By Wendell Dubberke
I drove a full sized GMC Yukon to Arlington, rented a two-axle, sprung, car trailer there and hauled a 1969 unrusted Avanti back to Iowa. Normally I use my 1986 Toyota 4X4 with a two-axle, unsprung car trailer for hauling cars. The Toyota/trailer setup is not compatible with the short joint spacings used by most transportation engineers. Instead of using profilometers to evaluate pavement smoothness, they should ride with me in the Toyota with a car or tractor loaded on the trailer and try to drink a cup of hot coffee.
Sometimes the concrete pavement slabs are warped and other times they are tilted. There does not seem to be much mixing of the two types on any single highway, so it must have something to do with design or quality of the pccp components. Iowa highway #30 east of Colo, Iowa warped badly after 15 years of service. Car drivers never complained about this section of highway, but semi drivers were bouncing all over in their cabs. The truckers were happy when the joint areas were eventually ground down to smooth this section of highway. This section of highway #30 was constructed using Ferguson coarse aggregate in the mix. The concrete ledge in the Ferguson quarry is made up of alternating beds of durable limestone (calcite) and less-durable gray, medium-grained dolomite. Finely disseminated, minor amounts of pyrite give the dolomite its gray color. The quality of the aggregate used in the mix may relate to warpage problems in pcc pavements. Tilting is a design problem, with short joint spacings being the main culprit. Tilting can occur even when tie bars are used in the initial construction.
Tilting and warping of pcc pavement panels are most easily observed when the sun is low and at the correct angle. Often, when driving over tilted panels, the thump of each joint is heard and felt, but when looking ahead at the pavement, the joint offset cannot be easily observed. However, the offset joints can be readily observed by looking in the rearview mirror. Also, when traveling on two-lane, tilted panel highways, the off-set joints can be easily observed in the left, oncoming lane.
It is difficult to understand why transportation system designers feel that they must use short joint spacings when building concrete pavements. Unless they are ready to admit that the quality of the concrete isn't what it used to be, it does not make any sense. In Iowa, old pavements made with: (1)coarser grained cements containing no organic additives, (2)mediocre aggregates by todays standards, (3)joint spacings of 40 feet or more, (4) no air entrainment, (5) no plasticizers or water reducers and (6) water/cement ratios between 0.5 and 0.6, often established service records of over 30 years, with some going over 60 years. Now days it seems that we are lucky to get 20 years of maintenance-free service from new pccp. Increased traffic counts and increased loads do not account for this discrepancy.
Some additional problems associated with short joint spacings are; (1)more deteriorated concrete, because initial deterioration usually starts at the joints, (2)increased initial construction costs, (3) increased maintenance costs and (4)increased rehabilitation costs, particularily if the the pavement is to be removed and crushed.
Over the past decade, pccp construction practices include the sawing of transverse joints at an angle. This is to promote lateral slipping if excessive expansion occurs in the slabs. The saw groove is only an inch or two deep. If the pavement cracks, it is supposed to follow the sawed groove. On a 10 inch thick pavement, approximately 8 inches would be a very rough surface, making it a little tough for lateral slipping. In addition, there is a considerable amount of steel in the joints which would restrict lateral movement. On the angled joints I have looked at, I have never seen lateral movement.
(This paragraph added on 03-09-2005) Todd, at the Iowa DOT, tells me that the main reason for the angled joint is to separate left and right axle loads when going from one panel to the next. This will introduce torque into the equation. It will be interesting to see how it works out.
Longer joint spacings and right angle joints (as well as continuously formed pavement with no joints) should be a viable construction method as long as high-quality pcc components (including the cement) are used in a well designed mix with proper placement.
(This paragraph added 03-09-2005) On a recent visit to the Iowa DOT, an engineer indicated that the reason for the early pcc deterioration of the Waterloo continuously formed barrier rail was due to the use of Class 1 (10 year service record) coarse aggregate in the mix. This same argument surfaced 20 years ago and was refuted by the chief geologist who pointed out that the same aggregate was used in the bridge deck and it was not suffering from very early deterioration. Since the aggregate is highly deicing-salt susceptible, the deck would show problems first, but did not. Even when used in salted highways, class 1 aggregate related pccp deterioration will not be visible for 8 or 9 years.
The local inspector reported the early deterioration taking place on the bridge rails near Waterloo. Central office pcc engineers looked at the problem and decided that there was too much water in the mix. It could be seen that in the worst deteriorated areas, the concrete was apparently trying to slump out during construction. The ready-mixed concrete, fed into the slipform machine, met specifications. The slipform machine operator had on-board water and sometimes sprayed water into the input chute.
This pcc early deterioration problem with bridge barrier rails is the same problem as the early deterioration with pcc pavements. Since the barrier rails are high-and-dry, and contain a considerable amount of steel, they are able to survive longer than pavements subjected to freeze/thaw cycles and traffic loads.
Why are the above situations considered a "too much water in the mix" problem? Prior to 1970, most concrete mixes could withstand water/cement ratios of 0.5 or more. Prior to 1940, mixes with water/cement ratios of 0.6 or more were durable. Most of the current pcc early deterioration problems are not related to excessive water/cement ratios (or aggregates for that matter) but rather to additives (water reducers, plasticizers, grinding aids, etc) ending up in the mix. The compatibility of different brands of these additives makes the problem even more complex. If a concrete mix designer feels that these additives are absolutely necessary, then the amount of water used in the mix will be extremely important as it relates to long-term durability. How is a concrete mix designer to know what is the correct amount of water for a mix containing grinding aids, water reducers and/or plasticizers? Surely not high early strength-----Dick Burrows has shown in his excellent report that high early strength can relate to brittleness in concrete, which in turn can relate to micro-shrinkage cracking which in turn can relate to later expansion due to delayed ettringite formation (DEF), alkali-silica reaction (ASR) and/or freeze/thaw cycles.
In the early 1950s, a situation occurred where long-term durability of pccp took a turn for the worse, even when known, high-quality ingredients were used in the initial concrete mix. It took about 10 years to determine that some previously durable coarse-aggregates were non-durable in concrete where deicing salts were applied. The obvious solution would be to stop using deicing salts, but the motoring public was now accustomed to the clear road policy. The other option, which was adopted, was to take into consideration salt susceptible aggregates for classification purposes. It should not have taken 10 years to identify the salt susceptable aggregate problem. The quality of the aggregates or cement had not changed. The only thing that changed was the use of air entrainment and the use of deicing salts, so the new, 1950s pccp deterioration problem had to be related to one of these two or possibly both.
The later, 1980s pcc deterioration problem occurred with no change in the quality of the aggregates. However, the the cements now are; (1) more finely ground, (2)contain grinding aids and early strength enhancers, (3)contain more kiln dust and (4)contain more sulfur. In addition, mix designers are using a variety of water reducers and plasticizers in concrete. Now then, where does a person look for the solution to the problem? Not alkali/silica susceptible aggregates, unless it can be proven that changes in modern cement chemistry and the additional use of water reducers or plasticizers can convert formerly durable aggregates into non-durable aggregates. Its a stretch, but it is possible that the use of cements producing too much ettringite and/or the use of aggregates associated with excessive silica gel, extrude their products into the pore system of the older concretes, but now extrude their expansive products into the shrinkage micro-cracks, inhibiting their capacity to heal themselves through the process of carbonation. The solution to this potential problem is to find better additives and not marginalize aggregates. If a cement or mix additive disrupts the integrity of the concrete matrix, don't look to aggregate quality to solve the problem.
If the ASTM test has not yet been modernized to measure all of the components in cement, the ring test method or the Duggan test method can be used to identify cement brands that are susceptible to early shrinkage micro-cracking.
Long-term pcc durability may be at a crossroad again if the use of water reducers, plasticizers and grinding aids are considered an absolute "must" situation. How much water? How much plasticizer? What brand? How much grinding aid is in the cement? What brand of grinding aid is in the cement? What other compounds, promoting high-early-strength, are in the cement? What kind of a long-term pcc durability is needed? 10 years? 20 years? 30 years? 40 years? A few of the representatives attending the concrete durability meetings said 10 years without maintenance was sufficient. A few pcc pavements in Iowa are over 60 years old, still in service, and not resurfaced. How much faith is warranted in correlations between laboratory quick-test results and long-term pcc durability? One concrete durability specialist may say how obvious it is that high early strength equates with super concretes while another says that high early strength pcc is brittle an therefore susceptible to early shrinkage micro-cracking.
Click here to view recent photos of the 23 year old, slip-formed, deteriorated, waterloo bridge pcc barrier.
(This paragraph added on 2/18/2006) After reading two Iowa DOT reports concerning slip-formed pcc barrier problems by Todd Hanson and one report concerning ettringite formation by Cody, Cody & Spry, a possible cause for the early type of deterioration, seen in the above mentioned photos and incorporating the observations from a variety of people, is presented here. The contractor was having problems getting the D-57 mix through the slip-form machine. Specifications called for a slump of less than 3/4 inch and 6% air. A low w/c mix with AEA @ 25 oz. per cwt produced concrete with only 5.5% air entrainment. Iowa DOT construction records for pcc barrier rails are not as complete as records for pcc paving. Consequently, the amount, if any, of water reducer or other chemical additives is unknown. The slip-form machine operator was seen spraying water on the concrete in the chute when workability probems were encountered. The top of the pcc, slip-formed barrier rails have a washed and wavy appearance where the deterioration occurred after a few years. Surface application of water could get down into vugs and tears in a mix that is too dry and contains too much AEA or other chemical admixtures. Vugs and tears do not necessarily relate to non-durable concrete, but if expansive gels or crystals form in the vugs and tears when an excessive application of surface water reacts with excessive AEA in the mix, a situation could exist where cracking could occur. A compound (sodium lauryl sulfate) used in some air entrainment brands and some detergents was found to form hydrous, expansive fiberous crystals as indicated in the report by Cody, Cody & Spry which can be found here. To solve this problem, the Iowa DOT now uses a better mix design which allows the mix to more easily move through the slip form machine.
While traveling through Missouri on I-35, I noticed pcc deterioration around the deck drain holes on many of the overhead bridges. A similar problem occurred on Iowa bridges. The detrioration problem is enhanced by the use of deicing salts on bridge decks.
Fine grained, Pennsylvanian age limestones, used for coarse-aggregate in pcc mixes, are among the suite of aggregates that are susceptible to deicing salts. Iowa and Missouri share many geological units in the Forest City Basin.
Observations, using the scanning electron microscope (SEM), have shown the damage brine can do to the outer edge of crystal and crystallite boundaries where the interlock takes place. When the interlock is disrupted, crystal integrity fails and further damage by other physical and chemical forces can occur.
In Iowa, most of the deck drain holes now have a pvc plastic liner that extends a foot or so below the drain hole. Some of these may have been a retro-fit, as I have seen them on older bridges too.
These comments do not have anything to do with highway observations, but while traveling through Texas, I was reminded of a trip I took to Austin, Texas a few years earlier to attend an ICAR meeting.
At the ICAR meeting, I attended a session relating to the proper methods for writing and accepting research proposals. Almost everyone seemed to agree that all research proposals should contain a detailed section dealing with the implementation of (positive?) research findings.
I have been in front of research funding groups requesting funds for a research proposal and I have attended research board meeting where others have requested funds for research endeavors.
I get very pessimistic when attending presentations regarding research progress reports and final reports where the presenter purports to show that all research findings fit nicely with research proposal suppositions. Were there other, unreported findings that conflict with the original proposal suppositions? I figure a researcher is lucky if half of the results fit the expectations.
If funding agencies are so sure that the research proposal is totally logical and will be validated with positive research findings, they might as well skip the research phase and go directly to implementation. When research funding agencies require detailed implementation plans as a part of the research proposal, they are, at least, saying two things; (1)the research damn well better produce positive results and (2) if it does produce positive results, more money will be made available for the implementation phase of this project. When you dangle a carrot like that in front of any researcher, you better hope that he/she has a considerable amount of integrity.
If I had my way, I would; (1)deal primarily with local researchers with known integrity, (2)Allow the research proposal to be as flexible as possible, (3) discourage implementation phases in the proposal, (4) encourage the inclusion of findings that do not support the expectations in the original proposal, (5) make it clear that negative findings will not hamper future funding requests (negative findings also relate to good research), (6)avoid research facilities where the amount charged for overhead is too high (deal directly with the person doing the research if possible), (7)avoid research facilities with conflicts of interest and (8)encourage the early processing of results while the research is in progress.
The reason research proposals should be as flexible as possible is because it is so easy to process research data while the research is in progress. The early processed data may indicate that a change in direction is warranted.
