yield (Brace )

@1025kN

1000
First Yield (Chord )
@1575kN

1000

Fracture
@2400kN

2000

500
Specimen CT8

Specimen CT9

0

0
0

9000

20
40
Displacement (mm)

0

4000

Ultimate Load
@7600kN

8000
7000

Initial Crack
@6320kN

5000
4000

20
30
40
Displacement (mm)

50

Initial Crack
@3200kN
Fracture
@2860kN

2000
First Yield (Brace )
@1485kN

First Yield (Chord )
@3300kN

3000

10

Ultimate Load
@3500kN

3000

Fracture
@6980kN
Load (kN)

6000
Load (kN)

60

1000

2000
1000

Specimen CT11

Specimen CT10
0

0
0

20
40
Displacement (mm)

60

0

20
40
Displacement (mm)

60

Figure 3.40 Load deformation curves for specimens CT8 to CT11

114