Pavement Thickness Pavement Thickness Evaluation Using Ground Evaluation Using Ground Penetrating Radar Penetrating Radar Dwayne Harris Dwayne Harris Presented for Final Exam Presented for Final Exam
Jan 03, 2016
Pavement Thickness Evaluation Pavement Thickness Evaluation Using Ground Penetrating RadarUsing Ground Penetrating Radar
Dwayne HarrisDwayne Harris
Presented for Final ExamPresented for Final Exam
OUTLINEOUTLINE
IntroductionIntroduction Fundamentals of GPR Fundamentals of GPR Interpretation of GPR dataInterpretation of GPR data Methodologies for Thickness EvaluationMethodologies for Thickness Evaluation GPR Data QualityGPR Data Quality Validation of MethodologiesValidation of Methodologies
IntroductionIntroduction
Background on pavement thickness Background on pavement thickness evaluationevaluation
Literature reviewLiterature review
Significance of Thickness Significance of Thickness InformationInformation
Pavement managementPavement management Pavement performance and remaining life estimates Pavement performance and remaining life estimates
require knowledge of pavement thicknessrequire knowledge of pavement thickness Setting maintenance and rehabilitation prioritiesSetting maintenance and rehabilitation priorities Main input in overlay design Main input in overlay design INDOT Major Moves $138,483,477 budgeted for 2006 INDOT Major Moves $138,483,477 budgeted for 2006
resurfacingresurfacing Thickness of uppermost surface course needed for Thickness of uppermost surface course needed for
mill and Fill resurfacing projects.mill and Fill resurfacing projects. Pavement thickness is needed for project level FWD Pavement thickness is needed for project level FWD
structural analysis structural analysis
National Pavement RehabilitationNational Pavement Rehabilitation
YearYear Urban Urban InterstatesInterstates
Rural Rural InterstatInterstat
Rural Rural RoadRoad
ExpenditureExpenditure
19981998 8.69%8.69%
PoorPoor
3.25%3.25%
PoorPoor
1.42%1.42%
PoorPoor
$36.3$36.3
BillionBillion
20032003 7.62%7.62%
PoorPoor
1.64% 1.64%
PoorPoor
0.76%0.76%
PoorPoor
$49.3$49.3
BillionBillion
ChangeChange 1.07%1.07% 1.61%1.61% 0.66%0.66% 36%36%
[Hartegen, 2005]
Technologies Used for Pavement Technologies Used for Pavement Thickness EvaluationThickness Evaluation
CoreCore– CostlyCostly– DestructiveDestructive– Provides a good ground truth record.Provides a good ground truth record.
Falling Weight Deflectometer (FWD)Falling Weight Deflectometer (FWD)– None Destructive None Destructive
Ground Penetrating RadarGround Penetrating Radar– Non DestructiveNon Destructive– Collected at Highway SpeedCollected at Highway Speed– Dense CoverageDense Coverage– Heavy Post ProcessingHeavy Post Processing
Related Work on Thickness EvaluationRelated Work on Thickness Evaluation
[Berge et al, 1986] initial pavement thickness studies[Berge et al, 1986] initial pavement thickness studies [Livneh and Siddiqui, 1992] mathematical model [Livneh and Siddiqui, 1992] mathematical model [Fernando, 2000; Scullion and Saarenketo, 2002] [Fernando, 2000; Scullion and Saarenketo, 2002]
automated interface identification automated interface identification [Al-Quadi et al, 2005] model expanded to three or more [Al-Quadi et al, 2005] model expanded to three or more
layerslayers SummarySummary There are multiple models available for pavement There are multiple models available for pavement
thickness evaluationthickness evaluation– The model selected for this study is utilized for a large The model selected for this study is utilized for a large
majority of the studiesmajority of the studies Current literature suggests using semi-automatic data Current literature suggests using semi-automatic data
interpretation methodologiesinterpretation methodologies
FundamentalsFundamentals
GPR trace and waveforms and data GPR trace and waveforms and data presentationspresentations
Mathematical modelMathematical model
Trace Example SR - 9
0
500
1000
1500
2000
2500
3000
0 20 40 60 80 100 120 140 160
Am
pli
tud
e0 1 2 3 4 5 6 7
Travel Time (nanoseconds)
Reflected Pulse Waveforms
0
500
1000
1500
2000
2500
3000
0 20 40 60 80 100 120 140 160
Sample #
Am
pli
tud
e
0 1 2 3 4 5 6 7
Radar pulse w aveform reflected off third interface
Positive phase wavefoms
Negative phase waveform
Starting State
SecondState
EndingState
Pulse Center
StartingState
Ending State
SecondState
Pulse Center
GPR Data B-scanGPR Data B-scan
Amplitude
500
1000
1500
2000
2500
3000
3500
Distance (feet)
Tim
e (N
anos
econ
ds)
Data: i65n160.yat
0 1000 2000 3000 4000 5000
2
4
6
8
10
12
14
16
18
Interface 1 Interface 2
8967
29
3.9162.948
1.276
0
500
1000
1500
2000
2500
3000
3500
0 20 40 60 80 100 120
Am
plit
ud
e
0 1 2 3 4 5
Time (nanoseconds)
Two way travel time in first layer
Two way travel time in second layer
1
2 3
3293
582
1337
1724
0
500
1000
1500
2000
2500
3000
3500
0 20 40 60 80 100 120
Sample #
Am
plit
ud
e
0 1 2 3 4 5
Radar pulse w aveform reflected off pavement surface
Radar pulse w aveform reflected off f irst
pavement boundary
Radar pulse w aveform reflected off second pavement boundary
Amplitude of surface return, A0
Amplitude of return off first boundary A1
*A0
*
EM Wave Propagation VelocityEM Wave Propagation Velocity
11.8/ sec m
R R
cV in nano
20
1 20
( )
( )p
Rp
A A
A A
0
R1
speed of light
Amplitude of Return off Metal Plate
Amplitude of Surface Return
Dielectric Constant Uppermost Surface Course
p
c
A
A
Principles of GPR Interface Principles of GPR Interface InterpretationInterpretation
An interface is defined as the anomaly in GPR data An interface is defined as the anomaly in GPR data occurring when the reflected waveforms from a occurring when the reflected waveforms from a physical pavement boundary are contiguous for a physical pavement boundary are contiguous for a group of sequential tracesgroup of sequential traces
The radar (EM) wave must propagate, to the The radar (EM) wave must propagate, to the interface and back.interface and back.
The radar wave must reflect off the interface with The radar wave must reflect off the interface with enough energy to be recorded.enough energy to be recorded.
The interface must be identified in the GPR The interface must be identified in the GPR record.record.
The source of the first interface isclassified as an HMA surface course
boundary
Do while currentinterface is HMA
Can the interface source beclassified as steel reienforcment
ClassifySource as
HMA surfacecourse
AnotherInterface ?
end
Classify sourceas steel
reineforcement
Anotherinterface ??
Do while currentinterface is steelreienforcement
end
Can the interface source beclassified as steel reienforcment
Classifysource as
steelreinforcement
Anotherinterface ??
Classifysource as
base of PCCboundary or
disregard
end
end
no
no
yes
yes
yes
no
no
no
Next interfaceclassify source as
HMA surfacecourse boundary
yes
Next interfaceclassify source as
steelreinforcement
yes
Classify source of secondinterface as an HMA surface
course boundary
The source of the first interfaceis steel reineforcement
Classify source of secondinterface as steel reinforcement
Two Interface Case ATwo Interface Case A
Amplitude
500
1000
1500
2000
2500
3000
3500
RP(Miles)
Tw
o-W
ay T
rave
l Tim
e (N
anos
econ
ds)
I-74 Validation Section C
11.5 12 12.5 13 13.5 14 14.5 15 15.5
2
4
6
8
10
12
14
16
18
base of HMA steel reinforcement
Two Interface Case BTwo Interface Case B
Amplitude
500
1000
1500
2000
2500
3000
3500
RP(Miles)
Tw
o-W
ay T
rave
l Tim
e (N
anos
econ
ds)
I-65 Validation Section North Bound
150 155 160
2
4
6
8
10
12
14
16
18
HMA interfaces
Methodologies for Thickness Methodologies for Thickness Evaluation (regional M1)Evaluation (regional M1)
Top layer methodologyTop layer methodology– Discontinuities are located in dataDiscontinuities are located in data– Interfaces are identified in the dataInterfaces are identified in the data– Regional dielectric constants are determined Regional dielectric constants are determined – Thickness values are calculated for each mileThickness values are calculated for each mile– Enhanced to calculate thickness using dielectric Enhanced to calculate thickness using dielectric
constants from individual tracesconstants from individual traces
Interface SelectionInterface Selection
Regional Dielectric ConstantsRegional Dielectric Constants
Thickness CalculationThickness Calculation
Every thickness pick is Every thickness pick is assigned the assigned the respective regional respective regional dielectric value.dielectric value.
Thickness Values Thickness Values Calculated.Calculated.
Average value Average value calculated for each calculated for each mile.mile.
Dielectric Regional
DielectricPick
ThicknessPick
Thickness Calculated
r
p
p
n
r
ppn
ε
tk
tk
tktk
Multiple Layer Methodology (M2)Multiple Layer Methodology (M2)
Determine the layers to be modeledDetermine the layers to be modeled Form data set of possible interfacesForm data set of possible interfaces Select interfaces to be modeledSelect interfaces to be modeled Calculate thickness valuesCalculate thickness values Present the thicknesses in a visually acute Present the thicknesses in a visually acute
format allowing for proper interpretationformat allowing for proper interpretation
Quality of GPR DataQuality of GPR Data
BlundersBlunders– Improper waveform selectionImproper waveform selection– Omitted pavement layersOmitted pavement layers
Systematic errorsSystematic errors– Travel time systematic errorTravel time systematic error– Velocity systematic errorVelocity systematic error
Random errorsRandom errors– Error propagationError propagation
Improper Waveform Improper Waveform Selection Selection
I-65 Study AreaI-65 Study Area13 Inches HMA Over PCC13 Inches HMA Over PCC
Amplitude
500
1000
1500
2000
2500
3000
3500
RP(Miles)
Tw
o-W
ay T
rave
l Tim
e (N
anos
econ
ds)
Data: I65 North Bound Transition 2
223.2 223.25 223.3 223.35 223.4 223.45
2
4
6
8
10
12
14
16
18
PCC overlay HMA overlay
HMA interfaces
before transition after transition
Amplitude
1350
1400
1450
1500
1550
1600
223.4 223.42 223.44 223.46 223.48
4
5
6
7
8
9
10
11
RP (miles)
Tw
o-W
ay T
rave
l Tim
e (N
anos
econ
ds)
Data: e86b005.yat
Base of HMA
Base of HMA lifts
after transition
330 mm (13 inches) of HMA overlay over rubblized JRCP
305 mm (12 inches) of concrete overlay over JRCP
190 mm (7.5 inches) of fiber modified HMA overlay over Cracked and Seated JRCP
MM 229.1
MM 223.4
MM 217.2
SB NB North to Chicago
MM 237.8
Interface SelectionInterface Selection
0 200 400
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
Time: b006b.PKS
Distance (feet)
Tra
vel T
ime
(nan
osec
onds
)
HMA prior to transition 1
0 500 1000
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
Time: b005b.PKS
Distance (feet)
HMA following transition 2
0 500 1000
4
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
Time: b002b.PKS
Distance (feet)
HMA prior to transistion 3
(a) (b) (c)
Positive Phase
0
500
1000
1500
2000
2500
3000
3500
4000
0 100 200 300 400 500
Am
pli
tud
e0 2 4 6 8 10 12 14 16 18
Time (nanoseconds)
N O P
Negative Phase (Twisted Trace)-1000
-500
0
500
1000
1500
2000
2500
3000
0 100 200 300 400 500
Am
pli
tud
e
0 2 4 6 8 10 12 14 16 18
Time (nanoseconds)
S
Difference in Dielectric Constant and Difference in Dielectric Constant and ThicknessThickness
0 200 400 6004
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9 Dielectric Constant of PCC for Positive and Negative Waveforms
Distance (feet)
Die
letr
ic C
onst
ant
PCC dielectric constantpositive phase waveform
0 200 400 6004
4.5
5
5.5
6
6.5
7
7.5
8
8.5
9
Distance (feet)
Die
lect
ric C
onst
ant
PCC dielectric constantnegative phase waveform
(a) (b) 0 200 400 600
9.5
10
10.5
11
11.5
12
12.5
13
13.5
14
14.5
HMA Overlay Thickness Positive and Negative Phase Waveforms
Distance (feet)T
hick
ness
(in
ches
)
HMA overlay thickness positive phase waveform
0 200 400 600
9.5
10
10.5
11
11.5
12
12.5
13
13.5
14
14.5
Distance (feet)
Thi
ckne
ss (
inch
es)
HMA overlay thickness negative phase waveform
(a) (b)
Positive PhaseDielectric constant
Negative PhaseDielectric Constant
Positive PhaseThickness
Negative PhaseThickness
Error Omitted Pavement LayersError Omitted Pavement Layers
Amplitude
1350
1400
1450
1500
1550
1600
223.4 223.42 223.44 223.46 223.48
4
5
6
7
8
9
10
11
RP (miles)
Tw
o-W
ay T
rave
l Tim
e (N
anos
econ
ds)
Data: e86b005.yat
Base of HMA
Base of HMA lifts
after transition
Omitted Pavement LayersOmitted Pavement Layers
0 200 400 600
11
11.5
12
12.5
13
13.5
14
14.5
15
HMA Section 2 Thickness Utilizing All Interfaces and Omitting Intermidiate Interfaces
Distance (feet)
Thi
ckne
ss (
inch
es)
HMA overlay thicknessomitting interfaces
0 200 400 600
11
11.5
12
12.5
13
13.5
14
14.5
15
Distance (feet)
Thi
ckne
ss (
inch
es)
HMA overlay thicknessall interfaces
(a) (b)
Thickness (Layers Omitted) Thickness (All Layers)
Travel Time Systematic ErrorTravel Time Systematic Error
0 200 400 600 800 1000 1200-1500
-1000
-500
0
500
1000
1500
2000
2500Height 11 Inches
Sample #
Am
plitu
de
t1
t2 (a) 0 200 400 600 800 1000 1200
-1500
-1000
-500
0
500
1000
1500
2000
Sample #A
mpl
ititu
de
Example Hieght 22 inches
t1
t2 (b)
Time Scale Plot
y = 16.752x + 49.71
R2 = 0.9999
y = 17.4105x + 64.004
R2 = 0.9979
0
100
200
300
400
500
600
700
800
0 5 10 15 20 25 30 35 40
Antenna Height (inches)
Dif
fere
nce
in
Sam
ple
s
Channel 2
Channel 1
Linear (Channel 2)
Linear (Channel 1)
Channel 1RMS=4.77
Channel 2RMS=0.99
Width of Waveform
60
70
80
90
100
110
120
130
0 5 10 15 20 25 30 35 40
Antenna Height (inches)
Wav
efo
rm W
idth
(sa
mp
les)
Channel 2 w idth
Channel 1 w idth
Channel 1Mean: 106.5STD: 12.45COV: 11.7%
Channel 2Mean: 86.5STD: 0.60COV 0.7 %
Velocity Systematic ErrorVelocity Systematic Error
Random Error Random Error
0
0 0
0
2 2 2
02 2 202 2 2
0 0 0
2 2 2 2 2 2 2 20 02 2 2
0 0
2 2 2 202 2
0
22
2 2
( )
2relative error
p
p p
p
ppD A A t
p p p
D A p A D A p Ap p
DA p A
p
A AtAtAc
A A A A A A
ct DA A A A
A A A A
A AD A A
0
D
0
2 2
Standard deviation thickness, Relative error, Speed of light
Metal plate amplitude, Surface return amplitude, Thickness
Variance surface return amplitude, Variance metal platp
D
p
A A
cD
A A D
2
e amplitude,
Travel time, Variance trave timett
Error SummaryError Summary
Improperly selecting waveforms is a significant Improperly selecting waveforms is a significant blunder sourceblunder source
Utilizing automated interface selection algorithm Utilizing automated interface selection algorithm increased the likelihood of this blunderincreased the likelihood of this blunder
Omitting pavement layers introduces errorsOmitting pavement layers introduces errors Channel 1 data not used due to large systematic error Channel 1 data not used due to large systematic error
is travel timeis travel time Velocity systematic errors propagate into thickness Velocity systematic errors propagate into thickness
errorerror Amplitude random error propagates to about 1% Amplitude random error propagates to about 1%
relative thickness error relative thickness error
Validation of MethodologiesValidation of Methodologies
Comparison with 3Comparison with 3rdrd party Software party Software Comparison of methodologies developedComparison of methodologies developed Thickness variationThickness variation Network thickness studyNetwork thickness study GPR thickness evaluation accuracyGPR thickness evaluation accuracy
Thickness ComparisonsThickness Comparisons
Seven pavement sections of three Seven pavement sections of three interstates. interstates.
Pavement sections of three state roadsPavement sections of three state roads Five pavement sections of two interstates Five pavement sections of two interstates
used for 3used for 3rdrd party comparison party comparison
Statistical Analysis (M2 vs TERRA)Statistical Analysis (M2 vs TERRA)
Population IntersectionPopulation Intersection Split into 50 or 25 foot subsectionsSplit into 50 or 25 foot subsections Normality, F test, and T-test analysisNormality, F test, and T-test analysis Explanation of T-test resultsExplanation of T-test results
Normality Analysis of Sub Section Populations
H0=Population Normally Distributed
Alpha=95%
Thickness Normality Analysis
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
I65 (50 ft) I74A (50ft) I74A (25ft) I74B (50ft) I74B (25ft) I74C (50ft) I74C (25ft) I74F (50ft) I74F (25ft)
Per
cen
tag
e F
aile
d N
orm
alit
y G
oo
dn
ess
of
Fit
Tes
t M2 Omni
M2 JB
M2 Lillifors
TERRA Omni
TERRA JB
TERRA Lillifors
Equality of Means and VarianceAnalysis of Sub Section PopulationsH0=Populations Have Same VarianceAlpha=95%H0=Populations Have Same MeansAlpha=99%
Thickness Equality of Means and Variance Test Results
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
I65 (50 ft) I74A (50ft) I74A (25ft) I74B (50ft) I74B (25ft) I74C (50ft) I74C (25ft) I74F (50ft) I74F (25ft)
Per
cen
tag
e W
her
e N
ull
Hyp
oth
esis
Rej
ecte
d
F-test
T-pooled variance
T-unequal variance
14.2
4.4
4.6
4.8
Val
ues
Column Number
-2 -1.5 -1 -0.5 0 0.5 1 1.5 24
4.2
4.4
4.6
4.8
5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
14.2
4.4
4.6
4.8
Val
ues
Column Number
-2 -1.5 -1 -0.5 0 0.5 1 1.5 24
4.2
4.4
4.6
4.8
5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
(a)
(d)
(c)
(b)1
4.8
5
5.2
5.4
Val
ues
Column Number
-1.5 -1 -0.5 0 0.5 1 1.54.5
5
5.5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
14.8
5
5.2
5.4
Val
ues
Column Number
-1.5 -1 -0.5 0 0.5 1 1.54.5
5
5.5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
(a)
(d)
(c)
(b)
Best Case Worst CaseI-65 T-test 8% Rejected
1
5.8
6
6.2
6.4
6.6
Val
ues
Column Number
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.55
5.5
6
6.5
7
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
1
5.8
6
6.2
6.4
6.6
Val
ues
Column Number
-2.5 -2 -1.5 -1 -0.5 0 0.5 1 1.5 2 2.55.5
6
6.5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
(d)
(c)
(b)(a)
1
7.7
7.8
7.9
8
8.1
8.2
Val
ues
Column Number
-1.5 -1 -0.5 0 0.5 1 1.57.5
8
8.5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
1
7.7
7.8
7.9
8
8.1
8.2
Val
ues
Column Number
-1.5 -1 -0.5 0 0.5 1 1.57.5
8
8.5
Standard Normal Quantiles
Qua
ntile
s of
Inp
ut S
ampl
e
QQ Plot of Sample Data versus Standard Normal
(d)
(c)
(b)(a)
Best Case Worst Case
I-74F T-test 72% Rejected
T-test ExplanationT-test Explanation
15 20 25 30 35 40 45 500
500
1000
1500
2000
2500
3000
3500
4000
I-74 Validation Section A Trace Group A
Sample #
Am
plitu
de
Amplitude M2 Amplitude TERRA
Difference
Summary M2 TERRA ComparisonSummary M2 TERRA Comparison
90% of the M2 and TERRA populations 90% of the M2 and TERRA populations have the same variance (alpha=95%)have the same variance (alpha=95%)
98% of the M2 and TERRA populations for 98% of the M2 and TERRA populations for I-65 have the same mean (alpha=99%)I-65 have the same mean (alpha=99%)
28% of the M2 and TERRA populations for 28% of the M2 and TERRA populations for I-74F have the same mean I-74F have the same mean
Methodology ComparisonsMethodology Comparisons
Effect of sample sizeEffect of sample size Effect of using regional dielectric constantEffect of using regional dielectric constant
Difference in Thicknesses (M2 - traces M1)
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
0 10 20 30 40 50 60 70 80
Thi
ckne
ss D
iffe
renc
e (i
nche
s)
I-65
I-74
SR-28
SR-47
SR-213
standard deviation
M2
I-69
Difference in Thickness Values (traces M1 - regional M1)
-2.5
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
2.5
0 10 20 30 40 50 60 70 80
Sample #
Th
ickn
ess
Dif
fere
nce
(in
ches
)
I-65
I-69
I-74
SR-28
SR-47
SR-213
standard deviation
traces M1
Network Thickness EvaluationOver 1,600 Miles EvaluatedUppermost Surface Course Thickness Evaluated with GPR Using Regional M1 MethodPavement Structure Thickness Evaluated with FWD
I - 65 South Bound Driving Lane
0
5
10
15
20
25
30
35
40
0 20 40 60 80 100 120 140 160 180 200 220 240 260
Reference Post, Miles
Pav
emen
t T
hic
knes
s, I
nch
es
GPR, Surface Course 1 Thickness GPR, Thickness to 2nd Interface
FWD, Surface Thickness FWD, Total Thickness
Network Thickness EvaluationNetwork Thickness Evaluation
A majority of the INDOT interstate system is A majority of the INDOT interstate system is 25 inches thick with an uppermost surface 25 inches thick with an uppermost surface course thickness of 5 to 7 inches of HMA.course thickness of 5 to 7 inches of HMA.
GPR provided reasonable estimates of the GPR provided reasonable estimates of the uppermost surface course thicknessuppermost surface course thickness
FWD provided reasonable estimates of the FWD provided reasonable estimates of the pavement structure thicknesspavement structure thickness
Thickness VariationThickness Variation
SectionSection NumberNumber MeanMean STDSTD CVCV
I-65I-65 25,67225,672 4.624.62 0.440.44 9.45%9.45%
I-69I-69 41,10841,108 6.486.48 0.570.57 8.72%8.72%
I-74AI-74A 16,58716,587 6.676.67 0.540.54 8.10%8.10%
I-74BI-74B 8,8108,810 3.743.74 0.400.40 10.67%10.67%
I-74CI-74C 15,70415,704 4.974.97 0.340.34 6.93%6.93%
I-74DI-74D 14,25014,250 7.277.27 0.580.58 7.94%7.94%
I-74FI-74F 21,42721,427 6.906.90 0.540.54 7.81%7.81%
SR-47SR-47 32,26032,260 5.705.70 0.390.39 6.78%6.78%
SR-213SR-213 6,2336,233 6.186.18 0.470.47 7.65%7.65%
SR-28SR-28 20,67020,670 6.666.66 1.361.36 20.49%20.49%
AverageAverage 9.45%9.45%
Average*Average* 8.23%8.23%
Published CV valuesPublished CV values
StudyStudy CVCV
LTPP HMALTPP HMA 6.83% to 12.66%6.83% to 12.66%
LTPP PCCLTPP PCC 2.36% to 5.19%2.36% to 5.19%
NCDOT HMANCDOT HMA 25% to 38%25% to 38%
Reported Accuracies of GPR Reported Accuracies of GPR Thickness EstimatesThickness Estimates
REPORTREPORT AccuracyAccuracy
Kansas DOT Kansas DOT 7.5% - 10%7.5% - 10%
SHRPSHRP 8%8%
Minnesota DOTMinnesota DOT 3% - 6.5%3% - 6.5%
Missouri DOTMissouri DOT 4% - 11.3% 4% - 11.3%
Kentucky DOTKentucky DOT 5.82% - 165.04%5.82% - 165.04%
Case Study ResultsCase Study Results
StudyStudy AccuracyAccuracy
I-65I-65
12 Inch Concrete12 Inch Concrete 4.5%4.5%
13 Inch HMA13 Inch HMA 2.0%2.0%
7.5 Inch HMA7.5 Inch HMA 13.2%13.2%
US41 NorthUS41 North
HMAHMA 8.8%, 5.2%8.8%, 5.2%
Concrete Concrete 8.8%8.8%
SR32ESR32E
HMAHMA 16.6%16.6%
Accuracy/CV ResultsAccuracy/CV Results
Study CV (8.23%) within published range of Study CV (8.23%) within published range of 2.36% to 38%2.36% to 38%
Study absolute accuracy range (2% to Study absolute accuracy range (2% to 16.6%) in within published range of 3% to 16.6%) in within published range of 3% to 23.4%23.4%
ConclusionsConclusions
M1 provides efficient acceptable M1 provides efficient acceptable thicknesses for the uppermost pavement thicknesses for the uppermost pavement surface coursesurface course
M2 provides accurate pavement M2 provides accurate pavement thicknesses for multilayer pavementsthicknesses for multilayer pavements
The expanded visualization tools of M2 help The expanded visualization tools of M2 help prevent interface interpretation blundersprevent interface interpretation blunders
Conclusions ContinuedConclusions Continued
Likelihood of interface interpretation blunders Likelihood of interface interpretation blunders increases when automated interface selection increases when automated interface selection and tracking algorithmand tracking algorithm
The process of evaluating pavement thickness The process of evaluating pavement thickness with GPR has not progressed to the point of with GPR has not progressed to the point of eliminating a trained GPR interpretereliminating a trained GPR interpreter
Study absolute accuracy range (2% to 16.6%) Study absolute accuracy range (2% to 16.6%) within published range of 3% to 23.4%within published range of 3% to 23.4%