Skip to contentUnited States Department of Transportation - Federal Highway Administration FHWA Home
Research Home   |   Pavements Home
Report
This report is an archived publication and may contain dated technical, contact, and link information
Publication Number: FHWA-RD-02-088
Date: May 2003

Evaluation of Joint and Crack Load Transfer Final Report

Previous | Table of Contents | Next

The FWD data collected for the SMP LTPP sections allow for analysis of the effect of season and time of day on joint LTE.  Analysis of LTE was performed for 19 SMP sections, with a total of 1,945 FWD passes.  The number of passes is much higher than the number of sections because the sections could be tested several times a day and several times per year. The deflection data were downloaded during the summer of 2001 from the LTPP database table MON_DEFL_DROP_DATA.  Information about sensor locations was obtained from database tables MON_DEFL_LOC_INFO and MON_DEFL_DEV_SENSORS (June 2001 release). The procedure described in chapter 3 was used to calculate LTE from the FWD deflection basins and determine representative LTE indexes for individual joints and for each FWD pass.

Effects of Time of Testing on Joint LTE

The collected FWD data allow for analysis of the effect of time of day on measured joint LTE.  Typically, several passes are conducted on SMP sections to study the variations that may occur over a single day.  Almost all sections showed dependence of calculated LTE on the time of the testing.  Figures 56 and 57 show the LTEs from approach (J4) and leave (J5) tests, respectively, for nondoweled section 163023 (Idaho) obtained from the three FWD passes conducted in September 1992 at 10:45 a.m., 1:10 p.m., and 3:15 p.m.  The results show great variation over the course of the day.  The lowest LTE values came from the first FWD pass for that day, when the calculated LTEs for several joints are approximately three times lower than from the 3:15 p.m. testing.  Figures 58 and 59 show daily variation in LTEs for doweled section 040215 (Arizona) in March 1996.  The lowest level of LTE for this section was measured at 10 a.m., whereas the highest one was measured after 2 p.m.

Similar effects were observed for many sections located in different climatic regions.  Figures 60 and 61 present comparisons of mean LTEs for the doweled and nondoweled SMP sections, obtained for the same day from the first and third FWD pass deflection data.  Significant differences in mean LTEs from different FWD passes were observed for many doweled and nondoweled sections.  Moreover, the nondoweled sections exhibited much higher variability in LTE than the doweled sections for both approach and leave tests.

To investigate daily variation in LTE further, coefficients of variation of the mean section LTE for the SMP LTPP sections from different FWD passes made on the same day of testing were computed.  Table 12 presents mean and maximum values of these coefficients of variation. 

The mean coefficient of variation of mean LTE was found to be less than 10 percent for all doweled SMP sections for both approach and leave tests.  However, of the 12 SMP doweled sections for which multiple FWD passes were available, a coefficient of variation in mean LTE greater than 10 percent was observed on at least 1 day of testing for 4 sections for test approach (J4) and for 5 sections for leave (J5) test.

1 m = 3.28 ft

Figure 56. Daily variation in calculated approach LTE, section 163023 (September 1992). Station, 130 to 170 meters, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the vertical axis. The figure is a scatter plot with three sites, which are falling weight deflectometer passes, conducted at three different times of the day. The sites are, 11:01 AM, TA equals 81, TS equals 73; 1:09 PM, TA equals 86, TS equals 90; and 3:02 PM, TA equals 86, TS equals 97. The first pass conducted has the lowest load transfer efficiency at 21 to 28 percent load transfer efficiency. The figure shows great variation of results throughout the day.

Figure 56. Daily variation in calculated approach LTE, section 163023 (September 1992).

1 m = 3.28 ft

Figure 57. Daily variation in calculated leave LTE, section 163023 (October 1992). Station, of 130 to 170 meters, is graphed on the horizontal axis. The load transfer efficiency, percent, is graphed on the vertical axis. The figure has three sites, which are falling weight deflectometer passes conducted at different times of the day; 10:44 AM, TA equals 75, TS equals 63; 1:12 PM, TA equals 77, TS equals 75; 3:14 PM, TA equals 82, TS equals 81. The figure shows great variation of results throughout the day. The first pass at 10:44 AM has the lowest load transfer efficiency from 25 to 42 percent of 135 to 164 meters.

Figure 57.  Daily variation in calculated leave LTE, section 163023 (October 1992).

1 m = 3.28 ft

Figure 58. Daily variation in calculated approach LTE, section 4_0215 (March 1996). Station, of 130 to 160 meters, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the vertical axis. The figure is a scatter plot with four sites that are falling weight deflectometer passes conducted at different times of the day; 10:05 AM, TA equals 72, TS equals 77; 11:36 AM, TA equals 77, TS equals 92; 1:03 PM, TA equals 86, TS equals 97; and 2:37 PM, TA equals 88, TS equals 102. The figure shows great variation of results throughout the day. The first pass has the lowest load transfer efficiency measured from 42 percent at 140 meters. The highest load transfer efficiency was measured after 2 PM.

Figure 58.  Daily variation in calculated approach LTE, section 4_0215 (March 1996).

1 m = 3.28 ft

Figure 59. Daily variation in calculated leave LTE, section, 4_0215 (March 1996). Station, of 130 to 170 meters, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the vertical axis. The figure is a scatter plot with four sites that are falling weight deflectometer passes conducted throughout the day. The times of the passes are 10:07 AM, 11:39 AM, 1:06 PM, and 2:39 PM. The results show great variation of results throughout the day. The first pass at 10 AM has the lowest load transfer efficiency of 48 to 68 percent from 132 to 153 meters. The highest measured load transfer efficiency was after 2 PM, from 81 to 92 percent.

Figure 59.  Daily variation in calculated leave LTE, section 4_0215 (March 1996).

Figure 60. Comparison of mean LTEs for doweled SMP sections from different FWD passes on the same day of testing. Pass 1 load transfer efficiency, of 10 to 110 percent, is graphed on the horizontal axis. Pass 3 load transfer efficiency, of 10 to 110 percent, is graphed on the vertical axis. There is a slope increasing at a 45-degree angle. The slope is pass 1 load transfer efficiency equals pass 3 load transfer efficiency. There are two sites that are approach tests and leave tests. Both approach test and leave test increase in pass 3 as pass 1 increase. The sites are clustered along the slope between 60 to 95 percent at pass 1 and 3. There is significance in mean load transfer efficiencies for doweled seasonal monitoring program sections.

Figure 60. Comparison of mean LTEs for doweled SMP sections from different FWD passes on the same day of testing.

Figure 61. Comparison of mean LTEs for nondoweled SMP sections from different FWD passes on the same day of testing. Pass 1 load transfer efficiency, of 10 to 110 percent, is graphed on the horizontal axis. Passes 3 load transfer efficiency, of 10 to 110 percent, is graphed on the vertical axis. There is a slope increasing at a 45-degree angle. The slope is pass 1 load transfer efficiency equals pass 3 load transfer efficiency. The figure has 2 sites, which are approach tests and leave tests. Both approach and leave test increase in pass 1 as pass 3 increases. The sites are all scattered above the slope. There is a significant correlation in mean load transfer efficiencies for nondoweled seasonal monitoring program. The nondoweled section showed higher variability than doweled sections for both approach and leave test.

Figure 61.   Comparison of mean LTEs for nondoweled SMP sections from different FWD passes on the same day of testing.

Table 12.  Coefficients of variation of the section mean LTEs from the same day of testing.

Section

State

Type
Joint

Approach (J4)

Leave (J5)

Mean

Max

Mean

Max

040215

Arizona

doweled

0.088

0.181

0.070

0.192

133019

Georgia

doweled

0.088

0.267

0.041

0.110

183002

Indiana

doweled

0.084

0.147

0.080

0.156

274040

Minnesota

doweled

0.062

0.104

0.063

0.102

320204

Nevada

doweled

0.025

0.047

0.019

0.034

364018

New York

doweled

0.031

0.058

0.043

0.079

370201

North Carolina

doweled

0.041

0.076

0.103

0.103

390204

Ohio

doweled

0.010

0.013

0.010

0.018

421606

Pennsylvania

doweled

0.025

0.060

0.030

0.059

484142

Texas

doweled

0.009

0.019

0.011

0.032

484143

Texas

doweled

0.022

0.060

0.027

0.070

893015

Quebec

doweled

0.007

0.012

0.007

0.015

063042

California

nondoweled

0.063

0.142

0.086

0.173

163023

Idaho

nondoweled

0.250

0.470

0.147

0.412

204054

Kansas

nondoweled

0.057

0.152

0.068

0.160

313018

Nebraska

nondoweled

0.243

0.478

0.307

0.447

493011

Utah

nondoweled

0.259

0.538

0.136

0.455

533813

Washington

nondoweled

0.031

0.045

0.022

0.032

833802

Manitoba

nondoweled

0.213

0.213

0.102

0.102

Much higher variability was observed for nondoweled sections.  Only three of seven nondoweled SMP sections showed a mean coefficient of variation less than 10 percent.  The remaining four sections exhibited a mean coefficient of variation greater than 20 percent for mean LTEs calculated from approach tests.  Three sections (163023, 313018, and 493011) exhibited a coefficient of variation in mean LTE greater than 40 percent on at least one day of testing for both approach and leave tests.  Only one section, 533813, did not exhibit significant daily variability in LTE.  Significant daily variability in subgrade k-values backcalculated from FWD center slab deflections (test J1) for this section was reported in another study by Khazanovich et al. (2001).

It is reasonable to conclude, therefore, that the time of day of FWD testing may significantly affect joint LTE. This effect is most likely due to temperature differences and the resulting joint movement and slab curling.  Therefore, accounting for the effect of PCC slab curling is very important for reliable interpretation of FWD deflection data.  The effects of temperature conditions and joint opening on measured LTE will be discussed further in chapter 7. Nevertheless, the results presented above suggest that it is important to conduct FWD basin testing early in the morning to measure the lowest level of LTE, or during cold times of the year, when temperature remains constant all day.

Effects of Season of Testing on Joint LTE

As expected, the season in which testing was performed was found to affect measured joint LTE.  This effect is highly confounded, however, with the time of testing.  To reduce the latter effect, only first FWD passes for each day of testing were considered in the analysis of seasonal variation in joint LTE.

For each SMP section, mean LTEs obtained from different days of testing for the first FWD pass on each day were analyzed.  Mean, maximum, and minimum values of those LTEs, as well as corresponding standard deviations and coefficients of variation, were determined.  The results of that analysis for doweled sections for FWD for approach and leave test are summarized in tables 13 and 14, respectively.  Tables 15 and 16 present corresponding results for nondoweled sections.

Significant seasonal variability in LTE, as well as between the LTEs from approach and leave tests, was observed for both doweled and nondoweled pavements.  Analysis of tables 13 and 14 shows that every doweled SMP sections exhibited mean LTE of greater than 90 percent for at least one day of testing.  At the same time, six sections exhibited an LTE lower than 60 percent (and four lower than 50 percent) from approach tests on at least one day of testing.  Although slightly lower variability in LTE was observed for leave tests of doweled sections, five sections exhibited lower than 60 percent LTE on at least one day of testing.

Table 13.  Seasonal variation of approach LTE (test J4) for doweled SMP sections.

Section

Mean LTE, percent

Minimum LTE, percent

Maximum LTE, percent

Standard Deviation, percent

Coefficient of Variation

40215

75.2

45.4

96.2

 13.6

0.18

133019

80.3

45.5

95.8

 15.2

0.19

183002

81.4

64.3

93.5

 10.5

0.13

274040

76.8

57.7

94.5

9.9

0.13

320204

88.4

80.4

98.6

  5.9

0.07

364018

80.6

59.9

91.5

  6.5

0.08

370201

72.1

41.8

96.8

19.3

0.27

390204

93.3

87.6

98.2

  2.7

0.03

421606

73.0

39.1

94.0

21.5

0.29

484142

90.3

83.1

97.0

  2.8

0.03

484143

89.1

79.0

96.2

  5.2

0.06

893015

90.5

76.6

102.7

  4.6

0.05


Table 14.  Seasonal variation of leave LTE (test J5) for doweled SMP sections.

Section

Mean LTE, percent

Minimum LTE, percent

Maximum LTE, percent

Standard Deviation, percent

Coefficient of Variation

40215

75.1

54.4

91.2

12.0

0.16

133019

84.6

56.1

94.8

 9.6

0.11

183002

82.6

64.6

93.4

10.6

0.13

274040

74.5

52.4

92.1

10.9

0.15

320204

88.8

84.0

96.0

 4.2

0.05

364018

78.4

62.5

92.0

 6.7

0.09

370201

79.3

58.4

96.6

13.8

0.17

390204

93.3

88.0

96.8

 3.0

0.03

421606

78.7

52.2

92.0

13.6

0.17

484142

90.8

84.8

95.4

 2.6

0.03

484143

88.9

78.6

98.1

 6.5

0.07

893015

90.4

81.9

95.4

 3.6

0.04


 Table 15. Seasonal variation of approach LTE (test J4) for nondoweled SMP sections.

Section

Mean LTE, percent

Minimum LTE, percent

Maximum LTE, percent

Standard Deviation, percent

Coefficient of Variation

63042

64.4

27.6

96.6

22.5

0.35

163023

51.1

24.9

91.8

20.8

0.41

204054

74.3

53.7

90.1

14.1

0.19

313018

38.1

23.8

85.8

15.8

0.41

493011

41.5

16.7

92.7

23.0

0.56

533813

72.7

37.6

95.6

18.0

0.25

833802

58.7

18.6

96.9

26.9

0.46


Table 16.  Seasonal variation of leave LTE (test J5) for nondoweled SMP sections.

Section

Mean LTE, percent

Minimum LTE, percent

Maximum LTE, percent

Standard Deviation, percent

Coefficient of Variation

63042

64.4

27.6

96.6

22.5

0.35

163023

51.1

24.9

91.8

20.8

0.41

204054

74.3

53.7

90.1

14.1

0.19

313018

38.1

23.8

85.8

15.8

0.41

493011

41.5

16.7

92.7

23.0

0.56

533813

72.7

37.6

95.6

18.0

0.25

833802

58.7

18.6

96.9

26.9

0.46

As expected, seasonal variability is much higher for nondoweled sections. Analysis of tables 15 and 16 shows that every nondoweled SMP section exhibited mean LTE greater than 85 percent for at least one day of testing.  In addition, of the 7 nondoweled sections, 5 and 4 exhibited lower than 30 percent LTE from approach and leave tests, respectively, on at least 1 day of testing.  Therefore, the same section may be ranked as having very good or very poor LTE, depending on the day the testing was performed.

Figures 62 through 68 show mean LTEs and mean PCC surface temperature (for the duration of a FWD pass) obtained from different days of testing for the first FWD pass on each day.  Analysis of these figures shows that, as a general trend, LTE increases when the PCC surface temperature increases.  However, when PCC temperature drops below the freezing temperature of water (0 ºC [32 ºF]), LTE may increase dramatically (see figure 63). 

It may be concluded that joint LTE depends significantly on the season and time of testing.  Therefore, the results of LTE measurements taken in hot summer weather are not useful.  Also, the poor correlations between design features and site conditions presented in chapter 4 may be explained by the fact that GPS sections were tested under a wide variety of climatic conditions.

F = 1.8C + 32

Figure 62. Seasonal variation in LTE and PCC surface temperature, section 63042. The date, from 1995 to 1997, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the left vertical axis. Temperature, of 0 to 50 degrees Celsius, is graphed on the right vertical axis. The figure has 3 sites measuring falling weight deflectometer passes; approach load transfer efficiency, leave load transfer efficiency, and surface temperature. Both load transfer efficiencies and surface temperature begin in high temperatures in1995 (35 to 45 degrees Celsius) and then they all decrease in mid-1995. In 1996, the leave load transfer efficiency stays below 60 percent load transfer efficiency and increases half way through the year, the approach load transfer efficiency fluctuates, and the surface temperature starts to increase. Generally, when the load transfer efficiency increases, the portland cement concrete surface temperature also increases. The load transfer efficiencies depends significantly on the season and the time of testing

Figure 62.  Seasonal variation in LTE and PCC surface temperature, section 63042.

F = 1.8C + 32

Figure 63. Seasonal variation in LTE and PCC surface temperature, section 163023. The date, from 1991 to 1994, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the left vertical axis. Temperature, from negative 10 to 40 degrees Celsius, is graphed on the right vertical axis. The figure has three sites; approach load transfer efficiency, leave load transfer efficiency, and surface temperature.  In the beginning of 1992, all three sites increase in temperature and start to decrease in the summer of that year. The surface temperature decreases to negative 4 degrees Celsius and at the same time the approach and leave load transfer efficiencies increase drastically. The general trend is when load transfer efficiency increase, then the surface temperature also increases, when the load transfer efficiency decreases then the temperature decreases. However, when the portland cement concrete surface temperature drops below freezing (0 degrees Celsius), then both load and approach load transfer efficiencies increase significantly.

Figure 63.  Seasonal variation in LTE and PCC surface temperature, section 163023.

F = 1.8C + 32

Figure 64. Seasonal variation in LTE and PCC surface temperature, section 204054. The date, from 1993 to 1996, is graphed on the horizontal axis. Load transfer efficiency, percent, is graphed on the left vertical axis. Temperature, of 0 to 50 degree Celsius, is graphed on the right vertical axis. The figure has three sites; approach load transfer efficiency, leave load transfer efficiency, and surface temperature. Beginning in mid-1993, both load transfer efficiencies start at 90 percent load transfer efficiency and 45 degrees. Surface temperature is at its highest temperature (30) at 60 percent. All three stay level until mid-1994 and then decrease until 1996. There is significance between approach and leave load transfer efficiency and surface temperature. Load transfer efficiencies changes when the portland cement concrete surface temperature changes.

Figure 64.  Seasonal variation in LTE and PCC surface temperature, section 204054.

F = 1.8C + 32

Figure 65. Seasonal variation in LTE and PCC surface temperature, section 313018. The date, from 1998 to 2001, is graphed on the horizontal axis. Load transfer efficiency, in percent, is graphed on the left vertical axis. Temperature, from negative 10 to 40 degree Celsius, is graphed on the right vertical axis. The figure has 3 sites, which are approach load transfer efficiency, leave load transfer efficiency, and surface temperature. Both load transfer efficiencies change when the surface temperature change. In 1998, all three sites start to increase in temperature from 33 to 35 degrees Celsius and decrease in mid-1998. In 1999, the three sites stay level and increase again in 2000. Load transfer efficiency increase, then the surface temperature also increases, when the load transfer efficiency decreases then the temperature decreases.

Figure 65.  Seasonal variation in LTE and PCC surface temperature, section 313018.

F = 1.8C + 32

Figure 66. Seasonal variation in LTE and PCC surface temperature, section 493011. The date, from 1993 to 1996, is graphed on the horizontal axis. The load transfer efficiency, in percent, is graphed on the left vertical axis. The temperature, from negative 10 to 40 degrees Celsius, is graphed on the right vertical axis. The figure has three sites; approach load transfer efficiency, leave load transfer efficiency, and surface temperature. In mid-1993, both load transfer efficiencies and the surface temperature begin to decrease in temperature beginning from 37 degrees Celsius. In 1994, the sites start to increase from 2 to 4 degrees until the middle of the year, then decrease again. In 1995, the surface temperature decreases to negative 3 degrees Celsius, the load transfer efficiencies increase to 13 degrees, and leave increases to 24 degrees. When load transfer efficiency increases, then the surface temperature also increases; when the load transfer efficiency decreases, then the temperature decreases. However, when the surface temperature drops below freezing, then the load transfer efficiencies increase dramatically.

Figure 66.  Seasonal variation in LTE and PCC surface temperature, section 493011.

F = 1.8C + 32

Figure 67. Seasonal variation in LTE and PCC surface temperature, section 533813. The date, from 1995 to 1997, is graphed on the horizontal axis. The load transfer efficiency, percent, is graphed on the left vertical axis. The temperature, from 0 to 50 degrees Celsius, is graphed on the right vertical axis. The figure has three sites; approach load transfer efficiency, leave load transfer efficiency, and surface temperature. All three sites begin at a high temperature in the middle of 1995. At the end of the year, the surface temperature drops to 7 degrees. The leave only drops to 39 degrees and the approach drops to 24 degrees that year. In spring of 1996, the surface temperature increases to 21 degrees while both load transfer efficiencies increase to 43 degrees. The surface temperature has a lower load transfer efficiency than approach or leave load transfer efficiency. When load transfer efficiency increases, then the surface temperature also increases; when the load transfer efficiency decreases, then the temperature decreases.

Figure 67.  Seasonal variation in LTE and PCC surface temperature, section 533813.

F = 1.8C + 32

Figure 68. Seasonal variation in LTE and PCC surface temperature, section 833802. The date, from 1993 to 1995, is graphed on the horizontal axis. The load transfer efficiency, percent, is graphed on the left vertical axis. The temperature, from negative 15 to 35 degrees Celsius, is graphed on the right vertical axis. The figure has three sites; approach load transfer efficiency, leave load transfer efficiency and surface temperature. All three sites begin in mid-1993 at 27 to 32 degrees Celsius. The sites decrease into the end of the year to negative 6 to 2 degrees. The surface temperature decreases further to negative 13 degrees in 1994, while the load transfer efficiencies increase to 30 degrees. In the spring the load transfer efficiencies decrease to negative 5 degrees and 4 degrees. After spring, all three sites increase in temperature. The figure shows that when the load transfer efficiency increases, then the surface temperature also increases; when the load transfer efficiency decreases, then the temperature decreases. However, when the surface temperature decreases below freezing, then the load transfer efficiency increases dramatically.

Figure 68.  Seasonal variation in LTE and PCC surface temperature, section 833802.

Previous | Table of Contents | Next


The Federal Highway Administration (FHWA) is a part of the U.S. Department of Transportation and is headquartered in Washington, D.C., with field offices across the United States. is a major agency of the U.S. Department of Transportation (DOT).
The Federal Highway Administration (FHWA) is a part of the U.S. Department of Transportation and is headquartered in Washington, D.C., with field offices across the United States. is a major agency of the U.S. Department of Transportation (DOT). Provide leadership and technology for the delivery of long life pavements that meet our customers needs and are safe, cost effective, and can be effectively maintained. Federal Highway Administration's (FHWA) R&T Web site portal, which provides access to or information about the Agency’s R&T program, projects, partnerships, publications, and results.
FHWA
United States Department of Transportation - Federal Highway Administration