U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Accelerating Innovation

FHWA Home / Accelerating Innovation / HFL

APPENDIX E: TUBE TESTING REPORT

DESCRIPTION OF TESTED SPECIMENS

For this study, 10 empty (i.e., no grout) and 16 grouted FRP pultruded tubes were evaluate. The shape of the cross section is shown in figure 15. Overall dimensions and thicknesses were measured in nine specimens; see table 21. It appears that variability of the dimensions is in the order of 0.01 to 0.06 inches.

Diagram. FRP tube cross section
Figure 15. Diagram. FRP tube cross section.

Table 21. Dimensions and thicknesses of FRP tubes (no grout).
Tube ID No. Weight
(lb)
Length
(inches)
t1
(inch)
t2
(inch)
t3
(inch)
t4
(inch)
t5
(inch)
t6
(inch)
t7
(inch)
W1
(inch)
W2
(inch)
H1
(inch)
#30 57.86 131.94 0.200 0.190 0.222 0.206 0.253 0.260 0.241 6.516 3.608 4.477
#31 57.86 131.91 0.200 0.195 0.230 0.208 0.251 0.252 0.248 6.519 3.620 4.478
#32 57.86 131.88 0.201 0.196 0.223 0.210 0.246 0.254 0.248 6.518 3.607 4.478
#33 58.08 132.09 0.201 0.201 0.224 0.207 0.252 0.253 0.241 6.521 3.594 4.482
#34 58.08 132.06 0.200 0.196 0.227 0.208 0.252 0.256 0.244 6.521 3.597 4.483
#35 57.75 131.84 0.200 0.196 0.231 0.210 0.245 0.256 0.245 6.515 3.599 4.477
#36 57.75 131.91 0.204 0.198 0.226 0.208 0.248 0.255 0.248 6.516 3.600 4.480
#37 57.86 131.91 0.203 0.195 0.223 0.207 0.248 0.254 0.250 6.518 3.604 4.482
#38 57.75 131.88 0.200 0.193 0.224 0.211 0.245 0.256 0.247 6.518 3.612 4.479
#00
(unlabeled)
56.87 131.41 0.203 0.197 0.230 0.196 0.242 0.255 0.239 6.519 3.566 4.474
 
Average 57.77 131.88 0.201 0.196 0.226 0.207 0.248 0.255 0.245 6.518 3.601 4.479
Max 58.08 132.09 0.204 0.201 0.231 0.211 0.253 0.260 0.250 6.521 3.620 4.483
Min 56.87 131.41 0.200 0.190 0.222 0.196 0.242 0.252 0.239 6.515 3.566 4.474
Range 1.210 0.68 0.004 0.012 0.009 0.016 0.012 0.008 0.012 0.006 0.054 0.008

Of the 16 grouted tubes, half were filled with a cementitious grout and the other half filled with an epoxy grout. Two different grouting configurations were explored: “wide” or “narrow” cells filled. Two different cross section positions were also evaluated: wide side up and wide side down. Figure 16 shows all four possible combinations. The specimens were also checked for dimensional stability, weight, and any visible defects. One specimen, #69, had a lower weight than similar specimens; the manufacturer indicated that it was partially filled with epoxy grout. Specimens were received on February 2, 2012.

Diagram. Grout configurations.
Figure 16. Diagram. Grout configurations.

Table 22. Details of FRP grouted tubes.
Tube ID No. Grout type Cell filled Weight(lbs) Length(in) Date filled cross-section check
Height H1 Width W1
#68-EW Epoxy 2.5 Wide 109.6 132.06 1/31/2012    
#69-EW Epoxy 2.5 Wide *91.6 132.00 Partially filled 4.462 6.523
#50-EW Epoxy 2.5 Wide 111.4 132.25 1/31/2012 4.470 6.522
#51-EW Epoxy 2.5 Wide 111.0 132.13 1/31/2012    
#52-EN Epoxy 2.5 Narrow 88.4 132.25 1/31/2012    
#53-EN Epoxy 2.5 Narrow 88.0 132.13 1/31/2012    
#54-EN Epoxy 2.5 Narrow 88.6 132.28 1/31/2012 4.478 6.513
#55-EN Epoxy 2.5 Narrow 89.2 132.13 1/31/2012 4.474 6.514
#56-CW cementitious Wide 104.2 132.19 1/20/2012    
#57-CW cementitious Wide 104.0 132.00 1/21/2012    
#58-CW cementitious Wide 103.8 132.00 1/22/2012 4.476 6.520
#59-CW cementitious Wide 104.2 132.06 1/23/2012 4.473 6.511
#60–CN cementitious Narrow 84.8 132.00 1/24/2012    
#61–CN cementitious Narrow 84.8 132.25 1/25/2012    
#62–CN cementitious Narrow 85.0 132.13 1/26/2012 4.476 6.522
#63–CN cementitious Narrow 85.0 132.19 1/27/2012 4.471 6.516

DESCRIPTION OF TEST SETUP

The FRP tubes were tested under three–point loading (hydraulic actuator at midspan). Steel rollers were used for support between a 10–foot span length. Vertical deformation was measured using one linear variable differential transformer (LVDT) and/or string pot. In some specimens, strain measurements were obtained from uniaxial and rosette strain gage configurations. Two load cells were used to collect load data (22–kip or 110–kip capacity). The 22–kip load cell was used in nondestructive tests to better monitor the low–range of load, whereas the 110–kip load cell was used when the tubes were tested up to failure. A steel plate (12 inches by 4 inches by 0.5 inches) was used under the hydraulic actuator to create a 4–inch–wide distributed load over the FRP specimen.

Photo and diagram. Testing setup.

Figure 17. Photo and diagram. Testing setup.

Photos. Load cells used for FRP testing     Photos. Load cells used for FRP testing
Figure 18. Photos. Load cells used for FRP testing.

TEST PROGRAM

FRP Tubes with No Grout

All 10 FRP tubes with no grout were tested using the test setup described above. Two of them were tested up to failure, and the remaining eight were loaded and unloaded in the elastic range with two different cross section positions (wide side up, WSU, and wide side down, WSD), to check for any changes in their elastic stiffness response. The FRP tube #38 was initially loaded elastically up to 2.5 kips, and then was loaded up to failure (WSU). Specimen #30 was instrumented with strain gages in three locations: top (SG1) and bottom (SG2) "flange" under point load, and shear span (SG 3–5) at 25.4 inches from point load. This specimen was tested to failure (WSU). Figure 19 shows the strain gages locations. Table 23 provides additional details on this test program.

Diagram. Strain gage locations for the FRP tube specimen #30.
Figure 19. Diagram. Strain gage locations for the FRP tube specimen #30.

Table 23. Test program–FRP tubes no grout.
Tube ID No. Instrumentation used in test Cross–section position Max or Failure Load (kips) Displacement at max or failure load (inch) Flexural stiffness k (lbs/inch) Tested date
#00 (unlabeled) LVDT & 22 kips load cell WSU 2 1.36 1419 11/22/2011
    WSD   1.38 1418 11/22/2011
#30 LVDT & 22 kips load cell &
 strain_pot & 2 strain gages & 1 strain rosette
WSU 5(fail)
crushing at top
3.72 1520 12/19/2011
#31 LVDT & 22 kips load cell WSU 2.5 1.64 1527 01/04/2012
    WSD   1.65 1546 02/02/2012
#32 LVDT & 22 kips load cell WSU 2.5 1.67 1525 01/04/2012
    WSD   1.67 1542 02/03/2012
#33 LVDT & 22 kips load cell WSU 2.5 1.65 1541 01/05/2012
    WSD   1.66 1547 02/03/2012
#34 LVDT & 22 kips load cell WSU 2.5 1.67 1528 02/03/2012
    WSD   1.64 1551 01/05/2012
#35 LVDT & 22 kips load cell WSU 2.5 1.68 1518 02/03/2012
    WSD   1.65 1534 01/05/2012
#36 LVDT & 22 kips load cell WSU 2.5 1.68 1522 02/03/2012
    WSD   1.64 1543 01/05/2012
#37 LVDT & 22 kips load cell WSU 2.5 1.67 1519 01/06/2012
    WSD   1.65 1539 01/06/2012
#38 LVDT & 22 kips load cell WSU 2.5 1.66 1529 02/03/2012
    WSD   1.66 1533 02/03/2012
  LVDT & 110 kips load cell WSU 4.5(fail)
crushing at top
3.49 1468 02/14/2012

 

The load–deflection behavior of FRP empty tubes is shown in figure 20. Results indicate that the flexural stiffness (k; calculated as the slope between 300 and 2,000 lb) of the FRP tubes #31–#38 is very similar: k =1533 lb/inch (on average) with a 1.2 percent variation. The flexural stiffness of the FRP tube #00 empty tube is 7 percent smaller: k=1419 lb/inch, possibly related to the smaller tube thickness (table 21 shows that the thickness of the walls, e.g. t4, t5, t7, of this tube is smaller than the other specimens). Flexural thickness values are tabulated in table 23.

Strain data from the test to failure of FRP tube #30 shows that the strain gage on the bottom “flange” (SG2) ha as a quite linear response up to failure load . By contrast SG1, near the top flange exhibited a nonlinear response beyond 4,000 lb of applied load. The strain gage rossette (SG3–5) displayed a very small magnitude of strain values, as expected.

Graphs. Load–deflection elastic response of FRP tubes with no grout (left); Load–deflection response up to failure (tubes #30 and #38) (right).     Graphs. Load–deflection elastic response of FRP tubes with no grout (left); Load–deflection response up to failure (tubes #30 and #38) (right).
Figure 20. Graphs. Load–deflection elastic response of FRP tubes with no grout (left); Load–deflection response up to failure (tubes #30 and #38) (right).

Graph. Load–strain response of FRP tube #30 (no grout, WSU).
Figure 21. Graph. Load–strain response of FRP tube #30 (no grout, WSU).

The FRP tubes with no grout (#30 and #38, both WSU) failed by crushing of the top flange under the point load. Figure 22 shows photographs of these specimens after failure.

Photos. Failure modes of FRP tubes with no grout. Photos. Failure modes of FRP tubes with no grout.
Specimen #30 Specimen #38

Figure 22. Photos. Failure modes of FRP tubes with no grout.

Grouted Tubes

Fifteen grouted tubes were tested using the same test setup described above; the FRP tube partially filled with grout (#69) was not tested. The tested specimens had two different grouts and two different grouting configurations. Table 24 shows the details of this test program, including position of the cross section, instrumentation used, failure loads, displacements and modes. Flexural stiffness was calculated between two load levels in the elastic range (prior to first crack). Because each specimen’s load range is different, this information has been included in table 24.

Only four specimens were tested to failure (FRP tubes #50, 52, 57, and 60). The remaining specimens were loaded and unloaded in the elastic range. Figure 23 shows the load-deflection curves of all FRP tubes (with and without grout) tested to failure. It should be noted that the grouted tubes had a higher flexural stiffness than the empty tubes. Epoxy grouted specimens had a higher flexural stiffness than cementitious grouted specimens.

Table 24 shows that the elastic flexural stiffness of epoxy grouted tubes is 34 percent larger, on average, than that of the non–grouted tubes, while the flexural stiffness of cement-based grouted tubes is only 18 percent larger than that of the non–grouted tubes. When the cementitious grout is in compression, the flexural stiffness of the grouted tube is 17 percent larger than that of the grouted tube where the cementitious-based grout is in tension, which in turn has a very small increment of stiffness over the empty tube condition. However, for epoxy grouted tubes, the elastic flexural stiffness has little variation (only 4 percent) whether the grout is subjected to compression or tension. Comparing the effect of grouting either the wide or narrow cell, the epoxy grouted tubes have a 9 percent larger elastic flexural stiffness when the narrow cell is filled (as opposed to the wide cell); the cement-based grouted tubes show little difference.

Specimen #56 had the lowest flexural stiffness (1,600 lb/in.). Analysis of the load-deflection curve indicates that the composite action in this tube vanishes at a very small load, and the specimen behaves as an empty tube. Figure 24 shows a comparison of the load-deflection behavior (in the elastic range) for specimen #56, #57 (similar configuration), and #31 (empty tube).

Table 24. Test program–grouted FRP tubes.
Tube ID No. Instrumentation used Cross–section position Max or failure Load(kips) Deflection at max or failure load (inches) Elastic stiffness (lbs/inch) [load range, lbs] Load and deflection at first crack (kips/inches) Tested date
#68–EW String pot &
 22 kips load cell
WSD 3.5 2.05 1926
[300–2550]
2.55/1.319 03/02/2012
#50–EW String pot &
 110 kips load cell
WSU 9(fail)
Failure at bottom
4.92 1920
[300–4500]
No crack sound heard 02/24/2012
#51–EW LVDT &
 22 kips load cell & 2 strain gages &
1 strain rosette
WSU 7.3 3.89 1993
[300–3500]
No crack sound heard 03/03/2012
#52–EN String pot &
 110 kips load cell
WSD 9.6(fail)
Failure at bottom
4.77 2234
[300–4800]
No crack sound heard 02/24/2012
#53–EN LVDT &
 22 kips load cell & 2 strain gages
WSD 7.8 3.84 2182
[300–3900]
No crack sound heard 03/03/2012
#54–EN String pot &
 22 kips load cell
WSU 3.8 2.41 2048
[300–1590]
1.59/0.756 03/02/2012
#55–EN String pot &
 22 kips load cell
WSU 1.8 0.92 2045
[300–1750]
1.75/0.838 03/02/2012
#56–CW LVDT &
 22 kips load cell & 2 strain gages &
1 strain rosette
WSU 5.9 3.89 1600
[300–3000]
No crack sound heard 03/03/2012
#57–CW LVDT &
 22 kips load cell
WSU 2.5 1.42 1970
[300–2000]
2/1.015 02/03/2012
WSU 1.68   0.4/0.242
LVDT &
 110 kips load cell
WSU 6.44(fail)
Failure at bottom
4.39     02/14/2012
#58–CW String pot &
 22 kips load cell
WSD 3.8 2.58 1728
[30–400]
0.4/0.221 03/02/2012
#59–CW LVDT &
 22 kips load cell
WSD 1.0 0.68 1687
[30–420]
0.42/0.244 03/02/2012
#60–CN String pot &
 110 kips load cell
WSD 7.5(fail)
Failure at bottom
4.60 2103
[300–3000]
3/1.431 02/24/2012
#61–CN LVDT &
 22 kips load cell & 2 strain gages
WSD 8.0 5.39 1850
[300–3000]
5.98/3.654 03/03/2012
#62–CN String pot &
 22 kips load cell
WSU 3.8 2.63 1646
[30–300]
0.3/0.174 03/02/2012
#63–CN LVDT &
 22 kips load cell
WSU 1.0 0.66 1714
[30–360]
0.36/0.205 03/02/2012

Graph. Load-deflection behavior up to failure for grouted FRP tubes.
Figure 23. Graph. Load-deflection behavior up to failure for grouted FRP tubes.

Graph. Load-deflection curve (up to 2,500 lb) of specimens #56, 57, and #31.
Figure 24. Graph. Load-deflection curve (up to 2,500 lb) of specimens #56, 57, and #31.

Four specimens were instrumented with strain gages. FRP tubes #53 and #61 had one strain gage at the top (SG1) and one at the bottom (SG2). FRP tubes #51 and #56 had, in addition to SG1 and SG2, a strain gage rosette (SG3-5) at 25.4 inches from the point load. Figure 19 shows locations of all the strain gages. Load-strain plots are show in figure 25. The two strain gage rosettes (FRP tubes #51 and #56) failed during testing, perhaps due to insufficient curing time of the adhesive used prior to test. Strain responses of these specimens confirm the load-deflection behavior, analyzed previously, showing that grouted tubes have a larger section modulus when the narrow cell is filled and subjected to compression. Epoxy grouted tubes have a larger section modulus than cement-based grouted tubes.

. Graphs. Load-strain responses of grouted FRP tubes. . Graphs. Load-strain responses of grouted FRP tubes.
Specimen #51-EW-WSU Specimen #53-EN-WSD
. Graphs. Load-strain responses of grouted FRP tubes. . Graphs. Load-strain responses of grouted FRP tubes.
Specimen #56-CW-WSU Specimen #61-CN-WSD

Figure 25. Graphs. Load-strain responses of grouted FRP tubes.

Failure mode for the grouted tubes originated with a tensile rupture of the bottom flange, regardless of the location of the grout (wide or narrow cell in compression). Cracks propagated vertically for all specimens, and also horizontally in specimen #50 (the location of this horizontal crack is 0.5 inches from the bottom, apparently at the splice between wall tube and the bottom horizontal section of the FRP tube). Figure 26 shows photographs of the damage regions for all four tested specimens.

. Photos. Photographs of failure modes of grouted FRP tubes. . Photos. Photographs of failure modes of grouted FRP tubes.
Specimen #50- EW-WSU Specimen #52-EN-WSD
. Photos. Photographs of failure modes of grouted FRP tubes. . Photos. Photographs of failure modes of grouted FRP tubes.
Specimen #57-CW-WSU Specimen #60-CN-WSD

Figure 26. Photos. Photographs of failure modes of grouted FRP tubes.

Specimen #60 also showed slip of the cementitious grout (top) during testing at one of the ends of the specimen (the other side was covered by the manufacturer with tape), indicating that the bond between the grout and the FRP tube had failed. The load-deflection curve has a significant load drop around 3,000 lb, probably related to the debonding initiation. Load-strain curves for a cementitious-grouted specimen also registered a significant load drop, in this case around 6,000 lb.

Photos. Cementitious grout slipping at the end of FRP tube #60. Photos. Cementitious grout slipping at the end of FRP tube #60.

Figure 27. Photos. Cementitious grout slipping at the end of FRP tube #60.

After testing to failure, specimens #50 and #57 were cut using a wet cutting block saw near the damaged region (around the center span) to look in detail at the debonded region. Inspection of the cut surfaces shows that debonding between the epoxy grout and the FRP tube (specimen #50) occurred along the lower interface, whereas cracks between the cementitious grout and the FRP tube (specimen #57) were present along the top interface, as shown in figure 28. Specimen #37 also had a vertical crack inside the cementitious grout, a possible indication of additional damage on this grouted region.

Photos. Cross sections of grouted tubes near midspan, after failure. Photos. Cross sections of grouted tubes near midspan, after failure.
Specimen #50- EW-WSU Specimen #57-CW-WSU

Figure 28. Photos. Cross sections of grouted tubes near midspan, after failure.

CONCLUSIONS

Based on the experimental program developed, the following conclusions can be drawn:

Page last modified on May 18, 2012.
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000