William B. Streett
 03:19 Feb121998 Re: impactecho : Recently I read a new paper by Sansalone, Lin and Streett : in the ACI Materials Journal NovDec 97 where 'a procedure : for determining pwave speed in concrete ...' is decribed. : At one point, a difference between the pwave speed, and : a socalled "impactecho pwave speed" of about 4 % is : mentioned and a recent paper of Lin and Sansalone is : cited for explanation. Since this paper is not available, : I reread a paper of 86 by Carino, Sansalone and Hsu (flaw : detection in concrete ...) where they find a difference : of about 10 % between the two velocities, but cannot : find an explanation for that.: My question is, what is the reason for this phenomenon, : what is the physical effect standing behind this differnce? : Who can give an explanation? Mr. Weiler has asked about confusing statements in the impactecho literature about the 4% difference between the true Pwave speed in concrete and what is sometimes called the 'apparent Pwave speed in a plate'. If one considers multiple Pwave reflections in a simple plate, the expected relationship between the wave speed Cp, the thickness T and the frequency f is: f = Cp/(2 T). This is a simple statement that the period of wave reflections is twice the thickness divided by the wave speed. In the early work on impactecho it was assumed that this relationship explains the impactecho response of a plate structure. However, in studies of the timedomain waveforms from early impactecho tests, it was observed that successive Pwave arrivals were slightly delayed, and as a result the frequency of multiple reflections is slightly lower than that calculated by the equation above. It was not until after extensive computer simulations of stress wave propagation and reflection in a variety of geometric shapes were carried that the developers of impactecho realized that multiple Pwave reflections excite certain transient modes of vibration, and that the correct relation between the wave speed, the frequency of the first mode, and a characteristic dimension A is f = bCp/(2A), where b is a "shape factor". For plates the mode of vibration is a thickness mode, the characteristic dimension is the thickness T, and the shape factor b = 0.96. The frequency of the first mode of thickness vibration in a plate is 4% lower than the frequency of multiple Pwave reflections predicted by the simple equation f = Cp/(2T). The values of b for different geometric shapes have been calculated from computer simulations using a 3D, dynamic, finite element simulation method, and they have been verified by repeated tests both in the laboratory and in the field. (The finite element codes used in the simulations are the DYNA3D and related programs developed at the Lawrence Livermore Laboratories in California by G. Goudreau, J. Hallquist and others.) For more information see references [39, 40, 41] in http://www.ndt.net/article/0298/streett/refer.htm See also Chapters 4, 21 and 22 of the book by Sansalone and Streett: http://www.impactecho.com/ImpactEcho/bullbrie.htm If one measures the Pwave speed using two transducers on the surface of a concrete structure, the true Pwave speed, Cp, is obtained. If one observes the thickness frequency of a solid plate of thickness T, and calculates the wave speed from the simple equation Cp = 2fT, the result is what we now call "the apparent Pwave speed in a plate". It is equal to 0.96 times the true Pwave speed. In impactecho tests on beams and columns, the same equation f = bCp/(2A) applies. In this case A is a characteristic crosssectional dimension, and the value of b depends on the crosssectional shape. For circular columns, for example, A is the diameter and b = 0.92. For a square column, A is the length of the square, and b = 0.87. For a rectangular column, A is the dimension in the direction of the impact, and the value of b is a function of the aspect ratioof the cross section, and its value ranges from about 0.75 to 0.96. The latter value is the limiting value as the shape approaches that of a plate (lateral dimensions at least 5 times the thickness). There is indeed some confusion on this matter in the published papers on impactecho, because it was not fully understood until after the early papers were published. In addition to the explanation in the book on impactecho, an explanation can be found in a recent paper by M. J. Sansalone, entitled, 'ImpactEcho: The Complete Story', in the November/December 1997 issue of the ACI Structural Journal, pp. 777786. Prof./Dr. William B. Streett Cornell University Ithaca, NY 12 February 1998
