Humpfinder

1
Hump-finder RPUG-2003, Texas, USA Christopher
R. Byrum, Ph.D., P.E. Soil Materials Engineers,
Inc. Plymouth, MI Office
2
Forward Engineering Estimating the behavior by
understanding the inputs.
3
Forward Engineering Estimating the behavior by
understanding the inputs. Inverse Engineering
Understanding the inputs by measuring the
behavior.
4
Forward Engineering Estimating the behavior by
understanding the inputs. Inverse Engineering
Understanding the inputs by measuring the
behavior. RPUGers INVERSE ENGINEERS
5
PAVEMENT PROFILE A 2-D Slice through a 3-D
Solution LWP RWP
Subbase
Base
Subgrade
6
Roughness Defined
Function z f(x)
Smooth
Roughness- The presence of curvature and
discontinuities causing a deviation from a smooth
condition.
7
(No Transcript)
8
(No Transcript)
9
Elevation Spectral Density of 500-feet of
Repeating 20-ft Arc Segments, no faults
L/1
L/2
L/3
L/4
L/5
10
(No Transcript)
11
IRI f (Faults, Avg. Slab Curv, and Curvature
Std.Dev.)
GPS3 JPCP
12
(No Transcript)
13
(No Transcript)
14
MOST OF INITIAL IRI IS IN HERE !!! VARIATION of
CURVATURE
15
What makes up the standard deviation of curvature
component of the profile, which affects IRI?
16
What makes up the standard deviation of curvature
component of the profile, which affects IRI?

• Humps and dips within the correct wavelengths.

17
What makes up the standard deviation of curvature
component of the profile, which affects IRI?

• Humps and dips within the correct wavelengths.
• Macro/mega texture and tinning.

18
What makes up the standard deviation of curvature
component of the profile, which affects IRI?

• Humps and dips within the correct wavelengths.
• Macro/mega texture and tinning.
• Construction method in general.

19
Two Profilers
20
Curvature ? 1/R, H f(R,s) Apparent Texture
Size
R
H
s
s
s 6, 12, 18..48
21
6-inch Interval IRI Profiles- Two Devices
22
The Texture Effect
GPS3 Range
Above the Ideal Texture Is Other Stuff
like Humps, dips and waves
Ideal Texture Size 6 ?
23
The Moving-Hump Finder
Allowable Shape Error about /- 30
moving
Compares 5 Profile points to an Ideal Shape (/-
error)
In Order for a Match

• Hump Height greater than Threshold
• Shape Error less than Threshold

24
For IRI Find Sin-Humps that match these
wavelengths
I Look For 4 humps 14 humps 24humps
Coupled Mode
Body Mode
Tire Mode
Based on Sayers Karamihas
25
Geometry of a Sin-Hump
1.0
0.7071
0.7071
0.0
0.0
?/4
?/2
3?/4
?
0
L 4, 14, and 24
½ (wavelength)
26
Hump-Finder Control Parameters
Moving Sin-Hump
Profile
Base-line
Ni
N2i
N3i
N4i
N
The Hump-Finder has three control parameters
27
Hump-Finder Control Parameters
Profile
Height
Base-line
Ni
N2i
N3i
N4i
N
The Hump-Finder has three control parameters
Height Elev. (N2i) Avg. Elev. (N, N4i)
28
Hump-Finder Control Parameters
Error 1
Profile
Height
Base-line
Ni
N2i
N3i
N4i
N
The Hump-Finder has three control parameters
Height Elev. (N2i) Avg. Elev. (N, N4i)
Error 1 Avg. Elev. (N, N2i) 0.7071(Height)
Elev. (Ni)
29
Hump-Finder Control Parameters
Error 2
Error 1
Profile
Height
Base-line
Ni
N2i
N3i
N4i
N
The Hump-Finder has three control parameters
Height Elev. (N2i) Avg. Elev. (N, N4i)
Error 1 Avg. Elev. (N, N2i) 0.7071(Height)
Elev. (Ni) Error 2 Avg. Elev. (N2i, N4i)
0.7071(Height) Elev. (N3i)
30
Hump-Finder Logic
Error 2
Error 1
Profile
Height
Base-line
Ni
N2i
N3i
N4i
N
Hump-Finder Logic
IF 1.) Height gt Hmin AND IF 2.) (Error 1)2
(Error 2)2 lt ErrorMax THEN Report ? Height at
point N2i ELSE Report ? Zero
31
CRCP IRI is about 80 in/mi, mostly from Humps
and Dips
1.5 0.0
Elev, in.
3.5 0.0
Imperf, mm
0.35 0.0
Hump Size, in
Distance, feet
0
500
32
24 Humps
Hmin 0.15 in. (3.8 mm)
ErrorMax .002 in2 ? /- .032 (30) error
33
1
34
2
1
35
2
1
3
36
2
4
1
3
37
2
4
5
1
3
38
2
4
5
6
1
3
39
CRCP IRI is about 80 in/mi, mostly from Humps
and Dips
1.5 0.0
Elev, in.
3.5 0.0
Imperf, mm
0.35 0.0
Hump Size, in
Distance, feet
500
0
40
CRCP IRI down to 50 in/mi after hump and fault
simulated grinding
1.5 0.0
Elev, in.
3.5 0.0
Imperf, mm
0.15 0.0
Hump Size, in
Distance, feet
500
0
41
Any Questions?
Michigans State Fossil Mastodon
42
(No Transcript)
43
(No Transcript)
44
(No Transcript)
45
MOST OF INITIAL IRI IS IN HERE !!! VARIATION of
CURVATURE
46
(No Transcript)
47
The Texture Effect
LTPP GPS3 Regression
Calculated
48

• MODEL CAPTURED TEXTURE EFFECT
• WITH StDev. Curvature (texture pattern).
• MODEL ERROR
• RANDOM LONGER WAVES

49
The Texture Effect
0.5 m/km Offset avg. Long Wave Length Stuff ?
model error
GPS3 Regression
0.6 m/km Texture StDev Curv.
Calculated
50
(No Transcript)
51
(No Transcript)
52
The Texture Effect
GPS3 Range
Ideal Texture Size 6 ?