Chapter 14

PSD Random Vibration Strength

and Fatigue Analysis of a CNG Tube
Trailer

Zhiqian Liu, Caifu Qian, and Zuzhi Chen

Abstract Tube trailer is a kind of gas transportation equipment, which is widely

used in transportation of various industrial gases including compressed natural gas
(CNG). However, due to its particularity of carrying high-pressure gas and operating
and parking on urban roads, the requirements for its safety performance are especially
concerned. In this paper, finite element random vibration analysis on a tube trailer
was performed according to the standard GB/T 4857.23-2012, random vibration
analysis under the “ISTA 3A vehicle power spectral density (PSD) curve”. The first
9 orders natural frequencies are obtained from modal analysis and it is found that
only the first natural frequency is covered in the speed limitation of 80 km/h for the
tube trailer. The maximum deformations in the first 3 orders all occur in the middle
of the frame of the tube trailer; Stress analysis under random vibration excited by
the specified PSD found that the maximum stresses corresponding the probability
values at 1σ, 2σ and 3σ are all less than the yield stress of the material, meaning
that the structure meets the strength requirements. Within the fatigue life under PSD
random vibration, the trailer’s fatigue damage cycle ratio D is less than 1, meaning
the structure meets the fatigue design requirements.

Keywords Tube trailer · Finite element analysis · Random vibration analysis ·

PSD

14.1 Introduction

Compared with pipeline transportation, tube trailers have the advantages of lower
investment and higher suitable for short and medium distance transportation.

Z. Liu · C. Qian (B)

Beijing University of Chemical Technology, Beijing 100029, China
e-mail: qiancf@mail.buct.edu.cn
Z. Chen
China Special Equipment Inspection and Research Institute, Beijing 100029, China

© The Author(s), under exclusive license to Springer Nature Switzerland AG 2021 173
J. Xu and K. M. Pandey (eds.), Mechanical Engineering and Materials,
Mechanisms and Machine Science 100,
https://doi.org/10.1007/978-3-030-68303-0_14
174 Z. Liu et al.

However, since tube trailers are mostly used on highways during land transporta-
tion, they sometimes shuttle or stop in densely populated areas. As the working
pressure is high and the medium are flammable and explosive, it will cause huge
human life and property losses once an accident occurs. Therefore, comprehensive
analysis and safety assessment of tube trailers is of very practical significance.
In recent years, there have been more and more relevant studies on tube trailers in
China and abroad. Yu et al. [1] studied the fatigue life of tube trailer gas cylinders, and
proposed to control the inertial load during the transportation. Liu et al. [2] discussed
and analyzed the reasons for the wear of the external thread of the cylinder bottle,
the external thread structure of the bottle mouth, and the arrangement position of
the cylinder. Wang [3] analyzed the failure modes of CNG tube trailer gas cylinders
comprehensively, and believed that to reduce the failure damage, it is necessary to
improve the fixing structure of the bottle ends. Gao et al. [4] analyzed the actual
problems of tube trailer gas cylinder under different working conditions. Dong et al.
[5] proposed that the standardization of design and manufacture is a prerequisite
for safe use, and perfecting and improving technical requirements, forming tech-
nical standards are inevitable for industry development. Zhang et al. [6] put forward
suggestions on risk prevention for tube trailers from the aspects of material selection,
shape optimization, compliance with standards and anti-corrosion methods to ensure
the safety of the product under normal road conditions and correct operation.
As a kind of transportable equipment, tube trailers will inevitably experience
bumps and vibrations during road operation, which places more stringent require-
ments on the strength and fatigue life of the structure under random vibration condi-
tions. It is quite difficult to perform accurate vibration analysis of tube trailers with
the conventional design method. Instead, it is effective to conduct random vibration
analysis using finite element method.
Different methods of random vibration analysis are available. Hao et al. [7]
performed random vibration analysis of the converter cabinet based on the ASD
(Acceleration Spectrum Density) spectrum values applied to all directions of the
structure. Yang et al. [8] conducted their analysis of the frame of Radix Ophiopogonis
harvester by applying the displacement spectral density of road roughness.
In this paper, ANSYS Workbench 19.0 was applied to overall model the key
structures of tube trailers (i.e. frames and gas cylinders). Random vibration analysis
was carried out under the excitation of ISTA 3A random vibration PSD curve from
the standard GB / T 4857.23-2012 “Basic Experiments of Transport Package Part
23: Random Vibration Test Method”. Stress analysis was conducted on the structure
based on 3σ theory and the strength and fatigue life of the structure under the random
vibration is evaluated and studied.

14.2 Random Vibration Analysis Concept

Random vibration analysis is also called power spectral density analysis [9], which is
a statistical method based on probability [10]. The power spectral density spectrum
14 PSD Random Vibration Strength and Fatigue Analysis … 175

represents a measure of the mean square value of random variables and is mostly used
for random vibration analysis. In many situations in reality, the load on the structure
and the corresponding time course are uncertain, so the method of transient analysis
is not applicable in the simulation calculation under this premise [11], continuous
transient response can only be described by a probability distribution function, i.e. the
probability corresponding to a certain level of response. The power spectral density
is the statistical result of the response of the structure under random dynamic load
excitation, and is a relationship curve between the power spectral density value and
the frequency value.
The traditional method is to repeat the design and experiment multiple times.
This method will lead to repeated design and excessive optimization, which is time-
consuming and inefficient [12]. In order to solve this problem regarding the shortcom-
ings of traditional design methods, the products that have preliminarily designed are
analyzed using ANSYS simulation software which has been widely used in various
analyses including modal and random vibration analysis and then the weak parts of
the products are reinforced [13].
Random vibration analysis belongs to frequency domain analysis, and modal
analysis must be performed to obtain the natural frequency of the structure before
performing random vibration analysis.

14.3 Finite Element Modeling of Tube Trailer

For simplification but without affecting the results a lot, the flanges of the connection
between the gas cylinder and the baffle and the details of the individual support plates
were simplified. The finite element model is shown in Fig. 14.1. The model size is
12192 × 2438 × 1400, the design temperature is −40–60 °C, and the fully loaded

Fig. 14.1 Finite element

geometry model of the tube
trailer
176 Z. Liu et al.

Table 14.1 Main properties

Material Elastic modulus Poisson’s ratio Density [kg/m3 ]
of the material
[MPa]
Q345R 2.09 × 108 0.3 7850

Fig. 14.2 Finite element

grid model of the tube trailer

mass is 28755 kg with the net weight being 25,500 kg and the loading capacity being
3255 kg.
The material used in the equipment is Q345R, and its main properties are listed
in the Table 14.1.
Element Solid186 is used to mesh the model. The grid is hexahedral with mostly
a size of 50 mm after performing irrelevant testing. The whole grid model is shown
in Fig. 14.2.
For the boundary conditions, the four corner supports at the bottom of the frame
are fully restrained for each analysis.
It should be pointed out that there are stresses at the tubes due to internal pressure,
but this paper mainly concerns the strength of trailer frames which is not affected by
the pressure. So the pressure inside the tubes is not considered in this analysis.

14.4 Modal Analysis

Performing modal analysis, the natural frequency results of the first 9 orders are
obtained and listed in Table 14.2.
Tire 12R22.5PR is widely used in tube trailers manufactured in China. According
to “GB/T 2977-2016 specifications, size, pressure and load of truck tires”, the
diameter of the tire is 1085 mm.
Based on Eq. (14.1):
14 PSD Random Vibration Strength and Fatigue Analysis … 177

Table 14.2 The first 9 order

Order Natural frequency [Hz]
modal analysis results
1 6.1198
2 7.9711
3 9.5705
4 11.199
5 11.621
6 11.954
7 12.74
8 12.903
9 12.983

2π
v = ωr = r = 2π f r (14.1)
T

where: v is the linear speed of the tube trailer, m/s;ω is the angular speed of the tube
trailer, rad/s; r is the radius of the tire, mm; T is the period, s; f is the frequency, Hz.
According to the notice issued by the National Development and Reform Commis-
sion (NDRC) on soliciting comments on the national standard “Technical Require-
ments for Vehicle Speed Limiting Systems”, the maximum speed of dangerous goods
transport vehicles shall not be higher than 80 km/h. Substituting the speed limit and
tire radius values into Eq. (14.1) yields Eq. (14.2):

1
80 × 3.6 = 2 × 3.14 × f × × 1085 × 10−3 (14.2)
2
Therefore, f = 6.52 Hz.
So the maximum frequency that can be allowed for a tube trailer with the tire
12R22.5PR under safe driving conditions is 6.52 Hz.
The above calculations show that in normal road operation, only the 1st order
natural frequency of 6.1198 Hz is covered in the speed limitation. But in order to
ensure the safety of the long-tube trailer in case of possible emergencies during
driving or speeding, the analysis is extended to cover the 3rd order, where the
calculated driving speed is 117.38 km/h.
Deformation contours for the first three order vibration modes are shown in
Fig. 14.3. It is seen that the maximum deformations for the first 3 orders are all
located in the middle of the frame of the tube trailer.
178 Z. Liu et al.

(a) 1st order vibration mode (b) 2nd order vibration mode

(c) 3rd order vibration mode

Fig. 14.3 Deformation contours at the vibration modes for the first 3 orders

14.5 PSD Random Vibration Analysis

14.5.1 Selection of PSD Loads

In practice, random vibration loads applied to the vehicle are mainly caused by uneven
road surfaces. The standard GB/T 4857.23-2012 “Basic Experiments on Transport
Packages Part 23. Random Vibration Test Methods” provides the experimental accel-
eration power spectral density curves and data in the relevant international standards
for random vibration. In this study, the acceleration power spectral density curve for
the vehicle random vibration according to ISTA 3A-2008 was selected as shown in
Fig. 14.4.
Figure 14.4 gives a PSD excitation, which is not representative of a particular
transport situation, but averaged from different weather, different road conditions,
different driving speeds, etc. and used as an input excitation in a real vibration test.
ANSYS Workbench 19.0 is applied to perform random vibration simulations of
the model. Using the Random Vibration module, the coordinates of the corresponding
points within Fig. 14.4 were entered as PSD G Acceleration for loading with the value
14 PSD Random Vibration Strength and Fatigue Analysis … 179

Fig. 14.4 Random vibration

PSD curves for model
loading

of the modal damping ratio being taken as 0.01. von Mises stress was obtained for
the results.

14.5.2 Strength Analysis

From the random vibration calculations, stress solutions under random loads can be
obtained according to different probability described by 1σ, 2σ and 3σ as shown in
Fig. 14.5 for the distribution. The most dangerous case is the solution occurring at
the 3σ case.
Stress distributions are shown in Fig. 14.6. The maximum stresses at 1σ, 2σ and
3σ are 93.256 MPa, 186.51 MPa and 279.77 MPa, respectively. Figure 14.7 shows
that the location of maximum stress of 279.77 MPa is located at the end of the frame,
where the crossbeam joins the longitudinal beam.
As the maximum stress value at 3σ of this structure is 279.77 MPa, which is less
than the yield stress of material Q345R (345 MPa), it is reached that the structure
meets the strength requirements under random vibration excited by the specified
PSD.

Fig. 14.5. 3σ normal distribution

180 Z. Liu et al.

(a) Stress distribution at 1σ (b) Stress distribution at 2σ

(c) Stress distribution at 3σ

Fig. 14.6 Stress solutions at 1σ, 2σ and 3σ

Fig. 14.7 Position of the maximum stress

14.5.3 Fatigue Analysis

Regarding fatigue calculations of structures under random vibration loads, for struc-
tures whose stresses obey a Gaussian distribution, it is usual to calculate the fatigue
of the structure by the Steinberg’s proposed three-interval method [14], that is, the
14 PSD Random Vibration Strength and Fatigue Analysis … 181

Mises stress is divided into three intervals according to a 3σ distribution, with each
interval corresponding to an occurrence probability of 68.3%, 95.4% and 99.7%.
As the stresses for the case greater than 3σ occur only 0.27% of the time, it is
assumed that they do not cause any damage.
From Miner’s law, the formula for the fatigue life under PSD random vibration
can be obtained as Eq. (14.3) [15]:
n 1σ n 2σ n 3σ
D= + + (14.3)
N1σ N2σ N3σ

In Eq. (14.3), D is the fatigue damage cycle ratio; n1σ, n2σ, n3σ is the actual
number of cycles under the corresponding 1σ, 2σ, 3σ level; N1σ, N2σ, N3σ is the
number of cycles under the corresponding 1σ, 2σ, 3σ solution, which can be found
in JB 4732. Since n 1σ = 0.6831v0+ T ,n 2σ = 0.272v0+ T ,n 3σ = 0.4331v0+ T [16], it is
easy to transform Eq. (14.3) into Eq. (14.4):

0.6831v0+ T 0.272v0+ T 0.04331v0+ T

D= + + (14.4)
N1σ N2σ N3σ

In Eq. (14.3), v0+ is the average frequency of vibration, Hz; T is the expected life
of the structure, s.
Based on relevant requirement, assuming that the value of expected life T is
1.08 × 104 s, the average frequency v0+ is 5 Hz, and the maximum stresses at 1σ,
2σ and 3σ are respectively 93.256 MPa, 186.51 MPa and 279.77 MPa. Based on JB
4732 C-1, the corresponding N1σ, N2σ, and N3σ can be calculated as 4.96 × 105 ,
3.06 × 104 and 8.23 × 103 , respectively.
Substituting these values into Eq. (14.4) yields Eqs. (14.5) and (14.6):
 
0.6831 0.272 0.04331
D= + + × 1.08 × 104 × 5 (14.5)
4.96 × 10 5 3.06 × 10 4 8.23 × 103

D = 0.834 < 1 (14.6)

Therefore, the structure can meet the fatigue design requirements.

14.6 Conclusions

In this paper, finite element simulation was performed to analyze the strength of
a tube trailer under PSD stochastic excitation. The main conclusions are drawn as
follows.
(1) The first 9 orders natural frequencies are obtained from modal analysis and only
the first natural frequency is covered by the speed limitation of 80 km/h for the
182 Z. Liu et al.

trailer. The maximum deformation at the first 3 orders is all located in the middle
of the frame of the tube trailer.
(2) Stress analysis under random vibration excited by the specified PSD found that
the maximum stresses corresponding the probability values at 1σ, 2σ and 3σ are
respectively 93.256 MPa, 186.51 MPa and 279.77 MPa, respectively. They are
less than the yield stress of the material, meaning that the structure meets the
strength requirements.
(3) Fatigue analysis found that in the fatigue life under PSD random vibration, the
trailer’s fatigue damage cycle ratio D is 0.834, less than 1, meaning the structure
meets the fatigue design requirements.

References

1. Yu, X., Song, B., Zhang, Z., Liu, Q.: Analysis of fatigue life for tube trailer cylinders. J. Hebei
Univ. Sci. Technol. 36(4), 407–412 (2015)
2. Liu, Z., Li, X., Zhu, B., Yue, Z.: Random vibration analysis of two-component vehicle hydrogen
supply system based on power spectral density. Packag. Eng. 40(1), 69–74 (2019)
3. Wang, Q.: Analysis and discussion of potential risks in application of natural gas tube trailer.
Chem. Eng. Machinery. 46(4), 372–374 (2019)
4. Gao, H., Lin, S., Li, D., Wei, H., Tao, F., Wu, B.: The road running research of tube trailer gas
cylinder based on finite element method. Fujian Chem. Ind. 12, 11–15 (2016)
5. Dong, H., Li, B., Bo, K., Chen, Z.: Review of the development of tube trailer cylinder on safety
technology. China Special Equipm. Saf. 30(8), 1–5 (2014)
6. Zhou, R., Zhang, J., Xu, G., Zhao, W.: Developing of safety accessories application technology
on long tube trailer. Shandong Chem. Ind. 44(24), 103–105 (2015)
7. Hao, Y., Feng, J., Hu, J.: Probabilistic distribution of Von Mises stress and strength evaluation of
structures under static-random vibration combined condition. J. Vibrat. Shock 39(5), 188–193
(2020)
8. Yang, Y., Liang, S., He, Y.: Random vibration analysis of an ophiopogon japonicus harvester
frame. J. Xihua Univ. (Natural Science Edition) 38(6), 87–91 (2019)
9. Woodcock, D.L.: An Introduction to Random Vibration. Robsinm J.D. Cambridge University
Press, Edinburgh (1963)
10. Zheng, M., Shen, F., Luo, P.: Vibration fatigue analysis of the structure under thermal loading.
J. Adv. Mater. Res. 853, 559–564 (2014)
11. Bai, Y., Qiu, E., Wang, H.: Random vibration fatigue life analysis and optimization design of
connector based on ANSYS. J. Sichuan Ordnance 40(11), 178–182 (2019)
12. Zhao, W., Zhang, Y.: Reliability analysis of random vibration transmission path systems. Mech.
Syst. Signal Process. 113, 77–89 (2018)
13. Xia, J., Li, G., Li, B.: Fatigue life prediction of Package-on-Package stacking assembly under
random vibration loading. Microelectron. Reliab. 71, 111–118 (2017)
14. Steinberg, D.S.: Vibration analysis for electronic equipment, 2nd edn. Wiley, New York (1988)
15. Sun, Y., Deng, J., Liu, Z.: Random vibration fatigue analysis of body parts. Internal Combustion
Engine & Parts 1, 49–51 (2019)
16. Wang, K., Shi, Y., Yan, G.: Random vibration analysis of unmanned electric vehicle frame
based on pavement power spectrum. Agric. Equip. Vehicle Eng. 58(6), 129–133 (2020)