The high percentage of debonding underlines the importance of quality assurance during installation。 In particular, installation should be avoided during rainy weather or in case of wet surfaces of the BEJ-tanking。 For activating the BEJ-tanking side surfaces, it is recom- mended to use a hot-air fan because it blows also away the weak zones of the pavement。 However, activation of the MA surface at the BEJ-tanking side surface has to be carried out carefully to avoid partial burning and overheating of the connecting PBM strip。 Too lit- tle heating on the other hand may not activate the surfaces sufficiently。

The use of bonding agents may be tricky and counterproductive。 This is particularly true for agents with solvents which should also be avoided for environmental and health rea- sons。 Adhesion test results in this investigation showed that the use of such bonding agents might also have negative mechanical effects, since the solvents may be partly absorbed by the mastic of the adjacent pavement and therefore not fully evaporate。 Hence, solvents may remain trapped in the interface between the BEJ filling and the pavement, thus wea-

kening the adhesion resistance and promoting debonding in winter time when the maxi- mum opening of the joint occurs。

The laboratory investigations described in the study revealed wide variations in the quali- ty of the same binder product。 This explains the sometimes pergent behaviour in differ- ent structures in spite of installing the same polymer-modified BEJ system by the same team。 However, it is interesting to note that certain binder test results in the lab were clearly related to the performance in the field and the ranking by the expert group。

Figure 8。 Locations of the BEJs (top left) and results of the damage assessment of the expert group at parapet (top right), in the border zone & emergency (bottom left) lane and in traffic lane (bot- tom right)

Figure 9 shows the change of rating and properties over the years for the BEJs in dense pavements with respect to heat resistance in the lab tests and gelpermeation chromatogra- phy (GPC) measurements of polymer degradation in the field。 Polymer degradation with GPC is defined here as the ratio between the UV signal peak of the original and the de- graded polymer (see Fig。 10)。 Also shown are the dashed lines of requirement, piding the zones with bad and good behavior。 The agreement between lab and expert assessment is particularly satisfying in case of GPC polymer degradation。 One specimen (TI,1 in Fig。 8) showed good expert rating of performance after 5 yrs and moderate polymer degradation in the field (red point), but bad rating in terms of heat resistance of the BEJ binder。 This was probably due to the negligible loading from the little joint movements of this bridge which may have compensated for the unfavourable heat resistance properties in practice。

The study brought much to light the influence of installation equipment and workman- ship。 Key requirements for the durability of BEJ systems were shown to include: a strong

bond between the BEJ filling and BEJ-tanking side, strong cohesion between aggregate and binder, maximum expansibility and load bearing capacity of the BEJ-filling in both warm and cold conditions and a void-free BEJ mixture。 Prerequisites for achieving these proper- ties include comprehensive training of staff in handling materials, equipment and test appa- ratus, as well as strict compliance with good practice, backed by a strict quality control re- gime。

Heat-Resistance GPC Polymer Degradation

2 3 4 5 6

Expert Rating of Field Performance

2。5 yrs