International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp.

14498-14504
© Research India Publications. http://www.ripublication.com

Verification of Over-Speed and Burst Margin Limits Inaero Engine Rotor

Coupling Along with Estimation of Low Cycle Fatigue Life

Srinivas Murthy1, Shivarudraiah2

Vivekananda Institute of Technology, Bengaluru, Karnataka 560074, India.
1
2
University Visvesvaraya College of Engineering, Bengaluru, Karnataka 560001, India.

Abstract coupling is identified to show irregular large-amplitude

vibrations in axial direction when excited at natural frequency
Turbine of an aero engine is considered as a key knowledge for
of coupling.
development in overall engine performance. High performance
turbines with augmented operational temperature require Considering various loads acting upon the disc and variation of
sophisticated design concepts. Weight reduction, increased the loads with respect to time are the factors that add to the
strength and stiffness are the major requirements for highly complexity of turbine disc design. Weight of disc plays a vital
stressed turbine disc in forthcoming aircraft engines. A major role in improving efficiency of the gas turbine[1]. Hence,
challenge during the design process of a modern low weight allows component to operate under plastic zone assuring the
turbine disc is to find rotor disc geometries that meet both, static safety of the component with design limits. In present work an
and fatigue requirement for both mechanical and thermal loads. attempt is made to understand design criteria’s used for the
The important functions of flexible coupling in turbomachines design of gas turbine disc running at speed of 12000 RPM and
are, transmit mechanical power from one shaft to another with operating at a temperature of 500oC. The finite element
constant velocity, reimburse for mis-alignment without analyses were carried out to check the mechanical and
affecting structural integrity, without generating excessive structural integrity of the disc in a systematic order using the
thrust on either shaft for axial movement with minimum power commercially validated FE package software. This includes,
loss. Aero engine rotor burst evaluation is one of the most
important problems to be taken care off, whenever it comes to  A sensitivity study for material model and its
designing a turbo machinery disc. The consequences of a behaviour at in-service condition.
failure can be intense, since the disc fragments into multiple  3-D analysis for estimation of over-speed margin
pieces and they are hurled away in all the possible direction at evaluation with safety factor
high speeds. In present work evaluation of safety limits and  Estimation of over-speed margin in rotating aero rotor
low-cycle fatigue (LCF) life estimation of an aero engine flange disc as per international authorities for integrity,
coupled disc through classical methods. By blending the blending the classical approach with FEA
terminologies with simulation engineering to arrive at a  Application of 3-D elasto-plastic strain to
probable interpretation of number of duty cycles is carried out. conventional equations to arrive at fatigue life of disc;
The methodology compares the fatigue parameters involved in Coffin-Manson method
evaluation of disc life at off-design condition through
sensitivity analysis. The design tool closely connects the flight DESIGN CONSTRAINTS
certification regulating agencies for safety in air transportation
vehicles. The off-design speed regulations through API and The major loads acting on turbine blades are, centrifugal and
MIL handbook for material specification are considered to thermo-mechanical loads. Thermal load is dominated by
carry out finite element analysis. centrifugal loads for every cycle; Hence, importance is given
for mechanical loads. In a bladed disk assembly, the disk
Keywords: Flexible couplings, rotor burst, low-cycle fatigue, happens to be the stronger section compared to disk. Replacing
API & MIL handbook, finite element analysis blade under failure is cost effective than replacing disks.
During operating at design speed and over-speed
circumstances, the average stresses obtained at cross sectional
INTRODUCTION areas has to be well within allowable design limits as per design
The functional necessities and features of flexing cupling is to rule for both blade and disk. To sidestep all these complexities
transmit rated torque without undergoing buckling, permanent an integrated bladed rotor coupled assembly is considered for
deformation i.e. to possess with high torsional rigidity. analysis. The constraints for present analysis is sequenced as
However, under conditions of misalignment, flexing coupling follows
element must have necessary flexibility to accommodate these 1. Design Parameters
situations without inducing excessive force and moment on
shoulder shaft, bearing and bolts. Misalignment is compensated 2. Behavior constraints
using laminated disc sets. Both of the requirements should be
achieved by maintaining stress levels which are safely in the
range of fatigue limit of flexing material. Metal-flexing

14498
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

 Design parameters MATERIAL MODEL

1. At the Bore, permissible hoop and equivalent stress is 95% The aero engine discs are subjected to high centrifugal and
of 0.2% proof stress. thermal loads Inconel is the material preferred for their
2. At the web, permissible hoop stress is 85% of 0.2% proof manufacturing because of corrosion resistant property. Hence,
stress for blade and disk material Inconel [2] is considered.
3. At the web, permissible radial and equivalent stress is 80%
of 0.2% proof stress.
4. At the rim, permissible hoop and equivalent stress is 67% Table 1: Material property of INCONEL718
of 0.2% proof stress
5. Axial and radial growth ≤ specified tolerance i.e., 1 mm Density 8190 kg/m3
Poison’s ratio 0.3

 Behaviour constraints Young’s modulus 209 GPa

1. The average section stress at cross section area of disk Tensile strength 1035 MPa
must not exceed allowable design limits
2. Average cross sectional stress should not cross the limit of Co-efficient of thermal 13.0*10-6 K-1
10% of blade root neck average stress expansion
3. Permissible AWMHS< 72% of 0.2% proof stress at peak
temperature in the disk. [3]
4. Permissible AWMHS< 64% of UTS at peak temperature BOLT-PRETENSION
in the disc. [3] Conventional method of designing a mechanical component is
5. Disc burst-speed ≥ 125% of maximum permissible steady used for the preliminary design considerations of the turbine
state speed for 12000 rpm. rotor.
6. Disc over-speed ≥118% of maximum permissible steady
state speed for 12000rpm.  Petersons chart is used to calculate the Kt effect at the
discontinuities which is finally used to decide on the initial
These constraints are not readily available in the program like
fillet radius to start with. [4]
ANSYS and these are the constraints which are used in aero
 Bolt pretention loads are calculated to prevent the
industry.
loosening of bolts during operation.
 No of bolt holes are decided based on the flange opening
between the discs and the torque getting transmitted
MATHEMATICAL MODEL
through the discs.
A cyclic symmetry sector of an integrated bladed disk assembly
is considered for present analysis. The model of three blades
with disk sector of 180 is modelled using the commercial Mathematical representation of blot pre-tension is given by
software. The FE model of bladed disk sector is as shown in Fi = C*As*Sp Eqn.1
Figure 1. For the present analysis, higher order elements are
considered to generate the finite element model. A matching
node pattern is maintained at the blade and disk faces, where LOAD CONSIDERATIONS
the load transfers between blade and disk takes place. The blade
and disk are considered to be made of similar material. 1. Rotational velocity Range
Rotational or angular velocity is defined as rate of change
of angular displacement and is a vector quantity which specifies
the rotational speed (angular speed) of an object and the axis
rotation.

2. Influence of the blades

The blades on the external ring of the disc causes an additional
traction load in the radial direction, since they behave like
concentrated masses under the effect of the centrifugal field.

Formulas for blade pressure:

Fc = m𝜔2rc Eqn.2
Figure 1: 3-D axis symmetry and mathematical model of
rotor coupling Fc∗Nb
PB = Eqn.3
𝜋Dt

14499
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

3. Thermal load From the figure 3 it is observed that maximum at mid span of
the bolt. The bolt pretention is provided in the bolts to increase
Loads due to change in temperature are produced by the non-
the contact stress in the flange retaining the tensile load parallel
uniform distribution of temperature in the disc under service
to the axis of the rotor. Due to centrifugal pull, each bolt is
conditions. The external region of the rotor that is closer to the
subjected to shear loading since each stage of the rotor tries to
hot gasses, in fact, is hotter than the internal area. This non-
radially grow outward resulting in combination of bolt bending
uniform temperature gradient causes a deformation of the
and shear. The first principal stress in bolt indicates the shear
material, along with the coefficient of thermal expansion.
loading dominance in the bolt and the equivalent stress
indicates the stress intensity due to bolt bending. The obtained
stresses are well with the design limits therefore the design is
Service conditions for disc: safe under these conditions
Turbine disc experiences centrifugal loads, thermal loads and
blade pressure under operating conditions. The loading
conditions of the disc for the present work are classified into
three cases: first case is by considering centrifugal force only,
second case is combination of centrifugal and blade pressure,
third case is by considering all three loads as mentioned above.
Figure 2 represents the schematic diagram of disc loading.

Figure 4: Variation of deformation in blot

The above figure 4 shows the total deformation obtained for the
cyclic symmetry analysis for the given Boundary condition.
The radial growth is 0.534 mm which is safer under allowable
growth.

Figure 2: Features of the disc

RESULT AND DISCUSSIONS

Figure 5: Variation of stress with respect to the length of the

disc

From figure 5 it is evident that hoop stress (maximum at the

disc bore) induced in the disc sector decrease as the length of
disc increase from bore to rim. The radial stress (minimum at
Figure 3: Variation of maximum stress in blot the bore and maximum in the web) increase in the bore region
and decreases in web region as the length increases. Since, the
hoop stress is concentrated at disc bore and plays a major role

14500
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

in disc failure. Hoop stress considered for evaluating the safety evaluating the safety margins. When passing from engineering
margins. The deformation induced in the turbine disc is less stress to true stress, the following relations must be applied:
than ‘1 mm’ which is desired result.
𝑡𝑟𝑢𝑒 = ln(1 + 𝜀𝑒𝑛𝑔 ) Eqn.5

EVALUATION OF OVER-SPEED AND BURST

MARGINS 𝜎𝑡𝑟𝑢𝑒 = 𝜎𝑒𝑛𝑔 (1 + 𝜀𝑒𝑛𝑔 ) Eqn.6
Over-speed is a situation where the aero rotor disc is forced to
From Eqn. 4 and Eqn. 5 a relationship between true stress and
rotate after its design limit. The significances of running the
engineering stress was obtained. By utilizing these relations, a
disk speed differ from engine type, model etc. The duration of
graph was plotted between stress and strain which is
over-speed is main factor which speed of disc is dependent
represented in figure 6.
upon.
Generally, in aero engine discs, even a moment of over-speed
results in reducing the engine life or leads to catastrophic
failure. Speed at which disc undergoes catastrophic failure is
known as burst speed.
The constraints for rotating speed of aero engine turbine disc
given by the two international agencies is given below [EASA
CS-E 840 AND FAR 33.27] [7]
a.120% of maximum allowable rotor speeds related with any of
ratings of excluding OEI ratings < 2 ½ minutes.
b.115% of the maximum allowable rotor speeds associated with
any OEI ratings < 2 ½ minutes.
c.105% of the highest rotor speed that is outcome from the
failure of part or system which is the most crucial with regard
to over-speeding
d. The burst speed is a highest speed that result from failure of
Figure 6: comparison between engginerring stress and true
any component or system in a respective installation of engine.
stress [2]
Gross-yielding approach at critical locations is followed for
evaluating the safety margins.
Robinson criteria fail to explain the relation between the true
Failure criteria for rotating disc stress and engineering stress. It also fails to explain the effect
of maximum hoop stress in the disc failure. Hence this
The determination of the burst-speed has brought to the
criterion is modified in several ways and one of the forms is
formulation of several theories and criteria that prescribe the
Hallinan criteria.
procedure to determine the rotational speed that causes failure.
Among them, two are used in this work: The Robinson’s
Criteria also called average hoop stress criterion; and
Hallinan criteria:
Hallinan’s criteria.
This criteria extends the use of the Robinson criteria
considering the maximum hoop stress, 𝜎c max introducing a
Robinson’s Criteria [8]: factor S [8] (it is the ratio of true stress to the engineering stress)
to weight the influence of the maximum stress over the mean
The Robinson criteria is a method which is developed for
stress. The ratio S is obtained from the graph figure 3 by
calculating the burst speed in the hoop mode knowing the
comparing true and engineering stress.
ultimate tensile strength, 𝜎UTS and the mean hoop stress, 𝜎c mean.
As per this criterion “burst occurs when the mean hoop stress The mathematical form of Hallinan criteria
on a disc section becomes equal to the nominal tensile strength
𝜎𝑈𝑇𝑆 𝜎𝑈𝑇𝑆 𝜎𝑈𝑇𝑆
of the material, determined from a uniaxial tensile stress”. 𝜔𝑏𝑢𝑟𝑠𝑡 = 0.95𝜔 [𝑆 (√ −√ )+√ ] Eqn. 7
𝜎𝑐,𝑚𝑒𝑎𝑛 𝜎𝑐,𝑚𝑎𝑥 𝜎𝑐,𝑚𝑎𝑥
The mathematical formulation of Robinson criteria,
These two criterions are similar to each other if the value of S
𝜎𝑈𝑇𝑆
𝜔𝑏𝑢𝑟𝑠𝑡 = 𝜔√ Eqn.4 is close to 1, such as for discs made of ductile materials.
𝜎𝑐,𝑚𝑒𝑎𝑛
Therefore, the equation becomes,
It is important to know that ultimate strength considered by
𝜎𝑈𝑇𝑆
Robinson Criteria is the engineering stress that differs from true 𝜔𝑏𝑢𝑟𝑠𝑡 = 0.95𝜔√ Eqn. 8
𝜎𝑐,𝑚𝑒𝑎𝑛
stress. ANSYS applies engineering stress to perform the
analysis since; the evolution of safety margins is to be In general, the concept of calculating the average stress is
performed by considering true stress. Due to this relation associated with the ductility of material and no difference
between these stresses is to be performed and utilized for

14501
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

occurs between the two formulations; while if the disc is made 3. Bi-linear Kinematic Analysis:
of brittle material it is more appropriate to weight the influence
The isotropic model indicates that, yield strength in tension and
of the maximum stress by applying the Hallinan Formula.
compression are same in the beginning, i.e. the yield surface is
symmetric about the stress axes; they remain equal as the yield
1. Linear Static Analysis: surface develops within plastic strain. In Isotropic hardening
the material hardens till it responds elastically. To overcome
From the behavior constraints mentioned over-speed ≥ 118%
this, alternate laws i.e. kinematic hardening laws is introduced.
of the 100% speed. Hence, for evaluating the safety margins
As per hardening laws, the material become softer in
121% speed (16940 rpm) is considered as per API standards,
compression and thus yield surface remains same shape and
𝜎𝑦𝑒𝑖𝑙𝑑 size but turns in stress space, which gives the real behavior of
Over speed Margin = √ Eqn.9 the material. Hence, this approach gives the real behavior burst
𝜎 𝑐,𝑚𝑒𝑎𝑛
speed obtained are acceptable.
For burst speed,
Particulars Case1 Case2 Case3
𝜎𝑈𝑇𝑆
𝜔𝑏𝑢𝑟𝑠𝑡 = 𝜔 √ Over-Speed limits 1.34 1.29 1.27
𝜎𝑐,𝑚𝑒𝑎𝑛
Burst Speed 181.37% 150.34% 149.75%

Particulars Case1 Case2 Case3

Over-Speed limits 1.34 1.31 1.29
Burst Speed 190.9% 158.7% 157.50%

Similarly, for other cases and methods of analysis similar

approach is carried out for calculating the over-speed margin,
the burst-speed.
Linearity is just an assumption which simplifies the modeling.
Every problem is bi-linear, but solving these models time
consumption is more when compare to the linear models.
Hence, analysis of linear models is carried out whenever
possible. However, the assumptions built into linearity must be
considered with every model. It is assumed here that the
ultimate stress is within the elastic limit and the there is no yield
i.e. the stiffness of the material remains constant in linear
analysis which is not real. But, the burst takes place in the disc Figure 7: OSM for different approachs with respect to
when the induced stress is equivalent to ultimate stress i.e. the change in speed
linearity of the material is lost. Therefore, the assumptions
made are not applicable for evaluating the burst speed. Hence,
the burst speed obtained from this approach is in-valid. The evaluated safety margins is plotted in the graph with
respect to the change in speed. For over-speed (121% as per
API standards) condtion the saftey margins obatined for the
2. Bi-linear Isotropic Analysis: disc are in the range of 32-34% which is as shown in figure 7.
In isotropic hardening when the material is subjected to loading
tension or compression the material deforms linearly till the
yield limit. Once the material is crossed the yield limit the
material doesn’t deforms uniformly. But in this case the
deformation of the material takes place by maintaining its
centre constant. This implies that stress while tension and
compression remains same in value but opposite in signs,
which is not possible in the real behaviour of the material.
Hence the obtained burst speeds by this approach are also
unacceptable.

Particulars Case1 Case2 Case3

Over-Speed limits 1.35 1.30 1.29
Burst Speed 186.14% 153.33% 152.73%

Figure 8: Burst-margin for different approachs with respect to

change in speed

14502
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

The graph plotted figure 8 is by considering CASE 3 loading

condition. The burst-speed obtained for the disc from different
approachs is in the range of 152.73-149.7% of the operating
speed which is as shown in figure 8.

Comparison of Burst speed between Robinson and Hallinan

criteria:
Hallinan criteria = 0.95*Robinson criteria [by using Eqn. (6) &
Eqn. (1)] [2]
Therefore,
Burst speed in Hallinan criteria = 0.95* 190.9% = 181.355%
Similarly, the above is used to calculate for all other cases.

Figure 10: Total strain induced in the rotor assembly

From Manson’s-Coffin’s Approach

∆𝜖 ∆𝜖𝑒 ∆𝜖𝑝
= +
2 2 2

𝜎𝑓′
= (2𝑁𝑓 )𝑏 + 𝜖𝑓′ (2𝑁𝑓 )𝑐
2𝐸

Nf = 296319.69 cycles

CONCLUSION
Figure 9: Burst limits varying with two different approaches  The sensitivity analysis and design checks conducted
through blending the classical equations and
methodologies has thrown light on various aspects
In the above figure 7, 8 and 9 represents linear, isotropic, namely design considerations, parametric and design
kinematic analysis respectively. It is observed that burst limits constraints along with behavior constraints to achieve
obtained from Hallinan is conservative and safe when compare design goal. A non-linear analysis is must for
to Robinsons limits. evaluating the safety limits as per regulations to
ensure integrity.

FATIGUE LIFE ASSESSMENT  From the present study for evaluating the safety limits
for disc a bi-linear kinematic approach is
Fatigue is a critical failure phenomenon as it is the reason to recommended. The safety limits obtained from this
more than 90% of all service failures of machine components. analysis provides a conservative result in a range of 6-
Therefore, fatigue life estimation is extremely important in 8% when compared to other two approaches
extremely loaded components design. The turbo machines are
the most critical parts of gas turbine engines. These parts have  The burst limits obtained from Hallinan’s criteria is
colossal kinetic energies that are at the peak during maximum considered since, the results are 5-7% conservative
thrust conditions. Such processes incite intense cyclic stresses when compare to Robinsons criteria. By incorporating
in turbo machines. Hence, the absence of life prediction leads this criterion in FE code would help in evaluating
to low cycle fatigue failures. The expected life of small gas burst limits.
engine is much less than large engines. Hence certain degree of  The estimated fatigue life of the disc obtained from
plastic deformation is permissible in this class of engines. Coffin-Manson and Universal Slope method is greater
Manson- Coffin’s method (Strain approach) [9] is the best than 1*e6 cycles. For, estimating the fatigue life of
suited life estimation method for the case under study. For such disc Coffin-Manson method is optimum since, it
application the strain approach is better than the stress approach provides a factor of safety 1.66 when compare to
which basically ignores plasticity. universal slope method. Universal Slope method is
utilized for approximation of fatigue life when there is
lack of material data.

14503
 International Journal of Applied Engineering Research ISSN 0973-4562 Volume 13, Number 20 (2018) pp. 14498-14504
© Research India Publications. http://www.ripublication.com

REFERENCES
[1] Lakshman kasina, Raghavan kotur and Govindaraji
Gnanasundaram “Minimum weight design of Aero
engine Turbine disks”- ASME 2015
[2] Harinath SP “Evaluation of Over-speed, Burst Margin
and Estimation of Low-cycle Fatigue Life of an Aero
Engine Disc” (IJIRAE) Issue 04, Volume 4, 2017.
[3] K. Kumar, S.L Ajit Prasad and Shivarudraiah “Strength
evaluation in turbo machinery Blade disk assembly at
constant speed”- IJAME 2010
[4] K. Kumar, Ajit Prasad and K Ramachandra, “Critical
issues in assessment of over speed and burst margin in
aero engine discs”, International Journal of Computer
Applications in Engineering Technology and Sciences,
Vol-2(1), 2010
[5] S.A Zamani, S.R Tahmasbpour Oman, B.Asadi and M.
Hosseinzadh “Numerical simulation and plane stress
analytical solution of rotating disk in high speed” –
Indian J.sci res-3 2014
[6] Maruthi B.H, M. Venkatarama Reddy, K.
Channakeshavalu “Finite element formulation for
prediction of over speed and burst margin limits in Aero-
margin” IJSCE 2012, 2231-2307, VOL -2, july 2012
[7] STEFANO CHIANESE “Safety factor against burst
speed of turbo machinery rotating discs” -2011
[8] W.N Barack and P.A Domas “An improved turbine disk
design to increase reliability of aircraft jet engines”
NASA CR-135033
[9] U. MURALIDHARAN “A Modified Universal Slopes
equation for estimation of fatigue characteristics of
metals” – journal of engineering materials and
technology (Jan 1998) 110(1):55

14504