Professional Documents
Culture Documents
Lecture 3 Mechanistic-Empirical Pavement Desing Flex and Rigid
Lecture 3 Mechanistic-Empirical Pavement Desing Flex and Rigid
MECHANISTIC EMPIRICAL
PAVEMENT DESIGN
Abubeker W. Ahmed
1
6/8/2018
c
c
c
c
2
6/8/2018
c
c
c
c
3
6/8/2018
DESIGN APPROACHES
Difference is in the nature of Performance Relation
Empirical
CBR
AASHTO
Mechanistic / Mechanistic empirical
Shell
Asphalt Institute
CalME
MEPDG /AASHTO ME
4
6/8/2018
EMPIRICAL- AASHTO
Statistical/Experimental (based on road tests)
Limited to the conditions/environment they are developed
Perhaps the most famous and influential project
Late 50’s road test in Illinois
$27 million (in 1960!) Value today?
"...to study the performance of pavement structures of known
thickness under moving loads of known magnitude and
frequency"
(still!) the primary source of experimental data when vehicle
damage to highways is considered
4.79 𝐺ൗ
𝑊𝑋 𝐿18 + 𝐿2𝑠 10 𝛽𝑥
𝐿𝐸𝐹 = = 𝐿2𝑥 4.33
𝑊18 𝐿𝑥 + 𝐿2𝑥 𝐺ൗ
10 𝛽18
5
6/8/2018
SERVICEABILITY-PERFORMANCE
The serviceability-performance concept is based on five assumptions:
Highways are for the comfort and convenience of the road users
Road user’s opinion of how a highway should perform is subjective
Serviceability can be expressed by the mean ratings of the users
There are physical characteristics of pavement which can be
measured objectively and can be related to road users ratings
Performance can be represented by serviceability history of a
pavement (ΔPSI) (the design serviceability loss)
SERVICEABILITY-PERFORMANCE
6
6/8/2018
(SN)
7
6/8/2018
8
6/8/2018
𝒍𝒐𝒈 ∆𝑷𝑺𝑰 ൘ 𝟒. 𝟐 − 𝟏. 𝟓
𝒍𝒐𝒈 𝑾𝟏𝟖 = 𝒁𝑹 𝑺𝒐 + 𝟗. 𝟑𝟔𝒍𝒐𝒈 𝑺𝑵 + 𝟏 − 𝟎. 𝟐𝟎 + + 𝟐. 𝟑𝟐𝒍𝒐𝒈 𝑴𝑹 − 𝟖. 𝟎𝟕
𝟎. 𝟒 + 𝟏𝟎𝟗𝟒ൗ
𝑺𝑵 + 𝟏 𝟓.𝟏𝟗
9
6/8/2018
BASIC PROCEDURE
Determine the traffic (ESAL)
Calculate the effective subgrade modulus (Mref)
Select the performance level (ΔPSI)
Solve for the required SN needed to protect the subgrade (Nomograph
or numerical)
10
6/8/2018
DESIGN INPUTS
Design Life (Historically 20 years- could vary)
Material Properties
Traffic (18 kip ESAL. 80 kN)
Reliability, Degree of certainty that the pavement will last the design
period
Uncertainty in, Traffic prediction, Performance prediction, Materials
& construction
RELIABILITY
Level of reliability should increase with traffic, traffic sensitivity,
public expectation
Reliability factor decreases the predicted number of ESALs that a
particular design can accommodate
Adjust traffic for reliability: Table 11.14 (Huang )
Zr S0
ZR = standard normal deviate. Table 11.15
S0 = combined standard error of the traffic prediction and
performance prediction, 0.49 for flexible pavements, 0.39 for rigid
11
6/8/2018
𝑼𝒇 = 𝟏. 𝟏𝟖 × 𝟏𝟎𝟖 × 𝑴−𝟐.𝟑𝟐
𝒓
12
6/8/2018
LAYER COEFFICIENTS
𝑆𝑁 = 𝑎1 𝐷1 +𝑎2 𝐷2 𝑚2 +𝑎3 𝐷3 𝑚3
13
6/8/2018
Can estimate the base layer coefficient from Figure 7.15 (Text Book
Huang) for untreated base, bituminous-treated base and cement-
treated base
For untreated base can also use the following (instead of
interpolating from the figure):
a = 0.249logE − 0.977
DRAINAGE COEFFICIENT
Unbound layers (base, subbase) are also adjusted based on moisture
conditions:
Quality of drainage measured as time required for drainage
Percent time moisture levels approach saturation
14
6/8/2018
15
6/8/2018
16
6/8/2018
Climate
Material
Geometry
Traffic Rutting
Fatigue cracking
Roughness index
17
6/8/2018
RESPONSE MODELS
Multilayer elastic theory
Nonlinear elastic theory
Viscoelastic theory
Finite element methods
MECHANISTIC-EMPIRICAL PROGRAMS
Shell Method
Asphalt Institute
MEPDG (ASHTOWARE)
CalME
18
6/8/2018
E 10.447 h
0.471 0.041 0.139 0.287 0.868
h E E K
2 1 2 1 3 1
h E h E K
3
1/ 3 1/ 3 0.868
E 1a 1a 1b 1b 1
h h
1
1a 1b
Fatigue cracking
tensile strain at the bottom of asphalt layer controls fatigue performance
N f
f2 f3
f 1 bt
E
Where
Nf = number of cycles to failure
bt = strain at the bottom of asphalt
E = modulus of the asphalt layer
f1, f2, and f3 are model parameters
19
6/8/2018
f
f2
N per . def 1 ztopsg
Where
Nper.def = number of cycles to failure
ztopsg = strain at the top of subgrade
f1, and f2 are model parameters
DAMAGE RATIO
For a known amount of traffic volume, a damage ratio is then
calculated for each mode of damage (fatigue and rutting)
𝑛 𝑚
𝑁𝑖,𝑗
𝐷𝑟 =
𝑁𝑓𝑖,𝑗
𝑖=1 𝑗=1
20
6/8/2018
AI calibrated the equations using AASHO road test data and developed charts
based on the empirical relations controls
21
6/8/2018
22
6/8/2018
AASHTOWARE
AASHTOWare Pavement ME Design is the next generation of
pavement design software, which builds upon the National
Cooperative Highway Research Program mechanistic-empirical
pavement design guide (MEPDG)
23
6/8/2018
AASHTOWARE DISTRESSES
Longitudinal cracking
z i
Thermal cracking using Paris
Law
Permanent deformation in c
bituminous bound layers, c
granular base layers and
subgrade are all considered
24
6/8/2018
N f
f2 f3
f 1 bt
E
Where
Nf = number of cycles to failure
bt = strain at the bottom of asphalt
E = modulus of the asphalt layer
f1, f2, and f3 are model parameters
PERMANENT DEFORMATION
Pavement layers are subdivided into
sublayers and the permanent strain is z i
calculated at the mid depth of each
sublayer
z i
n c
p i z
i 1
p i c
25
6/8/2018
ε aT N ε
p 1
a2 a3
−𝑙𝑛 𝑁 𝑙𝑛 𝑁 𝛽×𝛾
𝛾𝑝 = 𝑒𝑥𝑝 𝐴 + 𝛼 × 1 − 𝑒𝑥𝑝 × 1+ × 𝑒𝑥𝑝 × 𝛾𝑒𝛿
𝛾 𝛾 𝜏𝑟𝑒𝑓
R
p ( N ) CN b
A R
Tseng & Lytton (1989) – Implementerad in AASHTOWARE (MEPDG)
( )
p ( N ) 0 r e N
c c+
26
6/8/2018
𝑎
𝜀𝑝1 = 𝛽1 𝑎1 𝑇1 2 𝑁 𝑎3 𝜀𝑟1
1ൗ
𝜀𝑝1 𝑎3
𝑁𝑒𝑞1 = 𝑎
𝛽1 𝑎1 𝑇2 2 𝑁 𝑎3 𝜀𝑟2
𝑁𝑖1 = 𝑁𝑒𝑞1 + ∆𝑁 𝑎 𝑎
∆𝜀𝑝 = 𝛽1 𝑎1 𝑇2 2 𝑁𝑖13 − 𝑁𝑒𝑞1
3 𝑎
𝜀𝑟2
ACCUMULATION OF DAMAGE
27
6/8/2018
THERMAL CRACKING
Cracking due to temperature may be either low temperature cracking or
thermal fatigue
Low-temperature cracking occurs when the tensile strains due to low
temperature is exceeds the tensile strength of the asphalt mix. Low-
temperature cracking is a common phenomena in flexible pavements in
cold climate
Thermal fatigue cracking is similar to load-induced fatigue cracking but
it is caused by the tensile strain in the asphalt layer due to daily
temperature cycle
The pavement surface cools down faster and with more intensity than the
core of the pavement structure, which causes thermal cracking to occur at
the surface of flexible pavements
Thermal cracks extend in the transverse direction across the full width of
the pavement
𝑙𝑜𝑔𝐶/ℎ
𝐶𝑓 = 𝛽1 × 𝑁
𝜎𝑠𝑡𝑑
E = stiffness, h = AC thickness, std = standard deviation of log of crack depth
N = standard normal distribution
28
6/8/2018
Triaxial tests
Plate loading test
Static shear test
Falling weight deflectometer test
29
6/8/2018
The resilient modulus and permanent modulus tests for granular materials
Cyclic loading (10hz, EN standard)
Multi stage loading (several stress paths)
1000
800
600
400
200
0
0 10 20 30 40 50 60 70 80 90 100
-79.5
-80
-80.5
-81
-81.5
-82
-82.5
-83
-83.5
-84
0 10 20 30 40 50 60 70 80 90 100
30
6/8/2018
The tests are conducted at several cyclic loading frequencies and temperatures
(Ex. EN standard, 0.1,1, 5,10 Hz frequencies
Test temperatures, -10, 0, 10, 20 C
For stress controlled sinusoidal loading test, t sin 2ft
0
t sin 2ft
0
Dynamic modulus
E
* 0
0
Complex modulus
E
*
cos i sin
0
0
31
6/8/2018
Dynamic modulus
5000
4500
4000 0 10 30 50
3500
3000
2500
2000
1500
1000
500
0
0.01 0.10 1.00 10.00 100.00
32
6/8/2018
33
6/8/2018
0
0.000 0.000 0.010 1.000 100.000
Arrhenius equation
1 1
𝐿𝑜𝑔 𝑎 𝑇 = 𝐶 −
𝑇 + 273 𝑇𝑜 + 273
WLF equation
𝐶1 (𝑇 − 𝑇0 )
𝐿𝑜𝑔 𝑎 𝑇 =
𝐶2 + (𝑇 − 𝑇0 )
Where, aT = fr/f is the shift factor, T temperature, T0 reference temp.
34
6/8/2018
The master curves for modulus and phase angle can be mathematically
modeled by a fitting function such sigmoidal function for modulus
35
12000 30
Dynamic modulus (MPa)
Dynamic modulus
Phase angle ()
25
Phase angle
8000 20
15
4000 10
0 0
1.E-06 1.E-04 1.E-02 1.E+00 1.E+02 1.E+04 1.E+06
Reduced frequency (Hz)
35
6/8/2018
10 cm
7.5 cm 4 cm
7.5 cm
15 cm
12.5 cm 26 cm 18 cm
36
6/8/2018
-800
-700
-600
-500
Strain (µm/m)
-400
-300
-200
-100
0
0 1000 2000 3000 4000
Number of cycles
37
6/8/2018
t t
t t
2500
2000
Load (N)
1500
1000
500
0
0 0.5 1 1.5 2 2.5 3
Time (sec)
38
6/8/2018
1000.00
y = 3612.1x-0.222
R² = 0.8388
Initial strain (µm/m)
100.00
y = 2679.3x-0.292
R² = 0.8711
10.00
1.00
1000 10000 100000 1000000
Number of cycles to failure
39
6/8/2018
Abubeker W. Ahmed
40
6/8/2018
41
6/8/2018
DESIGN APPROACHES
Empirical
AASHTO
Mechanistic
4.79 𝐺ൗ
𝑊𝑋 𝐿18 + 𝐿2𝑠 10 𝛽𝑥
𝐿𝐸𝐹 = = 𝐿2𝑥 4.33
𝑊18 𝐿𝑥 + 𝐿2𝑥 𝐺ൗ
10 𝛽18
42
6/8/2018
Design Period
Traffic –what changes? (EALF Table 6.7)
Reliability
Based on functional classification
Overall standard deviation (S0=0.25 – 0.35)
Performance criteria
ΔPSI = PSI0–PSIt
43
6/8/2018
Material Properties
Effective Modulus of Subgrade Reaction (k)
Need to convert subgrade MR to k for cases:
Without Subbase
With Subbase
Shallow bedrock
𝑀𝑅
𝑘=
18.8
44
6/8/2018
45
6/8/2018
𝟑.𝟒𝟐
Month k Ur 𝑼𝒓 = 𝑫𝟎.𝟕𝟓 − 𝟎. 𝟑𝟗𝒌𝟎.𝟐𝟓
1 500 33.5
2 450 32.5 σ𝑼𝒓
𝑼𝒓 =
𝒏
Then user chart or equation above to calculate keff
corresponding to the mean relative damage
46
6/8/2018
𝑬𝑪 = 𝟓𝟕𝟎𝟎𝟎 𝒇𝒄
Modulus of Rupture (S’c):
Third-point loading @ 28 days
LOSS OF SUPPORT
Reduction of keff by a factor LS to
account for erosion and/or
differential soil movement
Best case scenario, LS=0 (Slab is in
full contact with subbase)
47
6/8/2018
48
6/8/2018
𝒍𝒐𝒈 ∆𝑷𝑺𝑰 ൘ 𝟒. 𝟓 − 𝟏. 𝟓
𝒍𝒐𝒈 𝑾𝟏𝟖 = 𝒁𝑹 𝑺𝒐 + 𝟕. 𝟑𝟓𝒍𝒐𝒈 𝑫 + 𝟏 − 𝟎. 𝟎𝟔 + +
𝟏 + 𝟏. 𝟔𝟐𝟒𝒆𝟕ൗ
𝑫 + 𝟏 𝟖.𝟒𝟔
𝑺𝒄 𝑪𝒅 𝑫𝟎.𝟕𝟓 − 𝟏. 𝟏𝟑𝟐
𝟒. 𝟐𝟐 − 𝟎. 𝟑𝟐𝑷𝒕 𝒍𝒐𝒈
𝟏𝟖. 𝟒𝟐
𝟐𝟏𝟓. 𝟔𝟑 ∗ 𝑱 𝑫 −
𝑬𝒄 𝟎.𝟐𝟓
𝒌
BASIC PROCEDURE
Determine the traffic (ESAL)
Calculate the effective modulus of subgrade reaction (keff)
Select the performance level (ΔPSI)
Solve for the required D needed to protect the subgrade (Nomograph or
numerical)
49
6/8/2018
EXAMPLE
SOLUTION BY NOMOGRAPH
50
6/8/2018
SOLUTION BY NOMOGRAPH
Use 10in
51
6/8/2018
52
6/8/2018
𝛅 = 𝑪 ∙ 𝑳 𝜶𝒕 ∆𝑻 + 𝜺
𝛅 𝟎.𝟐𝟓
𝑳=𝑪 = 𝟎.𝟔𝟓 =894’’
𝜶𝒕 ∆𝑻+𝜺 𝟓.𝟓×𝟏𝟎−𝟔×𝟔𝟎+𝟏×𝟏𝟎−𝟒
= 75ft
Lengths typically between 30’-100’
JRCP REINFORCEMENT
If (when) concrete cracks, steel picks up
Steel reinforcement are used for control of temperature cracking
Do not increase structural capacity of the slab
Used to increase joint spacing
To tie the cracked concrete together
𝒇𝒂 𝜸𝒄 𝑳𝒉
𝑨𝒔 =
𝟐𝒇𝒔
53
6/8/2018
Climate
Material
Geometry
54
6/8/2018
RESPONSE MODELS
Finite element methods differing in the adopted foundation types
Liquid foundation (Winkler foundation)
Solid foundation (Boussinesq foundation)
Layered foundation (Burmister foundation)
MECHANISTIC-EMPIRICAL PROGRAMS
MEPDG (ASHTOWARE)
KENSLAB
55
6/8/2018
log N f f
S
f 1 2
Where c
DAMAGE RATIO
For a known amount of traffic volume, a damage ratio is then
calculated for each mode of damage (fatigue)
𝑛 𝑚
𝑁𝑖,𝑗
𝐷𝑟 =
𝑁𝑓𝑖,𝑗
𝑖=1 𝑗=1
56