MHz-ultrasound in air : a physical miracle?


In the case of non-contact ultrasonic testing no use is made of a coupling gel nor is the sample immersed in liquid. Air is the only coupling medium. This creates problems which at first sight seem to be insurmountable. Not only there is a big difference in acoustical impedance between air and sample, but also the extreme high absorption of ultrasound by the air.

The low acoustical impedance of air

The acoustical impedance (A.I) of a medium is a measure for the resistance of a medium against penetrating sound. It is comparable to the concept of electrical impedance in the case of electronic circuits. The definition is as follows:

Z = acoustical impedance = density x ultrasonic velocity

Z is expressed in kg/m2.s [1 kg/m2.s = 1 Rayl; 106 kg/m2.s = 1 MRayl]

  air Z = 420 Rayl
  water Z = 1.500.000 Rayl = 1,5 MRayl  
  aluminium Z = 17 MRayl
  steel Z = 45 MRayl

If sound is normally impinging from a medium with acoustical impedance Z1 on a medium with acoustical impedance Z2, by the difference in A.I., only a fraction of the sound will penetrate in the 2de medium; usually the major part of the sound is reflected.

The transmitted fraction of sound energy is a function of both Z1 and Z2 and is defined as the transmission coefficient T. In the case of normal sound incidence, T is given by

Transmissioncoefficient T =  4.Z1.Z2 (Merk op dat T = 1 indien Z1 = Z2)
  medium 1  medium 2 T loss [dB]  
  water  aluminium  0.30 5.3
  air  aluminium  0.0001 40.3
  water  steel 0.13 9
  air  steel 0.000037 44.5

If sound is launched from water into aluminium, 30% of the sound is transmitted and 70% is reflected. However, if the same sound beam enters the aluminium from air, only 1/10000 of the sound energy is transmitted into the aluminium, which is a loss of 40 dB!!!

With respect to steal, the difference is still more striking: from water 13% of the sound energy can penetrate into steal, but from air less than 0.004%, which implies a loss of 44.5 dB!!!

Of course, the same losses occur when the ultrasound leaves the medium to continue its path in the air! Hence, total losses due to the bridging of A.I. differences can mount up to 80 dB or more!!!


An additional problem is created by the strong absorption of ultrasound in air, at least for frequencies higher than 250 kHz.
As a consequence, ultrasound of e.g. 1 or 2 MHz can propagate in air over a distance of not more than a few centimeter... However, in all examples presented on this website, these few centimeter are sufficient to enable a powerful non-contact ultrasonic measurement technique.

To illustrate the absorption, the animation below shows the evolution of an amplitude distribution of an ultrasonic field created by a 2 MHz transducer (type ULTRAN) with an active diameter of 25 mm as a function of the distance travelled in air. The measurements have been performed by means of a needle transducer (type DAPCO), a hydrophone which is designed to operate in liquid media but, thanks to its high sensivity, also can measure sound fields propagating in air.

As one can see, the ultrasonic field is collapsing gradually over a short distance of some centimeter. In addition, it should be noted that the indicated amplitude units (the same scale has been used in all frames) already are the result of a logaritmic operation. Hence the losses in dB are much greater than the scale is suggesting. Over a distance of 1 and 5 cm a loss of almost 40 dB was measured.

Taking into account all these difficulties, it hardly seems possible to consider contactfree ultrasonic NDT-techniques operating at frequencies higher than 0.25 MHz... It is demonstratred abundantly in the present website that the contrary is true!

Print Print Laatste update 23 mrt 2012