Proceedings of Mechanical Engineering Research Day 2018, pp. 42-43, May 2018
__________
© Centre for Advanced Research on Energy
P(VDF-TrFE) piezoelectric sensor for Internet of Things application Khoon-Keat Chow1,2,*, Swee-Leong Kok2,*, Kok-Tee Lau3, Ali Mohammed Abdal-Kadhim2
1) Department of Electrical Engineering, Politeknik Ungku Omar,
Jalan Raja Musa Mahadi, 31400 Ipoh, Perak, Malaysia. 2) Faculty of Electronics and Computer Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]; [email protected]
Keywords: piezoelectric sensor; P(VDF-TrFE); Internet of Things (IoT)
ABSTRACT – This paper presents the conversion of
mechanical vibration into electrical energy using
P(VDF-TrFE) piezoelectric sensor. The sensor consisted
of three materials with two layers of electrode in
between of P(VDF-TrFE) layer on top of the flexible
substrate as PET. The 2D model was conducted by
using COMSOL Multiphysics 5.1. Results of simulation
shown maximum of displacement and stress were
obtained 1.01×107 N/m2 and 0.3mm at resonance
frequency of 131Hz under acceleration of 1g with
maximum output voltage of 20.2mV. This P(VDF-TrFE)
piezoelectric sensor can be used for the evolution of
internet of things (IoT) where IoT system can be
connected with number of sensors.
1. INTRODUCTION
Nowadays, the greatest challenge faced by
advances technology in wireless sensor networks and
internet of things application is energy. Most of these
devices and sensor are portable and powered by
conventional battery but their life span of battery is
short. Furthermore, the process of replacing battery is
complicated task because these electronic systems could
shut down at any time and some devices can be placed
in remote area such as under bridge or structural sensor
[1-2].
To overcome the battery replacement issue, energy
harvesting is an alternative solution to harvest energy
from ambient environment from mechanical vibration
into electrical energy using piezoelectric materials such
as lead zirconate titanate (PZT), polyvinylidene fluoride
(PVDF), poly(vinylidene fluoride) trifluoroethylene
P(VDF-TrFE) and aluminium nitride (AIN) [3-4]. Due
to the flexible structure application, polymer
piezoelectric materials like PVDF and P(VDF-TrFe) are
the option to be used in energy harvesting application
usage. Previous researcher harvested 0.95 mV PVDF
wafer active sensor [5]. In this study demonstrates the
finite element method (FEM) simulation for P(VDF-
TrFE) piezoelectric as a solution of battery-less for
sensor used in internet of things application. The stress
generation, displacement, resonance frequency and
voltage output have been studied to optimize the
physical parameters of this piezoelectric sensor.
2. METHODOLOGY
A 2D model was designed and simulated using
Comsol Multiphysics 5.1 to analyse the electrical and
mechanical properties. This P(VDF-TrFE) piezoelectric
sensor was simulated based on electrostatics and
structural mechanics interface. The 2D model was
meshed using quadrilateral elements with 1050 fine
elements and minimum meshing size of 0.012, as shown
in the Figure 1.
Figure 1 2D meshed geometry.
Figure 2 Mechanical properties: Resonant frequency for
displacement and von Mises stress contour.
In the mechanical properties, the maximum
displacement was 0.3mm and the maximum von Mises
stress generated about 1.01×107 N/m2 at the resonance
frequency of 131 Hz which its shown in the Figure 2.
Then, frequency domain analysis was able to generate
voltage output about 20.2mV at 131Hz as shown in
Figure 3.
Chow et al., 2018
43
Figure 3 Electrical properties: Voltage output at
resonance frequency.
3. RESULTS AND DISCUSSION
After the COMSOL simulation results, the P(VDF-
TrFE) piezoelectric sensor was fabricated and the
process of fabrication steps have been reported in
previous research paper [6]. Then, it was demonstrated
by using IoT application as shown in the Figure 4 and
the prototype system of P(VDF-TrFE) piezoelectric
sensor was able to generate output voltage about 743mV
when a force of 5N from index finger.
Figure 4 An image and block diagram of the system.
4. CONCLUSION
In this finite element simulation, the electrical and
mechanical properties of P(VDF-TrFE) piezoelectric is
successfully performed. The simulation results shown
maximum stress and displacement were obtained
1.01×107 N/m2 and 0.3mm. It’s also generates maximum
output about 20.2 mV where under acceleration of 1g at
at resonance frequency of 131 Hz. This study shows that
P(VDF-TrFE) piezoelectric can be used as a sensor by
using IoT application without using battery source for
sensor application.
ACKNOWLEDGEMENT
The authors would like to thank the Ministry of
Higher Education of Malaysia for the research grant of
and research grant of PRGS/1/2016/TK10/FKEKK-
CETRI/02/T00016 and UTeM-Industry Matching
GLUAR/IMPRESSIVE/2017/FKEKK-CETRI/I00024
and the support facility provided by Advanced Sensors
and Embedded Control Systems Research Group
(ASECs), UTeM.
REFERENCES
[1] Chuang, C. H., Lee, D. H., Chang, W. J., Weng, W.
C., Shaikh, M. O., & Huang, C. L. (2017). Real-
time monitoring via patch-type piezoelectric force
sensors for Internet of Things based
logistics. IEEE Sensors Journal, 17(8), 2498-2506.
[2] Seah, W. K., Eu, Z. A., & Tan, H. P. (2009, May).
Wireless sensor networks powered by ambient
energy harvesting (WSN-HEAP)-Survey and
challenges. In Wireless Communication, Vehicular
Technology, Information Theory and Aerospace &
Electronic Systems Technology, 2009. Wireless
VITAE 2009. 1st International Conference on, 1-5.
[3] Vatansever, D., Hadimani, R. L., Shah, T., &
Siores, E. (2011). An investigation of energy
harvesting from renewable sources with PVDF and
PZT. Smart Materials and Structures, 20(5), 1-6.
[4] Lavrik, N. V., Sepaniak, M. J., & Datskos, P. G.
(2004). Cantilever transducers as a platform for
chemical and biological sensors. Review of
Scientific Instruments, 75(7), 2229-2253.
[5] Lin, B., & Giurgiutiu, V. (2006). Modeling and
testing of PZT and PVDF piezoelectric wafer
active sensors. Smart Materials and
Structures, 15(4), 1085-1093.
[6] Chow, K. K., Kok, S. L., & Lau, K. T. (2017).
Fabrication and characterization of Piezoelectric P
(VDF-TrFE) thick film on Flexible
Substrate. APRN Journal of Engineering and
Applied Sciences, 12(10), 3347-3351.
Figure 5 An image of the prototype.
Proceedings of Mechanical Engineering Research Day 2018, pp. 44-45, May 2018
__________
© Centre for Advanced Research on Energy
Study of milling parameter effect of AISI 4340 alloy steel using FEM simulation
Afifah Juri, Shalina Sheik Muhamad, Jaharah A. Ghani*, Che Hassan Che Haron
Department of Mechnical and Materals Engneering, Faculty of Engineering and Built Environment,
Universiti Kebangsaan Malaysia, 43600 Bangi, Selangor, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: AISI 4340; thirdwave advantedge; cutting force; cutting Temperature
ABSTRACT – Nowadays, there is significant progress
in the development of simulation and modeling for
machining operations. The aim of this paper is to
investigate the effect of milling parameters in terms of
cutting temperature and cutting force by using Finite
Element Method (FEM). Simulations were performed
using Thirdwave AdvantEdge of the AISI 4340 steel
using uncoated carbide cutting tool. The milling
parameters used were cutting speed of 180- 220 m/min,
feed rate of 0.1- 0.2 mm/tooth and depth of cut at 0.3-
0.7 mm under dry condition. Analysis of variance
(ANOVA) was executed to find a significant parameter
that affects the machining responses. The analysis of the
result shows that cutting speed is the most significant
parameter affecting the cutting temperature and cutting
force on the end milling process.
1. INTRODUCTION
Finite element method is a very useful tool to
study the cutting process of the material. Thirdwave
AdvantEdge software has been used for machining
process simulation. It is a typical program written
specifically for machining simulations. It can generate
simulation results such as temperatures in the tool- chip
interface, as well as cutting forces; during the chip
formation much faster than using costly and time
consuming experiments [1]. Several studies have been
done on modeling and simulation using Thirdwave
AdvantEdge software in analyzing the cutting
temperature and cutting force during the machining
process. Kadirgama et al., [2] used Thirdwave
AdvantEdge software when machining Hastelloy C-
22HS with carbide coated cutting tool to determine
temperature and heat generated at the tool tip. Qasim et
al. [3] used FEM to find optimal parameters of the
orthogonal cutting process of AISI 1045 steel in order to
reduce the cutting forces. It is shown from ANOVA that
the depth of cut and feed rate are the most significant
factor affecting the cutting force.
The problems during machining are due to heat
generation and temperature associated with it. In metal
cutting, the magnitude of the temperature and friction
occur at the tool-chip interface is a function of the
cutting parameters. The study of cutting temperature
generated and cutting force are important due to their
effect on the machining responses. Therefore, this study
is to determine the temperatures at the tool and chip
interface and cutting force to determine the optimum the
cutting condition.
2. METHODOLOGY
The commercial FEM software of Third Wave
AdvantEdge (v6.4) was used to simulate the milling
process in a two dimensional (2D). Figure 1 shows the
schematic of an orthogonal cutting condition model.
Figure 1 Simulation model.
The cutting parameter for this study is shown in
Table 1. Nine sets of simulation combinations were
generated by using Taguchi design of experiment. By
applying Taguchi, the S/N ratios were calculated from
temperature and resultant force using larger is better and
smaller is better respectively.
Table 1 Cutting parameter and their levels.
Cutting parameters Level 1 Level 2 Level 3
Cutting speed
(m/min) 180 200 220
Feed (mm/tooth) 0.1 0.15 0.2
Axial depth of cut
(mm) 0.3 0.5 0.7
The workpiece material was high strength low
alloy steel AISI 4340. The operation is simulated using
insert carbide of DNMA432 type that has a nose angle
of 55 deg and without the use of coolant. The tool was
defined to be a rigid body which considers thermal
transfer for modeling the cutting temperature field. The
workpiece was meshed for a maximum number of
24000 nodes. The maximum and minimum element
sizes for both workpiece and insert were set at 0.1 mm
and 0.02 mm, respectively. The mesh refinement factor
was set at the value of 2, and the coarsening factor was
set at 6.
Juri et al., 2018
45
3. RESULTS AND DISCUSSION
3.1 Cutting temperature
The isothermal contours of the temperature
distributions for dry condition are shown in Figure 2
(Vc: 180 m/min, fz: 0.1 mm/tooth, ap: 0.3 mm). Table 2
shows the cutting temperature obtained by simulations.
Figure 2 Simulation of cutting temperatures at Vc: 180
m/min, fz: 0.1 mm/tooth, ap: 0.3 mm.
Table 2 Cutting temperature obtained by simulation.
Vc
m/min
fz
mm/tooth
ap
mm
Cutting
temperature (0C)
180 0.10 0.3 959.99
180 0.15 0.5 975.748
180 0.20 0.7 948.033
200 0.10 0.5 953.346
200 0.15 0.7 962.458
200 0.20 0.3 978.136
220 0.10 0.7 969.642
220 0.15 0.3 1010.93
220 0.20 0.5 982.21
Table 2 shows the highest temperature is generated
at cutting speed of 220 m/min, feed rate 0.15 mm/tooth,
and depth of cut 0.3 mm. While the lowest temperature
generated is at cutting speed 180 m/min, feed rate 0.2
mm/ tooth, and depth of cut 0.7 mm. In order to
minimise force generates are at the combination of low
speed (180 m/min), low feed (0.1 mm/tooth) and high
depth of cut (0.7 mm). Anova analysis found that the
cutting speed contributes 43.55 %, followed by depth
of cut of 28.08% and feed rate of 26.25% to the
temperature generated. Therefore, cutting speed has the
major effect on temperature generated, which is similar
finding with Das et al. [4].
3.2 Cutting force
Table 3 shows the cutting force in Fx and Fy
obtained from the simulations and calculated resultant
force FR.
From Table 3, the highest cutting force FR is
generated at cutting speed of 220 m/min, feed rate 0.15
mm/tooth, and depth of cut 0.3 mm. While the lowest
cutting force is generated at cutting speed 180 m/min,
feed rate 0.2 mm/ tooth, and depth of cut 0.7 mm. In
order for maximise temperature generates during the
milling the variables are the combination of high speed
(220 m/min), medium feed rate (0.15 mm/tooth) and
low depth of cut (0.3 mm). ANOVA analysis shown that
the cutting speed contributes 75.73%, followed by feed
rate of 12.71% and depth of cut of 11.18 % of the
cutting force generated. Therefore, the cutting speed is
the most influential effect affecting the cutting force.
Table 3 Cutting force obtained by simulation.
Vc
m/min
fz
mm/tooth
ap
mm
Fx
N
Fy
N
FR
N
180 0.10 0.3 444.99 816.09 972.23
180 0.15 0.5 528.72 862.12 972.98
180 0.20 0.7 539.79 878.76 967.44
200 0.10 0.5 490.66 840.21 975.74
200 0.15 0.7 532.21 871.94 994.34
200 0.20 0.3 489.63 839.94 1002.50
220 0.10 0.7 558.93 889.29 997.74
220 0.15 0.3 493.42 845.78 1028.17
220 0.20 0.5 577.11 926.93 1013.37
4. CONCLUSION
From this study, it can be concluded that the FEM
simulation is able to help the researcher to evaluate the
machinability of a particular material such as AISI
4340. From the results obtained it can be concluded that
the cutting speed is the most significant parameter
affecting the cutting temperature and cutting force.
ACKNOWLEDGEMENT
The authors would like to thank the Government
of Malaysia and Universiti Kebangsaan Malaysia for
their financial support under a
FRGS/1/2016/TK03/UKM/01/1 and GUP-2017-048
Grants.
REFERENCES
[1] Maňková, I., Kovac, P., Kundrak, J., & Beňo, J.
(2011). Finite element analysis of hardened steel
cutting. Journal of Production Engineering, 14, 7-
10.
[2] Kadirgama, K., Noor, M. M., & Rahman, M. M.
(2010). Heat Distribution Analysis In End-Milling.
National Conference in Mechanical Engineering
Research, 326-335.
[3] Qasim, A., Nisar, S., Shah, A., Khalid, M. S., &
Sheikh, M. A. (2015). Optimization of process
parameters for machining of AISI-1045 steel using
Taguchi design and ANOVA. Simulation
Modelling Practice and Theory, 59, 36-51.
[4] Das, S. R., Nayak, R. P., & Dhupal, D. (2012).
Optimization of cutting parameters on tool wear
and workpiece surface temperature in turning of
AISI D2 steel. International Journal of Lean
Thinking, 3(2), 140-156.
Proceedings of Mechanical Engineering Research Day 2018, pp. 46-47, May 2018
__________
© Centre for Advanced Research on Energy
Constrained MPC based Laguerre network to control IPA positioning system
Siti Fatimah Sulaiman1,2,*, M.F. Rahmat1, A.A.M. Faudzi1,3, Khairuddin Osman2, A.R. Azira2
1) Faculty of Electrical Engineering, Universiti Teknologi Malaysia,
81310 UTM Johor Bahru, Johor Bahru, Malaysia. 2) Faculty of Electronics and Computer Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
3) Centre for Artificial Intelligence and Robotics, Universiti Teknologi Malaysia,
81310 UTM Johor Bahru, Johor Bahru, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Constrained model predictive control; pneumatic system; position control
ABSTRACT – This study proposed a Model Predictive
Control (MPC) based Laguerre network as a strategy to
control the Intelligent Pneumatic Actuator (IPA)
positioning system. This study mainly aimed to
investigate the effect of considering constraints of IPA in
MPC algorithm towards the performance of IPA
positioning system. Simulation and experimental test
using MATLAB/Simulink demonstrated that the
proposed strategy was feasible, where it managed to
control the IPA positioning system well. The most
noticeable result from this study was the inclusion of
constraints in the controller algorithm effectively reduced
the overshoot in the system response, however, makes the
system relatively slower.
1. INTRODUCTION
Over the past few years, predictive control has
attracted increasing attention in controlling particular
systems due to its ability to predict the future behaviour
of the system. The suitability of predictive control to
control the IPA positioning system used in this study were
also reported in [1-3]. However, all the controllers
reported in those studies did not take into account the
prescribed limitations of the IPA system. Considering the
limitation of the IPA system when developing controller
was very crucial as it can affect the overall performance
of the system, and non-compliance with the prescribed
limits may also cause damage to the IPA system and its
components, especially when implementing the system
in real-time environment. MPC is the only controller in
the predictive control family which has an advantage of
considering system’s constraints in its algorithm. Wakasa
et al. [4] and Daepp [5] in their studies have revealed that
including the system’s constraint in MPC algorithm
significantly improved the performance of the respective
systems. Therefore, based on these facts, this study
proposes the development of constraint MPC as a new
strategy for controlling the IPA positioning system. The
main purpose of this study is to investigate the effects of
considering constraints in the MPC algorithm towards
the performance of the IPA positioning system.
2. METHODOLOGY
In this study, the dynamics of IPA system was
modeled using system identification approach and
procedures involved during modeling the system was
similar as described in [6-7]. Figure 1 illustrates the
schematic diagram of the IPA system used in this study.
The cylinder stroke movements were manipulated by the
duty-cycle of Pulse-Width Modulator (PWM) signals to
drive the valves. If the PWM model receives a positive
signal from the controller, it converts the signal into
equivalent PWM signal and sends that signal to Valve 1
to perform extension. If the PWM model receives
negative signal, the model sends the signal to Valve 2 to
perform retraction.
Chamber 1Chamber 2
Valve 1
Valve 2
Pressure
sensor
Cylinder
0.6 MPa
Exhaust
Optical
sensorExtend
Retract
Stroke
Figure 1 IPA system schematic diagram.
The MPC control law based Laguerre network
which consider the constraints on the valves control
signals of IPA system can be realized as:
−255 ≤ 𝑀𝜂 + 𝑢(𝑘 − 1) ≤ +255 (1)
Meanwhile, 𝑀 in Equation (1) can be represented as:
𝑀
=
[ ∑ 𝐿1(𝑖)
𝑇𝑘−1
𝑖=002
𝑇 ⋯ 0𝑚𝑇
01𝑇 ∑ 𝐿2(𝑖)
𝑇𝑘−1
𝑖=0⋯ 0𝑚
𝑇
⋮ ⋮ ⋮ ⋮
01𝑇 02
𝑇 ⋯ ∑ 𝐿𝑚(𝑖)𝑇𝑘−1
𝑖=0 ]
Note that 0𝑘𝑇 is a row vector with dimension as in
𝐿𝑘(0)𝑇 . +255 and −255 are the limits on the control
signals of Valve 1 and Valve 2, respectively, 𝜂 is the
Sulaiman et al., 2018
47
optimal solution based on minimization of the cost
function, and 𝑢(𝑘 − 1) is the previous control signal. In
this study, 𝑀𝜂 + 𝑢(𝑘 − 1) represents the controller
signal and can be denoted as 𝑢𝑚𝑝𝑐 . Figure 2 illustrates the
block diagram of the IPA positioning system with the
proposed control strategies.
Reference
(mm)Input valves
Linear
dynamic
MPC
(unconstrained/
constrained)
Stroke
position
(mm)+
-
IPA system
Umpc
Figure 2 IPA positioning system block diagram.
3. RESULTS AND DISCUSSION
The performances of MPC with and without
considering valve’s constraints in their control strategies
were presented in Table 1. In this study, the performances
of both controllers were observed based on their ability
to control the IPA positioning system at the actuator mid-
stroke position (100 mm).
Table 1 Performance comparison of UMPC and CMPC
Method Criterion UMPC CMPC
Simulation
(sim)
𝒕𝒓(s) 0.141 0.533
𝑶𝑺(%) 5.219 1.031
𝒆𝒔𝒔(mm) 0 0
Experiment
(exp)
𝒕𝒓(s) 0.669 0.972
𝑶𝑺(%) 12.535 0.420
𝒆𝒔𝒔(mm) 0.040 0.010
*UMPC is unconstrained MPC, CMPC is constrained
MPC, 𝑡𝑟 is rise time, 𝑂𝑆 is overshoot, and 𝑒𝑠𝑠 is steady-
state error.
The results in Table 1 is in the lines of earlier
literature [1-3] that found a predictive control is very
suitable to be used to control the IPA positioning system,
where it successfully provides a highly accurate
positioning control (𝑒𝑠𝑠(𝑠𝑖𝑚) = 0 and 𝑒𝑠𝑠(𝑒𝑥𝑝) ≈ 0).
From the results also, it is apparent that including
constraints in the MPC algorithm affect the 𝑡𝑟 and 𝑂𝑆 of
the system performance. Figure 3 illustrates a
comparative close-up view of the positioning system step
response based on the results in Table 1.
Figure 3 UMPC and CMPC responses at midstroke
position.
A comparison of the closed-loop responses of
UMPC and CMPC in Figure 3 shows that the 𝑂𝑆 in the
system response reduced with the inclusion of constraints
in the control algorithm. However, the system
performance becomes slower with the inclusion of
constraint in the algorithm as it would require more
computational effort to optimize the cost function
compared to the UMPC, consistent with the findings as
reported in [4-5].
4. CONCLUSION
This study was devoted to assess the capability and
effect of adding constraints in the MPC algorithm on the
IPA system’s position. Simulation and experimental
results demonstrated that the proposed control strategy
was feasible to be implemented and was found to be an
effective technique to reduce the overshoot in the
system’s transient response. The overshoot decreased
with the inclusion of constraints in the controller
algorithm. However, the proposed method resulted in a
less aggressive positioning response of IPA system as it
requires more computational effort to optimize the cost
function, compared to the unconstrained case. Further
study will investigate a suitable method to improve the
speed response of the system.
ACKNOWLEDGEMENT
The authors would like to acknowledge Universiti
Teknikal Malaysia Melaka (UTeM), Universiti Teknologi
Malaysia (UTM) and Ministry of Higher Education
(MOHE) of Malaysia for their support.
REFERENCES
[1] Faudzi, A. A. M., Mustafa, N. D., Osman, K.,
Azman, M. A., & Suzumori, K. (2012). GPC
controller design for an intelligent pneumatic
actuator. Procedia Engineering, 41(2012), 657-663.
[2] Faudzi, A. A. M., Mustafa, N. D., Azman, M. A., &
Osman, K. (2014). Position tracking of pneumatic
actuator with loads by using predictive and fuzzy
logic controller. Applied Mechanics and
Materials, 529(2014), 259-266.
[3] Osman, K., Faudzi, A. A. M., Rahmat, M. F.,
Hikmat, O. F., & Suzumori, K. (2015). Predictive
functional control with observer (PFC-O) design
and loading effects performance for a pneumatic
system. Arabian Journal for Science and
Engineering, 40(2), 633-643.
[4] Wakasa, Y., Sasaki, R., Tanaka, K., & Akashi, T.
(2007). Servo control of pneumatic systems
considering input and output constraints. IEEE
International Conference on Control Applications,
2007, 765-770.
[5] Daepp, H. G. (2016). Constrained model predictive
control for compliant position tracking of
pneumatic systems. Georgia Institute of
Technology.
[6] Sulaiman, S. F., Rahmat, M. F., Faudzi, A. A. M., &
Osman, K. (2016). Linear and nonlinear ARX
model for intelligent pneumatic actuator system.
Jurnal Teknologi, 78(6), 21-28.
[7] Sulaiman, S. F., Rahmat, M. F., Faudzi, A. A. M.,
Osman, K., Sunar, N. H., & Salim, S. N. S. (2017).
Hammerstein model based RLS algorithm for
modeling the intelligent pneumatic actuator (IPA)
system. International Journal on Advanced
Science, Engineering and Information
Technology, 7(4), 1457-1463.
Proceedings of Mechanical Engineering Research Day 2018, pp. 48-49, May 2018
__________
© Centre for Advanced Research on Energy
Stepwise regression for kenaf reinforced polypropylene composite M. Noryani1,2, S.M. Sapuan1,*, M.T. Mastura3, M.Y.M. Zuhri1, E.S. Zainudin1
1) Department of Mechanical and Manufacturing Engineering, Faculty of Engineering, Universiti Putra Malaysia,
43400 UPM Serdang, Selangor, Malaysia.
2) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Kenaf reinforced polypropylene composite; stepwise regression
ABSTRACT – Stepwise regression is an alternative in
statistical modelling. This paper discusses the parameters
that influence the performance score (PS) of kenaf
reinforced polypropylene composite (KRPC). It was
found the tensile strength, Young’s modulus and flexural
strength are the parameters that influence the materials
performance of KRPC. The model adequacy checking
was done by plotted the normality and regression
probability standardized residual.
1. INTRODUCTION
The demand of natural fibre as a plastics additive
with the polymer in automotive applications is driving
increasing year by year. The increment by 60% of natural
fibre demand was confirm by natural fibre industry [1].
Kenaf fibre was a favorite fibre in automotive application
due to the availability, ability and flexibility of them. A
good agreement with a better performance on mechanical
and thermal properties can be produce by the composite
compare to single material [2]. There are various
methodologies that can be used to analyze the material
performance and there are several parameters could
influence the material performance. The design engineer
should identify the significant parameters to optimize the
cost and time doing testing the function of desire
automotive component design. In this study, the best
statistical modelling is introduced with significant
parameters to predict the performance of KRPC using a
novel statistical approach.
2. METHODOLOGY
Secondary data of KRPC on strength and flexural
mechanical properties is referred. Table 1 shows the
tensile strength (TS), Young’s modulus (YM), flexural
strength (FS) and flexural modulus (FM). These
independent variables are the regressors to the dependent
variables in this study which is performance score (PS)
that can be calculated by using Equation (1). Where xi is
the mechanical properties of KRPC while n is the number
of samples. The best finding based on material
decomposition such as fibre loading and types of fibre
with polypropylene (PP) are considered. Statistical
package for social sciences (SPSS) software is used to
analyze the data. Statistical measurement such as
coefficient of correlation (r), determination of coefficient
(R2), adjusted determination of coefficient (AdjR2) and
analysis of variance (ANOVA) are used to identify the
significant parameters that influence PS.
1
n
ijiPS x
== 1,2,...,i n= (1)
Table 1 The data of KRPC [1], [3-6].
Mechanical Properties
TS (MPa) YM (GPa) FS (MPa) FM (GPa)
32 1.2 58 2.3
58 1.3 71 3.6
46 2.1 58 4.0
62 6.5 58 3.3
28 6.3 61 3.3
46 6.9 61 3.3
44 4.1 61 3.3
54 4.8 61 3.3
3. RESULTS AND DISCUSSION
The best statistical model from possible fifteen
models is selected based on the highest value of
coefficient of correlation and adjusted determination of
coefficient while the P-value of ANOVA testing should
less than 0.05. The best model of KRPC is selected with
three significant mechanical properties which are tensile
strength, Young’s modulus and flexural strength as
shown in Equation (2). This model shows the highest
value of Adj R2 indicate 99.8% the variation of the PS of
the KRPC is explained by tensile strength, Young’s
modulus and flexural strength. There is significant
difference in the model since the P-value of ANOVA is
less than standard error (α=0.05). The effect of
mechanical properties as a single and multiple parameter
in the possible constructed statistical model is shown in
Table 2.
ˆ 1.220 1.015 1.023 1.021y TS YM FS= + + + (2)
Equation (2) can be used as an inferential statistic
from the sample to predict the PS with the information of
TS, YM and FS. This model can benefit to design engineer
in manufacturing automotive component application
especially in composite industry.
Noryani et al., 2018
49
Table 2 Fifteen possible statistical model of KRPC.
Parameter r R2/ Adj
R2
ANOVA
(P-value)
TS 0.958 0.918 0.000
YM 0.130 0.017 0.579
FS 0.530 0.281 0.177
FM 0.503 0.253 0.204
TS, YM 0.961 0.894 0.002
TS, FS 0.987 0.963 0.000
TS, FM 0.962 0.896 0.002
YM, FS 0.608 0.117 0.316
YM, FM 0.511 0.035 0.469
FS, FM 0.647 0.186 0.257
TS, YM, FS 0.999 0.998 0.000
TS, YM, FM 0.965 0.878 0.000
TS, FS, FM 0.988 0.957 0.001
YM, FS, FM 0.689 0.080 0.416
TS, YM, FS, FM 1.00 0.997 0.20
The adequacy checking was done on this final
model of KRPC shows in Figure 1 and Figure 2. The
normality plot of the observed and expected data display
a good dispersion. There is no positive and negative
skewed in the data. The residual is plotted lie near exactly
along a straight line. The observed value is approximate
to the expected value. There is no obvious model defect
or nonlinearity pattern. The estimation process by using
this model can increase the trustworthy to the user and it
can reduce the uncertainties and error.
4. CONCLUSION
The stepwise regression is used to identify the best
model represent the performance score of KRPC in this
study. Tensile strength, Young’s modulus and flexural
strength are the significant mechanical properties that
influence the performance score of KRPC. The best fit
model with highest r, Adj R2 and significant P-value can
be used to others to estimate the performance score of
KRPC.
ACKNOWLEDGEMENT
The authors would like to thank Universiti Putra
Malaysia for the opportunity doing this study as well as
Universiti Teknikal Malaysia Melaka and Ministry of
Education of Malaysia for providing the scholarship
award and grant scheme Hi-COE (6369107) to the
principal author in this project.
REFERENCES
[1] Akil H. M, Omar M. F., Mazuki A. A. M., Safiee S.,
Ishak Z. A. M., & Abu Bakar A. (2011). Kenaf fiber
reinforced composites: A review. Mater. Des., 32(8–
9), 4107–4121.
[2] Radzi A. M., Sapuan S. M., Jawaid M., & Mansor
M. R. (2017). Influence of fibre contents on
mechanical and thermal properties of roselle fibre
reinforced polyurethane composites. Fibers Polym.,
18(7), 1353–1358.
[3] Ku H., Wang H., Pattarachaiyakoop N., & Trada M.
A review on the tensile properties of natural fiber
reinforced polymer composites. Compos. Part B,
42, 856–873.
[4] Mansor M. R., Sapuan S. M., Zainudin E. S.,
Nuraini A. A., & Hambali A. (2014). Rigidity
Analysis of Kenaf Thermoplastic Composites
Using Halpin-Tsai Equation. Appl. Mech. Mater.,
548, 29–33.
[5] Asumani O. M. L., Reid R. G., & Paskaramoorthy
R. (2012). The effects of alkali-silane treatment on
the tensile and flexural properties of short fibre non-
woven kenaf reinforced polypropylene composites.
Compos. Part A Appl. Sci. Manuf., 43(9), 431–
1440.
[6] Zampaloni M., Pourboghrat F., Yankovich S. A.,
Rodgers B. N., Moodre J., Drzal L. T., Mohanty A.
K. & Misra M.. (2007). Kenaf natural fiber
reinforced polypropylene composites: A discussion
on manufacturing problems and solutions. Compos.
Part A Appl. Sci. Manuf., 38(6), 1569–1580.
Figure 1 Normality plot of KRPC.
Figure 2 Normal probability plot of regression
standardized residual.
Proceedings of Mechanical Engineering Research Day 2018, pp. 50-51, May 2018
__________
© Centre for Advanced Research on Energy
Simulation of tool flute geometry influences the micro-end milling operation
N.A. Norrdin1,*, J.B. Saedon2, M.S. Kasim3
1)Faculty of Mechanical Engineering, Universiti Teknologi MARA Cawangan Pulau Pinang,
Jalan Permatang Pauh, 13500 Permatang Pauh, Pulau Pinang Malaysia. 2) Faculty of Mechanical Engineering, Universiti Teknologi MARA, 40450, Shah Alam, Selangor, Malaysia.
3) Faculty of Manufacturing Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100, Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: 3D FEM; micro-end milling; tool flute geometry
ABSTRACT – This paper presents a 3D FE model for
micro-end milling operation on hardened AISI D2 cold
work tool steel in order to investigate the effect of
different tool geometry of cutting flute numbers (two,
four, six and eight flutes) on the Von-Mises stress
distribution and cutting force analysis in three directions
Fx, Fy and Fz. The machining parameters used in the
simulation work was the cutting speed of 50m/min, depth
of cut of 55 µm, feed per tooth is of 2 µm /tooth. The
predicted cutting forces were compared with the result
trends from the literature to verify the accuracy of 3D FE
model. The results obtained indicate that tool flute
geometry had a slightly impact to cutting forces.
1. INTRODUCTION
Micro-end milling is a tool-based material removal
process; hence it very much relies on the performance of
the micro-milling tools. Chatelain and Zaghbani [1] used
different combinations of three cutters with different
geometry in order to find commonly stable cutting
condition. Common defects related with micro end mills
are the geometric deviations, poor determination of the
tool cutting edge and also the cutting force. This problem
is encountered when scaling down the tool. The micro-
tools are normally fabricated with an edge radius of 1-5
μm [2]. The tool edge radius need to be smaller than the
chip thickness dimension[3]. Additionally, Fang et al. [4]
studied various type of micro tool geometries in order to
obtain the lowest cutting force and higher rigidity during
machining. This paper introduces a 3D FEM model for
micro-end milling to study on the Von-Mises stress
distribution and cutting force. Nevertheless, based on the
previous studies [1-4], it reveals that the tool geometry
can become as an obstacle that can limit the capability in
micro-machining. Therefore, the aim of this paper is
investigate the effect of cutting tool geometry (two, four,
six and eight flutes) on the Von-Mises stress distribution
and cutting force analysis in micro-end milling process.
2. METHODOLOGY
The 3D model is a dynamic thermo-mechanical
FEA using explicit integration in Abaqus/Explicit. The
tool rotates around the Z-axis and the workpiece feed in
the cutting direction at the same time, as shown in Figure
1(a). The workpiece moves in the direction X and fixed
in the direction Y. The workpiece is set as deformed body
and rigid constraints are applied to the tool. The mesh
type of the workpiece is C3D8R which defined as a
hexahedral with eight node coupled bricks and first order
linear interpolation solid element and the tool is C3D4
which defined as a tetrahedron with four nodes linear
interpolation solid element, as shown in Figure 1(b). The
arbitrary Lagrangian Eulerian (ALE) formulation has
been adopted for the workpiece to reduce distortions
during simulations. The cutting tool used in this study
was 0.5mm of diameter with varying geometry in terms
of number of flutes and helix angle as listed in Table 1.
Figure 1 Top view of FEM setting (a) boundary
conditions and (b) meshing of the workpiece and tool.
Table 1 Tool geometry criteria.
Tool Criteria
T1 2-Flute, Diameter = Ø 0.5 mm, Helix angle =
15°, Rake angle = 0°
T2 4-Flute, Diameter = Ø 0.5 mm, Helix angle =
15°, Rake angle = 0°
T3 6-Flute, Diameter = Ø 0.5 mm, Helix angle =
30°, Rake angle = 0°
T4 8-Flute, Diameter = Ø 0.5 mm, Helix angle =
30°, Rake angle = 0°
3. RESULT AND DISCUSSION
3.1 Von-Mises stress distribution
It was noted that all the simulated maximum value
of stress developed at the shear zone at the tool-chip
interface for all cutting tools geometry however the
values are dissimilar. A primary shear zone was able to
be seen on Von-Mises contours as shown in the Figure 2.
As expected, the stress distribution contours show that
the lower value of Von-Mises stress occurs on the
machined surface and increasing towards the cutting
Norrdin et al., 2018
51
edge of the tool. This is happened due to when the cutting
tool edge in the shear zone, it involves higher force to
create chip formation.
Figure 2 Predicted Von-Mises stress contour for T1 tool
in orthogonal view.
Figure 3 shows the decreasing of Von-Mises stress
values was identified when adding the number of cutting
flute to the cutting tool. Nevertheless, the stress
distributions are still in the identical average. This finding
is consistent with the finding of Pratap and Patra [5] in
their FEM model to analyze the Von-Mises stress
distribution and cutting force for Copper material
workpiece.
Figure 3 Predicted Von-Mises stress magnitude on T1,
T2, T3 and T4.
3.2 Cutting force predictions
The cutting forces provide a good indication of
cutting tool performance. It can be summarized the
simulated cutting force variations in the steady state for
each cutting tool geometry of T1, T2, T3 and T4, as
shown in Figure 4. When comparing the cutting force
trends among all the cutting tool geometry, it showed that
the forces generated from the small number of flutes
higher compared to the high number of flutes, giving the
change of 15%, 19% and 10% for Fx, Fy and Fz,
respectively. This is due to fact that when cutting with
higher number of flutes, the tools become more rigid and
feed faster. Since they make less change, it can be said
that the forces of three directions are not sensitive to the
change of number of tool flute since their functions are
to cut the workpiece during the roughing and finishing
process. The addition of tooth at the cutting tool
smoothers the force profile making the fluctuation less
sudden (T4 compare to T1). This is because when adding
the number flutes in tool geometry is most probably due
of the reduction of the high contact surface of cutter tooth
in the cutting zone [6]. Therefore, this can have led to
lower rate of material removal at each tooth, which
significantly decreased the amplitude of the cutting
forces. The comparisons were performed with the finding
of Fang et al. [4] and Davounedinejad et al. [6] which
reveals globally similar trends for Fx and Fy components
in terms of the curve shapes. However, some
discrepancies in the magnitude of forces were observed,
due to different workpiece material.
Figure 4 Simulated cutting forces in three directions, Fx,
Fy and Fz on tool geometry effect,
4. CONCLUSION
The following observation can be concluded on the
influence of tool flute geometry on the simulated Von-
Mises stress and cutting force; (i) the Von-Mises stress
distribution is not significantly influenced by the change
of number of flutes in the cutting tool. (ii) a slightly
reduction in cutting forces was observed as the number
of flute increases. This model will be used as a platform
system which can access by various users for further
studies.
REFERENCES
[1] Chatelain, J. F., & Zaghbani, I. (2011, September).
Effect of tool geometry special features on cutting
forces of multilayered CFRP laminates.
In Proceedings of the 4th International Conference
on Manufacturing Engineering, Quality and
Production Systems (MEQAPS’11), 15-17.
[2] Bissacco, G., Hansen, H. N., & Slunsky, J. (2008).
Modelling the cutting-edge radius size effect for
force prediction in micro milling. CIRP Annals-
Manufacturing Technology, 57(1), 113-116.
[3] Jin, X., & Altintas, Y. (2012). Prediction of micro-
milling forces with finite element method. Journal
of Materials Processing Technology, 212(3), 542-
552.
[4] Fang, F. Z., Wu, H., Liu, X. D., Liu, Y. C., & Ng, S.
T. (2003). Tool geometry study in
micromachining. Journal of Micromechanics and
Microengineering, 13(5), 726-731.
[5] Pratap, T., & Patra, K. (2014). Modeling and
analysis of cutting forces in micro end milling. In
5th International & 26th All India Manufacturing
Technology, Design and Research Conference
(AIMTDR 2014), 1-5.
[6] Davoudinejad, A., Tosello, G., Parenti, P., &
Annoni, M. (2017). 3D Finite Element Simulation
of Micro End-Milling by Considering the Effect of
Tool Run-Out. Micromachines, 8(6), 187-206.
3229
3209 3207
3171
3140
3150
3160
3170
3180
3190
3200
3210
3220
3230
3240
T1 T2 T3 T4
Vo
n M
ises
str
ess
(M
Pa
)
Tool
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
T1 T2 T3 T4
Fo
rce, (N
)
Tool
Fx Fy Fz
Proceedings of Mechanical Engineering Research Day 2018, pp. 52-54, May 2018
__________
© Centre for Advanced Research on Energy
Influence of dimple textured surface on hydrodynamic pressure distribution via computational fluid dynamic
Haniff Abdul Rahman*, Jaharah A. Ghani, Wan Mohd Faizal Wan Mahmood, Rasidi Rasani
Department of Mechanical and Material Engineering, Faculty of Engineering and Built Environment,
Universiti Kebangsaan Malaysia, Bangi, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Fluid mechanic; surface texturing; hydrodynamic pressure
ABSTRACT – Surface texturing has been widely used
to reduce friction between sliding component due to its
ability to trap lubricant, generating hydrodynamic
pressure and increasing load capacity to reduce friction.
In this paper, Finite Volume Method has been used to
study the influence of dimple surface textured via turning
process on pressure distribution for future piston skirt
application. Three cases were simulated: flow along
minor dimple axis, flow along major dimple axis and
flow along untextured surface. Dimple with flow along
major axis was observed to provide the largest positive-
pressure distribution throughout the model compared to
two other cases.
1. INTRODUCTION
In order to improve engine efficiency and reducing
fuel consumption, friction reduction proved to be a vital
step. Approximately 25% of fuel consumption are due to
friction loss in ICE by which piston/cylinder system
contributed to half of it [1]. Surface texturing is an
effective method to counter the friction loss in sliding
component. Earliest work has been done by Hamilton et
al. [2] in regard of mechanical seal. They found that
micro asperity could generate hydrodynamic pressure
and lift, thus creating load carrying capacity. In another
study, using a pin on disk test, Kovalchenko et al. [3] also
observed a friction reduction trend by adding dimples
even at the lowest sliding speed. Similar trend was also
seen by [4]. They stated that textured surface delayed the
transition from hydrodynamic to mixed and boundary
lubrication, for all speed tested. Recently, numerical
method has been implemented by researchers to give in
depth view on a fluid mechanics of the textured surface.
Han et al. [5] modelled a single spherical cap between
two parallel surfaces using CFD. Pressure curve was
found to be asymmetric throughout the dimple, which
results in net pressure build up. Similar result was
obtained by Menon et al. [6] using hemispherical and
semi-ellipsoidal dimple. It was concluded that divergent-
convergent portion of textured surface able to generate a
pressure gradient that helped in supporting load thus
reducing friction.
2. METHODOLOGY
In this study, Ansys-CFX software has been used to
simulate the pressure distribution between smooth top
surface and dimple textured bottom surface. Modelling
method was inspired by [6]. Dimple's parameter was
based on previous study by Dali et al. [7] using turning
process for further used in piston application. The dimple
width is 1.6308𝜇𝑚 on major axis (a) and 0.1984𝜇𝑚 on
minor axis (b). The depth is 63.43𝜇𝑚. Figure 1 shows the
model geometry of dimple in this study.
Figure 1 Geometry with flow along minor dimple axis.
Periodic boundary condition was assigned at the
inlet and outlet. Top plate is smooth with wall velocity U
= 0.39m/s at y coordinate, and bottom plate is static with
texture. Fluid properties are according to real engine
lubricant. The condition is set to isothermal,
incompressible, laminar and steady. Cavitation
phenomenon is not yet considered.
In this study, three cases were simulated, case 1:
flow along minor dimple axis, case 2: flow along major
dimple axis and case 3: flow along minor dimple axis.
Both case 1 and 2 was simulated using area ratio of 9.5%
for comparison purposes. Grid independence test was
done on case 1, by which grid size of 580000 element
was sufficient for simulation due to error with regard to
higher element is less than 5%.
3. RESULTS AND DISCUSSION
3.1 Pressure distribution
Figure 2 shows the pressure distribution on top mid
plane wall for three different cases in which pressure is
plotted against dimensionless distance x.
From Figure 2, sudden rise is seen for all three cases
since the beginning of the plate as it starts moving.
Passing around the x=10 mark, all three cases show a
declination in pressure. This is following the Bernoulli
principle by which as velocity increases, pressure
decreases. Case 1 appears to experience the most drop-in
pressure at this mark followed by case 2 and case 3. This
is due to the flow already reached the divergent part of
the dimple, causing a significant drop in pressure as
Outlet
Inlet
U
Rahman et al., 2018
53
stated by [8]. However, passing the x=40-mark, case 2
shows the largest positive pressure built up throughout
the plate followed by case 2. While for case 1, further
reduction in pressure is exhibited thus providing no load
carrying capacity at this stage. The finding from the
graph is in accordance with study by Menon et al. [6] in
which dimple with flow along major axis provides the
maximum positive pressure compared to minor axis thus
providing highest load capacity and lift force.
Figure 2 Pressure distribution for three different cases.
3.2 Pressure contour
Figure 3 shows the pressure contour on the bottom
plate whereby the dimple is located for case 1 and 2.
(a)
(b)
Figure 3 Pressure distribution for (a) flow along minor
dimple axis and (b) flow along major dimple axis.
From Figure 3, it is observed that negative pressure
drop occurs at the diverging part of both dimples.
Gradual increase in pressure is then seen along the
dimple. This is in agreement with the study by Yu et al.
[9] that found a positive pressure accumulated towards
the converging part for a single dimple. For flow along
minor axis, faster transition from negative to positive
pressure is seen, compared to major axis. However, as the
flow passes the dimple region, the pressure decreases
earlier compared to the major axis with longer width. The
increase in dimple width that the fluid flow through,
provide a larger positive-pressure distribution throughout
the plane, thus results in net load capacity. This provides
an insight on how dimple size is important in generating
hydrodynamic pressure.
4. CONCLUSION
In conclusion, the dimple structure produced via
turning process on the sliding surface able to increase the
hydrodynamic pressure, which results in increasing load
capacity, to help reducing contact and friction between
plate. The orientation of dimple has an effect towards
pressure distribution, as flow parallel to the major axis
provides the largest pressure rise to this case. However,
the present study only acknowledged a single dimple
case, while for future piston skirt application, surface
area is larger. Thus, a study of the area ratio needs to be
done in the future to further optimize the effect of this
dimple on pressure distribution.
ACKNOWLEDGEMENT
This project is supported by the Government of
Malaysia and Universiti Kebangsaan Malaysia under
FRGS/1/2016/TK03/UKM/01/1 and GUP-2017-048
Grants.
REFERENCES
[1] Ryk, G., Kligerman, Y., Etsion, I., & Shinkarenko,
A. (2005). Experimental investigation of partial
laser surface texturing for piston-ring friction
reduction. Tribology Transactions, 48(4), 583-588.
[2] Hamilton, D. B., Walowit, J. A., & Allen, C. M.
(1966). A theory of lubrication by
microirregularities. Journal of Basic
Engineering, 88(1), 177-185.
[3] Kovalchenko, A., Ajayi, O., Erdemir, A., Fenske,
G., & Etsion, I. (2005). The effect of laser surface
texturing on transitions in lubrication regimes
during unidirectional sliding contact. Tribology
International, 38(3), 219-225.
[4] Borghi, A., Gualtieri, E., Marchetto, D., Moretti, L.,
& Valeri, S. (2008). Tribological effects of surface
texturing on nitriding steel for high-performance
engine applications. Wear, 265(7-8), 1046-1051.
[5] Han, J., Fang, L., Sun, J., & Ge, S. (2010).
Hydrodynamic lubrication of microdimple textured
surface using three-dimensional CFD. Tribology
transactions, 53(6), 860-870.
[6] Menon, D. P., Anil, P. M., & Kulkarni, P. S. (2015).
An analysis on the influence of oil pocket shape and
distribution on the reduction of friction in
hydrodynamic lubrication. In Proceedings of
the17th Annual CFD Symposium, Bangalore.
[7] Mohd Dali, M. N. A., Ghani, J. A., Che Haron, C.
H., & Hassan, S. (2017). Fabrication of dimple
-4
-3
-2
-1
0
1
2
3
4
5
0 20 40 60 80 100
PR
ESSU
RE
(PA
)
DIMENSIONLESS Y
Minor Axis Major Axis No Dimple
Rahman et al., 2018
54
structured surface of A390 Al-Si alloy using turning
process. Industrial Lubrication and
Tribology, 69(3), 348-354.
[8] Gropper, D., Wang, L., & Harvey, T. J. (2016).
Hydrodynamic lubrication of textured surfaces: A
review of modeling techniques and key
findings. Tribology International, 94, 509-529.
[9] Yu, H., Wang, X., & Zhou, F. (2010). Geometric
shape effects of surface texture on the generation of
hydrodynamic pressure between conformal
contacting surfaces. Tribology Letters, 37(2), 123-
130.
Proceedings of Mechanical Engineering Research Day 2018, pp. 55-56, May 2018
__________
© Centre for Advanced Research on Energy
The role of stiffener in resisting buckling of externally pressurized cone-cylinder intersection
M.S. Ismail1,3*, O. Ifayefunmi2, S.H.S.M. Fadzullah1
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100, Durian Tunggal, Melaka, Malaysia. 2) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100, Durian Tunggal, Melaka, Malaysia. 3) Excellence and Professional Division, Polytechnic Education Department, 62100, Putrajaya, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Cone-cylinder intersection; external pressure; shell Buckling
ABSTRACT – This paper aims to examine the
influence of stiffener position on the buckling
performance of externally pressurized ring-stiffened
cone-cylinder intersection of B/t = 5. Three different
stiffener locations are analyzed, they are: (i) cone-
cylinder junction, (ii) cone mid-section, and (iii)
cylinder mid-section. Buckling strength of the stiffened
cone-cylinder intersections were obtained with the aid
of FE analyses for all the cases above. Cone-cylinder
intersections are modelled as elastic-perfectly plastic of
Hiduminium alloy (HE-15). Bifurcation study shows
that the intersection area is much stronger than any
other parts of the assembly. Shell reinforced with
stiffener produces stronger structure by 23% of
increment.
1. INTRODUCTION
Shells buckling failure is a common event in the
industrial field and continuously attracts numerous
engineers/designers in finding the appropriate solution
to prevent it. Cylinders with conical end are widely used
in many engineering industries. Their application can be
found in submersibles, missiles, reducer and silo as well
as nuclear industries. Commonly, cylinders with conical
end are combined with a simple weld joint that is known
as intersection. In designing this kind of intersection, it
is necessary to take into account the structural capability
in resisting the buckling occurrence caused by the
excessive load during operation. This buckling
occurrence indeed can be catastrophic. In general,
cylindrical shell with conical ends, subject to external
pressure, may collapse in two difference ways. First, the
shell will locally collapse within either or both the
cylindrical or conical portions. Secondly, the cylindrical
and conical parts may collapse simultaneously.
In the early 1970s, tremendous amount of works
have been done in investigating the buckling
performance of externally pressurize cylinder shells
with conical end closure. Aylward et al. [1] has
presented some experimental and theoretical results of
nearly perfect steel cone-cylinder intersection.
Theoretical results were obtained by using BOSOR 3. In
another study by Galletly et al. [2], the buckling
behaviour of cone-cylinder combinations subjected to
uniform external pressure was examined through
experimental and numerical approach.
Generally, it is believed that shell component can
be further strengthened through reinforcement with
additional stiffener. The stiffener may be positioned
either vertically (stringer) or circumferentially (ring)
along the shell’s axis line. Stiffener often appears with
numerous profiles, this include; flat, angle, tee, box and
many more. However, flat stiffener is frequently used in
engineering design and application. It seems that for
cone-cylinder intersection, the stiffener (ring-type) is
positioned at the cone-cylinder junction, as the shell
junction is assumed to be the weakest part of the entire
assembly. Nonetheless, little work has been done on the
buckling characteristic of stiffened cone-cylinder
intersection. To-date, most recent work referred to the
study of internally pressurized ring-stiffened cone-
cylinder intersections by Teng and Ma [3].
This paper contributes on the understanding of the
influence of stiffener location on the buckling
performance of cone-cylinder intersection with a ring-
stiffener of B/t = 5. The present work is reported in
terms of numerical modelling analysis using the
ABAQUS software package, with the aim to
compliment and validate the experimental results
reported by Galletly et al. [2].
2. METHODOLOGY
First, preliminary calculations were carried out to
validate the experimental data presented in Galletly et
al. [3]. Two types of analysis were employed: (i)
bifurcation eigenvalue analysis, and (ii) nonlinear static
RIKS analysis. The shells consist of vertex cone angle,
α = 45○, 60○ and 75○ with series of L/D = 0.5, as
illustrated in Figure 1 (a). Specimens were assumed to
be made from Hiduminium Alloy (HE-15) with material
properties of E = 75.152 GPa, σyield = 434.37 MPa and υ
= 0.32, with the shell detail information as given in
Table 1. Next, to examine the role of stiffener at
different shell location, three different stiffener locations
are analysed, they are: (i) cone-cylinder junction, (ii)
cone mid-section, and (iii) cylinder mid-section. The
ring-stiffener of B/t = 5, as given in Figure 1 (b) is used
in the present analysis. The above shells were subjected
to external pressurize. All the analyses were carried by
using ABAQUS finite element software package.
3. RESULTS AND DISCUSSION
3.1 Validation study
Figure 2 illustrates the validation results of tested
Ismail et al., 2018
56
shells. The agreement between experiment and FE
analysis are very satisfactory. It may be summarizing
that the nonlinear static, RIKS analysis is found to be
within 1% - 6%. Somehow, the bifurcation eigenvalue
analysis shows that the discrepancy is in the range of
5% - 10%. It is also demonstrated that the bifurcation
analyses overestimate the shells buckling strength by
1% - 5% in comparison to the nonlinear static, RIKS.
However, model G3 was found to be only marginally
affected by both analyses.
Table 1 Detail of tested models.
Model r
[mm]
t
[mm] α [○]
B
[mm]
tstiff
[mm]
G1
68.58 1.3716
45
5 1 G2 60
G3 75
Figure 1 Dimension of (a) stiffened cone-cylinder
intersection and (b) stiffener.
Figure 2 Series of shells validation results.
3.2 The role of stiffener at different shells locations
Figure 3 and 4 illustrate the role of stiffener at
different shell locations for bifurcation eigenvalue and
nonlinear static RIKS analyses. The bifurcation analysis
indicates that the eigenmode formation take place at the
cone region for all tested models. Stiffened the conical
area (Case 2) produces stronger shells with a maximum
increment of 23% obtained, as presented in Figure 3.
Nonetheless, it appears that Case 1 and 2 (stiffener at
intersection and cone mid-section) calculate almost
identical pressure load for model G3.
In contrast, the nonlinear static RIKS analysis,
Case 1 (stiffener at shells intersection) proves to be
more effective in strengthening the shells load bearing
capability (Figure 4). The improvement of the shells
strength is calculated to be in the range of 1% - 6% for
all cases. Apparently, the role of stiffener seems to be
fairly insignificant for all models under Case 3 (stiffener
at cylinder mid-section). This insignificant is probably
due to the fact that the conical part of the assembly is
strongly affected by the buckling formation (similar to
the bifurcation study). Thus, stiffened the cylinder does
not seems to make any difference in improving the
shells load bearing capability.
Figure 3 Results of bifurcation eigenvalue analysis for
each case.
Figure 4 Results of nonlinear static, RIKS analysis for
each case.
4. CONCLUSION
The correlation obtained in the validation study is
very satisfactory as the results agreement is within 10%.
The role of stiffener appears to have a large effect on
cone-cylinder shells buckling strength, as it strengthen
up the shells to an increase of 23%. In contrast, the
bifurcation study shows that the conical part is much
weaker than the junction of cone-cylinder. This finding
can be further verified through experimental study.
5. REFERENCES
[1] Aylward, R. W., Galletly, G. D., & Moffat, D. G.
(1975). Buckling under external pressure of
cylinders with toriconical or pierced torispherical
ends: a comparison of experiment with
theory. Journal of Mechanical Engineering
Science, 17(1), 11-18.
[2] Galletly, G. D., Aylward, R. W., & Bushnell, D.
(1974). An experimental and theoretical
investigation of elastic and elastic-plastic
asymmetric buckling of cylinder-cone
combinations subjected to uniform external
pressure. Ingenieur-Archiv, 43(6), 345–358.
[3] Teng, J. G., & Ma, H. W. (1999). Elastic buckling
of ring-stiffened cone-cylinder intersections under
internal pressure. International Journal of
Mechanical Sciences, 41(11), 1357–1383.
Proceedings of Mechanical Engineering Research Day 2018, pp. 57-58, May 2018
__________
© Centre for Advanced Research on Energy
Impact behaviour of lightweight metal component repair using aluminium particles with high pressure cold spray process and low-
pressure cold spray process A. Manap1,*, Siti Nurul Akmal Yusof1, N.F. Afandi1, Savisha Mahalingam2
1) College of Engineering, Universiti Tenaga Nasional, 43000 Kajang, Selangor, Malaysia.
2) Institute of Sustainable Energy, Universiti Tenaga Nasional, 43000 Kajang, Selangor, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Aluminium; high pressure cold spray; low pressure cold spray
ABSTRACT – This study focuses on impact behaviour
of lightweight metal repair using aluminium powder by
high pressure cold spray process (HPCS) and low-
pressure cold spray process (LPCS). The aluminium
particles impacting on aluminium substrates using
LPCS process deformed slightly with the smallest
flattening ratio that leads to less pore formation between
the particles and resulted in good coating quality.
Moreover, LPCS deposition experiences compressive
stress that ensures longer component lifetime due to its
positive effect on the fatigue life and wear resistance
application. The overall results denote that LPCS
process is better for repairing lightweight metal
component than HPCS process.
1. INTRODUCTION
In the past decades, the world is facing the
challenge of global warming caused by the emission of
greenhouse gases such as carbon dioxide (CO2) through
human activities. Lightweight metal such as aluminium
(Al) applied tremendously in transportation can be used
to reduce CO2 emissions. However, defects may occur
in Al components such as crack, corrosion and wear.
Thus, repairing the components is an effective way to
lessen the global impact on the environment that can
save energy consumption and cost where, repairing the
components costs less than by replacing with new ones
[1]. Cold spray (CS) technique is a new approach to
repair all defects in lightweight materials. There are two
types of CS techniques which are high pressure cold
spray (HPCS) and low pressure cold spray (LPCS).
However, HPCS has limitation with lightweight
components. One of the common problem include
dimensional error. Al coating deposited using LPCS has
higher hardness because of peening effect [2]. Thus
LPCS is better for lightweight metal deposition. This
work aims on the study of impact behaviour of Al
particles impacting on different lightweight substrates
such as Al, titanium (Ti) and magnesium (Mg) by
smoothed particle hydrodynamics (SPH) simulation.
2. METHODOLOGY
The SPH modeling of the CS process was
performed using an in-house research program
developed in FORTRAN. The impact of Al powder on
Al, Mg, and Ti substrate is simulated using SPH. The
Johnson-Cook parameters and Gruneisen equation of
state are presented in Table 1.
Table 1 Material properties of Al, Mg, Ti.
Properties (Unit) Al Mg Ti
Density, ρ (g/m3) 2710 1.738 4520
Shear Modulus (GPa) 68.9 17 116
Heat capacity (J/kg/K) 904 1020 528
Reference temperature, T0
(K) 300 300 300
Melting temperature, Tm
(K) 916 923 1650
JC parameter, A, B, C, n,
m (MPa)
148.4
345.5
0.001
0.183
0.895
532
229
0.0294
0.032
1.00
806.6
481
0.319
0.019
0.655
Gruneisen parameter 2 1.07 1.23
Intercept Us-Up curve c
(m/s) 5386 5920 4573
Slope Us-Up curve, S 1.338 1.38 1.536
3. RESULTS AND DISCUSSION
3.1 Impact behavior of Al on Al
Figure 1 and 2 shows the deformation behavior of
Al multiple particle impact on the Al substrate using
LPCS and HPCS processes, respectively.
Figure 1 Deformation behaviour of Al multiple particle
impact on Al substrate by LPCS process.
The low velocity from LPCS process caused less
intensive deformation in lower particle and leads to a
good bond formation. On the other hand, the HPCS
process caused intensive deformation and created more
Manap et al., 2018
58
pores between the particles and forms poor coating.
Figure 3 shows the residual stress of Al on Al by
using LPCS and HPCS. Stress turns to compressive
stress at greater depth due to the large peening effect of
the Al impacting on the substrate that leads to good
coating deposition [4].
Figure 2 Deformation behaviour of Al multiple particle
impact on Al substrate by HPCS process.
Figure 3 Residual stress of Al multiple particle impact
on Al substrate by LPCS and HPCS processes.
3.2 Impact behavior of Al on other substrates
Figure 4 and 5 shows the deformation behavior of
Al multiple particle impact on the Ti and Mg substrates
using LPCS and HPCS processes, respectively. Since,
Mg is lighter and less hard than Al, more deformation
was formed in the substrate. On the other hand, more
deformation formed on particles on Ti substrate than Mg
due to almost all kinetic energy dissipated into Al
particle than in the Ti substrate.
Figure 4 Deformation behaviour of Al multiple particle
impact on (a) Ti (b) Mg substrates by LPCS process.
Figure 6 shows the residual stress of Al on Ti and
Mg by using LPCS and HPCS. From Figure 6, LPCS
experiences compressive stress that ensures longer
component lifetime.
Figure 5 Deformation behaviour of Al multiple particle
impact on (a) Ti (b) Mg substrates by HPCS process.
Figure 6 Residual stress of Al multiple particle impact
on Ti and Mg substrates by LPCS and HPCS processes.
4. CONCLUSION
In conclusion, Al particles impacting on Al
substrate using HPCS process deformed intensively that
leads to greater porosity formation between the particles
by experiencing tensile stress and resulted in poor
coating quality. Meanwhile, Al on Al created denser
coating with better bond formation during impact using
LPCS. Therefore, LPCS process is better for repairing
aluminium component than HPCS process.
ACKNOWLEDGEMENT
The authors acknowledge the financial supports by
the Malaysian Ministry of Higher Education (Grant
number: FRGS20160105).
REFERENCES
[1] Cobden, R., & Banbury, A. (1994). Aluminium:
Physical Properties, Characteristics and Alloys.
Training in Aluminium Application Technologies
Lecture 1501. European Aluminium Association.
[2] Lee, J. C., Kang, H. J., Chu, W. S., & Ahn, S. H.
(2007). Repair of damaged mold surface by cold-
spray method. CIRP Annals - Manufacturing
Technology, 56(1), 577-580.
[3] Manap, A., Okabe, T., & Ogawa, K. (2011).
Computer simulation of cold sprayed deposition
using smoothed particle hydrodynamics. Procedia
Engineering, 10, 1145-1150.
[4] Lee, H., Shin, H., Lee, S., & Ko, K. (2008). Effect
of gas pressure on Al coatings by cold gas dynamic
spray. Materials Letters, 62(10-11), 1579-1581.
Proceedings of Mechanical Engineering Research Day 2018, pp. 59-60, May 2018
__________
© Centre for Advanced Research on Energy
Computational modelling for autonomous vehicle navigation using stereo vision sensor
Rostam Affendi Hamzah1,*, Melvin Gan Yeou Wei2, Nik Syahrim Nik Anwar2, Ahmad Fauzan Kadmin1,
Shamsul Fakhar Abd Gani1, Mohd Saad Hamid1, Saifullah Salam1, Nadzrie Mohamood1
1) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Faculty of Electrical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Autonomous vehicle; computational modelling; stereo vision
ABSTRACT – This paper proposes a new computational
modelling for autonomous vehicle navigation (AVN)
using a stereo vision sensor. The modelling process
requires four important stages which are matching cost
computation, cost aggregation, optimization and
refinement stage. The result will be a depth map which
will be used for the AVN. This map contains the depth
information and obstacle locations. Based on the standard
benchmarking evaluation system from the KITTI, the
proposed work produces an accurate result and very
competitive with recent published works in the KITTI
database.
1. INTRODUCTION
In recent years, the vision system has been widely
used for the autonomous vehicle navigation (AVN). In
AVN technology, the stereo vision system is a
trustworthy perception of real world obstacle detection
system for dynamic environments [1]. In order to achieve
good performance, the AVN requires accurate depth
information and detection to make the system more
reliable and safe to be used. Hence, the important part of
the computational modelling needs to be very precise and
robust against the complex structure of real environment
and weather conditions [2]. There were several reliable
methods such as SGM [3], DWBF [4] and SED [5].
These computational modelling produces a map that
contains the depth information and objects detection.
2. METHODOLOGY
Figure 1 shows a framework of the proposed work.
It starts with input stereo images, the main framework
and the output of the objects detection. The four main
stages which are explained as follows:
2.1 Matching cost computation (MCC) – Stage 1
The first stage of the propose work is using
normalised cross correlation (NCC). The NCC
effectively reduces the preliminary mismatched errors.
This stage produces preliminary differences data between
the right and left images.
NCC(𝑥, 𝑦, 𝑑) =∑ 𝐼𝑙(𝑥,𝑦).𝐼𝑟(𝑥,𝑦−𝑑)(𝑥,𝑦)∈𝑤
√∑ 𝐼𝑙2(𝑥,𝑦).∑ 𝐼𝑟
2(𝑥,𝑦−𝑑)(𝑥,𝑦)∈𝑤(𝑥,𝑦)∈𝑤
(1)
Where {Il,Ir,x,y,d,w} denote {left image, right image, x-
coordinate, y-coordinate, depth data, window size}.
Figure 1 The proposed framework.
2.2 Cost aggregation (CA) – Stage 2
The CA is the second stage of the propose
framework. This stage reduces noise from the MCC data
to make the results more consistent. The propose work in
this article uses the guided filter (GF). The GF was
developed by He et al. [6] which was capable to reduce
the noise and preserve the edges of objects matching.
The GF kernel is given by Equation 2.
𝐾(𝑝,𝑞)𝐺𝐹 (𝐼) =
1
𝑤2∑ (1 +
(𝐼𝑝−𝜇𝑐)(𝐼𝑞−𝜇𝑐)
𝜎𝑐2+𝜀
)(𝑝,𝑞)∈𝑤𝑐 (2)
Where {w,p,q,c,I,µ,σ,ε} represented by {window support
size, coordinates of (x,y), neighbouring coordinates,
center pixel of w, reference image (left input image),
mean value, variance value, constant parameter}. The GF
is used in this article due to fast processing which relies
on the image pixels and better filtering effect near the
object edges. The final equation of this stage is given by
Equation 3.
Stage 1
Normalized
Cross Correlation (NCC)
Stage 2
Guided Filter
(GF)
Input
Stage 3
Winner-takes-all
(WTA)
Stage 4
Median Filter
(MF)
Output
Depth information /
objects detection
Hamzah et al., 2018
60
𝐶𝐴(𝑥, 𝑦, 𝑑) = 𝑁𝐶𝐶(𝑥, 𝑦, 𝑑)𝐾(𝑝,𝑞)𝐺𝐹 (𝐼) (3)
where 𝑁𝐶𝐶(𝑥, 𝑦, 𝑑) is the input from MCC stage and
𝐾(𝑝,𝑞)𝐺𝐹 (𝐼) represents the GF kernel with left image as a
reference image in this article.
2.3 Depth optimization (DO) – Stage 3
The third stage of the framework is the DO which
minimizes the data selection on a location and
represented it with disparity or depth value. Generally,
the local based stereo video matching algorithm is using
Winner-Takes-All (WTA) strategy [11]. The WTA uses
the minimum value of 𝐶𝐴(𝑥, 𝑦, 𝑑) and represented the
same location with the depth value. The DO stage is
given by Equation 4:
𝑑(𝑥, 𝑦) = argmin𝑑∈𝐷𝐶𝐴(𝑥, 𝑦, 𝑑) (4)
2.4 Refinement (RF) – Stage 4
The RF is the last stage of the modelling framework.
This stage is also known as depth refinement stage which
reduces the noise and invalid depth values on the results
from the DO. Fundamentally, the used of second filtering
process at this stage is to increase the efficiency and
accuracy of the depth map. The proposed work in this
article is using the median filter which works as a second
filtering process to increase the accuracy. The final depth
result is determined from the Equation 5:
𝑑𝑓𝑖𝑛𝑎𝑙 = 𝑚𝑒𝑑 ∑ 𝑑(𝑥, 𝑦)𝑤∈(𝑖,𝑓) (5)
Where the 𝑑𝑓𝑖𝑛𝑎𝑙 is the final depth value at the coordinate
of (x,y), med denotes as a median filter and 𝑑(𝑥, 𝑦) is the
depth value from DO.
3. RESULTS AND DISCUSSION
This section presents the experimental results and
performance of the proposed work. All of the analysis
were implemented using a computational platform with
the features of CPU i7-5500, 8G RAM and GPU
GTX550. The standard benchmarking dataset have been
used from the KITTI Vision Benchmark. This dataset was
developed by Menze and Geiger [7] which consists of
200 training images. The stereo images were recorded
from real environment of autonomous vehicle navigation
using a stereo vision system. Hence, it contains very
complex and challenging images. The performance is
measured based on the bad pixel percentage of the
nonocc and all error attributes. These attributes contain
the background (bg) and foreground (fg) accuracy of
objects detection. The parameters {wncc,wgf,ε,wmed} were
used in this article with the values of
{7×7,9×9,0.0001,11×11}. Table 1 displays the
quantitative results of the proposed work and other
published methods based on the KITTI dataset for
accuracy comparison. The proposed work produces the
lowest average error with 8.71% and 19.95% for nonocc-
bg and nonocc-fg respectively. Additionally, the all-bg
and all-fg are also the lowest average error with 8.98%
and 20.04% respectively. It shows that the proposed
work in this article is accurate and competitive with other
established methods.
Table 1 The average results based on the KITTI.
Method Nonocc (%) All (%)
bg fg bg fg
Proposed 8.71 19.95 8.98 20.04
SGM [3] 11.12 18.84 11.93 20.57
iGF [4] 17.76 20.14 18.61 21.69
SED [5] 24.67 39.95 25.01 40.43
4. CONCLUSION
In conclusion, the proposed work in this article
produces accurate results based on the standard
benchmarking evaluation system. It also shows the
competitiveness of the proposed computational
modelling with other methods in Table 1. Hence, the
proposed framework in this article can be used as a
complete model for the AVN.
ACKNOWLEDGEMENT
This project is supported by Universiti Teknikal
Malaysia Melaka. (grant number:
PJP/2018/FTK(13C)/S01632).
REFERENCES
[1] Freundlich, C., Zhang, Y., Zhu, A.Z., Mordohai, P.
and Zavlanos, M.M. (2017). Controlling a robotic
stereo camera under image quantization noise. The
International Journal of Robotics Research, 36(12),
1268-1285.
[2] McGuire, K., de Croon, G., De Wagter, C., Tuyls,
K. and Kappen, H. (2017). Efficient optical flow
and stereo vision for velocity estimation and
obstacle avoidance on an autonomous pocket drone.
IEEE Robotics and Automation Letters, 2(2), 1070-
1076.
[3] Schuster, R., Bailer, C., Wasenmüller, O. and
Stricker, D. (2018). Combining stereo disparity and
optical flow for basic scene flow. arXiv preprint
arXiv:1801.04720, 1-10.
[4] Hamzah, R. A., Kadmin, A. F., Hamid, M. S., A
Ghani, S. F. and Ibrahim, H. (2018). Improvement
of stereo matching algorithm for 3D surface
reconstruction. Signal Processing: Image
Communication, 65, 165-172.
[5] Peña, D. and Sutherland, A. (2016). Disparity
Estimation by Simultaneous Edge Drawing.
Proceedings of Asian Conference on Computer
Vision, 124-135.
[6] He, K., Sun, J. and Tang, X. (2013). Guided image
filtering. IEEE Transactions on Pattern Analysis
and Machine Intelligence, 35(6), 1397-1409. ook
[7] Menze, M. and Geiger, A. (2015). Object scene flow
for autonomous vehicles. Proceedings of the IEEE
Conference on Computer Vision and Pattern
Recognition, 3061-3070.
Proceedings of Mechanical Engineering Research Day 2018, pp. 61-62, May 2018
__________
© Centre for Advanced Research on Energy
Characterising durability for solid and honeycomb plate under constant loading using Finite Element Analysis
N.M.A. Arifin*, S. Abdullah, S.S.K. Singh, A.H. Azman
Centre for Integrated Design for Advanced Mechanical System (PRISMA),
Faculty of Engineering & Built Environment, Universiti Kebangsaan Malaysia, 43600 UKM Bangi, Selangor, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Durability; Finite Element Analysis; fatigue life
ABSTRACT – This paper presents the durability
characteristics for solid and honeycomb cantilever
plates under constant loading using the Finite Element
Analysis (FEA). The aim is to investigate the impact on
fatigue life. A geometrical model for the plates of solid
and honeycomb was developed using durability
assessments aided by a computer software. A
comparison of these simulations was done following the
FEA of fatigue life.
1. INTRODUCTION
Durability is an item's ability to withhold the
intended use for an appropriate timeframe. In the
automotive industry, the FEA method is often used for
evaluating durability. Determining the material
durability is an important factor in selecting engineering
material before designing a product. Fatigue is a major
failure in the structure of a material, pertaining to the
cracks in components that occur due to repeated load
cycles. Assessments of the mechanical durability of
mechanical products rendered estimations of the S-N
stress life or the ε-N strain life based on the mean values
of the cavity load fatigue test [1].
2. METHODOLOGY
The objective of this paper is to investigate how
‘part geometry’ can influence the maximum stress found
at critical points for plates and understand how this
influences the strain concentration factor. The FEA
programs eliminate the conventional method, which
takes a long time to solve deflections and high stress
locations in complex parts [2]. Table 1 shows the
properties of the material used for comparative
simulation [3] between the solid plate and the
honeycomb plate. A fatigue assessment using the strain
life durability approaches is performed. LMS Virtual
Lab Durability is a finite element software for analysing
parameter generation [4]. In durability analysis, using
this finite element software, the load is charged at
4000N on the material of the geometrical model for
plates of solid and honeycomb (as shown in Figure 1).
The results are scheduled for maximum life expectancy
until failure against the number of cycles.
The second simulation is used to analyse fatigue
and determine the durability parameters performed with
the same structural state but a different load of 2350N.
The schematic diagram of the finite element base
durability analysis is show in Figure 2.
Table 1 Properties of SAE1045
Material Properties Value
Modulus Young (GPa) 207
Poisson Ratio 0.3
Yield Strength (MPa) 1515
Ultimate tensile strength (MPa) 1584
Figure 1 Geomertical model for plates.
Figure 2 Final element-based durability analysis.
Arifin et al., 2018
62
Fatigue analysis can be carried out using one of the
three basic methodologies: stress life, strain life and
crack growth. However, for this study, the analysis is
made based on strain life, for material damage cycle D
is 1 / Nf, where Nf is the number of cycles corresponding
to the median fatigue life. Damage to the number of n
cycles is nD = n/N and is shown in Equation (1).
(1)
Where D is the fatigue damage, ni the number of load
cycles and Ni the number of cycles until failure at the
load level.
3. RESULTS AND DISCUSSION
Figure 3 and Figure 4 show the simulation results,
i.e. the difference between 2 plates: solid and
honeycomb. As observed in table 2, the fatigue life for
solid plate is 2.14 x 105 cycles, while that for the
honeycomb plane is 2.19 x 105 cycles. For a honeycomb
plate, the damage caused by a reduced impact is
difficult to be obtained because the deformation in
honeycombs is in the form of bending of the cell walls
[5]. This result clearly indicates that the solid plate can
withstand a larger load than the honeycomb plate.
(a)
(b)
Figure 3 Fatigue life assessments for (a) solid 4000N
and (b) honeycomb 2350N
4. CONCLUSION
The performance of plates was estimated for
various dimensions of the solid and honeycomb plates.
On comparing the fatigue life gained, the cantilever
plate deflections with a force on the free end of the solid
and honeycomb plates were determined as 11.4mm and
6.64mm, respectively.
(a)
(b)
Figure 4 Translational displacements for (a) solid
4000N and (b) honeycomb 2350N.
Table 2 Comparison of life and damage characteristic
for solid and honeycomb plate.
Plate Solid Honeycomb
Loading (N) 4000 2350 4000 2350
Fatigue
damage 4.15 x 10-6 - 5.57x 10-4 4.56 x 10-6
Fatigue life 2.14 x 105 - 1.79 x 103 2.19 x 105
Translational
displacement
(mm)
11.4 6.69 11.3 6.64
Von Mises
Stress (MPa) 674 396 1440 848
REFERENCES
[1] Gates, N. R., & Fatemi, A. (2018). Multiaxial
variable amplitude fatigue life analysis using the
critical plane approach, Part II: Notched specimen
experiments and life estimations. International
Journal of Fatigue, 106, 56-69.
[2] Smith, J., Medar, P., & MR, I. A. (2014). Finite
Element Analysis and Fatigue Life Estimation of
Plate with Different Stress Levels. International
Journal of Advance Research and Innovation, 2(3),
613-617.
[3] Karthik, J. P., Chaitanya, K. L., & Sasanka, C. T.
(2012). Fatigue life prediction of a parabolic spring
under non-constant amplitude proportional loading
using finite element method. International Journal
of Advanced Science and Technology, 46, 143-156.
[4] Łukasiewicz, M., Kałaczyński, T., Liss, M., &
Kanigowski, J. (2014). The LMS Virtual. Lab
application in machines technical state
analysis. Journal of Polish CIMAC, 9(3), 59-66.
Griese, D., Summers, J. D., & Thompson, L.
(2015). The effect of honeycomb core geometry on
the sound transmission performance of sandwich
panels. Journal of Vibration and Acoustics, 137(2),
1-11.
Proceedings of Mechanical Engineering Research Day 2018, pp. 63-64, May 2018
__________
© Centre for Advanced Research on Energy
Conceptual design and computational analysis of new school desk Muhammad Ikman Ishak*, Nurul Syifa’ Ahmad Zohri, Wan Nur A’tiqah Wan Draman, Suhaimi Shahrin,
A.H.M. Haidiezul, A.K. Mohamad Syafiq, Norsyahadah Yeop Wasir, Noor Diyana Dahlan, Bakri Bakar
Faculty of Engineering Technology, Universiti Malaysia Perlis, Level 1, Block S2, UniCITI Alam Campus,
Sungai Chuchuh, 02100 Padang Besar, Perlis, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Conceptual design; finite element analysis; school desk
ABSTRACT – This study aims to design and analyse a
new school desk which may reduce the number of the
unattended desks by transforming them into a blocking
panel to separate one family area to another owing to
privacy reason during the flood season. A series of
design development and evaluation phases were
undertaken. The results depicted that the proposed
design promoted convincing stress and displacement
distributions. This could be due to the presence of
highly stable placement and position of the legs as well
as the configuration of the desk top to bear the applied
load.
1. INTRODUCTION
Malaysia regularly experiences natural disasters
which commonly cause widespread destruction and loss
of life[1, 2]. One of the frequent natural disasters
attacking Malaysia is floods and it has usually forced
the affected communities to leave their home for a safer
place. Schools are commonly used as temporary relief
centre in flood-stricken states to accommodate the
victims. It is however a significant privacy issue raised
in performing personal activities due to lack of enclosed
area available. As the victims will stay at the relief
centre for a period of times, there is a huge necessity to
have panels which separating one family area to another
and this could be achieved by the use of the unattended
school desks. Most of the student learning desks are put
left aside to provide large spaces for comfort. The
conventional school desks used nowadays are heavy in
weight which makes them hard to be lifted and moved
besides consuming large spaces for storage due to
limited downsizing mechanism. Therefore, it is an
essential for the present study to develop a new design
of school desk and analyse it via three-dimensional (3-
D) finite element analysis (FEA).
2. METHODOLOGY
There were two ethnographic studies performed at
a few selected schools to determine the user needs and
product design specifications. The findings of the first
study revealed that those schools used the same type of
desks which are mainly made of rubber wood.
Moreover, some of the desks were in poor condition
with incomplete and broken parts. Whilst, the second
study was conducted to specifically investigate the real
situations faced by the flood victims at the selected
relief centres. A thorough observation was made on the
arrangement of classroom furniture especially the desks.
The placement of the desks inside or outside of the
classroom had consumed large spaces which may
considerably be wasted.
Several interview sessions and questionnaires
circulation were also performed with relevant personnel.
Among the vital responses obtained were the desk
design must be ergonomic, durable and long-life span
materials used, attractive colour, reasonable weight,
adjustable height, stable, and simple downsizing
mechanism.
All the information collected were used in the
design development stages which are starting from
constructing List of Metrics, followed by Needs-Metrics
Mapping, Competitor Benchmarking Information,
Target Specifications, and Final Specifications. Thus,
the finalised needs for the desk design are as the
following: multipurpose, adjustable height, light-weight,
portable, having compartments, low maintenance, safety
features, durable, and attractive colour.
The determination of target product specifications
has led to the next main phase to be undertaken which is
Functional Analysis. Consequently, there were three
different concepts of desk had been successfully created
as shown in Figure 1.
Figure 1 Three-dimensional model of (a) Concept A, (b)
Concept B, and (c) Concept C.
The desk design in Concept A was developed to be
70 cm in height, 70 cm in length, and 51 cm in width. It
is completed with a flip drawer which installed using
hinges to obtain a gentle closing mechanism. All the
four legs are detachable from the base of the desk and
they may also be adjusted to set into different heights.
Whilst, Concept B shares the same main dimensions
with Concept A, however, the drawer implies the
mechanism of pull-and-push where it is embedded with
Ishak et al., 2018
64
roller and trail. For the leg, it can be flipped for a
convenient storage. Concept C, on the other hand,
having similar leg design and overall dimensions with
Concept A, and completed with an opened drawer.
Based on the results of concept evaluation process,
Concept A had recorded the highest score compared
with another two. The structural performance of
Concept A was then verified via computational 3-D FEA
in terms of stress and displacement distributions. All
FEA models were assumed to be isotropic,
homogeneous, static, and linearly elastic throughout the
analysis[3].A vertical load of 981 N was applied on the
top surface of the desk and all bottom flat surfaces of
the legs were fixed.
3. RESULTS AND DISCUSSION
It was clearly observed that Concept A has more
superior features than those of Concept B and Concept
C. The legs in Concept A may easily be adjusted to set
the desk in a few heights. The users have to pull out the
adjustable part of the leg at the bottom to increase the
length. The original height of the desk is 70 cm and it
could be increased up to 75 cm by using the adjustable
legs. It also was exhibited that Concept A promotes the
most convenient way to detach the legs from the base in
order to transform the desk into a panel by merely
rotating them and stored in the inner drawer slots. There
is a low tendency of the legs to fall out from the drawer
due to tight closing attachment provided by the hinges.
As the proposed product is intended to tackle the
privacy issue raised at the relief centres, thus, Figure 2
illustrates the steps of downsizing the desk in
transforming it into a series of blocking panel.
Figure 2 (a) Detachment of the legs. (b) Storage of the
legs. (c) Development of the blocking panel.
The results of analysis showed that the highest
stress value within the desk structure was generated in
Leg 3 with 2.22 MPa as depicted in Figure 3a. A similar
pattern of stress distribution was found for the other legs
where the top connecting parts seemed to sustain a
greater stress level.It was also clearly shown that a
wider stress concentration region developed in the
frontal legs (Leg 1 and 2) as compared to the back legs
(Leg 3 and Leg 4). The greatest stress value was
recorded within the leg body could probably be due to
high carbon steel modulus of elasticity of 200 GPa used
as compared to the wood. Moreover, the maximum
stress level generated within the legs has no tendency to
the part failure as carbon steel is known can tolerate
stresses up to 900 MPa[4]. The displacement results of
the desk structure were in contrast with the stress
outcome where the desk top part produced the greatest
displacement value (30.83 mm downwards) in
comparison to other parts as exhibited in Figure 3b.
Figure 3 The (a) stress and (b) displacement dispersions.
4. CONCLUSION
It is suggested that the new design of school desk
comprises one desk top part which completed with a
tightly-closed drawer and four easy-detachable legs for
the transformation of the desk into blocking panel. The
structural performances of the proposed desk model
were found to satisfactorily withstand the applied load
in terms of stress and displacement values and
dispersions.
ACKNOWLEDGEMENT
Appreciation is given to Faculty of Engineering
Technology, Universiti Malaysia Perlis.
REFERENCES
[1] AlBattat, A. R. & MatSom, A. P. (2014).
Emergency planning and disaster recovery in
Malaysian hospitality industry. Procedia - Social
and Behavioral Sciences, 144, 45-53.
[2] Janius, R., Abdan, K., & Zulkaflli, Z. A. (2017).
Development of a disaster action plan for hospitals
in Malaysia pertaining to critical engineering
infrastructure risk analysis. International Journal
of Disaster Risk Reduction, 21, 168-175.
[3] Ishak, M. I., Khor, C. Y., Jamalludin, M. R., Rosli,
M. U., Shahrin, S., Yeop Wasir, N., Zakaria, M. S.,
Yamin, A. F. M., Dahlan, N. D., & Wan Draman,
W. N. A. (2017). Conceptual design of automotive
compressor for integrated portable air conditioning
system. MATEC Web Conf., 97, 01040.
[4] Odusote, J. K., Ajiboye, T. K., & Rabiu, A. B.
(2012). Evaluation of mechanical properties of
medium carbon steel quenched in water and oil.
Journal of Minerals and Materials
Characterization and Engineering, 11, 859-862.
Proceedings of Mechanical Engineering Research Day 2018, pp. 65-66, May 2018
__________
© Centre for Advanced Research on Energy
Simulation of aluminum cylindrical cup in deep drawing process A.F.M. Yamin*, A.S. Abdullah, N.S. Abdullah, H. Ghafar, S.N.A.M. Halidi, H. Yusoff
Faculty of Mechanical Engineering, Universiti Teknologi MARA (UiTM) Penang,
13500 Permatang Pauh, Penang, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Deep drawing; finite element analysis; Johnson-cook
ABSTRACT – This study investigates a mechanic
behavior of cylindrical cup during deep drawing process.
The assembly consists of 78 mm diameter of Aluminum
blank with 1 mm thickness, 40 mm diameter of punch
and 42 mm cavity of conical die and blank holder. The
travel displacement of the punch is limited to 35 mm. The
inelastic strain behavior of the cup is predicted using a
quasi-static Johnson-Cook model. The response of the
simulation was compared with the experimental
methodology to address the capability of numerical
technique. Results show that good agreement in terms of
punch forces and displacements between experimental
work and numerical simulation. At 26.5 mm punch travel
displacement, the critical location of the cylindrical cup
is located near to the fillet radius of the conical die. Such
condition resulting maximum stress and inelastic strain
values which are 308 MPa and 0.7024 respectively.
1. INTRODUCTION
Deep drawing is one of the widely used sheet metal
forming process. The successful rate of drawing product
depends on several factors such as blank material, tools
geometry, blank holder force etc. Most of the industries
relies on empirical methodology to find the best process
condition in which time consuming and costly.
Alternatively, numerical technique such as finite element
(FE) simulation can be adopted as it is more cost-efficient
and reliable.
In order to analyse the dynamic process experienced
by conical cup, FE simulation can be done. The FE model
was developed according to experimental works done by
Moshksar et al. [1]. Results of the simulation and
experiment were then compared in term of punch load
and displacement. Better prediction of FE simulation was
expected. Further results in term of stress and inelastic
strain distribution were then discussed to describe the
mechanic behaviour of the conical cup during this
process.
2. MATERIAL MODEL
Since, the blank sheet experienced high and
complex deformation [2], the capability of inelastic
model to predict the plastic response is crucial. A
Johnson-Cook model was used to estimate the inelastic
strain under such condition as describe in Equation 1.
𝜎𝑦 = 𝐴 + 𝐵𝜀𝑝𝑙𝑛 (1)
where A, B and n are Johnson-Cook material constants,
εpl is equivalent plastic strain and σy is equivalent yield
stress. Material properties and Johnson-Cook parameters
of the Aluminium alloy used in this study is summarized
in Table 1. The material properties and parameters were
found by curve fitting the hardening Equation (1) with
tensile test data done by Moshksar et al [1].
Table 1 Material properties and Johnson-cook
parameters of aluminium alloy [1].
Property / Parameter Value
Elastic Modulus, E (GPa) 114
Poisson Ratio, ν 0.33
Parameters of Johnson-Cook plasticity model
Quasi-static yield stress, A (MPa) 33.56
Strain hardening modulus, B (MPa) 177.1
Strain hardening exponent, n 0.4368
3. FINITE ELEMENT MODEL OF DEEP
DRAWING PROCESS
The FE model of deep drawing process of
cylindrical cup was based on an experimental works done
by Moshksar et al [1]. The model consisted of 78 mm
diameter of Aluminium blank with 1 mm thickness, 40
mm diameter of rigid punch and 42 mm cavity of rigid
conical die and rigid blank holder. For this analysis, the
punch and conical die nose radius were set to 6 mm and
4 mm respectively.
Figure 1 Axis-symmetry model of deep drawing process
of cylindrical cup.
Due to symmetry of the geometry, loading and
boundary conditions, the model was simplified to axis-
symmetry model as illustrated in Figure 1. Throughout
the simulation, all contact friction between surfaces were
included in the model and assumed to be 0.3 for friction
coefficient. The rigid punch travel from 0 mm (top
surface of the blank) to 35 mm at negative-2 direction.
Yamin et al., 2018
66
Fixed boundary condition (U1 = U2 = UR3 = 0) was
imposed at the conical die. The blank-holder force was
set to 1 kN to prevent wrinkling of the cup in the
simulation.
4. RESULTS AND DISCUSSIONS
Results of FE simulation are presented and
discussed in terms of force-displacement response of the
punch and distribution of stress and inelastic strain in
cylindrical cup.
4.1 Punch force-displacement
Figure 2 shows the force-displacement curves
between experiment [1] and current work. The predicted
model correlated well with the experiment until 28 mm
punch displacement. However, after 28 mm tool stroke,
the numerical value of the punch force is under predicted
compared to the experiment. This is likely due to some
portion in cylindrical cup starts to experience damage and
consequently degrading the load carrying capacity of the
material. Since, the material model used in this study was
not considering the damage behaviour of the cup, the cup
will keep hardened even beyond the fracture limit.
Figure 2 Punch force-displacement curves between
experiment and simulation during deep drawing
process.
4.2 Stress and inelastic strain distribution
The effective stress and inelastic distribution of the
cylindrical cup during the deep drawing process is
illustrated in Figure 3. The stress and inelastic strain
ranges from 0 to 310 MPa and 0 to 0.75, respectively.
Due to complex deformation of the cup, the critical
location is different depending on the punch stroke. At
the beginning of the simulation, only a minor stress
occurs at the blank due to engagement between conical
die and rigid blank holder. After punch displaces, a
significant change of stress and inelastic strain
distribution across the cylindrical cup.
It is observed that, the highest stress and inelastic
strain distribution of the blank is located closely to the
die nose radius. At 26.5 mm punch displacement, the
highest value of stress and inelastic strain are predicted
which are 308 MPa and 0.7024 respectively. Notice that,
the minimum stress and inelastic strain distribution of the
cup is located near to the contact surface of the punch up
to the punch fillet radius. Since less plastic deformation
happens at this area, less chance of crack will be initiated
and subsequently crack growth.
Figure 3 Evolution of stress and inelastic strain
distribution of the blank during deep drawing process.
5. SUMMARY
The response of a cylindrical cup during deep
drawing process has been examined using FE simulation.
Results show that;
a) Good agreement between experiment and simulation
in terms of punch force and displacement,
b) At 26.5 mm punch displacement, the maximum
stress and inelastic strain at the critical point are 86
MPa and 0.032 respectively,
c) The highest stress and inelastic strain distribution of
the cylindrical cup is located near to the die nose
radius,
d) The bottom depth of the cup is the minimum stress
and inelastic strain distribution predicted in the
simulation.
REFERENCES
[1] Moshksar, M. M., & Zamanian, A. (1997).
Optimization of the tool geometry in the deep
drawing of aluminium. Journal of Materials
Processing Technology, 72(3), 363-370.
[2] Raju, S., Ganesan, G. & Karthikeyan, R. (2010).
Influence of variables in deep drawing of AA 6061
sheet. Transaction of Nonferrous Metals Society of
China 20, 2010, 1856-1862.
Punch Displacement
(mm)
Stress (MPa)
Inelastic Strain
0.0
10.0
17.5
26.5
35.0
Proceedings of Mechanical Engineering Research Day 2018, pp. 67-68, May 2018
__________
© Centre for Advanced Research on Energy
Initial validation of RULA-Kinect system – Comparing assessment results between system and human assessors
Radin Zaid Radin Umar1,*, Chai Fong Ling1, Nadiah Ahmad1, Isa Halim1, Fatin Ayuni Mohd Azli Lee1,
Nazreen Abdullasim2
1) Fakulti Kejuruteraan Pembuatan, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Fakulti Teknologi Maklumat dan Komunikasi, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Ergonomics; Rapid Upper Limb Assessment; RULA – Kinect system
ABSTRACT – Automated smart system has been an
emerging trend in Industry 4.0. A RULA-Kinect system
has been developed to automate the traditional Rapid
Upper Limb Assessment (RULA). A validation study was
conducted to compare RULA assessment between the
system and human assessors. T-Test results showed that
the RULA scores of 9 tasks are comparable (no statistical
significant differences at α = 0.05) between developed
system and 10 novice human assessors. The promising
initial results demonstrated the potential to automate
RULA process, simplifying and improve efficiency of
assessment, and thereby is in line with the direction on
Industry 4.0.
1. INTRODUCTION
Smart integrated system, which is one of the
emerging trends in Industry 4.0, has been used in many
industries such as manufacturing and healthcare [1-2].
Sensitive, cheap, and easily accessible sensors have
played a big role as an enabler of systems’ integration and
automation. Smart system results in a more efficient
management of big data, and consequently enhance the
ability to measure the overall workplace improvement
programs. Human Factors and Ergonomics, which
focuses on interactions between human and workplace
system may benefit from application of Industry 4.0.
Postural assessment of workers has been one of the
core responsibilities for occupational ergonomists.
Ergonomic assessment tools have conventionally relied
on video, paper and pencil. One of the most common
assessment tools is Rapid Upper Limb Assessment
(RULA), developed by McAtemney and Corlett [3].
RULA has been widely used for assessing work posture
in different industries and countries [4-6]. Digitalization
and automation of RULA can simplify the assessment
process and improve the assessment efficiency, which
directly aligned with the direction of Industry 4.0.
A RULA-Kinect system has been developed by the
research team to automate RULA assessment at
workplaces. The hardware consisted of Kinect Xbox 360
camera and computer, while software system consisted of
customized programming algorithm developed using
Microsoft Visual Studio. The system is sensitive enough
to capture real time postural angle data, at a maximum
rate of 30 frames per second. The algorithm calculates the
RULA scoring for postural data of each frame. Muscle
and force / load data can be directly input in the system’s
Graphical User Interface (GUI), before producing RULA
scores. The computerized calculation of the scoring
system helps to simplify the calculation task and
minimize manual computing errors. This manuscript is
aimed to describe the initial validation process of the
system, through comparison of RULA scores between the
system and the traditional method.
2. METHODOLOGY
This initial validation study compared generated
RULA scores between the RULA-Kinect system and the
conventional method on 9 different tasks from actual
workplace activities. In preparing for the validation
process, RULA-Kinect system was used to capture
RULA scores of each task at actual worksite. At the same
time, videos of each task were separately captured for
traditional RULA to be conducted by human assessors at
later times using video, paper and pencil technique.
The traditional RULA assessments were then
conducted among 10 novice human assessors with
technical and engineering background. Each assessor
received training on RULA and were given practical
exercises as well as an examination prior to being asked
to conduct actual RULA evaluations. The video of each
task captured was shown to each human assessor. An
assessor would choose specific work posture to evaluate
by pausing the video, before calculating the RULA scores
using the traditional RULA form. As a result, each RULA
assessment conducted by human assessors is independent
of each other.
The RULA scores from specific posture chosen by
the human assessors were directly compared to RULA
scores of similar posture assessed by the RULA-Kinect
system. To minimize bias, each human assessor did not
have access to RULA scores generated by the RULA-
Kinect system. Descriptive analysis was conducted to
compare RULA scores between traditional assessment
and Kinect-RULA system. In addition, two samples T-
test was conducted using SPSS statistical software
package to compare between the results.
3. RESULTS AND DISCUSSION
In general, the RULA scores between the system
and assessors are comparable across the different tasks.
Figure 1 shows an example of the posture assessed in one
of the tasks by one of the human assessors using the
traditional RULA vs. RULA-Kinect system. Example of
RULA scores from traditional and system across subjects
for one of the tasks is demonstrated in Table 1.
Radin Umar et al., 2018
68
Figure 1 Posture of technician operating lathe machine
assessed using traditional RULA (left) and RULA-
Kinect system (right) for Subject 4 (S4).
Table 1 Example of RULA scores between Traditional
and Kinect-RULA system of lathe machine operation
between human assessor and Kinect-RULA system.
Subjects
RULA score
Manual
assessment
Kinect-RULA
assessment
S1 4 4
S2 4 3
S3 3 4
S4 4 3
S5 2 3
S6 2 3
S7 2 4
S8 4 4
S9 3 3
S10 2 3
The average T-Test analysis showed no significant
differences (at α = 0.05) in RULA scores between human
assessors and RULA-Kinect system, as shown in Table 2.
This indicates the developed system is capable to produce
results of RULA assessment similar to the results
produced by human assessors. The findings showed the
same trend of potentials with similar system developed
by other researchers [7-8].
4. CONCLUSION
This study provides an early validation of the
developed RULA-Kinect system. The results
demonstrated the early potential of the system to
automate RULA assessment. However, it should be
noted that the study only involves novice human
assessors. Follow up system validation requires
assessment comparison between the system and expert
human assessors.
ACKNOWLEDGEMENT
This project is supported by Universiti Teknikal
Malaysia Melaka and Ministry of Education Malaysia
(grant number: PJP/2016/FKP-AMC/S01501).
REFERENCES
[1] Brettel, M., Friederichsen, N., Keller, M., &
Rosenberg, M. (2014). How virtualization,
decentralization and network building change the
manufacturing landscape: An Industry 4.0
Perspective. International Journal of Mechanical,
Industrial Science and Engineering, 8(1), 37-44.
[2] Bates, D. W., Saria, S., Ohno-Machado, L., Shah,
A., & Escobar, G. (2014). Big data in health care:
using analytics to identify and manage high-risk and
high-cost patients. Health Affairs, 33(7), 1123-
1131.
[3] McAtamney, L., & Corlett, E. N. (1993). RULA: a
survey method for the investigation of work-related
upper limb disorders. Applied Ergonomics, 24(2),
91-99.
[4] Gandavadi, A., Ramsay, J. R. E., & Burke, F. J. T.
(2007). Assessment of dental student posture in two
seating conditions using RULA methodology–a
pilot study. British Dental Journal, 203(10), 601-
605.
[5] Moghaddam, S. R., Khanjani, N., & Hasheminejad,
N. (2012). Evaluating risk factors of work-related
musculoskeletal disorders in assembly workers of
Nishabur, Iran using rapid upper limb assessment.
Journal of Health and Development, 1(3), 227-236.
[6] Chyuan, J. Y. (2007). Ergonomic assessment of
musculoskeletal discomfort among commissary
foodservice workers in Taiwan. Journal of
Foodservice Business Research, 10(3), 73-86.
[7] Plantard, P., Shum, H. P., Le Pierres, A. S., &
Multon, F. (2017). Validation of an ergonomic
assessment method using Kinect data in real
workplace conditions. Applied Ergonomics, 65,
562-569.
[8] Nahavandi, D., & Hossny, M. (2017). Skeleton-free
RULA ergonomic assessment using Kinect sensors.
Intelligent Decision Technologies, 11(3), 275-284.
Table 2 T-test analysis to compare RULA scores
between two methods among novice assessors.
Tasks Method Mean SD P-
value
1. Copying
documents
TR 3.6 0.56 0.074
KRS 3.2 0.42
2. Stacking
shelves
TR 5.9 1.29 0.060
KRS 6.8 0.42
3. Operating
milling
machine
TR 4.8 1.75 0.133
KRS 3.8 0.02
4. Operating
lathe machine
TR 2.8 1.23 0.180
KRS 3.4 0.52
5. Replacing
parts
underside of
vehicle
TR 5.6 1.17
0.318 KRS 5.1 0.99
6. Maintaining
engine gasket
TR 3.2 0.63 0.057
KRS 3.2 0.42
7. Printing
silkscreen
TR 3.2 0.63 1.000
KRS 3.2 0.42
8. Operating
perspex
machine
TR 4.5 0.85 0.285
KRS 4.0 1.15
9. Labelling
part
TR 4.7 0.82 0.148
KRS 5.3 0.95
Note: Sample size (n) = 10.
TR = Traditional RULA. KRS = Kinect RULA system.
Proceedings of Mechanical Engineering Research Day 2018, pp. 69-70, May 2018
__________
© Centre for Advanced Research on Energy
Yaw angle effect on the aerodynamic performance of hatchback vehicle fitted with combo-type spoiler
Kwang Yhee Chin1, Cheng See Yuan1,2,*
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Yaw angle; spoiler; aerodynamics
ABSTRACT – This study investigated the aerodynamic
performance of the combo-type spoiler in yawing
conditions using RANS-based CFD method. To date, a
majority of the tests performed on spoiler were done in
a straight-ahead driving condition. However, the effect
of spoiler is most demanded during cornering for
stability reason. The results show the spoiler is the main
contributor to the overall downforce generated on the
vehicle, but its performance deteriorates with increasing
yaw angles.
1. INTRODUCTION
Basically, rear spoiler is an aerodynamic device
added externally to the trailing edge of the roof or trunk
of a vehicle to alter the air movement around the
vehicle. Ever since its practicability on racing car has
been proven, a variety of spoiler types has been
investigated extensively in previous studies [1-2].
Generally, there are two types of rear spoilers, namely
strip-type and wing-type spoilers. Besides, there are
spoilers in the market nowadays featuring the
combination of the two configurations. This type of
spoiler will be designated as combo-type spoiler for
convenient in the discussions in this study.
The effectiveness of spoilers has been the interest
in various studies due to fuel and energy consumption,
vehicle stability, and racing speed concerns. However,
publication on the effectiveness of combo-type spoiler
made available to the public was scarce. Despite the
scarcity, there was a study on this type of spoiler
reported by Gerhardt and Kramer [3] on the
aerodynamic optimization of a Group-5 racing car.
There is a 30s reduction of lap time on Nürburgring
racing track. The results reported are produced using
wind tunnel experimentation and practical experience.
Hence there was no specific data recorded on how the
spoiler contributed to the reduced lap time.
Moreover, despite yaw angle affects the vehicles’
aerodynamic performance, as shown in previous studies
[4-5], majority of the studies on rear spoilers did not
reported on this influences. To fill in the gap, the present
study investigated the aerodynamic performance of the
combo-type spoiler in yawing conditions.
2. METHODOLOGY
2.1 Model
In order to conduct this study, the wing spoiler is
utilizing airfoil profile of NACA 0018 combine with a
strip spoiler. The angle of attack of both spoilers is 5°. This spoiler is modeled and mounted on simplified road
vehicle geometry, namely Ahmed model [6] as shown in
Figure 1. The slant angle for this model is 35°, which is
a typical angle for most hatchback vehicle. The wing
spoiler has a chord length of 69mm, which is the result
of the scale ratio of 6.61% of the length of Ahmed
model, resembling the length of wing spoiler in reality.
The yaw angles investigated are from 0° to 12°, at 4° increments.
Figure 1 Ahmed model with combo-type spoiler.
2.2 Meshing
The model was meshed with the computational
domain being discretized into unstructured and
prismatic cells. The result of grid convergence study
indicates that the mesh is sufficiently refined at around
2337141 cells. The first prismatic cell layer thickness
around the model surface was 0.5 mm. The
corresponding y+ ranges from around 1.2 to 80.
2.3 CFD setting
The effect of yaw angle on the aerodynamic forces
of Ahmed model with a wing spoiler was investigated
using numerical simulation method. The commercial
finite-volume solver, ANSYS Fluent 16 was used to
calculate all the results obtained in this study using the
Reynolds-averaged Navier-Strokes (RANS) approach.
The boundary condition for inlet was set to be
uniform flow with inlet velocity of 40 m/s. The
Reynolds number was 2.7 x 106 corresponding to the
model’s length. On the other hand, the boundary
condition for outlet was set to be having zero gauge
pressure. The walls on top and at both sides of the
domain were set as symmetry boundary condition, while
the ground and model surfaces were set as no-slip wall.
Validation of the numerical method was done
through comparing the drag coefficient Cd of the Ahmed
model at increasing yaw angle obtained by the present
CFD method and the experimental work by Bello-
Chin et al., 2018
70
Millan et al. [4]. The two curves produced are in very
good agreement (with the maximum difference of 5.2%
at 20° yaw angle). The Cd values by each method are
normalized by their respective Cd values at 0° yaw.
3. RESULTS AND DISCUSSION
Table 1 shows the force coefficients simulated in
the present study (i.e. with spoiler) compared with the
results published by Meile et. al. [7] (i.e. without
spoiler). Hence, when Ahmed model was fitted with a
combo spoiler, there was a 1% and 600% of reduction
for Cd and Cl respectively.
Table 1 Cd and Cl of Ahmed model with and without the
combo spoiler at 0˚ yaw for 𝑈∞=40 m/s.
Figure 3 shows the effect of yaw angle on the total
Cd and Cl of the model. Both aerodynamic force
coefficients increased with increasing yaw angle. Note
that the body axis system adopted by this study is such
that Cd is defined as the aerodynamic force component
parallel to the longitudinal axis of the model. The graph
demonstrated that when the vehicle is no longer
travelling in a straight path, its aerodynamic
performances deteriorate.
Figure 3 Effect of yaw angle on Cd and Cl of Ahmed
model with wing-strip-combo-type spoiler.
Furthermore, Figure 4 shows the graph of Cd and
Cl against yaw angle, indicating the effect of yaw angles
on aerodynamic forces of the spoiler. It shows favorable
trend of decreasing for Cl with a decrease of 27.7%. In
contrary, Cd showed the opposite trend with an increase
of 6.1% at 8˚ yaw and 2.29% at 12˚ yaw. The values
recorded for the aerodynamic force coefficients proved
that the spoiler does contribute to lower the lift
coefficients for vehicles, even during non-zero-yaw
conditions but failed to do the same for the drag
coefficients.
In regard to Cd, the contribution of the spoiler to
the overall Cd just accounted to at most 6.39% at 0˚ yaw
and diminished to 4.24% at 12˚ yaw. Hence, its effect to
the overall Cd was insignificant. However, as for the Cl
values recorded, the proportion contribution of the
wing-strip-combo-type spoiler to the overall Cl recorded
at least 49.5% at 12˚ yaw and at most 52.6% at 0˚ yaw.
In addition, the spoiler is the only component
contributed to creating downforce. Hence, the spoiler
was reason for the negativity of the overall lift
coefficients.
However, despite the Cl of spoiler decreases with
yaw angle (see Figure 4), the overall Cl increases
instead. This may due to the fact that other than the
spoiler, all other components of the model, especially
the roof, increasingly contributed to positive lift.
Figure 4 Effect of yaw angle on Cd and Cl of wing-strip-
combo-type spoiler.
4. CONCLUSION
This study investigated the effect of combo-type
spoiler fitted onto simplified vehicle geometry in
different yaw angles using CFD simulations with RANS
approach. The results show that adding spoiler onto
vehicle did reduce both drag and lift coefficients.
Besides, not much yaw angle effect was found on Cd,
but for Cl, the spoiler actually provides favorable
influence. However, the overall Cl still increase with
yaw angle which could be due to increasing contribution
on positive lift from other components of the vehicle
model.
ACKNOWLEDGEMENT
This project is supported by Universiti Teknikal
Malaysia Melaka (UTeM) and Ministry of Higher
Education under FRGS/1/2015/TK03/FKM/02/F00273.
REFERENCES
[1] Cheng, S. Y., & Mansor, S. (2017). Rear-roof
spoiler effect on the aerodynamic drag
performance of a simplified hatchback model.
Journal of Physics: Conference Series, 822(1), 1-6.
[2] Kodali, S. P., & Bezavada, S. R. I. N. I. V. A. S.
(2012). Numerical simulation of air flow over a
passenger car and the Influence of rear spoiler
using CFD. International Journal of Advanced
Transport Phenomena, 1(1), 6-13.
[3] Gerhardt, H. J., Kramer, C., AmmerschlÄger, T., &
Fuhrmann, R. (1981). Aerodynamic optimization
of a group-5 racing car. Journal of Wind
Engineering and Industrial Aerodynamics, 9(1-2),
155-165.
[4] Bello-Millán, F. J., Mäkelä, T., Parras, L., Del
Pino, C., & Ferrera, C. (2016). Experimental study
on Ahmed's body drag coefficient for different yaw
angles. Journal of Wind Engineering and
Industrial Aerodynamics, 157, 140-144.
[5] Meile, W., Ladinek, T., Brenn, G., Reppenhagen,
A., & Fuchs, A. (2016). Non-symmetric bi-stable
flow around the Ahmed body. International
Journal of Heat and Fluid Flow, 57, 34-47.
[6] Ahmed, S. R., Ramm, G., & Faltin, G. (1984).
Some salient features of the time-averaged ground
vehicle wake. SAE Transactions, 473-503.
[7] Meile, W., Brenn, G., Reppenhagen, A., & Fuchs,
A. (2011). Experiments and numerical simulations
on the aerodynamics of the Ahmed body. CFD
letters, 3(1), 32-39.
-0.20
0.00
0.20
0.40
0 4 8 12
Cd
an
d C
l [-]
Yaw angle, Ψ [°]
Total Cl Total Cd
-0.10
-0.05
0.00
0.05
0 4 8 12
Cd
an
d C
l [-]
Yaw angle, Ψ [°]
Cl Cd
Force Coefficients Without spoiler With spoiler
Cd 0.276 0.273
Cl 0.013 -0.065
Proceedings of Mechanical Engineering Research Day 2018, pp. 71-72, May 2018
__________
© Centre for Advanced Research on Energy
Impact of roof shape on the wind pressure difference between the Windward and Leeward Façades of a building
Zhongyu Goh2, Cheng See Yuan1,2*
1) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Cross-ventilation; CFD; roof shape
ABSTRACT – The present study investigated the impact
of different roof shapes on the natural ventilation
performance of an isolated low-rise building by using
Computational Fluid Dynamics (CFD). The Gable,
Pyramid and Shed roof were chosen for the study. The
Realizable 𝒌 − 𝜺 turbulent model was adopted in the
CFD simulations. The wind which obeyed power law
equation was set to approach the building model at an
angle perpendicular to the front building surface.
1. INTRODUCTION
Adequate air ventilation provide thermal comfort in
a building and at the same time reduce the possibility of
Sick Building Syndrome (SBS) among the residents as
stated in previous study [1]. According to an estimation
by Spiru and Simona [2], people in urban areas tend to
spend up to 90% of their time in indoor environments
especially work place. A previous study [3] mentioned
that over reliance on mechanical ventilation on a global
scale will cause enormous amount of burden towards the
environment and energy suppliers. According to Schulze
and Eicker [4], several studies had showed that natural
ventilation was able to save 17% of energy consumption
by mechanical ventilation in a targeted building at Meiji
University, Tokyo. Previous study [5] shows that the
formation of natural ventilation relies heavily on air
velocity and air flow pressure difference. For cases where
only insignificant indoor and outdoor temperature
difference occur, air flow pressure difference determines
the performance of natural ventilation as stated in
previous study [6]. Hence, the objective of this study is
to investigate the impact of roof shape on the natural
ventilation performance of a Building.
2. METHODOLOGY
In this study, ANSYS FLUENT 16.0 commercial
software was used to perform simulations. The building
model length, L, width, W and height, He was fixed at
6.6m, 6.6m and 6m respectively while maximum roof
height for all roof shapes was 7.65m. The computational
domain was 126 m (Length), 54 m (Width) and 54 m
(Height). The distance between the front façade of the
building model and domain inlet was 42m. The domain
was discretized into 1.4 million tetrahedral elements with
4 hundred thousand nodes and finer elements were
adopted on regions around the building model.
Realizable 𝑘 − 𝜀 turbulent model was used and all
the transport equation were discretized using a second-
order upwind scheme. The COUPLED algorithm was
used for pressure-velocity coupling. At the inlet, the
mean streamwise velocity of the approaching flow
obeyed a power law with an exponent of 0.25 as shown
in Equation 1.
UPL(z) = Un (𝑍
𝑍𝑛)𝛼 (1)
Where, Zn is reference height, Un is reference velocity at
reference height and 𝛼 is power-law index.
Zero static pressure was applied for the domain
outlet. Both of the sides and the top of the domain were
applied symmetry boundary conditions. The wall
functions on the ground were altered for roughness
height, ks of 1.0mm and roughness constant, Cs of 1.0.
Grid Independent Test was done to identify the optimum
grid quality for simulation. Gable, Pyramid and Shed roof
were simulated with wind approach angles perpendicular
to the building model front façade. The velocity profile
for each roof shapes were plotted. The wind pressure
difference between two points, P1 and P2 which situated
on windward and leeward façade respectively, 1.5m from
the ground were compared for each roof shape to identify
the best natural ventilation performance potential.
3. RESULTS AND DISCUSSION
Figure 1 shows three dimensional modelling of the
building models with each roof shape: Gable, Pyramid
and Shed.
3.1 Wind velocity profile
Figure 2 shows the velocity profile of the mean
velocity, U1 of each roof shape generated by CFD. All
roof shapes generated significant backflow at the region
around leeward façade as seen in wind velocity profile at
x/He = 1.0 in Figure 2. Overall wind velocity profile for
Gable configuration was the largest and it indicated the
wind had the least resistance flowing over the building
surface. On the windward side, Shed configuration
created smallest wind velocity profile which shows the
airflow around the building went through highest
resistance. While for the leeward side, Pyramid
configuration generated the largest airflow resistance
around the building due to the smallest wind velocity
profile.
.
Goh et al., 2018
72
Figure 1 Gable, Pyramid and Shed Three Dimensional
Modelling.
Figure 2 Gable, Pyramid and Shed Wind Velocity
Profile Comparison.
3.2 Wind pressure difference
The average wind pressure data were extracted from
the windward and leeward surface of the each building
model. The wind pressure data obtained were tabulated
in Table 1 to calculate the wind pressure difference. Both
Gable and Shed roof produced similar wind pressure
measurements on the windward façade. It was suggested
that same roof cross sectional area contributed to the
similarity. Meanwhile, the Shed roof had 43% wind
pressure difference with the Pyramid roof on the
windward facade. On the leeward facade, the wind
pressure difference between the Gable, Pyramid and
Shed roof did not exceed 9%. At the end of the result,
Shed roof came up with highest wind pressure difference
with Gable roof in second and Pyramid roof in the last
place. The wind pressure difference between windward
and leeward which the windows shall be located plays an
important role in encouraging natural ventilation in the
interior of the building. Thus, by comparing the wind
pressure difference, it is possible to analyse the natural
ventilation performance impact by roof shape.
Table 1 Wind Pressure Data for Each Roof Shapes.
Windward
Facade
(Pa)
Leeward
Facade
(Pa)
Pressure
Difference,
∆P (Pa)
Gable 1.9795 -1.2909 3.2704
Pyramid 1.0822 -1.2894 2.3716
Shed 1.9242 -1.3964 3.3206
4. CONCLUSION
As a conclusion, the roof shape significantly affect
the pressure difference between the windward and
leeward facades of a building. The percentage of
difference is up to about 40% between Shed and Pyramid
configurations. The results suggest that the Shed roof has
the highest natural ventilation performance potential,
followed by Gable roof and lastly, Pyramid roof.
ACKNOWLEDGEMENT
This project is supported by Universiti Teknikal
Malaysia Melaka (UTeM) and Ministry of Higher
Education under FRGS/1/2015/TK03/FKM/02/F00273.
REFERENCES
[1] Norhidayah, A., Chia-Kuang, L., Azhar, M. K., &
Nurulwahida, S. (2013). Indoor air quality and sick
building syndrome in three selected
buildings. Procedia Engineering, 53, 93-98.
[2] Spiru, P., & Simona, P. L. (2017). A review on
interactions between energy performance of the
buildings, outdoor air pollution and the indoor air
quality. Energy Procedia, 128, 179-186.
[3] Omrani, S., Garcia-Hansen, V., Capra, B. R., &
Drogemuller, R. (2017). Effect of natural
ventilation mode on thermal comfort and
ventilation performance: Full-scale
measurement. Energy and Buildings, 156, 1-16.
[4] Schulze, T., & Eicker, U. (2013). Controlled natural
ventilation for energy efficient buildings. Energy
and Buildings, 56, 221-232.
[5] Burnett, J., Bojić, M., & Yik, F. (2005). Wind-
induced pressure at external surfaces of a high-rise
residential building in Hong Kong. Building and
environment, 40(6), 765-777.
[6] Yuan, C. S. (2007). The effect of building shape
modification on wind pressure differences for cross-
ventilation of a low-rise building. International
Journal of Ventilation, 6(2), 167-176.
Proceedings of Mechanical Engineering Research Day 2018, pp. 73-74, May 2018
__________
© Centre for Advanced Research on Energy
Development of machining simulation application using visual basic programming in NX CAM system environment
Mohamad Hafiz Mohamad, Muhammed Nafis Osman Zahid*
Faculty of Manufacturing Engineering, Universiti Malaysia Pahang, 26660, Pekan, Pahang, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Visual Basic programming (VB); Computer-Aided Manufacturing (CAM); simulation
ABSTRACT – This paper presents the integration of
visual basic programming in NX Computer-Aided
Manufacturing (CAM) system with 4th axis milling
simulations as machining routines. A customized
graphical user interface (GUI) was developed to
simplify the simulation process planning and reduce the
dependency on user’s experience while developing the
machining program in NX CAM system. The simulation
operation construction code was recorded by using
journaling tool that available in NX CAM. Then the
code is modified in visual basic program to build
custom machining simulation applications. The results
indicate that the developed programs are capable to
optimize 4th axis machining simulation by reducing the
processing steps and time with minimum process
planning tasks.
1. INTRODUCTION
Simulation in manufacturing is defined as the
imitation routines of the selected operation in real
processes for pre-evaluation purposes. The behavior of
machining processes and response parameter is studied
by developing a simulation model for cutting operation
before proceed into real machining. The simulation is
carried out to identify the issue or problem at early stage
of machining [1]. It is important to investigate the
machining processes by simulating the operation to
ensure the result is similar as expected. Simulation can
be carried out in CAM software or direct on the
machine control panel. Besides that, simulation analysis
also permits the user to identify the effect of changes
and act as a design tool to develop a new system [2]. A
part of that, it is also can be used to analyze different
machining scenarios, not only rapidly but also without
any risk, damage and waste of workpiece. In Computer-
Aided Manufacturing (CAM), the efficiency of planning
task and process execution are crucial factors to develop
machining routines for simulation purpose. Process
planning in CNC machining is directly influence the
processing time, procedure, operator skill and operation
cost [3].
2. METHODOLOGY
In this study, a visual basic programming language
was used as a basis for graphical user interfaces (GUI)
development and machine code customization. The
developed GUI was embedded with journaling code
generated from NX CAM system. Journaling is a tool
that available inside NX CAM where it allows user to
record, edit and replay back all the interaction during
NX sessions [7]. The instruction tasks during machining
program developments are recorded separately with
different parameter setup for each operation. The
recorded codes are translated into visual basic script
files. Then it was modified to remove the code
stickiness. The modification allows user to input certain
parameters such as, cutting orientation, cutting
parameter (spindle speed, feed rate, and depth-of-cut),
tool diameter and workpiece diameter. Two different
GUI programs were developed to handles different
simulations and operation analysis. Roughing operation
GUI used to build roughing machining operation and
aims to remove large amounts of material rapidly from
the workpiece to produce part geometry close to the
desired shape. Finishing operation GUI construct
finishing machining operation and the purpose is to
achieve final geometry of the machined parts with a
good surface finish. In order to illustrate the overview of
simulation operation, the differences in process planning
between conventional and proposed approaches can be
seen as shown in Figure 1. Manual approach is
conventional methods that are typically used to build
machining programs and requires a significant user
intervention and efforts to execute the repetitive
processes [8]. Some parameters and settings for each
operation need to be changes in order to run simulations
with a few constant parameters. In this study, certain
level of automation is expected to be embedded in the
operation build-up routines. The proposed approach is
an improvised method by developing a custom
application to build a NX CAM program with the
addition of several automation elements.
Figure 1 Comparison of simulation approaches.
Mohamad and Osman Zahid, 2018
74
The simulation will run continuously without
requiring user intervention between the geometry in 4th
axis machining operation. Consequently, if there has
much geometry in one operation, the program will
automatically loop the simulation to the next operation
efficiently. Journaling program codes are recorded
through the tool in NX CAM starts from “Create
Geometry” (level-2) to “Create Operation” (level-5).
Some parameters values that need to be set in each level
have been simplified and grouped in GUI program
window. Through this method, the proposed approach
has managed to reduce the processing step from 7 steps
to just 4 steps.
3. RESULTS AND DISCUSSION
The proposed simulation application was validated
by machining several 3D CAD models as shown in
Figure 2. Table 1 reveals the results of the proposed
approach in assisting the process planning of machining
program developments in NX CAM systems.
Figure 2 3D CAD simulation models.
Table 1 Results of processing time required to construct
a machining operation programs using conventional and
proposed approach.
No. Total
operations
Processing time
(min) Impro
vemen
t
rate
(%)
Conventio
nal
approach
(min)
Proposed
approach
(min)
1 4 Roughing
2 Finishing 16.78 2.98 82.2%
2 4 Roughing
2 Finishing 14.48 2.38 83.5%
3 4 Roughing
2 Finishing 16.27 2.45 84.9%
4 4 Roughing
2 Finishing 15.82 2.80 82.3%
4. CONCLUSION
This paper has discussed the integration of visual
basic programming in NX Computer-Aided
Manufacturing (CAM) system for the application of 4th
axis machining. From the study, the developed
applications managed to execute, control and develop
machining simulation programs efficiently with
minimum processing steps. The results show that
proposed approach successfully reduces processing time
up to 84.9% of improvement rate.
ACKNOWLEDGEMENT
We acknowledge with gratitude to Ministry of
Higher Education Malaysia for providing a financial
support under Research Acculturation Grant Scheme
(RDU151406), which realize this research project.
REFERENCES
[1] Anderberg, S. (2009). A study of process planning
for metal cutting (Doctoral dissertation, Chalmers
Reproservice).
[2] Banks, J., Carson, J. S., & Nelson, B. L. DM
Nicol.(2010). Discrete-Event System Simulation.
5th ed., Prentice Hall, 2010.
[3] Frank, M. C. (2007). Implementing rapid
prototyping using CNC machining (CNC-RP)
through a CAD/CAM interface. Proc. Solid Free.
Fabr. 112–123.
[4] Osman Zahid, M. N., Case, K., & Watts, D. M.
(2017). Rapid process planning in CNC machining
for rapid manufacturing applications. Int. J. Mech.
Eng. Robot. Res., 6(2), 118–121.
[5] Moi, M. B. (2013). Web Based Customized
Design (Master's thesis, Institutt for
produktutvikling og materialer).
[6] Zhao, J., Zhang, D. H., & Chang, Z. Y. (2011). 3D
model based machining process planning.
Advanced Materials Research, 301, 534-544.
[7] Siemens. (2014). Getting Started with SNAP, no.
October. Siemens Product Lifecycle Management
Software Inc.
[8] Turley, S. P., Diederich, D. M., Jayanthi, B. K.,
Datar, A., Ligetti, C. B., Finke, D. A., ... & Joshi,
S. (2014, January). Automated process planning
and CNC-Code generation. In IIE Annual
Conference. Proceedings (p. 2138). Institute of
Industrial and Systems Engineers (IISE).
Proceedings of Mechanical Engineering Research Day 2018, pp. 75-76, May 2018
__________
© Centre for Advanced Research on Energy
Study the electromyography (EMG) technique for rehabilitation purpose Nurul Muthmainnah Mohd Noor1,*, Mohamad Saddam Mohamad Baharudin2
1) Faculty of Mechanical Engineering, Universiti Teknologi MARA, Cawangan Pulau Pinang,
Kampus Permatang Pauh, 13500 Pulau Pinang, Malaysia. 2) Faculty of Mechanical Engineering, Universiti Teknologi MARA Shah Alam, 40500 Selangor, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Electromyography; rehabilitation; applications; sensor; muscle
ABSTRACT – Electromyography (EMG) is a
technique for evaluating and recording the electrical
produced by skeletal muscles. This method can be used
as a communication tool between human and machine.
The detection, processing and classification of EMG are
very desirable because it allows a more standardized
and precise evaluation of the neuropsychological,
rehabilitation and assistive technological findings. This
paper aims to study and analyze the EMG from arm
muscle by using the shield EMG-EKG circuit. The
EMG signal was acted as an input, and then the data
was acquired by interfaced with MATLAB software.
There are five subjects who involved in the experiment
with 2 males and 3 females. These EMG signals will be
used as an algorithm to the user for any rehabilitation
purpose.
1. INTRODUCTION
Electromyography is the discipline that deals with
the detection, analysis, and use of the electrical signal
that emanates from contracting muscles [1]. It is a
technique for evaluating and recording the electrical
activity produced by skeletal muscles and study the
muscular function through the generated electrical
signal that produce when muscle has any activity such
as contraction or movement of muscle. The small
electrical currents are generated by muscle fibers prior
to the production of muscle force. These currents are
generated by the exchange of ions across muscle fiber
membranes, a part of the signaling process for the
muscle fibers to contract. EMG can be achieved by
using a highly and precision technology device is called
as electromyograph and produce the EMG signal it
called as an electromyogram. The electromyograph
purpose is to detect the electrical potential that
generated by the muscle cells when these cells are
electrically or neurologically activated [2]. The potential
difference that obtained in the muscle fiber can be
registered in the surface of the human body through
surface electrodes due to the biological tissues
conducting properties [3]. It is the electrical expression
caused by neuromuscular activation during muscular
contraction, depicting the current detected at the specific
location that is produced by the ionic flow that across
the muscle fibre membranes and transmitted through
intervening tissues. The motor unit is the most
elementary functional unit of a muscle, generating a
motor unit action potential (MUAP) when activated.
Repeated continuous activation of motor units generates
motor unit action potential trains (MUAPT) that are
superimposed to form the EMG signal [4]. Collecting
EMG signals emanated from the human body using
electrodes has become a routine procedure both in
rehabilitation engineering and medical research.
2. METHODOLOGY
Figure 1 shows the flowchart in collecting the
EMG signals. The EMG data was collected using Shield
EMG-EKG by Olimex (Olimexino-328). This board
was interfaced with the Arduino IDE software.
Therefore, the result of signal was displayed through
MATLAB software. By using this circuit, it is also
already built-in with high voltage protection, filter and
rectifier and smoothing to avoid the noise signal. The
placements of three Ag/AgCl electrodes were attached
to the arm as shown in Figure 2, where at point A, point
B (as a reference) and point C for ground. The selection
this electrode because it is common used in detection of
EMG signals. It has electrolyte for conducting the
electrical signal produce by skeleton muscle.
Figure 1 Flowchart of collecting EMG signals.
2.1 Experiment setup
For this project, the experiment was setup as
shown in Figure 3. The three electrodes were attached to
Muthmainnah and Mohammad Baharudin, 2018
76
the user’s arm. The EMG signal was reflected at point A
meanwhile at point B and C for reference only. Then the
activity of muscle signal either in moving or resting was
recorded by EMG-EKG circuit board and the signal was
displayed on the scope on the MATLAB software. There
are 5 readings were taken for each subject.
Figure 2 Placements of electrodes.
Figure 3 Experiment setup for collecting EMG data.
3. RESULTS AND DISCUSSION
3.1 EMG signal
Figure 4 shows the EMG signal that was displayed
on the scope in MATLAB software. This result shows
the muscles when doing activity and without activity (at
rest). The value of EMG was measured in mV scale
unit. When muscle in rest condition, the EMG signal
was shown the smooth line compared to the muscle has
doing activity. The applied pressure on the muscle or
movement of muscle also will increase the electrical
potential difference, so that the value of EMG signal
was varied but in constant.
Figure 4 The EMG signal data for muscle activity.
In this study, there are two experiments were
carried out. i) Using the hand gripped (the mass ~
21kg). In this experiment, each the subject should grip
the gripper for 1 s and rest and grip again for another 1
s. ii) Using the 3kg mass. For the second experiment,
the subject should make the movement of muscle with
the mass of 3 kg in 2 s. Figure 5 and Figure 6 show the
reading of EMG signals for five subjects. Subject 1 and
Subject 2 are female and the rest are male. From the
figures, the reading for both experiment are slightly
same for each other. In this experiment, each subject
was needed to grip the gripper for 100 s. Then rest for
100 s and continued with other 100s. Therefore from the
graph, the highest value for each subject was same.
From this result, it can develop the algorithm for
controlling any external devices such as prosthetic arm
or powered wheelchair.
Figure 5 The EMG signal data for muscle activity-hand
gripped.
Figure 6 The EMG signal data for muscle activity –
mass 3kg.
4. CONCLUSION
In a conclusion, the main objective for this study
has been successfully developed and achieved in order
to collect the data from arm muscle using the EMG
circuit board by Shield EMG-EKG and Arduino
Olimexino-328. The result of EMG signal that obtained
has been compared with the different moving activities
of muscles. Most of each data signals have their own
pattern of waveform of the EMG signals. The type these
patterns of the EMG signal with several subjects is
important to formulate the algorithm for controlling
external devices.
REFERENCES
[1] De Luca, C.J. (2006) Electromyography
Encyclopaedia of Medical Devices and
Instrumentation, John Wiley Publisher, 98-109.
[2] David, M. Blake, Procedure Offered for Lexington
Neurology General Services. Lexington, KY.
[3] De la Rosa, R., Alonso, A., Carrera, A., Durán, R.,
& Fernández, P. (2010). Man-machine interface
system for neuromuscular training and evaluation
based on EMG and MMG signals. Sensors, 10(12),
11100-11125.
[4] Soderberg, G. L., & Cook, T. M. (1984).
Electromyography in biomechanics. Physical
Therapy, 64(12), 1813-1820.
Proceedings of Mechanical Engineering Research Day 2018, pp. 77-78, May 2018
__________
© Centre for Advanced Research on Energy
Preliminary study of future lightweight aircraft structure during survivable crash event
A.S. Abdullah1,2,*, A.F.M. Yamin1, H. Ghafar1, H. Yusoff1, R. Othman1, S.N.A.M Halidi1, N.S. Abdullah1, H. Sharudin1
1) Faculty of Mechanical Engineering, Universiti Teknologi MARA, Cawangan Pulau Pinang,
Kampus Permatang Pauh, 13500 Pulau Pinang, Malaysia. 2) ARTeC, Universiti Teknologi MARA, Cawangan Pulau Pinang,
Kampus Permatang Pauh, 13500 Pulau Pinang, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: ABAQUS/Explicit; aircraft crashwothiness; 7075-T6
ABSTRACT – The objective of the study was to develop
a reliable finite element model (FEM) of the main
structure of fuselage during survivable crash event using
ABAQUS/Explicit. For validation, the Johnson-Cook
material model for the aluminum alloy 7075-T6 was
compared to the experimental stress-strain curves. The
results of the crash simulation indicated that this
preliminary FEM development can be reliably used for
further crash simulation that involves advanced material
as the main structure of the fuselage.
1. INTRODUCTION
New type of advanced materials namely fiber metal
laminates (FML) has started to be used in the high-
performance lightweight aircraft and plenty of
commercial aircrafts have already incorporated high
strength-to-weight ratio composite in the main structure
[1,2]. As new materials being introduced, the structural
integrity and crashworthiness of the aircraft during
survivable crash must not be compromised. To simulate
the crash of such a complex structure that consisted of
many structural parts, one can start with a preliminary
simulation on the crashworthiness of the main structure,
namely fuselage frames which carries the main load of
the aircraft. This study focused on establishing a reliable
finite element (FE) model of crash simulation of the
fuselage frame using ABAQUS/Explicit.
2. FINITE ELEMENT MODEL
The frame structure in this paper is based on the
previous study done by Abdullah [3]. In that paper, the
preliminary study was on semi-monocoque frame of
Boeing 737 but with the absence of the passenger’s floor
as shown in Figure 1a. The absence of the floor
underestimated the integrity of the structure. Figure 1b
shows the fuselage frame with passenger’s floor being
studied in this work. The upper outer radius of the frame
is 1.88 meter meanwhile the lower outer radius is 1.80
meter. Boeing Company website and Niu provided the
details of the frame’s geometry. [1,4].
The fuselage structure was discretized by shell
element, S4R with hourglass reduced integration. The
connection between floor and frame was modelled as tie
connection which represented rigid connection without
damage model. Eight mass elements were modelled on
the passenger’s floor at the location of the seat tracks to
represent the loading due to the weight of passengers and
seats. The mass element was 53 kg at each point. The
impact velocity with downward direction of the fuselage
was set as 9 m/s. The vertical component of the impact
velocity represents the vertical impact speed during a
survivable crash scenario [5]. With this impact velocity,
18 kJ of impact energy was generated during the crash
event. Penalty contact method was applied to define
contact between fuselage structures to the rigid surface
that represent the ground. The same contact method
defines the contact between the fuselage structures
themselves.
Figure 1 Fuselage frame.
2.1 Material model
Both frame and passenger’s floor of the fuselage
were made of aluminium alloy (AA) 7075-T6. Johnson-
Cook plasticity and damage models were used to model
AA 7075-T6 for this simulation. In order to ensure
reliable plastic model, the Johnson-Cook plasticity model
was compared with the experimental stress-strain curve
as shown in Figure 3 [6]. Table 1 provides the material
properties, plastic parameters and damage parameters of
AA 7075-T6.
3. RESULTS AND DISCUSSION
The approximation of plastic hardening using
Johnson-Cook plastic model was verified by the
experimental stress-strain curves as illustrated in Figure
2. This indicates that the material model used is reliable
to simulate the crash as material properties play a big role
in determining how the structure deform, fail and absorb
the impact energy.
Figure 3 illustrates the crash of the fuselage frame.
It started with high stress concentration at the lower part
of the semi-monocoque frame as in Figure 3a in which
immediately followed by buckling at that area as shown
in Figure 3b. The buckling progresses at various spots as
shown in Figure 3c, d and then followed by large
structural displacements and rotations which causes the
fuselage frame to sustain an amount of crush magnitude.
At 194 ms, it is observed that the fuselage starts to
rebound as the elastic strain energy is released from the
(a) (b)
Abdullah et al., 2018
78
fuselage structure. The crushing distance observed is
1.71 meter which is almost half of the original height of
the fuselage as shown in Figure 4. Important note is that
the crushing stopped at the height of the passenger’s
floor. Figure 5 illustrates the energy balance of the
fuselage during crash.
Figure 2 Validation of Johnson-Cook approximation by
experimental data [3].
Figure 3 (a) t = 2.5 ms; high stress concentration at few
spots, (b) t = 3.75 ms; buckling initiated at high stress
concentration spots, (c) t = 10 ms; buckling progress.
(d) 32.5 ms, buckling progress at various spots, (e) t =
125 ms; further crushing, (f) t = 194 ms, fuselage
rebound.
Figure 4 Crushing of fuselage frame.
Figure 5 Energy balance of the fuselage frame.
Table 1 Material properties and related Johnson-Cook
plastic and damage parameters [6]. Material properties
Density [kg/m3] 2934
Young’s modulus [GPa] 59.8
Poisson’s ratio 0.33
Parameter Notation
Hardening parameters
Static yield stress [MPa] A 473
Strain hardening modulus [MPa] B 210
Strain hardening exponent n 0.3813
Strain rate coefficient C 0.033
Thermal softening exponent m 1.56 Melting temperature [K] θmelt 750
Damage parameters
d1 0.3714
d2 -0.1233 d3 -1.9354
d4 0.0101
4. SUMMARY
(a) The crushing distance is 1.71 meter
(b) During crash, few spots within the fuselage
frame experience high-stress concentration
then followed by buckling. Then the buckling
progresses and caused large structural
deformation and rotation.
(c) The crushing stopped just before the
passenger’s floor indicating that the safety
envelop for the occupants has not been
penetrated.
REFERENCES
[1] 2016 Boeing, “The 787 Dreamliner family,” 1995.
[Online]. Accessed: May. 1, 2015.
[2] A. S, “Technology | Airbus, a leading aircraft
manufacturer,” airbus. [Online]. Accessed: May 1,
2015.
[3] Abdullah, A. S., Yamin, A. F. M., Ghafar, H., Halidi,
S. N. A. M., Ab Hamid Pahmi, M. A., & Ismail, N.
I. (2017). Structural integrity of aluminum alloy
7075-T6 fuselage frame during crash event.
Proceedings of Mechanical Engineering Research
Day, 2017, 84-85.
[4] Niu, C. (1988). Airframe structural design:
practical design information and data on aircraft
structures. Conmilit Press.
[5] Abromowitz, A., Smith, T. G., & Vu, T. (2000).
Vertical drop test of a narrow-body transport
fuselage section with a conformable auxiliary fuel
tank onboard. 2000.
[6] Zhang, D. N., Shangguan, Q. Q., Xie, C. J., & Liu,
F. (2015). A modified Johnson–Cook model of
dynamic tensile behaviors for 7075-T6 aluminum
alloy. Journal of Alloys and Compounds, 619, 186-
194.
-10,000
0
10,000
20,000
0 50 100 150 200Ener
gy b
alan
ce
(Jo
ule
)
Time (ms)
KineticPlasticElastic
0
100
200
300
400
500
600
700
0.01 0.03 0.05 0.07 0.09 0.11 0.13
Equ
ival
ent
stre
ss (
MP
a)
Equivalent plastic strain
Equivalent stress vs equivalentplastic strainJ-C approximations
(a) (b)
(c) (d)
(e) (f)
Proceedings of Mechanical Engineering Research Day 2018, pp. 79-80, May 2018
__________
© Centre for Advanced Research on Energy
Fluid structure interaction simulation of large deformation and added-mass effect using OpenFOAM
Mohamad Shukri Zakaria1,2,3,*, Haslina Abdullah5, Kamarul Arifin Ahmad3,4,*
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3) Faculty of Engineering, Universiti Putra Malaysia, 43400 Serdang, Selangor, Malaysia. 4) Mechanical Engineering Department, College of Engineering, King Saud University,
P.O. Box 800, Riyadh 11421, Saudi Arabia. 5)Faculty of Mechanical & Manufacturing Engineering, Universiti Tun Hussein Onn,
Parit Raja, 86400 Parit Raja, Johor, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Fluid structure interaction (FSI); Arbitrary Lagrangian Eulerian (ALE), OpenFOAM, simulation
ABSTRACT – Large deformation and added mass
effect (i.e., closed density between vessel and blood)
dominate the Fluid Structure Interaction (FSI)
simulation instability on the numerical algorithm.
Therefore, in this article, we provide numerical study of
such problem using FSI partition coupling approach.
The FSI solver were develop using CFD Open source
solver library OpenFOAM, in which Arbitrary
Lagrangian Eulerian (ALE) Finite Volume Method
(FVM) solver for fluid with automatic mesh motion and
updated lagrangian FVM solver is used for elastic solid.
The robustness of the solver as well as its accuracy is
compared to the monolithic solution of classical FSI
benchmark test case of flapping flag attaches on the
back of cylinder. Finally, implications of the results and
future research directions on efficient FSI simulation
especially in heart valve applications are also presented.
1. INTRODUCTION
Stability of the numerical algorithm in FSI for
elastic structure is challenging. It required either strong
coupling or the use of a monolithic FSI model due to the
onset of a stability problem known as added mass effect.
This term is normally applied in literature to indicate the
instability of the FSI algorithm when the density of the
solid is almost similar to that of the surrounding
viscous, incompressible fluid. This phenomenon is not
observed in other branches of engineering problems,
such as aero-elasticity, but this issue is prominent in
biomechanics applications, such as heart valves. In this
scenario, the density of the leaflets is almost similar to
that of blood [1-2]. Furthermore, the very low moment
of inertia of the valve leaflets, owing to the added mass
effect, induces numerical instabilities in the fluid-
structure interaction algorithm that compromises its
stability and prevents it from converging [3].
Conventional partition approaches cannot be
employed when added mass effect is significant, even
under strong coupling algorithm, the coupling is not
stable. However, with the aid of Aitken relaxation
technique, the number of iteration needed for
convergence per time step could be significantly
reduced [3]. Alternatively, the monolithic approach
must be applied to prevent numerical instability.
Therefore, in this paper, the monolithic ALE method is
using to simultaneously solve the governing equations
of the flow and structure with a single solver together
with Aitken relaxation algorithm
2. METHODOLOGY
2.1 Numerical method
In this article, continuity and momentum equation
for in- compressible laminar flow is solved using in
ALE formulation. The large solid deformation solid is
govern by the Piola-Kirchoff stress-strain formulation. It
is important to noted that the largest solid movement is
adjacent to the moving boundary, potentially leading to
local deterioration in mesh quality. Ideally, largest
deformation should be confined to the internal part of
the mesh, where it causes less distortion. This can be
achieved by prescribing variable diffusivity in the
Laplacian. To compute this grid velocity while
considering the conservation principle and avoiding the
loss of mass and momentum, the space conservation law
(SCL) is applied.
The collocated 2nd order finite volume method on
a deforming mesh is used for the space discretization of
fluid flow equation, while 2nd order backward scheme
is used for temporal discretization. The solution
procedure make use of segregated procedure on a
moving mesh with Pressure-implicit with splitting of
operators (PISO) pressure-velocity coupling algorithm
used in fluid flows.
Number of mesh used in this study is 20940 with
time step of 0.001s. This setup yield CFL < 0.2 since we
have noted that larger stability parameters while
increasing proportionally the time step size also
increased the initial error of the fluid- structure
interaction algorithm that needed more iterations to
converge; this resulted in an overall increase of the CPU
time.
3. NUMERICAL RESULTS
The following test case extends the previously
detected classical stationary flow around cylinders in
Zakaria et al., 2018
80
CFD to an FSI test by attaching an elastic flag to the
back of the cylinder. This FSI benchmark problem was
posed by Turek and Hron [4] and is challenging because
two main issues are involved that require fully coupled
FSI, namely, the finite deformation of a beam structure
during fluid interaction and the added mass effect
attributed to densities of solid and fluid, which are close
to each other.
The sketch problem is depicted in Fig. 3. The
prescribed parabolic velocity profile at the inlet channel
is derived as follows:
Where is average inflow velocity. Three test cases were
considered, namely, FSI1, FSI2, and FSI3, as shown in
Table 1. Each case has its own challenge: FSI1 is
subject to added mass effect; FSI2 possesses a large
deformed structure; and FSI3 combines both features.
However, for feasibility of current method, only the
most complex case, namely FSI3 will be tested.
Figure 3 Geometry of the test case for the flow around a
cylinder with elastic flap. The dimensions applied are
those used in [4].
Table 1. Parameter setting for FSI test cases.
FSI1 FSI2 FSI3
Solid density 1000 1000 1000
Poisson coefficient 0.4 0.4 0.4
Young modulus 1.4e-6 1.4e-6 1.4e-6
Fluid density 1000 1000 1000
Fluid viscosity 1 1 1
Average inflow velocity 0.2 1 2
Reynolds number 20 100 200
(a) (b)
(c) (d)
Figure 4. Fluid-structure simulation with large
deformations in the solid and added mass effect for (a)
1s (b) 1.5s (c) 2.075s and (d) 2.152s.
As we can see on result despite in Fig. 4, the solid
undergoes nonlinear and severe deformation, which in
turn affect the fluid flow. This simulation is well agreed
by the experimental work of Turek and Hron [4]. Both
fluid velocity and stress distribution is clearly seen. One
also can easily obtain the displacement of the tip of the
solid.
4. CONCLUSION
Numerical results are presented which proved
strong implicit coupling approach can handle large
deformation and added mass effect problem which is
inherent in FSI systems such as tissue heart valves. The
used of single solver implement in OpenFOAM is
feasible for this type of problem. Although the test case
used is somewhat trivial with heart valve model, it could
easily have extended as long as the method is validated.
As expected, moving mesh in ALE method is
suffered in term of computational time. The CPU time
taken is 20hrs for only 2.54s real time. However, the
result could be used as an additional benchmarking of
current author for further code development such as
using fix grid method.
5. REFERENCES
[1] Zakaria, M. S., Ismail, F., Tamagawa, M., Aziz, A.
F. A., Wiriadidjaja, S., Basri, A. A., & Ahmad, K.
A. (2017). Review of numerical methods for
simulation of mechanical heart valves and the
potential for blood clotting. Medical & Biological
Engineering & Computing, 55(9), 1519-1548.
[2] Zakaria, M. S., Ismail, F., Wiriadidjaja, S., Aziz,
M. F. A., Tamagawa, M., Basri, A. A., & Ahmad,
K. A. (2017). Numerical simulation of mechanical
heart valve with coherent vortex shedding in
OpenFOAM. Proceedings of Mechanical
Engineering Research Day, 2017, 68-69.
[3] Borazjani, I., Ge, L., & Sotiropoulos, F. (2008).
Curvilinear immersed boundary method for
simulating fluid structure interaction with complex
3D rigid bodies. Journal of Computational
Physics, 227(16), 7587-7620.
[4] Turek, S., & Hron, J. (2006). Proposal for
numerical benchmarking of fluid-structure
interaction between an elastic object and laminar
incompressible flow. Fluid-Structure
Interaction (pp. 371-385). Springer, Berlin,
Heidelberg.
Proceedings of Mechanical Engineering Research Day 2018, pp. 81-82, May 2018
__________
© Centre for Advanced Research on Energy
Preliminary results of numerical simulation in pre-combustion chamber (PCC) engine
M.N. Norizan1,*, S.I. Sazman1, M.I.M. Zainudin1, F.A. Munir1,2,*, A.R. Saleman1,2, F. Idral1,2, M.S. Yob1,2
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia 2) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia
*Corresponding e-mail: [email protected]
Keywords: Numerical simulation; pre-combustion chamber
ABSTRACT – There are two combustion chambers in
the spark ignition engine. First, the main combustion
chamber and another one is pre-combustion chamber
(PCC). This setting is utilized in Compressed Natural Gas
(CNG) engine to improve the combustion efficiency.
Generally, the size of PCC is much smaller than the main
combustion chamber. The air is going through the main
chamber during compression stroke and enters the PCC.
At the same time, mixture of gaseous is being directed
into combustion chamber and ignited using the spark
plug. In this research, concept designs of PCC were
developed in Design Modeler in ANSYS Fluent Version
16.0. The simulation of combustion of compressed
natural gas is performed by utilizing the similar software.
Each design of the PCC was analyzed for their flow
behavior for the velocity vector, kinetic rate of reactions
and static temperature.
1. INTRODUCTION
The focus alternative fuel in this study is natural gas
with add pre-combustion chamber in internal combustion
engine. Natural gas is also gained from fossil fuel, which
is similar to liquid fuel and diesel. However, natural gas
can be considered as renewable energy due to the recycle
of methane gas [1-2]. However, the power obtained from
natural gas is not as high as liquid fuel. Hence, that is why
pre-combustion chamber must be added in internal
combustion engine to increase engine power [3].
The main substance inside the natural gas is
methane. Methane can be considered as renewable fuel
and apply in internal combustion engine. The
implementation of natural gas in internal combustion
engine produces low emission but at the main problems
is that it produced lower performance [4, 5]. This study
focused on the numerical simulations of different design
of PCC for a single cylinder engine.
A PCC air-fuel ratio (A/F) was performed at each
engine operating condition by varying the fuel supply
pressure to the PCC. An analytical model was developed
to estimate the mass flow between the PCC and main
cylinder, and mass flow out of the PCC through a sample
extraction apparatus [6].
More specifically, a novel staged pre-chamber
yielded significant reductions in NOx and total
hydrocarbon emissions by promoting stable pre-chamber
and main chamber ignition under fuel-lean conditions [7].
Precise fuel control was also critical when balancing low
emissions and engine efficiency.
2. RESEARCH METHODOLOGY
Figure 1 shows the CFD simulation step of this
study.
Figure 1 CFD simulation step.
2.1 Meshing
Meshing is probably the most important part in any
of the computer simulations because it can show drastic
changes in obtained results. Meshing is defined as giving
the geometry multiple nodes or grid generation. Meshing
is performed with a variety of tools & options available
in the software. Figure 2 shows the completed meshing
on the geometry. The results are calculated by solving the
relevant governing equations that are continuity,
momentum, energy and species equations. The pattern
and relative positioning of the nodes affects the
computational time.
Figure 2 Geometry meshing.
2.2 Solver setup: Boundary condition
There are several boundary conditions that need to
be set up. This boundary condition is important because
the fluid flows need to be defined with parameters desired
with certain magnitude. Three boundary conditions are
set which are the velocity inlet, pressure outlet and
combustor wall.
2.3 Solver setup: Post processing
The final simulation step is the post-processing.
Post-processing step enables the handling of the
Norizan et al., 2018
82
information that is obtained from the computed solution
and also observations and examinations of the outcome
can be simply conducted. After finishing the computation,
the ‘Result’ tab from the workbench window is selected
where the CFD-Post window appeared. There a lot of
data that can be sort out from the simulation
3. RESULTS AND DISCUSSION
The velocity vector of the flow at each of designs of
pre-combustion chamber were analysed to determine the
maximum velocity vector of the flow as depicted in
Figure 3.
Figure 3 Velocity vector.
This kinetic rate of reaction as shown in Figure 4
helps the combustion rate to be faster and at the right
place. From Fluent, the flame propagation can be seen
clearly and the position of flame propagation started can
be found during analyzation of kinetic rate of reaction
contours.
Figure 4 Kinetic rate of reaction.
From ANSYS post-processing module, static
temperature contour can be extracted as depicted in
Figure 5. and analysed by researcher which one has the
highest static temperature.
Figure 5 Static temperature.
4. CONCLUSIONS
When the inlet velocity magnitude has the highest
velocity vector, then the combustion will be faster and
better performance of spark ignition engine can be
obtained. Apart from that, the kinetic rate of reaction is
related to the flame propagation. From the result, the
starting point of flame propagation which is the ignition
point can be discovered. The higher the air inlet velocity
magnitude, the longer distance of flame propagation
happened from inlet wall.
REFERENCES
[1] Alvarez, C. E. C., Couto, G. E., Roso, V. R., Thiriet,
A. B., & Valle, R. M. (2017). A review of
prechamber ignition systems as lean combustion
technology for SI engines. Applied Thermal
Engineering, 128, 107-120.
[2] Korakianitis, T., Namasivayam, A. M., & Crookes,
R. J. (2011). Natural-gas fueled spark-ignition (SI)
and compression-ignition (CI) engine performance
and emissions. Progress in energy and combustion
science, 37(1), 89-112.
[3] Esfahanian, V., Salahi, M. M., Gharehghani, A., &
Mirsalim, M. (2017). Extending the lean operating
range of a premixed charged compression ignition
natural gas engine using a pre-
chamber. Energy, 119, 1181-1194.
[4] Tahir, M. M., Ali, M. S., Salim, M. A., Bakar, R. A.,
Fudhail, A. M., Hassan, M. Z., & Muhaimin, M. A.
(2015). Performance analysis of a spark ignition
engine using compressed natural gas (CNG) as
fuel. Energy Procedia, 68, 355-362.
[5] Barzegar, R., Shafee, S., & Khalilarya, S. (2013).
Computational fluid dynamics simulation of the
combustion process, emission formation and the
flow field in an in-direct injection diesel
engine. Thermal Science, 17(1), 11-23.
[6] Gingrich, J. W., Olsen, D. B., Puzinauskas, P., &
Willson, B. D. (2006). Precombustion chamber
NOx emission contribution to an industrial natural
gas engine. International Journal of Engine
Research, 7(1), 41-49.
[7] Crane, M. E., & King, S. R. (1992). Emission
reductions through precombustion chamber design
in a natural gas, lean burn engine. Journal of
Engineering for Gas Turbines and Power, 114(3),
466-474.
Proceedings of Mechanical Engineering Research Day 2018, pp. 83-84, May 2018
__________
© Centre for Advanced Research on Energy
Thermal-stress analysis of the corrugated metal gasket under high temperature load
W.S. Widodo1,2,*, M.A. Choiron2, Hambali Arep@Ariff1, Mohd Shukor Salleh1
1) Faculty of Manufacturing Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Faculty of Engineering, University of Brawijaya University, Jalan Veteran, Ketawanggede, Lowokwaru,
Ketawanggede, Kec. Lowokwaru, Kota Malang, Jawa Timur 65145, Indonesia.
*Corresponding e-mail: [email protected]
Keywords: Corrugated metal gasket; thermal stress; finite element
ABSTRACT – The paper presents thermal stress
simulation and analysis of the corrugated metal gasket
under high temperature environment (200oC – 400oC).
High temperature condition resulting thermal stresses
and contact stresses which affects sealing performance of
the gasket. Sealing performance of the gasket depends on
the contact stress and contact length between the gasket
and flanges. The study consists of development of CAD
model and simulation of the finite element model in
ANSYS software to obtain deformation, temperature and
stress distribution of the gasket. The simulations show
that contact stresses and contact length between the
gasket and flanges increase when temperature increases.
1. INTRODUCTION
Gaskets as seal elements play an important role in
piping connection in some industries such as food
industries, automotive, oil and gas industries, as well as
in pressure vessel piping systems and distillation column
installations. Saeed et al. [1] studied about corrugated
metal gaskets named Super Seal Gasket using Finite
Element (FE) simulation and Taguchi method to optimize
the geometric model and the dimensions of the gasket.
Choiron et al. [2] focused on design optimization based
on contact stresses of the corrugated metal gasket.
Choiron et al. [3] also evaluated the effects of contact
area to leak performance of the corrugated metal gasket
based on FE simulation and helium leak test experiment.
Ushijima et al. [4] examined the deformation model of
the corrugated metal gasket and put highlight on
geometrical transformation on flat area and convex area
of the gasket. Nurhadianto et al. [5] investigated the
effects of forming process during manufacturing phase to
contact stresses of the gasket.
The research investigated the effect of high
temperature (200oC – 400oC) to the contact stress and
contact length between the corrugated metal gasket and
flanges. Previous researches [1-3] proved that sealing
performance of a gasket mostly affected by contact
stresses and contact length between a gasket and flanges,
higher contact stresses and contact length will improve
sealing performance of a gasket. Investigation of this
problem consists of several steps. Firstly, the governing
equation were developed to obtain the correlation
between thermal loads and stresses in the gasket and its
effects to the gasket deformations. Secondly, the physical
model of the gasket and flanges were recreated in CAD
software. Lastly, the CAD model was transferred as a FE
model in ANSYS software. By applying appropriate
boundary conditions, mechanical behavior of the gasket
as response to bolts pretension and thermal load can be
simulated and investigated.
2. METHODOLOGY
In this study, a corrugated metal gasket is assembled
in the bolted flanges piping connection where a hot fluid
flows, as shown in the Figure 1.
Figure 1 General gasket installation.
The temperature of the fluid is simulated within the
range of 200oC – 400oC. The gasket and flanges are made
of stainless steel SUS304 with material properties as
shown in Table 1.
Table 1 Properties of SUS304.
Tensile strength, σty (N/mm2) 398.5
*Modulus of elasticity, E (N/mm2) 210000
*Conductivity, k (W/m.K) 16.2
*Coeff of thermal expansion, α (μm/m.K) 17.3
*ANSYS database.
2.1 Finite element model
The CAD model was designed in Solidwork and
imported into ANSYS workbench as a 2D model to
reduce the memory space and time required to perform
the simulation. The meshing/discretization process for
the gasket-flanges model was done by using ANSYS
Mechanical workbench software. The finite element
model for the gasket-flanges consists of 459 nodes and
325 (PLANE 77) elements. The boundary condition is
shown in Figure 2.
Widodo et al., 2018
84
3. RESULTS AND DISCUSSION
3.1 Bolt-pretension loading
Figure 3 shows the deformed model and stress
distribution of the gasket during tightening the bolt of the
gasket-flange assembly. The deformed model obtained
from this simulation then was transformed as a new
model for the steady-state heat transfer simulation
process.
Figure 2 Boundary condition for FE model.
Figure 3 Gasket deformation and stress distribution
under bolt pretension loading.
3.2 Thermal loading
Figure 4 shows temperature distribution of the
gasket-flange assembly. The temperature distribution
from this simulation then be used as thermal loading and
be converted as structural loading in the gasket.
Figure 4 Temperature distribution at gasket-flanges.
3.2 Thermal stress
Figure 5 shows the gasket deformation as an effect
of thermal load at 200oC. Temperature for the model was
varied from 200oC–400oC to obtain behaviour of gasket
deformation when temperature increased.
Figure 5 Deformation of the gasket at T= 200 oC.
Table 2 presents measured contact length and
maximum stress at three regions of gasket (A, B, C)
obtained from the simulation with temperatures varies
from 200oC to 400oC.
Table 2 Contact length and contact stress at different
temperatures.
Temp (oC) Contact length (mm) Contact stress (MPa)
A B C A B C
200 0.998 1.085 1.002 386.3 301.4 298.7
250 1.065 1.092 1.013 393.6 304.7 301.5
300 1.069 1.107 1.015 394.8 309.2 303.9
350 1.073 1.112 1.019 399.6 311.5 307.2
400 1.081 1.124 1.026 404.6 314.7 309.8
It can be seen from here that the region B
experienced the highest contact length while region A
experienced the highest stress. The stress even exceeds
the ultimate strength of the material thus indicate that the
material starts to fail/crack, but this only occurred in the
very small area, whereas the others area experienced
much lower stress.
4. CONCLUSION
Simulation results show that contact length and
contact stress between the gasket and flanges increase as
temperature increases. Previous researches proved that
higher contact length and contact stress improve the
sealing performance of the gasket. This implies that
higher temperature will increase the sealing performance
of the gasket. On the other hand, higher temperature also
produces higher stress which exceeds the tensile strength
of the material, as a result at certain temperature the
gasket will be crack and the sealing performance will be
fail.
ACKNOWLEDGEMENT
The authors would like to thank Universiti
Teknikal Malaysia Melaka for the continuous support on
this research project.
REFERENCES
[1] Saeed, H. A., Izumi, S., Sakai, S., Haruyama, S.,
Nagawa, M., & Noda, H. (2008). Development of
new metallic gasket and its optimum design for
leakage performance. Journal of Solid Mechanics
and Materials Engineering, 2(1), 105-114.
[2] Choiron, M. A., Haruyama, S., & Kaminishi, K.
(2011). Optimum Design of New 25A-size Metal
Gasket Considering Plastic Contact
Stress. International Journal of Modeling and
Optimization, 1(2), 146-150.
[3] Choiron, M. A., Haruyama, S., & Kaminishi, K.
(2011). Simulation and experimentation on the
contact width of new metal gasket for asbestos
substitution. International Journal of Aerospace
and Mechanical Engineering, 5(4), 283-287.
[4] Ushijima K., Haruyama S., Kaminishi K., Chen
Dai-Heng (2011). Study on deformed mode of thin
metal gasket based on experimental and FEM.
Proceeding of the 8th International Conference on
Innovation and Management, 302-306.
[5] Nurhadiyanto, D., Choiron, M. A., Haruyama, S., &
Kaminishi, K. (2011). Contact Width Evaluation of
New 25A-size Metal Gasket Considering Forming
Effect. 8th International Conference on Innovation
and Management, 296-301.
Proceedings of Mechanical Engineering Research Day 2018, pp. 85-86, May 2018
__________
© Centre for Advanced Research on Energy
A significant improvement of vehicle body responses using limited state feedback controller for active suspension system
Mohd Hanif Harun1,3,*, Ridhwan Jumaidin2,3, Adzni Md Saad1,3, Fauzi Ahmad1,3, Mohd Zakaria Mohamad Nasir2,3
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Vehicle suspension; quarter vehicle model; limited state feedback controller
ABSTRACT – This paper investigates the performance
of passive and active vehicle suspension. It introduces
two degrees of freedom (2-DOF) mathematical
equations of a quarter vehicle mode which was
governed based on Newton’s second law. Two
difference control strategies were applied on this quarter
vehicle suspension namely Proportional-Integrated-
Derivative (PID) controller and Limited-State-Feedback
(LSF) controllers. A 5 cm road bump is considered as a
road disturbance in this study. The results show that
both controllers were able to reduce unwanted vehicle
body motions in terms of vehicle body acceleration and
suspension travel. From the study, it can be concluded
that the performances of both controllers in improving
ride quality of vehicle body are much better compared
to the passive system.
1. INTRODUCTION
Comfortable is one of hundred important criteria to
be considered by all manufacturer around the world in
designing a vehicle. The functions of vehicle suspension
is to minimize the vehicle body vibration caused by
road surface, to support a vehicle body and for vehicle
handling. Many researchers and academicians nowadays
doing a research on vehicle suspension focusing on
improving vehicle ride and handling qualities [1-3].
Due to lack of attitude control and high cost in
implementing active suspension on vehicle, it inspires
and motivates the researchers to consider the use of low
cost of active suspension system by simplifying an
outer-loop controller of vehicle suspension. According
to Kumar [4], the electronically controlled active
suspension system can potentially improve ride comfort
as well as stability of the vehicle. This paper I organized
as follow: The first section is an introduction on vehicle
suspension system, followed by quarter vehicle
modelling and control structure of active suspension in
second and third sections. The forth section presents the
performance evaluations of active suspension system
and the last section contain some conclusion.
2. QUARTER VEHICLE MODEL
The quarter vehicle model of the passenger vehicle
consists of a 1/4 sprung mass (vehicle body) mass
connected to unsprung mass and presented as a 2-DOF
system. The sprung mass is represented as a plane and
allowed to displace in vertical direction, while the
unsprung mass is allowed to bounce vertically with
respect to the sprung mass. Figure 1 shows the hematic
diagram of 2-DOF quarter vehicle model and Table 1
shows the parameters of vehicle suspension.
Figure 1 Schematic diagram of a 2-DOF quarter vehicle
model.
Table 1 Suspension parameters.
Definition Value
Sprung mass ( )sm 466.5 kg
Unsprung mass ( )um 49.8 kg
Spring stiffness ( )sk 5,700 N/m
Damping coefficient ( )dc 850 Ns/m
Wheel spring stiffness ( )tk 192,000 N/m
Wheel damping coefficient ( )tc 12,000 Ns/m
Force balance analysis on sprung and unsprung masses
can be written as:
asudsuss Fxxcxxkzmb
−−+−= )()(....
(1)
)()()(....
utdsusu xwkxxcxxkzmsub
−+−−−−=
ad Fxwc
u+−+ )(
.
(2)
Where uuusss xxxxxx ,,,,, are the sprung and
unsprung mass displacements, velocities and
accelerations, while w is the road input disturbance.
Harun et al., 2018
86
3. CONTROL STRUCTURE
3.1 PID controller
Figure 2 below shows the structure of the
controller connected to the plant. Plant is a system that
is going to be controlled and the controller is the
selected control method that provides the excitation for
the plant.
Figure 2 Basic controller design of PID.
An error signal (e), also known as the different
between desired input and actual output will be sent to
the controller.
3.2 LSF controller
LSF is a control scheme that use two simple gains
control the vehicle body displacement and damper
displacement. There is an advantage of using this
controller such as it can control the vehicle body and
damper displacement simultaneously.
Figure 3 Basic controller design of LSF.
4. RESULTS AND DISCUSSION
4.1 Vehicle body acceleration
Figure 4 is the time response of vehicle body
vehicle acceleration. From the figure, it shows that the
PID and LSF control can reduce the magnitudes with
the best performance as an active suspension system
with an active system to become stable is faster than
passive with 3.1 m/s2 for LSF and 3.9 m/s2 for PID
controller against 4.5 m/s2 for passive system.
4.2 Suspension travel
It can be clearly seen that the suspension travel of
active suspension with PID and LSF controller are much
higher than the passive suspension system as shown in
Figure 5. This is due to the quick force applied by the
actuator in response to the signal from the controller, but
the vibrations become stable are faster than the passive
system due to the initial stage of transient vibration
which increasing of displacement amplitude [4].
Figure 4 Vehicle body acceleration.
Figure 5 Suspension travel.
5. CONCLUSION
In conclusion, the performance of LSF strategy
shows the superiority over PID scheme and passive
system in a quarter car model. This is due to an ability
of LSF controller to control both body and damper
displacement. Both controllers are able to reduce or
eliminate both amplitude and settling time of unwanted
body motions.
ACKNOWLEDGMENT
The authors would like to thank Universiti
Teknikal Malaysia Melaka for the financial support
provided through Journal Publication Incentive Grant
(project number JURNAL/2018/FTK/J00004).
REFERENCES
[1] Emam, A. S. (2015). Fuzzy Self Tuning of PID
controller for active suspension system. Advances
in Powertrains and Automotives, 1(1), 34-41.
[2] Kashem, S. B. A., Roy, S., & Mukharjee, R. (2014,
May). A modified skyhook control system (SKDT)
to improve suspension control strategy of vehicles.
Informatics, Electronics & Vision (ICIEV), 2014,
1-8.
[3] Ahmad, F., Hudha, K., & Harun, M. H. (2009).
Pneumatically actuated active suspension system
for reducing vehicle dive and squat. Jurnal
mekanikal, 28(1), 85-114.
[4] Kumar, M. S., & Vijayarangan, S. (2007).
Analytical and experimental studies on active
suspension system of light passenger vehicle to
improve ride comfort. Mechanics, 65(3), 34-41.
0 1 2 3 4 5 6 7 8 9 10-4
-3
-2
-1
0
1
2
3
4
5
time (sec)
Vehic
le B
ody A
ccele
ration (
m/s
2)
Body Acceleration of Peugeot 206 with different type of Suspension System
Passive
PID
LSF
0 1 2 3 4 5 6 7 8 9 10-0.06
-0.04
-0.02
0
0.02
0.04
0.06
0.08
time (sec)
Suspensio
n T
ravel (m
)
Suspension Travel of Peugeot 206 with different type of SUspension System
Passive
PID
LSF
Proceedings of Mechanical Engineering Research Day 2018, pp. 87-88, May 2018
__________
© Centre for Advanced Research on Energy
The effect of spin and friction on reaction forces in a soccer ball impact: A computational study
Mohd Hasnun Arif Hassan*, Zahari Taha
Innovative Manufacturing, Mechatronics & Sports Lab (iMAMS), Faculty of Manufacturing Engineering,
Universiti Malaysia Pahang, 26600 Pekan, Pahang, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Soccer ball; finite element; ball spin; coefficient of friction
ABSTRACT – This study investigates the effect of spin
and coefficient of friction on the reaction forces in a
soccer ball impact. The analysis was conducted by means
of finite element (FE) method. A validated soccer ball FE
model was used. Angular ball velocity and the coefficient
of friction between the ball and the impacted surface were
varied in the simulations. The normal and friction forces
were the output parameters of the simulations. The results
show that the normal force is neither influenced by the
coefficient of friction nor the angular velocity, while the
friction force is influenced by all impact variables. This
study shows that a soccer ball impact can be influenced
by several external factors.
1. INTRODUCTION
Soccer is one of the most popular sports in the
world. The total number of people who are actively
involved in soccer are about 4% of the total population of
the world [1]. This is a contact sport, which involves
many contacts between players and also the ball. One of
the popular manoeuvre in soccer is ball heading. The
player uses his/her head to hit the ball to another
teammate or hit the ball into the net to score a goal.
Several studies have suggested that soccer heading might
cause traumatic brain injury [2-5]. Measurement of brain
motion experimentally is very difficult, thus a
computational method like the FE analysis is a very
useful tool in studying such case.
During the game, the ball impacts various types of
surface. Simulation studies for instance by means of FE
method have been conducted to study and understand the
mechanics of the ball impact [6-8]. The output of the
simulations are the ball impact characteristics, namely
the contact time, rebound velocity and maximum
longitudinal deformation. The input of the simulation, on
the other hand, is the inbound velocity. Nonetheless,
these studies employed only linear ball velocity without
any spin, which is
The objective of this study is to quantify the
reaction forces due to the ball impact using the FE
method. In the abovementioned studies, the only input
was the inbound velocity. In this study, we introduce a
spin to the ball by defining both linear and angular
velocities. This study serves as an initial study before
simulating the soccer heading manoeuvre of a spinning
ball. Several parameters, namely the inbound velocity,
the ball spin and the coefficient of friction were varied to
investigate the influence of each parameter on the
reaction forces. The following sections detail the
methodology of the study and the results obtained.
2. METHOD
This investigation was conducted computationally
using a validated finite element model of soccer ball. The
model was developed by Taha and Hassan [8], which
utilises a composite sphere shell geometry. The soccer
ball model comprises of two layers, namely the inner
rubber bladder and an outer composite panel. The
material properties of each layer were obtained from
Price et al. [6].
The model was validated by Taha and Hassan [8]
through a dynamic impact test and a drop test on a force
platform that measures the reaction force upon impact.
Nonetheless, their simulation was performed without any
ball spin. Thus, this study aims to extend the work of
Taha and Hassan [8] by introducing spin to the ball before
it impacts the rigid surface as shown in Figure 1.
Figure 1 Soccer ball impacting rigid floor with spin.
To define the velocity of the ball, a reference point
was created at its centre of mass. The reference point was
coupled to the ball’s surface using the structural coupling
method. The structural coupling method couples the
translation and rotation of the reference node to the
translation and the rotation motion of the coupling nodes.
By doing this, the motion of the ball is governed by the
motion of the reference point. Therefore, the velocity and
the angular velocity of the ball were defined at the
reference point, which in turn will cause the ball to have
the same motion. The velocity of the ball was varied from
9, 12, 15 and 18 m/s, while the angular velocity was
varied from 5, 10, 25 and 50 rad/s.
The impact between the ball and the rigid surface
involves a contact. Thus, a general contact was defined.
The friction between the contact surfaces was defined
using penalty method. This method permits some relative
motion of the surfaces when they should be sticking. The
coefficient of friction was then defined as 0 (frictionless),
0.3, 0.6, and 0.9, and varied one at a time for the
parametric analysis.
3. RESULTS AND DISCUSSION
Figure 2 and Figure 3 show the magnitude of total
forces due to frictional stress for different ball spins and
Hassan and Taha, 2018
88
coefficients of friction. It is seen that the friction force is
influenced by all parameters.
Figure 2 Friction forces with respect to the ball spins.
Figure 3 Friction forces with respect to the coefficient of
friction.
The larger the ball spin and coefficient of friction,
the larger the friction force. In Figure 2, it is observed that
the 5 and 10 rad/s spins have different friction forces
pattern compared to that of 25 and 50 rad/s spin.
Nonetheless, the contact time is almost the same in all
cases, regardless of the magnitude of the angular velocity.
Figure 4 Normal forces with respect to the ball spins.
Figure 5 Normal forces with respect to the coefficient of
friction.
Further, Figure 4 and Figure 5 depict the magnitude
of total forces due to contact pressure for different ball
spins and coefficients of friction. It was found that the
ball spin and coefficient of friction were found to not
influence the normal forces as shown in both figures.
4. CONCLUSIONS
This study uses the FE method to investigate the
influence of the ball spin and coefficient of friction on the
friction force and normal force. The friction force was
found to be influenced by all parameters. The larger the
magnitude of these parameters, the larger the resulting
friction force. The magnitude of normal force, as
obtained from the simulations, was not influenced by the
ball spin and coefficient of friction. This study provides
an initial insight of the impact characteristics of a
spinning ball. Future study will be looking into the effect
of ball spin on the head motion in a soccer heading
manoeuvre.
ACKNOWLEDGEMENT
This study was supported by the Universiti
Malaysia Pahang (UMP) internal grant RDU1603106.
REFERENCES
[1] FIFA.com. (2006). Big Count - FIFA.com.
https://www.fifa.com/mm/document/fifafacts/bcof
fsurv/bigcount.statspackage_7024.pdf
[2] Lipton, M. L., Kim, N., Zimmerman, M. E., Kim,
M., Stewart, W. F., Branch, C. A., & Lipton, R. B.
(2013). Soccer heading is associated with white
matter microstructural and cognitive
abnormalities. Radiology, 268(3), 850-857.
[3] Matser, J. T., Kessels, G., Lezak, M. D., & Troost,
J. (2001). A dose-response relation of headers and
concussions with cognitive impairment in
professional soccer players. Journal of Clinical and
Experimental Neuropsychology, 23(6), 770–774.
[4] Naunheim, R., Bayly, P., Standeven, J., Neubauer,
J., Lewis, L., & Genin, G. (2003). Linear and
angular head accelerations during heading of a
soccer ball. Medicine and Science in Sports and
Exercise, 35(8), 1406–1412.
[5] Zhang, M. R., Red, S. D., Lin, A. H., Patel, S. S., &
Sereno, A. B. (2013). Evidence of cognitive
dysfunction after soccer playing with ball heading
using a novel tablet-based approach. PloS One,
8(2), 1–4.
[6] Price, D., Jones, R., & Harland, A. (2006).
Computational modelling of manually stitched
soccer balls. Proceedings of the Institution of
Mechanical Engineers, Part L: Journal of
Materials Design and Applications, 220(4), 259–
268.
[7] Price, D., Jones, R., & Harland, A. (2007).
Advanced finite-element moxzddelling of a 32-
panel soccer ball. Proceedings of the Institution of
Mechanical Engineers, Part C: Journal of
Mechanical Engineering Science, 221(11), 1309–
1319.
[8] Taha, Z., & Hassan, M. H. A. (2017). A reaction-
force-validated soccer ball finite element
model. Proceedings of the Institution of
Mechanical Engineers, Part P: Journal of Sports
Engineering and Technology, 231(1), 43-49.
Proceedings of Mechanical Engineering Research Day 2018, pp. 89-90, May 2018
__________
© Centre for Advanced Research on Energy
Statistical process control as a traceability tools for industry 4.0 Norazlin Nasir1,2,*, Ahmad Yusairi Bani Hashim1, Mohamad Hafidz Fazly Md. Fauadi1, Teruaki Ito2
1) Faculty of Manufacturing Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Institute of Technology and Science Tokushima University,
Minami-Jyousanjima 2-1, Tokushima-shi, Tokushima, 770-8506, Japan.
*Corresponding e-mail: [email protected]
Keywords: Industry 4.0; statistical process control; traceability
ABSTRACT – Industry 4.0 brings integrative concepts
where all the manufacturing processes including
resources are connected and automatically information
exchange. The main issues are controlling and
traceability the original sources to ensure the data gain
is eminently trustable. In manufacturing applications,
statistical process control seems fit to be implemented
by showing the process trend. Based on that, engineers
are able to evaluate the process control for just-in-time
and the process change can be done for improvement.
Eventually, at the end of this paper the SPC as a
Traceability tool is proposed which may ease and
expedite the decision-making process in production.
1. INTRODUCTION
Internet of things (IoT) becomes a pillar of
Industry 4.0 (I4). In IoT environment, trust is a best
practice to overcome the data leakage issues. The
trackback of file-centric approached has increased the
data traceability, real-time record, accountability and
transparency in IoT environment [1].
Traceability is an information path between two
different domain platform that allows changes, decision
and eases the maintenance. Traceability also a
transformation model that enables to keep track the
information evolvement and its original source by
providing the trackback record in different-time [2]. The
traceability tool makes the man/operator to expedite the
decision-making without any delay. Traceability i.e for
the residential user may not be a concerned instead of
the high-quality timing source. However, when
manufacturing required automation process that worked
under internet network to transmit the information
between synchronized distinct and devices that correlate
with time, the real-time source is necessary [3].
2. STATISTICAL PROCESS CONTROL AS A
TRACEABILITY TOOL
Since the statistical process control (SPC) is
typically used in industry and it’s also become a reason
to be explored deeply in this study. The p-chart is one of
the others charts in SPC which represented by a ratio of
the number of non-conforming units to the total number
of units in the sample set. The probability having the
non-conforming units in the set of 𝑛 is known as 𝑝 and
the experimental data can be evaluated from Equation 1
where in the fraction of the non-confirming units, 𝑝𝑖 in
𝑖𝑡ℎ sample set, 𝑘 is known as the number of the sample
set.
=
=
k
i
ipk
p
1
1 (1)
The p-chart can be prepared with constant n and
graphically represented the confidence limits with: pCL =
(2)
( )
−−=
n
pppLCL
13 (3)
( )
−+=
n
pppUCL
13
(4)
2.1 Boundaries condition
The graphical p-chart representation also able to
show the lack of control scenarios which may occur
under various circumstances that evidenced of the un-
behavior trend. ASTM International suggested nine
conditions of un-behavior trend that can be detected if
one or more points for un-behavior trend are violated
[4]. However, only five conditions are considered for
further investigation. The conditions are:
a) Any single point, 𝛽 or more those beyond 3𝜎
which is upper control limit (UCL), or lower
control limit (LCL), 𝛽 ± 3𝜎.
b) Two consecutive points, 2𝛽 beyond 2𝜎 from the
mean or centreline either above or below, 2(𝛽 ±𝜎).
c) Four out of five consecutive points, 4𝛽 beyond 1𝜎
from the mean or centreline either above or below,
4𝛽 ± 1𝜎.
d) Six or more points, 6𝛽 that consecutive higher or
lower with no change in direction, 6𝛽 ± �̅�.
e) Eight or more points, 8𝛽 in a row on one side of
the centerline, 8𝛽 ± �̅�.
3. METHODOLOGY
The pseudocode as shown in Table 1, is chosen as
a unified language to share the procedure in SPC chart
development. n≤100 as a set of data input for the SPC
that needed to be analyzed. The size of data must not
more than 100 for optimal evaluation. To plot the SPC
chart, the value of LCL must be less than 0. β is any
single point of data plotted in SPC chart where the value
is bigger than mean, �̅� but less than LCL and UCL to
show that the process is under control. If the value of β
is similar to �̅� means that the process is under control, return (x,y)=(0,0).
Nasir et al., 2018
90
Table 1. The pseudocode for SPR chart development.
Pseudocode: SPR Chart
�̅� =∑ 𝑝
∑ 𝑛, 𝐶𝐿 = �̅�, 𝐿𝐶𝐿 = �̅� − 3 (
�̅�(1−�̅�)
𝑛), 𝑈𝐶𝐿 = �̅� + 3 (
�̅�(1−�̅�)
𝑛),
1𝜎 = 𝐶𝐿 +(𝑈𝐶𝐿−𝐶𝐿)
3, 2𝜎 = 1𝜎 +
(𝑈𝐶𝐿−𝐶𝐿)
3
if LCL<0 then
plot CL, UCL, LCL, 1𝜎 and 2𝜎 else
if β<100 then
return (x,y)=(0,0)
end
end
4. A CASE STUDY
4.1 Ceramic valve
Company XYZ producing the ceramic valve with
the excellent surface. Most of the ceramic valve is used
for sliding component, potentiometers, isolators and
faucet valves. Generally, the ceramic valve
manufacturing processes begin with ceramic powder
pressing. No quality activities for powder pressing and
the output measurement details are relied on machine
setting and a second process which is firing/curing.
After the firing process, the product sample needs to be
measured to ensure the machine setting is correctly
done.
Based on the pattern shown in Figure 1, some
adjustment would be done for improvement when the
uncertainty is found. The SPC help in showing the
powder press machine pattern. Even the pattern showed
the process is under control, but the process needs to be
monitored carefully. It is proven by Batch 3 in figure 2
where most of the point is consecutive higher compare
with Batch 1 and 2.
The next process after firing is grinding and
lapping. For grinding process, the thickness does not
give the quality issues in the production process.
Lapping is a process to give the flat surface with
roughness up to 0.0001 in precision. The lapping relies
on two factors which are flatness level for diamond plat
and the material handling during process input. Those
factors are related with working knowledge and skill.
The result obtained shows that the surface roughness for
Batch 1, Batch 2 and Batch 3 is under control. However,
there are some uncertainty that may contribute to defect
if without any monitoring and control action taken.
Figure 1 The uncertainty found in Batch 1 and Batch 2.
Even the pressing process is under control but without
trigger action, it may contribute to defect.
Figure 2 The consecutive points are higher compared to
Batch 1 and Batch 2.
Figure 3: The lapping result for ceramic valve where the
worker knowledge and skill play a major role in quality
of surface roughness.
4.2 Discussion
Company XYZ was chosen is because of they still
implement the manual production as for production
type. Both process firing and lapping are critical in
quality factors. Furthermore, company XYZ still does
not implement any Industry 4.0 tools which able to help
them in ease the monitoring process. By having the SPC
as a Traceability tool, it will expedite the analysis
process for decision making and minimize the cost of
defects.
5. CONCLUSIONS
In this paper, the SPC as a Traceability tool is
emphasized where the real data from manual production
type is used for validation. The SPC is aimed to be
operated through the graphical user interface as a
service where the visual graph may ease on analyzing
the production process instead of bulky data and paper.
For the future work, the data from semi-auto and fully
automated production is required for validation purpose.
In order to recognize the SPC pattern just-in-time, the
Hidden Markov Model is emphasized by considering
the boundaries conditions.
REFERENCES
[1] Qin, J., Liu, Y., & Grosvenor, R. (2016). A
categorical framework of manufacturing for
industry 4.0 and beyond. Procedia Cirp, 52, 173-
178.
[2] Bhatt, G. D. (2000). An empirical examination of
the effects of information systems integration on
business process improvement. International
Journal of Operations & Production
Management, 20(11), 1331-1359.
[3] Herter, J., & Ovtcharova, J. (2016). A Model based
Visualization Framework for Cross Discipline
Collaboration in Industry 4.0 Scenarios. Procedia
CIRP, 57, 398-403.
[4] Neubauer, D. V. (2010). Manual on Presentation of
data and control chart analysis. ASTM
International.
Proceedings of Mechanical Engineering Research Day 2018, pp. 91-92, May 2018
__________
© Centre for Advanced Research on Energy
Real time object customization in cad software via visual basic programming
Zainal Fahmi Zainol Abidin, Muhammed Nafis Osman Zahid*
Faculty of Manufacturing Engineering, Universiti Malaysia Pahang, 26600, Pekan, Pahang, Malaysia.
Corresponding e-mail: [email protected]
Keywords: 3D CAD modelling; object customization; graphical user interface (GUI)
ABSTRACT – This paper outlines a development of
graphical user interface (GUI) for object customization
in computer aided design (CAD) by utilizing Visual
Basic (VB) programming language in NX10 CAD
software. Major works involve development of the object
customization tool in the form of new GUI which
provides a set of editable parameters section. A
customized graphical user interface (GUI) was developed
to simplify the manual customization process. Initially,
customization work is translated into programming codes
via advanced tool available in the NX10. The recorded
codes translated into visual basic script files and then is
modified to create a functional GUI. The results revealed
that the developed programs are capable to simplify
drawing work in CAD by reducing the drawing steps and
time.
1. INTRODUCTION
In conventional approach, customization process
for a model designed in CAD requires user to modify the
parameter manually. User needs to reopen the file, select
the area that need to be customized and make the editing
process by key-in the desired input. For recurring work
towards the same model, it consumes a lot of steps
needed to complete the sketch. Other than that, in current
industrial environment, production time becomes a
crucial factor. Considering overall product development
cycle, 80% of the production time is wasted in design
process [1]. Based on the previous studies [2-7], time
saving can be greatly reduced during the designing stage
of the product. In this project, real time object
customization has been developed by integration of
Visual Basic Programming studio and Unigraphics NX
10 software. NX 10 has a smart feature called NX Flow,
and NX Open Common Application Programming
Interface (API) that allows the integration of custom
software applications.
2. METHODOLOGY
In this study, object customization was executed by
using Microsoft Visual Basic 2010 (VB). Integration
between NX 10 and VB was possible due to NX Open. It
works through the Common API of NX 10. NX Open
provides all programming languages to assist the
development of software to improve automation and
fusion of the tasks. The customization work involves four
main steps that need to be executed. The block diagram
for this process is shown in Figure 1.
Figure 1 Block diagram for object customization
process.
2.1 Journaling
A function inside NX 10 called Journaling allowed
a series of actions performed on the interface being
recorded. A file of .net language code was automatically
produced by NX 10 describing what had been carried out
in programming language (visual basic, JAVA or C++) as
shown in figure 2. Based on that, it produces a scripted
file from an interactive session of NX which can be run
and replay again later. These sessions will be edited and
enhanced with distinct programming instruction for
example to construct graphical user interface component.
Figure 2 Sample of VB language programming recorded
by NX10 Journaling tool.
2.2 Visual basic studio
Figure 3 below shows the GUI that has been created
for the object customization. The GUI has been
developed in Visual Basic Studio using windows form
application. Visual basic language programming that has
Recording
journalling in
NX 10
3D part
modelling
GUI design in
VB
Executing .exe
file in NX10
Zainol Abidin and Osman Zahid, 2018
92
been recorded by NX10 Journaling tool will be open in
the Visual Basic software and then the designing process
was executed.
Figure 3 Object customization graphical user interface
(GUI) created using visual basic studio.
3. RESULTS AND DISCUSSION
Experiment was conducted using developed
program and are classify into two levels which is
classified as expert and beginner. Expert user represented
by worker with more than 3 years of CAD experience
while beginner user represented by worker with below
than 3 years of experience in CAD software. Object
selected for this customization process is guitar model
sketched in NX-10. The guitar’s curve surface will be
modified using GUI that has been developed. The curve
surface will be adjusted to the most comfortable curve
that will fit user body. For this guitar shape, the
customization is made on B-spline curve on the guitar
model. Developed GUI allow user to customize the B-
spline point that has been plotted for the curve shape as
shown in Figure 4.
Figure 4 Developed GUI is activated in NX10 system.
Figure 5 Customization on the B-spline control point.
From Table 4, object customization manages to
increase 93% time saving for common user and 50% time
saving for daily user in editing a guitar model when using
real-time customization method. Time saving efficiency
has been measured by dividing time recorded using GUI
and without using GUI.
Table 1 Experimental results.
Guitar
model
Time
recorded
without using
GUI (sec)
Time
recorded with
using GUI
(sec)
Time
saving
(%)
Expert 40 20 50
Beginner 300 20 93
4. CONCLUSION
This approach is considered as an alternative way
from editing the sketch manually into real time
customization that allows user to view real time changes
on the object while manipulating the GUI. As conclusion,
the developed program managed to improve the
efficiency of object customization and shorten the design
processing time in CAD system.
ACKNOWLEDGEMENT
We acknowledge with gratitude to Ministry of
Higher Education Malaysia for providing a financial
support under Research Acculturation Grant Scheme
(RDU160130), which realize this research project.
REFERENCES
[1] Koli, P. S., & Patil, S. K. (2017). Customization of
3D CAD model for piston fixture using nx software.
International Journal of Scientific & Engineering
Research, 8(4), 221-228.
[2] Alsop, L. (2010). Intelligent 3D cad modelling of a
diseased carotid. Thesis.
[3] Wade, B. (2011). Automated solution to the
CADMAT project. Thesis.
[4] Camba, J. D., & Contero, M. (2016). Parametric
CAD modeling: An analysis of strategies for design
reusability. Computer-Aided Design, 74, 18-31.
[5] Monedero, J. (2000). Parametric design: a review
and some experiences. Automation in
Construction, 9(4), 369-377.
[6] Mok, H. S., Kim, C. H., & Kim, C. B. (2011).
Automation of mold designs with the reuse of
standard parts. Expert Systems with
Applications, 38(10), 12537-12547.
[7] Bodein, Y., Rose, B., & Caillaud, E. (2014). Explicit
reference modeling methodology in parametric
CAD system. Computers in Industry, 65(1), 136-
147.
Proceedings of Mechanical Engineering Research Day 2018, pp. 93-94, May 2018
__________
© Centre for Advanced Research on Energy
Power spectral density-based analysis of secondary suspension parameters effect on railway vehicle ride quality
Mohd Hanif Harun1,3*, Ridhwan Jumaidin2,3, Adzni Md Saad1,3, Fauzi Ahmad1,3, Faizul Akmar Abdul Kadir1,3
1) Faculty of Mechanical Engineering, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Faculty of Engineering Technology, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3) Centre for Advanced Research on Energy, Universiti Teknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia.
*Corresponding e-mail: [email protected]
Keywords: Railway vehicle suspension; frequency response analysis; power spectral density
ABSTRACT – The objective of this paper is to study
the effects of suspension parameters on ride quality of
railway vehicle in frequency response analysis.
Seventeen degrees of freedom (17-DOF) railway
vehicle lateral model was used and it governed based on
Newton’s second law. The value of spring stiffness and
damping coefficient for secondary suspension are varied
in order to realize which suspension elements give more
effect to vehicle ride comfort. All simulation analysis
was made using MATLAB/Simulink software. The
analysis shows that the secondary lateral damper was
able to reduce unwanted motion and improve comfort
level simultaneously.
1. INTRODUCTION
The function of suspension system in railway
vehicle can be divided into two categories; first is for
wheel-set guidance and stability (primary suspension),
and another is for ride quality and passenger comfort
(secondary suspension). A comfort level can be analysed
by means of the response of railway vehicle body in
terms of body acceleration. It can be measured in both
time response and frequency responses analysis. This
paper presents the analysis of simulation results of
railway vehicle body vibration in frequency domain.
There are three criteria’s considered in the frequency
response analysis which are body lateral, roll and yaw
acceleration as proposed by [1].
There are two levels in railway vehicle suspension
system and can be categorized based on the location of
the suspension component namely primary and
secondary suspension [2]. The function of primary
suspension is to connect wheel-sets to the bogie and
enhance bogie stability. The secondary suspension is
located between bogie and vehicle body. The effect of
secondary suspension systems on ride quality of railway
vehicle will be studied in this paper. The study is
important in order to examine which suspension
element gives more effect on ride quality in the railway
vehicle.
2. 17-DOF RAILWAY VEHICLE MODEL
2.1 Mathematical Model
A mathematical model of 17-DOF railway vehicle
model was developed based on Newton’ second law as
described in previous study [3]. A schematic diagram of
the suspension model can be seen in Figure 1 and 2
below.
Figure 1 Top view of suspension system.
Figure 2 Front view of suspension system.
3. RESULTS AND DISCUSSION
Figure 3 shows the frequency responses of railway
vehicle body accelerations for all kind of suspension
parameters. Based on the figure, the lateral and roll
body accelerations of railway vehicle have two
resonance peaks, except for yaw acceleration which has
three peaks. For lateral acceleration, when the spring
stiffness is decreased, the lateral acceleration is also
decrease at the first peak, however at the second peak
the response is the same with others. Just like the
response of the roll acceleration, it also has two peaks
which the first peak is looks like the suspension with
lower spring stiffness still has lower vehicle body
response. Moreover, there are three peaks of the railway
vehicle body response for the yaw acceleration. The
system with lower spring stiffness has a better
Harun et al., 2018
94
attenuation ability especially at the second peak but for
almost the whole frequency range, the responses are
almost the same for all different suspension parameters.
From the figure, it can be concluded that the ability
of the secondary spring is not superior in improving ride
quality of railway vehicle body since the main function
of this component is to support the vehicle body.
Nowadays most of the modern railway vehicle
technology is not fitted with the secondary spring and it
is replaced with an air-bag suspension system.
Figure 3 PSD of the body acceleration after varying the
secondary spring stiffness: (a) lateral acceleration,
(b) roll acceleration, and (c) yaw acceleration.
Figure 4 depicts the PSD graph of the railway
vehicle body lateral, roll and yaw accelerations. As
discussed in time response analysis, by increasing the
value of the damping coefficient at the secondary level,
significant improvement occurred on ride quality of
railway vehicle. For all vehicle body accelerations, the
system with higher damping coefficient has three peaks,
which the first and second peaks are the natural
frequency of the system and it is occurred below 1 Hz.
The third resonance peak of the system which is
occurred at the frequency of 2.3 Hz shows that the
system has minimum response of vehicle body
accelerations.
Figure 4 PSD of the body acceleration after varying the
secondary damping coefficient: (a) lateral acceleration,
(b) roll acceleration, and (c) yaw acceleration.
4. SUMMARY
The simulation test was done by varying the lateral
suspension parameters which consists of secondary
springs and dampers. From the result analysis, the
secondary damper gives more effect on ride quality of
railway vehicle body compared to the secondary spring
element.
ACKNOWLEDGMENT
The authors would like to thank Universiti
Teknikal Malaysia Melaka for the financial support
provided through Journal Publication Incentive Grant
(project number JURNAL/2018/FTK/J00004).
REFERENCES
[1] Wang, D. H., & Lao, W. H. (2009) Semi-active
suspension systems for railway vehicles using
magnetorheological dampers. Part II: Simulation
and analysis. Vehicle System Dynamics, 47(1),
1439-1471.
[2] Goodall, R. M., & Mei, T. X. (2006) Active
Suspensions. in Handbook of Railway Vehicle
Dynamics, S. Iwnicki, Ed., ed London: CRC Press
Taylor & Francis. 328-357.
[3] Harun, M. H., Jamaluddin, H., Rahman, R. A.,
Hudha, K., & Wan Abdullah, W. M. Z. (2014).
Analysis of primary and secondary lateral
suspension of railway vehicle system. Journal of
Mechanical Engineering, 11(1) 19-40.
10-1
100
101
102
0
0.01
0.02
0.03
0.04
0.05
0.06
Frequency (Hz)(a)
PS
D [(m
/s2)2
/Hz]
Decrease by 50% of k2y
Decrease by 25% of k2y
k2y = 2800 N/m
Increase by 25% of k2y
Increase by 50% of k2y
10-1
100
101
102
0
0.01
0.02
0.03
0.04
0.05
Frequency (Hz)(b)
PS
D [(r
ad
/s2)2
/Hz]
Decrease by 50% of k2y
Decrease by 25% of k2y
k2y = 2800 N/m
Increase by 25% of k2y
Increase by 50% of k2y
10-1
100
101
102
0
1
2
3
4
5
6x 10
-4
Frequency (Hz)(c)
PS
D [(r
ad
/s2)2
/Hz]
Decrease by 50% of k2y
Decrease by 25% of k2y
k2y = 2800 N/m
Increase by 25% of k2y
Increase by 50% of k2y
10-1
100
101
102
0
0.02
0.04
0.06
0.08
0.1
0.12
Frequency (Hz)(a)
PS
D [(m
/s2)2
/Hz]
Decrease by 50% of c2y
Decrease by 25% of c2y
c2y = 2800 Ns/m
Increase by 25% of c2y
Increase by 50% of c2y
10-1
100
101
102
0
0.02
0.04
0.06
0.08
0.1
Frequency (Hz)(b)
PS
D [(r
ad
/s2)2
/Hz]
Decrease by 50% of c2y
Decrease by 25% of c2y
c2y = 2800 Ns/m
Increase by 25% of c2y
Increase by 50% of c2y
10-1
100
101
102
0
1
2
3
4
5
6
7
8x 10
-4
Frequency (Hz)(c)
PS
D [(r
ad
/s2)2
/Hz]
Decrease by 50% of c2y
Decrease by 25% of c2y
c2y = 2800 Ns/m
Increase by 25% of c2y
Increase by 50% of c2y
Proceedings of Mechanical Engineering Research Day 2018, pp. 95-97, May 2018
__________
© Centre for Advanced Research on Energy
Preliminary thermal simulation analysis of building via IES<VE> software
Afiqah Ngah Nasaruddin1,3,*, Tee Boon Tuan1,2, Musthafah Mohd Tahir1,2, Teruaki Ito3
1) Faculty of Mechanical Engineering, UniversitiTeknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 2) Centre for Advanced Research on Energy, UniversitiTeknikal Malaysia Melaka,
Hang Tuah Jaya, 76100 Durian Tunggal, Melaka, Malaysia. 3)Department of Mechanical Engineering, Graduate School of Technology, Industrial and Social Sciences,
Tokushima University, 2-1 Minamijosanjima-cho, Tokushima-shi, 770-0855, Japan.
*Corresponding e-mail: [email protected]
Keywords: Thermal environment; building information modeling; computational fluid dynamics
ABSTRACT – Attaining thermal environment was a
severe issue to be considered in building design.
Besides that, in the case of existing building, thermal
environment is crucial to be monitored for the purpose
of optimization. With that in mind, IES<VE> software
is selected as a platform to simulate the thermal
condition. This paper aims to present the steps involved
prior to implementation of thermal analysis. From the
simulation parameters such as radiance level, solar gain,
air temperature, moisture content and relative humidity
was identified accordingly. The value of building
predicted chiller load in condition described in the
simulation study is 81.7139 kW.
1. INTRODUCTION
Building simulation is intentionally governed as a
tool for prediction corresponding to several what-if
scenarios for both indoor and outdoor environment [1].
Though researchers are having the unalike parameter of
interest and focus output. Their output somehow will
lead to precise building’s thermal characteristic profile
development and possibility for optimization based on
the data mining respectively [2]. Thermal analysis that
is being conducted on existing building aims to obtain a
baseline. In contrary, if it is conducted during the design
stage, the purpose is to enable early prediction based on
several possibilities of phenomena [3].
Existence of multiple software platforms that
support architectural model development along with
building performance analysis are the outcome of the
advancement in the computer aided design (CAD) [4].
Numerous software are available that support thermal
simulations analysis such as EnergyPlus, TRNSYS,
Ecotect, Green Building Studio and many more [5].
IES<VE> is applied to conduct thermal simulation and
analysis of the building for this case study. In general,
IES<VE> provides user with numbers of modules that
serve for major element in building simulation such as
HVAC, daylighting and flow dynamic. This paper
provides the information related to workflow of
preliminary analysis of thermal simulation on selected
building model which is library building.
2. METHODOLOGY
The research consists of several steps which
generally described to exist between pre-simulation and
post-simulation stages. For instance, it is essential to go
through data gathering steps in pre-simulation stages to
developed thermal profile of studied building and
considering any relevant parameter that having a great
influence on the simulation. The workflow for the
simulation designed, and analysis are shown in Figure 1.
Figure 1 Thermal analysis simulation workflow
The information required in running thermal
analysis simulation comprises of building layout,
detailed on surface volume any opening such as air vent,
window and door. In addition, considering the internal
load contributor such as the daylighting, occupant and
weather condition contribute to much accurate analysis
[6]. Initially, the building profile was developed by
gathering the information as in Table 1.
Table 1 Building descriptions: Library building.
Properties Description
Total Area 10,063.68 m2
level 4
Roof Reinforced-concrete metal roof
Window Uncoated single glazing ¼ inch thick
Wall 230mm thick brickwall plaster on both side
Table 2 describes the cooling capacity of water cooled
screw chiller model that currently operated at library
building as for 2017. In term of operation all three
chillers are scheduled to rotate their duty consecutively.
With only two chillers will be operating at one time
while the other will serve as a backup and provide a
long-lasting lifetime.
Ngah Nasaruddin et al., 2018
96
Table 2 Chiller cooling capacity and input power (2017)
Model Cooling capacity Input
power COP
TR kW
YEWS
100SA50D 102 359 68.8 5.21
YEWS
130SA50D 170 598 108 5.53
YEWS
210SA50D 210 738 134 5.51
The building model is developed accordingly as
shown in Figure 2, without assigning any detail of
furniture and building services such as lighting system
or existing HVAC system. ApacheSim was then
employed to perform the preliminary thermal analysis.
The analysis will provide information such as indoor
thermal comfort conditions by verifying the contributed
heat load in general [4]. It is being complimented by
using other IES module namely RadianceIES module
and Suncast module. The solar analysis is based on
Melaka weather data throughout the year was
conducted. It provides the accurate sunpaths and output
of solar radiation transmitted into the building that
eventually affects the building total heat load. On top of
that, change in radiation depends upon the temperature
and surface characteristic of the object [7].
Figure 2 Library model.
3. RESULTS AND DISCUSSION
For convenient data gathering and analysis, the
library space is divided into eight sections as labelled in
Figure 2. Among the data gathered was the mean value
parameter as listed in Table 3. From Table 3 it is shown
that the maximum mean value for solar gain in a year is
located at section 4 having value of 2.6536 kW.
Table 3 Mean value
Section
Mean parameter
Solar
gain
(kW)
Air
temperatur
e (̊c)
Moisture
content
(kg/kg)
Relative
humidity
(%)
S1 1.8950 23.00 0.01389 78.00
S2 0.0000 23.00 0.01387 77.92
S3 0.0001 23.00 0.01387 77.92
S4 2.6536 23.00 0.01389 78.00
S5 0.1212 23.00 0.01387 77.93
S6 0.1139 23.00 0.01387 77.92
S7 0.1169 23.00 0.01387 77.92
S8 0.1101 23.00 0.01387 77.92
Figure 3 also shows the predicted building chiller
load though the exact occupant comfort while other heat
load contributor does not take into consideration. Given
the parameter showns by the 3D graph between the
time, month and the chiller load in Figure 3, it is known
that the peak building chiller load is 81.7139 kW on
March 2nd at 15.30 p.m.
Figure 3 Predicted building chiller load.
The thermal environment does not only being
influence by HVAC system operated within the facility.
In addition, how the building envelope was designed to
withstand the external weather condition by avoiding
excess heat gain turns out as a crucial issue that
correspond to thermal comfort for an indoor
environment. Thermal simulation is being applied with
the concern of reflecting the analysis with important
aspects such as air heating load, and cooling load. The
simulation using IES<VE> is expected to be reliable
considering that they are having 14 high degree of
certainty
4. CONCLUSION
Based on this study, performance simulation is the
constructive instruments that can demonstrate the
wellbeing of building services by identification of while
ensuring the indoor environment thermal comfort are
achieved.
ACKNOWLEDGMENT
The authors gratefully acknowledge the financial
support from UniversitiTeknikal Malaysia Melaka
(UTeM), Malaysia and Tokushima University, Japan.
REFERENCES.
[1] Hemsath, T. L. (2013). Conceptual energy
modeling for architecture planning and design:
impact of using building performance simulation
in early design stages. Proceedings of BS2013: 13th
Conference of International Building Performance
Simulation Association, 376-384.
[2] Bahar, Y. N., Pere, C., Landrieu, J., & Nicolle, C.
(2013). A thermal simulation tool for building and
its interoperability through the building
information modeling (BIM)
platform. Buildings, 3(2), 380-398.
[3] Akadiri, P. O., Chinyio, E. A., & Olomolaiye, P. O.
(2012). Design of a sustainable building: A
conceptual framework for implementing
sustainability in the building
sector. Buildings, 2(2), 126-152.
[4] Zhu, D., Hong, T., Yan, D., & Wang, C. (2013). A
detailed loads comparison of three building energy
modeling programs: EnergyPlus, DeST and DOE-
2.1 E. Building Simulation, 6(3), 323-335.
1
4
3
4
2
4
4
4
5
4
6
4
8
4 7
4
Ngah Nasaruddin et al., 2018
97
[5] Cho, Y. K., Alaskar, S., & Bode, T. A. (2010).
BIM-integrated sustainable material and renewable
energy simulation. In Construction Research
Congress 2010: Innovation for Reshaping
Construction Practice, 288-297.
[6] Clarke, J.A., and Hensen, J.L.M. (2015). Clarke, J.
A., & Hensen, J. L. M. (2015). Integrated building
performance simulation: Progress, prospects and
requirements. Building and Environment, 91, 294-
306.
[7] Patra, P.K., and Tripathi, D.P. (2014). Monitoring
the room and controlling its temperature and light
industry from a remote location over a network.
International Journal of Advanced Computing
Technique and Applications, 2(2), 96-101.