Constant amplitude fatigue tests with welded specimens under fully reversed four-point bending as well as under axial loading have shown that vibration stress relieving does not lead to a fatigue life improvement of welded parts when compared to the as-welded state. Thus, a substitution of thermal stress relieving by a vibration treatment is not successful. This was also proved by residual stress measurements in the welded parts studied in this paper. NOMENCLATURE A, = elongation f = frequency F = force k = slope of the S-N-curve Nf = number of cycles to final failure P, = probability of survival R = ratio between minimum and maximum stress R, = tensile strength RN,2 = Feld strength t = vibration time ua = stress amplitude e= residual stress An = nominal stress range AnB = stress amplitude

In this study, the feasibility of low-cost Cu-sintering technology for power electronics packaging and the effect of sintering conditions on the bonding strength of the Cu-sintered joint have been evaluated. A Cu paste with nano-sized Cu powders and a metal content of ~78% as a high-temperature bonding material was fabricated. The sinter-bonding reactions and mechanical strengths of Cu-sintered joints were evaluated at different sinter bonding pressures, temperatures, and durations during the sintering process. The shear strength of the Cu-sintered joints increased with increasing sintering pressure. Good interfacial uniformity and stable metallurgical microstructures were observed in the Cu joints sintered at a high sintering pressure of 10 MPa, irrespective of the sintering time. It was confirmed that a high-pressure-assisted sintering process could create relatively dense sintered layers and good interfacial uniformity in the Cu-sintered joints, regardless of the sintering temperatures being in the range of 225–300 °C. The influence of the sinter bonding pressure on the shear strengths of the Cu-sintered joints was more significant compared to that of the sintering temperature. Durations of 10 min (at 300 °C) and 60 min (at 225 and 250 °C) are sufficient for complete sintering reactions between the Si chip and the direct bond copper (DBC) substrate. Relatively good metallic bonding and dense sintered microstructures created by a high sintering pressure of 10 MPa resulted in high shear strength in excess of 40 MPa of the Cu-sintered joints.

No abstract

In this paper, we evaluated the solderability of thin electroless nickel-electroless palladium-immersion gold (ENEPIG) plating layer for fine-pitch package applications. Firstly, the wetting behavior, interfacial reactions, and mechanical reliability of a Sn-3.0Ag-0.5Cu (SAC305) solder alloy on a thin ENEPIG coated substrate were evaluated. In the wetting test, maximum wetting force increased with increasing immersion time, and the wetting force remained a constant value after 5 s immersion time. In the initial soldering reaction, (Cu,Ni) 6 Sn 5 intermetallic compound (IMC) and P-rich Ni layer formed at the SAC305/ENEPIG interface. After a prolonged reaction, the P-rich Ni layer was destroyed, and (Cu,Ni) 3 Sn IMC formed underneath the destroyed P-rich Ni layer. In the high-speed shear test, the percentage of brittle fracture increased with increasing shear speed.

The results of fatigue tests on 40 tubular T and X-joints are presented. The chord diameters of the joints are 168, 457 and 914 mm. The most important conclusions are that the lifetime decreases with increasing joint size and that in the high cycle rance (N ? 107) some test results of the large joints fall below the AWS-X curve. INTRODUCTION The weather conditions which induce a fatigue crack are more severe in the North Sea than in the Gu1f of Mexico, Because of these severe conditions, a research project, sponsored by the European Coal and Steel Community (ECSC) on the fatigue behaviour of steel in offshore structures, entitled "Fatigue and corrosion fatigue behaviour of offshore steel structures", has been carried out. The research has been carried out in the U.K., France, Germany, Italy and the Netherlands. There is a special cooperation between Norway and the participating ECSC countries. In the programme, flat plate specimens and tubular joints have been tested. In addition, theoretical work (e.g. finite element analysis on tubular joints) has also been carried out. An interim report of the program was presented at a siminar in November 1978 in Cambridge. In this paper, the results of tests on tubular joints, carried out in the Netherlands under the sponsorship of the ECSC, are discussed. Also, the results of additional work (strain measurements, crack measurements and calculations) sponsored by MaTS (Netherlands Marine Technological Research) are discussed. TEST PROGRAMME Forty tests on tubular joints were carried out by the Stevin Laboratory of Delft University of Technology and the Institute TNO* for Building Materials and Building Structures (IBBC-TNO). T- and X-joints were tested with chord diameters of 168, 457 and 914 mm. Table I gives a survey of the test programme. Three small T-joints were loaded with in plan8 bending. All other joints were axially loaded. One X- joint was tested with both chord and brace axially loaded. Some joints have one (Tl and Xl) or two (T2) additional unloaded braces. Two joints are tested with a random load and all others with a constant amplitude load (R = Fmin/Fmax = 0 or -1). Four joints have been tested in artificial seawater, one of them with cathodic protection. TEST SPECIMENS AND EQUIPMENT The test specimens were made by a manufacturer with experience in offshore structures. The specimens were welded in a fixed position with both chord and brace horizontal. Fig. 1 shows the weld shape of a large joint. The steel quality of the joints is in accordance with the standards mentioned in table II. Fig. 2 shows the test rig for the large joints. To avoid secondary effects, all supports in the test rigs are hinges. STRAIN MEASUREMENTS General Before starting the fatigue tests, the strain distribution and the maximum (hot spot) strains in the joint were measured under a static load smaller than the fatigue load. The results are related to the nominal strain in the loaded brace (SNCF = Elocal/Enominal).