CHAPTER II.
RESEARCH METHODS AND MATERIALS
2.1 Evaluation of Waterproofing Systems
Two different waterproof bituminous concrete systems were examined in this research study. The first system, consistent with current timber bridge construction practice, uses a preformed waterproofing membrane placed on a preservative treated wood deck and overlaid with a bituminous concrete wearing surface (Figure-2.1). The second system consists of a treated wood deck overlaid with a base course of bituminous concrete, a waterproofing membrane, and finally, a bituminous concrete wearing surface (Figure-2.2). The second system was considered because, in spite of the use of dense graded surface asphalt concrete mixtures, the current practice of placing a waterproofing membrane directly on top of a flat and level bridge deck may lead to stripping of the asphalt cement from the aggregate due to the buildup of moisture in the asphalt concrete directly above the membrane. In addition, any water that is allowed to remain stagnant above the membrane for an extended period of time will be more prone to penetrating any
membrane weaknesses and thereby potentially reduce the functional service life of the wood deck and other bridge components. The bituminous concrete base course can also be used to construct a deck cross-slope into the pavement system at the level of the waterproofing membrane to allow any water that penetrates the wearing surface to flow to the edges of the bridge deck and off the bridge without wetting other bridge members.
Figure-2.2 Proposed Timber Bridge Pavement Configuration
The testing regime used in this research to evaluate watertightness and bond performance incorporated three parameters: three waterproofing membranes, two wood preservative
treatments, and two environmental degradation conditions induced by temperature cycling in a moisture saturated condition. Control groups were also evaluated for each study parameter. Duplicate specimens were prepared and tested for each of the study parameters. A total of 160 specimens were constructed and tested as shown in the study matrix (Figure-2.3).
Watertightness of each waterproof asphalt wearing surface system was determined by measuring the electrical impedance across a test specimen perpendicular to the direction of bond orientation in the pavement materials. The bond strength between each material of both paving systems was assessed using a shear test apparatus designed and built for this study. Using a constant deformation rate, load was applied to one side of the system bond line while the side opposite was held in a fixed position. Maximum sustained load by the system was recorded for each specimen along with the location of failure relative to the bond line.
In addition to the laboratory constructed specimens, three drilled cores were taken from a bridge deck located on Creekside Drive in East Pennsboro Township, Pennsylvania. The deck was constructed using the new system design proposed in this research using two asphalt mixtures and a waterproofing membrane. The asphalt mixtures, membrane, and wood species incorporated
into this bridge were different from those included in the laboratory constructed specimens. At the time the cores were taken, the bridge was approximately one year old.
2.2 Specimen Design and Materials
Specimen design was based on the geometry of specimens used in ASTM D 1559
"Resistance to Plastic Flow of Bituminous Mixtures Using Marshall Apparatus". In this test, hot mix asphalt is compacted into the shape of a 4 inch diameter by 3 inch thick cylinder using a steel mold and a standardized compaction hammer. For the specimens in this research, this basic shape was subdivided into two distinct and separate cylindrical sections 1.5 inches thick to allow for the use of different materials in each lift with a waterproofing membrane placed in between (Figure-2.4). By using this specimen configuration, the bond strength and watertightness of an entire asphalt concrete wearing system was examined.
Figure 2.4 Lab Constructed Test Specimen
To examine current paving practice for timber bridges, a group of specimens was
constructed with a wood bottom section and a compacted surface hot-mix asphalt concrete in the top section.
All wood was machined from quarter sawn nominal "2-by" southern pine lumber that was either left untreated, treated with creosote, or treated with pentachlorophenol. Untreated wood was used to establish a baseline for future research using other treatment methods, though the performance of CCA treated lumber is likely to be very similar. It should be stressed that the use of untreated wood in an actual field application is not recommended. Minimum preservative treatment retention levels were specified as 12lb/ft3 for creosote and 0.60lb/ft3 for
treatment parameters for the treated boards are included in Appendix-A. In the test specimens, quarter sawn lumber was chosen to simulate the grain orientation observed at the paving surface of typical glulam deck panels.
The surface asphalt concrete mixture, consisting of asphalt cement and crushed limestone aggregate no larger than 0.5 inches, was manufactured to Virginia Department of Transportation specifications for mixture type SM-2AL at a local batch plant. A complete mix design is included in Appendix-B. In addition, an independent analysis of the asphalt concrete used in this study was performed by the Virginia Transportation Research Council. A copy of this analysis is also included in Appendix-C. This mixture is commonly used for secondary roads in Virginia which is where the timber bridges considered in this study are typically located.
Test specimens for the waterproof asphalt concrete wearing surface system proposed in this research study consisted of a compacted base hot-mix asphalt concrete in the bottom cylindrical section and compacted surface hot-mix asphalt concrete in the top section.
The base asphalt concrete mixture, consisting of asphalt cement and crushed limestone aggregate no larger than 1 inch, was manufactured to Virginia Department of Transportation specifications for mixture type BM-2 at a local batch plant. The mixture proportions are included in Appendix-B. In addition, an independent analysis of the asphalt concrete used in this study was performed by the Virginia Transportation Research Council. A copy of this analysis is included in Appendix-C. This mixture is commonly used as a base course for secondary roads. The asphalt concrete surface mixture was the same material used in the wood-asphalt specimens.
Three different waterproofing membranes were employed in the laboratory specimens constructed for both systems. They included Bituthene 5000 manufactured by W.R Grace,
Petrotac manufactured by Phillips Fibers Corporation, and Protectowrap M400A manufactured by the Protectowrap Company. Control specimens were constructed with no membrane between the wood and surface asphalt concrete layers to assess baseline watertightness and shear strength parameters for an unprotected system.
Properties of weight per unit area and unit weight were determined using a one foot square section of each membrane and a balance capable of measuring to the nearest 0.0001 pounds. Thicknesses were determined by obtaining a one foot square section of each membrane, cutting it into four equal size square sections, and measuring around the entire perimeter a total of eight times (one at each corner and one at the midpoint of each side) with a micrometer capable of measuring to the nearest 0.001 inch. Membrane measurement data is included in Appendix-D.
Measured dimensional properties and physical observations for each membrane are included in Table-2.1. The thickest membrane was determined to be Bituthene 5000 and the thinnest was Petrotac, though the thicknesses of Petrotac and Protectowrap M400A were very similar. The densest membrane was Protectowrap M400A and the least dense was Petrotac.
Table-2.1 Measured Dimensional Properties and Physical Observations for Membranes Membrane Thickness (in) Weight/Unit Area (lb/ft2) Unit Weight (lb/ft3) Observations (at room temperature) Bituthene
5000
0.070 ±0.004
0.3841 66.15 Most flexible of three membranes. Very sticky.
Petrotac 0.061
±0.005
0.2855 55.87 Stiffest of three membranes. Similar to heavy construction felt.
Least sticky. Protectowrap
M400A
0.064 ±0.002
0.4341 81.36 Similar to Bituthene 5000 except for fibrous material present on top
surface. Slightly less flexible than Bituthene.
Load-strain properties were determined for each membrane using the procedure outlined in the draft of a proposed major revision to ASTM D 2523 "Standard Test Method for the Load-Strain Properties of Roofing and Waterproofing Membranes" using 1 inch by 6 inch cut
specimens. Five specimens of each membrane type and orientation were tested after being conditioned for 16 hours at 73.4ºF, 50% R.H. followed by 16 hours at 0ºF, 50% R.H. The two orientations considered for each membrane were 0º and 90º relative to the longitudinal axis of the membrane roll. Each specimen was kept at 0ºF until placement into the loading apparatus and each test was completed within 10 minutes of being placed in the tensile testing machine. The deformation used to determine strain was based on the change of distance between loading grips, so any slippage is implicit in the results.
A summary of load-strain properties is presented in Table-2.2 and complete test data is included in Appendix-D. For tensile load in the 0º orientation, both Bituthene 5000 and Petrotac were approximately 19% lower than Protectowrap M400A. In comparison to individual tensile loads in the 0º orientation, tensile load in the 90º orientation was 9% lower for Protectowrap M400A, 10% lower for Petrotac, and 19% lower for Bituthene 5000. Strain at maximum load was highest for Protectowrap M400A and lowest for Bituthene 5000. Loading orientation was significant to strain at maximum load only for Petrotac where the result was 30% lower for the 0º orientation. Energy at maximum load was highest for Protectowrap M400A and lowest for Bituthene 5000.
For the Creekside Drive bridge, the cross-section of the bridge deck and pavement system consisted of the following (substrate to wearing surface): a creosote treated red maple glulam deck, Pennsylvania Department of Transportation asphalt concrete base mixture BC-BC, Royston 10A waterproofing membrane, and finally, Pennsylvania Department of Transportation asphalt concrete surface mixture ID-2. The cross section of this configuration is similar to the one presented in Figure-2.2.
Table-2.2 Summary of Load-Strain Properties for Membranes
Membrane Maximum Load
(lbs) Strain at MaximumLoad (%)
Energy at Maximum Load
(lbs•in)
Mean Std. Dev. Mean Std. Dev. Mean Std. Dev.
Bituthene 5000 0º 91.6 2.5 19.6 0.8 43.3 2.4 90º 76.7 3.0 19.8 1.6 40.4 3.7 Petrotac 0º 91.9 2.5 55.2 2.8 145.2 10.4 90º 83.2 3.8 71.6 5.4 158.5 15.2 Protectowrap M400A 0º 109.2 5.0 76.1 8.9 257.6 39.3 90º 100.5 8.3 72.0 4.1 219.7 19.9
Assay measurements taken from the bridge deck indicate an average final creosote retention level of 15.8 lb/ft3 in the bridge deck. The deck moisture content was 9.4% at the time of bridge construction. An independent analysis of the asphalt concrete mixtures used on this bridge was performed by the Virginia Transportation Research Council, a copy of which is included in Appendix-C.
2.2.1 Specimen Construction
Specimen construction consisted of several discrete steps including: preparation of 4 inch diameter by 1.5 inch thick wood specimens, cutting 4 inch diameter circles out of membrane sheets, assembling the specimen (asphalt compaction), and constructing a perimeter seal/ponding dam.
2.2.2 Wood Preparation
Nominal 2"-by-12" kiln dried southern pine was obtained directly from an eastern Virginia saw mill in 16' lengths. Each piece of lumber was ripped in half to yield two quarter sawn 2"-by-6" boards and then cross cut into more manageable 4' lengths to allow for shipment to a
Figure 2.5 Drill Press and Hole Saw Used to Obtain Wood Cylinder Substrates
After the boards were pressure treated (where applicable), 4 inch diameter by 1.5 inch thick cylinders were cut using a 4-1/8 inch O.D. hole saw attached to a drill press, see Figure-2.5. Special care was taken to avoid surface defects (i.e., knots, splits) and assay core holes when sawing the cylindrical specimens.
2.2.3 Membrane Preparation
To cut the 4 inch diameter circles from the membranes, a punch was fabricated using a short section of 4 inch O.D. steel tube with a sharpened edge ground against the outside perimeter (Figure-2.6). The membrane was unrolled onto a clean wood surface with the removable backing material facing upward. The punch was placed on top of the membrane and struck on top several times with a heavy hammer until the cutting edge passed cleanly through the membrane material. 2.2.4 Assembly and Compaction of Specimens
Assembly of the test specimens was designed to simulate the rigors that a membrane faces during a typical bridge deck paving operation. With this in mind, asphalt concrete pavement was compacted based on ASTM D 1559 "Resistance to Plastic Flow of Bituminous Mixtures Using Marshall Apparatus". As stated earlier, this test uses a steel mold and a standardized hammer to compact hot mix asphalt into a 4 inch diameter cylinder.
Initially, asphalt concrete test specimens were made in accordance with the standard test procedure to determine the theoretical maximum density for both types of asphalt concrete used in this research. Since only 1.5 inch thick lifts were needed, some preliminary specimens were compacted to determine the mass of hot mix asphalt and the number of hammer blows necessary to reach the theoretical maximum density for each pavement material. Though minimum
requirements vary for in place pavement densities based on mixture type, a pavement with a density at the theoretical maximum exceeds the densities required by the Virginia Department of Transportation and most other transportation agencies. Through trial and error, it was
determined that 700 grams of asphalt concrete compacted with 50 blows applied to one side yielded a 1.5 inch thick lift at the theoretical maximum density for both SM-2AL and BM-2.
To begin specimen assembly, the steel molds, miscellaneous tools, and previously obtained hot mix asphalt were heated to approximately 275ºF in a large convection oven. The compaction hammer was placed on a hot plate to bring its temperature up to 275ºF.
For the wood base specimens, a mold was removed from the oven, wood was placed inside at the bottom, and a membrane (if applicable) was pressed onto the wood. Surface mix asphalt concrete was then measured out to the desired mass, placed in the mold on top of the membrane, and spaded 25 times around the perimeter and 10 times in the center. At this point, the mold was fixed in the compaction machine, the temperature of the hot mix asphalt checked to be sure it exceeded 250ºF, the compaction hammer put in place, and compaction initiated. After 50 hammer blows, the mold was removed from the machine and allowed to cool for
approximately 30 minutes before it was removed from the mold.
For the specimens using BM-2 as the base, the mold was removed from the oven, a piece of release paper was placed in the bottom of the mold, and base mix asphalt concrete was
measured out to the desired mass and placed in the mold. Compaction of the base mix proceeded in a similar manner to that used for the surface mix. The base was allowed to cool for
approximately 20 minutes, a membrane (if applicable) was placed on the base lift, and then a surface mix was added and compacted.
2.2.5 Edge Seal and Ponding Dam
To reduce edge effects, a system consisting of mastic asphalt, a membrane, a PVC plastic cylinder, and a clamp was used to seal the perimeter of each specimen after cooling for at least 24 hours, see Figure-2.7. The system was devised to: allow moisture to enter and exit the specimen from the top and bottom only; allow ponding of the specimen with a 0.5% NaCl solution to saturate the surface asphalt concrete thus providing a conductive path should a leak be present in the membrane; and prevent visco-elastic creep of the asphalt concrete during high temperature environmental exposure. The top surface of each specimen was ponded for the duration of environmental exposure conditioning except for the brief period of time when watertightness and moisture content measurements were being made. In addition to the edge wrap around the system, a plastic or foil wrap (depending on temperature exposure) was fastened over each specimen to prevent the escape of moisture from the system.
Application of Mastic at Section Interface
Application of Containment Membrane
Application of PVC Plastic Surround and Clamp
2.3 Obtaining Creekside Drive Bridge Deck Cores
Three cores were sawn from the bridge constructed using the system design proposed in this research at the locations presented in Figure-2.8. The specimens were obtained using a water cooled 4 inch I.D. diamond core saw as shown in Figure-2.9. For two of the three specimens taken, the core drill penetrated the full depth of the asphalt paving system at which point the core sample immediately broke loose from the treated wood deck indicating that no bond existed between the wood and base asphalt mixture. On the third sample, the base asphalt cracked and broke loose approximately 2 inches before the core drill reached the wood deck, but beyond the waterproofing membrane. The measured depth of surface and base asphalt for each sample is listed in Table-2.3.
After each core sample was removed, the resulting hole in the bridge deck was patched using a cold mix patching asphalt. Special care was taken to replace the removed membrane with a new piece at the same height as the existing membrane.
Once the raw core samples were returned to the laboratory, four inch diameter by three inch thick cylindrical test specimens were prepared by trimming the base asphalt layer to a height of 1.5 inches using a water cooled diamond blade saw. The surface layer for each sample was approximately 1.5 inches as paved, so no trimming was necessary. Once the standard specimens were prepared, they were sealed in the manner described in Section-2.2.5.
Table-2.3 Asphalt Lift Thicknesses for Creekside Drive Bridge Core Samples Specimen
Number Surface AsphaltThickness (in.) Base Asphalt Thickness (in.) BRA.1 1-3/8" to 1-1/2" 4-7/8"to 5"
BRA.2 1-3/8" to 1-1/2" 2-1/2" to 2-3/4" (Broken While Coring) BRA.3 1-3/8" to 1-1/2" 4-1/4" to 4-1/2"
2.4 Environmental Degradation Exposure Cycling
Where applicable, each specimen was exposed to temperature cycle ranges of 0 to 40ºF or 90 to 140ºF in accordance with cycle rates described in ASTM C 666 "Standard Test Method for Resistance of Concrete to Rapid Freezing and Thawing". These two ranges were chosen to simulate the temperature effects of cold weather and warm weather climates in the United States.
An ASTM C 666 rapid freeze thaw chamber utilizing mechanically controlled refrigeration and heating units was used to achieve the desired cold temperature exposure cycles, see Figure-2.10. One complete cold temperature cycle was considered to be the change in temperature at the center of an asphalt-asphalt control specimen starting at 40ºF, decreasing to 0ºF, and returning to 40ºF. The internal temperature of an asphalt-wood control specimen was monitored in conjunction with the asphalt-asphalt control
Figure 2.8 Bridge Core Locations
specimen and temperature data from both was recorded using thermocouples attached to a data acquisition system.
For the warm temperature exposure cycles, an electronically controlled convection oven was employed, see Figure-2.11. One complete warm temperature cycle was
considered to be the change in temperature at the center of an asphalt-asphalt control specimen starting at 90ºF, increasing to 140ºF, and returning to 90ºF. As in the cold temperature chamber, the internal temperature of an asphalt-wood control specimen was also monitored. Data from both controls was recorded using thermocouples attached to a data acquisition system.
The length of each exposure cycle in both chambers was typically 4-6 hours. The cycle length was primarily determined by the thermal properties of the asphalt-asphalt control specimen and chamber loading (i.e., the number of specimens in the chamber). 2.5 Watertightness Test and Procedures
Watertightness of each waterproof asphalt wearing surface system was assessed by
measuring the electrical impedance across a test specimen perpendicular to the direction of bond orientation in the pavement materials. Electrical impedance was measured with a meter using a 12 Volt, 97.4 Hertz square wave alternating current power supply. The maximum measurable impedance with this test instrument is 1.1 MΩ. Test electrodes were fabricated using 12 gauge copper wire embedded in sponges wetted with a 0.5% sodium chloride electrolyte solution. A concrete block weighing 14.48 pounds was placed on the top sponge to provide a consistent pressure on the electrodes for each test. The complete test apparatus is shown in Figure-2.12.
All watertightness tests were conducted at ambient temperature (65ºF) and where applicable, measurements were taken after seven days of ponding with a 0.5% sodium chloride solution, and after 100 and 200 environmental degradation exposure cycles. After exposure to environmental exposure cycling, each specimen was allowed to stabilize to room temperature for a 24 hour period prior to testing. Watertightness tests were performed on the three bridge core samples after they were ponded for seven days with a 0.5% NaCl solution. Prior to watertightness testing, the edge seal clamp on each
specimen was tightened to help eliminate any conductive path if leaking occurred. 2.6 System Bond Test and Procedures
The bond strength between each material of the paving systems was assessed using a shear test apparatus in which load was applied at a constant strain rate parallel to the direction of bond (Figure-2.13). Load was applied on one side of the system bond line while the side opposite was held in a fixed position. Fabrication drawings for the bond strength test apparatus are presented in Appendix-E. System bond strength tests were performed at 0, 100, and 200 environmental degradation exposure cycles.
The bond test strain rate was 0.1 inches/minute and all system bond tests were conducted at an ambient temperature of 72ºF. Shear load and deformation were recorded for each specimen. The location of failure relative to the bond line and maximum
sustained load were also observed and recorded. The load caused by the weight of the test fixture is added to the dial load for all reported results. For the three bridge core samples, system bond strength measurements were taken immediately after the watertightness tests were performed.
