Quality Control and Specification Compliance for Asphalt Pavements

Many of the specifications that are now in use by the Federal agencies and states are based upon the laws of probability, not absolute limits. They are called Statistical Specifications, and use the law of probability to assess specification compliance. They are quite mathematical and not very well understood even by many of those who are using them.  It must be discussed because it is no longer the wave of the future but in use throughout the country. Small details can cause specification noncompliance even when the product furnished is of excellent quality.  The lack of understanding of the laws of probability by those who are writing specifications has caused many of the problems which result in penalties.

What does statistical mean?

At first, one may say “well, he must be talking about gathering lots and lots of data.”  That is not what is meant by statistical, although data is needed to develop a sound statistically based specification.  The word “statistical” and the word “probability” are closely related.  The introduction of statistics is the introduction of the theory of probability and concerns chance.  Some of the top statisticians in the country are undoubtably working in Las Vegas, Atlantic City and in many Indian casinos to make sure that the games offered will assure that those who play them will, on the average, lose.   The same laws that govern gambling govern statistical specification including, with some specifications, assurance that the contractor will not, on the average, get full pay.

The specifications generally are based upon the assurance that for a single pay item with a 100% pay factor, on the average, the contractor who is actually within specification 100% of the time would get paid 95%. Thus any specification that has a maximum pay factor of 100%  has a penalty build within it which is, by the way, calculable. To overcome this problems many specifications will have as a possible “bonus” an extra 5%. However this is really not a bonus but a way to make the specification fair.

Precision and Accuracy

Before I go any further, I need to clear up a misunderstanding. The words Precision and Accuracy are often used interchangeably, however their meanings are quite different.  Accuracy refers to how close the average of the test values represents the actual or true value.  Precision, on the other hand, refers to how close repeated tests on the same sample are to each other.  It is therefore possible to have accurate but imprecise test data or to have very precise but inaccurate data.  As an example, the data may be scattered and, when plotted out, would resemble a shotgun pattern, but the mean of the data would be close to the true value.  The data is therefore accurate, but imprecise (however such imprecise data requires many tests to obtain accuracy).  We also can have a narrow shotgun pattern, which would give us a group of data that was very close to the true value, which would be accurate and precise.

On the other hand, a piece of equipment might very accurately reproduce itself but does not give an accurate answer.  It’s like shooting a very narrow pattern at a target, but with the pattern way up into one of the corners, such as would happen if the sight of the rifle is off.  While the rifle is shooting with precision, it is not accurate.  An example of how this can happen in a laboratory would be the case when a thermometer has a split mercury.  The thermometer might be reading, say, 10° high.  If that thermometer was used to control the viscosity bath, all of the viscosities would be reported very high.  If we ran those viscosities over and over, we would get very close to the same answer as the precision of the viscosity test is quite good, but the answer would be wrong in all cases.  The method would be very precise but not accurate.

What can the specifying engineer do if he wishes to be assured that the actual values lie within a narrower range that can be justified by the confidence limits for a single test?  That is quite easy.  Run duplicate or higher replicate test and use the average of the replicates.  The standard deviation of duplicates on % asphalt is 0.20 rather than 0.28 while the standard deviation for quadruplicates is 0.14.  One way replication has been achieved without increasing the testing load is to use running averages of, say, the last five tests.  The other way is to use lot sizes of 3-7 samples and use the lot average and range of data.

In the early days specifications were generated that would have a number of different criteria with the pay factor that of the lowest value. This gets expensive. If there are two criteria each with a 5% chance of being out of specification compliance when they were actually in, there is a 10% chance that one or the other will be out of specification compliance thus even though they were actually in specification on both criteria. The pay would be 90%. If there were 5 criteria, the pay would be only 77%. I was on one project in which I calculated that there was only one chance out of 10^15 that the contractor could get full pay on all lots. That case was settled in favor of the contractor just prior to the trial.

When limits are not really limits

Before the introduction of statistical specifications if the specification for, say, % asphalt was 4.5-5.5, an average of 4.4 would be fine. Now 4.5 and 5.5 would be the lower and upper limits for calculating pay factors. An average of 4.4 from a lot size of 4 would have a pay factor of 92% (assuming a standard deviation of 0.28).

These are just brief comments of a complicated but important area of mathematics.

Robert L. Dunning www.petroleumsciences.com, chemistdunning@gmail.com


Asphalt Construction Nightmares

You have a project going well when all if the sudden things change. How can this happen? You are buying the asphalt from the same source, it meets specification, but it just doesn’t work well. What could have happened?

There are various factors that can address that problem:

Change of Crude Oil Source. First, every crude produces its own distinct asphalt. York and Halstead characterized a multitude of asphalts in the late 1950s and Rostler and White produced a set of punch cards as “finger prints” of the various asphalt sources. References available by request. I have those cards. I mention this as a refinery may switch crudes. The asphalts would still meet specifications, but may behave entirely different. In addition, many asphalts are being modified with polymers (rubber) in different ways with each system having their own distinct set of properties. Some examples.

Asphalt Susceptible Water. In the middle 1970s a contractor was having a problem with stripping which even 1% antistrip would not help. The mixes when soaked turned brown which indicated that water was being absorbed. It turned out that the crude from which the asphalt was produced had been treated with caustic soda which made the naphthenic acids into soaps. As asphalts from caustic treated crudes made lousy emulsions, they also had non-caustic treated crude in stock. Switching to the non-treated crude solved the problem. Asphalt from the treated asphalt caused other problems so the refiner switched to lime treating of the crude. That greatly improved the asphalt making it resistant to stripping.

Changing Grading System. When I first got involved in asphalt technology (1959), asphalts were graded by penetration at 77°F. (How far a needle would penetrate the asphalt in dmm.) The grade mainly used was 85/100. Later the grading system was changed to viscosity at 140° F after an aging test to mimic the effect of mixing on the asphalt consistency. The contractors were told that the grade AR 4000 would replace the 85/100 grade, implying that that no changes in practice needed to be made. However in Southern California the contractors claimed that the AR 4000 asphalt didn’t work the same. I was at a meeting with the contractors, oil companies and the asphalt institute, the latter two of whom stated flatly that the asphalts were the same. Period. In a broad sense they were telling the truth. Averaging all of the available asphalts, the AR 4000 was the same as an 85/100. However with the asphalts in California, the range of penetrations for AR 4000 from the various crudes varied from 40/50 to 120/150. In Southern California a AR 4000 was actually a 60/70 penetration grade. Once the contractors were finally told that, they were fine with it because they also would have known how to handle the 60/70 asphalt if someone would have told them that was what it was. In the roofing industry some specifications actually specify the crude source. It might be well for a contractor to get a guarantee that the crude source will not be changed.

Product Trading among Oil companies. Oil companies trade products, primarily to reduce shipping costs. One example was one company trading California Coastal crude to another for asphalt emulsion in the northwest. But they can also trade asphalts if there are supply problems at the refinery.

Replacing an Approved Product with Another during Construction. Another thing that can happen is that the supplier submits one material for approval then switches the ingredients to lower the cost. The product may still meet specifications, however will not perform that well in the field.

If retains are kept during construction and there is a costly problem, the asphalts could be analyzed. One method is the Rostler analysis ASTM D 2006, (removed) and the Clay Gell procedure ASTM 2007. (We only do the Rostler procedure because the solvents from this method are easier to handle.) There are other new unpublished procedures that can be used including Fourier transform infrared analysis.

Robert L. Dunning, chemistdunning@gmail.com, www.petroleumsciences.com


Slurry Seals

The application of a seal coat has a number of functions however one of the most important is to waterproof the pavements, protecting them from water damage and oxidation. If pavements were sealed early in their life, e.g. within a year, the pavements would last a lot longer.

Under slurry seals I am including standard Slurry Seal, Microsurfacing, the Rubberized Emulsion-Aggregate Slurry (REAS), a special seal developed to stop emission of radon gas from uranium tailing piles, and the parking lot seals.

None of the above seals will prevent crack propagation when place upon cracked pavements. However there are other seals which will be discussed in a later blog that will repair cracked pavements.

Standard Slurry Seals. Slurry seals are mixes of asphalt emulsions and a well graded aggregate that are placed over a pavement to waterproof and to replace the fines that have raveled away. They come with three gradations, Type I which is the finest, Type II, which is the one mostly used and Type III, which is the coarsest. The exact gradations are usually those specified by the International Slurry Seal Association (ISSA), however often certain agencies will adjust them. The first slurry seal that I saw was mixed in a concrete mixing truck and was made with an SS-1h asphalt emulsion. Today the emulsions that we work with are quick set, which means that in a short time after mixing with the aggregate, the mix sets, kicking out the water and allowing traffic on the slurry. They are now mixed and placed by special trucks that can carry water, emulsion, aggregate and additives. The chemistry is intriguing, but too complex to discuss here.

Microsurfacing. Microsurfacing is a more robust type of slurry which uses primarily the Type III coarse gradation however Type II will also be used. It is designed to fill in areas in which there is more severe damage. The test criteria are more challenging than for a simple slurry seal.

REAS. The Rubberized Emulsion-Aggregate Slurry was developed to be able to be able to use tire buffing in a slurry seal like surface coating. The formulation is quite complex and covered by certain patents. It is specified in the Standard Specifications for Public Works Construction (Green Book) of Southern California.

Seal to Stop Emission of Radon Gas from Uranium Tailing Piles. This was developed on a special project but has not been put into practice as far as I know. It is based upon Slurry Seal quick set technology. Radon has a half life of only about 3 ½ days. If it takes more than 30 days for radon to diffuse through the coating, it is essentially blocked. We used helium gas in our research then turned the technology over to our client who confirmed our data with radon.

Parking Lot and Drive Way Seals. These seals are designed to primarily water proof, fill in minor loss of matrix fines and, actually, be black. The compositions are proprietary but essentially are emulsion based with fine fillers. A specification for this type of seal is also in the Standard Specifications for Public Works Construction (Green Book) mentioned above, however there are other specifications available from suppliers. They do an excellent job in preserving pavements.

Robert L. Dunning. chemistdunning@gmail.com, www.petroleumsciences.com.


Superpave Myths

There is a lot of attention of binder properties on the performance of asphalt pavements. Various types of polymer are added to asphalt in the drive to improve properties.
It is easy to forget about gradation in evaluating a new additive to, say, reduce tenderness and rutting. However the prime reason that polymers are added to asphalt is to change the temperature susceptibility so that the binder appears to be more viscous at the higher temperatures than neat asphalt and softer at lower temperatures than neat asphalt. These additives may make have other benefits, however gradation is extremely important. Here are some Superpave Myths:

  • Rutting is solved by asphalt modification
  • A Maximum Density Line drawn from 100% passing to Zero has value
  • There is a “forbidden area” that the gradation curve must avoid
  • Gradations that go above the “forbidden area” are good as those that go below


Rutting is a gradation problem, not a binder problem. While modifying the binder with additives may show delayed rutting in the tests, the only sure solution is to assure that the coarse aggregate carries the load. There are forest roads that are open graded mixes bound together with a CMS-2s asphalt emulsion (an emulsified cutback) that perform very well. While modifiers may slow the rutting down, the only sure thing is the strength of interlocking rocks. If asphalt is modified so that it cannot stress relieve by flowing, non-load associated cracking will surely occur.

Maximum Density Line

Another myth, in my opinion after decades of providing mix designs, is that a “maximum density” line going from the 100% passing point to zero has value. It does not. Let’s call this the false maximum density line (FMDL) as it has no relationship to the gradation in the mix designs and is thusly of no value.

Forbidden Area

The myth of the so-called “forbidden area” of the gradation is essentially silly and does not provide any benefit. It was based upon an FMDL which has no resemblance to the reality of the gradations in mix designs. It was an academic after-thought that ended up in the specifications. Aggregates with a high rugosity may very well have gradations that could go through that area yet perform very well.

Specifying Gradations above the FMDL

Such gradations will have a hump in them. That is an invitation for rutting since the large aggregate cannot interlock to provide strength. The load has to be carried by a sand mastic.

The above myths may have a nice academic feel, but are useless in solving problems in the field.

Mixes that are Strong and Compact Easily

The tool we have found to be beneficial for strong easily compacted mixes is based upon gradations in which the coarse aggregate interlocks and carries the load. We first draw a straight line from the sieve size that first retains aggregate to the -#200 on the 0.45 power graph. For this discussion, let’s call that the true maximum density line (TMDL). That line should be the specification maximum, i.e., under no circumstance should the gradation plot go over that line. The slope of the line drawn from the sieve that first retains aggregate to the #4 should be greater (coarser) than that of the TMDL. It should not be too coarse, however, in order to avoid segregation problems during construction.  Ideally the slope from the – #4 to the – #200 should be close to a straight line, with a slope less that the TMDL.

If the gradation in about the # 30 sieve goes over the TMDL, the mix will be tender and tend to rut. We call such mixes “over sanded”. There was a pavement a number of years ago that was rutting during construction. By using the above principals, not only was the rutting stopped, when the principals were place into specifications, the rutting specification for the agency could be lowered below the national norm. With certain non-modified asphalts, compaction is nearly impossible with over sanded mixes.

Another problem area in the compaction curve is below the # 30 and above the #200.If there is an overabundance of some material and lack of other, there will be a “hole” in the gradation which needs to be filled, usually with a fine sand. Without the sand as a filler, additional asphalt has to be used resulting in compactions problems.

I did a Gram-Schmidt orthogonalization of gradation data and found that only about 3-4 sieve sizes were truly independent thus having only a few sieve sizes in the specification is wise, especially when statistical specifications are used. However it is also wise for those who do mix designs to have used as many sieve sizes in the gradations as practical so that problems in the -#30 to -#200 region of the gradation can be identified and fixed.

 Criteria for a Rut Resistant, Easily Compacted Economical Hot Mix

First, select a gradation that meets the TMDL guide lines. Within these guidelines for, say, a ½” nominal mix, adjust the gradation and -#200 material so that the effective asphalt content is between 4-5 %, the film thickness is 7-10 microns and the VMA is no higher than 0.5% above specifications. For every 1% the VMA is above the specifications about 0.4 % additional asphalt is required. (For the nominal ½” mix, the VMA specification is higher than it should be at 14%. Hveem mixes for this mix specified by the FHWA was 13% for decades and there is no technical reason for it to be higher.)

For more information or help, contact Robert L. Dunning at chemistdunning@gmail.com  or www.petroleumsciences.com.


Stopping Stripping

In construction the major task is to stick things together and keep things stuck together. In building a house boards are stuck together with nails and with some sort of adhesive at times. With hot mixed asphalt concrete we want the cement (asphalt) to glue the rocks together and for the rocks to stay glued together. We don’t want water slipping in between (or wetting) the rocks, causing damage by replacing the asphalt cement. With asphalt concrete we call the water damage, stripping, and it is very serious.

When things become unstuck from water damage, the cause is loss of adhesion. The asphalt or coating is simply lifted off by the more aggressive power of water, the sustainer of life and the great destroyer; a blessing and bain!

In the absence of water, the separation of the adhesive from the substrate is most often a cohesive failure although rarely identified as such. As an example, bubbles on a coating on a Portland cement concrete (PCC) or asphalt concrete will generally be defined as an adhesive failure. However, if the bubbles are opened up one will generally see part of the substrate stuck to the bottom of the coating or part of the coating stuck to the substrate; a cohesive failure. The failure occurs at a week boundary layer in either the coating (glue) or substrate.

With PCC the water wets not only the outside surface but also the surfaces of all the pores. The presence of the water causes damage by changing the very composition of the Portland cement itself. In doing so its cementing ability is degraded. With asphalt products including seal coats, the general cause is water simply replacing the asphalt binder at the surface. One exception is with chip seals in those cases where a dusty aggregate is used. In that case, the asphalt applied as a rapid set emulsion coats the dust rather than the rocks. With products in which the asphalt resists stripping, rocks break under stress before the asphalt interface is affected or the asphalt flows apart.

Additives are used to stop stripping by altering the chemistry at the asphalt aggregate interface. They work well, however whether the effect is permanent is questionable. There are data available that indicates that the protection does fade away with time for certain additives, resulting in water damage later in the life of the pavement.

The effect of water on the pavements, whether asphalt based or PCC, is not just of academic interest. The pavements lives are drastically diminished if there is damage from water. No matter how sophisticate the mix design is or how elegant the chemistry of the modification of an asphalt binder is, if water is allowed to cause damage all is blowing in the wind. The stripping tests are the most important part of the design procedure.

Symptoms of pavement damage are increased cracking, especially in the wheel path, potholes, and raveling at the surface, especially with standing water on the pavement. I would suggest that water damage would also accelerate damage from studded tires.

There are data that show that water damaged is accelerated by the presence of salt or magnesium chloride which suggests that the stripping tests should be adjusted to run the tests in the presence of those materials in areas that use them because of snow and ice on the roads.

The three main products used as antistrips are:

  • Amines
  • Hydrated lime and quicklime, (even if quicklime is used, it becomes hydrated when it becomes in contact with water. However, in a project on a Marine base in the early 1970s a project did successfully add quicklime directly to the mix in a pug mill.)
  • Ionic organosilicon compounds

Discussion of the additives and the basic chemistry will be discussed in more detail in an article at www.petroleumsciences.com at a later date.

Robert L. Dunning, chemistdunning@gmail.com


Optimizing Pay Factors

The ideal goal of a paving project is to place a pavement with a gradation that won’t rut, contains 4-5% asphalt above that absorbed, have a film thickness of 7-10 microns on the aggregate, is compacted well with no more than 8% voids and is not damaged by water. In addition, the asphalt must be able to stress relax well enough when it is cold so the pavement doesn’t crack. Although many won’t agree, it really doesn’t matter whether Marshall, Hveem or Superpave mix design is used since all the mix design really does is to determine how much asphalt should be used (although with coarse mixes a 6” Marshall is advised).

Finally the contractor wants full pay.

It needs to be reemphasized that if the pavement is susceptible to damage by water, the life of the pavement will be greatly diminished. It is my opinion that the most severe stripping test should be used, and it might be well to run the test in a salt brine as well as in water. And the test should be run on samples with 6-8% voids, never at the mix design void content. The question of how long the protection of a particular antistrip lasts should be considered also, however that is for another day.

When statistical specifications are being used, there are two basic factors that are necessary to get full pay; accuracy and precision. The pay factor is determined from a ratio consisting of a value that measures how accurately the test data represents the job mix formula divided by a value that measures how precise the data are, i.e., how big the spread of data, usually in a lot of four tests.

First, accuracy. The first parameter measured is the distance within the specification limits of the mean of the test data from the control limits. To achieve accuracy, several factors must come into play. These are some examples:

The aggregate gradations must be accurately know and uniform. That means that sampling must be done properly, and it is not easy to get an accurate sample. The gradation of the job mix formula must accurately reflect the material that is going into the pavement.

Any bias in the determination of the % asphalt must be known exactly, i.e. the correction factor for the furnace must be accurately determined, and that determination should be for the furnace being used.

Nuclear gages used for compaction must be calibrated, and it would be well if they were calibrated against cores.

To get the best resistance to rutting, a coarse gradation should be used, however mixes with coarse gradations have a tendency to segregate. Which brings us to the second important factor in calculating the pay factor; precision.

Precision measures how close replicate tests are to each other. Hot mix tends to segregate in a silo and as it comes out behind the paving machine. The mix will tend to be finer behind the tunnels and coarse at the edges and between the tunnels. Thus improper sampling can cause a loss of precision. If sampling is suspected to be a cause of variability, a plot of the #4 data vs. % asphalt can be helpful. If the % asphalt increases with increased % passing the #4, segregation is the cause, not variability of the gradation of the aggregate feed.

Contractors who get 105% pay factors pay close attention to both how close their mixes are to the job mix formula and how uniform the mix is. Note. There is a calculable penalty built into specifications for which the maximum pay factor is only 100%.

For more information on how specifications work, to get help in maximizing pay factors, or on seminars on various aspects of asphalt technology contact Robert L. Dunning, chemistdunning@gmail.com at Petroleum Sciences, Inc., www.petroleumsciences.com.

stripping, accuracy and precision, rutting, statistical specifications

Fundamental Causes of Cracking, Potholes, Raveling, and Rutting in Asphalt Pavements

There are many discussions throughout our country about the serious problems with our pavements. There are some basic material properties and environmental conditions which are fundamental in how pavements perform that need to be recognized. The list below is based on over 50 years of experience. We are able to readily determine the cause of pavement failure from visual inspections with very good reliability.

  • While the tensile strength of a rock might be as much as 10 Mpa, the tensile strength of compacted rocks (aggregate) in a road is zero.
  • The strength of a rock in compression is 20-30 times greater than its tensile strength.
  • The compressive strength of aggregate greatly depends on the gradation. If the gradation is over sanded so that the large rocks cannot interlock, traffic will cause the mix to flow even in compression.
  • Failure can occur when materials containing aggregate are no longer in compression.
  • Asphalt is a liquid and an adhesive and works the best when it can act as a liquid.
  • Although the tensile strength of asphalt is low, it is the only source of strength when a pavement is under tension. That strength increases with increased shear rate (traffic speed) and decreases with increased temperature.
  • The asphalt in a pavement must be able to relax tensile stresses faster than they build up or the pavement will come apart (crack) once the tensile strength of the asphalt is reached.
  • Water greatly reduces the strength of asphalt pavement materials.
  • Most aggregates like water better than asphalt and will reject asphalt if given the chance. That is called stripping.
  • While additives can be added to asphalt so that a mix can meet specification requirements, there are data suggesting that the protection with certain popular additives will not be permanent.
  • The presence of salt or magnesium chloride in wet pavement makes stripping problems worse.

A pavement will not form cracks unless it is in tension. An asphalt pavement is compressible. If it undergoes a rapid cycle in tension, and that tension does not exceed the strength of the asphalt, the pavement can rebound. However if the tension is prolonged, the asphalt will flow allowing the rocks to separate thus forming cracks. Although generally the cracks will remain open, on roads with heavy traffic, the traffic will knead the cracks back together in hot weather.  Potholes form in areas where there is serious cracking. The presence on water can make any damage worse. Raveling occurs at the surface and is greatly increased if water is present. Again traffic causes tensile stresses to pull the aggregate from the surface.

In brief, cracking, potholes and raveling are all caused by tension on the asphalt bond between aggregate particles.

Rutting has four causes, studded tires, under designed pavement, under-compacted subgrade and poor gradation of the aggregate. The problem with studded tires has not been solved. With under-designed pavements, the structural strength is not strong enough to prevent failure of the subgrade causing a rut in the surface. With under-compacted subgrade, traffic on the pavement will complete the compaction resulting in a rut or a low spot. If the aggregate gradation is over sanded, the mix will flow away from the tire track and leave a rut.

A more detailed discussion can be found on www.petroleumsciences.com at a later date.

Robert L. Dunning, chemistdunning@gmail.com