Site hosted by Build your free website today!

Concrete Deterioration and Testing; Observations, Comments & Opinions

By Wendell Dubberke

ASTM C666 Durability Test
Duggan Test Method
Alkali-Silica Reaction (ASR)
Alkali-Carbonate Reaction (ACR)
Delayed Ettringite Formation (DEF)
Fly Ash & Granulated Blast Furnace Slag use in PCC
Plasticizers, Water Reducers And Retarders
SEM Images & Chemistry from Texas PCC Box Beams
Traveling & Gawking at Highways in 2003
Evaluating Aggregates for PCC

I am not a consultant. I am a retired Iowa DOT employee. I worked for the Iowa DOT for 30 years as a geologist. The first 20 years was spent mostly in the area of classifying aggregates for use in concrete and asphalt. The next five years was spent in developing aggregate classification test methods using x-ray analyses (XRF & XRD), Thermal analysis (TGA) and pore system analysis.

The coarse-aggregate pore system was analyzed by pushing water, at low pressure (35psi), into the pores of 9000 grams of 3/4 inch, oven dried, crushed aggregate. The volume of water, injected into the pores, is read at one minute and again at 15 minutes. The test is non-destructive. This simple pore system analysis test method is sufficient to identify those carbonate aggregates that contain an extensive capillary-sized pore system. Since carbonate pore size diameters correlate strongly with grain size diameters, an indication of grain size can also obtained. Early in my career, concrete durability gurus told me (and I believed) the pore size and volume in concrete aggregates related directly to freeze/thaw susceptibility. If the previous statement is true (unfortunately, it is not true for field concrete), then why not measure the pore system directly? With this thought in mind, the Iowa pore index test method was developed. As it turns out, pore index test results (like ASTM C666 test results) are good for about 1/3 of the information needed to evaluate carbonate coarse-aggregates for use in Iowa PCCP. In Iowa, the worst PCCP coarse-aggregate performers are those that contain inferior chemical characteristics along with an extensive capillary sized pore system that provides a very large internal surface area for adverse chemical reactions to occur. In some cases, crystallites and crystals in fine grained (approximately 25 micron diameter) carbonate aggregates, can completely loose their interlock (particularly when deicing salts are used) and then become freeze/thaw susceptible. Images (polished sections treated with NaCl)from the scanning electron microscope showed the deicing salt to be particularly agressive on the perimeter of the crystallites and crystals. With fine grained carbonates, the crystal interlock between pores is short (10-15 microns). In Iowa, the limestones (calcite) aggregates that perform poorly when used in PCCP (receiving deicing salts) contain more strontium than normal performing limestone aggregates. The overall amount of strontium is quite small in either case. The outside edge of crystals may contain aragonite which is more unstable than calcite. Aragonite can hold more strontium than calcite.

The last five years was spent investigating the initial cause of the early deterioration of some pcc pavements. X-ray (XRD & XRF) equipment, UV spectrometers, thermogravimetric (TGA) equipment and low-vacuum scanning electron microscopes (SEM) were used in this endeavor. These observations & conclusions are my own. Other Iowa DOT researchers were investigating a variety of areas in an attempt to answer the question of why some of the portland cement concrete pavements (PCCP), designed for a 30 year service life were starting to fail in 4 years.

The observations should be accurate. The conclusions should be questioned just as the conclusions of other pc concrete investigators need to be questioned. Albert Einstein said the important thing is to never stop questioning. Just because some concrete guru says something is true doen't make it true. In science, you are supposed to go where the numbers take you, whether you like the direction or not. Unfortunately, many "scientists" operate in a manner similar to preachers and politicians---- that a firm, sincere belief can override sound judgement.

Many of the tests used to evaluate concrete and concrete components were developed in the 1930s and 1940s. Much time and money has been spent trying to fine-tune these tests during the last 40 years. Very few concrete or aggregate investigators ever questioned the validity of some of these old test methods. It was relatively easy to get research funds directed towards the minor modification of existing test methods. Obtaining research funds that would be used for questioning basic assumptions concerning these old test methods was more difficult to obtain.

With a few exceptions, the end-point users of concrete seem to have an aversion to the use of sophisticated equipment for the evaluation of concrete components. At a meeting in St. Louis. attended by concrete and aggregate researchers, a show of hands indicated that nearly everyone thought that the use of a scanning electron microscope was not needed for petrographic analysis of PCCP. A wide variety of analytical equipment is now available to do direct testing of concrete and concrete components. Over the past 10 years, the manufacturers of these analytical machines have made them extremely reliable and user friendly. Iowa DOT experience with some modern analytical equipment is covered in links to other pages found at the top of this page.

As the end-point users of concrete move towards vendor certification of concrete and concrete components, it is more important than ever to have the capacity to fully evaluate the concrete and all of its components at the point of incorporation into the project. Some people, in charge of inspection, think vendor certification along with minor inspection is an adequate safeguard for the production of quality pc concrete--it is not. Vendor certification programs can work if-- (1) the final product and its components are fully tested and meet adequate specifications and (2) penalties (and enforcement) are adequate to keep vendors and contractors on thier toes.

The reliance upon the vendor certification program in conjunction with ready-mixed concrete producers can be used to illustrate a potential problem. I do not want to give the impression that I'm picking on ready-mixed concrete producers. Most of them produce a quality product. But if you are a user of ready-mixed concrete, how do you gauge the long-term quality of the product? Early compressive strength? Pretesting some of the components you assume will go into the final product? Certification of the individual components that will go into the final certified product? Vendor certification--have to trust someone? Flexural strength? What else?

A step in the right direction would be to wash the plastic concrete over a screen and test the components using modern testing techniques. The quality and gradation of the coarse and fine aggregate could be verified. The amount of fly ash could be verified. Verifying the amount of cement clinker grinding aids, water reducers, retarders, plasticizers and other pozzolans is very difficult, but with the use of modern spectographic equipment it should be possible. The need to qualify and quantify all of the components in concrete is urgent. How can direct correlations of PCCP service records to concrete components be accomplished if some of the components remain hidden? Indirect correlations, now available, seem to be implicating some of these unmeasured components in concrete. Manufacturers of these unmeasured components in concrete say that the amount used is so small that it can not have a significant impact on long-term durability. I would submit that the potency rather than the amount is whats critical. A clinker grinding aid must be potent if it is to maintain flowability in a cement silo over time considering the weight involved.

Recently, in an attempt to get a handle on the relationship of concrete components and test results to the early deterioration of PCCP, an outside concrete investigative group was hired. The latest technology in correlation programming was used. It was a cooperative effort with several states participating. The Iowa DOT does as well as any DOT in maintaining construction and materials records. However, when the daily reports (central PCCP mix plant) were analyzed in detail and computerized, errors and other shortcomings were obvious. No matter how sophisticated the correlation program is---garbage in, garbage out. How do we tell this sophisticated correlation program how much (and type) surfactant and/or dispersent (clinker grinding aids, water reducers, etc.) are in the PCCP. If the inspectors write down 3 ounces of a lignosulfonate water reducer per sack of cement in the daily report, is that good enough? A potassium carbonate extraction test showed approximately 10 ounces of lignosulfonate water reducer per sack of cement in a highly deteriorated section of PCCP and this was after a period of 6 years when water would have most likely removed some of the initial lignosulfonate.

Again--if every component in the plastic or hardened concrete is measured and evaluated, the workload placed upon the inspection staff can be lessened or possibly eliminated, but not until then. In respect to the taxpayers who are paying for all of this, you can't eliminate inspectors and hope that someday in the future adequate test methods will arrive or worse yet, that they are not presently needed.

The best way for PCCP to compete with asphalt pavements is to have a product (pavement) that will perform at least twice as long before maintenance is required. Currently, many PCCP engineers and designers are trying to compete with asphalt, in the area of initial cost, by lowering the quality of their product. The cost lowering of PCCP usually involves the using of less cement in the PCCP design. If high early strengths can be maintained with a reduction in the amount of cement used in the mix, PCCP designers feel they have done their job. Apparently, it is assumed that high early strength equals long-term PCCP durability. Mix designs, developed by PCC paving contractors in competition with asphalt paving contractors, also go for an acceptable early strength with the lowest possible amount of true cement in the mix. Only if the PCC paving contractor can absolutely guarantee a 30/40 year PCCP service life, can the contracting authority at the DOT give a two for one price differential in regard to PCC paving vs. asphalt paving. The argument that only a few of the PCC pavements are beginning to deterirate in 5 years will not work. How many PCC pavements will begin to fail in 15 years? 20 years?

In an effort to reduce the initial cost of PCCP, designers at the Iowa DOT implemented the following- (1) Allowed a reduction of 5% cement in the PCCP mix when water reducers were used. It was argued that early strength could be maintained under these circumstances, (2) Allowed the use of cement containing 17% calcined kaolinite formulated at the cement plant, (3) Allowed the use of 15% - 20% replacement of cement with fly ash, (4) Allowed the use of granulated blast furnace slag (GBFS) replacement of some cement, (5) and in some cases allowed for the multiple use of the above items in a single PCCP mix.

In addition to all of the above, the cement manufacturers now want to add carbonate fines (approximately 8%) to their final product. Both dolomites and calcites would be acceptable. Would argillaceous and/or cherty carbonates be acceptable?

The Iowa DOT implements severe penalties on aggregate producers who have too much fine material (minus #200) in their coarse aggregate, even if the material (carbonate) is the same as the parent material. That the DOT would require the aggregate producer to remove fine carbonate material by specification and then accept cement containing up to 8% of possibly inferior carbonate material would be idiotic.

Some in-house research at the Iowa DOT indicates that nearly pure calcite fine material (up to 5%) from the Alden quarry could inhance durability factor (ASTM C666) results. However, dolomite fine material from one of the best dolomite quarries (coarse-grained pure dolomite up to 5%) related to very poor durability factor test results. Reagent grade calcium carbonate (made by a precipitation process) was also tried at the 5% level. It had an adverse effect on durability factor results.

The philosophy behind the move towards the direct testing of pcc aggregates (and other pcc components as well) is that if an aggregate consistently fails when used in PCCP, then there must be a chemical ond/or physical component in the aggregate that can be directly measured with modern techniques. The direct testing (XRF, XRD, TGA, pore analysis, etc) can be accomplished in a matter of minutes as opposed to indirect testing (freeze-thaw tests, etc) that can take days, weeks, months or years to accomplish. Many assumptions, concerning the material being tested, are being made with the use of indirect test methods. Some of them are invalid.

Before 1980, most of the early pc concrete deterioration could be related to deficiencies within the coarse aggregate used in the pcc mix. At the Iowa DOT, records regarding components used to construct PCCP, have been kept since the 1920s. The records were computerized long ago and conseqently it was easy to run correlations relating PCCP service records to the aggregate's source which includes the ledge wihin a quarry. Correlating concrete test methods, such as the ASTM C666 durability test, directly to PCCP service records was a routine operation.

A poor quality aggregate will relate to early PCCP deterioration regardless of the other components in the PCCP mix. The Iowa DOT was never able to identify any fine-aggregate (sand) source that consistently related to early PCCP deterioration (based on service records) even though some of the sands contained significant (15%) questionable material (chert, iron spal, etc)in the plus #30 size fraction. The minus #30 size fraction is mainly durable quartz. These correlations were made directly to PCCP field service records. At the Iowa DOT, service record overrides any laboratory test method. The reason for this is that no completely satisfactory test method for evaluating concrete or aggregate could be found. Iowa DOT personnel were also able to correlate PCC and aggregate test results to service records. These correlations contain a large number of observations and therefor should command more respect than subjective, isolated opinions involving only a few observations. Even though the correlation results may be indirect (for example, not being able to measure the amount of clinker grinding aid in cement), they can give direction to concrete/aggregate investigators for future research.

The best correlation of aggregate test results to PCCP service records involved the use of a rudimentary program (fuzzy logic?) to combine (assign weights) to XRD, XRF, TGA and pore system test results, performed on coarse-aggregate, generating an overall quality number.

I would encourage other concrete and aggregate researchers to generate a web page. A considerable amount of in-house research must have been accomplished in other agencies. Quite often the personnel involved with the in-house research are too busy with routine operations to officially publish their findings or perhaps their supervisors discouraged official publication for a variety of reasons. Other researchers may want to read about your work.

Click on subjects at the TOP of this page for additional comments. Some of the pages are still under construction.