Sunday, 22 May 2016

Conception of a physical model

Hello readers!

As part of my research project, it is expected to compare the numerical outputs obtained from a FE model with experimental data collected from a physical model test. This scheme is called to be the best weapon to evaluate the uncertainties inherent to the current analysis methodology. Such a physical model is be able to test different configurations in terms of clearance, fuel loading, ground accelerations, friction coefficients, etc. In addition, it allows to simultaneously testing more than one rack. This approach represents a big progress in relation to the up to date experiences outlined in the previous post, since here we evolve from a 1-rack to a 2-rack model. This upgrade will definitely take into account the coupling forces interacting between 2 free-standing racks.

Preliminary sketch of the physical model for rack seismic testing

Once identified the objective, the first step in the conception of a physical model is setting the scale. According to the Buckingham PI theorem, this kind of physical phenomenon can be described by only 3 physical variables. In other words, it is enough to set scale factors for 3 independent variables, considering that the scale of the remaining variables come straightforward. The final scale factors were can be explained as follows:
  • Scale 1/1 in densities due to the difficulties in scaling water density.
  • Scale 1/1 in accelerations due to the difficulties in scaling the gravity.
  • Scale 1/3 in geometrical dimensions due to the space and weight limitations.

The second step involves the structural design of the steel frame, the vibration table and the water pool. They should represent the features of the real ground location and therefore be infinitely stiff and undeformable. However, it would result too expensive so we just focused to avoid resonant amplification effects. To do so, a modal analysis of the whole vibration system was performed to assess the natural vibration frequencies and local stiffness were corrected where necessary.

In a final step, the dynamic analysis of the mocked racks was performed following the ENSA methodology. Maximum sliding displacement and impact forces were assessed and the numerical outputs were used to check the operability of the model and set its working range preventing dangerous impacts.

The physical model is still under construction but hereafter you can find a video of the preliminary alignment test.

See you down here soon!