investigation of the modified lottman test to predict the stripping performance of pavements

8/18/2019 Investigation of the Modified Lottman Test to Predict the Stripping Performance of Pavements

1/70

Report No. COOT

OTO R 93 3

INVESTIG TION OF THE

MODIFIED

LOTTM N TEST

TO PREDICT THE

PERFORM NCE OF

IN

COLOR DO

Timothy Aschenbrener

Colorado Department of Transportation

4340

East

Louisiana

Denver. Colorado 80222

Robert

B. McGennis

The

Asphalt Institute

Lexington

Kentucky

Final Report

April 1993

Prepared

in cooperation

with the

U S Department

of

Transportation

Federal

Highway of Administration

STRIPPING

P VEMENTS


2/70

The

contents of

t h i s repor t r e f l ec t

the

views of

authors

who are

respons ible

for the fac t s and

the

accuracy

of the data

presented

herein .

The

contents do not necessar i ly

r e f l ec t

the of f ic ia l

views

of

the

Colorado

Department

of

Transporta t ion

or

the

Federal Highway Adminis trat ion. This

repor t

does not cons t i tu te

s tandard

spec i f ica t ion o r

regula t ion .


3/70

CKNOWLEDGEMENTS

Larry

Lamar,

Dave

Gallegos, Brad Black,

Ken

Medina, Ju l ie

Le

and

Kim

Gilbert of the Colorado Department of

Transportat ion

CDOT) performed a l l of

the te s t ing

for t is study.

I

am also

grateful for the numerous Region Materials personnel

who

sampled

and delivered the

3

tonnes of aggregate used in the study.

The

CDOT Research Panel provided

many

excel lent

comments and

suggestions

for

the

study.

t included:

Gerry

Peterson

and

Jay

Goldbaum CDOT-Region 1 Materials) , Bob

LaForce CDOT-Staff

Materials) ,

Donna

Harmelink

CDOT-Research),

and

Mark

Swanlund

FHWA-Colorado

Division .

Special Thanks to

the

expert panel of Colorado

asphal t

paving

experts

who

provided

numerous

ideas

and

suggestions which

made

t i s

study more informational:

Bud

Brakey Brakey Consult ing

Engineers),

Jim Fife

Western

Colorado Testing),

Darrel

Holmquist

CTL/Thompson),

Joe

Proctor Morton

Internat ional) , and Eric

west

Western Mobile .


4/70

Technical Report Documentation Page

1.

Report No.

2. Government Accession No.

3

Recipient s Catalog No.

CDOT-DTD-R-93-3

4. Title and Subtitle

Investigation of the Modified Lottman Test to Predict the

Stripping Performance of Pavements in Colorado

5 Report

Date

Aj)ril1993

6.

Performing Organization

Code

File No 13.07

7. Author(s)

Timothv Aschenbrener and Robert

B

McGennis

9. PCI·forming

Organization Name

and Address

Colorado Department of Transportation

4201 East Arkansas Avenue

Denver Colorado 80222

12. Sponsoring Ageney

Name

and Address

Colorado

Department

of Transportation

4201 East Arkansas Avenue

Denver, Colorado 80222

15. Supplementary Notes

8. Performing

Organization

Rpt.No.

CDOT-DTD-R-93-3

10.

Work Unit

No. (TRAIS)

11. Contract

or

Grant No.

13. Type of

Rpt

and

Period

Covered

Final

Renort

14. Sponsoring Agency Code

Prepared in Cooperation with the U.S. Department of Transportation Federal

Highway Administration

16.

Abstract

Moisture damage to

hot

mix asphalt pavements has been a sporadic

but

persistent

problem in Colorado even though laboratory testing is performed to identify moisture

susceptible mixtures. The laboratory conditioning was often less severe than the conditioning

the hot mix asphalt pavement encountered in the field.

Twenty sites of known field performance with respect to moisture susceptibility, both

acceptable and unacceptable, were identified. Material from these sites were tested using

seven versions of the modified Lottman test (AASHTO T 283) and the boiling water test

(ASTM D 3624).

For modified Lottman testing, two levels of severity for conditioning laboratory samples

were identified that correlated well with conditions in the field.

For

mixtures placed under

high traffic, high temperatures, high moisture, and possible freeze, the severe laboratory

conditioning defined in the report should be used.

The

milder laboratory conditioning

defined in this report

is

appropriate for low traffic sites.

17. Key Words

Stripping

Lottman

Case Histories

19.5ccm·ity ClassiC. (rcpOl·t)

Unclassified

Boiling Water Test

Moisture Susceptibility

20.Seew·ity ClassiC. (page)

Unclassified

18.

Distribution Statement

No Restrictions: This report is

available to the public through

the National Technical Info.

Service.

Sprindield VA

22161

21. No. of Pages

7

22.

Price


5/70


6/70

VI.

TEST RESULTS

ND

DISCUSSION

• • • • • 21

VII.

VIII

Analysis

of

Untreated

Aggregate Quali ty

• • •

22

Inf luence-Freeze/No

Freeze/30-Minute

Satura t ion 22

Comparison

of

TSR s with Actual Performance • • • • • 25

Analys is

of

Swell

29

Analysis

of

Short

Term

Aging

• •••

31

Analysis

of

Lime Addit ion

• • • •• 34

Analysis

of the

Boil ing Water Test ••

35

MODIFIED LOTTM N REPEATABILITY

• • • • • • • • • • • •

36

C SE

HISTORIES

OF

OTHER INVESTIGATIONS •

37

Stuart

1986 37

Kennedy

1983

38

Parker

1988

39

Summary

39

IX

CONCLUS IONS 40

X IMPLEMENT TION

• •••• • ••••••••

43

XI.

RECOMMEND TIONS FOR

FUTURE

RESE RCH

• • • • • • • • • • • • 44

XII REFERENCES 45

v


7/70

LIST OF

T BLES

Table Page

Number Number

1 Si tes Used in This Study . • • • • • • • • . . . • . . . . . . . . . . . . • • 2

2

Aggregate

Gradations

and

Optimum

Asphalt

Contents

for HM Mixtures

Used in

This Study• • •

10

3 Comparison of the Original Lottman

to

Currently

Recommended

Procedures

. • • • • • . . . • . . . • • • • • • • • • . • • • • • 11

Experimental

Grid

for the Stripping Study

• • • . • . • • . 19

5 Comparison of Tensile

Strength Ratios

• • 23

6

Comparison of

Pavements

of Known

Field Performance

with AASHTO

T

283 Severi ty Level

2B)

••••

25

7 Comparison of Pavements of Known Field Performance

with AASHTO T 283

with No Freeze

Severi ty Level

2C) • . . • • • • • . . . . . . . • . . • • . • • • • • • • • • • 25


with

AASHTO T

283 with

a

30-Minute

Saturat ion

Severi ty Leve l l ) • . • . . . . • • . • • • . . . . . • • . • . • . • • • • • . .

25

9 Summary of Swell After Saturat ion

. • • • • • • • • • • • • • • • •

28

10

Summary of

Swell

After Conditioning • . • • • • • • • • . . • • •

31

11 Summary of Dry Tensile

Strengths

and Tensi le

Strength

Ratios

Using

Various

Lengths

of

Short-

Term Aging STA) • ••••••• ••••••••••••••••

32

12

Boil ing Water Test Results • . . • • • • • • • . . • . . . . • • • • • . .

36


with

the Boiling

Water Test

ASTM D 3625)

• • • • . • • . •

36

14 Comparison

of

the Lottman Test Level 2A) to

Actual Performance •• • • •• ••• 37

15 Comparison of the Lottman Test

Level

2C) to

Actual Performance

8

16 Comparison of the Lottman Test Level l )

to

Actual Performance • . . • . . . . • • • • ·

• •

38

17 Comparison

of

the Lottman Test Level

2B) to

Act

ual Performance 9

18 Comparison of the Lottman Test

Level

2C) to

Actual Performance 9

v i


8/70

LIST

O

FIGURES

Figure

Page

Number Number

1 A High

Maintenance

Mix

Experiencing

Raveling

f te r 15 Months •

2 A

Disintegrator After

6

Months

• • • • • • • • •

4

3

The

Surface

of

a Pavement

Requiring

Complete

Rehabi l i ta t ion

in 12

Months • • • • • • • • • • • • • •

6

4 A Core Showing

Str ipping

Below

the

Surface from

the

Pavement n igure

6

5 Environmental Conditions

for the

Pavement a t

Agate

Requiring

Complete

Rehabil i tat ion

f ter 12

Months

• 7

6 Environmental Conditions for the Pavement a t Cedar

Point

Requiring Complete

Rehabi l i ta t ion

f ter 11 Months 7

7 Environmental Conditions for the Pavement a t Arriba

Requiring Complete

Rehabil i tat ion

f te r 1 Months • • • • • 8

8 Environmental Conditions for the Pavement a t

Limon

Requiring

Complete

Rehabil i tat ion

f ter 1 Months • 8

9

Ranked

Order

of

Tensile Strength Ratios

from

SHTO

T 283 with a 30-Minute Saturat ion • • • • • • • • • •

27

1

Relat ionship

of

Swell After

Conditioning Versus

Tensi le Strength Ratio from

Samples Tested According

to

SHTO

T 283 with 30-Minute Satura t ion • • • • •

30

11 Influence of Short-Term

Aging on

ry Tensi le Strength • ••••••• 33

12 Influence of

Short-Term

Aging on

Tensi le Strength

Ratio • • • • • • • • • • • • • • • • • • • •

33

v i i


9/70

PPENDICES

ppendix • • Summary of Laboratory Test Data

ppendix

B

• • • •

Comparisons

of

Different

Versions

of the

Modified Lottman

Test

from

Other

Invest igators

vi i i


10/70

I INTRODU TION

Moisture damage,

otherwise

known as s tr ipping , to hot mix

asphal t HMA) pavements has been

a

sporadic but pers is tent

problem on

projec ts

in Colorado. After

premature

moisture

damage was observed on a

specif ic

project in

July

of

1991, the

Asphalt Ins t i tu t e was requested

to

perform a jo in t study with

the Colorado Department of

Transportation

CDOT). The

resul ts

of the jo in t study

were

reported by

McGennis (1)

in october of

1992. Recommendations included:

-

evaluating aggregates of

known

f ield performance with

several versions of the

moisture

suscep t ib i l i ty t es t

used by Colorado,

-

evaluating aggregates of

known

f ie ld performance

without

lime

or l iquid anti-s tr ipping

addit ives,

-

evaluating the sand

equivalent t e s t

-

implementing

a

bet ter

P200 management

st ra tegy during

construction,

l imi t ing the quantity

of

P200

in

HMA,

and

-

mill ing ruts instead of using level ing

courses.

This report

presents resul ts

from

the

f i r s t

two

recommendations;

a l l

of

the

other recommendations have been invest igated and

implemented. The

purpose of

th is

report s

to

compare HM

pavements of known f ie ld performance with resu l ts from

various

laboratory

moisture

susceptibi l i ty

t e s t s .

The laboratory

t e s t s

invest igated

were

the modified

Lottman

AASHTO

T

283)

and

various versions, and

the

boil ing water t e s t ASTM D

3625).

1


11/70


12/70

Good Some aggregate

sources in

Colorado have a

good

his tory of providing pavements

tha t

re s i s t moisture

damage.

Seven dif fe ren t aggregate sources with a history of excel lent

performance were selected for invest igat ion.

A

specif ic

project using

each

aggregate

source was then

studied

in deta i l

for

th is

invest igat ion.

High

Maintenance

These pavements have

received

an

exceptionally high level

of

maintenance.

Although

pavements

in th i s category are s t i l l in service

a f te r

two to f ive

years,

the i r

performance i s

considered

unacceptable

when

compared

to

the i r design l i f e .

The

maintenance required to

address

problems

from

moisture

damage to

the

HM

pavements

included

overlays and s ignif icant patching

of

potholes . A 15-month old

pavement

tha t required an overlay

on some

sections i s

shown

in Fig.

1 .

Dis integrators

There

are several

aggregate sources used

in

HM

pavements

tha t

have a notorious

history of

severe

moisture

damage.

A 6-month

old

pavement

tha t

disintegrated i s

shown

in Fig. 2. Since

contractors

have not used

these

aggregate

sources

on

CDOT

projects for many years ,

specif ic

mix

designs for

the

dis in tegrators

were di f f i cu l t

to

obtain.

The

mix designs with

the

aggregate sources thought

to

be

disintegrators

were reproduced

as

closely

as

possible

with

the

help

of

experienced, long-term employees of

the CDOT

Complete Rehabil i tat ion

Several pavements in

Colorado

required

complete

rehabil i ta t ion

when le ss than

two years old,

and

often when

le ss

than

one

year

old.

The

moisture

damage

was

re la ted to

a

unique

pavement design

feature, ru t res i s tant


13/70

Fig . I .

A

"High Maintenance" Mix R t p ~ r ; . i . p . n c i n g Rave l ing

Af t e r

I '

~ ' : o . n t l 1 . n .

Fig . 2 .

A

Dis in t eg ra to r

Af t e r

Months .


14/70

composite pavement,

tha t

ut i l ized a plant mixed seal

coat

PMSC) as described and evaluated by Harmelink

(4).

HM

pavements

di rec t ly

below the PMSC exhibited severe moisture

damage.

The

pavement

surface (Fig.

3)

and a core showing

the

moisture

damage

tha t occurred jus t below the surface

(Fig.

4)

are

shown

for

a

pavement requir ing

complete

rehabi l i ta t ion

a f t e r 12 months.

Even

though

the

PMSC was a

contribut ing

£actor to

the

dis t ress in the underlying

HMA,

the HM was s t i l l

considered to

be

susceptible to

moisture damage

since t fa i led

so

quickly.

The weather

condit ions

tha t contributed to the fa i lure of the

pavements

requiring

complete

rehabil i ta t ion

examined

in

th i s

study are shown in Figures 5 through

8.

The temperature i s

the

monthly mean maximum

temperature,

i e the average of the daily

high temperatures.

The precipi ta t ion i s

the

to ta l

accumulation

for

the month. The f i r s t month and

year

in

each

f igure

represents the

end of construction, and

the

f ina l

month

and

year in

each f igure

represents the time of

fa i lu re

Pavements

requiring complete rehabil i ta t ion a l l fa i led when

high

levels of precipi ta t ion

occurred in

the

hot tes t

part

of

the summer

Even

though

a l l

pavements in Colorado

are

subjected to freeze cycles, the

severe

moisture damage did

not correspond

with

freezing conditions.

The instantaneous

fa i lu res were direc t ly re la ted to a

simultaneous

combination

of

high temperature,

high

moisture, and

high

t r a f f i c

The

environmental data

used in

th is report was obtained from

the

weather

s ta t ion

located

closest

to

each

project

and

reported

by

the

National Oceanic

and

Atmospheric

Adm inistrat ion s

National Climatic

Data

Center.

5


15/70

Fig . 3 . The Surface o f a Pavement Requir ing

Comolete Rehab i l i t a t i on in 12 Months

F ig

4 A Core Showing St r ipp ing Below

the

Surface

from · t h r ~

P q N ~ m e n t

in Figure 3.

6


16/70

Environmental Conditions

Agate

100

Temperature

recipitation

~ - : : : : - - - - - - - - - - - , 4 . 5

90

80

70

60

50

40

30

20

10

8 9

10

12 2

3

4

88

89

Month and Year

_ Temperature

~

recipitation

5 6

4

2

1 5

1

Fig .

5 . Environmenttd

Condi t ions f o r t h e Pavement a t

Agate Requir..' ; ng Complete Rehab i l i t a t ion

A f t e r

12 : 1 ( o I · : h ~ .

Environmental Conditions

Cedar Point

Temperature recipitation

100 , - - - - - - - - - - - _ _ 4 5

90

4

80 3 5

70 3

60

2 5

50

2

40 1 5

30

20

10

9 10 11

2

2 3 4 5 6 7

89 90

Month and Year

_ Temperature

~

recipitation

0.5

o

Fig .

6

Environmenta l Condi t ions

fo r t he Pavement a t

Cedar Po in t Requir ing nComplete

R e h a b i l i t a t i o n

n

A f t e r

11

Months.

7


17/70

Environmental onditions

100

90

80

70

60

50

40

30

20

10

Arriba


,

i i i

'

,

•

8 9

10 2 1 2

3

4

5 6 7 8 9

10 2 1 2

3

4 5

6 7

I

89

I

90

I

91

I

Month and Year

_


4.5

4

3.5

3

2_5

2

1.5

0.5

o

Fig . 7 . Environmenta l Condit ions fo r

t h e Pavement a t

Arr iba Requi:o::ing "Complete R eh ab i l i t a t i o n

n

Afte r

24.

Mon-ths

.

Environmental onditions

Limon


100 ---

--------------------

------------.4.5

90 4

80

3.5

70 3

60 2.5

50 2

40 1.5

30

20

10

10

2

90

34

9

Month and Year

5

_ Temperature recipitation

6 7

0.5

o

Fig . 8 . n v i r o n ~ q n t a l Condi t ions fo r t h e

Pavement

a t

Limon Ren"; .:: "-:-·;_:'1q "Complete R eh ab i l i t a t i o n

n

Afte r 10

on-'.:hn_

8


18/70

I I I . T ST

METHO OLOGY

The

original

mix design used

a t

each s i t e was

ident i f ied

Information

retr ieved

included

the aggregate sources

percentage of each

component

aggregate stockpile

component

and combined aggregate gradat ions optimum asphalt content

asphal t cement source and grade

and

anti s t r ipping treatment.

I t

was

not

possible to

use

the exact

aggregates and

asphal t

cements

from

the

original projects placed two to ten

years ago.

So

virgin aggregates

from the original

sources

used a t each

s i t e

were

sampled.

Addit ional ly

recent ly

produced

asphal t

cements

and anti s t r ipping treatments were

obtained

from the

original suppl iers

of materials

to the s i t e s .

The aggregates

from each s i te were then

blended

to

match

the

gradation used

on

the

project

as closely as possible.

A

mix

design

was then performed

to val idate the

optimum

asphal t

content

from

each

s i t e When

the optimum

asphalt content

of

the

new

mix

design

matched the

optimum

asphal t

content of the

original

mix

design

the moisture susceptibi l i ty t es t ing

proceeded. When the

optimum asphalt content of the

new

mix

design did

not

match

the

optimum

asphal t

content of the

original

mix

design t

was assumed the

aggregates had

changed

and the

new optimum asphalt

content was used.

o optimum

asphal t

contents used

in

th i s study varied

by more than 0.2

from the

original

designs.

The

aggregate

gradations

and

optimum asphal t

contents

of

the

HM mixtures used for th i s study are

shown

in

Table

2.

9


19/70

Table 2 Aggregate

Gradations and Optimum Asphalt

Contents

for RN Mixtures Used in This Study

Gradation mm and inches)

s i t e AC 19.0

12.5

9.50 4.75 2.36 0.60 0.30 0.15

0.08

3/4 1 /2

3/8

#4 #8 #30 #50 #100 #200

1

5.5 100

87

72

51

45 26

18

10

7.0

2 4.5

100

87

74

53

42

24

15 10

6.6

3

5.3

100

93

77

53

37 21

14

9

5 .9

4

4.9

100

88

66

50

40

21

14

8

5 .1

5

5 .0

100

94

80

52 41 31

18

10

7.1

6

6.0

100

88

51

37

22 14

10

5.9

7 5.7

100

91 74 49 3 7

18

12

8 4.7

8 4.8 100 94

77 49 38

24

18

12

8.1

9 5.9

100

96

62

41

25 13 10

6.1

10

5.0

100

86

77

55

43

26 18

13

8.6

11 4.9

100

97

57

40

21

15

11

7.8

12

5.0

100 86

76

54

42 25

18

13

8.4

13

5 .7 100

86

78

60 45 22 15

9 5.7

14 5 .3

100 86

78

63

47

25 16 10

7.7

15

5.6 100 85 76

62

49

27

18

13

8.3

16

5.4

100

88 79

61

50

30

20 13

8.3

17 5.6 100

95 72

44 24

17

12

7.3

18

5.6 100

95

70

39

21

15 11

7.2

19 5.5 100

96

93

83 69

32

20

14 11.

7

20 6.5

100

96

80

50

42 26 18 12

8 .3

IV. DEVELOPMENT

OF

THE

MO IFIE

LOTTM N

PROCEDURE

A modif ied Lottman

procedure i s

most commonly used

by

the CDOT

for mois ture

suscep t ib i l i t y t e s t ing of HMA.

Since nat ion-wide

exper ts have made numerous

modif ica t ions

to the procedure ,

understanding ts h i s to r i ca l development

i s

c r i t i c a l . In Table

3, the

or ig ina l procedure developed by Lottman

5,6 ,7) i s

compared

to

the two

most

commonly used vers ions today

AASHTO

T 283 and ASTM D

4867).

10


20/70

TABLE

3.

COMPARISON OF THE ORIGINAL LOTTMAN

5)

TO CURRENTLY

RECOMMENDED PROCEDURES

I

Or ig ina l

Modified

Lottman

Lottman

Ref.

S)

AASHTO T 283

ASTM

D

4867

Short-Term None

Loose mix:

None

Aging 16

hrs @

60·C

Compacted mix:

72-96 hrs

@

2S·C

Air Voids 3 to 5

6 to 8 6 to

8

Sample

Random

,Average

a i r Average

a i r

Grouping voids of

two

voids

o f

two

subse t s

should

subse t s should

.

be equal

be equa l

Satu ra t ion

100

55 to 80 55 to 80

Freez

e

15

hrs

@

-18·C

Min.

16

hrs

Optiona l :

@

-18·C

15

hrs @ -18·C

Hot water

Soak

24

hrs

@

60·e

24

hrs @

60·C 24

hrs

@

60

G

C

Strength

Ind i r e c t

In d i r e c t

Ind i r e c t

Proper ty

t ens ion o r

t ens ion t ens ion

diamentra l

modulu s

Loading Rate

1 .6

mm/min.

51 . rom/min.

'

51

mm/min.

@

13

G

C

@ 25·C

@

25·C

Prec i s ion

10

fo

r TSR

None

8 ps i

fo r

Statement i nd i r e c t

for

a Sing le

t e n s i l e

Operator s t r en g th

wet o r

dry)

Not spec i f i ed , bu t

r ep resen ta t ive

o f

a t yp ica l value

encountered .

11


21/70

n i t i a l Study

From

1968 through

1982, Lottman

(5,6,7,8,9,10,11) developed

a new

t es t procedure

to

identify

the moisture susceptibi l i ty of an HMA

The

procedure was

ver i f ied with aggregates

of

known

f ie ld performance and through

the construction of t e s t sections.

In 1970, Lottman (10) reported tha t

moisture damage

could

resul t from

excess

pore pressures tha t developed

in

the HM

pavement

from t r a f f i c and thermal

expansion.

Therefore,

the

moisture suscep t ib i l i ty t es t procedure included conditioning

phases

tha t created pressure

within

the

a i r

voids of the HM

sample.

The

condit ioning

included high

levels

of

sa tura t ion

and a freeze cycle to create pore pressure.

In

1974, Lottman (5) calculated indirect t ens i le s trength and

modulus

ra t ios as

the

value from

the

conditioned sample divided

by

the value from the unconditioned sample.

Conditioning

included vacuum sa tura t ion

followed

by e i ther single or

multiple cycles

of freezing

and hot-water soaking. Loading

ra tes

for

the indirect

tens i le

s trengths

were examined

a t 1.6

mm/min. (0.065 in . /min.)

a t

13

0

C

(55

0

F and 3.8 rom/min.

(0.15 in . /min.) a t 23

0

C

(73

0

F).

Lottman

(6)

reported the procedure in

1978,

and Lottman (7)

f inal ized

the

procedure in 1982. Testing parameters for

the

or ig inal Lottman procedure are

shown

in Table 3. In 1982,

f ie ld

evaluations on

eight t e s t

sections in seven s ta tes

including one

in

Colorado, provided validat ion of

the Lottman

moisture

suscep t ib i l i ty

t es t

(7).

12


22/70

Loading

Rate

In 1979,

Maupin

(12)

performed a study

to

implement the

1978

Lottman t e s t

procedure

for

Virginia.

For convenience

of using exis t ing

equipment, he

recommended a

loading ra te of 51 mm/min. 2 in . /min.)

a t

25

0

C 77

0

F) instead

of

Lottman s recommendation of 1.6

mm/min. (0.065 in. /min) a t

13

0

C (55

0

F).

No

s t a t i s t i ca l ly

s ignif icant difference

was

measured in the tens i le strength ra t ios between the

two

loading

ra tes

a t

the corresponding temperatures.

Testing in Colorado

was

performed

to

determine the difference

in the 51 mm/min. and 5.1 mm/min. loading ra tes Tensile

strength

ra t ios

prduced

from

the

two

different ra tes

were

iden t ica l

Dry strengths

using

the fas ter

ra te were 2.5 to 3

t imes

higher than those produced from the slow loading ra tes

The

two

modified Lottman procedures most commonly

used

today

recommend a loading ra te of 51 mm/min. (2 in . /min.)

a t

25

0

C

(77

0

F), as

shown in

Table 3.

Air

Voids Lottman (6) recommended compacting

the

laboratory

sample

to

match the projected a i r

voids

tha t would be in

the

HMA

pavement

a f te r approximately 6 years (3 to 5 a i r voids).

In order

to

improve

the

Lottman procedure,

Tunnicliff

(13,14,15) recommended modification of the ta rget a i r voids.

The sample should

be

compacted between 6

and 8 a i r voids

to

simulate

the

in-place voids of the HMA pavement

soon.

The two modified

Lottman

procedures

most

commonly used today

recommend l imit ing

a i r

voids

between

6

and

8 ,

as

shown

in

Table

3.

13


23/70

Saturation sample can

be damaged

i f

t swells

during

the

vacuum

sa tura t ion

process.

Based on

t es t ing of

aggregates of

known f ie ld performance,

Jimenez

16) indicated

tha t

swelling

during

vacuum

sa tura t ion

was

re la ted to s t r ipping

suscep t ib i l i ty Coplantz 17)

vacuum

saturated

samples

for

30

minutes a t a pressure of 610

of

mercury

to

provide

100

saturat ion. Vacuum

sa tura t ion

alone did not appear

to

in i t i a te

a st r ipping mechanism. Kennedy

18) found

excessive vacuum

saturat ion alone did

not

create

str ipping

unless the

aggregate

had shown poor st r ipping performance.

Stuar t

19)

performed

te s t ing

on

mixtures

of

known st r ipping

performance, both acceptable

and unacceptable. Based

on

t e s t

resul t s

from

the study, there was no conclusive evidence tha t

high

saturat ion

or

over-saturation

adversely effected

the t e s t

resul t s

The Lottman procedure 7) with high

sa tura t ion

and

the

modified Lottman procedure

developed

by Tunnicliff

15)

with

par t ia l

sa tura t ion were

comparable.

Dukatz

20) performed tes t ing that indicated no conclusions

could

be made

on

the effec t of

saturat ion

and

swell on

the

t ens i le

strength

of the

conditioned

sample. Various samples

saturated to high and low levels had

high

and low t ens i le

strengths.

In

order to determine i f par t ia l saturat ion could

predic t

pavement

performance, Tunnicliff

15) tes ted

the

eight samples

Lottman

7)

used for f ie ld verif ica t ion 6 years a f te r

Lottman).

He

concluded

tha t

l imi t ing sa tura t ion levels

correlated well with the

Lottman

procedure 7) tha t

allows

over-sa tura t ion

and

swell.

14


24/70

Tunnicl i f f 13) contended tha t damage from excessive

saturat ion could resul t in

low t ens i le

strength

ra t ios

even i f

the HM

was

not

moisture

susceptible. Although no t e s t

resul ts

were presented, past published

unreferenced)

and unpublished

l i t e ra tu re allegedly

supported

his claim. Over-saturat ion

of

one sample by

Tunnicliff

14)

indicated

tha t over-saturat ion

may be too severe.

The swell of

the

HM should

be

measured. The

two

modified

Lottman

procedures

most commonly used today recommend

l imit ing

the

level of saturat ion between

55

and 80 , as shown

in

Table 3.

Freeze Cycle

Tunnicliff

14,15)

also

recommended eliminating

the freeze cycle so

the

t e s t would be quicker and

easier

to

perform

for

f ield verif ica t ion

Lottman

5)

indicated tha t

the

freeze

cycle

with

high

sa tura t ion

predicted

the

str ipping

susceptibi l i ty of HM

pavements, even when the HM pavements in the f ie ld were

not

exposed

to freezing condit ions. Stuart 19)

indicated tha t

e i ther

the high sa tura t ion of

low

a i r voids with a freeze cycle

recommended

by Lottman

7)

or

the par t ia l

sa tura t ion of high

a i r voids

with no freeze

cycle

in

the modified

Lottman

procedure 15) would be

comparable.

I t was l ike ly tha t the

freeze was required

to

apply a s t ress in the sample as

discussed

by Professor

B.M. Gallaway

with Graf

21).

The

modified

Lottman

procedure

recommended

by

AASHTO

and

ASTM

allows

the freeze

cycle

to be optional,

as shown in

Table 3.

15


25/70

S h o r t ~ e r m Aging

he

time

of

exposure of the HM

sample

to

a

high

temperature

had

a s ignif icant

effec t on the

freeze-thaw

pedestal

moisture

susceptibi l i ty

t e s t

resul t s

he longer

the

HM

sample had exposure to a

high

temperature, the more

res i s tant t was to moisture

damage

(Graf,

21).

he

high

temperature

exposure increased the aging of the asphal t

cement

and provided

bet ter

coating, or

wetting .

Testing to

i sola te

the

two

variables indicated the increased resistance

to

moisture

damage was

primarily

re la ted

to the bet ter coating of

the

aggregate.

The

modified

Lottman

procedure

recommended

by AASHTO

requires

short- term aging whereas

the ASTM version has no short- term

aging,

as

shown

in Table

3.

Although

Lottman

(6) original ly

recommended the short-term aging speci f ied in AASHTO Lottman's

(7) final ized procedure

did not

mention short-term

aging.

Sample Grouping In 1987, Dukatz (20) reported the potent ial

for

tremendous

var iabi l i ty

in tens i le

s t rength ra t ios i f

the

average a i r

voids

in the conditioned and unconditioned

samples

were not equal. Ranges

in

the t ens i le s t rength

ra t ios were

as

high as 0.40.

When the

the

average

a i r voids

of

the

samples

in

the

conditioned and

unconditioned

groups were equal,

the

t ens i l e s t rength ra t ios were within

a

range

of 0.08.

he

two modified

Lottman procedures most commonly used today

recommend

the average

a i r voids of the conditioned and

unconditioned sample groups be equal, as

shown in

Table 3.

16


26/70

Mult iple Freeze Thaw Cycles . Several

researchers

have

invest igated

the

use

of multiple

freeze-thaw

condit ioning

cycles

and

have had varied conclusions. hen Lottman 5)

original ly invest igated the

procedure,

multiple

freeze-thaw

cycles

were

examined but

over-predicted

damage and were

considered too time

consuming

for pract ica l i ty

Lottman

concluded one freeze-thaw cycle

was adequate.

Scherocman 22) reported tha t multiple freeze-thaw cycles would

provide a

greater different ia t ion between the t ens i le strength

ra t ios

for

mater ia ls

with various levels

of

moisture

suscep t ib i l i ty

by Coplantz

17).

This conclusion i s supported

by

data

presented

To

the contrary, Kennedy 23) reported

tha t

ra tes of deter iorat ion in tens i le s trength

ra t ios

using

multiple freeze-thaw cycles were

not

s t a t i s t i ca l ly different

for

various

types of

materials . Materials with

higher

t ens i le

strength ra t ios a f te r

one

cycle would have

higher

t ens i le

strength

ra t ios a f te r

multiple cycles . The slopes or ra tes of

deter iorat ion of the

tensi le

s trength ra t ios were constant .

There i s tremendous var iabi l i ty in

conclusions drawn from

data

analysis of samples

tes ted

with multiple

freeze-thaw

cycles.

Additionally, t es t ing

samples with

multiple

freeze-thaw

cycles

requires an

addit ional

amount

of laboratory

t es t ing

time

tha t

may

not be

readi ly

available prior to paving.

Therefore,

multiple freeze-thaw

cycles were not

examined in

th i s

study.

~ h

Speci f i cat ion

The minimum

specif ied

t ens i le

s trength

ra t io

to

ensure

an

HM

pavement

will

perform

acceptably with

regard to

moisture susceptibi l i ty has varied.

Part of

the

reason

accounting for

the

specif icat ion to

vary

has been the

changes

in

the t es t

procedure. Based on

t es t ing of samples of

known f ie ld

performance,

tens i le

s trength ra t ios have been

17


27/70

recommended

a t

0.70 by

Lottman

(5)

and

Maupin

(12),

and a t 0.80

by

Lottman (7),

O Connor

(24),

and

Stuart

(19). A

survey of

s ta te

speci f ica t ions by

Tunnicliff

and Root (13) and

Hicks

(4)

revealed most s ta tes used 0.70 to

0.75

but

speci f ica t ions

ranged from 0.60 to 0.80. Computer

simulations

have resulted

in a t ens i le

strength

ra t io

recommendation of

0.85

by

Lottman (25).

Dukatz (20)

recommended a minimum t ens i le s trength a f te r

conditioning. A

very

weak

sample would not be

accepted.

This

also

prevented

acceptance of a mixture tha t had a low

unconditioned

t ens i le s trength ,

sometimes

caused

by

the

addition

of

l iquid ant i s t r ipping

addi t ives . Tunnicliff

(15)

recommended addi t ional research

be

performed to ident i fy a

minimum t ens i le

strength requirement.

v.

TEST PROCEDURES

Two different

t e s t procedures were investigated for

th is

study:

the AASHTO T 283 and the boi l ing

water

tes ts . Five different

var ia t ions

of

the

procedure were

used. Two

additional

t es t s

were performed: without any treatment for moisture

suscep t ib i l i ty and with hydrated lime. The procedures

used

to

inves t iga te

the

moisture

susceptibi l i ty of the HM

pavements

of

known f ie ld performance are described below. The experimental

grid of t e s t s performed on

samples

from the various s i tes i s

shown

in

Table 4.

Standard SHTO 283

The

materials

from

a l l

of the

s i t e s

in th i s

study

were tes ted

with

the

standard

procedure

(AASHTO T 283).

I t

includes

short- term aging, freezing, and

l imits on a i r

voids

(6

to

8 )

and

saturat ion

(55

to 80 ).

18


28/70

\

I

1

SHTO

T

283

X

SHTO

T

283

X

No f reeze

SHTO

T

283

X

30 min

sa t

SHTO T 283

X

No STA

SHTO

T

283 X

Double

STA

SHTO

T

283

X

No

addi t ive

SHTO

T 283

X

Lime

STM D 3625 X

Boi l ing Water

ST

A =Short Term Aging

T BLE 4

EXPERIMENTAL GRID

FOR THE STRIPPING

STUDY

GOOD PERFORMERS

HIGH

COMPLETE

M INTEN NCE

REHABILITATION

2

3

4

5 6

·7

8 9

10

11

12

13

14

15

16

X

X X X

X

X

X ·

X X X X X

X

X

X

X

X X X X

X

X

X

X

X X X X X

X

X

X

X X X

X X X

X

X

X

X X X X

X X X X X

X X X

X

X

X

X X X

X X X X X

X X X X X X X X X

X

X X X X X X X X X X X X X X X

X

X X X X

X X

X X

X

X

X X X X

X

X

X

X

X

X

X

X X X X

X X

X

X

DISINTEGRATORS

17

18

19

20

X

X

X

X

X

X

X

X

X

X X

X

X X X X

X X X X

X X X X

X X X X

X

X

X

X


29/70

S H ~

8l

No

Freeze) .

In

order to determine

i f

the

actual

pavement

performance could be predicted

without

the

freeze

cycle, the materials from a l l

of

the s i tes

in

th i s study

were tes ted without the freeze cycle.

SHTO

8l

lO-Minute Vacuum Saturat ion) . Some

inves t iga tors (17,18,20,26,27)

performed the modified Lottman

t e s t by

vacuum sa tura t ing a

sample

with

7 a i r voids

for 3

minutes. The procedure was performed with a vacuum sa tura t ion

of 3

minutes under

a

pressure

of 61

mm

of

mercury.

Consequently, the degree of

saturat ion was not controlled.

SHTO 283 No Short-Term Aging). The materials from

a l l

of

the

s i t e s in

th i s

study were tes ted without short- term

aging.

SHTO

283 Extra

Short-Term

Aging).

When HM i s produced

for a projec t

in Colorado,

a

sample i s

obtained

and delivered

to the Central Materials Laboratory for tes t ing.

After

delivery, the sample i s reheated for sp l i t t ing into the correct

sample size and reheated a

second

time for compaction.

In

to ta l the sample i s reheated approximately 4

to

8 additional

hours.

The ef fec t of the additional short- term

aging

was

invest igated. The

materials

from

a l l

of the s i t e s in

th i s

study

were

tes ted with

an additional

short- term aging of 5

hours

a t

121

C

(250

F

on

the loose

mix.

SHTO

283

No

Anti-Str ipping

Treatment). Numerous

aggregates

in

Colorado have

moisture

susceptibi l i ty

problems.

Anti-str ipping treatments in the form of lime and l iquid

anti s t r ipping additives have

commonly

been used

in Colorado.

The use of

lime

began

on

a

regional

basis in

the

early 1960 s

(28). The

DOT

specif ied

the use of l iquid anti s t r ipping

2


30/70

addi t ives

in

a l l

mixtures in

approximately

1983.

Even HM

with

l iquid

anti s t r ipping additives had

continued

problems

with

moisture susceptibi l i ty .

The

CDOT

then began

requiring

hydrated

lime in a l l

mixtures by

1 of the weight of the

aggregate in

1990.

Some

aggregate sources have modified Lottman t es t

resul t s tha t

are

very

low

when no ant i s t r ipping treatment

i s

provided.

McGennis (1) has

indicated

tha t

ant i s t r ipping

t reatments

should be able

to improve a

marginal HM

mixture,

but should

not

be expected

to

overcome severe

deficiencies. The

mater ia ls

from

a l l

of the s i tes in

th is

study

were tes ted

with

no

anti s t r ipping

treatment

to

determine

the

baseline

moisture

suscep t ib i l i ty potent ia l of the untreated HMA.

SHTO

T

283

With Lime). Some of

the

HM pavements

tha t

exhibited moisture dis t ress

were not

t rea ted with

hydrated

lime. The potent ia l moisture

susceptibi l i ty of

these

mater ia ls

with 1 hydrated lime by weight of the aggregate

was

investigated. I f an HM of

known

f ie ld performance did not

contain

hydrated

lime. When

constructed,

the procedure

was

performed

on

material from the s i t e with

hydrated lime.

oi l ing Water Test . Several studies have indicated the

boil ing

water

t es t (ASTM

D

3625) has

accurately

indicated the moisture

susceptibi l i ty performance of HM

pavements.

The t e s t

used

for

th i s

study involved

immersion of the sample in boil ing water

for

1 minutes. A

retained

coated area

over

95

i s usually

required.

VI.

TEST

RESULTS

ND DIS USSION

Results

from each

var ia t ion in

the AASHTO

T 283

t es t are

21


31/70

tabulated

in

Appendix A The

data includes a i r

voids,

saturat ion,

swell

a f te r

saturat ion and

conditioning,

conditioned and unconditioned

t ens i le

strengths, and

tens i le

s t rength

rat ios .

Analys is

o Untreated Aggregate Ouality .

Many

of the

aggregates t es ted

in

th i s study are

moisture

susceptible based

on known

f ie ld performance. Modified Lottman

t e s t

AASHTO T 283) resul t s were very

poor

when no l iquid

ant i - s t r ipping addi t ives or lime were used.

The

t ens i le

strength

ra t ios

of

samples t es ted

without treatment

are

shown

in Table 5.

Just enough treatment with l iquid an t i - s tr ipping

addit ives

or

hydrated lime

was

used

on

each

projec t

so

the

HM

samples would pass

the

t es t ; unfortunately condit ioning in the

f ie ld was more

severe than

conditioning in the laboratory,

and

the pavements fai led.

I t i s cr i t i ca l tha t the conditioning in the

laboratory

vacuum

saturat ion,

freeze,

hot-water

soak) be

equal to or greater

than

the

sever i ty

of condit ioning expected in the f ie ld. This i s

especial ly

important

when

marginal aggregates are used which

require

treatment with

l iquid

ant i - s t r ipping

addi t ives or

hydrated

lime. I f conditioning in the

laboratory i s less

severe

than condit ioning in the

f ie ld

an

engineer could

erroneously

assume

an

HM mixture would have good

f ie ld

performance for 10 or 20 years .

Field condit ions

re la t ing to

moisture damage

are

high t r a f f ic

high temperature,

high

moisture,

and possibly

freeze.

Inf luence

o

Freeze/No Freeze/30 Minute

Saturation.

Tensile

strength ra t ios for

samples

tested according to

AASHTO

T 283

freeze) , AASHTO

T 283

no freeze), and AASHTO

T 283

30-minute

sa tura t ion

and freeze) are

shown

in Table 5 .

No

s ta t i s t i ca l ly

22


32/70

s ign i f i c a n t d i f fe rence was found between AASHTO T 283 performed

with and without the f reeze

cycle .

By

using

the

30-minute sa tu ra t ion

and

f reeze , the

t e ns i l e

s t r e n g t h r a t i o s were s i g n i f i c a n t l y lower than t he samples

t e s t e d with p a r t i a l sa tu ra t ion AASHTO T

283

f reeze

and

no f reeze .

Table

5. Comparison o f Tensi le S t rength Ra t ios .

Tens i le

St reng th Rat ios

AASHTO

T 283)

S i t e Freeze No Freeze

30-Minute

Freeze

Sa tu ra t ion Note

1)

and Freeze

1.02

0.85 0.98

0.37

2

1.20

1.25 1.05 0.70

3 1.11 1.22 1.20

0.37

4

1.06 0.97 1.05

0.49

5 1.10

1.07

0.97 0.92

6 0.83

0.91

0.74

0.40

7

0.97

0.89

0.86

0.90

8

0.94

0.91

0.69

0.21

9

0.95 0.90

0.72 0.40

10 0.84 0.93

0.68

0.70

11

1.11

0.96 1.09

0.38

12

1.01

1.07 0.81

0.60

13 0.69 0.64 0.56 0.45

14 0.32 0.34 0.21

0.30

15

0.53 0.46 0.32

0.35

16

0.82

0.70 0.76 0.44

17

0.65

0.51 0.30 0.55

18

0.89 0.92 0.86

0.49

19

0.22

0.20

0.28

0.24

20

0.59 0.49 0.26 0.37

Avg.

0.84 0.81 0.72 0.48

S.D.

0.27 0.29 0.31 0.20

Note 1: The HMA sample was t e s t e d without t he use o f

l i q u i d

a n t i - s t r i pp ing

add i t ives

o r

hydrated

l ime.

23


33/70

In

separate

s tudies , Coplantz 17) and Dukatz 20) invest igated

samples compacted

to

approximately 7 a i r voids, Appendix B.

They

compared resu l ts from samples

tha t

were vacuum saturated

for

30 minutes

with

a freeze and par t i a l ly saturated

with

no

freeze.

They

found

tha t

samples

par t i a l ly

saturated

and

not

subjected

to

a freeze cycle

had tensi le strength ra t ios tha t

were two

to

three

t imes

higher than samples ful ly saturated and

subjected

to

a freeze cycle.

A modified Lottman t es t can be performed in

several

different

manners. Based

on

resu l ts from th i s study and others

15,17,19,20,27) shown

in

Appendix B, a

ranking of

t e s t s

in

decreasing

order

of

sever i ty

i s

l i s t ed

as

follows:

1) 30-minute saturat ion, 7 a i r voids,

freeze,

2A 30-minute

saturat ion,

4 a i r voids, freeze,

2B) 55-80

saturat ion, 7

a i r voids,

freeze,

2C) 55-80

saturat ion,

7

a i r

voids, no freeze.

All levels of sever i ty include hot-water soaking. Severi ty

Levels 2A, 2B, and 2C a l l

appear

to

provide approximately equal

resul ts . Severi ty Level 1 produces s ignif icantly lower t ens i le

strength

ra t ios .

The condit ioning

in

the laboratory should be

equal to

or

greater than

the conditioning in the f ield.

I t i s l ike ly tha t

different levels of f ie ld conditioning exis t throughout

Colorado. The

most

severe f ie ld conditions

are hypothesized

to be a

function of:

1) high t r a f f i c

2)

high

temperature,

3)

high levels of

moisture,

and

4)

possibly

very

low

temperatures.

High

levels of moisture

should

not

be determined on an annual

basis .

Based upon the performance of

pavements

requiring

complete

rehabil i ta t ion the amount of precipi ta t ion received in

a

high temperature

period i s more relevant.

24


34/70

The sever i ty level

of the modified

Lottman

t e s t

performed

in the

laboratory

should correlate with

the

sever i ty of the

f ie ld

condit ions. I f dif ferent

levels

of f ie ld condit ioning exis t

throughout

Colorado, then different

levels

of laboratory

condit ioning

should be

used.

Comparison

of Tensile Strength

Ratios with

Actual

Performance.

A

comparison of the various

severi ty levels of

the

modified

Lottman

t e s t

with pavements of known f ie ld performance i s shown

in

Tables 6, 7, and 8. A minimum

t ens i le

s trength

ra t io

of

0.80

was used.

Table

6.

Comparison

of

Pavements

of

Known

Field

Performance

with

SHTO

T 283 Severity

Level

28).

Good High

Complete Disint .

Maint.

Rehab.

Pass

7 5 1

1

Fail 3 3

Table 7.

Comparison of Pavements of Known

Field Performance

with SHTO

T 283

with

No

Freeze

Severi ty

Level

2C).

Good High

Complete Disint .

Maint.

Rehab.

Pass

7

5 1

ai l 4

3

Table

8. Comparison

of

Pavements of Known

Field Performance

with

SHTO

T 283 with a

30-Minute

Vacuum Saturation

Severi ty Level

1).

Good

High Complete Disint .

Maint. Rehab.

Pass

6

2

1

Fail

1

3

4 3

25


35/70

Based

on

these resul t s the

best predictor of

the

pavements

studied in

th i s

invest igat ion i s the modified Lottman t e s t

AASHTO T 283) with a 30-minute vacuum saturat ion. A plo t of

the t ens i le strength ra t ios in ranked order i s shown in

Fig.

9.

Although

a minimum

t ens i le

s trength

ra t io

of

0.80

was

used

to

develop Tables 6-8 and

Figure

9, considerat ion should

be

given

to

using a minimum t ens i le

strength

ra t io of 0.85,

as

recommended

by Lottman 25).

I t

i s known tha t many HM

mixtures

in Colorado

with

good

performance

would fa i l

the

sever i ty Level t e s t .

The

laboratory condit ioning specif ied for a par t icu la r

projec t

should

re la te

to the anticipated

f ie ld

condit ions.

Two

levels

of

laboratory

condit ioning should

exist . Severity

Level

should

be used

for high t r a f f i c s i t e s with severe environmental

conditions. Severi ty Level 2C should be used

for

low

t r a f f i c

s i t e s without

extreme environmental

condit ions.

Some out l ie rs

exis t

and

are

worth noting. Si te 11 i s a

quarried

source and very

f ine par t ic les conglomerated

onto

the

large par t ic les during processing of the aggregate. In order

to

minimize

the

generation of

dust,

a

large quantity of

water

i s sprayed on the aggregates. The dust

stays

out of the a i r

but remains

on

the aggregate.

After

the

water

dries the f ine

par t ic les can

be

removed by scraping an aggregate with

your

f inger

nai l .

I t

i s l ike ly tha t the

one cycle

of loading

applied by the modified

Lottman

i s not severe enough

to

penetra te the conglomeration. A t es t with multiple

cycles

would bet ter

identify

th i s

problem.

s i t e 6

has been

previously studied

and reported

by

McGennis 1). Identical t e s t s performed by the CDOT using

mater ia ls sampled

one

year apart were

substant ia l ly dif fe ren t .

26


36/70

o

a

I

CJ

Z

w

a

m

l l

J

m

z

w

f

0.4

0.3

0 21

1

3 4 2 1 5 18 7 12 16 6

9

8 10 3

5

7

19

20 14

SITE NUMBER

FIG.9 Ranked Order of Tensile Strength Ratios from AASHTO T 283 with a 30 Minute Saturation.

27


37/70

The di f fe rence

i s

not bel ieved to be

caused by r epea t ab i l i t y .

t i s wel l known t h a t the quarr ied source i s highly var i ab le

and conta ins seams of silt

and sha le .

When l a rge quan t i t i e s of

shale

are quar r ied , the ma ter ia l i s

l i ke ly to

be

highly

suscept ible

to

moisture, as

ind ica ted

by CDOT t e s t i n g performed

in 1992 Ref. 1 . When no shale

i s

quar r ied ,

the ma te r i a l

i s

l i k e l y to be marginal ly acceptab le as

ind ica ted

by

the

r e s u l t s

in t h i s

s tudy. Si tes

10 and 12 a l so had the same quar r ied

mater ia l .

Table 9. Summary of Swell Afte r

Sa tu ra t ion .

Swell

(

After Satura t ion ,

AASHTO

T

283

Si t e

Freeze

No 30-Minute No STA Ext ra

Freeze Sa tu ra t ion

STA

1

-0 .2

0.0

+0.1

-0 .2 0.0

2

-0 .1

-0 .2

0.0 -0 .2

-0 .1

3

-0 .4 -0 .2 -0 .4 +0.1 -0 . 4

4 -0 .3

-0 .3

-0 . 1

-0 .3

0.0

5 -0 .2 -0 .2 0.0

-0 .2

-0 .2

6

0.0

0.0

+0.1

-0 .1

-0 .2

7 -0 .6 -0 .3

-0 .4 -0 .2

-0 .3

8

-0 .5

-0 .4

+0.3

-0 .3 -0 .3

9

-0 .6

-0 .1 0.0 -0 .5 -0 .5

10

+0.2

-0 .6 -0 .3

0.0

0.0

11

-0 .7

-0 .7 -0 .2 -0 .2 -0 .7

12

-0 .5 -0 .3

-0 .1

-0 .6

-0 .8

13 -0 . 2

-0 .4

+0.4

-0 .6 -0 .3

14 -0 .3

+0.4

+0.8

0.0

+0.3

15

-0 .1

-0 .3 +0.3 0.0

0.0

16

0.0

-0 .1

+0.1

-0 .2 -0 .2

17 -0 . 2

-0 .2

0.0

-0 .2

-0 .2

18

+0.1

-0 .8 -0 .2 -0 .2

0.0

19

+0.3

-0 .1

+1.0 +0.2 +0.2

20

-0 .3

+0.2

-0 .2

-0 .1 -0 .5

STA

= Short-Term Aging

28


38/70

I f a

more judicious

aggregate

processing scheme

were

pract iced a t the quarries tha t provided

the

aggregates

for s i t e s

10,

11,

12,

and 16, a higher qual i ty HM

pavement could be produced

with

these aggregates.

Analysis o Swell Swell

was

measured

and

calculated a f te r

satura t ion

Table 9) and

a f te r

conditioning Table 10)

according to

the formulas

in

ASTM

D

4867.

For

samples saturated between

55

and 80 , there was no swell

more

than 0.5 a f te r

satura t ion.

For samples saturated for 30

minutes, swell more

than

0.5

af ter

satura t ion only occurred

on

two

samples

Sites

14

and

19): both

had

very

low

t ens i le

s t rength ra t ios and poor f ie ld

performance.

Swell often occurred a f te r conditioning. The amount of

swell

was

direc t ly

re la ted to the t ens i le s t rength ra t io

and known

f ie ld performance.

The higher

the

swell

a f te r conditioning;

the

lower

the t ens i le

s t rength

ra t io

and the

more

moisture

damage in the f ield A plot of

swell

a f te r conditioning and

t ens i le s t rength ra t io

i s shown in

Fig.

10 for

samples

with

30-minute

satura t ion.

The sample

can

be damaged by swell i f

allowed

to saturate for

30 minutes, but those

materials also

fa i l in

the

f ie ld based on

data in th i s study and data

reported

by

Jimenez 16) and

Kennedy 18).

Data

from Coplantz 17), Stuart 19) and Dukatz

20) indicated

tha t

excessive vacuum satura t ion

alone

did not

appear to in i t i a te

a s t r ipping

mechanism.

29


39/70

1.3

1.2

1 1

0

1

«

a:

I

0.9

CJ

0.8

z

w

a:

0.7

j)

w

0.6

J

j)

Z

0;5

0.4

0.3

o

O . 2 - - - - - - - - - - ~ - - ~ ~ - - - 4 - - ~ - - - ~ - ~ ~ ~

1

o

1 2

4 5

6

SWELL ( ) AFTER CONDITIONING

FIG.10 Relationship

of

Swell After Conditiorung Versus Tensile Strength Ratio from Samples Tested According

to AASHTO T 283 with 30-Minutes of Saturation.

3


40/70

Table

10

ummary

o

Swell After Conditioning

Swell

( After Condit ioning,

AASHTO

T

283

Si t e

Freeze

No 30-Minute No

STA

Ext ra

Freeze

Satu ra t ion

STA

1

+1.1 +1.3 +1.4

+1.7

+1.3

2

+0.2 +0.2

+0.4 +0.5

+0.2

3

+0.2

+0.3 +0.2

+0.6

0.0

4

+0.6 +0.6 +0.6 +0.7 +0.7

5

+0.5 +0.6 +0.8 +0.5 +0.7

6 +0.8 +0.6

+1.1

+1.0 +0.8

7 -0 .3

+0.4

+0.3 +0.4 +0.3

8 +1.3 +1.8 +1.8 +1.1 +1.2

9

+0.9 +1.0 +1.0

10

+0.3

+0.4 +1.0 +0.6

+0.2

11

0.0

+0.4

-0 .1

+0.3

+0.3

12

+0.5

+0.6 +0.1 +0.6

+0.3

13

+0.9

+1.7

+1.9 +1.4

+1.4

14

+3.4

+3.9

+4.6

+4.0

+3.8

15 +1.5

+1.9

+2.4

+1.6 +2.0

16 +0.8

+0.5

+1.1 +0.5 +1.3

17

+3.3

+3.9

+4.8

+3.6

+3.4

18

+1.3 +1.0 +1.3

+1.5 +1.8

19

+8.2 +8.1

+3.3 +9.4 +10.4

20

+3.8

+4.2 +5.9

+4.3 +4.3

STA = Short-Term Aging

Analysis o S h o r t T e ~ Aging The t ens i l e

s t r eng ths o f

uncondit ioned

samples

dry

t ens i l e

s t r eng ths

and t e n s i l

e

s t r eng th r a t io s us ing

th ree di f fe ren t

l eve l s of shor t - term

aging are shown in Table

11.

These

va lues

were

normal ized

to

the

values ob ta ined

from

the shor t - te rm

aging

spec i f i ed in

AASHTO T

283

and are

plo t ted

in

Figs . 11 and 12.

31


41/70

Figures

11

and 12 a r

e

box p lo t s . The l i ne i n the cen te r ~

the

box i s

the average .

The box encloses

p lus

and minus

one

s t anda rd dev ia t ion o f da ta .

The

whiskers

t ha t

extend

out

of

each

end of t he box are

the

range of

the da ta .

Table

11. ummary o f Dry Tens i le Strengths and ensi le Strength

Ratios

Using Various Lengths

o f

Short-Term Aging

STA).

Dry

St reng th kPa

Tens i le

Strength Ratio

Si t e

No

STA

Standard Extra

No STA

Standard Extra

ST

STA

ST

STA

440

500 620 1.00

1.02

0.87

2

460

490

7

00

1.13

1 .

20

1 .16

3

540

610 660

1.02 1.11 1.22

4

500

530 560

1.19 1.06

1.04

5

460

590

670

1.04 1.10

1.03

6

520

570

700

0.85

0.83

0.83

7

540

640

700

1.08 0.97

1.01

8

670 690 710

1.00

0.94 1.02

570 660 760 0.94

0.95 0.87

10

470

550

750

0.88

0.84

0.93

11

630 720

830

1.27

1.11

0.98

12

490

480

600

1.02 1. 01

0.90

13

650 660

880

0.72 0.69 0.74

14

680

700

900

0.35

0.32 0.29

15

700

680

800 0.55

0.53

0.53

16

610 680

770

0.90 0.82

0.61

17 400

460

490

0.74

0.65 0.65

18 420 460

540 0.93 0.89

0.84

19

500

550

590

0.15 0.22 0.16

20 550 550 700

0.51 0.59

0.49

Avg.

540

589

697 0.86 0.84

0.81

S.D.

91

86

110 0.29 0.27

0.28

32


42/70

0

....

"001

«

§ E

o

Zl l l

I-

.£:1

....

...

01'"

mE

..

/)

....

(/)"0

QI a

:="0

f/) C

C «

QI'"'

I-(/)

)o,QI

.... :

0 '

9

"0

N .-

: :

01

«1


43/70

The

HM samples

with longer aging have higher dry

t ens i le

strengths, Fig

11.

I f a dry

t ens i le

strength

i s

specif ied,

the

length

of

short-term

aging

must

also be specif ied. The dry

tens i le

strength

did

not

discriminate amoung any s i tes . The

wet t ens i le

strength

was

only

able

to

discriminate

between

good

and dis int igra tor s i t e s

a t

about 45 kPa

65 ps i .

In

most

cases,

the t ens i le

strength

ra t io remained constant

with increases in

aging,

Fig. 12.

However,

in one case

Si te

16), the t ens i le strength ra t io dropped because the dry

s t rength increased dramatical ly, and

the

wet strength did not

change. The t ens i le s trength ra t io i s

generally

insensi t ive

to the

length of

aging.

Kennedy

27) also determined

the

effect of short- term aging

on

the t ens i le

strength

ra t io was not s ignif icant .

By

eliminating

short-term

aging,

the time required

for

t es t ing

could be

shortened s ignif icantly .

Specifying a t ens i le strength ra t io appears to be superior

to

an absolute requirement

on

a t ens i le strength of a conditioned

sample.

The

inf luence of short- term

aging i s negated

when a

ra t io

i s used. In

the

f ie ld , condit ioning i s a function of

plant type, s torage time

in

the s i lo ,

haul time,

etc . with

a l l

of

the

f ie ld variables,

t

i s di f f icul t

to

quantify

the

amount of

short-term

aging

an

HM mixture

receives.

Analys is o

Lime Addit ion

Using SHTO

T

283, a l l

HM

samples

except one of

the

disintegrators , Si te 19) te s ted in th i s

study

had

acceptable

tens i le

s trength

ra t ios

when

lime

was

used. I t i s not

clear i f

the

addition

of

lime

would have

provided

good f ie ld performance

because the

sever i ty

Level 1

t e s t was

not

used.

4


44/70

Analysis o f

the

oi l ing Water

Test

The

resul t s

of the

boi l ing

water t e s t

are shown in Table 12.

Five

people

rated th

e

samples

and

the

t e s t resu l ts

are

very subject ive . A

large

sca t ter

in the

resul ts

was obtained

between each evaluator.

The boi l ing

water t e s t i s not an

idea l

t e s t because the resul t s

are subject ive. Addit ionally

the resul ts do

not consider

the

void s tructure permeabil i ty

or

gradation of the HM mixture.

In some

mixtures the

t r a f f i c loads are carr ied by

the

f ine

str ipping-susceptible

aggregates

and f ie ld performance

s

poor.

In

other mixtures with the same str ipping-susceptible

aggregates

performance

may

be

good

i f the

gradation

allows

the

t r a f f i c loads

to

be

carr ied

by the coarse nonstr ipping

susceptible aggregates.

The permeabil i ty of the sample

determined

by the void st ructure

should also be a factor in determining

the

suscep t ib i l i ty to

moisture

damage.

The boil ing water

t es t does

not

consider the

void s t ructure

since

the t e s t i s performed

on

a loose

mixture.

Results are summarized in Table 13 using a cutoff of 95 . The

boil ing t e s t i s a very severe

t e s t .

Most a l l of the samples

fa i led

regardless

of

known f ie ld performance.

35


45/70

Table

12. Boiling Water Test Results .

Si te

Category Coated Aggregate

1

Good 35

2

65

3

40

4

95

5

45

6

80

7

90

8 High Maintenance

55

9

95

10

80

11

65

12

95

13

Complete Rehab.

90

14

80

15

55

16

95

17

Disin t igra tors

50

18

50

19

40

20

75

Table

13. Comparison of Pavements of Known Field Performance

with the Boiling Water

Test

ASTM

D 3625).

Good High Complete Disint .

Maint. Rehab.

Pass

1

2

1 0

ail 6

3 3

VII. MODIFIED LOTTMAN REPEATABILITY

The

Colorado DOT

performed

an

invest igat ion to

determine

the

amount

of var iabi l i ty in the

ind i rec t

tens i le s t r ipping

t e s t

within the CDOT Central Materials

Laboratory

29).

A

s ingle operator

standard

deviat ion

in the

t ens i le s trength

36


46/70

ra t io

was

0.04. Maupin 30) performed

a

study in

Virginia

on

the var i ab i l i t y of the indirect t ens i le s trength r a t io One

standard deviation was 0.035. The indirect t ens i le

strength

ra t io i s

very repeatable within one

laboratory.

VIII . C SB H I S ~ R I S OF OTHER IRVESTIGATORS

The Lottman moisture suscep t ib i l i ty t e s t

can

be

performed

a t

various

levels

of severi ty.

The ColoradQ s i tes

with high

t ra f f ic in th i s study

are best predicted

with severi ty Leve l l

I t

was of i n t e res t to determine the

level

of severi ty tha t

predicted actual

performance

of o ther

pavements

reported

in

the

l i t e ra tu re Four case his tor ies in the l i t e ra tu re were

analyzed

to determine the level of

severi ty

of the Lottman t e s t

tha t predicted actual

pavement

performance.

Stuart

1986.

Stuart 19)

tes ted

materials from 14 s i tes with

good,

s l ight and

severe

f ie ld

performance with

respect to

moisture damage. The s i t e s were from

Georgia,

Maryland,

Mississippi

and

Utah and were

tes ted

with the

original

Lottman

t e s t Level

2A

and

the

Lottman

t es t

modified by Tunnicl i f f

Level 2C). Both t e s t s worked acceptably as shown in Tables 14

and

15. A minimum t ens i le strength ra t io of 0.80 was used.

Table 14. Comparison o the Lottman Test Level 2A

to Actual

Performance.

Actual Pavement Performance

Good

Slight Severe

Lottman

Pass

6

2

0

Test

Level 2A

Fai l

0 2 4

37


47/70

Table 15. Comparison

of

the Lottman

Test Level

2C)

to

Actual

Performance.

Actual Pavement Performance

Good Sl igh t Severe

Lottman

Pass

5 1

0

Tes t

Level 2C) Fa i l

1

3

Kennedy

1983. Kennedy

18)

analyzed e igh t s i t e s in Texas

with good

and

bad performance from moisture damage. Although

the

Lubbock

s i t e was or ig ina l ly c l a s s i f i ed as

good,

t was

changed

to

bad

when

a

low

area revealed s igns of

moisture

damage. The most severe vers ion L e v e l l ) of the Lottman t e s t

had good cor re la t ion as shown in Table 16.

A

minimum t e n s i l e

s t reng th

r a t i o o f 0.80 was used.

Table 16. Comparison of the

Lottman Test L e v e l l )

to Actual Performance.

Actual

Pavement Performance

Good Bad

Lottman

Pass 2

0

Tes t

L e v e l l )

Fai l

1

5

Kennedy 1983. One s i t e in Texas with poor performance was

analyzed

by Kennedy 26) . The

most severe ve rs ion L e v e l l )

o f

the Lottman t e s t showed good cor re l a t ion .

38


48/70

Parker

1988.

Parker

31) tes ted

five

aggregates

from Alabama

with good,

moderate, and

poor

performance his tor ies using the

Lottman t e s t

as

modified

for Levels

B and 2C. Correlat ion was

poor as shown in Tables 17 and 18. A minimum t ens i le st rength

ra t io

of 0.80 was used.

Table 17. Comparison

o the

Lottman Test Level 28)

to Actual Performance.

Actual

Pavement

Performance

Good Moderate

Poor

Lottman

Pass

0 1

0

Test

Level

2B Fai l 2 0 2

Table

18. Comparison

o the Lottman Test Level 2C)

to Actual

Performance.

Actual Pavement

Performance

Good

Moderate

Poor

Lottman

Pass 1 1 1

Test

Level

2C

Fai l

1

0

1

Summary. The modified Lottman and numerous versions

do

appear

to

have a

reasonably

good

corre la t ion with mixtures

of known

f ie ld performance.

However, the re la t ionship i s not ideal

When using a t e s t tha t does

not

ideal ly re la te

to

actual f ie ld

performance, t

i s

reasonable

tha t a large

factor

of

safe ty

be

applied

in

establ ishing the severi ty level of

the

t e s t

procedure

and

the specif ica t ion value.

39


49/70

IX. ON LUSIONS

The

aggregates

and

asphal t

cements used

for

th i s

study were

from the

same sources

but

were not the exact material

tha t was

used on each

project .

The proper

performance of

hot

mix

asphalt HMA) pavements i s

not

sole ly dependent on the material propert ies; improper

construction

or s t ruc tura l design

could also cause problems

with

the

HM

pavement.

For

the

fa i lures

studied

in

th i s

investigation, t i s not clear

how much

of the fa i lure could be

at tr ibuted

to

materia ls , construction, or s t ru

investigation of the modified lottman test to predict the stripping performance of pavements

Documents