2022 Ijat NVH
2022 Ijat NVH
2022 Ijat NVH
net/publication/362605451
CITATION READS
1 290
4 authors, including:
Wieslaw Fiebig
Wroclaw University of Science and Technology
70 PUBLICATIONS 306 CITATIONS
SEE PROFILE
All content following this page was uploaded by Damian Pietrusiak on 23 August 2022.
-DNXE:UyEHO'DPLDQ3LHWUXVLDN:LHVáDZ)LHELJDQG0DWHXV]&]HFKRZVNL
9HU\ILQHORFDOPHVK±FRQWDFWDUHD
Figure 7. Model assembly and boundary conditions.
The same force time signal was implemented in
experimental testing with an electrodynamic shaker
(subsection 3.2).
Figure 6. Worm wheel mesh definition. 6LPXODWLRQ
Simulation was conducted in two time steps. First the
preload of springs, bushing and vibration isolating
component of the assembly which modal modes may
components. Second step consisted of application of
contribute to resonances and deteriorate the system NVH
operational load ± force time trace. Preliminary results
performance.
indicated that the noise and vibration problems are triggered
by the impact load observed as a dynamic collision between
,QWHUDFWLRQV'HILQLWLRQ%&V
CRBRACKPINION pair.
When all structural components are modelled and Identification of the causing phenomenon (Figures 8 and
assembled, proper definition of boundary conditions and 9) is relatively straightforward. The main challenge in the
interactions need to be defined. As the simulation consist of definition of simulation FEM dynamic model, where the
several structural components which mostly interacts in the NVH is the main target of the results simulation, is the
model with contact, any simplification of BC’s which can proper representation of impact between the two
be introduced will improve the model performance components namely investigations of contact stiffness.
(convergence, calculation time) significantly. In the EPS When assessing the global dynamic behavior of the
structure considered in this paper, boundary conditions components (e.g. deflection shape) the local contact stiffness
were defined as follows (Figure 7) where some of theme
where realized with of the connectors:
1 ± assembly of bushing and two Orings, 2, 3 ± fixed
support of the bearing outer ring, 4 ± spring and Oring 5$&. 3,1,21
supporting CRB, 5 ± one side fixed slot connector, 6 and 7
± dampers/ vibration isolators supporting the worm, 8 ±
electric motor (inertia, electromagnetic resistance). The
stiffness of the component 1, which is subassembly of the :250
bushing and Orings, was identified in separate simulation
&5%
where the assembly was loaded by normalized load in all
6DOF. With that approach the complex prestress Figure 8. Contact separation.
simulation and nonlinear interaction between those
components was substituted by the bushing connector
corresponding to the stiffness identified in the separate
simulation. Thanks to that, the main simulation (dynamics)
was less time consuming and with greater convergence.
The red arrow in Figure 7 represents place where external
load was applied to the structure. Load was applied as a
force signal in the time domain. Force value and character
of its variability was acquired during track testing on chosen
road surface types, vehicle speeds and vehicle platforms. Figure 9. Impact between components.
$872027,9((/(&75,&32:(567((5,1*6<67(0619+3(5)250$1&(,19(67,*$7,216
DQGQRLVHJHQHUDWLRQLVDQDO\]HGWKHORFDOLPSDFWEHWZHHQ ϴ͘нϬϰ
FRPSRQHQWVLVWKHPDLQDQDO\]HGSKHQRPHQRQ ϲ͘нϬϰ
LQWURGXFHQRQSURSRUWLRQDOO\KLJKG\QDPLFUHVSRQVHRIWKH Ϯ͘нϬϰ
;ŵŵͬƐϮͿ
FRPSRQHQWV0RUHRYHULQWKHSURGXFWLRQ(36DVVHPEO\WKH Ϭ͘нϬϬ
VXUIDFHV RI UDFN ± SLQLRQ FRQWDFW LV FRYHUHG ZLWK JUHDVH ͲϮ͘нϬϰ
ZKLFKLVLQWURGXFLQJVWURQJGDPSLQJ'XHWRWKHIDFWWKDW Ͳϰ͘нϬϰ
UHDOFRQWDFWVWLIIQHVVDVZHOODVGDPSLQJFRHIILFLHQWZHUH Ͳϲ͘нϬϰ
UDFN SHUSHQGLFXODU GLUHFWLRQ DUH SUHVHQWHG LQ )LJXUH ƚĞƐƚͺĐŽŶĨŝŐͺϰ ƚĞƐƚͺĐŽŶĨŝŶŐͺϱ
where;
WHVWBFRQILJB ± VWLII ZRUP ORQJLWXGLQDO VXSSRUW OLQHDU )LJXUH$FFHOHUDWLRQUHVSRQVHPHDVXUHGRQ&5%LQWKH
JHQHUDOFRQWDFWVWLIIQHVV GLUHFWLRQRILWVWUDQVODWLRQ±VLPXODWLRQVUHVXOWV
WHVWBFRQILJB±QRQOLQHDUHODVWLFORQJLWXGLQDOVXSSRUWRI
WKHZRUP ZRUPGXPSHUV OLQHDUJHQHUDOFRQWDFWVWLIIQHVV
WHVWBFRQILJB ± ZRUP GXPSHUV QRQOLQHDU JHQHUDO (
WHVWBFRQILJB
FRQWDFWVWLIIQHVV (
(
VWLIIQHVVYLVFRXVGXPSLQJRQ&5%VSULQJ (
WHVW FRQILJB ± ZRUP GXPSHUV OLQHDU JHQHUDO FRQWDFW
(
VWLIIQHVVYLVFRXVGXPSLQJRQ&5%VSULQJ
7KLVSUHVHQWVKRZWKHPRGHOLVVHQVLWLYHIRUPHQWLRQHG (
D E F
6KDNHU
6WLQJHU
(36
Figure 12. EPS ± Rack mode shapes at a) 260(Hz), b)
720(Hz), c) 830(Hz). )L[WXUH
D E F
Figure 13. EPS ± Housing mode shapes at a) 320(Hz), b)
730(Hz), c) 790(Hz).
geometry, modify the mesh size and validate the chosen
element models.
Figure 13 presents first three mode shapes of the EPS
housing. Once again the natural frequencies are positioned
out of the excitation range. Similarly to the rack
measurements, this data can be used to optimize the FEM
system dynamics simulation. Those experiments validates Figure 15. Figure 1 Acceleration measurement points.
numerical model (rack) and proves that modeling of the
housing is not necessary (see section 2) while it stiffness is
60
much greater than the internal components stiffness. RACK
CRB
40
7HVWLQJRQWKH5LJ 20
(m/s2)
car mounting position and insures proper tie rod angle -60
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
position in relation to the excitation force vector at the (s)
shaker.
Excitation signals depend on the test procedure, in some Figure 16. Exemplary acceleration signals acquired at two
test cases random or sine signals are used. In a specific car different points at the EPS test track excitation.
platform tests, the input force signals are acquired at
various points of car prototype during test track runs and
then replicated at the test rig by means of FDR (Field acceleration measurement points are selected based on
Data Replication). Excitation force generated on the engineering and product experience.
electrodynamic shaker is applied to the tested specimen via Acceleration signals presented in Figure 16 can be used
a stinger attached to the tie rod joint which resembles the to indicate the vibration origin and noise causes. Detailed
EPS real operating conditions and allow to better noise and analysis of system input and output signals can directly
vibration troubleshoot. System response is registered with connect the test track occurrence such as braking, turning
accelerometers and sound pressure transducers. The etc. to vibration problems and its place.
$872027,9((/(&75,&32:(567((5,1*6<67(0619+3(5)250$1&(,19(67,*$7,216
$FRXVWLF
&DPHUD
(36
)L[WXUH
)LJXUH $FRXVWLF SKRWR ± 1RLVH ORFDOL]DWLRQ DW )LJXUH $FRXVWLF SKRWR ± 1RLVH ORFDOL]DWLRQ DW
IUHTXHQF\RI +] +]
-DNXE:UyEHO'DPLDQ3LHWUXVLDN:LHVáDZ)LHELJDQG0DWHXV]&]HFKRZVNL