behaviour and the second to a stiffer one. These settings are generally referred to as normal, comfort and sport. The latter is a prototypic vehicle, equipped with a novel sky-hook active suspension system, designed in the framework of a specific research project [15] and implemented on the vehicle by removing some components and replacing others, in a word modifying the mass, but also the stiffness distribution of the car, besides the damping one. The suspension behaviour is effectively governed by a controller which logic depends on several parameters as the car body heave, pitch and roll velocities and the wheel-body relative velocities. The car speed and the steering wheel angular position and velocity are other parameters that the control strategy can be influenced by, during operation. On these two cars, two different in-operation type of tests have been performed, (i) proving ground tests on two different road profiles, running at different car speeds and (ii) laboratory tests performed by using a four poster test rig. A four poster test rig basically consists of four actuators, one for each wheel, able to reproduce a running condition on a certain road profile on a car that actually stands still on them at null speed. On the test rig, the excitation signals sent to the four shakers have been firstly the ones needed to simulate the same roughness profiles considered in the road test campaigns. Secondly four uncorrelated random signals have been utilised, to have a useful reference condition in which the basic OMA assumption was verified. By OMA-processing the acquired signals in the different testing conditions and by comparing the obtained results, it is shown in what follows how this tool can be effectively utilised to verify the operation and the performance of the two suspension systems. OMA, indeed, becomes a useful designing tool in this field, since it can be utilised to check the actual behaviour of a prototype at an intermediate or final stage of its optimisation. SENSOR LOCATIONS 24 sensors on the car body (blue), including 4 sensors on the suspensions, body side 4 sensors on the suspensions, wheel side (black), 2 sensors on the front seat rails (red), 1 sensor on the steering wheel (green) Fig.1 The geometry of the analysed system with the sensor locations represented 2 TEST CAMPAIGN DESCRIPTION For both the cars considered, the semi-active suspension equipped car and the active suspension equipped one, four poster and proving ground tests have been performed, as anticipated. The test campaign on each car lasted about three days. Two different road profiles have been considered for both the test typologies, namely (i) the Blue-kay track and (ii) the Florida one. The car velocity on the former profile was 60 km/h, on the latter 80 km/h. In the case of four poster tests, also random excitation has been considered, as said. In this case four uncorrelated white sequences have been sent to the four hydraulic actuators. Different maximum amplitudes, in terms of shaker displacement have been considered, in particular 1, 3, 5 mm in the frequency range from 0 to 30 Hz and 10 mm in that from 0 to 15 Hz. The complete accelerometer layout consisted of 31 sensor locations. In each location the three components of the point acceleration have been measured, by using tri-axial piezoelectric ICP accelerometers. The output signals have been acquired and recorded by using LMS SCADAS front-ends, connected in master-slave configuration to achieve the needed number of channels and to a laptop PC, running the LMS Test.Lab analysis suite. In Fig.1, the geometry of the sensor layout is reported. The different colours identify the sensors placed on (i) the car body (containing also that on the suspension dampers, body side), (ii) the suspension dampers, wheel side, (iii) the front (driver and passenger) seat rails and (iv) the steering wheel. In Fig.2, several images of the tests campaign performed on one of the two cars are reported, showing the acquisition set-up. In particular some of the sensor locations listed above are depicted and the two front-ends utilised. 314
RkJQdWJsaXNoZXIy MTMzNzEzMQ==