The authors of the journal article do not understand Modal analysis and should not have included the deformation in Tables 7 and 8. Have you taken the free course on Modal analysis yet?
There are many ways you can show that corroded pipes have a frequency shift compared with uncorroded pipes. You can use Modal analysis, Harmonic Resposne analysis or Random Vibration analysis. Modal does not take any input load, which is why the deformation output magnitude should not be used the way it was in that paper. The modal deformation is only to understand the mode shape, the magnitude is arbitrary.
Random Vibration analysis takes a PSD input load and computes the 1 sigma deformation in the structure. You can plot the response (output) PSD of a vertex. Take the free course in Random Vibration.