Debonding of the joint-pavement interface and stripping of joint materials is re- lated to adhesion property of the binder。 The normal bond tests and shear bond tests can be used for investigation of adhesive bonding strength。 Effects of water and aging also can be considered by using water immersion specimen and aged bin- der。 Various anti-stripping additives can be used to promote the adhesion between bitumen and aggregate if required [40–42]。 In order to ensure a strong bond be- tween the joint and asphalt pavements, adhesion tests were suggested to carry out on field produced mold specimens in Switzerland。 The used specimens were 100 mm diameter samples which were cored in situ glued to semicircle grips for tensile  testing。  Tests  were  done  with  a  deformation  rate  of  100 mm/min    at

—20 °C。  Minimum  adhesion  strength  of  2。0 MPa  was  required  [14]。  Apart   from

bonding strength, fatigue resistance should also be investigated since wheel tire passage on the joint-pavement interface can result in high interface stress。

3。2。 Aggregate

Flaky aggregates should be avoided in APJs。 The preferred aggregated should have good polished stone value, aggregate abrasion value, aggregate impact value and aggregate crushing value。 Surface characteristics of the aggregate should pro- mote proper adhesion。 According to ASTM D 6297-01, the used aggregate should be crushed, washed, dried, pre-weighed and pre-packaged [10]。 The aggregate size and gradation is not specified, but usually dependent on the requirements of the manufacturers。 Table 4 gives the relevant requirements on aggregate properties in India and UK。 In Indian specification, use is made of single size aggregate。 The nominal maximum size of aggregate depends on the depth of the joint, for example,

12。5 mm for depths of joints up to 75 mm and 20 mm for joints of depths larger

rials must have a much higher failure strain at low temperatures compared to tra- ditional asphalt mixtures。 In this case, a value larger than 1% at —10 °C is recommended。 Minimum tensile strain at failure should be proposed on the basis of application temperature as well as bridge     movement。

The low-temperature cyclic joint movement simulator (JMS) was used to eval- uate the fatigue crack resistance [13,14,22]。 It consisted of a fixed and a sliding ta- ble, which was moved horizontally by a hydraulic piston, thus opening and closing the gap of the joint in a displacement controlled way。 The force exercised during this movement was measured by a load cell。 In Switzerland, the APJ systems were tested with the JMS at —20° ± 2 °C under a cyclic horizontal saw tooth joint move- ment of maximal 25 mm at a strain rate of 10 mm/h。 The strain was found in the order of 8% at a joint expansion of 20 mm [14]。 Research done by Park showed that the surface horizontal strains were in a range from 5% to 10% at a joint movement of 25 mm and a temperature of 20 °C [13]。 EMPA performed a low temperature behav- ior research on four innovative joints provided by Dutch research project on durable silent joints [22]。 The test temperature was controlled at —10 °C and the joints were opened to 13 mm and were closed to the original length at a same rate of 10 mm/h。 In total, 10 tension–compression cycles were applied。 The applied force was mea- sured and the surface strain of joint filling materials was also calculated。 It was found that the strain distribution along the joint width ranged from 1% to 10% at the maximum joint opening of 13 mm。 Most of the strains  were in the  order of 3%。 Cracking and debonding was found in the joint systems with hard  binders。

The near full-scale joint performance validation tests were carried out at Uni- versity of Wyoming to compare the previously-determined material characteriza- tion, failure modes, finite element analysis and proposed design guidelines。 The validation  program  consisted  of  a  series  of  laboratory  tests  that  simulate     a