Recent observations of windturbines suggest that resonance is a serious issue [7–10]. Field measurement indicates that the resonance issue is caused by thechange of its foundation stiffness after several years of service [2].Therefore, the design problems are: (1) prediction of the longterm tilt in the wind turbine due to the change in the soil proper-ties owing to irregular and asymmetric cycling, (2) long term shiftin natural frequency of the systemand howclose can the frequencybe with respect to the forcing frequencies. This is particularly im-portant for ‘‘soft–stiff’’ design as any increase/decrease will impacton the forcing frequency causing higher fatigue damage [11].Due to its previous successful application in OWTs, monopile isstill the prevailing foundation option for supporting OWT for wa-ter depth of less than 30 m in standard soils (sand, soft and stiffclay). More than 75% of the OWTs in Europe (i.e., UK, Denmark,Germany, and Netherlands) are supported on monopiles [1]. Thispaper therefore investigates the long-term dynamic behavior ofmonopile supported OWT through a series of small scale tests. Thisis in contrast to the tests carried out in other researchers’work, seefor example Refs. [12,13],where the dynamics of the problemis notconsidered and fatigue type problem is investigated.Derivation of the correct scaling laws constitutes the first stepin an experimental study. Every physical process or mechanismcan be expressed in terms of non-dimensional groups and thefundamental aspects of physics must be preserved in the designof model tests. In this paper, the main principle of scaling relatedto OWTs comprises of geometrical and mechanical similaritiesbetween scaled model tests and prototype. And this part has beendiscussed in detail in Refs.
[1,14,15] and the readers can refer tothose publications for more information.Dynamic loading system Previous research [1,14–16] on dy-namic testing as shown in Fig. 3(a) used an actuator to apply all thecyclic and dynamic loads at one location (denoted by yc in Fig. 3(b))and the methodology to find out the load is shown in Fig. 3(b) [1].After applying a user defined number of cycles, the actuator is dis-connected to obtain the natural frequency through free vibrationtest. Then, the actuator needs to be reconnected to apply the nextset of cyclic loads. This causes not only some amount of inconve-nience to the testing but also unavoidable disturbance to the soilaround the foundation. Furthermore, the actuator can only provideone directional regular cyclic loads. But in reality, due to the mis-alignment of wind and wave, the cyclic load is always multidirec-tional. This led to the development of an innovative cyclic loadingsystem used in this paper.The physics behind this innovative device is simple and fol-lows the concepts of centripetal forcing, i.e., for a body of massm, which is rotating about a center in a circular arc of radius rat a constant angular frequency ω, the mass will exert an extraforce acting towards the center of rotation in the magnitude ofFn (Fn = mrω2, see Fig. 4(a)). Figure 4(b) shows the final design ofthe device and it produces a harmonic loading in two perpendiculardirections when two masses (m1 and m2) complete one revolution  (shown in Figs. 4(c) and 4(d)). Thus, the simple and economic de-vice is comparable to an actuator, but with an obvious advantageof producing two directional two-way cyclic loading, which indi-cates amisalignment of wind and wave in real field. The frequencyff (Hz) of the applied loads is determined by the voltage U (V) thatdrives themotor of the device. The amplitude and frequency of thecyclic loads can be adjusted by replacing the masses (m1 and m2)on the gears and the voltage output.Test preparation and proceduresDynamic tests are carried outin a model tank (1150 mm long, 950 mm wide, and 600 mm high).The soil used in this experiment is Red Hill 110 silica sand, whichis quite typical of that encountered in the North Sea, and it is rep- Table 1Properties of the red hill silica sand.Items ValuesSpecific gravity Gs 2.65Median particle diameter D50/mm 0.144Internal friction angle/(°) 36.0Dry unit weight/(kN • m3) 16.8Maximum void ratio emax 1.035Minimum void ratio emin 0.608Relative density 0.63Shear modulus G/MPa 10.0Table 2Detailed parameters of the model turbine.Part Length/mm Diameter/mm Thickness/mm Weight/kgModel pile 450.0 43.0 2.0 3.3Model tower 1000.0 43.0 2.0 2.2Top mass – – – 1.8resentative of the soil along the southeast coastline of China [3].Some basic properties of this sand is given is Table 1. The sand bedis prepared by pouring sand from a hopper, maintaining the samerate of flow and the same height. The final thickness of the sandbed is approximately 500 mm. Relative density of the sand bed is63% asmediumdense sand. Shearmodulus of the sand bed G is ob-tained as approximately 10 MPa by the in-situ shear wave veloc-ity method. Based on the scaling laws, set-up of the experiment isshown in Fig. 5. Themodel turbinemainly consists of 3 parts,moredetailed information is given in Table 2. The innovative cyclic load-ing device is connected to a volt generator, and mounted on top ofthe tower. 单桩海上风力发电机英文文献和中文翻译(2):http://www.751com.cn/fanyi/lunwen_36963.html