Moisture susceptibility of HMA mixtures
The term "stripping" is applied to hot-mix asphalt (HMA) mixtures that exhibit separation of asphalt binder film from aggregate surfaces due primarily to the action of moisture and/or moisture vapor. During the last 10 years, I have investigated four cases of premature (less than two years after construction) failure of HMA overlays resulting from stripping. The following conclusions were drawn and recommendations made from those four case histories:
First, as has been said by many pavement engineers, three things are important for pavements: drainage, drainage and drainage. Stripping of asphalt courses will not occur in absence of moisture and moisture vapor.
Next, if subsurface drainage of the pavement is inadequate, moisture and/or moisture vapor can move upwards due to capillary action and/or the "greenhouse" effect and saturate the asphalt courses. Stripping can then occur due to cyclic pore water pressure generated by heavy traffic, which will physically scour the asphalt binder off the aggregate surface.
Also, if an open-graded friction course (OGFC) is used as a wearing course, the underlying asphalt surface course has a potential to strip.
The following recommendations are made to minimize stripping in the underlying asphalt course:
- Do not use a surface course mix with more than 12.5 mm maximum nominal size;
- If possible, delay the placement of OGFC for two summers so that the surface of the underlying mix gets kneaded by traffic and becomes practically impermeable to water from the OGFC;
- If the placement of OGFC cannot be delayed, apply a uniform emulsion fog seal to completely fill the surface voids just prior to the placement of OGFC; and
- Use an effective anti-stripping agent in the underlying surface course mix.
There is a need to develop a reliable and realistic laboratory test method to predict moisture susceptibility of HMA mixtures. It was observed in the four case histories that the asphalt pavements were near 100% saturated with water (not 55-80% saturated as specified in ASTM D4867 or AASHTO T283) and the cyclic pore pressure generated by the traffic mechanically scoured the asphalt binder off the aggregate surface. A laboratory test procedure that simulates such conditions will be more realistic.