No abstract

A new system of four synchronized eccentric mass vibrators has been used in forced vibration tests on dams. Corresponding calculations have been made using two-dimensional and three-dimensional finite elements, and for the gravity dam a proposed'l empirical formula for the fundamental period of vibration has been assessed. For Wimbleball Dam (buttress), it was found that buttresses are not completely independent; in fact, appreciable energy transfer occurs between them. From two visits, when the reservoir level was 21 m and 7 m below spillway level, it was shown that the fundamental frequency and response per unit force applied were reduced as the water level rose, whereas the transfer of energy between buttresses increased. Since the upper frequency range of the exciters is 10 Hz, only the resonance frequencies below this value could be excited. Five modal frequencies were measured at Llyn Brianne Dam (rockfill). For the finite element calculations, the variation of material properties with depth was assessed from P wave velocities measured at the surface in conjunction with empirical formulae. The three-dimensional calculations gave valuable indications of mode shape and the existence of a vertical mode in the early frequencies. A visit to Baitings Dam (gravity) suggests that of the three types measured, graviti dams are the easiest to understand in their dynamic behaviour.

No abstract

No abstract

Since the Paper was written, a third visit has been made to Wimbleball, further three-dimensional calculations have been made for Llyn Brianne using a new finite element mesh, and a successful series of tests has been carried out on a 180 m high arch dam in Switzerland. Wimbleball60. The second visit was planned for March 1979 when it was hoped that the reservoir would be at spillway level, but in fact it was 7 m below. With the continuing co-operation of the Consultants Rofe, Kennard and Lapworth, and the Contractors Bovis, together with the two Water Authorities, a third visit was organized in April 1980, when the water level was 0.75 m below the spillway. Before this visit, a small but important improvement was made to the excitation system by the manufacture of a set of weights half the size of the existing sets. The need for these had become apparent during previous tests where, at frequencies beyond about 7.5 Hz, a single set of the original weights took the generated sinusoidal force beyond the design limit. The use of these half-weights gave a better signal/noise ratio during the third visit to Wimbleball.61. Figure 22 is the revised version of Fig. 3, and includes the April 1980 results for frequency sweeps along the crest when buttresses 4, 8 and 12 were separately excited in turn. Fig. 3 shows the expected reduction in resonance frequency with increase of water level, as well as reduction in response per unit of force applied. Fig. 22 shows that the response is further reduced by the increase in water level from 7 m to 0.75 m below the spillway, but the April 1980 curves show a small increase in resonance frequency rather than a further decrease to that observed between October 1978 and March 1979. 62.It was assumed (4 16) that the decrease in amplitude and frequency of the responses between October 1978 and March 1979 was caused by the buttresses behaving more as a monolith than as individual entities. It was expected that these frequency and amplitude trends would continue with the continued raising of the water level in the reservoir with, perhaps, an enhanced effect caused by the added mass of water participating in the vibrations. The results