With its short heating times, ultrasonic welding is a highly promising technique for joining thermoplastic (TPC) to thermoset (TSC) composites, to prevent thermal degradation of the thermoset adherend. A neat thermoplastic coupling layer is co-cured on the surface to be welded to make the TSC “weldable”. For welding CF/PEEK to a TSC adherend, it would be logical to use PEEK as the coupling layer. However PEEK and epoxy are not miscible with each other, therefore a bond created after co-curing of these two materials is not reliable. PEI on the other hand is known to be miscible to most epoxy systems at high temperatures and PEEK polymers, hence it is an excellent candidate for the coupling layer material. The other necessary element for ultrasonic welding is the energy director (ED), a neat TP film placed at the interface to help promote heat generation through preferential frictional and viscoelastic heating. Usually EDs are made from the same material as the TP matrix, but in this case ED can be either PEI or PEEK. Mechanical testing and fractographic analysis showed that the usage of a PEEK ED is the most successful approach. This research is part of the European project EFFICOMP.

Adhesive bonding is a highly desirable joining technique to join composites to metals. The surfaces of both composite and metal substrates have to be carefully treated before bonding them together, in order to avoid interface failure between the adherend’s surface and adhesive. This paper describes the surface pretreatments on carbon fiber reinforced plastic (CFRP) and Titanium for the manufacturing of adhesively bonded CFRP-Titanium joints. Different treatments were applied in order to roughen and activate both substrate surfaces. The quality of the surface pretreatment using different treatment methods was initially checked by contact angle measurements. Destructive tests on the bonded specimens after various surface pretreatments, including those which provided the lowest contact angle, were performed to validate the mechanical performance of the surface treatment on the bond quality. The test procedure and results on adhesively bonded CFRP-CFRP specimens and Titanium-Titanium specimens will be presented and discussed. 100% cohesive failure in both CFRP-CFRP and Titanium-Titanium joint types guarantees the high quality of the adhesively bonded joints, and proves that the respective surface pretreatments on CFRP and Titanium excludes adhesive failures in bonded CFRP-Titanium joints.

The goal of this study is to investigate new designs of composite bonded joints in order to improve their strength under tensile loading. Multiple stacked overlaps are compared with single overlap designs. The concept of multiple stacking is well known as ply-interleaving technique for co-curing dissimilar materials. For a secondary bonding process, a similar concept is used in tongue-and-groove joints. However, it is so far limited to one stacking level due to the complexity of the design. By means of thin unidirectional layers, the tongue-and-groove design is expanded further to two stacking sequences and applied to secondary bonding of CFRP adherends.
Single lap joints of 12.7 and 25.4 mm overlap length were compared to finger joints with 1 and 2 overlaps of 12.7 mm overlap length, stacked through the thickness. Specimens were tested according to ASTM D-5868-01. The initial and final failure load were recorded.
The study shows that for the same overlap length in a multi-stacked configuration, there is a potential for higher average lap strength, in comparison with an increase in overlap length of a single overlap. This effect might be mainly due to the reduction of secondary bending moment and by load distribution over multiple interfaces.

Orthotropic steel bridges experienced early fatigue failures of several welded connections in the steel deck plate. Solutions to enlarge the fatigue life of the existing movable decks include the bonding of a second steel plate of 6 mm to the existing 12mm deck. The adhesive is a thin epoxy layer, vacuum infused between the two steel plates. This renovation technique was for the first time applied on the orthotropic deck of the movable Bridge Scharsterrijn. In this paper, the results from static measurements performed on the old and renovated deck are presented. The tests were carried out with a calibrated truck positioned on specific locations of the deck. The resulting strains in the deck plate at each location were recorded using strain gauges installed on the deck.
Strain influence lines were drawn for each strain gauge position and wheel load. After analyzing the results before and after the renovation, it can be concluded that the strain values decrease considerably after renovation. The strains measured at 15 mm from the welded connection between deck plate and stiffeners web reduce about 60 % in the deck plate and 40 % in the stiffeners web. As this is one of the critical details for the fatigue life of orthotropic bridges it can be concluded that this renovation technique seems to be a promising solution to extend the fatigue life of orthotropic decks of movable bridges.

Deck plates of orthotropic steel bridges experienced early and threaten fatigue cracks at the heavy vehicle lane. Previous research developed a new stiffer surface layer for renovations of the fixed bridges based on reinforced high performance concrete. The study presented in this paper focuses on a possible renovation system for movable bridges in which a second steel plate is added to the existing bridge deck. The properties and durability of the interface layer between the two steel plates strongly influence the response and efficiency of the structure. The study focuses on two solutions, thin epoxy layer namely Bonded Steel Plates and thick polyurethane layer namely Sandwich Steel Plates. Structural calculations were carried out based on analytical solutions using
Classical Laminate Plate Theory and First order shear Deformation plate Theory. The different parameters of the renovations structures were varied and the results for the two solutions are compared. Based on the weight restrictions and geometrical properties of the existence deck plate, one can choose the most efficient interface layer from the material available, i.e., the lightweight structure solution that provides the increase of stiffness required for the renovation.

There is a tendency to use more and more High Strength Steel (HSS) elements in civil engineering structures. The rules described in Eurocode 3 for bolted connections in bearing can be applied on joints of plates of steel grades up to S700. However, these rules are based on test data of connections with 8.8 and 10.9 class bolts in mild steel plates. In fact “strong” bolts in “weak” steel plates. With the use of S690, S960 or even higher grade plates, in combination with conventional bolts, this changes to “weak” bolts in “strong” steel plates.

In this study, a series of tests was carried out using specimens designed according to the rules of Eurocode 3, part 1-8 “Design of joints”. The aim of the study was to investigate whether or not those rules are adequate for high strength steels.

The experimental programme consisted of ten different types of specimens of single bolt joints made with steel grade S690. End and edge distances were varied. In total, thirty tests were performed (three tests per each different type of specimen). The test results show that the rules given by Eurocode 3 are conservative using steel grade S690, mainly when edge distance is smaller than 1.5d0. Therefore, a corrected function for the k1 factor of the bearing resistance formula given by Eurocode 3 is proposed. The proposed correction is based on a statistical evaluation of the test results according to Annex D of EN1990: Basis of Design (formerly Annex Z of Eurocode 3: Design of Steel Structures). This correction was made in the k1 factor, since the main differences between experimental values and theoretical values were found in tests specimens with different edge distances. The test results further show that using HSS plates, the minimal values of edge and
end distance can also be reduced from 1.2d0 to 1.0d0.