NDT.net • April 2004 • Vol. 9 No.04


A. M. Abdelhay, and I. M. I. Mubark
Associate Professors, Production Engineering Department
Faculty of Engineering, Helwan University, Helwan, Cairo 11792-Egypt.

Corresponding Author Contact:
Email: Abdelhay1953@Yahoo.Com


This experimental study used ultrasonic pulse - echo as a non-contact technique for measuring the surface roughness parameters of machined parts made from steel materials. Ultrasonic parameters such as the attenuation coefficient and reflection coefficient were employed to surface roughness. In concept, this study also, can be used to show the influence of rough surface on the measurements of the ultrasonic technique as a nondestructive testing NDT. It is important in ultrasonic NDT to understand quantitatively the influence of surface roughness of a test object on an ultrasonic echo signal.

Measurements were made of surface roughness, ultrasonic wave attenuation, and the reflection coefficient of backwall echo sound waves. The following results were found, first, surface roughness greatly degraded the normal incident and reflected echoes from the backwall of tested surfaces, and such degradation or attenuation is in direct proportional to it. Second, the ultrasound reflection coefficient is inversely proportional to the degree of surface roughness. Either of the ultrasonic attenuation coefficient or the reflection coefficient can be faithfully describe and estimate the level of surface roughness with small error.

KEYWORDS: surface roughness; ultrasonic testing; ultrasonic attenuation; ultrasonic reflection coefficient; NDT, quadratic models.


Surface topography is of great importance in specifying the function of a surface. A significant proportion of component failure starts at the surface due to either an isolated manufacturing discontinuity or gradual deterioration of the surface quality. Typical, these problems of lower surface integrity, lead to in service stress corrosion and fatigue failure. The most important parameter describing surface integrity is surface roughness.

In the manufacturing industry, surface must be within certain limits of roughness. Therefore, measuring surface roughness is vital to quality control of machining workpiece, especially if it is non - contact method compared to the direct conventional method; which uses stylus type devices. In the direct contact method, measurements are obtained using a stylus drawn along the surface to be measured: the stylus motion perpendicular to the surface is registered. This registered profile is then used to calculate the roughness parameters. This method requires interruption of the machine process, and the sharp diamond stylus may make micro-scratches on surfaces (i.e. destructive damaging effect).

Several attempts [1-5] have been reported of studying non-contact techniques for the assessment of surface roughness. Of these worth mentioning works, the study by Hilton [4], where he used two orthogonally polarized laser beams to produce two independent speckle patterns that are imaged and auto-correlated to deduce the roughness of the surface being illuminated. Roberts and Briggs [5], have used several techniques such as the acoustic emission and incident X-ray scattering for the characterization of surface roughness and sub-surface damage. Bilgen and Rose [6] theoretically analyzed and discussed the problem of the signal noise of back scattering generated by rough surface using modeled techniques. Their study showed that early time variance in the ultrasonic signal, is independent of the transducer type and its radius. Also, they reached a conclusion, that by using a focused ultrasonic probes lead to a reduction in the variance in the signal due to scatters in the focal zone of the scanned zone within the tested material.

It is the purpose of this paper to experimentally study the correlation between both the acoustic attenuation and reflection coefficients of longitudinal ultrasonic waves propagated in a steel bar having different degrees of surface roughness.


A stress longitudinal wave propagates or traveling through engineering materials will lose energy for a variety of reasons (i.e. . Surface rough of the contact surface with the ultrasonic probe is one reason). This behavior can account for a loss in amplitude as will as for change in the reflected or echo stress waves. Such energy loss of the stress wave pulse is treated very thoroughly in several of the references [7-10].

There are three basic processes that account for loss of stress pulse energy, namely, beam spreading, absorption, and scattering [11]. The interest here is focused on the stress wave attenuation; which serves our research purpose. Attenuation is generally expressed in the form

P = P0 • eaL (1)

P0 the original stress wave pressure at a source or reference location
P the pressure level at second reference location
a the attenuation coefficient (dB/mm)
L the distance between the two reference locations (mm)
The relative sound pressure level (SPL) of a propagating wave is

SPL= 20 log ( P0 / P ) (2)

Considering two points in the path of an ultrasonic stress wave, the sound pressure level loss for a wave passing between point 1 an 2 is given by

SPL1 - SPL2 = 20 log ( P1 / P2 ) (3)

Namely, the two successive reference locations of interest are the incident contact surface and the backwall surface of the tested workpiece. These two surfaces are L distance apart (i.e. the thickness of tested workpiece).

From Eq. (1) and (3) one can write

a L = 20 log ( P1 / P2 ) (4)

If the ultrasonic probe used for the acoustic measurements is of transmitter/receiver (T/R ) type, the stress wave travels twice the "L" distance, so Eq. (4) becomes

a = 10/L log ( P1 / P2 ) (5)

This previous formula Eq. (5)- is used in the calculations of the ultrasonic attenuation coefficient due to surface roughness, keeping all other variables unchanged.


Two 30 mm square cross-sectional bars were cut to 200 mm length each. These bars which are designated as A and B, were from 0.25%C hot rolled steel bar (St. 42). Experiments and measurements are carried - out in three phases. Namely, the machining operations phase, the surface roughness measurements phase, and finally the ultrasonic measurements phase.

3.1 Preparation of Tested Surfaces
Generation of machined surfaces with different values of surface roughness was achieved using shaping machine fitted with sharp nosed - bend tool ( 8° positive rake angle, and 4° relief angle with 90° nose angle). Machining operations were performed for 20 X 30 mm top surface area, along the bar. Different feeds ranging from 0.2 to 1.2 mm/stroke were used for bar "A" with cutting speed of 18 m/min. While, bar "B" was machined with different cutting speeds ( from 3 to 28 m/min) and constant feed of 0.3 mm/stroke. A depth of cut of 0.5 mm was maintained and used for all machined surfaces.

3.2 Assessment of Surface Roughness
Once all surfaces were machined and thoroughly cleaned, they were measured for defining their surface roughness parameters or descriptors. Surtronic3+ - model surface roughness tester was used in these measurements. Three statistical surface roughness descriptors or parameters that give the average behavior of the surface height of the machine surface asperities were recorded (at Lc =0.8 mm and Ln =0.4 mm. These surface parameters are: the arithmetic average roughness Ra, the root mean square roughness Rq, and the peak-to-valley height or roughing depth Ry. ). These parameters are of the most effective surface roughness measures that are most commonly adopted in general engineering practice[12]. These surface roughness parameters give a good general description of the height variations in the tested surface.

3.3 Ultasonic Measurements
Ultrasonic inspection of the different machined surfaces is performed using the standard procedure of " Normal beam Pulse - Echo " procedure, using the measuring setup shown in Fig.1. A straight beam of 3.5 MHz frequency transducer probe (Krautkramer 12.5 mm effective diameter) was used in these measurements. This integrated data acquisition and waveform analysis system consists of transmitter/receiver (T/R) ultrasonic instrument (NDT 801D Model), a digital storage oscilloscope (TEKTRONIX- TDS 210 -60MHz), with alphanumeric CRT read out and a software package (WaveStar- Ver 2.1) for analyzing the acquired ultrasonic waveform signal.

A coupling medium is normally required at the interface between the probe and the tested surface, in order to introduce the ultrasonic energy through them. Petroleum jelly was used as a thin layer coupling medium; which does not significantly contribute to any energy losses [11]. A 500 grm pressure force was maintained in holding the probe against the tested surface for all test runs.

Fig 1: Typical laboratory setup for the ultrasonic and data acquisition system.


Results of generating machined surfaces with different levels or degrees of surface roughness are shown in Fig. 2. These plots show the influence of increasing feed motion (Fig. 1a) on surface roughness. Also, it indicates the reduction of surface roughness by the increase in cutting speeds (Fig. 1b). What is noticed that the rate of change in the average roughness Ry is relatively quite higher compared to Rq and Ra for both feed and speed. For this reason, Ry is going to be used in this work as a sensitive surface roughness descriptor.

These numerical values for surface roughness parameters, are then used as independent variable, upon which the acoustic ultrasonic attenuation coefficient will depend. Graphs in Fig. 3. illustrate the pulse - echo waveforms recorded for different surface conditions. Backwall echoes for the as received surface (Fig. 3b) shows a little energy loss or pulse attenuation due to the tested materials' properties, and partially due to the initial surface roughness of the tested surface. In one hand fine finished surface (Fig. 3a) shows gradual loss in sound waves compared to rough finished ones (Fig. 3c). On the other hand, the same surfaces examined from the opposite or back side, shows multiple reflections accompanied the original echoes; which is clear at high surface roughness levels ( See Fig. 3c l right graph).

Using the results of machining operations (i.e. Fig. 2) and the calculated attenuation coefficients " a " of the ultrasonic backechoes (using Eq. (5)), graph of Fig. 4 was obtained. Backecho amplitudes of the second, third and fourth echoes were employed in these calculations of a's. Averages of three ultrasonic measurements were used in the calculations of the ultrasonic attenuation coefficients.

It is evidence from the graph of Fig. 4, that increasing in surface roughness leads to an appreciable increase in attenuation of the ultrasound waves. This behavior can be attributed to the severe loss in stress wave energy upon transmission across the contact surface asperities. Also, the thickness of the coupling medium filling the gap between the surface of the transducer probe and the tested


Fig 2: Generated different levels of surface roughness through (a) changing machining feeds, and (b) changing machining cutting speeds.

Machined Surface Contact Side

Back Surface
( a) Test Surface no. A1 and no. A1 Bk

Machined Surface Contact Side
( b ) As Received Surface.

Machined Surface Contact Side

Back Surface
( c ) Test Surface no. A8 and no. A8 Bk.
Fig. 3. The Pulse - Echo of the ultrasonic waveforms, for (a) a fine finished surface, (b) the as -received surface, and ( c ) a rough finished surface.

surface, is increased with rougher surface than with finer finished ones. Such gap results in additional loss in sound wave energy transmitted to the tested material, causing more attenuation to take place.

For each test specimen, the amplitude of the back-wall echo received from its rough surface was compared with the amplitude of the backwall echo received from the reference surface of the as received surface in order to determine reflection coefficient " a ". This reflection coefficient is defined as the ratio of the amplitude of the backwall echo received from the rough surface of the specimen under test divided by the amplitude of the backwall echo received from the as - received surface (reference surface).

Fig 4: The surface roughness effects on attenuation of ultrasonic waves. Fig 5: Ultrasonic wave reflection coefficients as a function of surface roughness.

Figure 5. Shows the results of the relationship between the reflection coefficient of sound waves and surface roughness. The results of this relation indicate that as the value of surface roughness increases, the value of the relative backwall echo amplitude at a given probe frequency (i.e. 3.5 MHz.) decreases. The possible reason for the observed decrease in the amplitude of ultrasonic back-wall echoes with the increase in surface roughness of the test specimen could be the interaction of impinging ultrasonic wave with the roughness (surface dilutions) on the specimen surface. Consequently, increased loss in wave energy due to the increase in reflection/scattering of the ultrasonic wave while transmitting took place, hence, resulted in the observed decrease in echo-amplitude.

A set of quadratic models for minimum error, was developed to estimate the roughing depth Ry as a function of each ultrasound parameter " a " and " b ". These models have the form:

Ry = C0 + C1a + a2 microns (6)

with 0.96 correlation coefficient, and

Ry = C0 + C1b + b2 microns (7)

with 0.91 correlation coefficient.

The model's coefficients; which are experimentally determined (i.e. Ci's) are given in Table 1.

Ultrasound Parameter Co C1 C2
a 0.0624 0.0008 -3 E-6
b 0.9628 0.9628 1E-5
Table 1. Quadratic models coefficients

To verify, the relationship of Fig. 5., a surface was machined with arbitrary cutting condition ( feed = 0.4 mm/stroke, and cutting speed = 13 m/min); which resulted in Ry = 124 microns. Ultrasonic measurements showed a reflection coefficient of 0.63 corresponds to Ry' of 118 microns (approx.) by using Eq. 7. The error was 4.7 % which is experimentally reasonable.


The effects of surface roughness of different machined surface were experimentally investigated in relation with some ultrasonic parameters such as the attenuation coefficient and the reflection coefficient. It was found that in one hand, surface roughness is greatly degraded the normal incident and reflected echoes from the backwall of tested surfaces, and they are in direct proportional to each other. On the other hand, the ultrasound reflection coefficient is inversely proportional to the degree of surface roughness. Either of the ultrasonic attenuation coefficient or/and the reflection coefficient can be faithfully describe and estimate the level of surface roughness with small allowable experimental error. As for ultrasonic measurements NDT technique, great attention must be put on the quality of the surface to be tested, since , surface roughness can render the process ineffective and inaccurate.


We wish to express our gratitude and thanks to our colleague Dr. O. M. Dawood, Associate Professor, Production Engineering Department, Helwan University, for his cooperation in carrying out the ultrasonic measurements, and for his sincere guidance and his many helpful discussions.


  1. Wadaka, S., Kimura, T. and Kameyama, S., " Analysis of bottom echo through rough test surface using equivalent circuit of ultrasonic transducer," The 1st Int. Conf. on NDE in Relation to Structural Integrity for Nuclear and Pressurized Components , 20 - 22 October 1998, Amsterdam, Netherlands.
  2. Krolikowski, J. and Szczepek, J., " Assesment of tangential and normal stiffness of contact between rough surfaces usinhg ultrasonic method," Wear , 160 , pp 253 - 258 (1993).
  3. Ogilvy, J. A. and Culverwell, I. D., " Elastic model for simulating ultrasonic inspection of smooth and rough defects," Ultrasonics, 29, pp 490-496 (1991).
  4. P. J. Hilton, " Image surface roughness using correlated speckle grain pairs," Report No. DICTA/IVCNZ97, Massey University, New Zealand, pp 349-354 (Dec. 1997).
  5. Roberts, S. G. and Briggs, G. A. D.," Characterization of surface roughness and subsurface damage," Final report on EPSRC, Oxford University ( Dec. 1996).
  6. Bilgen, M. and Rose, J. H., " Focused Ultrasonic Probes and the Effects of Surface Roughness on Material Noise," in Review of Progress in Quantitative Nondestructive Evaluation, edited by D. O. Thompson and D. E. Chimenti (Plenum, NY), Vol. 13B, pp. 1769- 1776 (1993).
  7. Nagy, P. B. and Rose, J. H., "Surface roughness and the ultrasonic detection of subsurface scatterers," J. Appl. Phys. 73, 566-580 (1993).
  8. Russell, M. D. Neal, S. P. and Boote, E. J., "Experimental estimation of the longitudinal-wave backscatter coefficients for ultrasonic interrogation of weak scattering materials," J. Acoust. Soc. Am. 93, 1267-1276 (1993).
  9. Nagy, B. and Adler, L., "Surface roughness induced attenuation of reflected and transmitted ultrasonic waves," J. Acoust. Soc. Am. 82, 193-197 (1987).
  10. Abdelhay, A. M. and Dawood, O. M.," Characterization of material's permanent deformation by ultrasonic technique." 7th Inter. Conf. On PEDAC, Alex, Egypt, Vol III, pp:1431-1441, Feb. (2001).
  11. Bray, D. E. andStanley, R. K., " Nondestructive Evaluation- A tool in design, manufacturing and service,"Mc Graw Hill Book Co., N.Y. (1989)
  12. Internet, "Metrology for Manufacturing", www.mfg.mtu.edu


Co, C1 , C3Quadratic model's coefficients
L Path length for ultrasound wave
Pi Stress wave energy
Ra, Ry, Rq Surface roughness parameters
a Ultrasonic attenuation coefficient
b Ultrasonic wave reflection coefficient

© NDT.net |Top|