Why coherence needs two hammer hits?

Hi,

I am curious how is the coherence calculated in Dewesoft, Modal test toolbox. According to your website, it is a auto/cross correlation function for input and output. Therefore I don't understand why it is "1" for all frequencies for the first hammer hit, and only valid values appear after the second hit.

I can see that you write on your tutorial that:

But what is the reason for this?

Dear Dan,

this is also described in an article on our webpage: https://dewesoft.com/daq/what-is-modal-analysis#modal-test-results

Averaging of Test Measurements

It is important to average multiple measurements taken at the same DOF locations. 

Why? Each measurement includes some random noise that can affect the determination of the resonance frequencies of the system and mode shapes. By averaging multiple measurements, the random noise components can be reduced. 

A random noise will be reduced more if a higher number of averages are chosen, and it is up to the engineer to decide a preferable number of measurement blocks to average. This decision is a balance between the desired quality of the results and the required test time. 

In general, a rule of thumb is to average across:

• 32 to 64 averages for shaker testing, and
• 4 to 8 averages for hammer impact testing.

Normally Modal Test application software lets the user specify the number of averages to use at each DOF measurement. For normal impact excitation testing (with one impact per measurement block), this means that the user should impact the set of DOFs locations each equal to the number of averages specified.  

Looking at Coherence functions is a primary validation tool when performing Modal Testing, but don’t be misled by a coherence value of 1 after a single measurement has been taken.

By definition, the coherence function is based on multiple measurements and therefore only provides useful data after averaging has been performed.

Averaging should be done on the data for the individual DOF locations and not across all DOFs.

Best regards,

Jernej Sirk