Fatigue Analysis of Rotary Cement Kiln Welded using FEM J. Teknologi 2016

Jurnal
Teknologi

Full Paper

Article history
Received
1 Fe b rua ry 2016
Received in revised form
24 Ma rc h 2016
Accepted
1 Aug ust 2016

FATIGUE ANALYSIS OF ROTARY CEMENT KILN
WELDED USING FINITE ELEMENT METHOD
Hasan Basri a*, Irsyadi Yania
aFaculty

of Engineering, University of Sriwijaya, Indonesia

*Corresponding author

hasanbas1960@gmail.com
Graphical abstract

Abstract
In this research, the implementation of Shell of Kiln problem has been discussed. The results
obtained in this research are taken based on the simulation and experimental work.
This research discusses only the simulation work. The simulation works are performed to
evaluate the fatigue of the shell of a kiln. To validate the characteristic of fatigue
problem simulation, the mechanical and thermal load are implemented on the shell of
a kiln. The results are analyzed in detail in this research for fatigue life for the shell of a kiln.
In this work, the shell of the kiln has been modeling by solid works and Fast2014. All the
boundary conditions (mechanical load, thermal load, Young’s modulus, density, etc.) must
be evaluated such as the actual condition. This simulation showed how the most relevant
aspects of the developed work presented in this report can contribute to the state-of-theart of the analysis fatigue life of Rotary Cement Kiln technique with innovative ideas and
strategies. It also reviews if the obtained results achieve the proposed objectives. The main
goal of the work presented in this research is to propose fatigue life analysis algorithm in
the shell of a kiln using finite element analysis. Due to the dimensionality of the problem
addressed, the research specification has to set limits to the applicability of the research by
selecting only mechanical load problems in rotary cement kiln tasks; and goal-seeking, to
predict the fatigue life simulation investigated. From the simulation, model and boundary

conditions are defined. Crack growth behaviour in the rotary kiln was predicted. As the
crack grows, the speed of crack depth increase.
Ke ywo rds: FEA, fatigue analysis, rotary kiln, crack simulation

Abstrak
Dalam kajian ini, permasalahan Kiln shell telah dibincangkan. Keputusan yang diperolehi
dalam kajian ini diambil berdasarkan simulasi dan kerja uji kaji. Kajian ini membincangkan
simulasi sahaja. Simulasi dilaksanakan untuk menilai keletihan shell. Untuk mengesahkan
ciri-ciri simulasi masalah keletihan, beban mekanikal dan haba dilaksanakan pada shell.
Umur kelelahan dianalisis secara terperinci dalam kajian ini untuk shell. Dalam kajian ini,
shell dibangun menggunakan software Solid works dan Fast2014. Semua beban mekanikal,
beban haba, modulus Young, ketumpatan, dan lain-lain perlu dinilai seperti keadaan
sebenar. Simulasi ini menunjukkan bagaimana aspek yang paling berkaitan yang
dibangunkan dibentangkan dalam laporan ini boleh menyumbang kepada statei-of-theart analisis keletihan teknik Cement Kiln Rotary dengan idea-idea dan strategi inovatif. Ia
juga mengkaji jika keputusan yang diperolehi mencapai objektif yang dicadangkan.
Matlamat utama kerja yang dibentangkan dalam kajian ini adalah untuk mencadangkan
keletihan algoritma analisis kehidupan di shell tanur menggunakan analisis unsur terhingga.
Dari simulasi, model dan keadaan, tingkah laku pertumbuhan retak dalam shell telah
diramalkan. Sebagai retak tumbuh, kelajuan peningkatan kedalaman retak bertambah.
Kata kunc i: FEA, fatigue analysis, rotary kiln, crack simulation.

© 2016 Penerbit UTM Press. All rights reserved

xx:x (201x) xxx–xxx | www.jurnalteknologi.utm.my | eISSN 2180–3722 |

2

Ha sa n Ba sri & Irsy a d i Ya ni / Jurna l Te kno lo g i (Sc ie nc e s & Eng ine e ring ) xx: x (201 x) xxx– xxx

1.0 INTRODUCTION
A kiln is basically an industrial oven, and although
the term is generic, several quite distinctive designs
have been used over the years, such a kiln in PT.
Semen Baturaja made about 1,200.000 tonnes of
clinker per year. The Company was first established
under the name of PT. Semen Baturaja (Persero) on
November 14, 1974. The Company specifically
engages in clinker production business with its
production center located in Baturaja, South
Sumatera, while the cement grinding and
packing process are operated in Baturaja Plant,

Palembang Plant and Panjang Plant to be distributed
to the Company’s marketing areas. The Company, in
its effort for business development, endeavors to
improve the existing equipment in order to achieve
the target of installed capacity amounting to 50,000
tons cement per year and to improve its
installed capacity.
Rotary kiln shell is a large-scale welded structure with
4.5 m in diameter and 75 m in length and produced by
welding thin cylindrical steel plate one by one. The
shell of the kiln is made of mild steel plate. Mild steel is
the only viable material for the purpose, but presents
the problem that the maximum temperature of the
feed inside the kiln is over 1400°C, while the gas
temperatures reach 1900°C. The melting point of mild
steel is around 1300°C, and it starts to weaken at
480°C, so considerable effort is required to protect the
shell from overheating (FLSmidth).
Padded plates are directly soldered to the shell in
the supporting rollers places to reduce their

concentrated stress. Crack are often initiated at these
welded joints, and the over long circumferential
crack are prevailing at welded joints near the
supporting rollers. However, Kikuchi et.al. (2010) has
predicted of two interacting surface cracks of
dissimilar sizes by finite element analysis. The
simulations were performed for fatigue crack growth
experiments and the method validity was shown on
this research. It was shown that the offset distance
and the relative size were both important parameters
to determine the interaction between two surfaces of
crack; the smaller crack stopped growing when the
difference in size was large. It was possible to judge
whether the effect of interaction should be considered
based on the correlation between the relative spacing
and relative size. In 2014, Fatigue crack growth
simulation in a heterogeneous material using finite
element method has generated by Kikuchi et.al.
Kikuchi has developed a fully automatic fatigue crack
growth simulation system using FEM and applied it to

three-dimensional surface crack problems, in order to
evaluate the interaction of multiple surface cracks,
and the crack closure effects of surface cracks. The
system is modified to manage residual stress field
problems, and the stress corrosion cracking process
is simulated.

To prediction of crack propagation under thermal,
residual stress fields using S-Version FEM (S-FEM), Kikuchi
was employed to solve a crack growth problem by
combining with the auto-meshing technique, this remeshing process of the local mesh becomes very
simple, and modeling of three-dimensional crack
shape becomes computationally easy. On the other
hand, in 2004, Irsyadi has developed visualization of
finite element analysis in 3D (C, C++, under
Linux/Fedora), with this system, analysis for extra
large problems such as fatigue life predictions
becomes easy and fast. Irsyadi, Kikuchi, and Kanto
employed numerical analysis of 3-D Surface Crack in
2006, and then, Irsyadi and Kikuchi was developed a

numerical analysis in the low carbon steel by finite
element method and experimental method under
fatigue loading. In this research, they were
predicted fatigue live of material under stresses.
Fatigue, or metal fatigue, is the failure of a
component as a result of cyclic stress. The failure
occurs in three phases: crack initiation, crack
propagation, and catastrophic overload failure. The
duration of each of these three phases depends on
many factors including fundamental raw material
characteristics, magnitude and orientation of applied
stresses, processing history, etc. Fatigue failures often
result from applied stress levels significantly below
those necessary to cause static failure.
The prediction of fatigue life of rotary cement kiln
welded shell is not completely understood and
fascinating, therefore it should be investigated. For
rotary kiln shell, cracks can grow with a complex
overloading conditions for over thousands of tons, and
then results in premature shell failure. The affecting

conditions
crack
growth
include
material
characteristics, initial crack size, service stresses, and
stress concentration due to overheated in hot spot
area, all these conditions are random. The fatigue life
of the welded shell during crack growth need to be
predicted numerically by using finite element analysis
and experimentally. T he fatigue life analysis attempts
to provide the answers to general questions, which
remain unclear on the evolution of the crack growth
and its impacts on failure in rotary cement kiln system.
In addition, the research is intended to bring our
knowledge of the simulation and experimental of the
fatigue life analysis in the welded joints of a rotary
cement kiln in PT. Semen Baturaja. The primary
objective of this study is to investigate the fatigue life
of the crack growth analysis in the welded joints of a

rotary cement kiln in PT. Semen Baturaja. We
approach the goal through a combined analysis of
fatigue growth observational data and numerical
model simulation.

2.0 METHODOLOGY
Some materials, such as steel, show an endurance
limit stress below which the fatigue life is essentially
infinite. Other materials may not show such behavior

3

Hasan Basri & Irsy adi Yani / Jurnal Te kno lo g i (Sc ie nc e s & Eng ine e ring ) xx: x (201 x) xxx– xxx

but an effective endurance limit may be specified at
some large number of cycles. the important
parameters to characterize a given cyclic loading
history are :

Stress range : σ = σmax - σmin

Stress amplitude: σa = 0.5 (σmax - σmin)
Mean stress : σm = 0.5 (σmax + σmin)
Load ratio : R = σmax/ σmin

The total fatigue life of a component can be
considered to have two parts, the initiatio n life and the
p ro p a g a tio n life . The stress-life approach just
described is applicable for situations involving
primarily elastic deformation. Under these conditions
the component is expected to have a long lifetime.
For
situations
involving
high
stresses,
high
temperatures, or stress concentrations such as
notches, where significant plasticity can be involved,
the approach is not appropriate.
The main geometrical characteristics of the rotary

kiln shell are shown in Table 1.
Table 1: The main geometrical characteristics of the
rotary kiln
Magnitude

Value

Cold real length
Inner diameter
Number of tires
Slope (in direction to outlet)
Maximum speed

75
4.5
3
3.5
3.5

Units
meters
meters
-%
rpm

The thicknesses of the shells along the different
sections of the rotary kiln are given in Table 2 and
Figure 1. In Table 2 zero is placed in the upper end
of the rotary kiln, called ‘inlet-I’. The distances
between supports, in millimeters, are given in Table
3 where ‘III-outlet’ denotes the lower end of the rotary
kiln.

Figure 1. Diameter, length, and thickness variation of
rotary kiln
Table 3: Distances between supports
Supports

Distance (mm)

Inlet – I
I – Girth Gear
I – II
II – III
III – Outlet

13,000
4,200
31,000
27,000
4,000

Tables 2, 3, Figure 1 and 2 is the rotary kiln (1) along
with the structural elements to rotate the kilns (3), (10),
(7), and (11) around its longitudinal axis. The kiln (1)
includes an elongated, cylindrical, rotating shell (2)
which has a feed end (8), an opposite discharge end
(5). The kiln (1) is erected so that the discharge end
(5) is at a lower level then the feed end (8) in order
to cause the material being processed. It travels
through the open processing zone to the discharge
end (5). The kiln shell (2) is supported by riding rings
or tires (3) that engage steel rollers (10) which are
supported on concrete piers (4) and steel frames (6).

Table 2: Thicknesses of the shells along the different
sections of the rotary kiln
Section
(mm)

Thickness
(mm)

Section
(mm)

Thickness
(mm)

0 – 1,800
1,800 – 2,500
2,500 – 5,500
5,500 – 8,100
8,100 – 9,700
9,700 – 9,900
9,900 – 11,855
11,855 – 14,190
14,190 – 16,525
16,525 – 18,860
18,860 – 21,195
21,195 – 23,530
23,530 – 25,865
25,865 – 28,200
28,200 – 30,000
30,000 – 32,200
32,200 – 34,000
34,000 – 35,260

60
60
90
70
40
28
28
28
28
28
28
28
28
28
40
60
40
28

35,260 – 37,595
37,595 – 39,930
39,930 – 42,265
42,265 – 44,100
44,100 – 46,435
46,435 – 48,770
48,770 – 51,105
51,105 – 53,440
53,440 – 55,400
55,400 – 56,900
56,900 – 59,400
59,400 – 61,000
61,000 – 63,200
63,200 – 64,800
64,800 – 68,100
68,100 – 71,400
71,400 – 73,650
73,650 – 75,000

28
28
28
28
28
28
28
28
28
40
40
40
60
40
25
25
25
25

Figure 2. Rotary kiln configuration
Materials used for the kiln are shown in Table 4. These
materials are used to build the kiln components. The
shell, tire, roller and pinion are the main components
in the kiln. However, the material has been modeled
as isotropic and linear, elastic temperature
dependent, according to the elastic properties of the
steel used in Table 4. The kiln shell is welded
circumferentially and longitudinally.

Ha sa n Ba sri & Irsy a d i Ya ni / Jurna l Te kno lo g i (Sc ie nc e s & Eng ine e ring ) xx: x (201 x) xxx– xxx

4

Table 4: Materials are used for the kiln
Component
Shells
Tires
Rollers
Pinion

Material
ASTM A.36
Low alloy steel casting
Low alloy steel casting
30 Cr Ni Mo 8 (ISO R 638 =
II-68 Type 3)

As mentioned earlier there are mainly four
methods to predict fatigue of welded components:






Nominal Stress
Structural Stress
Effective Notch Stress
Linear Elastic Fracture Mechanics (LEFM)

The effects of welding residual stresses, R-ratio, wall
thickness and improvement techniques are included
in this research. In case of variable amplitude loading,
Palmgren- Miner´s linear damage rule is used when
the design methods nominal, structural and effective
notch stress are applied.

Figure 3. Model of the kiln shell

3.0 RESULTS AND DISCUSSION
The results obtained in this research are taken based
on the simulation and experimental work. This chapter
discusses only the simulation work. The simulation
works are performed to evaluate the fatigue of the
kiln shell. To validate the characteristic of fatigue
problem simulation, the mechanical and thermal
load are implemented on the shell of kiln. The results
are analyzed in detail in this chapter for fatigue life for
the kiln shell.
In this work, the kiln shell has been modeling by
Autodesk Inventor 2014. All the boundary conditions
(mechanical load, thermal load, Young’s modulus,
density, etc.) must be evaluated such as the actual
condition. The simulation of the kiln shell is only
conducted to determine the fatigue life of the kiln
shell.
The development of modeling is the initial work
that marks the natural condition of the designed
software into the simulation platform. In this phase all
algorithms of simulation condition are implemented
using a standard Finite Element Analysis. Within this
development model, the architecture is validated
and the main functions and algorithms are tested.
During the design, by using static analysis which has
been implemented in the kiln shell. The model of the
kiln shell, boundary conditions and initial crack as
shown in Figure 3 and 4 are used in order to achieve
robust performance, high reliability and minimal
computational cost.

(a)

(b)

(c)

(d)

(e)

(f)

Figure 4. Boundary Conditions (BC) and Initial Crack
Considering the cracked elastic plate shown on
the following page, with Young’s modulus E and
Poisson’s ratio ν. The height is small compared to the
dimensions a and c , and the thickness is small enough
that p la ne stre ss conditions. If the crack grows all the
way across the plate and the plate separates into two
pieces, the system will have zero strain energy. So the
strain energy decreases as the crack grows.
Calculating the change in strain energy of the plate if
the crack grows by some amount dr. Using the same
assumptions used previously to calculate the total
strain energy.
Considering of a finite element analysis,
characteristics of a kiln shell are a pretension and a
mating part contact (Figure 4.a). The pretension can

5

Hasan Basri & Irsy adi Yani / Jurnal Te kno lo g i (Sc ie nc e s & Eng ine e ring ) xx: x (201 x) xxx– xxx

generally be modeled with static loading, thermal
deformation, a constraint equation, or an initial strain.
For a thermal deformation method, the pretension is
generated by assigning virtual different temperatures
and thermal expansion coefficients to the shell, clinker
and temperature of gases. In this work, in order to
generate a finite element model for the kiln shell with
clinker and temperature of gases (Figure 4.b and 4.c),
two kinds of models are introduced. All the proposed
models are taken into account above primary
characteristics such as a pretension effect and a
contact behavior between shell, clinker and
temperature of gases (see Figure 4.d). The prediction
of the crack propagation are considered on crack
emanating from contact surface between kiln shell
and clinker and welded joint on the kiln shell surface
(Figure 4.e). On this Finite Element model, the final
mesh consisted of 40.856 elements and 45.245 nodes.
The elements along the direction of crack advance
had a length of 28 mm. Following the stress and
deformation simulation, fatigue crack growth was
modeled by repeated loading (mass of clinker and
thermal gases), unloading, advancing the crack and
then the loading again.
Under static analysis, stress and deformation in the
kiln is obtained as shown i n F i g u r e 5 . From the
simulation results, it is found that t h e critical area is
occurred in the kiln. B y u s i n g the static value,
subsequent fatigue analysis can be performed.

Figure 5. Stress, displacement and strain of the static
analysis
Since local stresses were far in excess of the yield
strength of the materials at the stitch welds, elasticplastic analyses were used to obtain the strain ranges
needed to perform a low cycle fatigue evaluation.
The evaluation was made using fatigue design
curves obtained by applying a factor of cycles to
the theoretical failure curves. Based on the design
curves, a cumulative fatigue usage had been
reached when failure occurred. This is consistent with
the knowledge that cracks had initiated and grown to
a critical size and propagated as fast fractures at this
usage as shown in Figure 6. Actual failure occurred
between the design fatigue curve and the theoretical
mean failure data for small polished laboratory test
specimens. Failure below the mean laboratory failure
curve is expected due to size effects, surface finish
effects, environmental effects and scatter in the data.

The calculated loads and stresses were used to
perform a low cycle fatigue evaluation b y using the
actual operating data. The operating temperatures
were lower than the temperatures anticipated in the
design specifications. There were unknown cycles
corresponding to the anticipated 450°C gas inlet
temperature upset condition, and only three cycles to
the normal operating temperature. Hence, the actual
operating cycles were used to evaluate the fatigue
damage.

(a)

(b)
Figure 6. Crack growth and Transition of stress
intensity factor

6

Ha sa n Ba sri & Irsy a d i Ya ni / Jurna l Te kno lo g i (Sc ie nc e s & Eng ine e ring ) xx: x (201 x) xxx– xxx

[4]

The results of the structural analyses of the original
design showed that the flexibility of the weight stresses
and reduced thermal bending stresses in the duct.
Hence, bending of the duct imposed high cyclic
loads and stresses on the support truss, buckling several
truss members. With properly designed duct supports.
Bending of the duct does not bend the supporting
truss. Moreover, the flexibility of the truss would have
no effect on dead weight stresses in the duct. This is
important since the typical design sequence involves
first designing the duct, and then using the resulting
weight to design the truss.

[5]

[6]

[7]
[8]
[9]

[10]

4.0 CONCLUSION
This research shows how the most relevant aspects of
the developed work presented in this report can
contribute to the state-of-the-art of the analysis
fatigue life of rotary cement kiln technique with
innovative ideas and strategies. It also reviews if the
obtained results achieve the proposed objectives.
The main goal of the work presented in this
research is to propose fatigue life analysis algorithm in
the shell of a kiln using finite element analysis. Due
to the dimensionality of the problem addressed, the
research specification has to set limits to the
applicability of the research by selecting only
mechanical load problems in rotary cement kiln
tasks and goal-seeking, to predict the fatigue life
simulation investigated.
From the simulation, model and boundary
conditions are defined. Crack growth behaviour in
rotary kiln was predicted. As the crack grows, the
speed of the crack depth increase.

[11]

[12]

[13]

[14]

[15]

[16]

[17]

[18]

References
[1]
[2]
[3]

Energy Efficiency Component EU-China. A reference book
for the industry. Beijing, 2009.
Cement kiln, http://en.wikipedia.org/wiki/Talk, 2010.
Coz Díaz JJ, Mazón FR, García Nieto PJ, Suárez Domínguez
FJ., Design and finite element analysis of a wet cycle
cement rotary kiln. Finite Elements in Analysis and Design;
39(1):17–42, 2002.

[19]

ABP, ASME boiler and pressure vessel code VIII division 2 alternative rules. The American Society of Mechanical
Engineers, 1995.
ABP, ASME boiler and pressure vessel code II - material
properties. The American Society of Mechanical Engineers,
1995.
ABP, ASME boiler and pressure vessel code VIII division 2 alternative rules, Appendix 5. Design based on fatigue
analysis, 1995.
Merk Blatter A., Technical rules for pressure vessels. Verlag
KG, Carls Heymanns, 1997.
Frederic O., The kiln book; materials, specifications and
construction. Chilton Book Company, 1997.
Liu XY., Specht E., Temperature distribution within the
moving bed of rotary kilns: measurement and analysis.
Chemical
Engineering
and
Processing.
Process
Intensification 2010;49(2):147–50.
Kikuchi M., Fatigue crack growth simulation in
heterogeneous material using S-version FEM, International
Journal of Fatigue 58, 2014.
Kikuchi M., Growth prediction of two interacting surface
cracks of dissimilar sizes, Engineering Fracture Mechanics
77, 3120–3131, 2010.
Kikuchi M., Crack growth analysis in a weld-heat-affected
zone using S-version FEM, International Journal of Pressure
Vessels and Piping 90-91, 2012.
Irsyadi Yani, Visualization of Finite Element Analysis in 3D (C,
C++, Under Linux/Fedora), Seminar PPI Jepang Tengah,
Toyohashi University of Technology, Japan.
Irsyadi Yani, Masanori Kikuchi, “Numerical Analysis in the
Low Carbon Steel by Finite Element and Experimental
Method under Fatigue Loading”, Proceeding of The 5th
International Conference on Numerical Analysis in
Engineering (NAE), 2007.
Irsyadi Yani, Masanori Kikuchi, Yasuhiro Kanto, ”Numerical
Analysis of 3-D Surface Crack”, International Conference of
Computational Mechanics and Numerical Analysis, 2006.
Irsyadi Yani “Analysis K-Value on the crack pipe under
Torsion” Jurnal Rekayasa Mesin, FT. Unsri Jurusan Teknik
Mesin.
Thamrin, I., Basri, H., “Exact and Numerical Solution of Pure
Torsion Shaft”, Proceeding of 5th International Conference
on Numerical Analysis in Engineering (NAE 2007). University
of Sumatera Utara (IC-STAR USU), ISSN: 979-96139-6-7,
pp.3.24-3.29, 2007.
Basri, H., “Numerical Solution of Fatigue Stress in Journal
Bearing”, Proceeding of 5th International Conference on
Numerical Analysis in Engineering (NAE 2007), University of
Sumatera Utara (IC-STAR USU), ISSN: 979-96139-6-7, pp. 8.18.7, 2007.
Chandra, H., Basri, H., “Analisis Fatigue Poros Pompa
Vakum“, (in Indonesian), Prosiding Seminar Nasional Sains
dan Teknologi 2007, Lembaga Penelitian, Universitas
Lampung, Bandar Lampung, 27-28 Agustus 2007.