ANALYSIS OF OKLAHOMA MIX DESIGNS FOR THE NCAT TEST TRACK USING THE BAILEY METHOD TRB 85 TH ANNUAL MEETING WORKSHOP PRACTICAL APPROACHES TO DESIGN OF HOTMIX ASPHALT DANNY GIERHART, P.E. ODOT BITUMINOUS ENGINEER
Mar 27, 2015
ANALYSIS OF OKLAHOMA MIX DESIGNS FOR THE
NCAT TEST TRACK USING THE BAILEY METHOD
TRB 85TH ANNUAL MEETING WORKSHOP
PRACTICAL APPROACHES TO DESIGN OF HOTMIX ASPHALT
DANNY GIERHART, P.E. ODOT BITUMINOUS
ENGINEER
PRESENTATION TOPICS
• BAILEY METHOD OVERVIEW
• CASE STUDY – ODOT’s NCATTEST TRACK MIX DESIGNS
• ODOT OPINION OF THEMETHOD AT THIS POINT
THE BAILEY METHOD
ACHIEVING VOLUMETRICS AND
HMA COMPACTABILITY
TRANSPORTATION RESEARCH CIRCULAR
Number E-C044
Bailey Method for Gradation Selection in Hot-Mix Asphalt Mixture Design
Vavrick, Huber, Pine, Carpenter, Bailey
October 2002
How Can the Bailey Method Help?
• In Developing New Blends:– Field Compactibility– Segregation Susceptibility
• In Evaluating Existing Blends:– What’s worked and what hasn’t?– More clearly define ranges referenced in the method
• In Estimating VMA/Void changes between:– Design trials– QC samples– Potentially Saves Time and Reduces Risk!
Originally developed in the 1980’s by Robert D. Bailey, a civil engineer now retired from Illinois DOT
The method focuses on how aggregate particles fit together
Aggregate PackingWhat Influences the Results?
• GRADATION- continuously-graded, gap-graded, etc.
• SHAPE- flat & elongated, cubical, round
- smooth, rough• SURFACE TEXTURE
• STRENGTH- resistance to breaking, abrasion, etc.
• TYPE & AMOUNT OF COMPACTIVE EFFORT- static pressure, impact, or shearing
REQUIRED LABORATORY TESTING
Sieve Size (mm) Raised to 0.45 PowerSieve Size (mm) Raised to 0.45 Power
100100
00
% P
assi
ng%
Pas
sing
Illustration of the Four Principles Illustration of the Four Principles – Predominantly Coarse – Predominantly Coarse
Aggregate MixAggregate Mix
5050
2020
8080
1010
3030
4040
6060
7070
9090
CoarseFine
1
2
34
PRINCIPLE # 1 – CATEGORIZE MIX AS PREDOMINANTLY “COARSE” OR “FINE”
Coarse particles create voids
Fine particles fill voids
Designation of coarse and fine particles is based on the Nominal Maximum Particle Size (NMPS).
Diameter = NMPS
Average Void Size = 0.22 x NMPS
Primary Control Sieve ≈ 0.22 x NMPS
Primary Control Sieve
Mixture NMPS
NMPS x 0.22
Primary Control Sieve
37.5mm 8.250mm 9.5mm
25.0mm 5.500mm 4.75mm
19.0mm 4.180mm 4.75mm
12.5mm 2.750mm 2.36mm
9.5mm 2.090mm 2.36mm
4.75mm 1.045mm 1.18mm
PCS determines the break between Coarse and Fine in the combined blend and if a given aggregate is a CA or FA
Chosen Unit Weight - CA(s)
LUW RUW
Coarse-Graded SMAFine-Graded
< LUW
< 90% 95-105% 110-125%
Coarse-Graded Mix
• Some particle-to-particle contact of CA
• Coarse and Fine fractions carry load
• Reduced FA strength acceptable
Fine-Graded Mix• Little to No particle-
to-particle contact of CA
• Fine fraction carries most of the load
• Increased amount of FA support needed
Sieve Size (mm) Raised to 0.45 PowerSieve Size (mm) Raised to 0.45 Power
100100
00
% P
assi
ng%
Pas
sing
Combined Blend Gradation – Combined Blend Gradation – Predominantly Fine Aggregate MixPredominantly Fine Aggregate Mix
5050
2020
8080
1010
3030
4040
6060
7070
9090
2
3
4
CoarseFine
PCS
New PCS = 0.22 x PCS
1
PCSNew NMPS
PRINCIPLE # 2 – ANALYSIS OF THE COARSE FRACTION OF THE BLEND
Smaller particles in the coarse fraction are still too large to fit into the voids created by the larger particles
The coarse fraction is the portion retained above the Primary Control Sieve (PCS)
PCS
NMPS
“Half” Sieve
• “Half” sieve = “half” of NMPS
• CA Ratio =
Where:% Half sieve = % passing the Half sieve
% PCS = % passing the PCS
• Adjusting CA Ratio– Alter volume blend of CA’s– Change CA source/gradation
% Half sieve - % PCS
100% - % Half sieve
PRINCIPLE # 2 is evaluated using the Coarse Aggregate Ratio
“pluggers”
“interceptors”
“interceptor” particles increase voids because they are large enough to prevent “plugger” particles both from packing together and from packing the fine fraction
CA Ratio EffectsFINE IN CONTROL COARSE IN CONTROL
Portion evaluated as new coarse fraction is smaller – less sensitive to changes
Portion evaluated as coarse fraction is larger – more sensitive to changes
Low New CA Ratio – Lower VMA & air voids
Low CA Ratio – Lower VMA & air voids
Coarse particles “floating” in fine particles – New CA Ratio does not relate to segregation, Old still does
Low CA Ratio – too many “pluggers”, mix prone to segregation
High New CA Ratio – too many “interceptors,” mix can be difficult to compact
High CA Ratio – too many “interceptors,” mix can be difficult to compact
CA Ratio Guidelines
COARSE IN CONTROL
NMPS 25.0mm 19.0mm 12.5mm 9.5mm 4.75mm
CA Ratio
0.70 - 0.85
0.60 – 0.75
0.50 – 0.65
0.40 – 0.55
0.30 – 0.45
FINE IN CONTROL
NMPS All Sizes
New CA Ratio
0.60 - 1.00
PRINCIPLE # 3 – ANALYSIS OF THE FINE FRACTION OF THE BLEND (COARSE PORTION)
The coarser fine particles also create voids which finer particles fill
The fine fraction is the portion passing the Primary Control Sieve (PCS)
PRINCIPLE #3 is evaluated using the FAc ratio
• Secondary Control Sieve (SCS)– View fine fraction as
a “blend”– New coarse and
fine break– SCS = 0.22 x PCS
• PCS generally serves as the maximum and NMPS of overall fine fraction
• FAc Ratio =
PCS
Fine Fraction
% SCS
% PCS
SCS
FAc Ratio Effects
COARSE IN CONTROLFINE IN CONTROL
0.05 increase in New FAc Ratio up to 0.50 results in an approximate 1% decrease in VMA and Air Voids
Once New FAc Ratio increases beyond 0.50 VMA begins to increase
As New FAc Ratio increases toward 0.50, compactability of fine fraction increases
0.05 increase in FAc Ratio up to 0.55 results in an approximate 1% decrease in VMA and Air Voids
Once FAc Ratio increases beyond 0.55 VMA begins to increase
As FAc Ratio increases toward 0.50, compactability of fine fraction increases
PRINCIPLE # 4 – ANALYSIS OF THE FINE FRACTION OF THE BLEND (FINE PORTION)
Again, the larger fine particles of this portion also create voids which the finest particles fill
Now looking at the finer portion of the fine fraction passing the Secondary Control Sieve (SCS)
PRINCIPLE #4 is evaluated using the FAf ratio
• Tertiary Control Sieve (TCS)– View fine part of fine
fraction as a “blend”– New coarse and
fine break– TCS = 0.22 x SCS
• SCS generally serves as the maximum and NMPS of fine part of fine fraction
• FAf Ratio =
PCSFine Fraction % TCS
% SCS
SCSTCS
FAf Ratio Effects
COARSE IN CONTROLFINE IN CONTROL
Once New FAf Ratio increases beyond 0.50 VMA begins to increase
As New FAf Ratio increases toward 0.50, VMA begins to decrease
Once FAf Ratio increases beyond 0.55 VMA begins to increase
As FAf Ratio increases toward 0.55, VMA begins to decrease
FAc & FAf Ratio Guidelines
COARSE IN CONTROL
NMPS All Sizes
FAc & FAf Ratio
0.35 – 0.50
FINE IN CONTROL
NMPS All Sizes
New FAc & FAf Ratio
0.35 – 0.50
Combined Blend EvaluationCoarse-Graded Mixes
1. CA CUW increase = VMA increase– 4% change in PCS 1% change in VMA or
Voids
2. CA Ratio increase = VMA increase– 0.20 change 1% change in VMA or Voids
3. FAc Ratio increase = VMA decrease– 0.05 change 1% change in VMA or Voids
4. FAf Ratio increase = VMA decrease– 0.05 change 1% change in VMA or Voids
Has the most
influence on VMA or Voids
Estimating VMA or VoidsCoarse-Graded Mix Example
• Trial #1– PCS = 38.2%
• 100% CA LUW
– CA ratio = 0.693
– FAc ratio = 0.492
– FAf ratio = 0.394
– AC = 4.6%– Air Voids = 3.4%– VMA = 12.6%
• Trial #2– PCS = 37.2%
• 102.5% CA LUW
– CA ratio = 0.725
– FAc ratio = 0.444
– FAf ratio = 0.412
– AC = 4.6%– Expected VMA?– Expected Air Voids?
Estimating VMA or VoidsTrial #2 vs. Trial #1
• PCS37.2% - 38.2% = - 1.0%
• CA ratio0.725 – 0.693 = + 0.032
• FAc ratio
0.444 – 0.492 = - 0.048
• FAf ratio
0.412 – 0.394 = + 0.018
• Increases VMA or Voids– 1.0/4.0 = + 0.25%
• Increases VMA or Voids– 0.032/0.2 = + .16%
• Increases VMA or Voids– 0.048/0.05 = +.96%
• Decreases VMA or Voids– 0.018/0.05 = - 0.36%
• Total Estimated Change:– Plus ~ 1.0% VMA
ODOT’S PERPETUAL PAVEMENT STRUCTURAL SECTIONS AT NCAT TEST TRACK
PLAN VIEW
SECTION 1 – 150’ SECTION 2 – 150’
25’ TRANSITION 50’ TRANSITION 25’ TRANSITION
PLAN VIEW
SECTION 1 – 150’ SECTION 2 – 150’
25’ TRANSITION 50’ TRANSITION 25’ TRANSITION
PROFILE VIEW
2” SMA w/PG 76-28
3” SuperPave 19.0mm w/PG 76-28
3” SuperPave 19.0mm w/PG 64-22
3” SuperPave 19.0mm w/PG 64-22 2” RBL w/PG 64-22
3” RBL w/PG 64-22
*RBL = RICH BOTTOM LAYER
ODOT’S PERPETUAL PAVEMENT STRUCTURAL SECTIONS AT NCAT TEST TRACK
AGGREGATE SUMMARY
16.3
7.4
6.8
Hanson
26.3
23.8
1.1
Martin Marietta
2.0
GMI Sand
Aggregate Type
Aggregate Shape
L.A. Abrasion
Micro Deval
Screenings P200
Rhyolite LimestoneRiver Sand
Very Angular
Angular
25.2
14.7
12.9
Dolese
Limestone
Angular Rounded
n/a
n/a
0
20
40
60
80
100
RBL MIX DESIGN INFORMATION
Hanson Hanson Dolese
5/8” Chips
Screenings
Screenings
35% 20% 45%
Pb 6.0
% Air Voids
2.0
% VMA 14.6
RBL MIX – EVALUATED AS A FINE-GRADED MIX
OLD CA RATIO – 0.875
NEW CA RATIO – 0.556
NEW FAc RATIO – 0.558
NEW FAf RATIO – N/A
For fine-graded mixes, the volume of the fine fraction exceeds the CA LUW voids. This value is less than 90% of CA LUW, and ensures that the fine aggregate is in control.
CHOSEN UNIT WT. = 78.9%
RBL MIX – EVALUATED AS A FINE-GRADED MIX
OLD CA RATIO – 0.875
NEW CA RATIO – 0.556
NEW FAc RATIO – 0.558
NEW FAf RATIO – N/A
For coarse-graded mixes, the preferred range is 0.50 – 0.65. For this fine-graded mix, the high CA Ratio indicates a low susceptibility to segregation.
CHOSEN UNIT WT. = 78.9%
RBL MIX – EVALUATED AS A FINE-GRADED MIX
OLD CA RATIO – 0.875
NEW CA RATIO – 0.556
NEW FAc RATIO – 0.558
NEW FAf RATIO – N/A
The preferred range is 0.60 – 1.00. The New CA Ratio is primarily controlled by the FAs rather than the CAs and its affect on the entire blend is therefore mitigated.
CHOSEN UNIT WT. = 78.9%
RBL MIX – EVALUATED AS A FINE-GRADED MIX
OLD CA RATIO – 0.875
NEW CA RATIO – 0.556
NEW FAc RATIO – 0.558
NEW FAf RATIO – N/A
The preferred range is 0.35 – 0.50. The value of 0.558 indicates a high dust/binder ratio (1.4 for this design) and a high mortar stiffness. Higher values →lower VMA.
CHOSEN UNIT WT. = 78.9%
RBL MIX – EVALUATED AS A FINE-GRADED MIX
OLD CA RATIO – 0.875
NEW CA RATIO – 0.556
NEW FAc RATIO – 0.558
NEW FAf RATIO – N/A
The tertiary sieve for 12.5mm fine-graded mixes would fall below the 0.075mm, therefore the FAf Ratio cannot be calculated.
CHOSEN UNIT WT. = 78.9%
2.0
1.0
3.0
0.0
6.0
4.0
7.0
5.0
9.0
8.0
-1.0 32 54 6 7
AIR VOIDS @ 6.0% BINDER – ACTUAL vs. ESTIMATED
RBL MIXACTUAL
EST.
% A
IR V
OID
S
TRIAL #
0
20
40
60
80
100
19.0mm SUPERPAVE MIX DESIGN INFORMATION
Hanson Hanson Dolese MM GMI1” Chips Screenings Screenings Stone Sand Sand
30% 25% 15% 20% 10%
Pb 4.3
% Air Voids
4.0
% VMA 13.6
19.0mm MIX – EVALUATED AS FINE-GRADED
OLD CA RATIO – 0.358
NEW CA RATIO – 0.681
NEW FAc RATIO – 0.517
NEW FAf RATIO – 0.332
This value is far less than 90% of CA LUW, and ensures that the fine aggregate is in control. However, such a low value indicates that the mix may be difficult to compact.
CHOSEN UNIT WT. = 50.7%
19.0mm MIX – EVALUATED AS FINE-GRADED
OLD CA RATIO – 0.358
NEW CA RATIO – 0.681
NEW FAc RATIO – 0.517
NEW FAf RATIO – 0.332
Even though this is a fine-graded mix, the low CA Ratio means that in the CA there is a higher % of “pluggers” than “interceptors,” indicating a potential problem with segregation.
CHOSEN UNIT WT. = 50.7%
OLD CA RATIO – 0.358
NEW CA RATIO – 0.681
NEW FAc RATIO – 0.517
NEW FAf RATIO – 0.332
The preferred range is 0.60 – 1.00. This mix falls within the preferred range, which means any compaction issues would likely not be attributed to this fraction.
CHOSEN UNIT WT. = 50.7%
19.0mm MIX – EVALUATED AS FINE-GRADED
OLD CA RATIO – 0.358
NEW CA RATIO – 0.681
NEW FAc RATIO – 0.517
NEW FAf RATIO – 0.332
The preferred range is 0.35 – 0.50. The value of 0.517 might indicate a tenderness problem if the mix contained a high % sand. However, this mix contains only 10% natural sand.
CHOSEN UNIT WT. = 50.7%
19.0mm MIX – EVALUATED AS FINE-GRADED
OLD CA RATIO – 0.358
NEW CA RATIO – 0.681
NEW FAc RATIO – 0.517
NEW FAf RATIO – 0.332
The preferred range is 0.35 – 0.50. However, the FA ratios are generally a problem only if both are high or both are low.
CHOSEN UNIT WT. = 50.7%
19.0mm MIX – EVALUATED AS FINE-GRADED
2.0
1.0
4.0
3.0
5.0
2
AIR VOIDS @ 4.3% BINDER – ACTUAL vs. ESTIMATED
19.0mm SUPERPAVE MIX
% A
IR V
OID
S
TRIAL #
ACTUAL
EST.
0
20
40
60
80
100
SMA MIX DESIGN INFORMATION
Hanson Hanson Dolese Boral
5/8” Chips
Screenings
Screenings
Mineral Filler
67% 13% 10% 10%
Pb 6.8
% Air Voids
4.0
% VMA 17.9
SMA MIX
CA RATIO – 0.398
FAc RATIO – 0.720
FAf RATIO – 0.843
This value barely falls within the preferred range of 110 - 125% of CA RUW. This indicates that the %CA, although acceptable, is on the low side for a SMA mix.
CHOSEN UNIT WT. = 110.0%
SMA MIX
CA RATIO – 0.398
FAc RATIO – 0.720
FAf RATIO – 0.843
This value falls within the preferred range of 0.25 – 0.40. Be careful interpolating the value for the “half sieve” on a 12.5mm SMA. It would be best to insert a ¼” sieve into the nest.
CHOSEN UNIT WT. = 110.0%
SMA MIX
CA RATIO – 0.398
FAc RATIO – 0.720
FAf RATIO – 0.843
This value falls within the preferred range of 0.60 – 0.85, indicating a good balance in the relative fractions of the fine aggregate.
CHOSEN UNIT WT. = 110.0%
SMA MIX
CA RATIO – 0.398
FAc RATIO – 0.720
FAf RATIO – 0.843
This value falls within the preferred range of 0.65 – 0.90. Typically, the higher the ratio, the greater P200.This mix was designed on the high side to decrease permeability potential.
CHOSEN UNIT WT. = 110.0%
4.0
3.0
6.0
5.0
7.0
2
AIR VOIDS @ 7.0% BINDER – ACTUAL vs. ESTIMATED
12.5mm SMA MIX
% A
IR V
OID
S
TRIAL #
ACTUAL
EST.
2.0
1.0
3.0
0.0
6.0
4.0
7.0
5.0
9.0
8.0
-1.0 2 3 4
EXAMPLE OF ACTUAL ODOT QC/QA PROJECT DATA AIR VOIDS – ACTUAL vs. ESTIMATED
19.0mm SuperPave Mix
ACTUAL
EST.
% A
IR V
OID
S
SAMPLE #
OUR THOUGHTS SO FAR:
• The Bailey Method principles make sense when reviewed in the context of previous mix design experience
• The Method provides a way to quantify changes that we have only made “educated guesses” at before
• Based on previous experience, the Method gives a reasonable indication of aggregate combinations which are susceptible to segregation and field compactability problems
OUR THOUGHTS SO FAR:
• Based on previous experience, the mixes that fall into the “Coarse-Graded” category are often too permeable
• The voids estimation process looks at gradation only, and is therefore “blind” to changes in aggregate shape and texture
• The voids estimation process performs better when working with aggregates of similar properties
OUR THOUGHTS SO FAR:
• Although the Bailey Method is a good tool, users must not forget the things they already know about the materials they are using
• The “default” values used in the void estimation process should vary depending on the types of aggregate used
• Each user should analyze historical data and interview field personnel to “calibrate” the method to their own materials
SOME TOOLS REQUIRE MORE PRACTICE AND EXPERIENCE THAN OTHERS…
THANK YOU!