PSI - Issue 29

3

Gian Paolo Cimellaro et al. / Procedia Structural Integrity 29 (2020) 142–148 Domaneschi et al. / Structural Integrity P o edi 00 (2019) 000 – 0

144

a

f

(2)

d  =

g

The step-by-step displacement and velocity are found by a Matlab code opportunely set (https://it.mathworks.com/products/matlab.html); in this way the va lues found for  d can be rela ted to the corresponding va lue of the input velocity. A specia l attention has been pa id to the measuring uncerta inties and errors occurring in the test. One of the most common errors arises by the misa lignment between the measuring axes (see Cimellaro and Domaneschi 2020), i.e. the norma l and the tangentia l ones. Figure 2 shows the possible misa lignment between the two axes, the accelera tion va lues recorded on X and Y axes are the components of gravity acceleration due to the a ccelerometer’s misalignment angles.

Figure 1 . Scheme of the dynamic test .

Figure 2 . Misalignment errors

Named a fx and a fy the friction accelera tion components a long the axes X and Y , respectively, the Error (E) which can affect theassessment of  d can bequantified by Equation (3):

2 a a a a − + f fx

2

(3)

fy

E

=

f

3. The performed test 3.1. The shaking table

The shaking table, shown in Figure 3a, consists of steel profiles connected through transversa l rectangular sections and an upper a luminium platform. Upon the steel profiles there are a luminium guides a llowing the motion, a long the longitudina l direction, of sliders that support two 600x500x10 mma luminium platforms. Each track has its own pla tform, which is moved by a linear electric actuator anchoredunder it. On the sma ll pla tforms, other two tracks and platforms are fixed. For the transversa l motion other two linear electric actuators are anchored under the a luminium platforms. A more deta iled description of the shaking table can be found in Cimellaro and Domaneschi (2020). The tuning of the motors, i.e. the checking of the initia l configuration of a ll the control parameters, is made through the software LinMot-Ta lk, which is a lso used to switch on the actuators and to bring them in the home position. The seismic input is sent to the shaking table through a myRIO device manufactured by Na tiona l Instrumen ts. This device is physica lly connected to the motors’ drivers and a lso to an accelerometer, which is located on the platform and a llows catching the actua l response of the system. The LabView code is used to set the input and output sampling ra tes, to generate a sinusoida l seismic signa l or to load a rea l one, to sca le it, to start and stop the motion and fina lly to compare the data obta ined from the accelerometer with the theoretica l ones. The accelerometers and the load cell are connected each other to