simulation m hods or nvh -d opm t of a oubl clutch ... · pdf filesimulation m hods or nvh -d...
TRANSCRIPT
Simulation Methods for NVH-Development of a Double Clutch
Transmission Gearbox
SIMPACK User Meeting 2014
Alexander Neubauer
Mercedes-Benz Cars Development
Daimler AG, Stuttgart
Alexander Neubauer | Daimler AG | 09 October 2014 2
Content of presentation
1. Gearbox Simulation in Driveline Developement
2. Overview Testrigmodels for NVH-Simulation and Methodology
3. Simulation of Gearwhining in the DCT Gearbox
4. Evaluation of Rattling Sensitivity of DCT Gearbox
5. Conclusion and Future Work
Vortrag Alexander Neubauer | DAIMLER AG, RD/FNP | 01.10.2014 3
3
Design Integration Design Matching
Big models, Low Discretisation
Low frequency range High frequency range
Compact models, High Discretisation
Simulationmodels - Design and Developement Driveline
1.1 Driveline Development - NVH Simulation
1.2 Overview of Noises in Automotive Gearboxes
Gearbox NoisesGearbox Noises
Loaded gearsLoaded gears
Gear whiningGear whining
Unloaded gearsUnloaded gears
Idle speedIdle speed Coast / Full loadCoast / Full load
RattlingRattling
source/reason:
• Fluctuation in stiffness (transmission error)
• Macrogeometry of gear (Spur bevel- and
Helix Gearwheel)
• Microgeometry (Gearteeth flank correction)
• High loadtorque (Pretension)
• Bending stiffness of the shafts
• Structural behavior of housing
• Oil viscosity and temperature
source/reason:
• Impulsive hits at the gear backlash limits
• Loose gear oscillation of unloaded gearwheels
(also synchrobodies) because of gear backlash
• Torsional vibration excitation of the driven gear
• Angular acceleration amplitude is responsible
for taking off loose gear from the driven flank
of the fix gear
• Oil viscosity and temperature
Alexander Neubauer | Daimler AG | 09 October 2014 4
2.1 Methodology for NVH-Simulation of DCT Gearbox
Alexander Neubauer | Daimler AG | 09 October 2014 5
SIMPACK- Pre SIMPACK-Post
• Taper roller bearings with
stiffness in 3D-Map from the
supplier (BEARINX-MAP)
• bearing properties from KISS-
SOFT interface (input from CAD)
0 1000 2000 3000Frequenz [Hz]
1000
1500
2000
2500
3000
3500
Drehzahl [1/min]
-50
-40
-30
-20
-10
0
10
20
Beschleunigung [dB/(m/s²)]
21 42
63KS1 Z
• Campbell plots for gear rattling
and whining
• 3D-map to descibe rattling
sensitivity of a gearbox for a
switched gear
• Creation of targets for stucture
borne sound level on defined
housing locations
• Operational mode shape
analysis of the gearbox housing
(excitation at bearing locations
with constant velocity)
• flexible gearbox housing
• modal materialdamping
• external nodes for evaluation of
structure borne sound
• rigid body elements for
gearshaft mounting
• flexible gearshafts
• switchable gears
• bearings with 6 DOF
• syncrobodies as dummy bodies
• rigid gearwheels with detailed
macro- and microgeometry
• gearwheels with material data
• analytical gearpair contact
simulation (ISO 6336)
1500 2000 2500 3000
Drehzahl [1/min]
-30
-20
-10
0
10
20
30
Beschleunigung [dB/(m/s²)]
KS2 X
940 Hz
2700 min-1
570 Hz
1600 min-1
• transmission error
• eigenfrequency analysis
• Detailled analysis of mechanism
• correlation with end-of-line
testrig measurements
• Evaluation of accelerations on
defined housing locations
• detailed analysis of gearorders
Alexander Neubauer | Daimler AG | 09 October 2014 6
DCT
Diff
Constant
Torque
velocity
at sideshaft
velocity
at sideshaft
DCT
Diff
Constant
Torque
Constant
Torque
Defining velocity with:
• Measured torsional excitation
• Sinusiodal excitation
��
�
2.2 Overview Testrig Models for NVH-Simulation
Targeting “whining“
Targeting “rattling“ 0 1000 2000 3000Frequenz [Hz]
1000
1500
2000
2500
3000
3500
Drehzahl [1/min]
-50
-40
-30
-20
-10
0
10
20
Beschleunigung [dB/(m/s²)]
21 42
63KS1 Z
2nd gearcoast
Rattling Testrig Whining Testrig
Generat gearbox excitation
2.3 Simulation Workflow
Alexander Neubauer | Daimler AG | 09 October 2014 7
SIMPACK Pre /Solving POST / Nastran
MBS-Simulation
Excitation from
GEAR-PAIR
+Dynamics from bending modes
of gearshafts
+Dynamic behaviour of gearbox
housing up to
6kHz
Transient excitation
forces
at shaft bearings
Export-script
transient excitation
F(t) from gear order
Frequency dependent
excitation
F(f) from gear order
(as RLOAD/DLOAD-
Table in NASTRAN)
FEM-simulation (NASTRAN)
operational mode shape
analysis
Evaluation housing
accelerations (in dB) on
defined housing locations in
gear order
MBS-simulation (SIMPACK)
Evaluation of response
Alexander Neubauer | Daimler AG | 09 October 2014 8
Defining all boundary conditions (given speed at
sideshaft, torque at inputshaft) from DCT EOL
testrig in series production
n [rpm]
M [Nm]
3.1 Introduction to DCT Gearbox Simulation Model
DCT End-of-line (EOL) SIMPACK MBS-Model
Dre
hza
hl
Dre
hm
om
en
t
Coast
Full-load3500 to
1000 rpm
5,5 sec
from– 73 Nm
up to – 15 Nm
5,5 sec
from+ 16 Nm
up to+ 164 Nm
Operation Program EOL Testrig (one gear)
Rigid
Differential
Flexible
Side shaft
Gearset with
Rigid gears and
Flex shafts
Flexible
housing
Alexander Neubauer | Daimler AG | 09 October 2014 9
3.2 Description of Gearset in DCT Gearbox Model
• Two part gearsets with flexible shafts
• Switchable gearstages and loose gears
• Definition of gearwheel Macro- and Microgeometry
• Analytical gearpair contact simulation (ISO 6336)
• Gearshaft bearings with 6 DOF
• Synchrobodies with dummys (mass & inertia)
• Rigid gearwheels with material properties (damping)
• Gear slicing for shifting of shaft and gearwheel
• Loss torque at each loose gear as damping value
• Taper roller bearings with 3D-maps for stiffness fromthe supplier (BEARINX-MAP)
• Bearing properties from KISS-SOFT interface (inputfrom CAD)
Contactmodelling with slicing
Topology of one gear stage
Input shaft loose gearfixed gear Output shaft
Force Element with Number Joint with 1 DOF about rot.axis
1500 2000 2500 3000
Drehzahl [1/min]
-30
-20
-10
0
10
20
30
Beschleunigung [dB/(m/s²)]
1500 2000 2500 3000
Drehzahl [1/min]
-30
-20
-10
0
10
20
30
Beschleunigung [dB/(m/s²)]
KS2 X KS3 X
Alexander Neubauer | Daimler AG | 09 October 2014 10
3.3 Correlation End-of-line to Simulation
good correlation between measured housing
structure borne sound level in the gear order at the
end of line (EOL) testrig and in MBS-simulation
Measurement end of line (EOL)
Model with nonlinear bearing stiffness (FE43)
Model with 3D-Map (BEARINX)
DCT, coast, 2nd gear order
Alexander Neubauer | Daimler AG | 09 October 2014 11
3.4 Mechanism of Gear Whining
NVH Problem:
Gear whining at coast in the gear order from
meshing between fixed and loose gear
2nd bending mode at 974 Hz (lateral bending)
with gearbox carrier resonance
1st bending mode at 607 Hz
Axial vibration of
inputshaft
bearing bed
560 HZ operational mode of housing
Mechanism:
Overlap gearmeshing with bending resonance of
both inputshafts (disadvantage in frequency of
gearpair with the teeth meshing frequency)
gear order
Gear order
940 Hz
2700 rpm570 Hz
1600 rpm
60dB
40dB
30dB
50dB
n [rpm]
Measurement at end-of-line
F [N]
L [dB]
Bearing forces simulation
F-X Inputshaft2 bearing
F-Y Outputshaft2 bearing
Alexander Neubauer | Daimler AG | 09 October 2014 12
Housing acceleration end of line
Housing acceleration engine testrig
Good correlation between measured
structure borne sound level at gearbox
housing at end of line and engine
testrig and also the air borne sound
level in the vehicle at co driver seatS2
S2, Mot –X(orientation of
gearbox in vehicle –
lateral build in)
3.5 Correlation Vehicle and Testrig Measurements
Air borne sound level co-driver seat
1600 rpm2600 rpmCoast 2nd Gear
Gear order
End-of-line & Engine testrig
Vehicle measurement
1600 rpm 2600 rpm
S2, Fzg -Y
Gear order
Alexander Neubauer | Daimler AG | 09 October 2014 13
3.6 Influence of Bending at Input shaft on Whining
Trail: Application of additional masses at inputshaft1 (solid shaft), for example130gr
on fixed gearwheels at two locations
Result MBS-Simulation Result End-of- line Testrig
Target: Shifting up/down of the bending eigenfrequencies from inputshaft 1, that the
first characteristic gear whining peak will decrease at 1600 rpm
-4 dBBaseline
Simulation
Changing mass at input shaft 1 has influence on
gear whining above all in lower engine speed
rpm
dB
rel. m
/s² Baseline
Measurement
-3 dB
dB
rel. m
/s²
rpm
Time [s]
Housing accelerations [dB]
velocity in [rpm] of
fixed gear
loose gear
Alexander Neubauer | Daimler AG | 09 October 2014 14
4. Evaluation of Rattling Sensitivity of DCTImpacts on flank with backlash are responding as housing acceleration
peaks (sinusoidal torsional excitation at the gearbox inputshaft)
� Created 3D-map is only valid
for one gear and load, variation
of speed and angular excitation
� 3D-map is an important criteria
for rattle sensitivity of the DCT
gearbox
rattle limit
Alexander Neubauer | Daimler AG | 09 October 2014 15
5. Conclusion and Future Work
0 1000 2000 3000Frequenz [Hz]
1000
1500
2000
2500
3000
3500
Drehzahl [1/min]
-50
-40
-30
-20
-10
0
10
20
Beschleunigung [dB/(m/s²)]
21 42
63KS1 Z
• complex eMBS-modell necessary
• great influence of the transferpath from
– flexible shafts to housing
– bearing stiffnesses and damping
• Good correlation simulation to measurements
with non linearity bearing stiffnesses as 3D-
maps (BEARINX)
• gearrattling 3D-map for one gear and load
created
• further investigation for rattling sensitivity with
comparison of structure borne sound level on
housing with real engine excitation response
with DMF.
• Modelling of additional excitation mechansim
for example from the gearbox oil pump gear
meshingFull load run up of a gearbox in the 2nd gear with a real
engine excitation of a 4-Cylinder Dieselengine without
DMF, Torque at inputshaft 300 Nm