Research on pre-tensioning process of low-pressure

turbine shaft assembly for aeroengines considering the

rabbet joint structure

Xiaodong Li1, Jin Chen2 , Kepeng Sun1, Liangguo Meng1, Qingkai Han1, Qingchao Sun1*,
1 School of Mechanical Engineering, Dalian University of Technology,
Dalian, PR China
2 AECC Commercial aircraft engine Co.,LTD,Shanghai,PR China

Abstract: Assembly is an important part of aero-engine manufacturing process.

Its workload accounts for more than half of the total aircraft generation work.
Aero-engine is in extreme service environment. Therefore, higher requirements
are put forward for the performance and reliability of its assembly and connection
system. In this paper, the influence of the change of the interference of the rabbet
joint structure on the stress of the joint surface of the rabbet joint structure, the
pressure of the joint surface of the shaft and disk and the pre-tightening force of
the bolt is obtained through simulation and analysis. The effect of the pre-tension
assembly process on improving the initial tightening state is clarified. A number
of pre-tension assembly process experiments under different process parameters
are completed and compared. The optimum pretension process parameters can
effectively improve the coaxiality of rotor assembly and the uniformity of bolt
pretension force.

Keywords: Rabbet joint structure，Simulation, Aero engine assembly，Inter-

ference，Pre-drawing process

1 Introduction

The assembly quality of aero-engine is directly related to various service

performance indicators, and which almost determines the final quality, manufacturing
cost and cycle of the engine. The assembly process is difficult and involves many
disciplines, which is one of the key bottlenecks of engine manufacturing in China.
Assembly is an important part in the process of aero-engine manufacturing. Its
workload accounts for more than half of the total aircraft generation. The control of
assembly accuracy is one of the key technologies in aero-engine manufacturing. The
aero-engine low-pressure rotor transmits the power of low-pressure turbine to the fan,
and the fan rotates to generate thrust, which makes the aircraft generate flight power.
The research object of this paper is the bolt connection between the low pressure turbine
shaft and the cone wall of the disc drum: the shaft disc and the cone wall is connected

*Corresponding author: Qingchao Sun ([email protected]).

ISSN 2572-4975 (Print), 2572-4991 (Online) 23
International Journal of Smart Engineering, Volume 3, Issue 1, 2019

by the 36 bolts. The connection of the shaft disc and cone wall has features of thin
wall ,many connecting bolts, complexl oad (axial load, bending load, torsional load,
etc.), high environment temperature and the high speed, which requires high connecting
quality. The connecting quality of the shaft and the cone wall is very important to the
engine. The connection quality has an important influence on the safety and reliability
of the whole engine [1-3].
Bolted connection is a typical non-linear structure. Although a lot of research has
been done on bolted connection at home and abroad, it mainly focuses on the sealing
connection ,such as pipeline. The research on complex boundary conditions, such as
low-pressure rotor connection bolts is still in its infancy. Generally, the bolted
connection is neglected and the general manual is used to determine the preload. The
connection quality of the rotor can not be guaranteed [4-6]. Cold/hot assembly is often
used for the rabbet joint structure interference assembly of rotors, but the mechanism
of assembly process is not clear, which leads to lower qualified rate of coaxiality of the
rotor after assembly, it can’t guarantee assembly quality, and often leads to poor
bonding state of the joint surface, uneven local stiffness of the connection and difficult
to predict. In order to ensure the high safety and reliability of aeroengine, the assembly
process should be improved on the basis of mechanism research, and the quality and
reliability of installation should be improved.
The main connection form of aeroengine rotor/stator is bolt- the rabbet joint structure
connection. The main function of the rabbet joint structure is rotor centering, while bolt
connection is the key of bearing capacity. When assembling rotor/stator, the assembly
quality often fails to meet the standard, and the trial-and-error rate is very high. Aero-
engine is in extreme service environment, which requires higher performance and
reliability of its assembly and connection system. The stiffness, pressure, uniformity
and assembly accuracy of the joints between the rotors are all important factors
affecting the service performance. The centring connection between the ends of the
rotors is the key and difficult point in the assembly process of the rotors/stators. In
actual assembly, the connection stiffness between the rotors, the pressure of the joints
and the assembly accuracy of the whole machine are all important factors affecting the
service performance. It is difficult to fully play its centring role, and even lead to the
overall assembly coaxiality accuracy decline. In this paper, the influence of the change
of the interference of the rabbet joint structure on the stress of the joint surface of the
rabbet joint structure, the pressure of the joint surface of the shaft and disk and the pre-
tightening force of the bolt is obtained through simulation and analysis. The effect of
the pre-tension assembly process on improving the initial tightening state is clarified.
A number of pre-tension assembly process experiments with different process
parameters are completed and compared. For the low-pressure turbine shaft connection
of aero-engine, combined with the pre-tensioning assembly process used in the hot-
packing process, after the low-pressure turbine shaft-disc heating assembly, pre-
tensioning is carried out with appropriate process parameters ,which can improve the
initial state of the bolt connection, and effectively improve rotor assembly coaxiality
and bolt preload uniformity.

24 ISSN 2572-4975 (Print), 2572-4991 (Online)

 International Journal of Smart Engineering, Volume 3, Issue 1, 2019

1.1 Analysis purpose

The main connection forms of low-pressure turbine shaft disk are interference fit of
rabbet joint structure and bolt connection. In this connection mode, the rabbet joint
structure mainly plays the role of centering, and the bolt connection plays the role of
transmitting torque.
The existence of the rabbet joint structure constraints makes the bolt group
connection of low-pressure turbine shaft disk different. On the one hand, the restriction
of the rabbet joint structure will produce stress on the joint surface, which will lead to
the geometric deformation of the disk surface; on the other hand, in the assembly
process, the existence of size deviation, shape and position error and installation error
will lead to inaccurate positioning between the low-pressure turbine shaft and the cone
wall, and uneven stress on the interference surface, which will cause the gap between
the joint surface and the deformation of the disk surface. Incorrect positioning and disk
deformation will lead to uneven distribution of pre-tightening force and unbalanced
mass of rotor disk when bolts are fastened, which will cause vibration in work. The
interference of the rabbet joint structure is the key parameter . Therefore, it is necessary
to study the influence of the change of the rabbet joint structure interference on the
tightening contact state.

1.2 Rabbet joint structure

The rabbet joint structure is a structure used for positioning when the low-pressure
turbine shaft and the cone wall are assembled. The low-pressure turbine shaft is first
assembled into the rabbet joint structure, and the axis of the shaft and the cone is caused
by a certain interference between the hole axes of the stop. Position in the circumferen-
tial direction to facilitate the next step of bolt tightening. The mating size of the rabbet
joint structure is 238.7 H7/p7, wherein the shaft plate is a concave end and the cone
wall is a convex end.

1.3 Finite element analysis of the rabbet joint structure

(1) Establish a low-pressure turbine shaft-cone wall finite element analysis model.

ISSN 2572-4975 (Print), 2572-4991 (Online) 25

International Journal of Smart Engineering, Volume 3, Issue 1, 2019

Fig.1. Three-dimensional model

(a) low-pressure turbine shaft; (b) cone wall; (c) the whole structure after assembly

(2) The interference amount of the rabbet joint structure is shown in Tab.1.Other
boundary conditions and bolt pre-tightening force were applied. Finally, data pro-
cessing and analysis were carried out to extract the results of the joint surface stress,
combined surface pressure of shaft and disc, bolt pre-tightening force, etc., so as to
analyze the influence of interference amount of the rabbet joint structure on the above
parameters[7-10].

Table 1. The interference of the rabbet joint structure

groups 1 2 3 4 5 6 7

Interference
0.00 0.004 0.02 0.04 0.06 0.08 0.096
(mm)

Fig. 2. Finite element analysis model

26 ISSN 2572-4975 (Print), 2572-4991 (Online)
 International Journal of Smart Engineering, Volume 3, Issue 1, 2019

(3) Results
The effect of the change in rabbet joint structure interference on the axial load.

Fig. 3. The position of chosed bolts for analysis

In the state where the working load is loaded, take 4 bolts at A, B, C, and D as shown
in Fig.3, where the connection of the A and C bolts is the x-axis (the bending moment
is around the x-axis), and the bolts connection between B and D is perpendicular to A
and C.

Fig. 4. Force of a bolt joint node

For any bolt, take the intersection of the center line and the vertical bisector of the
adjacent bolt as the sample point, take 4 points in the circumferential direction, take the
average normal stress and the shear stress value, and draw a curve.

ISSN 2572-4975 (Print), 2572-4991 (Online) 27

International Journal of Smart Engineering, Volume 3, Issue 1, 2019

Table 2. Bond surface stress value

Interference Average shear stress Average normal stress

(mm) (MPa) (MPa)

0.004 11.92593 5.797342

0.01 11.92273 5.851008

0.02 11.88497 6.014063

0.03 11.78777 6.032823

0.04 11.66596 6.05108

0.048 11.43381 6.094819

Fig. 5. Results of finite element analysis

With the increase of the interference, the normal stress increases and the shear stress
decreases. Under the state of working, the pre-tightening forces of the 4 bolts at A, B,

28 ISSN 2572-4975 (Print), 2572-4991 (Online)

 International Journal of Smart Engineering, Volume 3, Issue 1, 2019

C and D as shown in Fig. 3 are extracted, and using the average pre-tightening force
value of the 4 bolts to draw the curve [11-12].

Table 3. Average pre-tightening force of bolts

Interference (mm) Average preload (N)

0.004 45406.55
0.01 45403.84
0.02 45399.93
0.03 45396.23
0.04 45392.78
0.048 45388.09

Fig. 6. Influence of interference on bolt preload

Analysis of the above figure shows that under the action of the working load, as the
interference increases, the pre-tightening force of the bolt gradually decreases.

2 Pre-drawing process experiment

2.1 Experimental object and experimental equipment

The experimental environment was a special positioning clamping system for low-
pressure turbine shaft disc, and a special specimen for low-pressure turbine shaft disc
experiment was used. In the experiment, ultrasonic patch bolts, ultrasonic measurement
system, copper heater and digital torque wrench were used.

ISSN 2572-4975 (Print), 2572-4991 (Online) 29

International Journal of Smart Engineering, Volume 3, Issue 1, 2019

Fig. 7. Ultrasonic measuring system

Fig. 8. Experimental specimen and ultrasonic bolt

2.2 Experimental program

（1）No pre-tension assembly experiment

The hot-packing process is used to heat the temperature of the low-pressure turbine
shaft recess to about 100 °C, push it into the mouth and measure it with a 0.03 mm
feeler gauge. If it cannot be inserted in a circle, it means that the joint is assembled
successfully. Wait for the door to cool to room temperature (about 15 minutes) and start
tightening the bolts.
Tighten the way: D-type ultrasonic bolts are used for tightening. The tightening
method is a single-step tightening of the torque method. Use 76Nm as the target torque
to tighten with a digital torque wrench. Mark the serial number of all bolts. Test the
pre-tightening force for each bolt, as the initial pre-tightening. Force, after all 36 bolts
are screwed out, wait 10 minutes until the bolt pre-tightening force is stable, measure
the corresponding pre-tightening force of each bolt, and record the final pre-tightening
force after 10mins.
（2） Verify pre-tension assembly experiment
The hot-packing process is used to heat the temperature of the low-pressure turbine
shaft recess to about 100 °C, push it into the mouth and measure it with a 0.03 mm
30 ISSN 2572-4975 (Print), 2572-4991 (Online)
 International Journal of Smart Engineering, Volume 3, Issue 1, 2019

feeler gauge. If it cannot be inserted in a circle, it means that the joint is assembled
successfully.
Number the bolts and the shafts. Starting from 12 o'clock, use a torque wrench
to tighten the 18 process bolts in the singular position to 10Nm according to the cross
method to complete the pre-tensioning. Use a torque wrench to tighten the bolts in the
double-position bolts according to the cross method. The method is tightened to 76Nm,
measuring the pre-tightening force of the corresponding bolt, and record it as the initial
pre-tightening force; unscrew the process bolt in the singular position, and use a torque
wrench to screw the working bolt to 76Nm according to the cross method, and measure
the pre-bolt of the corresponding bolt. Tight force, and recorded as the initial preload;
check all bolts in time to verify that it reaches 76Nm; wait 10min until the bolt preload
is stable, measure the corresponding preload of each bolt, and record it as 10min tight
force.
（3） Pretensioned assembly experiments under different process parameters
The experimental scheme is the same as the above experiment, the number of bolts
in the pre-drawing process and the corresponding torque should be changed to verify
the influence of the pre-drawing process on the evenness of the pre-tightening force of
the final bolts with different process parameters.

2.3 Data processing analysis

（1）Experimental data analysis of predrawing contrast of bolts with or without

process

Fig.9. No bolt with pretensioning Fig.10. Bolts with pre-tensioning

The average preload of 36 bolts is used as the pre-tightening target value. The final
average preload force change is 2329.78N when no pre-tensioning of the process bolt
is taken. The average preload force change value after the pre-tensioning assembly pro-
cess is 1312.69N. Without pre-tensioning, the pre-tensioning force change value of the
pre-tensioning is reduced by 56.34%. Therefore, the pre-tensioning of the shaft by the
process nut can effectively reduce the influence of the rabbet joint structure on the dis-
persibility of the pre-tightening force.
（2）Experimental data analysis under different predrawing process parameters
ISSN 2572-4975 (Print), 2572-4991 (Online) 31
International Journal of Smart Engineering, Volume 3, Issue 1, 2019

Compare and analyze the uniformity of pre-tightening force of bolts without pre-
tensioning process and under different pre-tensioning process parameters:

Fig.10. Preload’s dispersibility of different number of bolts with pretensioning

Based on the above experimental results, the pretension of process bolts can effec-
tively reduce the dispersion of pretension force of bolt connection, and the more the
number of pretension bolts, the smaller the dispersion of pretension force.

Table 4. Pretension test results

Number of pre-tensioned Pre-pulling torque Final preload force vari-

bolts （Nm） ance

0 -- 5.4872E+07

18 10 1.5211E+07

9 18 2.2411 E+07

4 30 2.3741E+07

32 ISSN 2572-4975 (Print), 2572-4991 (Online)

 International Journal of Smart Engineering, Volume 3, Issue 1, 2019

3 Conclusions

The interference amount of the rabbet joint structure has a certain effect on the pre-
tightening force of the bolt, and with the increase of the interference amount, the pre-
tightening force of the bolt gradually decreases.
Pretensioning process is adopted to assemble low-pressure turbine shaft disc of aero-
engine, which can effectively reduce the influence of the rabbet joint structure on the
dispersion of bolt pretightening force.
In the pre-drawing process, the more the number of bolts, the smaller the dispersion
of the final pre-tightening force.

Acknowledgement

This work is supported by the Major special basic research projects (No. 2017-VII-
0010-0105), Shanghai Sailing Program(No.19YF1452400) and the Fundamental Re-
search Funds for the Central Universities of China (No.DUT18LAB04).

References
1. Liang Z. Research on aeroengine assembly digitalization system[J]. Computer Engineering
& Design, 2007.
2. Huang Min, Wang Zhao, Yi Jun. Study on Probability Distributions of Processing and As-
sembling Promoters of Aero-Engine Critical Part[J]. Applied Mechanics and Materials,
2011, 290:107-111.
3. Xu T. Analysis on Assembly Precision of Important Aircraft Engine[J]. Science and tech-
nology economic guide, 2017(18):116.
4. AStephen J T, Marshall M B, Lewis R. An investigation into contact pressure distribution
in bolted joints[J]. Proceedings of the Institution of Mechanical Engineers, Part C: Jour-
nal of Mechanical Engineering Science, 2014, 228(18): 3405-3418.
5. Oskouei R H, Keikhosravy M, Soutis C. Estimating clamping pressure distribution and stiff-
ness in aircraft bolted joints by finite-element analysis[J]. Proceedings of the Institution of
Mechanical Engineers, Part G: Journal of Aerospace Engineering, 2009, 223(7): 863-
871.
6. Nassar S A, Abboud A. An improved stiffness model for bolted joints[J]. Journal of Me-
chanical Design, 2009, 131(12): 121001.
7. Kim J, Yoon J C, Kang B S. Finite element analysis and modeling of structure with bolted
joints[J]. Applied mathematical modelling, 2007, 31(5): 895-911.
8. Yang G, Hong J, Wang N, et al. Member stiffnesses and interface contact characteristics of
bolted joints[C]. 2011 IEEE International Symposium on Assembly and Manufacturing,
2011: 1-6.
9. Pedersen N L, Pedersen P. On prestress stiffness analysis of bolt-plate contact assemblies[J].
Archive of Applied Mechanics, 2008, 78(2): 75-88.
10. Al-Huniti N S. Computation of member stiffness in bolted connections using the finite ele-
ment analysis[J]. Mechanics based design of structures and machines, 2005, 33(3-4): 331-
342.

ISSN 2572-4975 (Print), 2572-4991 (Online) 33

International Journal of Smart Engineering, Volume 3, Issue 1, 2019

11. Shigley J E. Shigley's mechanical engineering design[M]. Tata McGraw-Hill Education,

2011.
12. Richtlinien VDI 2230 Blatt 1. Systematic calculation of high duty bolted joints[S]. VDI-
Richtlinien, 2003:1-169.

34 ISSN 2572-4975 (Print), 2572-4991 (Online)