You are on page 1of 44

Profiling of rough terrain

by Carl Becker
SCOPE
Introduction
Literature survey
Profiling methods
Profiling concepts Profiling concepts
Profiled terrains
Simulations
Conclusion
INTRODUCTION
Simulations are used to verify designs
Cost effective design verification, prior to prototype
building
Compare simulation results with actual field tests
Simulation requirements
Model may be simplified
Constraints
Input data
Simulation is nothing like recycling
Simulation => Garbage in = garbage out
Correct input data is very important
Model => Land Rover Defender (validated by M Thoresson)
Input => Profile of rough terrain
INTRODUCTION
Three profiling concepts were built and verified
Profile a section of the Belgian with each concept profiler
Compare profiles from concept profilometers
Displacement Spectral Densities Displacement Spectral Densities
Obtain profiles of frequently used rough terrains
Profiles of the rough terrains at Gerotek test facilities
Use profile as a Road Definition File in simulation with
the use of a previously validated model
Compare simulation results with actual field tests
LITERATURE SURVEY
Profiling methods
Inertial profilers (2-D)
Laser profilers (3-D)
Aerial photography (3-D) Aerial photography (3-D)
Profiling methods
Inertial profilers
(2-D)
Profiling methods
Inertial profilers
(2-D)
High speed profiler
Profiling methods
Laser profilers
(3-D)
Vehicle Terrain
Measurement System
DGPS
Inertial Measurement
Unit
Laser scanner
Profiling methods
Aerial
photography
(3-D)
Profiling methods
2-D and 3-D profilometers are available to profile
smooth roads at relative high speed
Terrains to be profiled are too rough for available
profilometers
All profilometers have errors All profilometers have errors
Gyro errors
Laser errors
GPS errors
Errors caused by the dynamics of vehicle
Profilometer Requirements
It is required that the profilometers are capable of
the following;
Profile rough terrain (vertical displacements > 25
mm). mm).
Minimum profile width of 2.5 m.
Profilometer must be light weight and easily
transported.
Effective profiling with minimal resources
required.
Profilometer Concepts
Mechanical profilometer
Photogrammetric profiling
Laser scanner
Mechanical Profilometer
A mechanical profilometer using arms pivoted to a reference
frame on one side with small wheels on the ends in contact
with the ground, nicknamed the Can-Can machine
Tilt sensor
Profiling arms Encoder
Potentiometers
Mechanical Profilometer
Profiles a 3 m wide section
Resolution = 10.18 mm x 100 mm
Profile an obstacle < 100 mm high
Accuracy less than 5 mm
Profiling fast and effective
Profile of Belgian paving
Mechanical profilometer
Photogrammetric Profilometer
Survey control points on profiled terrain (Pro Mapping)
Calibrate camera to determine the lens distortions, focal length and size of the image
Sequential photos are taken perpendicular to terrain a with digital camera mounted
on a tripod
Photos overlap one another by 60%
Mapping done by CAD Mapping on a digital photogrammetric workstation
Photogrammetric Profilometer
Profiles a 2.5 m wide section, depending on height of camera
Resolution = 10 mm x 10 mm
Profile an obstacle > 150 mm high
Accuracy 3 mm to 10 mm, depending on camera
Profiling resource intensive
-0.4
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
-0.2
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
1.8
0
0.2
0.4
Width [m]
Concept D test section
Distance [m]
H
e
i
g
h
t

[
m
]
Profile of Belgian paving
Photogrammetric profilometer
Distortion in camera lens
Laser Scanner Profilometer
Laser Distance Sensor, S80-MH-5 Data Sensor, used to measure the profile of the
terrain from a vertical distance of 2m
The Data Sensor is mounted in a purposely build gimball which is mounted on a
tripod
The gimball enables the Laser Distance Sensor to rotate about two perpendicular
axes. The rotation is controlled by two Stepper motors
Stepper Motors
Laser
Laser Scanner Profilometer
Profiles a 2.4 m wide section
Resolution = ave 60 mm x 60 mm, limited by step size
Profile an obstacle > 150 mm high
Accuracy 7 mm
Profiling time consuming, laser sensitive to profiled surface
500x360x58 mm
Profile of Belgian paving
Laser scanner profilometer
Belgian paving
Displacement Spectral Densities
DSD calculated as follows:
With
( )
( )
2
2
xx
X F
S F
F

With
( )
( )
2
*
X Fast Fourier Transformof theroad profile x d
F Stepin Frequency
X F X X


=
=
=
Belgian paving
Displacement Spectral Densities
10
-2
10
0
10
2
/
c
y
c
l
e
s
/
m
]
DSD of profiled tracks

Class A road
Class D road
Class H road
50m Belgian paving, Photogrammetry
40m Belgian paving, Laser
40m Belgian paving, Can-Can
10
-3
10
-2
10
-1
10
0
10
1
10
2
10
-12
10
-10
10
-8
10
-6
10
-4
Spatial Frequency [cycles/m]
D
i
s
p
l
a
c
e
m
e
n
t

S
p
e
c
t
r
a
l

D
e
n
s
i
t
y

[
m
2
/
c
y
c
l
e
s
/
m
]

Displacement Spectral Densities
vs. Straight line fit
10
-3
10
-2
10
-1
/
c
y
c
l
e
/
m
]
Belgian paving PSD, Can-Can profile
Belgian paving, Can-Can profile
Inverse Power Law,S
x
= Ax
-n
0.05 cycles/m
10 cycles/m
10
-2
10
-1
10
0
10
1
10
-7
10
-6
10
-5
10
-4
Spatial Frequency [cycles/m]
D
i
s
p
l
a
c
e
m
e
n
t


S
p
e
c
t
r
a
l

D
e
n
s
i
t
y
[
m
2
/
c
y
c
l
e
/
m
]
Profiled Terrains
Fatigue Track
Profiled Terrains
Corrugations
Profiled Terrains
Potholes
Profiled Terrains
Ride and Handling Track
4.2 km long, profiled in 1 day.
1000
400 600 800 1000 1200 1400 1600 1800 2000 2200
1000
1200
1400
1600
1800
2000
Profiled Terrains
Ride and Handling Track
Close-up on low mobility Ride and Handling Track
Profiled Terrains
Rough Track
Profile top Plato of Rough Track (800 m)
3200 3300 3400 3500 3600 3700
800
850
900
950
1000
Profiled Terrains
Rough Track
Structural stiffness of Can-Can too low
Profiled Terrains
Rough Track
Close-up of rough track
Profiled Terrains
The Can-Can Machine is considered an efficient
profilometer due to:
High profiling speed and efficiency
Ease of data processing Ease of data processing
Accuracy of the profilometer
Low operating costs
Simulation Model
Front Suspension
Simulation Model
Rear Suspension
Simulation Model
Complete Model on RDF Belgian paving
Simulation Model
Simulation validation over a trapezoidal bump
(M Thoresson)
Simulation Model
Actual test @ 15 km/h Simulation @ 15 km/h
Simulation Model
Simulation results FFT unfiltered
0.25
0.3
Simulation FFT and Land Rover FFT of Right Rear Body @ 15 km/h
Simulation Right Rear Body
Land Rover Right Rear Body
0 2 4 6 8 10 12 14 16 18 20
0
0.05
0.1
0.15
0.2
Frequency [Hz]
A
m
p
l
i
t
u
d
e

[
m
]
Simulation Model
Simulation results filtered FFT (8 Hz low-pass)
0.3
0.35
Simulation FFT and Land Rover FFT of Right Rear Body @ 15 km/h
Simulation Right Rear Body
Land Rover Right Rear Body
0 1 2 3 4 5 6 7 8 9 10
0
0.05
0.1
0.15
0.2
0.25
Frequency [Hz]
A
m
p
l
i
t
u
d
e

[
m
]
Simulation Model
Simulation Model
BS 6841
[m/s
2
]
Ride comfort
>2.5 Extremely
uncomfortable
RMS of the vertical accelerations
2.5
3
3.5
RMS_SIM_LF
uncomfortable
1.25-2.5 Very
uncomfortable
0.8-1.6 Uncomfortable
0.5-1.0 Fairly
uncomfortable
0.315-0.63 A little
uncomfortable
<0.315 Not
uncomfortable
0
0.5
1
1.5
2
15 26 40 57 73
Speed [m/s]
R
M
S

[
m
/
s
^
2
]
RMS_SIM_LF
RMS_Landy_LF
RMS_SIM_RR
RMS_Landy_RR
RMS_SIM_LR
RMS_Landy_LR

Conclusion
Three profilometers verified
Can-Can Machine most efficient profilometer
Obtained accurate 3-D profiles of frequently
used test tracks used test tracks
Straight line approximation for DSD not valid
approximation
Good correlation between actual tests and
simulations on profiled terrain
Questions?

You might also like