Standard Method of Test for

Determining the Damage

Characteristic Curve of Asphalt
Mixtures from Direct Tension
This document is only for acquaintance

Cyclic Fatigue Tests

AASHTO Designation: TP 107-14

American Association of State Highway and Transportation Officials

444 North Capitol Street N.W., Suite 249
Washington, D.C. 20001

Больше стандартов - [Link]

 Standard Method of Test for

Determining the Damage Characteristic

Curve of Asphalt Mixtures from Direct
Tension Cyclic Fatigue Tests

AASHTO Designation: TP 107-141

This document is only for acquaintance

1. SCOPE
1.1. This test method covers procedures for preparing and testing asphalt concrete mixtures to
determine the damage characteristic curve via direct tension cyclic fatigue tests.

1.2. This standard is applicable to laboratory prepared specimens of mixtures with nominal maximum
size aggregate less than or equal to 37.5 mm (1.48 in.).

1.3. This standard may involve hazardous material, operations, and equipment. This standard does not
purport to address all safety problems associated with its use. It is the responsibility of the user of
this procedure to establish appropriate safety and health practices and to determine the
applicability of regulatory limitations prior to use.

2. REFERENCED DOCUMENTS
2.1. AASHTO Standards:
 PP 60, Preparation of Cylindrical Performance Test Specimens Using the Superpave Gyratory
Compactor (SGC)
 PP 61, Developing Dynamic Modulus Master Curves for Asphalt Mixtures Using the Asphalt
Mixture Performance Tester (AMPT)
 R 30, Mixture Conditioning of Hot Mix Asphalt (HMA)
 R 62, Developing Dynamic Modulus Master Curves for Asphalt Mixtures
 T 342, Determining Dynamic Modulus of Hot Mix Asphalt (HMA)
 TP 79, Determining the Dynamic Modulus and Flow Number for Asphalt Mixtures Using the
Asphalt Mixture Performance Tester (AMPT)

2.2. Other Document:

 NCHRP Report 629, Equipment Specifications for the Simple Performance Test System,
Appendix E, October 16, 2007.

3. TERMINOLOGY
3.1. Definitions:

3.1.1. complex modulus (E*)—a complex number that defines the relationship between stress and strain
for a linear viscoelastic material.

TS-2d TP 107-1 AASHTO

Больше стандартов - [Link]

 3.1.2. cyclic pseudo secant modulus (C*)—the secant modulus in stress–pseudo strain space for a single
cycle. This pseudo modulus differs from C because it is computed using a steady-state assumption
and is used only with cycle-based data.

3.1.3. damage (S)—the internal state variable that quantifies microstructural changes in asphalt concrete.

3.1.4. damage characteristic curve (C versus S curve)—the curve formed when plotting the damage on
the x-axis and the pseudo secant modulus on the y-axis. It defines the unique relationship between
the structural integrity and amount of damage in a given mixture.

3.1.5. dynamic modulus (|E*|)—the norm of the E*, which is calculated by dividing the peak-to-peak
This document is only for acquaintance

stress by the peak-to-peak axial strain measured during the steady-state period.

3.1.6. failure cycle (Nf)—the cycle in which the measured phase angle drops sharply after a stable
increase during cyclic loading.

3.1.7. phase angle (φ)—the angle, expressed in degrees, between an applied sinusoidal stress and the
resulting sinusoidal strain measured during the steady-state period.

3.1.8. pseudo strain (εR)—a quantity that is similar to strain but does not include time effects. Pseudo
strain is calculated by solving the convolution integral of the strain and E(t).

3.1.9. pseudo secant modulus (C)—the secant modulus in stress–pseudo strain space.

3.1.10. relaxation modulus (E(t))—the quotient of the stress response of a material with time to a constant
step amplitude of strain.

4. SUMMARY OF METHOD
4.1. A controlled and repeated cyclic loading is applied to a cylindrical asphalt concrete specimen until
failure. The applied stress and on-specimen axial strain response are measured and used to
calculate the necessary quantities. The relationship between the damage (S) and the pseudo secant
modulus (C) is determined and expressed as the damage characteristic curve.

5. SIGNIFICANCE AND USE

5.1. The damage characteristic curve represents the fundamental relationship between damage and
material integrity for asphalt concrete mixtures. This property is independent of temperature,
frequency, and mode of loading. Combined with the linear viscoelastic properties of asphalt
concrete, the damage characteristic curve can be used to analyze the fatigue characteristics of
asphalt concrete mixtures.

5.2. Damage characteristic curves can also be combined with additional pavement response models to
predict the fatigue behavior of in-service asphalt concrete mixtures.

TS-2d TP 107-2 AASHTO

Больше стандартов - [Link]

This document is only for acquaintance

Figure 1—General Schematic of Direct Tension Test Setup

6. APPARATUS
6.1. Direct Tension Test System—Consists of a testing machine, environmental chamber, and
measurement system. Figure 1 presents a schematic of the test setup for the fatigue test.

6.2. Asphalt Mixture Performance Tester—An Asphalt Mixture Performance Tester or system meeting
or exceeding the requirements of Equipment Specifications for the Simple Performance Test
System, NCHRP Report 629, Appendix E.

6.3. External Conditioning Chamber (optional)—An environmental chamber for conditioning the test
specimens to the desired testing temperature. The chamber shall be capable of controlling the
temperature of the specimen over a temperature range of 5 to 25°C (41 To 77°F) to within ±0.5°C
(±1°F). The chamber shall be large enough to accommodate at least a single test specimen and a
monitoring specimen with thermocouples mounted at the center and on the surface for temperature
verification.

6.4. Measurement System—Shall be fully computer-controlled and capable of measuring and recording
the time history of the applied load and axial deformations. A system that uses the characteristics
shown in Table 1 has been found to be sufficient:

Table 1—Accuracy and Resolution of Equipment Measurement System

Measurement Range Accuracy Resolution
Load 0.12 to 13.5 kN Error ≤1.0% ≤0.0012 kN
Machine deformation At least 12 mm Error ≤0.03 mm ≤0.0025 mm
On-specimen deformation Not specifieda Error ≤1.0% Not specifiedb
a
A range of about 7000 με is recommended.
b
A resolution less than 7.5 με is recommended.

TS-2d TP 107-3 AASHTO

Больше стандартов - [Link]

 6.5. Axial Deformation Measurements—Axial deformations shall be measured using sensors mounted
between gauge points that are glued to the specimen, an example of which is shown in Figure 2.
The deformations shall be measured at two locations 180 degrees apart, three locations 120
degrees apart, or four locations 90 degrees apart.
Note 1—It is recommended, but not specified, that these sensors be the loose-core LVDTs (linear
variable differential transformers) type. If the gauge-head type (spring-loaded LVDTs) is chosen,
care should be taken to ensure that the spring force is not so strong as to force the gauge points
apart. This effect can be measured by monitoring the deformation of a sample that is lying on its
side and is conditioned to the appropriate temperature measured using a spring-loaded LVDT that
has been attached to the sample.
As long as the spring force is confirmed to be not strong enough to cause deformation on the
This document is only for acquaintance

sample, gauge-head type (spring-loaded) LVDTs can be advantageous and convenient. They can
also be used for performing the dynamic modulus test according to TP 79.
Note 2—Using four sensors set at 90 degrees apart has an advantage over using fewer sensors,
because if one sensor does not function properly, that sensor and the sensor 180 degrees opposed
to it can be discarded, and the remaining two sensors can be used to determine the average
deformation.
Note 3—The reliability of sensors should be verified with the supplier or manufacturer.

Figure 2—General Schematic of Gauge Points (not to scale)

6.6. Loading Platens—Are required above and below the specimen to transfer the load from the testing
machine to the specimen. The diameter of the loading platens should be within ±0.5 mm
(±0.02 in.) from the target specimen diameter of 100 mm. Generally, these platens should be made
of hardened or plated steel, or anodized high strength aluminum. Softer materials will require
more frequent replacement. Materials that have linear elastic modulus properties and hardness
properties lower than that of 6061-T6 aluminum shall not be used. The face of each load platen
shall be grooved to provide better adhesion between the glue and plate. The top loading platen
shall be designed so that it can be mated to the test machine without inducing loading eccentricity.
A locking ball joint that will lock when tightened into the machine as shown in Figure 3 is one
potential design. If the end plate gluing apparatus is sufficiently stiff, it may be possible to install
the test specimen without the need for such a locking ball joint.

TS-2d TP 107-4 AASHTO

Больше стандартов - [Link]

 Note 4—Grooves in the end plates can improve the overall adhesion between the glue and the
end plates. It is recommended that if grooved, the grooves be approximately 0.6 mm deep and
spaced laterally by a distance of approximately 1.9 mm. The grooving pattern can be either
concentric circles or a spiral pattern. A V-shaped groove or a cross-hatched groove is suggested.
The 60-degree point threading bit has been used successfully for making the grooves.

6.7. End Plate Gluing Apparatus—Should be available for gluing the end plates to the asphalt concrete
specimen. The device should self-center the specimen within the end plates and ensure that no
eccentricity exists between the specimen and end plates. It shall be possible to clamp or otherwise
hold the gluing apparatus at a fixed height for a period of at least 4 h. Figure 4 shows a gluing jig
that has been used successfully. If no jig is available, the specimen can be glued directly in the test
This document is only for acquaintance

machine. To do this, the gluing should be performed with the machine in load control. The applied
compressive load should not be greater than 0.045 kN (10 lb).

Figure 3—Example Locking Ball Joint Design

TS-2d TP 107-5 AASHTO

Больше стандартов - [Link]

This document is only for acquaintance

Figure 4—End Plate Gluing Apparatus

7. HAZARDS
7.1. Standard laboratory safety precautions must be observed when preparing and testing asphalt
concrete specimens.

8. TESTING EQUIPMENT CALIBRATION

8.1. The guidelines provided in TP 79 shall be followed to ensure that the test equipment and on-
specimen measurement devices are properly calibrated.

8.2. If any of the verifications yield data that do not comply with the accuracy specified, the problem
shall be corrected prior to further testing.

9. TEST SPECIMENS
9.1. Direct tension testing shall be performed on test specimens cored and cut from larger Superpave
gyratory compacted specimens. Test specimens with acceptable properties have been obtained
from gyratory samples 150 mm (6 in.) in diameter and a minimum of 180 mm (7.1 in.) in height.
The average diameter of test specimens shall be 100 to 104 mm (4 to 4.1 in.) in diameter with a
standard deviation of less than 0.5 mm (0.02 in.). The average height of the test specimens shall be
between 127.5 and 132.5 mm (5.0 and 5.2 in.), respectively. Test specimens should be prepared in
accordance with PP 60 with the exception of the allowance for 130-mm-height specimens.
Note 5—The specimen height that should be used and the cutting scheme will depend on the air
void gradient produced by the specific compactor and the capabilities of the sawing equipment.
For specimens compacted to a height of 180 mm, it has been found that 25 mm should first be cut
from the bottom face of the specimen. The bottom face of the specimen is the one that was at the
bottom of the gyratory compaction mold. Then, the remaining length should be cut so that the
final specimen height is between 127.5 and 132.5 mm.

TS-2d TP 107-6 AASHTO

Больше стандартов - [Link]

 9.2. Aging—Laboratory-prepared mixtures shall be temperature-conditioned in accordance with the
4-h, short-term oven-conditioning procedure outlined in R 30. Field mixtures need not be aged
prior to testing.

9.3. Prepare three test specimens at the target air void content ±0.5 percent.

9.4. Replicates—One replicate specimen should be tested at each input condition.

Note 6—Sample storage—If test specimens will not be tested within 2 days, it is recommended
to wrap each specimen in polyethylene or seal it in a polyethylene bag and store in an unlit and
environmentally protected storage area at temperatures between 5 and 25°C (40 and 77°F).
Specimens shall not be stacked during storage. To eliminate the effects of aging on the test results,
This document is only for acquaintance

it is recommended that specimens be stored no longer than 2 weeks prior to testing.

10. TEST SPECIMEN INSTRUMENTATION PROCEDURE

10.1. Thoroughly clean all end plates by first heavily brushing the face of each platen using either a
hand operated wire brush or one attached to a standard electric drill. After cleaning the platen with
the wire brush, wipe the surface clean of any dust by using a towel dipped in acetone or similar
solvent.

10.2. Using the same towel, but with only a small amount of solvent, wipe the ends of the specimen
clean of any residual dust.

10.3. Weigh out an appropriate amount of adhesive to adhere the end plates and specimen to one
another. The gluing process will require approximately 10–20 minutes, so prepare an adhesive that
is appropriate for this length of working time.
Note 7—Steel putty, such as Devcon 10110, has been found to be satisfactory for attaching the
end plates. Devcon 10240 has also been found satisfactory, but the overall working time is less
with the 10240 than it is with the 10110, so more careful attention will need to be paid to the
gluing process.
Note 8—Approximately 120 g of the steel putty has been found sufficient for 100-mm-diameter
samples.

10.4. Fill in any surface voids in the top and bottom surfaces of the specimen with the adhesive.

10.5. Divide the remaining adhesive into quarter measures and spread evenly between the end plates and
specimen end faces (that is, 1/4 to the top plate, 1/4 to the bottom plate, 1/4 to the bottom face, and
1
/4 to the top face). Insert and secure the end plates into the gluing jig and gently place the
specimen on top of the bottom end plate, as close as possible to the center. Tighten the jig to bring
the end plate and specimen into the center and lower the top plate into position. Do not allow the
top plate to rest completely on the specimen because the adhesive viscosity may be too low to
provide resistance. The final glue thickness should be approximately 1 mm (0.04 in.) thick. Wipe
or scrape away any excess material before the glue stiffens. Allow the adhesive to reach its initial
set before moving the specimen from the jig.
Note 9—The initial set of the Devcon 10110 requires approximately four hours. This setting time
can be tested by pressing against the adhesive with the thumbnail. If the thumbnail does not indent
the glue, then the initial set has been reached.

10.6. Remove the specimen from the gluing jig, taking special care to support the specimen from the
bottom. Avoid lifting the specimen by the end plates so as not to put any tension stress on the
adhesive.

TS-2d TP 107-7 AASHTO

Больше стандартов - [Link]

 10.7. Attach the mounting studs for the axial sensors to the sides of the specimen using epoxy cement.
The fixtures should meet or exceed those specified in Table 2.
Note 10—Quick-setting steel putty, such as Devcon 10240, has been found to be satisfactory for
attaching the studs.

Table 2—Minimum Requirements of Specimen Deformation Measurement System

Property Value
Gauge point contact area 80 mm2 ± 10 mm2 (0.12 in.2 ± 0.016 in.2)
Dimension of the gauge point in the direction of the gauge length 10 mm ± 2 mm (0.40 in. ± 0.08 in.)
Mass of mounting system and transducer 80 g max
This document is only for acquaintance

10.8. The gauge length for measuring axial deformations of 100-mm (4-in.) diameter samples shall be
70 mm ± 1 mm, measured center to center of the gauge points. Suitable alignment and a spacing
fixture may be used to facilitate the mounting of the axial deformation measuring hardware. The
gauge length shall be measured between the stud centers, that is, on center. All sensors shall be
mounted using the same gauge length.

10.9. Allow the adhesive to reach full cure before testing. Follow the manufacturer’s recommendation to
determine the time needed to full cure.

11. PROCEDURE
11.1. This test procedure is designed to first test the specimens, each of which is individually
instrumented with sensor brackets, at a specific temperature and frequency in oscillation mode to
obtain linear viscoelastic fingerprints and then at the desired frequency and temperature until
failure.

11.2. A total of three tests need to be done at three different strain levels.

11.3. An appropriate testing temperature can be selected based on the 98 percent reliability climatic
Performance Grade (PG) determined based on LTPPBind at the location of interest, but not
exceeding 21°C. The testing temperature in degrees Celsius should be determined as the average
of the high- and low-temperature climatic PG temperatures minus 3°C. If the calculated testing
temperature is greater than 21°C, use 21°C as the testing temperature. For example, the testing
temperature for the LTPPBind high temperature of 64°C and low temperature of –22°C is 18°C.

11.4. Optional: If using a separate chamber to temperature-condition the samples, continue to Section
11.5. If using the AMPT to temperature-condition the samples, skip to Section 11.7.

11.5. Insert all specimens to be tested and a specimen instrumented with surface and midspecimen
thermocouples into the temperature chamber.

11.6. Allow the test specimens to reach the specified testing temperature ±0.5°C (±1°F) by monitoring
the two thermocouples on the instrumented specimen. When the temperature has equalized, these
two measurements should not differ by more than ±1°C (±2°F). If these two measurements differ
by more than this value and it is certain that equilibrium has been reached, then record the mid-
specimen temperature and continue to the next step, but recalibrate the thermocouple reader or
replace the thermocouples before the next test specimen.

11.7. Insert the specimen into the testing machine, making sure to tighten the specimen securely to the
bottom support. When the specimen is firmly in place, bring the actuator into position and apply
approximately 0.09 kN (20 lbs) of seating load. Quickly secure the actuator with the upper loading
platen using a locking ball joint or screws, making sure not to shear the specimen unintentionally.

TS-2d TP 107-8 AASHTO

Больше стандартов - [Link]

 Note 11—For the sample ball joint shown in Figure 3, it is beneficial to fabricate the top end
platen and the cap on the locking ball joint with a series of holes for using opposing spanner
wrenches. After hand tightening the locking cap on the ball joint, insert a spanner wrench into the
locking cap while also inserting a spanner wrench into the top platen. While holding the top platen
stationary, turn the locking cap spanner wrench to tighten the ball joint. Do not tighten the top
platen by using a scissor action on the two spanner wrenches.

11.8. Reduce the load on the specimen to 0 kN ± 0.01 kN (0 lb ± 2 lb). Leave the testing machine in
load control to allow for free expansion and contraction.

11.9. Attach the sensors to the sample and position them at approximately the zero position.
This document is only for acquaintance

Note 12—It may be more convenient to attach the sensors while tightening the specimen to the
machine’s bottom support rod.

11.10. Allow the specimen to reach the specified testing temperature ± 0.5°C (± 1°F). If the machine
chamber has enough room, a monitoring specimen with thermocouples mounted at the center and
on the surface may be used to determine when the specimen reaches the specified test temperature.
The equilibrium time will vary depending on the available equipment (e.g., separate temperature
conditioning chamber), ambient conditions, and test temperature. If room does not exist for both
the monitoring specimen and the test specimen, either wait 4 h before testing or calibrate this time
separately by using a monitoring specimen alone in the AMPT environmental chamber.
Note 13—It has been found that, if the specimen is conditioned in a separate temperature-
conditioning chamber and the steps presented in Sections 11.7 through 11.9 can be completed
within 5 min, 1 h is enough to bring the specimen temperature back to the test temperature in the
AMPT chamber. Having two sets of LVDTs and temperature-conditioning the specimen with the
LVDTs mounted on it help in reducing the time required to complete the steps shown in
Sections 11.7 through 11.9.

11.11. For temperature calibration, measure the temperatures of the surface and the center of the
monitoring specimen and record the average of the two values. When the temperature has
equalized, these two measurements should not differ by more than ±1°C (±2°F). If these two
measurements differ by more than this value and it is certain that equilibrium has been reached,
then record the mid-specimen temperature and continue to the next step, but replace or recalibrate
the thermocouple before the next experiment.

11.12. Adjust, balance, and zero the electronic measuring system, as necessary.

11.13. Perform a fingerprint dynamic modulus test at the frequency of 10 Hz and at the target test
temperature. The fingerprint test shall be performed in the tension-compression mode of loading.
If using equipment with the necessary feedback controls (e.g., AMPT equipment) and the
appropriate control software, which should be capable of measuring and recording the time history
of the applied load and axial deformations during the test and can control on-specimen strain
automatically, follow steps presented in Sections 11.14 through 11.16. Otherwise follow the steps
shown in Sections 11.17 through 11.20.

11.14. In the equipment control software, enter a target strain range of 50 to 75 microstrains.

11.15. Start the fingerprint test and have the machine apply a few cycles at a load amplitude lower than
the estimated target loads in Table 3.

11.16. Calculate the load level necessary to achieve 50 to 75 microstrains using the results of these first
few cycles and then apply this load level for 50 cycles. Compute the dynamic modulus for the last
five cycles according to the method recommended in T 342 and TP 79. Skip to Section 11.21.

TS-2d TP 107-9 AASHTO

Больше стандартов - [Link]

 11.17. Identify the initial target load for fingerprint testing, Pdynamic, from Table 3 based on the testing
temperature, or if the dynamic modulus of the mixture is known, then the value may be calculated
using Equation 1.
Pdynamic= 60 × 10−6 × E * × πr 2 (1)
where:
Pdynamic = peak-to-peak load for the fingerprint test (kN or lb);
|E*| = dynamic modulus for the mixture of interest at the temperature and frequency of the
fingerprint test (kPa or psi); and
r = radius of the specimen (m or in.).
This document is only for acquaintance

Table 3—Estimated Target Loads for Fingerprint Test

Pdynamic
Temperature (°C \ °F) kN lbs
10 \ 50 4.71 1060
15 \ 59 3.56 800
20 \ 68 3.01 675

11.18. Apply the initial target load to the specimen at a frequency of 10 Hz for a total of 5 s and compute
the dynamic modulus for the last five cycles according to the method recommended in T 342 and
TP 79. If the strain levels obtained are between 50 and 75 microstrains, then record this value as
|E*|fingerprint and go to Section 11.21; otherwise, continue to Section 11.19.

11.19. Divide the peak-to-peak strain levels obtained from this test by 60 microstrains and denote this
value as Padj.

11.20. Divide the initial target load level identified in Section 11.17 by Padj and designate the result as
Pdynamic. Then, rerun Section 11.18 with the new Pdynamic until the on-specimen strain is between
50 and 75 microstrains. If the peak-to-peak strain exceeds 150 microstrains, discard the specimen.
Note 14—The specimen should be allowed to rest a total of 5 min between repetitions of
Section 11.20.

11.21. Compute the strains based on actuator displacements using Equation 2.

δ
εact =act (2)
L
where:
εact = peak-to-peak strain amplitude calculated from the actuator displacement;
δact = peak-to-peak actuator displacement; and
L = specimen height (mm or in.).

11.22. Compute the machine compliance factor using Equation 3.

εact
K= (3)
εos
where:
εos = peak-to-peak on-specimen strain amplitude calculated from Section 11.16 or 11.18; and
K = machine compliance factor.

11.23. Allow the specimen to rest for a period of 20 to 45 min following the fingerprint testing.

TS-2d TP 107-10 AASHTO

Больше стандартов - [Link]

 Note 15—If the testing machine does not have an automatic transfer capability between the load
control and displacement control modes, the rest period should be applied after manually
transferring the machine into displacement control mode.

11.24. Start the first cyclic fatigue test with the peak-to-peak on-specimen strain amplitude of
300 microstrains (εos1) and compute the actuator displacement value using Equation 4.
Note 16—The strain level calculated from the actuator displacement will not necessarily be the
same strain that the specimen undergoes. Machine compliance factors are likely to be significant
and specific to the testing equipment. A factor of 5 to 10 for machine compliance is reasonable to
expect.
δ = (ε * K ) * L (4)
This document is only for acquaintance

where:
δ = displacement amplitude level to be used at test temperature (mm or in.);
ε = target on-specimen strain level; and
L = specimen length (mm or in.).

11.25. Rezero the sensor readings to compensate for any slight drift that may have occurred during the
fingerprint test.

11.26. Perform a constant positive movement actuator oscillation fatigue experiment at a frequency of
10 Hz and at the amplitude calculated from Equation 4, e.g., a pull–pull actuator displacement test.
Record the load and all sensor displacements as a function of time for at least the first half of the
first cycle of loading (from zero to first peak) at a rate of 1000 samples per second. For the rest of
the applied cycles, it is sufficient to acquire only the cycle number, peak and valley values of
force, and the peak and valley values of sensor displacement. Alternatively, users may acquire the
full time history data for force and displacement and postprocess these data to obtain the peak and
valley values, or record the peak-to-peak values of stress, peak-to-peak values of strain, β values
(Section 12.19) and K1 values (Section 12.23) by cycle number. For the latter option, an
acquisition rate of 1000 samples per second is suggested. Stop the test when propagated micro-
cracks form one clear macrocrack. This macrocrack can be visually seen on the specimen surface
and will cause the specimen to break into two completely separate parts.

11.27. Repeat steps in Sections 11.1 through 11.26 on the remaining two specimens. The peak-to-peak
on-specimen strain levels of the second specimen (εos2) and the third specimen (εos3) in
microstrains can be found from Table 4 based on the resultant number of cycles to failure of the
first test (Nf1) and can then be converted to the actuator displacement using Equation 4.
Note 17—If the resultant number of cycles to failure of the first test (Nf1) is less than 500, the first
test has to be discarded and redone with the peak-to-peak on-specimen strain amplitude (εos1) of
250 microstrains.

Table 4—On-Specimen Strain Levels for the Second and Third Specimens
Case εos2 εos3
500 < Nf1 < 1,000 εos1 – 100 εos1 – 150
1,000 < Nf1 < 5,000 εos1 – 50 εos1 – 100
5,000 < Nf1 < 20,000 εos1 + 50 εos1 – 50
20,000 < Nf1 < 100,000 εos1 + 100 εos1 + 50
100,000 < Nf1 εos1 + 150 εos1 + 100

TS-2d TP 107-11 AASHTO

Больше стандартов - [Link]

 12. CALCULATIONS
12.1. This section presents a standard procedure for calculating the pseudo strain, pseudo secant
modulus, and damage for the fatigue tests. All the calculations in this section and in the
appendices can be automatically performed using ALPHA-Fatigue software.

12.2. Determine the E(t) Prony coefficients from the dynamic modulus and phase angle measured using
T 342 or PP 61.

12.3. Compute the specimen-to-specimen normalization parameter using Equation 5 and denote this
parameter as the DMR (Dynamic Modulus Ratio).
This document is only for acquaintance

E * fingerprint
DMR = (5)
E * LVE
where:
|E*|fingerprint = dynamic modulus determined from Section 11.14 or 11.16 (kPa or psi);
|E*|LVE = average representative dynamic modulus for the mixture of interest at the
temperature and frequency of interest (kPa or psi), and computed by Equation 6;
and
DMR = dynamic modulus ratio, which is the specimen variability compensation
parameter.
N 2 2
 Em ωR2 ρ m2   N Em ωR ρ m 
E* LVE =
=
 E∞ +
 m 1=
∑  + ∑ 
ωR2 ρ m2 + 1   m 1 ωR2 ρ m2 + 1 
(6)

where:
ω = angular frequency used in the fingerprint experiment;
aT = time–temperature shift factor for the fingerprint test temperature;
ωR = reduced angular frequency, Equation 7, used in the fingerprint experiment; and
E∞, Em, ρm = Prony coefficient terms.
ωR =ω * aT (7)

12.4. Separate the data into two parts. The first part, referred to as data set 1, comprises the data for the
first half of the first loading path (from zero to first peak stress). The second part, referred to as
data set 2, comprises the rest of the data.

12.5. For data set 1, average all sensor readings and compute the average strain for all data points using
Equation 8.

TS-2d TP 107-12 AASHTO

Больше стандартов - [Link]

 Note 18—If using four sensors and one provides unusual readings or fails during the experiment,
do not include that sensor in the average. Also, remove the sensor that is 180 degrees opposed to
that sensor.
δ
ε= (8)
GL
where:
ε = average axial strain;
δ = average axial displacement measured by the sensors (mm or in.); and
GL = sensor gauge length used in Section 10.8 (mm or in.).
This document is only for acquaintance

12.6. Compute the axial stress using Equation 9 for each data point in data set 1.
F
σ= (9)
πr 2
where:
σ = axial stress (kPa or psi);
F = axial force measured by the load transducer (kN or lb); and
r = specimen radius, Section 9.1, (mm or in.).

12.7. Compute the reduced time for each data point in data set 1 using Equation 10.
t
tR = (10)
aT
where:
aT = time–temperature shift factor at a given temperature;
t = time measured from the experiment (s), and
tR = reduced time (s).

12.8. Compute the pseudo strain for each data point in data set 1 using the state variable formulation
shown in Equation 11.
1  n +1 N n +1 
ε ( )=
R n +1
ER
 ηo +

∑
m =1
ηm 

(11)

where:
εR(n+1) = pseudo strain at the next time step;
ER = reference modulus, a value of 1 should be chosen;
η = elastic component of the pseudo strain (Equation 12);
ηm = pseudo strain contribution of Prony element m (Equation 13);
n = time step used in the calculation;
ε = strain calculated for the current or subsequent time step using Equation 8;
ΔtR = duration of the reduced time step, tRn+1–tRn; and
tR = reduced time.

η0n +1 = E∞ εn +1 ( ) (12)

 εn +1 − εn 
ηnm+1 e −∆tR
= ρm n
ηm + Em × ρm  −∆t
 1 − e R
ρm 
 (13)
 ∆ t R 

TS-2d TP 107-13 AASHTO

Больше стандартов - [Link]

 12.9. Compute the normalized pseudo secant modulus for each data point in data set 1 using
Equation 14.
σ
C= (14)
ε R * DMR

12.10. Compute the change in damage, ΔS, for each time step using Equation 15.
Note 19—Due to inherent electronic interference (data noise) during data acquisition, a few
sequential data points may have positive ΔC values. A few of these spurious data points do not
negatively affect the overall value of damage (S), but they do complicate the calculation. An
efficient method that accounts for these spurious data points is the piecewise function shown in
This document is only for acquaintance

Equation 17. This piecewise function can be included in a spreadsheet by using the if(…) function.
α
 DMR R 2
( )  α+1 1
∆Si = − ε ( Ci − Ci −1 )  ( ∆tR ) α+1 (15)
 2 
where:
Ci = pseudo secant modulus at the current time step;
Ci – 1 = pseudo secant modulus at the previous time step;
ΔtR = change in the reduced time step; and
α = continuum damage power term related to material time dependence, Equation 17.
 DMR α

( )
α+1
 − ( Ci − Ci −1 ) 
2 1
∆Si =
εR ( ∆tR ) α+1 Ci ≤ Ci −1
(16)
 2 

 0 Ci > Ci −1

1
α= +1 (17)
n
where:
n = maximum log-log slope of the relaxation modulus.

12.11. Determine the damage at each time step using Equation 18.
N
Si
= ∑ ∆S
i =1
i (18)

where:
Si = cumulative damage at the current time step; and
ΔSi = incremental damage for all time steps to be summed from the initial time step, i = 1, to
the current time step, N.

12.12. Define the damage at the final point in data set 1 as Sdataset 1.

12.13. Compute the peak-to-peak strain for each sensor and each cycle in data set 2 using Equation 19.
δipeak − δivalley
εipp = (19)
GL
where:
i = index to denote that the calculation is performed for all four sensors;
εipp = peak-to-peak axial strain for sensor, i;
δipeak = peak axial displacement measured by sensor, i (mm or in.); and
δivalley = valley axial displacement measured by sensor, i (mm or in.).

TS-2d TP 107-14 AASHTO

Больше стандартов - [Link]

 Note 20—The algorithms in T 342 or PP 61 may also be used for this purpose.

12.14. For each cycle in data set 2, average all sensor strains and denote this strain as the test peak-to-
peak strain amplitude, εpp.
Note 21—If using four sensors and one provides unusual readings or fails during the experiment,
do not include that sensor in the average. Also, remove the sensor that is 180 degrees opposed to
that sensor.

12.15. Compute the peak-to-peak stress for each cycle in data set 2 using Equation 20.
Fpeak − Fvalley
σ pp = (20)
This document is only for acquaintance

πr 2
where:
σ = axial stress (kPa or psi);
Fpeak = peak axial force measured by the load transducer (kN or lb);
Fvalley = valley axial force measured by the load transducer (kN or lb); and
r = specimen radius, Section 9.1, (mm or in.).
Note 22—The algorithms in T 342 or PP 61 may also be used for this purpose.

12.16. Compute the peak-to-peak pseudo strain for each cycle in data set 2 using Equation 21.
ε Rpp =ε pp * E * LVE (21)
where:
εRpp = peak-to-peak pseudo strain.

12.17. Compute the cyclic pseudo secant modulus for each cycle in data set 2 using Equation 22.
σ pp
C* = (22)
ε Rpp * DMR

12.18. Reduce the number of data points in data set 2 using the filtering scheme shown in Section X1.
Define this data set as reduced data set 2.
Note 23—Depending on the specific details of the data acquisition system, it may be more
convenient to acquire data at a reduced rate to avoid having to follow the data reduction procedure
in Section X4. However, because the specific amount of data reduction depends on many site-
specific factors, it is not possible to provide rigid guidance as to what this rate should be. The
recommended procedure is to acquire data for all cycles and apply the method in Section X4 to
determine the most efficient data acquisition rate for the specific system used. After this self-
calibration phase, additional mixtures can be tested at the more efficient data acquisition rate.

12.19. Compute the functional form factor, β, for each cycle in reduced data set 2 using Equation 23.
Fpeak + Fvalley
β= (23)
Fpeak + Fvalley

12.20. Compute the tension amplitude pseudo strain for each cycle in reduced data set 2 using Equation
24.
R β +1 R
εta
= ε pp (24)
2
where:
εRta = tension amplitude pseudo strain.

TS-2d TP 107-15 AASHTO

Больше стандартов - [Link]

 12.21. Compute the time within a cycle when tensile loading begins, tb, for each cycle in reduced data set
2 by using Equation 25.
cos −1 (β )
tb = (25)
62.83

12.22. Compute the time within a cycle when tensile loading ends, te, for each cycle in reduced data set 2
by using Equation 26.
2 π − cos −1 (β )
te = (26)
62.83
This document is only for acquaintance

12.23. Compute the form adjustment factor for each cycle in reduced data set 2 using Equation 27.
Equation 27 should be solved for each cycle, but generally β does not change significantly after
the first few cycles, and a constant value may be applied after this transient period. Values of K1
have been tabulated for typical values of β and α in Table 5.

Table 5—Compiled K1 Values for Typical Material and Test Conditions

Alpha
Beta 4.333 4.077 3.857 3.667 3.500 3.353 3.222
–0.5 0.277 0.285 0.293 0.300 0.306 0.312 0.318
0.0 0.263 0.271 0.278 0.285 0.291 0.297 0.302
0.2 0.256 0.264 0.271 0.277 0.284 0.289 0.295
0.4 0.248 0.256 0.262 0.269 0.275 0.280 0.286
0.6 0.238 0.245 0.252 0.258 0.264 0.269 0.274
0.8 0.225 0.231 0.238 0.243 0.249 0.254 0.259
1.0 0.189 0.195 0.200 0.205 0.209 0.214 0.218

 2 α te 
1  1  ( ) 
2α
=K1
te − tb  β + 1  ∫ β − cos ( 62.83* t ) dt (27)
 tb 
Note 24—Equation 27 can be solved with sufficient accuracy using a suitable numerical
technique with the number of time steps equal to 100.

12.24. Compute the average reduced time for each cycle in reduced data set 2 using Equation 28 or
Equation 29 depending on the data acquisition history.
1  tpeak + tvalley 
tR =   (28)
aT  2 
where:
tpeak = time at the peak force; and
tvalley = time at the valley force.
1 N 
tR = (29)
aT 10 
where:
N = cycle number.

12.25. Compute the change in damage, ΔS, for each cycle in the reduced data set 2 using Equation 30.
Note 25—Even with data reduction, a few sequential data points may have positive ΔC values. A
few of these spurious data points do not negatively affect the overall value of S, but they do
complicate the calculation. An efficient method that accounts for these spurious data points is to

TS-2d TP 107-16 AASHTO

Больше стандартов - [Link]

 use the piecewise function shown in Equation 31. This piecewise function can be included in a
spreadsheet by using the if(…) function.
α
 DMR R 2
( )  α+1 1 1
∆Sn = − εta ( C *n −C *n −1 )  ( ∆tR ) α+1 ( K1 ) α+1 (30)
 2 
where:
C*n = the cyclic pseudo secant modulus at the current analysis cycle;
C*n – 1 = the cyclic pseudo secant modulus at the previous analysis cycle; and
ΔtR = the change in the average reduced time between analysis cycles.
This document is only for acquaintance

 DMR α

( )
α+1
 − ( C *n −C *n −1 ) 
2 1 1
∆S n =
εtaR ( ∆t R ) α+1 ( K1 ) α+1 C *n ≤ C *n −1
(31)
 2 

 0 C *n > C *n −1

12.26. Determine the damage at each analysis cycle using Equation 32.
N
S=
n Sdataset1 + ∑ ∆S
n =1
n (32)

where:
Sn = cumulative damage at the current analysis cycle; and
ΔSn = incremental damage for all analysis cycles to be summed from the initial analysis cycle
step, n = 1, to the current time step, N.

12.27. Combine the damage and pseudo secant modulus from for each time step in the first cycle,
Sections 12.9 and 12.12, with the cyclic pseudo secant moduli and damage values from Sections
12.17 and 12.26, into a single dataset.

12.28. Determine the damage characteristic relationship by fitting one of the following equations to the
plot of the pseudo secant modulus and damage from all of the fatigue tests.
b
C = e aS or (33)
z
C = 1 − yS (34)
where:
a,b = the fitting coefficients for the exponential model; and
y, z = the fitting coefficients for the power model.

13. REPORT
13.1. Report the following for each specimen tested:

13.1.1. Test temperature;

13.1.2. The fingerprint dynamic modulus, |E*|fingerprint;

13.1.3. DMR value; and

13.1.4. C versus S curve coefficients (a and b in Equation 33 or y and z in Equation 34).

TS-2d TP 107-17 AASHTO

Больше стандартов - [Link]

 14. KEYWORDS
14.1. Axial deformation; complex modulus; cyclic fatigue; damage characteristic curve; direct tension;
DMR; dynamic modulus; failure cycle; fingerprint; modulus; phase angle; Prony coefficients;
relaxation modulus; pseudo secant modulus; pseudo strain; specimen deformation; strain.

APPENDIXES
(Nonmandatory Information)
This document is only for acquaintance

X1. CALCULATING THE RELAXATION MODULUS PRONY

COEFFICIENTS IN SPREADSHEET SOFTWARE
X1.1. This section presents a standard procedure for calculating the relaxation modulus, E(t), from the
dynamic modulus, |E*(ω)|, and phase angle, θ(ω), as measured according to T 342 or PP 61
protocol. The equations are presented in a way that allows the use of standard spreadsheet
software functions. These equations are specialized for Microsoft Excel software because of its
general availability. Users will need to have some familiarity with the Excel Solver utility.
Advanced users may find it beneficial to develop a stand-alone application for performing the
analysis.

X1.2. The basic approach for determining the relaxation modulus is the collocation method. For this
method, it is assumed that the relaxation modulus can be represented by Equation X1.1. The
coefficients in Equation X1.1 (Em, ρm, and E∞) can be established by solving the linear equation
shown in Equation X1.2. The following steps can be followed to accomplish this goal. For
Appendixes X1.9 through X1.17, users can refer to Figure X1.1 for an example of the spreadsheet
format.
N −t
E (=
t ) E∞ + ∑i =1
Em e ρm
(X1.1)

where:
E(t) = relaxation modulus as a function of time, t, (kPa or psi);
E∞ = long-time equilibrium modulus, (kPa or psi);
Em = modulus of Prony term number m, (kPa or psi);
ρm = relaxation time of Prony term m (s); and
N = number of Prony terms used.
Em
= [ B ]mi
−1
× {E ′}norm (X1.2)
where:
[B]mi = relaxation kernel matrix for determining the Prony coefficients; and
{E′}norm = shifted storage modulus array for determining the Prony coefficients.

X1.3. Compute the storage modulus, E′, for each temperature and frequency combination measured in
either T 342 or PP 61 via Equation X1.3.
 θ×π 
E′
= E * × cos   (X1.3)
 180 
where:
E′ = storage modulus (kPa or psi);

TS-2d TP 107-18 AASHTO

Больше стандартов - [Link]

 |E*| = dynamic modulus determined via experiment (kPa or psi); and
θ = phase angle determined via experiment (degrees).

X1.4. Compile the frequency, temperature, and storage moduli in a spreadsheet, formatted in a series of
three columns.

X1.5. Compute the angular frequency from the physical frequency using Equation X1.4.
ω= 2 × π × f (X1.4)

X1.6. Because measured data contain some variability, a smoothing process is needed to obtain reliable
coefficients. If the data was gathered using the T 342 protocol, fit the sigmoidal function to the
This document is only for acquaintance

storage modulus versus angular frequency data by optimizing the coefficients in Equation X1.5
and Equation X1.10 simultaneously. If the data were gathered using the PP 61 protocol, fit the
storage modulus versus angular frequency data by optimizing the coefficients in Equation X1.6
and Equation X1.11 or Equation X1.12 simultaneously. For this error minimization, use the sum
of squared errors shown in Equation X1.13. Designate this fit as “Log fit” and set aside the
optimized coefficients.
β
log ( E ′ ( ω, T ) ) = log ( E ′ ( ωR ) ) = κ + δ+ γ log ( ωR )
(X1.5)
1+ e
where:
E′(ω,T) = storage modulus at a particular temperature and angular frequency (kPa or psi);
E′(ωR) = storage modulus at a particular reduced angular frequency (kPa or psi);
ωR = reduced angular frequency, Equation X1.9 (rad/s); and
κ, β, δ, γ = fitting coefficients.
max − κ
log ( E ′ ( ω, T ) ) = log ( E ′ ( ωR ) ) = κ + δ+γ log ( ωR )
(X1.6)
1+ e
where:
E′(ω,T) = storage modulus at a particular temperature and angular frequency (kPa or psi);
E′(ωR) = storage modulus at a particular reduced angular frequency (kPa or psi);
ωR = reduced angular frequency, Equation X1.9 (rad/s);
max = defined by Equation X1.7; and
κ, δ, γ = fitting coefficients.

  VMA   VFA xVMA   1 − Pc

max = Pc  A 1 −  + B  + (X1.7)
  100   10, 000    1 − VMA  
 100
 VMA 
 + 
 A B( VFA) 
 
where:
Pc = defined in Equation X1.8 (kPa or psi);
VMA = voids in mineral aggregate, %;
VFA = voids filled with asphalt, %;
A = 4,200,000 for prediction in psi or 29,000,000 for prediction in kPa; and
B = 435,000 for prediction in psi or 3,000,000 for prediction in kPa.

TS-2d TP 107-19 AASHTO

Больше стандартов - [Link]

 0.58
 435, 000(VFA ) 
 20 + VMA 
  (X1.8)
Pc = 0.58
 435, 000(VFA ) 
650 +  
 VMA 
ωR =ω * aT (X1.9)
where:
aT = time–temperature shift factor at a given temperature, as shown in Equation X1.10,
Equation X1.11, or Equation X1.12.
( ) ( )
This document is only for acquaintance

α1 T 2 −Tref
2
+α2 T −Tref
aT = 10 (X1.10)
where:
T = temperature (°C or °F);
Tref = reference temperature (°C or °F); and
α1, α2 = fitting coefficients.

Ea  1 1 
 − 
2.303* R  273.15 + T 273.15 + Tref 

aT = 10 (X1.11)
where:
T = temperature (°C);
Tref = reference temperature (°C); and
Ea = activation energy determined by curve fitting (J/mol); and
R = universal gas constant (8.314 J/(K*mol)).

 A+VTS log(T ) A+VTS log(Tref ) 

c 10 −10 
aT = 10  
(X1.12)
where:
T = temperature (°Rankine);
Tref = reference temperature (°Rankine); and
A = intercept of the viscosity temperature susceptibility plot;
VTS = slope of the viscosity temperature susceptibility plot; and
c = fitting coefficient.
M 2
=OFlog ∑ log ( E′ )
j =1
fit j − log ( Emeas
′ )j
 (X1.13)

where:
OF = objective function;
E′fit = storage modulus predicted from Equation X1.5 for temperature and angular frequency, j;
E′meas = storage modulus calculated from Equation X1.3 for temperature and angular frequency, j;
and
M = total number of temperatures and frequencies measured in T 342 or PP 61.

X1.7. If using data obtained via T 342, refit the sigmoidal function to the storage modulus and angular
frequency data again, but this time use the weighted sum of the squared errors function shown in
Equation X1.14 and optimize the coefficients of Equation X1.5 only. For this optimization, use the

TS-2d TP 107-20 AASHTO

Больше стандартов - [Link]

 time–temperature shift factors determined from the Log fit. Designate this fit as “Weight fit” and
set aside the optimized coefficients.
M 2
OFweight = ∑ ( E ′
j =1
)
meas j × log ( Efit

′ ) j  
′ ) j − log ( Emeas
  (X1.14)

where:
OF = objective function;
E′fit = storage modulus predicted from Equation X1.5 for temperature and angular frequency, j;
E′meas = storage modulus calculated from Equation X1.3 for temperature and angular frequency, j;
and
This document is only for acquaintance

M = total number of temperatures and frequencies measured in T 342.

X1.8. If using T 342 data, locate the intersection of the Log and Weight fits by an appropriate
mathematical or numerical means and designate this frequency as ωc.
Note X1—A quick method to find the intersection is to plot both fits as a function of reduced
angular frequency, as shown in Equation X1.5, on the same graph and visually determine a close
range of intersection. Then, set up a small function using both sets of coefficients tied to the same
frequency cell. Make the difference between the two values zero by changing the value of the
frequency cell. Constrain the optimization to the range determined by the visual inspection.

X1.9. In a new spreadsheet, create an artificial data set of storage modulus values spaced one decade
apart by using Equation X1.5 (for AMPT based data) or the piecewise sigmoidal function given in
Equation X1.15 (for T 342 based data). Enter the artificial values of the reduced angular frequency
in the first column, Column A, starting at Row 3, and the storage modulus values in the second
column, Column B, starting at Row 3.
Note X2—This artificial data set should cover a reduced angular frequency range of at least two
decades before and after the measured data. For the example shown in Figure X1.1 a 5°C
reference temperature was chosen and this range was approximately 1 × 108 to 1 × 10–8 rad/s. If
choosing a different reference temperature, then the range of angular frequency values will need to
be determined from a visual inspection of the modulus versus reduced frequency plots.

 β Log rad
 κ Log + δLog +γ Log log( ωR )
for 0 ≤ ωR ≤ ωc
 1+ e sec
log ( E ′ ( ωR ) ) = (X1.15)
κ β Weight rad
Weight + for ωc < ωR ≤ ∞
 δ +γ
1 + e Weight Weight
log( ωR ) sec
where:
E′(ωR) = storage modulus at a particular reduced angular frequency (kPa or psi);
ωR = reduced angular frequency, as obtained from Equation X1.9;
ωc = critical reduced angular frequency (Appendix X1.8);
κLog, βLog, = fitting coefficients determined from the Log fit (Appendix X1.6); and
δLog, γLog
κWeight, βWeight, = fitting coefficients determined from the Weight fit (Appendix X1.7).
δWeight, γWeight

X1.10. Determine the long-time equilibrium modulus (E∞) using Equation X1.16 and enter the value in
the spreadsheet in Column B, Row 1.
κ Log
E∞ = 10 (X1.16)

TS-2d TP 107-21 AASHTO

Больше стандартов - [Link]

 X1.11. In Column C, subtract the long-time equilibrium modulus from each of the artificial storage
modulus values that are in Column B, and designate the resulting values as E′norm, as shown in
Equation X1.17.

′ = E ′ − E∞
Enorm (X1.17)

X1.12. Skip two columns and enter values for the relaxation time of Prony term m (ρm) coefficients into
the sixth column, Column F. There should be the same number of ρm values as there are artificial
frequencies. The ρm coefficients should be spaced one decade apart, exactly the same as the
artificial angular frequencies. Use values of 2 × 10x to 2 × 10y, where x and y are the same power
values used for the artificial angular frequencies for these coefficients. So, for example, if the
This document is only for acquaintance

reference temperature is 5°C, then the artificial angular frequencies would be 1 × 108, 1 × 107, 1 ×
106…1 × 10–7, and 1 × 10–8, and the ρm values would be 2 × 108, 2 × 107, 2 × 106…2 × 10–7, and
2 × 10–8.

X1.13. Starting in the seventh column, Column G, and at Row 2, duplicate the artificial angular frequency
values, but in row format, and denote these values as angular frequencies, ωi.

X1.14. For each cell of the matrix formed by the ωi and ρm values, Bmi, enter Equation X1.18.
ωi 2ρm2
Bmi = (X1.18)
ωi 2ρm2 + 1
where:
Bmi = relaxation kernel matrix for determining the Prony coefficients, same as in Equation
X1.2;
ωi = reduced angular frequency at the temperature and angular frequency condition, i (rad/s);
and
ρm = relaxation time of Prony term, m (s).

Note X3—This calculation can be simplified by using the absolute and relative referencing
properties in Microsoft Excel. Enter the function, as shown in Equation X1.19, into the first
element in matrix Bmi (Column G, Row 3). Then, fill this equation to the remaining cells of the
matrix, Bmi.
B11 = ( G$2 ^ 2 × $F3 ^ 2 ) / ( G$2 ^ 2 × $F3 ^ 2 + 1) (X1.19)

where:
B11 = reference to the value of the first element (Row 1 and Column 1) in the Bmi matrix;
G$2 = reference to the value of spreadsheet Column G, Row 2, with absolute referencing to
Row 2; and
$F3 = reference to the value of spreadsheet Column F, Row 3, with absolute referencing to
Column F.

X1.15. Use the rows below the final row of the Bmi matrix for finding the inverse of the Bmi matrix. Copy
the values of the ρm starting at Row 3 and Column G and paste a few rows below the final cell of
the Bmi matrix. Copy the values of the ωi values and paste immediately above the just pasted ρm
values.

X1.16. Select all cells bound by these just pasted values of ρm and ωi, Column G, Row 24 through
Column W, Row 40. While these cells are being selected, enter the following in the upper left-
most cell: =minverse (. Then, select all of the values of the Bmi matrix and enter the close
parenthesis, ). Hold the Control and Shift keys while pressing Enter. After pressing Enter, the

TS-2d TP 107-22 AASHTO

Больше стандартов - [Link]

 values for the inverse of the Bmi matrix should appear. If users experience difficulty, try an Internet
search for “Solving Linear Equations with Excel.” Denote this matrix as Bmi–1.

X1.17. Below matrix Bmi–1 ,copy and paste the value of ρm. Select the cells in the column immediately
beside these just pasted values of ρm. While keeping these cells selected, enter the following into
the cell immediately beside the first value of ρm: =mmult (. Then, select all of the values of the
Bmi–1 matrix, insert a comma, and select all rows of the E′norm array. Then, enter the close
parenthesis, ). Hold the Control and Shift keys while pressing Enter. The resulting values are the
Em coefficients given in Equation X1.1.
This document is only for acquaintance

TS-2d TP 107-23 AASHTO

Больше стандартов - [Link]

 TS-2d
This document is only for acquaintance
TP 107-24
AASHTO

Figure X1.1—Example of a Spreadsheet Format for Calculating Prony Coefficients

Больше стандартов - [Link]

 X2. CALCULATING PSEUDO STRAIN IN SPREADSHEET FORMAT
X2.1. This section presents a standard procedure for calculating the pseudo strain given the strain,
reduced time, and E(t) Prony coefficients. The equations are presented in a way that allows the use
of standard spreadsheet software functions, but are specialized for Microsoft Excel software
because of its availability in most engineering offices. Advanced users may find it beneficial to
develop a stand-alone application for performing the analysis. A sample spreadsheet format is
given at the end of this section in Figure X2.1.

X2.2. In a new spreadsheet, insert the long-time equilibrium modulus (E∞) value in the upper left cell.
This document is only for acquaintance

X2.3. In the row immediately below the E∞ value, insert the time–temperature shift factor, aT, for the test
of interest. This value can be determined by inserting the test temperature into Equation X1.10, or
by employing an equivalent model.

X2.4. In the row immediately below the aT value, insert the relaxation times, ρm, with column headings.
The values of ρm can be copied and pasted from the spreadsheet developed in Appendix X1.
However, the values should be pasted in a single row.

X2.5. In the row immediately below the ρm values, insert the Prony coefficients, Em, with column
headings. The values of Em can be copied and pasted from the spreadsheet developed in Appendix
X1. However, the values should be pasted in a single row.

X2.6. Skip two rows below the Em values and enter column titles for the following. Users may select any
order, but this section assumes the following order.
Column A – Time (s)
Column B – Stress (kPa or psi)
Column C – Strain
Column D – Reduced Time (s)
Column E – ΔStrain
Column F – ΔReduced Time (s)
Column G – η0
Column H – η1
Column I – η2
•
•
•
Column X – η17
Column Y – Pseudo Strain

X2.7. Paste the values of time, stress and strain into the appropriate columns.

X2.8. Compute the reduced time for each row using Equation 11.

X2.9. Compute the change in strain for each row by subtracting the value of the strain in the previous
row from the current row. For the first row, use the strain value for that row as the Δstrain.

X2.10. Compute the change in reduced time for each row by subtracting the value of reduced time in the
previous row from the current row. For the first row, use the reduced time value for that row as the
Δreduced time.

TS-2d TP 107-25 AASHTO

Больше стандартов - [Link]

 X2.11. Compute the elastic state variable, η0, for each row by multiplying the strain for that row by the
long-time equilibrium modulus, E∞.
Note X4—This calculation can be performed conveniently by using absolute referencing to the
cell containing the E∞ value and then pasting the formula in the remaining cells.

X2.12. Compute the state variable for the first time step of each Prony term by first entering Equation
X2.1 into Column H, Row 10.

1  ∆ε1   −∆t1R ρ1 
1 E1 × ρ1 
η=  ∆t1  1 − e  (X2.1)
 R
This document is only for acquaintance

where:
η11 = value of the first Prony term state variable at the first time step;
E1 = Prony coefficient value for the first Prony term;
ρ1 = relaxation time value for the first relaxation time;
Δε1 = change in strain occurring at the first time step; and
ΔtR1 = change in time occurring at the first time step.
Note X5—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X2.1, as shown
in Equation X2.2.
cellH10 = A$6 × A$4 × ( $E10 / $F10 ) [1 − e ^ ( −$F10 / A$4 )] (X2.2)
where:
H10 = an indication that everything after the equals sign would be entered into cell H10;
A$6 = reference to the value of spreadsheet Column A, Row 6, with absolute referencing to
Row 6;
A$4 = reference to the value of spreadsheet Column A, Row 4, with absolute referencing to
Row 4;
$E10 = reference to the value of spreadsheet Column E, Row 10, with absolute referencing
to Column E; and
$F10 = reference to the value of spreadsheet Column F, Row 10, with absolute referencing
to Column F.

X2.13. Fill this formula through Column X.

X2.14. Compute the state variable for the remaining time step of each Prony term by first entering
Equation X2.3 into Column H, Row 11.
2
−∆t R ρ1  ∆ε2   −∆t
2
ρ1 
=η12 e × η11 + E1 × ρ1  2 
1− e R (X2.3)
 ∆t  
R

where:
η11 = value of the first Prony term state variable at the first time step;
η12 = value of the first Prony term state variable at the second time step;
E1 = Prony coefficient value for the first Prony term;
ρ1 = relaxation time value for the first relaxation time;
Δε2 = change in strain occurring at the second time step; and
ΔtR2 = change in time occurring at the second time step.

TS-2d TP 107-26 AASHTO

Больше стандартов - [Link]

 Note X6—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X2.3, as shown
in Equation X2.4.
cellH11 = e ^ ( −$F11 / A$4 ) × H10 + A$6 × A$4 × ( $E11 / $F11) [1 − e ^ ( −$F11 / A$4 )] (X2.4)
where:
cellH11 = an indication that everything after the equals sign would be entered into cell
H11;
A$6 = reference to the value of spreadsheet Column A, Row 6, with absolute
referencing to Row 6;
A$4 = reference to the value of spreadsheet Column A, Row 4, with absolute
This document is only for acquaintance

referencing to Row 4;
$E11 = reference to the value of spreadsheet Column E, Row 10, with absolute
referencing to Column E;
$F11 = reference to the value of spreadsheet Column F, Row 10, with absolute
referencing to Column F; and
H10 = reference to the value of spreadsheet Column H, Row 10, with completely
relative referencing.

X2.15. Fill this formula through Column X and for all rows.

X2.16. Compute the pseudo strain by summing all of the Prony term state variables and the elastic state
variable using Equation 12. Because the reference modulus is equal to 1, the summation function
can be used directly for this purpose. An example for the first time step is given in Equation X2.4.
This same equation can be copied for all rows.
cellY9 = sum ( G10 : X10 ) (X2.5)

TS-2d TP 107-27 AASHTO

Больше стандартов - [Link]

 TS-2d
This document is only for acquaintance
TP 107-28
AASHTO

Figure X2.1—Example of a Spreadsheet Format for Calculating Pseudo Strain

Больше стандартов - [Link]

 X3. CALCULATING THE CONTINUUM DAMAGE POWER TERM, Α, IN
SPREADSHEET FORMAT
X3.1. This section presents a standard procedure for calculating the continuum damage model power
term, α, using the Prony coefficients determined in Appendix X1. The power term is related to the
log-log slope of the relaxation modulus. This value is found analytically using the Prony series
representation of the E(t) in Equation X1.1 and is fit using a Gaussian distribution. The equations
are presented in a way that allows the use of standard spreadsheet software functions. These
equations are specialized for Microsoft Excel software because of its availability in most
engineering offices. Users will need to have some familiarity with the Excel Solver utility.
This document is only for acquaintance

Advanced users may find it beneficial to develop a stand-alone application for performing the
analysis.

X3.2. The log-log derivative of the relaxation modulus in the Prony series is equal to Equation X3.1,
where the terms have been defined in previous equations. Equation X3.1 can be recast into
Equation X3.2 for easier interpretation. A sample spreadsheet format is given at the end of this
section in Figure X3.2.
N −t
t
d log E ( t )
∑ −E
m =1
me
ρm
×
ρm
= and (X3.1)
d log ( t ) N −t
E∞ + ∑E
m =1
me
ρm

d log E ( t ) 1
N
 −t
t 
d log ( t )
= ∑ −
 m
E ( t ) m =1 
E e ρm
× 
ρm 
(X3.2)

where:
E(t)m = the component of the relaxation modulus related to element m.

X3.3. In a new spreadsheet, insert the long-time equilibrium modulus (E∞) value in the upper left cell.

X3.4. Below this row, insert three rows with the titles of n, μ, and υ. In the column immediately adjacent
to these titles, enter the values of 0.4, 2, and 4, respectively. The values in these cells define the
shape and magnitude of the Gaussian distribution and will be optimized to minimize the error
between the analytical calculation and the actual Gaussian distribution.

X3.5. In the row immediately below the title, υ, enter the title, OF. The objective function value will be
entered in the cell immediately adjacent to this title (Appendix X3.21).

X3.6. In the row immediately below the OF title, insert the relaxation times, ρm, with column headings.
The values of ρm can be copied and pasted from the spreadsheet developed in Appendix X1.
However, the values should be pasted in a single row.

X3.7. In the row immediately below the ρm values, insert the Prony coefficients, Em, with column
headings. The values of Em can be copied and pasted from the spreadsheet developed in
Appendix X1. However, the values should be pasted in a single row.

X3.8. Skip two rows below the Em values and enter the title “dt-Multiplication Factor” in Column A. In
the same row, but in Column B, enter a value of 1.2.

TS-2d TP 107-29 AASHTO

Больше стандартов - [Link]

 X3.9. Skip two rows below the “dt-Multiplication Factor” row and enter column titles for the following.
Users may select any order, but this section assumes the following order.
Column A – Reduced Time (s)
Column B – E∞
Column C – E(t)1
Column D – E(t)2
•
•
•
This document is only for acquaintance

Column S – E(t)17
Column T – d log E(t)1
Column U – d log E(t)2
•
•
•
Column AJ – d log E(t)17
Column AK – Log-Log Derivative
Column AL – Fit Derivative
Column AM – Error

X3.10. Build an array of times with 200 elements. The most appropriate value for the first time in this
array will depend on the exact material being analyzed and on the reference temperature chosen in
Appendix X1.6. For this step, use a value of 1 × 10–6. The times should be spaced equally in
logarithmic space, which can be accomplished by setting the current time value equal to the
previous time value multiplied by the dt-Multiplication factor. In Appendix X3.18, this
multiplication factor and the initial time will be adjusted to more optimal values.
Note X7—This calculation can be performed efficiently by using absolute cell referencing. For
the cell immediately below the initial time and for the format shown in Figure X3.2, Equation
X3.3 is used to perform this operation. This function can be filled down for 200 rows.
cellA17
= A16 × B$13 (X3.3)

X3.11. Starting at the row where the initial time was input, fill all 200 rows of Column B with the value
of the long-time equilibrium modulus, E∞.

X3.12. Compute the relaxation modulus for all 200 rows and for each Prony term by first entering
Equation X3.4 into Column C, Row 15.
−t
ρ1
E11 = E1e (X3.4)
where:
E11 = the component of the relaxation modulus related to the first Prony term at the first time
value;
E1 = the Prony coefficient value for the first Prony term;
ρ1 = the relaxation time value for the first relaxation time; and
t1 = the first value of time (1 × 10–6).
Note X8—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X3.4, as shown
in Equation X3.5.

TS-2d TP 107-30 AASHTO

Больше стандартов - [Link]

 −$ A15
cellC=
15 A$9 × e A$7
(X3.5)
where:
cellC15 = an indication that everything after the equals sign would be entered into cell C15;
A$9 = reference to the value of spreadsheet Column A, Row 9, with absolute referencing to
Row 9;
A$7 = reference to the value of spreadsheet Column A, Row 7, with absolute referencing to
Row 7; and
$A15 = reference to the value of spreadsheet Column A, Row 15, with absolute referencing
to Column A.
This document is only for acquaintance

X3.13. Paste this function until Column S and for the 200 rows of times.

X3.14. Compute the partial logarithmic derivative of the relaxation modulus for the first value of time for
each Prony term by first entering Equation X3.6 into Column T, Row 15.
−t
t
(E )
1
1
der
− E1e
=
ρ1
×
ρ1
(X3.6)

where:
(E11)der = component of the logarithmic based relaxation modulus derivative related to the
first Prony term at the first time value;
E1 = Prony coefficient value for the first Prony term;
ρ1 = relaxation time value for the first relaxation time; and
t1 = first value of time.
Note X9—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X3.6, as shown
in Equation X3.7.
− A$9 × e ^ ( −$A15 / A$7 ) × ( $A15 / A$7 )
cellT15 = (X3.7)
where:
cellT15 = an indication that everything after the equals sign would be entered into cell
T15;
A$9 = reference to the value of spreadsheet Column A, Row 9, with absolute
referencing to Row 9,
A$7 = reference to the value of spreadsheet Column A, Row 7, with absolute
referencing to Row 7, and
$A15 = reference to the value of spreadsheet Column A, Row 15, with absolute
referencing to Column A.

X3.15. Paste this function until Column AJ and for the 200 rows of times.

X3.16. Compute the logarithmic derivative of the relaxation modulus for each of the 200 times by using
Equation X3.2 and the values computed in Appendixes X3.12 and X3.14.
Note X10—This calculation can be completed by using the algorithm shown in Equation X3.8.
cellAK 16 = sum ( T16 : AJ16 ) / sum ( B16 : S16 ) (X3.8)

X3.17. Plot the logarithmic derivative as a function of time with the time plotted as a logarithm, as shown
in Figure X3.1.
Note X11—The fluctuations seen in this graph are normal and are a result of the exponential
Prony series.

TS-2d TP 107-31 AASHTO

Больше стандартов - [Link]

This document is only for acquaintance

Figure X3.1—Example of a Plot of the Logarithmic Slope of the Relaxation Modulus as a Function of Time, E(t)

X3.18. Adjust the initial time and dt-Multiplication factor to obtain a mostly symmetric curve.
Note X12—For the distribution shown in Figure X3.1, the initial time has been adjusted to
2 × 10–7 s and the dt-Multiplication Factor is equal to 1.2.

X3.19. Insert the formula for the Gaussian distribution for log time versus log derivative of the relaxation
modulus, as indicated in Equation X3.9, into Column AL for all 200 time steps created in
Appendix X3.10.
−( log ( t ) − µ )
d log E ( t ) 2υ 2
= n×e (X3.9)
d log ( t )
where:
n = the amplitude of distribution;
μ = the centroid of distribution; and
υ = the standard deviation of distribution.

X3.20. In Column AM, compute the squared difference, Equation X3.10, between the values in Column
AK and the values in Column AL.
2
  d log E ( t )     d log E ( t )   
=SD j     −     (X3.10)
  d log ( t )     d log ( t )   
 Gauss  j  calc  j 

where:
j = the index to denote that the calculation is performed for each time value;
SDj = the squared difference at time step, j;
 d log E ( t ) 
  = the log derivative of E(t) calculated in Appendix X3.19; and
 d log ( t ) Gauss
 d log E ( t ) 
  = the log derivative of E(t) calculated in Appendix X3.16.
 d log ( t ) calc

TS-2d TP 107-32 AASHTO

Больше стандартов - [Link]

 X3.21. Insert a formula to sum all of the squared differences into the cell immediately adjacent to the cell
titled, OF (objective function), as shown in Equation X3.11.
200
OF = ∑ SD
j =1
j (X3.11)

X3.22. Use the Microsoft Solver function (or equivalent optimization tool) to minimize the OF value
while adjusting the values of n, μ, and υ.

X3.23. Determine the α value using Equation X3.12.

1
This document is only for acquaintance

α= +1 (X3.12)
n

TS-2d TP 107-33 AASHTO

Больше стандартов - [Link]

 TS-2d
This document is only for acquaintance
TP 107-34

Figure X3.2—Example of a Spreadsheet Format for Calculating the α Value

AASHTO

Больше стандартов - [Link]

 X4. PROCEDURE FOR REDUCING CYCLIC DATA IN SPREADSHEET
SOFTWARE
X4.1. This section presents a standard procedure for determining the appropriate data reduction and
filtering scheme to reduce cyclic data so that numerical errors are minimized. Variability and
measurement error can introduce significant errors in the damage calculation, particularly when
damage grows slowly (as in fatigue testing). To avoid this problem, sequential cycles should not
be used in the calculation of damage. Instead, damage should be calculated using cycles with some
space between them. For example, damage should not be calculated with real data using cycles 1,
2, 3, 4, etc.; instead, it should be calculated using cycles 1, 20, 40, etc. A universal rule for the
This document is only for acquaintance

appropriate cycle spacing depends on many factors specific to the test and measurement
equipment available and the test conditions used. This section presents a standard method to
determine the appropriate cycle spacing for the specific conditions under consideration.

X4.2. The basic method of data reduction is the moving average data reduction filter. In this data
reduction scheme, the term “filter ratio” is used to denote the extent of data filtering. The filter
ratio is the data reduction rate of the chosen filter. For example, a filter ratio of 20 indicates that
each group of 20 cycles is reduced to one data point.

X4.3. The most efficient data filtration scheme does not need to be uniform in its application; in fact, it
is likely that early cycles will be filtered less than later cycles. This situation arises because the
material state generally changes very rapidly early in the controlled crosshead cyclic test and then
changes more uniformly in the later stages. As a general rule, the cycles from the beginning of the
test until 10 percent of Nf are early cycles and those from 10 percent Nf until Nf are late cycles.

X4.4. Assume a filter ratio and average a number of cycles equal to that filter ratio. For example, if a
filter ratio of 20 is assumed, then average cycles 1 through 20. Averaging should include all
quantities, including time. Then, from the cycle immediately after the last cycle included in the
previous average; average a number of cycles equal to the filter ratio. For this example, one would
average cycles 21 through 40, then average cycles 41 through 60, etc. At any point the filter ratio
may be changed to best optimize the filtration.
Note X13—As a rule of thumb, apply a filter ratio of 10 to the early cycles and a filter ratio of
100 to later cycles.

X4.5. To help judge the filtration, compile a list of filtered cyclic pseudo modulus values.

X4.6. Compute the slope of the cyclic pseudo modulus versus the cycle for each compiled value,
according to Equation X4.1.
 C *n −C *n −1
 dC *   for Cn ≤ Cn −1
  = ∆N (X4.1)
 dN n  0 for C > C
 n n −1

where:
C*n = the pseudo secant modulus at the current cycle;
C*n–1 = the pseudo secant modulus at the previous cycle in the filtered data set; and
ΔN = change in the cycles.
Note X14—For n = 1, use the last value of C from data set 1 as C*n–1.

X4.7. Compute a quantity, referred to as the “assembled pseudo modulus” and denoted by C*asm for each
data element in the filtered data set by using Equation X4.2.

TS-2d TP 107-35 AASHTO

Больше стандартов - [Link]

 −1  dC * 
C *nasm =
C *nasm +  * ∆N (X4.2)
 dN n
n −1
Note X15—For n = 1, use the last value of C from data set 1 as C *asm .

X4.8. Apply a moving average filter without data reduction to the C*asm, C*, and N data compiled in the
previous step.
Note X16—An appropriate filter ratio is likely to be between 10 and 20.

X4.9. Compute the difference between the C*asm and C* computed in Appendix X4.8 using
Equation X4.3 and plot this difference with time.
This document is only for acquaintance

= C * −C *asm
difference (X4.3)

X4.10. If the filter ratios have been appropriately chosen, the difference between C*asm and C* should be
constant for the majority of the loading history. The difference itself is not critical, but it is
important that the difference does not change appreciably with the cycle number. Sample plots of
the difference between these two quantities versus cycle number are given in Figures X4.1 and
X4.2.
Note X17—Figure X4.1 shows the effect of the filtering from Section X4.8. The series labeled
“No Filter” has been compiled using Equations X4.1 through X4.3 without any data reduction
filter or moving average filter. The series labeled “Appendix X4.8 Only” has been compiled using
only the moving average filter described in Appendix X4.8, but without the data reduction filter
described in Appendix X4.4. A filter ratio of 20 was used for this calculation. In this case, the
filtration performed in Appendix X4.8 alone is not sufficient because the difference between C*
and C*asm continues to increase (see inset for better clarity).
Note X18—Figure X4.2 shows the effect of different data reduction filtering processes. When
compared with Figure X4.1 (inset), the differences have been substantially reduced by including
the data reduction filter. The series labeled “Heavy Filter” was created using a data reduction filter
ratio of 10 for the early cycles and a filter ratio of 100 for the later cycles. The series labeled
“Light Filter” was created by using a data reduction filter ratio of 2 for the early cycles and a ratio
of 10 for the later cycles. In both cases a moving average filter ratio (Appendix X4.8) of 20 was
used. The light filter results are at the threshold between acceptable and not acceptable filtration.

TS-2d TP 107-36 AASHTO

Больше стандартов - [Link]

This document is only for acquaintance

Figure X4.1—Example of Underfiltered Cyclic Data

Figure X4.2—Example of Threshold-Filtered and Optimally Filtered Cyclic Data

X5. PROCEDURE FOR CALCULATING DMR IN SPREADSHEET

SOFTWARE
X5.1. This section presents a standard procedure for calculating the DMR. This section makes reference
to the sample layout given in Figure X5.1.

X5.2. In a new spreadsheet, label and insert the test temperature into the upper left cell (label in cell A1
and quantity in cell B1).

TS-2d TP 107-37 AASHTO

Больше стандартов - [Link]

 X5.3. In the row immediately below alpha, enter the |E*|fingerprint value for the replicate under
investigation (cell B2).

X5.4. In the row immediately below the |E*|fingerprint value, enter the time–temperature shift factor for the
test temperature entered in Appendix X5.5 (cell B3). Use the appropriate equation chosen during
Appendix X1.6 to calculate the shift factor.

X5.5. In the row immediately below the time–temperature shift factor, enter the test frequency in Hertz
(cell B4).

X5.6. In cell B5, calculate the reduced test frequency by multiplying cell B3 by cell B4:
This document is only for acquaintance

cellB5
= B3 × B4 (X5.1)

X5.7. In cell B6, convert the reduced frequency in Hz to reduced frequency in radians per second:
cellB6= B5 × 2 π (X5.2)

X5.8. Enter a blank row.

X5.9. Insert the long-time equilibrium modulus and label (E∞) value in first cell (label in A8 and value in
B8).

X5.10. In the row immediately below E∞, insert the relaxation times, ρm, with column headings. The
values of ρm can be copied and pasted from the spreadsheet developed in Appendix X1. However,
the values should be pasted in a single row.

X5.11. In the row immediately below the ρm values, insert the Prony coefficients, Em, with column
headings. The values of Em can be copied and pasted from the spreadsheet developed in
Appendix X1. However, the values should be pasted in a single row.

X5.12. Skip two rows below the dt-Multiplication Factor row and enter column titles for the following.
Users may select any order but this section assumes the following order:
Column A – E′ (kPa or psi)
Column B – E″ (kPa or psi)
Column C – |E*| (kPa or psi)
Column C – E′1
Column D – E′2
•
•
•
Column X – E′21
Column Y – E″1
Column Z – E″2
•
•
•
Column AS – E″21

X5.13. Compute the storage modulus for each individual Prony term by first entering Equation X5.3 into
Column D, Row 17.

TS-2d TP 107-38 AASHTO

Больше стандартов - [Link]

 E1ω2Rρ12
E1′ = (X5.3)
ω2Rρ12 + 1
where:
E′1 = the component of the storage modulus related to the first Prony term;
E1 = the Prony coefficient value for the first Prony term;
ρ1 = the relaxation time value for the first relaxation time; and
ωR = the reduced angular frequency calculated in Appendix X5.7.
Note X19—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X5.3, as shown
This document is only for acquaintance

in Equation X4.3.
A$13 × ( $B$6 ) ^ 2 × ( A$11) ^ 2
cellD17 = (X5.4)
( $B$6 ) ^ 2 × ( A$11) ^ 2 + 1
where:
cellD17 = an indication that everything after the equals sign would be entered into cell
D17;
A$13 = reference to the value of spreadsheet Column A, Row 13, with absolute
referencing to Row 13;
A$11 = reference to the value of spreadsheet Column A, Row 11, with absolute
referencing to Row 11; and
$B$6 = reference to the value of spreadsheet Column B, Row 6, with absolute referencing to
Column B and Row 6.

X5.14. Fill this function to the right until column X.

X5.15. Compute the loss modulus for each individual Prony term by first entering Equation X5.5 into
Column Y, Row 17.
E1ωRρ1
E1′′ = (X5.5)
ω2Rρ12 + 1
where:
E″1 = component of the storage modulus related to the first Prony term;
E1 = Prony coefficient value for the first Prony term;
ρ1 = relaxation time value for the first relaxation time; and
ωR = reduced angular frequency calculated in Appendix X5.7.
Note X20—This calculation can be performed conveniently with the appropriate use of relative
and absolute referencing. For the format suggested in this section, enter Equation X5.5, as shown
in Equation X5.6.
cellY17 =  A$13 × ( $B$6 ) × ( A$11)  / ( $B$6 ) ^ 2 × ( A$11) ^ 2 + 1 (X5.6)
where:
cellY17 = an indication that everything after the equals sign would be entered into cell Y17;
A$13 = reference to the value of spreadsheet Column A, Row 13, with absolute referencing
to Row 13;
A$11 = reference to the value of spreadsheet Column A, Row 11, with absolute referencing
to Row 11; and
$B$6 = reference to the value of spreadsheet Column B, Row 6, with absolute referencing to
Column B and Row 6.

TS-2d TP 107-39 AASHTO

Больше стандартов - [Link]

 X5.16. Fill this function to the right until column AS.

X5.17. Compute the total Storage modulus according to Equation X5.7 or in terms of the spreadsheet
shown in Figure X5.1 by Equation X5.8.
N
Em ω2Rρm
2
′ E∞ +
E= ∑ 2 2
m =1 ωRρ m +
1
(X5.7)

cell=
A17 $B$9 + sum(D17 : X17) (X5.8)

X5.18. Compute the total Loss modulus according to Equation X5.9 or in terms of the spreadsheet shown
This document is only for acquaintance

in Figure X5.1 by Equation X5.10.

N
Em ωRρm
E ′′ = ∑ω ρ 2 2
m =1 R m +
1
(X5.9)

cell=
A17 $B$9 + sum(D17 : X17) (X5.10)

X5.19. Compute the LVE |E*| according to Equation 7 or in terms of the spreadsheet shown in Figure
X5.1 by Equation X5.11.
cellC17 sqrt ( A17 ) ^ 2 + ( B17 ) ^ 2 
= (X5.11)

X5.20. Compute the DMR according to Equation 6 or in terms of the spreadsheet shown in Figure X5.1
by Equation X5.12.
cellB7 = B3 / C17 (X5.12)

TS-2d TP 107-40 AASHTO

Больше стандартов - [Link]

 TS-2d
This document is only for acquaintance
TP 107-41

Figure X5.1—Example of a Spreadsheet Format for Calculating DMR

AASHTO

Больше стандартов - [Link]

 X6. PROCEDURE FOR CALCULATING DAMAGE VARIABLE, S, FOR
FATIGUE TEST IN SPREADSHEET SOFTWARE
X6.1. This section presents a standard procedure for calculating the damage variable for a fatigue test. A
sample spreadsheet format is given at the end of this section in Figure X6.1.

X6.2. This step can be completed after the pseudo strain has been calculated for the first half of the first
loading cycle via the procedure detailed in Appendix X2. This appendix demonstrates the
calculations carried out as part of Sections 12.9 through 12.25.
This document is only for acquaintance

X6.3. In a new spreadsheet label insert the alpha value determined from Section X3 into the upper left
cell (label in cell A1 and quantity in cell B1).

X6.4. In the row immediately below alpha enter the |E*|fingerprint value for the replicate under
investigation (cell B2).

X6.5. In the row immediately below the |E*|fingerprint value, enter the test temperature in degrees Celsius
(cell B3).

X6.6. In the row immediately below the temperature, enter the time–temperature shift factor for the test
temperature entered in Appendix X5.5 (cell B4). Use the appropriate equation chosen during
Appendix X1.6 to calculate the shift factor.

X6.7. In the row immediately below the time–temperature shift factor, enter the test frequency in Hertz
(cell B5).

X6.8. In cell B6, calculate the reduced test frequency by multiplying cell B4 by cell B5:
cellB6
= B5 × B4 (X6.1)

X6.9. In cell B7, enter the value of DMR determined from Appendix X5 for the particular test of interest.
It may be more convenient to link the DMR sheet developed in Appendix X5 to this spreadsheet,
but it is not necessary.

X6.10. Skip one row below the DMR values and enter column titles for the following. Users may select
any order, but this section assumes the following order:
Column A – Time (s)/Cycle
Column B – Reduced Time (s)
Column C – Stress/Peak-to-Peak Stress (kPa or psi)
Column D – Peak Force (kN or lb) Reduced Time (s)
Column E – Valley Force (kN or lb)
Column F – Strain/Peak-to-Peak Strain
Column G – Beta
Column H – Pseudo Strain/Pseudo Strain Peak-to-Peak
Column I – Pseudo Strain-ta
Column J – K1
Column K – C
Column L – dS
Column M – S

TS-2d TP 107-42 AASHTO

Больше стандартов - [Link]

 X6.11. Insert a zero for the first value of all columns except C. For the first value of C (cellK10), enter a
value of 1.

X6.12. For the first half cycle of loading, compute pseudo strain from the method outlined in Section X2
and then copy and paste the time into column A (starting at row 10), copy and paste the stress into
column C (starting at row 0). You may also paste these values into column D and E, but it is not a
requirement. Copy and paste the strain for the first loading cycle into column F (starting at row
10). For all rows containing the first half cycle, you may replace the contents of column G and
column J with a 1 or leave it blank. Finally, copy and paste the values of pseudo strain calculated
by the method of Section X3 into column H. You may also paste these values into column I or you
may leave column I blank.
This document is only for acquaintance

Note X21—In the example spreadsheet shown in Figure X6.1, the number of data points in the
first half of the first cycle is 50.
Note X22—You may also develop a spreadsheet that links the various quantities together if you
do not wish to copy and paste values.

X6.13. Compute the reduced time for all points in the first half of the first loading cycle by dividing the
time in column A by the time–temperature shift factors.
Note X23—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first reduced time shown in Figure X6.1, Equation X6.2 is used to perform this
operation. This function can be filled down for all rows of the first half loading cycle.
cellB11
= A11 × B$4 (X6.2)

X6.14. Compute the C value for all points in the first half of the first loading cycle by Equation 15.
Note X24—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first C shown in Figure X6.1, Equation X6.3 is used to perform this operation.
This function can be filled down for all rows of the first half loading cycle.
cellK11 C11 / ( H11 × B$7 )
= (X6.3)

X6.15. Compute the dS value for all points in the first half of the first loading cycle by Equation 17.
Note X25—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first dS shown in Figure X6.1, Equation X6.4 is used to perform this operation.
This function can be filled down for all rows of the first half loading cycle.
  B$2
 
 B$7 H11 2 K10 − K11  B$2 +1 1
cellL11 = if  K11 > K12,   − ( ) ( ) ( B11 − B10 ) B$2 +1 ,0 (X6.4)
  2   
   

X6.16. Compute the damage at each time step of the first half of loading by summing all damage prior to
that time step plus the increment of damage at that time step, e.g., Equation 19.
Note X26—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first S shown in Figure X6.1, Equation X6.5 is used to perform this operation.
This function can be filled down for all rows of the first half loading cycle. When filled down, the
equation in cell L20 would read as shown in Equation X6.6.
cellM11 = sum(L$10 : L11) (X6.5)
cellM20 = sum(L$10 : L20) (X6.6)

X6.17. Move to analysis of data set 2 and insert the filtered cycle data from Section X4 starting below the
just analyzed first half loading data.

TS-2d TP 107-43 AASHTO

Больше стандартов - [Link]

 X6.18. Paste the cycle number into column A, paste the peak-to-peak stress values, Equation 21, into
column C, paste the maximum force values into column D, paste the valley force values into
column E, paste the peak-to-peak strain values, Equation 20 into column F.

X6.19. Compute the reduced time for all points in data set 2 by dividing the cycle number in column A by
the reduced frequency.
Note X27—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first reduced time for data set 2 shown in Figure X6.1, Equation X6.7 is used to
perform this operation. This function can be filled down for all rows of data set 2.
cellB60 = A60 / B$6 (X6.7)
This document is only for acquaintance

X6.20. Compute beta for all points in data set 2 by using Equation 24.
Note X28—For the value of the first beta for data set 2 shown in Figure X6.1, Equation X6.8 is
used to perform this operation. This function can be filled down for all rows of data set 2.
D60 + E60 / ( ABS ( D 60 ) + ABS ( E60 ) )
cellG60 = (X6.8)

X6.21. Compute the Peak-to-Peak Pseudo Strain for every row in data set 2 by Equation 22.

X6.22. Compute C for all points in data set 2 by using Equation 24.
Note X29— This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first C for data set 2 shown in Figure X6.1, Equation X6.9 is used to perform this
operation. This function can be filled down for all rows of data set 2.
cellK60 C60 / ( H60 × $B$7 )
= (X6.9)

X6.23. Compute the tension–amplitude pseudo strain for all points in data set 2 by using Equation 25.
Note X30—For the value of the first tension amplitude pseudo strain for data set 2 shown in
Figure X6.1, Equation X6.10 is used to perform this operation. This function can be filled down
for all rows of data set 2.
cellI60 = ( ( G10 + 1) / 2 ) × H60 (X6.10)

X6.24. Compute K1 for all points in data set 2 by using Equation 28, by lookup for each cycle with
Table 5, or by regression with the data in Table 5.

X6.25. Compute the dS value for all points data set 2 by Equation 32.
Note X31—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first dS shown in Figure X6.1, Equation X6.4 is used to perform this operation.
This function can be filled down for all rows of the first half loading cycle.
  B$7 B$2
B$2 +1  
 
 − ( I60 ) ( K59 − K60 ) 
2
 
cell
= L60 if  K59 > K60,   2  ,0 (X6.11)
  1 1  
  ( B11 − B10 ) B$2 +1 ( J60 ) B$2 +1  

X6.26. Compute the damage for each row in data set 2 by summing all damage prior to that time step plus
the increment of damage at that time step, e.g., Equation 19.

TS-2d TP 107-44 AASHTO

Больше стандартов - [Link]

 Note X32—This calculation can be performed efficiently by using absolute cell referencing. For
the value of the first S shown in Figure X6.1, Equation X6.12 is used to perform this operation.
This function can be filled down for all rows of the first half loading cycle.
cellM60 = sum(L$10 : L60) (X6.12)
cellM20 = sum(L$10 : L20) (X6.13)
This document is only for acquaintance

TS-2d TP 107-45 AASHTO

Больше стандартов - [Link]

 TS-2d
This document is only for acquaintance
TP 107-46
AASHTO

Figure X6.1—Example of a Spreadsheet Format for Calculating Damage Variable

Больше стандартов - [Link]

 X7. PROCEDURE FOR ESTIMATING ON-SPECIMEN STRAIN LEVELS IN
SPREADHSHEET SOFTWARE
X7.1. This appendix presents a procedure for estimating strain levels for fatigue tests based upon the
modulus of the test material. This section provides guidelines for testing materials with unknown
fatigue characteristics.

X7.2. Estimate the on-specimen strain amplitude for 1000 cycles to failure by substituting |E*|fingerprint or
|E*|LVE for the study mixture into Equation X7.1.
This document is only for acquaintance

ε1000 = 1257e−0.00015*|E*| (X7.1)

where:
|E*| = either |E*|LVE or ||E*|fingerprint (MPa); and
ε1000 = initial test peak-to-peak strain magnitude for 1000 cycles to failure in microstrain.

X7.3. Depending on the testing equipment, it may be necessary to convert the on-specimen target strain
magnitude to an actuator displacement using Equation 5.

X7.4. After performing the experiment at the target strain magnitude from Section X7.2, record the
number of cycles to failure as Nf,target1000 and the actual initial peak-to-peak strain magnitude as
ε0,target1000 in microstrain.

X7.5. Estimate the target strain magnitude for 10,000 cycles to failure using the quantities from Section
X7.4 by using Equation X7.2.
−1
  5.5
 
 10000 
ε10000 =
N  (X7.2)
f ,target1000
 −5.5 

 (
ε0,target1000 ) 


where:
ε10000 = the initial test peak-to-peak strain magnitude for 10,000 cycles to failure in microstrain.

X7.6. Depending on the testing equipment, it may be necessary to convert this on-specimen target strain
magnitude to an actuator displacement using Equation 5.

TS-2d TP 107-47 AASHTO

Больше стандартов - [Link]