By Wendell Dubberke
Back in the 1930s and 1940s, the Iowa DOT specifications for pcc aggregate was very lax. For example, crushed stone from the Glory Quarry was shipped all over the state because the quarry was located near a railroad between Cedar Rapids and Waterloo. The Devonian age, crushed stone was produced from argillaceous, dolomitic, limestone ledges. The Glory stone had a mediocre pccp service record until the use of deicing salts began in the early 1950s. The service record became extremely poor where deicing salts were used. Consequently, specifications for the acceptance of gravel and quarry stone, for use in pccp, were strengthened.
The argillaceous dolomites and limestone are both ASR and freeze/thaw susceptable as they usually fail the ASTM C666-B test method and can cause surface pop-outs prior to any freeze/thaw cycles. Iowa and most other states identified and eliminated these ASR susceptable aggregates by specification, rigorous testing and trial and error in the 1950s and 1960s.
In Iowa, pcc coarse aggregate can contain up to 5% non-complying particles because extensive service record information indicates that a good pcc matrix can handle moderate amounts of silica gel. Sometimes aesthetics can be a problem if ASR particles are too near the pccp surface and cause popouts, even then, experience has shown that the integrity of the pcc slab can be maintained if the percentage of non-complying particles is kept at less than 5%.
After all of the effort, in the 1950s and 1960s, to identify and eliminate ASR aggregates for use in pcc, the ASR horse has reared its ugly head again in the 1990s. Cements with only 0.7% equivalent alkali are deemed high alkali. Pcc coarse aggregates containing only minor amounts of reactive particles, and with excellent prior service records, are now being blamed for pcc failures occurring a very short time after placement. The same thing is true for the fine aggregates (sands) where minor amounts of reactive particles are being blamed for early pccp failure. This is mostly nonsense. Fast aggregate quality test methods must correlate with long-term service records, if they do not, the test method should be suspect.
Large particles of tripoli chert (sedimentary) and crystoballite are extremely alkali-silica reactive when used in pcc and can cause large popouts shortly after placement. These same particles, when finely ground, can become beneficial pozzolans for use in pcc. As with flyash and silica fume, these finely ground chert or crystoballite particles are alkali-silica reactive. They can generate silica gel around each small particle, which a good pcc matrix can easily absorb. This evenly distributed silica gel can do a good job of mopping up stray potassium and sodium ions. If size matters, does it then stand to reason that sand sized particles should be less reactive than coarse aggregate particles? A good pcc matrix should be able to handle minor amounts of silica gel generated from faulty sand particles. The size and percentage of faulty sand particles could be a controlling factor.
A better explanation for much of the faulty pccp seen in the 1990s is microcracking caused by the use of shrinkage prone cements and/or mix designs that use too much finely ground cement. Shrinkage cracking occurs first, then water more easily enters the pcc through the microcracks. Alkali-silica reactive particles that, under normal circumstances, would be relatively inert due to the lack of water and being surrounded by a durable matrix are now able to easily expand and additionally crack a previously weakened matrix. A microcracked matrix also enhances the development of expansive ettringite which also needs access to water. In states where both available-moisture and freeze/thaw cycles occur, the results from freeze/thaw damage will overwhelm the damage from ASR or DEF in a short amount of time. In many cases, if the microcracking had not occurred first, the damage from the other mechanisms would be minor.
(The following 5 paragraphs added on 7-5-2007) Recently Dick Burrows emailed some photographs showing the surface cracking taking place on the Denver International Airport (DIA) runways. The 13" thick, pcc slabs are being replaced after less than 10 years of service. According to the Denver newspaper, DIA officials are blaming alkali/silica reactive particles, in the aggregate, for the early pcc failure. The DIA officials also say the use of a potassium acetate deicer made conditions worse.
Crushed Glory quarry limestone (argillaceous & dolomitic), discussed earlier on this page, was used extensively in Iowa pccp in the 1930s & 1940s. When deicing salts began to be used on highways in the early 1950s, pccp made with Glory stone failed in less than 10 years. Surface pop-outs and extensive scaling occured on the pccp surface. Internal cracking was taking place as well.
In Iowa, the specification for chloride deicing salt is that it must be at least 70% sodium chloride. When a random sample was tested, the remaining 30% contained a considerable amount of potassium. Sodium chloride and potassium chloride are very soluble in water. Sodium and potassium ions should penetrate pccp easily and function in a manner similar to cement alkalis as far as ASR is concerned. In Iowa, a 50 year old pcc sidewalk, made with pit-run gravel, treated for the first time with deicing salt, showed fresh pop-outs where the deicing salt was applied.
The surface photos of the DIA map-cracked pcc showed no pop-outs. If these photos are representative of the whole problem, then the diagnosis of ASR being the initial culprit is tenuis. How can there be enough alkali/silica reactive particles to generate enough expansive silica gel to internally crack a 13" slab, and yet not cause pop-outs or scaling on the surface where matrix tensile strength is less and where moisture and deicing salt components are higher?
Concerning the Glory stone used in Iowa pccp in the 1930s & 1940s. Some pcc investigators would say that the poor pcc service record of this stone relates to alkali carbonate reactivity (ACR) rather than ASR. The Glory quarry stone is an argillaceous dolomitic limestone. The clay particles are microscopic and evenly distributed throughout the stone. It is possible that the silicon, in the clay particles, relates to the problem and that it is therefore closer to an alkali/silica reaction. The clay fraction of the Glory stone runs between 10%-20%. Before the introduction of deicing salts (1950), pccp made with glory stone could go 20 years without patching, generating a mediocre service record.
