U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
Federal Highway Administration Research and Technology
Coordinating, Developing, and Delivering Highway Transportation Innovations
|This report is an archived publication and may contain dated technical, contact, and link information|
Publication Number: FHWA-HRT-01-167
Date: April 2005
Structural Factors of Jointed Plain Concrete Pavements: SPS-2—Initial Evaluation and Analysis
Chapter 4. Experimental Design Versus Actual Construction
One of the main objectives of this study is to identify confounding factors introduced into the SPS-2 experiment by virtue of construction deviations or other factors not accounted for in the original experimental design. It is important to evaluate the variables that are considered as key design factors in the SPS-2 experiment and to determine if they meet the parameters established in the design factorial. Additionally, two SPS guideline reports established specific site-selection criteria and key variable construction guidelines.(11,12) The guidelines in both reports were developed to control the quality and integrity of the SPS-2 experiment results and findings, and therefore should be included in the construction adequacy evaluation.
This chapter evaluates the design and the actual construction of key variables identified in the experiment design factorial and the above-mentioned guidelines. This includes the following:
The experimental design specified that the SPS-2 sites be located in four specific climates:
The main purpose of this requirement was to obtain representative SPS-2 sections in widely varying climates, with a geographic distribution across the continental United States Table 28 shows a summary of the design requirements and actual precipitation data. All of the sites meet the criteria, except two that were supposed to be in dry areas (Kansas and North Dakota). The Kansas site is much wetter than the design limit, and the North Dakota site is just barely wetter than the limit.
What effect will these deviations have on achieving the objectives relative to climate? Analysis of the data will utilize the actual precipitation, not dry or wet variables. The only limitation is that the performance from the Kansas site will represent an area with greater precipitation than desired (819 mm versus 508 mm maximum or dry area); however, this site is still much drier than the corresponding wet sites. Kansas has 819 mm annual precipitation, and the other wet sites range from 865 to 1,380 mm with an average of 1,068 mm.
The freezing index data are shown in table 29. As shown, all sites meet the criteria for freeze and non-freeze based on the annual freezing index criteria.
The SPS-2 experimental design called for half of the sites to be constructed on coarse-grained subgrade soils, and the other half to be constructed on fine-grained soils. Furthermore, it was required that all test sections at one site must be constructed on soils classified as same soil type, either fine-grained or coarse-grained.
Table 30 provides a comparison of the designated versus constructed subgrade types for all SPS-2 projects. Information from both cores taken from constructed pavements (TST_L05B table) and construction surveys (SPS2_LAYER table) is provided for comparison purposes. As indicated, for 11 of 13 SPS-2 projects, the subgrade soils are approximately uniform for all the core sections within the project. Furthermore, the soil types are now consistent between the designated and the constructed after correcting the subgrade type of the Washington SPS-2 project from fine-grained to coarse-grained. Further evaluation of the site data is needed to assess the significance of this finding.
For the Colorado site, the project was designed as a coarse-grained subgrade soil. However, four sections within the project were found to be constructed on fine-grained sandy clay soil. For the Nevada site, the project was designed as a coarse-grained soil. However, 9 out of the 11 sections were constructed on fine-grained sandy silt soil. Further evaluation is needed of the site data to assess the significance of this finding.
In the original SPS-2 experimental design, traffic was incorporated as a covariant. The traffic rate of at least 200,000 ESALs per year was required. The required annual ESAL and actual ESALs per year are compared in table 31. As shown, this requirement was met for most of the sites and years, with exceptions of the annual traffic for Iowa 1997. The annual ESAL data are not completely available at the time of analysis for five SPS-2 sites (38 percent). The wide range of traffic loadings between sites will need to be fully considered in any comparative analysis between sites.
The SPS-2 experimental design specifies two levels for concrete slab thickness: 203 mm and 279 mm. The SPS-2 construction guideline requires that the concrete slab thickness should be constructed within "6.4 mm. Many sections did not meet this guideline. Therefore, for practical reasons, "12.7 mm was used as the thickness tolerance or the design range. Table 32 compares designed versus constructed or measured mean PCC thicknesses from table TST_L05B. Thirty-
Note: Bolded numbers are outside the design required range.
8 of 143 SPS-2 sections (27 percent) fall outside of the design ranges (design value ±12.7 mm), with 4 sections having below-range values and 34 sections having above-range values. Twelve sections, all at the Wisconsin SPS-2 site, do not have thickness information in TST_L05B table at the time of analysis.
The frequency distributions of the tested slab thickness from table TST_L05B are provided in figure 4 for 203-mm design cells, and figure 5 for 279-mm design cells. As shown, 203-mm cell design sections have more scatter slab thickness distribution. Both the 203-mm and 279-mm mean thickness distribution shows a skew toward higher-than-designed thicknesses.
The SPS-2 experimental design specifies two levels for concrete flexural strength at 14 days: 3.8 MPa and 6.2 MPa. Table TST_PC09 was examined to compare the designed and constructed flexural strength values. The 14-day concrete flexural strength data were found for 11 SPS-2 sites; North Carolina and North Dakota sites' flexural strength information was not available at the time of analysis.
The design versus constructed SPS-2 PCC slab flexural strength comparison results are given in table 33. For the 3.8 MPa design cells, 7 of the 11 sites (64 percent) have average tested flexural strength values 10 percent outside of the design range, and 1 site's values were 20 percent outside of the design range. For the seven sites that are 10 percent outside of the design range, two sites are below the design range (3.8 MPa) and five are above. For the 6.2 MPa design cells, 5 of the 11 sites (45 percent) have average tested flexural strength 10 percent outside of the design range, and 3 sites' data were 20 percent outside of the design range. All of these five sites fall below the design value of 6.2 MPa.
The frequency distributions of the tested flexural strength values are provided in figure 6 for 3.8 MPa design cells, and in figure 7 for 6.2 MPa design cells. As shown, the distribution of the 3.8 MPa design cells is closer to a normal distribution, while the distribution of the 6.2 MPa design cells is very skewed to the right.
Field studies have shown that PCC continues to gain strength over many years. The 1 year strength data may be more indicative of the actual strength over the 20-year pavement evaluation period than the 20-day data. The differences in strength levels at 1 year are very important. Time-series plots were generated for concrete strength, as shown in figures 8 to 10. For most sites, the time-series plot of the concrete strength remains more or less parallel between 3.8 MPa and 6.2 MPa cells. The frequency distributions of the 1-year modulus of rupture values are shown in figures 11 and 12. They remain two distinct distributions with some overlay.
Statistical t-tests were performed on both the 14-day concrete strength and 1-year concrete strength, and the results are presented in table 34. Even though the mean difference of the strength measurements decreases from 1.71 MPa at 14 days to 1.24 MPa at 1 year, the strength differences between the lower and higher strength concrete were still very significant at 1 year specimen age. This finding indicates that overall concrete strength values of the 6.2 MPa cells are still significantly higher than those of the 3.8 MPa cells at 1 year of pavement age.
Figure 4. Frequency distribution of the mean PCC slab thickness for SPS-2 203-mm cells.
Figure 5. Frequency distribution of the mean PCC slab thickness for SPS-2 279-mm cells.
Figure 6. Frequency distribution of the 14-day modulus of rupture for SPS-2 3.8-MPa cells.
Figure 7. Frequency distribution of the 14-day modulus of rupture for SPS-2 6.2-MPa cells.
Figure 8. Time-series plot of modulus of rupture for SPS projects in Arizona, Arkansas, and Colorado.
Figure 9. Time-series plot of modulus of rupture for SPS projects in Delaware, Iowa, and Kansas.
Figure 10. Time-series plot of modulus of rupture for SPS projects in Nevada, Ohio, and Washington.
Figure 11. Frequency distribution of the 1-year modulus of rupture for 3.8-MPa cells.
Figure 12. Frequency distribution of the 1-year modulus of rupture for 6.2-MPa cells.
The concrete strength factor should be examined in the future analysis to determine if it affects pavement performance. The actual strength measurements (instead of the target strength levels) should be used in any analysis due to the variation in strength of any given section.
The following base types and thicknesses are specified in SPS-2 experiment design:
IMS table TST_L05B was used to compare the designed versus constructed base types and thicknesses. The base types were confirmed to be constructed as designed for all the sections with base type information. For the base thicknesses, "13-mm tolerance was used for the design ranges. The comparison results are provided in table 35. Twenty out of 131 SPS-2 sections (15 percent) have representative base thicknesses outside the design range, with 3 sections having base thicknesses below the design range and 17 above the design range.
Note: Bolded numbers are outside the design required range.
Note: Bolded numbers are outside the design required range.
Edge drains were required for SPS-2 sections with PATBs. IMS table SPS_GENERAL contains drainage information for SPS-2 sections. Records were found for 130 SPS-2 sections in this table, and drainage designations were found to be as designed for all the sections.
The SPS-2 experimental design specifies two levels for lane width: standard lane width of 3.66 m, and widened lane width of 4.27 m. The lane width information contained in IMS table SPS-GENERAL was examined for the designed versus constructed data, as shown in table 36. Three of the 131 SPS-2 sections (2 percent) have different lane width values from the design specifications (sections 19-0216 and 19-0219 in Iowa, and 26-0220 in Michigan).
The experimental design specifications and the actual construction data of the key experimental factors for the SPS-2 project sites are summarized in table 37. As shown in the table, most SPS-2 sections meet the experimental design criteria for the large majority of the design factors. Most deviations from the experimental design are found for the concrete slab thickness and 14-day flexural strength.
A summary of experimental specifications versus as-constructed data for each SPS-2 project is provided in table 38.
Of the 13 SPS-2 projects, only the Wisconsin SPS-2 project does not have enough data in the IMS database to be evaluated. Eight projects can be characterized as good to excellent when comparing designed versus constructed data, while the remaining four projects are considered poor to fair.
= Indicates as-constructed value meets as-designed criteria.
- = No data