[ Ultrasonic Application Workshop ]
[ Follow Ups ] [ Post Followup ] [ Help ]
Re: Model generated by an inverse Fast Fourier Transformation
Posted by: Hermann Wuestenberg , E-mail: Hermann.Wuestenberg@bam.de, on May 16, 1997 at 16:37:18:In Reply to: Model generated by an inverse Fast Fourier Transformation posted by : Rolf Diederichs on May 12, 1997 at 10:44:23:
: Concerning the article "Approximative Modeling for the Practical Ultrasonic Inspections" by H.Wüstenberg, A.Erhard.
: The A-Scans are in the model generated by an inverse Fast Fourier Transformation.
: What are the advantages and drawbacks of this approach?: Rolf Diederichs
Pleas find here an first attempt of answering the following question:
1. The model can be linked to well known pulse spectra of UT-probes,
which commonly are available as data measured and also - depending
from the manufacturer - calculated by a KLM-quadrupol model or
similar equivalent circuits for the probe. The influence of the
bandpass filter characteristic of the receiving amplifier within the
UT-equipment can easely be studied.
2. The phase behaviour of reflection or mode conversion factors and of
the point source directivities for the different wave modes and its
influence can be considered without any difficulty
3. This enables us to cut the totality of the process into different
parts and to multiply or add them in order to get the transfer
function of the whole system.
4. The necessary number of frequency samples to be calculated can be
adapted to the extension of the time domain to be considered for the
A-Scan. E.g. many of the examples in the above mentioned contribution
are using 64 samples in the frequeny domain and 128 in the time
domain, which corresponds e.g. to 16 mm sound path length. That means
that the use of the inverse FFT enables us to reach very short
computational times.
The drawbacks have to listed by someone else favorising a time domain
approach.
greetings
and Frohe Pfingsten
Hermann Wuestenberg