Failure Analysis of A Re-Design Knuckle
Failure Analysis of A Re-Design Knuckle
Failure Analysis of A Re-Design Knuckle
Received: 6 September 2018 – Revised: 9 July 2019 – Accepted: 3 August 2019 – Published: 26 August 2019
Abstract. In this study, a systematic design process is carried out for the design of the knuckle. A systematic
method is proposed for the design and analysis of a lightweight steering knuckle in an electric vehicle. In the
proposed approach, a finite element (FE) model of the knuckle is constructed based on an inspection of the
suspension and steering requirements of the target vehicle and the results of a kinematic analysis. A two-stage
topology optimization method is then applied to refine the material distribution within the FE model in such
a way as to minimize the knuckle weight. Finally, FE simulations are performed to evaluate the strength of
the knuckle under road impact conditions and to determine the fatigue life of the knuckle for four ISO 8608
road classes (A–D). The results show that the optimized knuckle has a weight of 3.64 kg (approximately 6.2 %
lighter than the original knuckle of the same strength and material) and achieves fatigue lives of 2.512 × 1011 ,
2.972 × 108 , 5.598 × 103 and 2.432 × 101 cycles for road classes A, B, C and D, respectively.
Table 1. Vehicle load factor and specific load conditions (Fan et al.,
2011). Unit: Newton.
Figure 7. Finite element model of knuckle. 4.2 Boundary conditions for fatigue life analysis
Figure 12. Road profile signal (class A) and corresponding input displacement signal to shock absorber hardpoint. (Note that vehicle velocity
V is equal to 40 km h−1 .)
Table 3. ISO road roughness classes. angle; Sg () is the roughness value of the road surface class;
δf is the frequency interval; f is the frequency; and n is the
Grade of roughness Sg () road surface roughness and usually has a value between 2
unit: 10−6 (m2 (cycles m−1 )−1 ) and 3.
Road Class Range Geometric Mean Consider the general road surface level (class A) as an ex-
ample and assume that the maximum speed of the target vehi-
A (Very Good) <8 4 cle is V = 40 km h−1 . Assume further that the road surface is
B (Good) 8–32 16 composed of h = 1000 sine waves with a frequency interval
C (Average) 32–128 64 of δf = 0.05 Hz. In other words, the road surface covers the
D (Poor) 128–512 256
frequency range of 0.05 to 50 Hz. According to Table 3, the
E (Very Poor) 512–2048 1024
F 2048–8192 4096
maximum value can be used as the input of the road grade,
G 8192–32 768 16 384 Sg (). Therefore, the Sg () values of classes A, B, C and
H > 32 768 D road roughness are 8, 32, 128 and 512, respectively. From
Eq. (3), the time domain signal of the road surface profile
(class A) can be obtained with the form shown in Fig. 12a.
Sg () is the roughness function of the road surface (cy- The corresponding input displacement signal applied to the
cle m−1 ). In addition, N1 is the frequency index related to shock absorber hardpoint is shown in Fig. 12b.
the frequency structure of the road spectrum and has a value
in the range of 1.75–2.25 with an average value of 2, N1 = 2.
Table 3 shows the ISO roughness of the eight road classes 5 Model verification
(ISO 8608, 1995).
The ISO specification for the road roughness has the form To ensure the validity of the FE model and confirm the feasi-
of a frequency signal. However, the input to the tire system bility of the designed knuckle, the deformation of the knuckle
used in the present fatigue analyses should have the form in the actual EV was measured under the application of a
of a time-domain signal. Thus, a Matlab program was used load of 450 N in the downward y direction at the upper shock
to convert the ISO-standard frequency domain signal into a mount with fixed constraints applied in the x, y and z direc-
time-domain signal in accordance with the following rela- tions at the lower control arm hardpoints and at the wheel
tionship (Soliman et al., 2006): center position of the knuckle (see Fig. 13). The results ob-
h
s tained using a dial gauge placed under the strut showed that
X V n−1
X (t) = 2 × Sg (0 ) × × δf the strut arm underwent a displacement of 0.0117 mm. Un-
j =1
fjn der the same load and constraint conditions, the FE model
× sin(2π fj tk + ψj ) (3) showed a displacement of 0.0113 mm (see Fig. 14). This
value is broadly similar to the experimental result, and hence
where X(t) is the displacement signal of the road profile and the basic validity of the FE model is confirmed (see Fig. 15).
is composed of multiple sine waves of different frequencies; Notably, the displacement predicted by the FE model for the
h is the number of sine waves, k is the current sampling point redesigned knuckle is less than that observed for the actual
of the signal; V is the vehicle speed (m s−1 ); ψ is the phase knuckle. Thus, it is further confirmed that the strength of the
Figure 13. Knuckle arm deformation measurement under applied load of 450 N.
Figure 17. Fatigue life analysis of knuckle for class A road surface.
Table 4. Fatigue life of knuckle for various road classes and vehicle
speed of 40 km h−1 .
7 Conclusions
at the point of maximum von Mises stress was taken and
the rain flow counting method was used to count the num- This study has proposed a systematic framework for the de-
ber of occurrences of the maximum stress. The fatigue life, sign and analysis of a lightweight knuckle for a commercial
Nf , was then calculated in accordance with the occurrences electric vehicle (EV). In the proposed framework, the design
ni of amplitude σa corresponding to the average stress σm . space is identified from an inspection of the steering and sus-
Finally, Miner fatigue theory was used to accumulate the fa- pension systems of the target EV and a hardpoints sensitiv-
tigue damage values under various fatigue strengths in order ity analysis. A finite element model of the knuckle is con-
to calculate the total fatigue life of the knuckle (see Fig. 16). structed in ABAQUS and two-stage topology optimization is
At a vehicle speed of 40 km h−1 , the maximum ampli- then performed to minimize the weight of the knuckle. Fi-
tudes of the road surface in the time-domain signal dia- nally, finite element simulations are conducted to analyze the
grams are approximately 10, 25, 50 and 100 mm for road strength of the knuckle and evaluate its fatigue life under four
classes A, B, C and D, respectively. Figure 17a shows the ISO 8608 road classes (A–D). The simulation results support
von Mises stress distribution in the re-designed knuckle for the following main conclusions:
the class A road surface. It can be seen that the von Mises The original knuckle has a weight of 3.88 kg. By con-
stress reaches a maximum value of 32 MPa at a point just trast, the redesigned knuckle has a weight of just 3.64 kg for
beneath the knuckle cone. Figure 17b shows the von Mises the same material and strength constraints. In other words,
stress history at the corresponding location. Substituting the the redesigned knuckle achieves a weight saving of approxi-
stress history results into FE-Safe, the fatigue life is found to mately 6.2 %.
be Nf = 2.512 × 1011 cycles. The fatigue life is greater than Under impact load conditions, the knuckle is capable of
1×106 cycles, and hence the knuckle can be assumed to have bearing 3.5 G in the vertical direction. The maximum von
Mises stress has a value of σe,max = 314.4 MPa and occurs ISO 8608: Mechanical Vibration – Road Surface Profiles – Report-
beneath the knuckle strut arm. The maximum stress is less ing of Measured Data, International Standardization Organiza-
than the yield strength of the knuckle material (GCD450). tion, Geneva, Switzerland, 1995.
Hence, the mechanical integrity of the knuckle under impact Muhamad, W. W., Sujatmika, E., Hisham, H., and Tarlochan, F.: De-
loads is ensured. sign Improvement of Steering Knuckle Component Using Shape
Optimization, International Journal of Advanced Computer Sci-
The fatigue life analysis results show that the redesigned
ence, 2, 65–69, 2012.
knuckle has fatigue lives of 2.512 × 1011 , 2.972 × 108 , Shelar, M. L. and Khairnar, H. P.: Design Analysis and Optimiza-
5.598 × 103 and 2.432 × 101 under ISO 8608 road classes tion of Steering Knuckle using Numerical Methods and Design
A, B, C and D, respectively. For general road classes (A and of Experiments, International Journal of Engineering Develop-
B), the fatigue life is greater than 1 × 106 cycles, and hence ment and Research, 2, 2958–2967, 2014.
the knuckle can be regarded as having an infinite life. Sivananth, V. and Vijayarangan, S.: Fatigue Life Analysis
and Optimization of A Passenger Car Steering Knuckle
under Operating Conditions, International Journal of Au-
Data availability. All the experimental data can be obtained from tomotive and Mechanical Engineering, 11, 2417–2429,
the author. https://doi.org/10.15282/ijame.11.2015.22.0203, 2015.
Soliman, A. M. A., Abd Alla, S. A., El-Mashed, Y. A., and
Hamid, M. S. A.: Improvement of Vehicle Ride Performance
Author contributions. YCC and HHH conceived of the pre- Using a Hydro-pneumatic Active Suspension System, SAE
sented idea. We developed the theory and CWW performed the Technical Paper No. 2006-01-1298, SAE International, USA,
CAE. YCC and HHH verified the analytical methods and super- https://doi.org/10.4271/2006-01-1298, 2006.
vised the findings of this work. All authors discussed the results Song, C. Y. and Lee, J.: Reliability-based Design Optimization
and contributed to the final manuscript. of Knuckle Component using Conservative Method of Moving
Least Squares Meta-models, Probabilist. Eng. Mech., 26, 364–
379, https://doi.org/10.1016/j.probengmech.2010.09.004, 2011.
Vijayarangan, S., Rajamanickam, N., and Sivananth, V.:
Competing interests. The authors declare that they have no con-
Evaluation of Metal Matrix Composite to Replace
flict of interest.
Spheroidal Graphite Iron for a Critical Component
Steering Knuckle, Materials and Design, 43, 532–541,
https://doi.org/10.1016/j.matdes.2012.07.007, 2013.
Financial support. This research was supported by MOST Wang, J. X., Ou, K. H., Liu, S. H., and Chen, J. H.: Simulation of
(project no. MOST 107-2637-E-020-001). knuckle strength based on multi-body dynamics, Chang’an Auto-
mobile Engineering Research Institute CAE Department, Chang
An, 2002.
Review statement. This paper was edited by Guimin Chen and Zoroufi, M. and Fatemi, A.: Fatigue Life Comparisons of Com-
reviewed by Shuang Xu and three anonymous referees. peting Manufacturing Process: A Study of Steering Knuckle,
SAE Technical Paper Technical, No:1999-01-0344, SAE Inter-
national, USA, https://doi.org/10.4271/2004-01-0628, 2004.
References
Bhardwaj, S., Ashok, B., Lath, U., and Agarwal, A.: Design and Op-
timization of Steering Upright to Reduce the Weight Using FEA,
SAE Technical Paper 2018-28-0081, SAE International, USA,
https://doi.org/10.4271/2018-28-0081, 2018.
Fan, P., Zhao, B., and Qiu, L.: The Analysis on Destruc-
tion Forms of Steering Knuckle, Third International
Conference on Measuring Technology and Mechatron-
ics Automation, IEEE, Shanghai, China, Vol. 3, 677–679,
https://doi.org/10.1109/ICMTMA.2011.740, 2011.