Grupo de Investigação em Vias de Statistical bituminous ... · Grupo de Investigação em Vias de Comunicação. Mechanical performance evaluation: StiffnessModulus+ PhaseAngle

PORTUGAL

ENGINEERING INSTITUTE OF COIMBRA

ISEC FCTUC

UNIVERSITY OF COIMBRA

POLYTECHNIC OF VISEU

Silvino Capitão António Baptista Luís Picado-Santos

Grupo de Investigação emVias de Comunicação

Statistical analysis of unconventional bituminous mixtures’ performance based on 4PB Tests

ESTV

Grupo de Investigação em Vias de Comunicação

High Modulus Bituminous Mixtures (HiMBM):

BinderType: 10/20 pen bitumenContent: 4.8% to 5.8%

AggregateType: graniteMax. Size: 20 mm

Pavement LayerBase

Unconventional bituminous mixtures tested

2


Hot‐mix Bituminous Recycled in plant (HMR):

BinderType: 35/50 pen + recoveredContent: 4.1% to 4.4%

AggregateType: limestone mainly Max. Size: 25 mm

% of recycled mixture 20, 30 and 40%

Pavement LayerBase


Mechanical performance evaluation: Stiffness Modulus + Phase Angle

3


σ(t)= σ0 sin(ωt)

ε(t)= ε0 sin(ωt ‐ φ)

Stiffness Modulus

StressStrain

Smix =

Phase Angleti – time between σ(t) and ε(t)tp- loading cycle time

360 p

i

tt

=φ

degree

Controlled displacement


Mechanical performance evaluation: Fatigue behaviour

4


BituminousLayers

Nadm

εt

Number of loading cycles

Stra

in (ε

t)

εt = a N‐b

Fatigue law obtained by the classical approach: specimen fails when required load to keep displacement level has decreased 50 per cent of its initial value.


HiMAM: evaluated parameters

5


AB

HiMAM: stiffness results (example for lab.A2: 5.3% bitumen content)

6


5 Hz0,1 Hz 1 Hz

10 Hz

Cov:9.7%9.6%10.0%11.5%

Cov:6.2%4.0%10.5%15.3%

Cov:5.4%6.5%4.8%‐‐‐

Cov:4.1%10.9%16.1%‐‐‐

Temperature (oC)

Stiffne

ssMod

ulus

(MPa)

εt = 120 μm/m

HiMAM: stiffness results (example for lab.A2: 5.3% bitumen content)


25ºC

45ºC

40ºC

15ºC

Cov:11.5%15.3%‐‐‐‐‐‐

Cov:10.0%10.5%4.8%16.1%

Cov:9.6%4.0%6.5%10.9%

Cov:9.7%6.2%5.4%4.1%

Frequency (Hz)

Stiffne

ssMod

ulus

(MPa)

εt = 120 μm/m

HiMAM: phase angle (example for lab.A2: 5.3% bitumen content)

8


Cov:‐‐‐‐‐‐‐‐4.5%7.2%

Cov:8.9%4.9%4.1%3.9%

Cov:9.6%6.5%9.3%7.5%

Cov:9.5%8.1%10.8%8.7%

Frequency (Hz)

PhaseAn

gle(o)

25ºC

45ºC

40ºC

15ºC

εt = 120 μm/m

HiMAM: statistical analysis of stiffness results

9


Specimens from trial sections

LaboratorySpecimens

t –test: The difference between the averages of stiffness modulus is not significant, for a 5% significance level.


Wilcoxon Signed Rank Test:hypothesis of identical populations is rejected for a 5% significance level.

HiMAM: regression analysis of stiffness results

10



bmaxbmix S x 33.6 x 94.16V x 50.853T x 90.331t x 10.215130231S +ε−−−−=

Smix (MPa) ‐mixture’s stiffness modulust (s) ‐ loading timeT (oC) ‐mixture temperatureVb (%) ‐ percentage of bitumen by volumeεmax (mm/m) ‐maximum strainSb (MPa) ‐ bitumen stiffness.

R2 = 0.89

Smix (MPa)

Traffic Speed (Km/h)

Example for a typical situation:

HiMAM: regression analysis of fatigue results (examples)

11



Classic Analysis

100

1000

1000 10000 100000 1000000 10000000

Load cycles

Stra

in(m

icron

s)

Lab.A3(bit=5.8%)

Lab.A2(bit=5.3%)

Lab.A1(bit=4.8%)

εt=a x Nb

Effect of aggregate inter-locking along

crack borders

70% stiffness reduction

R2=0.93

R2=0.904

R2=0.947

HiMAM: regression analysis of fatigue results (examples)

12



100

1000

1E3 1E4 1E5 1E6 1E7Number of load cycles

Stra

in (m

icron

s)

A2.2 Vv=1.95 %

A2.1Vv=1.79%

Lab.A2Vv=1.73%

205.4 214.2 22

8.0

207.7

248.8

220.8

50

100

150

200

250

300

350

A1.1 A1.2 A2.1 A2.2 A3.1 A3.2Trial sections

ε 6(m

icron

s)

εt = a N -b ε6 strain which induces specimen ruin

after 1 million load cycles

(bit= 5.3%)


HMR: test parameters

13


HMR: stiffness results (example for mix D: 4.3% bitumen content)

14


25oC

40oC

15oC

Stiffne

ssMod

ulus

(MPa)

εt = 100 μm/m

Frequency (Hz)

Stiffne

ssMod

ulus

(MPa)

Frequency (Hz)

Specimens from pavement Specimens from laboratory

40oC

25oC

15oC

Higher stiffness values:Better control of the mixing process...

HMR: regression analysis of fatigue results (example for mix D: 4.3% bitumen content)

15


εt = 100 μm/m

Strain(μm/m

)

Load cycles

Specimens from pavement Specimens from laboratory

25oC

Better performance:Better control of the mixing process...

25oC

Load cycles

Strain(μm/m

) ε6 = 258

ε6 = 205

16

CONCLUSIONS• HiMBM: The prediction models, bearing in mind their limitations, can be used within the range of characteristics studied for HiMBM, for mixture composition studies and for design purposes (stiffness modulus model)

• It is adequate assessing fatigue resistance of HiMBM from specimens produced in laboratory instead of building trial sections, particularly for mix design purpose (Parameter ε6 seems to be adequate for ranking mixtures during the mix design process)

• HMR:It was found that mixtures produced in laboratory have a better mechanical behaviourthan mixtures produced in plant. This underlines the fact that better fabrication and lay‐down control are significant issues.

17

CONCLUSIONS

• HMR:Hot‐mix bituminous recycled in plant is possible, in technological terms, with 40% of reclaimed material incorporation by total weight.

Finally, it can be affirmed that mechanical behaviour of unconventional mixtures similar to HiMBM and HMR can be evaluated from four‐point bending tests.

18

PORTUGAL

Grupo de Investigação emVias de Comunicação

ISEC FCTUC

UNIVERSITY OF COIMBRA

POLYTECHNIC OF VISEU

Silvino Capitão António Baptista Luís Picado-Santos

ESTV

Thank You for Your Attention.

Documents

Grupo de Investigação em Vias de Statistical bituminous ... · Grupo de Investigação em Vias de Comunicação. Mechanical performance evaluation: StiffnessModulus+ PhaseAngle