Study of Long-Term Pavement Performance (LTPP): Pavement Deflections

Chapter 7. Suggested Computed Parameters and FWD Testing Protocols

The Initial Working Data Files for the Project

As mentioned near the beginning of this report, the originally provided database files were reorganized to facilitate further analyses. Specifically, day files were created under directories organized by region and test section; these files consisted of all the data for each date of test at each LTPP section. These day files were changed in accord with the findings of this study to include the deletions, changes, and notes reported above. All of these files were written to a computed parameter data CD, which consisted of normalized and averaged FWD load-deflection data, with data errors and anomalies identified, changed, or deleted from the files as appropriate.

The Modified (Computed Parameter) Individual Data Files

As mentioned previously, each day file was identified by a six-character date (yymmdd). After each date, in the seventh character of the file name, the letter “n” or “m” was employed. The letter “n” stands for normalized, while “m” stands for modified after normalization. Finally, each normalized and/or modified day file has the extension “*.txt” since these are comma-delimited files. Occasionally, “u” or “x” was used instead of the usual “n” or “m.” These two letters denote that a note only is present, with no data. Such notations are further explained below.

The “n” files denote those day files that have been screened for inconsistent basin type load-deflection errors. They have been normalized to the target load levels for each drop height, while each line or record represents the average of the deflection readings for a given test point and drop height. The field for drop sequence number has been replaced with a field indicating the number of drops (1, 2, 3, or 4) used to create the averages. The “n” files have no other changes or notes attached. The “m” files denote that some additional change or changes have taken place. In all these cases, there is a note in the last record of the “m” file that explains the changes or deletions (in some cases) or warns the user of potential problems (in other cases). These pre-autumn 1998 data files presently exist on a single data CD (approximately 100 megabytes).

However, it was decided that these files would not be incorporated into the LTPP database. Therefore, no computed parameters of pavement deflections exist at this time in the LTPP database. It is, however, still possible that the “n” files could eventually be consolidated into database tables so that they are available for general use.

FWD Test Protocol Recommendations

Based on the findings of this report and other considerations, the following LTPP FWD testing procedure changes were recommended in 1999:

1. Install the Dynatest Edition 25 Field Program, in lieu of the currently employed Edition 20.
2. Utilize all nine available deflection sensors, in lieu of the currently employed seven sensors.
3. Place sensors 1 through 9 permanently at the following positions: 0, 203, 305, 610, 914, 1219, 1524, 1829, and –305 mm, respectively. These positions will permit all types of tests with no maneuvering of sensors. The corresponding U.S. customary sensor positions, in inches, are: 0, 8, 12, 18, 24, 36, 48, 60, and –12, respectively.
4. With a simple software change to Edition 25, request Dynatest to allow for the recording of both the infrared (IR) pavement surface temperature at the time the F1 key is pressed, and also at the time a test sequence ends; the current procedure only records the latter measurement. One of the many available fields in the comma-delimited “Station” line of Edition 25 can be used for the extra temperature value (along with the usual time-of-departure IR temperature reading).
5. Continue utilizing three seating drops, as previously, at drop height 3; however, record the peaks of these three seating drops for potential use in analyzing the pavement’s hardening or softening properties, which may be related to pavement performance.
6. It is adequate to utilize only three drop heights for all types of bound-layer tests, whether PCC or AC, and for all types of LTPP testing, whether General Pavement Studies (GPS), Specific Payment Studies (SPS), or the Seasonal Monitoring Program (SMP). These drop heights should be 2, 3, and 4, or approximately 9, 12, and 16 kips, respectively.
7. It is adequate to utilize only three drops per drop height for all types of bound-layer tests, whether PCC or AC, and for all types of LTPP testing, whether GPS, SPS, or SMP.
8. It is adequate to store only one full load-deflection time history for each test point, for all types of bound layer tests, whether PCC or AC, and for all types of LTPP testing, whether GPS, SPS, or SMP. If it is not possible to record more than one full time history per test point, this full time history should be the last drop at the highest utilized drop height.
9. Utilize the same spacing between test points as is currently used for the various types of experiments (GPS, SPS, and SMP).
10. Do not eliminate or delete FWD operator comments from the level E database.
11. Make sure both IR temperatures (at factory calibration settings) and manual temperatures are monitored according to the original protocols, and that the temperature measuring equipment is operating properly. Reemphasize the importance of correct protocol in measuring indepth temperatures.
12. Furnish the regions with the transformed basin or SLIC procedure software for all types of data storage files. This screening tool should help correct or eliminate data errors well before they reach level E in the database, whether these data errors may be in the form of misplaced sensors, faulty sensors, or sensor holders.
13. Emphasize the importance of checking sensor spacings and the stability of the sensor holders, magnets, and other equipment (including the center sensor) prior to testing at each LTPP test section.
14. All other current FWD testing and QA/QC protocols and procedures—especially those involving accuracy of actual sensor positions and the correct method of conducting pavement temperature measurements—should remain the same.

Further, an analysis was conducted to develop a procedure to determine if more frequent LTPP testing should be conducted as a pavement ages and exhibits more distress. The approach was to use a representative measure of the deflection basin, adjust this measure for pavement temperature, and determine whether definitive trends in the selected deflection measure exist.

The following specific analysis approach was used:

1. After a review of various deflection basin measures, it was decided to use the AREA measure to represent the overall characteristics of the deflection basin. The AREA measure is a calculation of the normalized area of a deflection basin. Facilitating the use of the AREA measure was the fact that a temperature adjustment procedure particularly pertinent to the AREA measure was available. The AREA, using U.S. customary units, has traditionally been defined as shown by the equation in figure 5:

where: d(n) = deflection at n inches (n * 25.4 mm) from the center of the load plate

Figure 5. Equation. AREA.

A list was developed of GPS–1 and GPS–2 test sections that had at least four different times (nonseasonal) of deflection testing between 1989 and 1998. Ten of these test sections were randomly selected for further analysis.

The AREA values were determined for each F3 test point (wheelpath testing at 7.6-m (25-ft) test intervals).

The average middepth temperatures during the approximately 1 hour of wheelpath testing were established, for each test visit and for each of the 10 sections.

The basin adjustment factor (TAF) for the AREA parameter was established using the procedure described in a previous FHWA study, Temperature Predictions and Adjustment Factors for Asphalt Pavement.⁽²⁾ Although the TAF is a function of the AC layer thickness and the latitude of the test section, it typically ranges in almost a linear manner from about 0.90 at 0 ºC (32 ºF) middepth temperature to about 1.1 at 40 ºC (104 ºF) middepth temperature, using the reference middepth temperature of 20 ºC (68 ºF). The temperature adjusted AREA is then given by the equation in figure 6:

Figure 6. Equation. AREA (@ 20 ºC).

All AREA values determined in step 3 were adjusted to account for the middepth pavement temperature as shown in step 5. The adjusted AREA values are listed in table 8.

A review of the results shown in table 8 indicates no definitive trends in average AREA values with time for the ten sections used in the analysis. Therefore, no specific recommendations can be provided at this time for considering changes in the frequency of deflection testing. Also, no acceptable procedures presently exist to adjust the maximum deflections for temperature at the time of test, although this may be possible in the near future. Finally, it may be necessary to include sections exhibiting significant cracking and/or rutting distress in the study sample, since it is possible that the deflections do not change appreciably until some pavement distress ensues.

The recommendations shown above were reported in 1998. Most (but not all) of these recommendations have since been implemented in connection with FWD upgrades, annual equipment servicing, and other events, and the FWD deflection testing protocols have been changed accordingly, where appropriate.