On Design of Ultrasonic Transducers and
Accuracy of Velocity Measurements
Nesvijski E. G.
Center of Technology, UFSM, Santa Maria, RS, Brazil
Corresponding Author Contact:
Email: email@example.com, Web: http://w3.ufsm.br/edouard/
Design of innovative ultrasonic transducers was recently discussed in the NDTnet. The majority of ultrasonic transducers presently manufactured are using flat-plate-contract (FPC) for contact with the testing material. Some coupling is necessary to provide acoustic contact with the material surface. One of the main parameters for the material characterization is wave velocity. Velocity values are used for evaluation of elastic constants, strength, or flaw detection. That is why accurate velocity measurements are very important for the NDT needs. Problems interconnecting design and accuracy of velocity measurements are going to be discussed in this papers.
Analysis of standard approaches
In the majority of cases velocity measurement means registration of the wave time of flight t between the "in" and "out" points. Then this time value has to be corrected taking into consideration "dead" time t0 of the equipment set-up. Velocity values are calculated on the base of distance L between the "in" and "out" points. A simplified concept of wave propagation is used here. It is very difficult to agree with this concept for the majority of cases. Nevertheless many standards and investigations are based on this concept and present velocity table values. Moreover, distance between centers of the transducers plates is accepted as distance L. This approach seems to be very artificial and inappropriate. It can be explained.
On one hand, the FPC transducers have quite complicated distribution of shifts on the plate surface depending on the transducer design and spectral pattern of acoustic (ultrasonic) signal. In accordance with this pattern the acoustic signal is transferred to the testing material by the entire plate surface. That is why it is difficult to expect that the concentration point of ultrasonic energy coincides only with the center of the transducer plate.
On the other hand, the FPC transducers have their own directional pattern, which, as a rule, has a main "petal" perpendicular to the plate surface and additional "petals" with differently oriented to the plate. That is why "linear geometrical" understanding of wave propagation is rather questionable. It is also possible to add that rough surface materials (stone, concrete, brick, some composites) make energy transfer randomly distributed on the surface in spite of the coupling.
Some typical schemes of ultrasonic testing are given in the figure below.
Figure. Typical schemes of ultrasonic testing, where L - distance between the transducers plates centers, D - diameter of the transducer plate, m and n - distances between the transducers centers and the material corner: a) "one axis through testing", where transducers are on two material surfaces; b) "angle testing"; c) "one surface testing"; d) "curve surface testing".
It is possible to assume that "in" and "out" are not concentrated in the center of the transducer plate, but distributed on the plate with uniform probability (this approach is usually accepted when there is no idea about type of distribution). In this case it is possible to calculate maximum values of velocity (1) for "one axis through testing" (a):
then minimal values of velocity (2) for "one axis through testing" (a):
It is also possible to calculate velocity variation (3):
It is possible to calculate maximum values of velocity (4) for "angle testing" (b):
and minimal values (5):
It is also possible to calculate velocity variation (6) for this case:
It is possible to calculate maximum values of velocity (7) "one surface testing" (c):
and minimal values (8):
then velocity variations (9) for this case:
Calculations for case d) differ very litter from calculation for the case a) and depend on the relation between plate diameter and the diameter of the curved surface.
Example of calculations of (1) - (9) for the following data using metric system
(diameter and distance in meters and time in seconds):
D = 0.035; L= 0.15; t = 0.000040; t0 = 0.0000018
Calculated values for the cases a), b), and c) are given in the Table:
The analysis of the FPC transducer design influence on the calculated velocity values present an extremely specific view on the existing problems. This is made to underline inaccuracy of measurements for velocity evaluation by the FPC transducers. The presented artificial models were introduced to explain possible principal evaluation errors. Unfortunately, many users close their eyes on this uncertain situation of velocity evaluation and present velocity values as indisputable constants.
There are a lot of other problems that are not discussed in this paper connecting the FPC transducer design with the problems of surface waves propagation, guided wave propagation, wave diffraction and interference on inhomogeneity of materials, which have their own important impact on velocity evaluation.
It seems that application of laser ultrasonic, air array, and dry point contact (DPC) transducers are free of some disadvantages that are demonstrated for the FPC transducers in this paper.
/DB:Article /AU:Nesvijski_E /CN:BR /CT:UT /CT:transducer /ED:2000-02