See discussions, stats, and author profiles for this publication at: [Link]

net/publication/289244766

Design and Analysis of Bellows for High Temperature Application as Per RCC-
MR

Article in Transactions of the Indian Institute of Metals · January 2016

DOI: 10.1007/s12666-015-0813-4

CITATIONS READS

9 1,113

3 authors, including:

s. C. S. P. Kumar Krovvidi Goutham Padmakumar

Indira Gandhi Centre for Atomic Research T.K.M. College of Engineering
57 PUBLICATIONS 148 CITATIONS 65 PUBLICATIONS 552 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by s. C. S. P. Kumar Krovvidi on 01 February 2021.

The user has requested enhancement of the downloaded file.

Trans Indian Inst Met (2016) 69(2):531–535
DOI 10.1007/s12666-015-0813-4

TECHNICAL PAPER

Design and Analysis of Bellows for High Temperature Application

as Per RCC-MR
S. C. S. P. Kumar Krovvidi1 • G. Padmakumar1 • A. K. Bhaduri2

Received: 16 October 2015 / Accepted: 9 December 2015 / Published online: 4 January 2016
Ó The Indian Institute of Metals - IIM 2016

Abstract Bellows are specially engineered components. 1 Introduction

Bellows find wide range of applications in FBR systems
such as in bellow-sealed valves and in piping systems for Bellows are specially engineered components. Bellows find
flexibility. Typical operating temperatures in FBR systems wide range of applications in FBR systems such as in
are around 550 °C, where failure modes such as creep and bellow-sealed valves and in piping systems for flexibility.
creep-fatigue interaction are significant. Conventional Typical operating temperatures in FBR systems are around
design codes for design of bellows such as standards of 550 °C, where failure modes such as creep and creep-fa-
EJMA do not extensively address the high temperature tigue interaction are significant. Conventional design codes
failure modes. In this present work, methodology adapted for design of bellows such as standards of Expansion joint
for design of bellows for high temperature application, manufactures association (EJMA) do not extensively
operating at 550 °C has been presented. The preliminary address the high temperature failure modes. A methodol-
design of the bellows for temperature below the creep ogy is suggested in standards of EJMA for designing the
range was completed by standards of EJMA, so that the bellows operating at high temperature, which is based on
failure modes such as fatigue, in-plane and column insta- extensive testing. In this method, minimum four bellows
bility of bellows can be addressed. Subsequently, the pre- need to be tested at high temperature till failure to validate
liminary design was analyzed using FEA as per the the design and hence it is very expensive. In this present
guidelines given by RCC-MR to address the failure modes work, methodology adapted for design of bellows for high
such as creep and creep-fatigue interaction. Elevated tem- temperature application, operating at 550 °C has been
perature testing of the bellows showed close agreement presented. The preliminary design of the bellows for tem-
between the design of the bellows by the above mentioned peratures below the creep range was completed by stan-
method and the experimental results. dards of EJMA. Subsequently, the preliminary design was
analyzed using FEA as per the guidelines given by RCC-
Keywords Bellows  High temperature design  MR to address the failure modes such as creep and creep-
Creep fatigue interaction. Elevated temperature testing of the
bellows showed close agreement between the design of the
bellows by the above mentioned method and the experi-
mental results. This paper details the preliminary design,
analysis of the preliminary design and estimation of the
high temperature life of the bellows as per RCC-MR,
& S. C. S. P. Kumar Krovvidi considering failure modes fatigue, creep and creep fatigue
krovvidi@[Link]
interaction. This paper also gives the comparison between
1
Fast Reactor Technology Group, Indira Gandhi Centre for the cycle life predicted and experimental results at elevated
Atomic Research, Kalpakkam 603102, India temperature.
2
Metallurgy and Materials Group, Indira Gandhi Centre for
Atomic Research, Kalpakkam 603102, India

123
532 Trans Indian Inst Met (2016) 69(2):531–535

2 Preliminary Design The preliminary design arrived based on EJMA stan-

dards is analyzed using linear elastic FE analysis. An
Preliminary design of the bellows is carried out using axisymmetric shell element is used for modeling one half
standards of Expansion joint manufactures association of a convolution. Symmetric boundary condition is
(EJMA) [1]. Standards of EJMA are well accepted for imposed at one end and displacement is given at other end
bellows operating below the creep range. Besides of the geometry. The Von-Misses stress intensity due to
addressing the fatigue due to cycling of the bellows, pressure load alone is given in Fig. 1 and due to combined
standards of EJMA address the failures which are specific pressure and displacement loading is given in the Fig. 2.
to bellows such as squirm and root bulging (column and in- Maximum value of the Von-Mises stress intensity in the
plane instabilities). The standards of the EJMA are more bellows due pressure alone is 51 MPa. The same due to
familiar to the manufactures of the bellows. Thus the combined pressure and deflection loading is 345 MPa. The
selection of the standards of EJMA for preliminary design methodology given in RCC-MR [3] is adapted for calculating
is justified. the number of cycles at elevated temperature. From the above
The equations given in EJMA are made in the form of a analysis, the preliminary design of the bellows selected meet
program in MATLAB to carry out the parametric study. the high temperature design criteria such as number of cycles
The design specifications of bellows used at elevated at elevated temperature, creep for the given hold time and
temperature in integrated cold trap using standards of creep-fatigue interaction as per RCC-MR.
EJMA is given in following Table 1. From the above analysis, total strain range without
Based on parametric study using EJMA, the optimum considering the creep component is ðDeel þ Depl Þ ¼
design arrived to meet the conflicting requirements such as 0:201 %. Number of cycles at 475 °C at the strain range is
the higher pressure (3 bar) with the low spring rate (10 kg/ 200,000. Considering the creep component, the number of
mm) and no in-plane and column instabilities is given in cycles reduces to 1005. In order to carryout testing in
Table 2. The procedure for preliminary design was pre- shorter duration, the above analysis was repeated for
sented in Ref. [2]. deflection of 30 mm and temperature of 600 °C. Stress
range corresponding to 30 mm movement is 490 MPa. A
hold time considered is 50 h. Total strain range computed
Table 1 Technical specifications of the bellows operating at elevated
temperature
is 0.52 %. The design number of cycles at elevated tem-
perature is 500. The procedure for finding the high tem-
S. no. Parameter Value perature life of the bellows as per RCC-MR is given in the
i. Type of the bellows Formed following paragraph and the calculated parameters are
bellows given Table 3.
ii. Material of construction SS 316L
iii. Design temperature 550 °C
3 Analytical Method for Finding Out High
iv. Design pressure 3 bar
Temperature Life of Bellows as Per RCC-MR
v. Maximum movement in service 30 mm axial
vi. Bellows free length 150 mm
The sequence of parameters to be computed for estimating
vii. Inside diameter 36
high temperature fatigue life, fatigue damage fraction,
viii. Axial spring rate required 100 N/mm
creep damage fraction and creep-fatigue interaction as per
ix. Hold time in hours 1500
RCC-MR is given below.
x. Number of cycles of operation during its life 500
1. Find stress range Dr from linear elastic analysis and
find De1 using De1 ¼ 23 ð1 þ tÞ Dr
E
2. Compute De2 , the plastic increase in the strain range
Table 2 Preliminary design of the bellows using EJMA
due to primary stress range. It is negligible unless
S. no. Parameter Value elastic follow-up is present.
i. Inside diameter of the bellows 40 mm
3. Compute De3 , Plastic increase in strain due to strain
ii. Depth of convolution (w) 6 mm
redistribution as per Neuber’s rule. It is computed by
the formula, De3 ¼ ðKe  1ÞDe1 . Ke is computed from
iii. Pitch (q) 6 mm
the graph (Fig. 3) given in RB-3261.123, RCC-MR.
iv. Ply thickness (t) 0.1 mm
4. Compute De4 , Plastic increase in strain due to
v. Number of plies (n) 3
triaxiality. It is computed by the formula,
vi. Number of convolutions 14
De4 ¼ ðKm  1ÞDe1 . Kt is computed from the graph
vii. Number cycles as per EJMA 210,000
(Fig. 4) given in RB-3261.123, RCC-MR.

123
Trans Indian Inst Met (2016) 69(2):531–535 533

Fig. 1 Von Mises stress

intensity in the bellows due
pressure loading (rP)

Fig. 2 Von Mises stress

intensity in the bellows due
pressure and deflection loading
(DQ)

Table 3 The values of parameters computed for estimating life of the

bellows at the elevated temperature
Parameter From loading or computed

Dr (MPa) 345
De1 0.0015
De2 Negligible
De3 ¼ ðKe  1ÞDe1 =0.000315 (Ke ¼ 1:21)
De4 ¼ ðKm  1ÞDe1 0.000195 (Km ¼ 1:13)
ðDeel þ Depl Þ ¼ De1 þ De2 =0.00201 = 0.201 %
þDe3 þ De4
Symmetrization factor Ks 0.5
Fig. 3 Graph (RB 3261.123, RCC-MR) for finding Ke
rk (MPa) 172.5
9 :4537
Decreep (%) ð1:155  10  1000
[Link] Þ ¼ 0:29 5. Compute the elastic and plastic strain
DeðTotal strain rangeÞ ¼ =0.49 %
ðDeel þ Depl Þ þ Decreep
ðDeel þ Depl Þ ¼ De1 þ De2 þ De3 þ De4
Dr
Allowable number of cycles (N) 1005
6. Find R ¼ 2r y
and Symmetrization factor Ks(R) from
Allowable time (T) for creep rupture for (rk =0:9) 100,000 h A3.46, RCC-MR as shown in the following Fig. 5.
in hours 7. Find the creep stress rk and creep strain Decreep
Maximum fatigue damage fraction: VA = (n/N) 500/1000 = 0.5 8. Compute total strain range equal to
Creep damage fraction: WA = t/T 1000/100,000 = 0.01 De ¼ ðDeel þ Depl Þ þ Decreep

123
534 Trans Indian Inst Met (2016) 69(2):531–535

Fig. 4 Graph (RB 3261.123, RCC-MR) for finding Kt

9. Obtain allowable number of cycles from design curve

of De versus N
10. Compute fatigue damage fraction VA and creep damage
fraction WA and check for their interaction as per the
creep fatigue interaction diagram as shown in Fig. 6.

Fig. 6 Creep fatigue interaction diagram (for austenitic stainless

4 Experimental Validation steels)

High temperature testing of the bellows was done at M/s

Metallic bellows India Pvt. Ltd., Chennai and the test setup
is shown in Fig. 7. To validate the analysis, accelerated
testing was performed as described above at 30 mm
movement and 600 °C. The testing chamber was maintained
with inert atmosphere with high purity argon to avoid oxi-
dation. The bellows failed at 720 cycles (drop in pressure
was noticed). After 200 cycles and 400 cycles, the bellows
were compressed by 15 mm movement and held for 24 h in
each step. Further 320 cycles were tested continuously.

5 Discussion

The preliminary design carried out for bellows is to address

the requirements such as the squirm, in-plane stability. The
preliminary design was analyzed as per linear elastic FEA Fig. 7 High temperature cycle life testing of bellows
and the following observations are noted.

Fig. 5 Symmetrization factor

for finding creep stress

123
Trans Indian Inst Met (2016) 69(2):531–535 535

1. Considering only the elastic and plastic components of EJMA and by analysis as per RCC-MR, when the creep
strains, i.e., without considering the creep strain, the component is excluded. This validates mutually the pre-
fatigue life of the bellows is 200,000. The fatigue life liminary design and the FE analysis for temperatures below
estimated for the bellows as per standards of EJMA is the creep range of the material. The design of the bellows
210,000. Standards of EJMA do not address creep and for high temperature life as per RCC-MR is conservative
the number of cycles computed based on analysis and in close agreement with the experimental results.
excluding creep and the fatigue life computed as per
standards of EJMA are in close agreement. Acknowledgments The author acknowledges M/s Metallic bellows
India Pvt. Ltd., Chennai for the valuable support given by them during
2. The elevated temperature life estimated as per RCC-
testing of the bellows.
MR is conservative and in close agreement with the
experimental results.
References

1. Standards of Expansion Joint Manufacturer’s Association, 9th

6 Conclusions edition (2009).
2. S. C. S. P. Kumar Krovvidi, S. Sathishkumar, M. G. Hemnath, V.
This paper presents the design of the bellows operating at Vinod, B. K. Sreedhar,G. Padmakumar, and K. K. Rajan, Design of
elevated temperatures, where the failure modes such as Bellows for High Temperature Sodium Application in Integrated
Cold Trap, OPE-2013, Mamallapuram (2013).
creep are significant. The fatigue life expectancy of bel- 3. Design and construction Rules for Mechanical Components of
lows decreases considerably when the operating tempera- Nuclear Installations, RCC-MR Section-I, Subsection-B and
ture is in creep range of the material. There is a close Appendix A3.
agreement between the cycle life predicted by standards of

123

View publication stats