The procedure of preparing samples using the Marshall method is common. It is currently available in EN 12697-30 and ASTM D 6926-04.
The biggest drawback of this method is the incompatibility of laboratory conditions and the realism of a construction site. This type of laboratory compaction consists of tamping a mix down with strokes of a predetermined compaction effort defined by the number of impacts on the face of a cylindrical sample.[47] The Marshall method was created several decades ago for designing an optimal content of binder in fine- and medium-graded mixes with continuous grading. Despite progress in technology, the impact method of compaction is still being widely used.
Besides the gap between the lab and the construction site, the most important issue is the energy used for compacting SMA mixtures. Let us have a look at the compaction effort of 2 x 50 (50 strokes on each face of a cylindrical sample) and another one of 2 x 75. Greater compactive efforts—namely, 2 x 75—are typically applied when designing asphalt concrete mixes for courses under heavy traffic. Such considerable compaction efforts may also be used for mixes of continuous grading. Gap-graded mixes with a strong coarse aggregate skeleton, like an SMA, may experience the following two subsequent stages of compaction:
• Compaction of the skeleton until stone-to-stone contact is achieved
• Additional compaction, which may cause the crushing of weaker grains (overcompaction)
The proper compaction of an SMA mix is achieved at the moment when the stone-to-stone contact is reached, which corresponds to a certain amount of com — pactive effort. That is why in most countries (e. g., Germany, the United States) the compactive effort of 2 x 50 has been determined as adequate, regardless of the traffic assignment. Though it happens rarely, in some countries the same rules are used for the specification of SMA as for asphalt concrete, specifically 2 x 50 for low — and medium-traffic loading and 2 x 75 for heavy traffic.
In some countries, additional tests of SMA sensitivity to overcompacting are required. In the Czech method (see Chapter 7), typical compaction (2 x 50) is used for design, and an additional test of 2 x 100 strokes is used to indicate the resistance of the aggregate mix to overcompaction. It is obvious that after such overcompacting some of the aggregates will be crushed. The requirement for the air void content (more than 2.5% with 2 x 100) in such overcompacted samples could help ensure that even in cases of overcompacting during rolling or when weak aggregates are used, the SMA layer will still have enough void space between the aggregate grains.
Research conducted in many countries (Boratyriski and Krzemihski, 2005; Brown and Haddock, 1997; Perez et al., 2004) into the compaction of SMA has demonstrated clear evidence that an excessive compaction effort put into Marshall samples leads to aggregate crushing and adverse volume changes in the mix.
