In the past, surface guided waves have not been used extensively for NDT. This article describes a surface guided acoustic mode with several inherent characteristics that make it particularly useful for detecting subsurface transverse flaws in railroad rail. These flaws have been previously in the "blind spot" of other ultrasonic NDT techniques. G.P. Singh Contributing Editor

 

Introduction
The first higher order mode of the Rayleigh wave was discussed by Sezawa in the early part of this century in context of seismological wave studies. These Sezawa, or M₂₁, or first higher order mode Rayleigh waves, have subsequently been used in the field of nondestructive testing of layered materials based on the development of the seismological model of the Sezawa waves by others. Application of these waves in the early detection of microcracks in used railroad rail can help prevent accidents and provide better utilization of resources in current use.

Railroads are a very efficient and economical way of moving bulk commodities over large distances. However, like any other mode of transportation, this mode has its fair share of problems. Table 1 shows some interesting data about the railroad industry over the past few decades. It can be seen from the data (AAR, 1993) that over a period of time the average carload and trainload tonnage has gone up substantially. The average length of haul has also increased. However, the amount of new rail laid has actually decreased significantly. This presents a challenging opportunity to research better and more efficient ways of keeping the current railroad rail track in service and to find better ways of testing those rails to pre-empt any causes for accidents. An important factor to be considered here is the testing of the railroad rail for the amount of cold work in the top layer and the presence of microcracks in that layer.

---

Testing the top layer of railroad rail using ultrasonic is, perhaps, one of the most efficient ways of nondestructively testing railroad rail.

---

 

The most common rails used range in weight from 39.5 kg/m (80 lb./yd) to 69.5 kg/m (140 lb./yd). Fifty one point one percent (AAR, 1993) of all rail in use today is of the 130-140 lb./yd range. The 110-119 lb./yd range rail makes up for about 25.9 percent of all rail in use. Of these, the heavier sections (i.e. 49.6 kg/m or 100 lb./yd and over) are typically used in main-line track, while the lighter sections are used for branch and switching-line service. The rail head (Bray, 1978) varies from 63.5 mm (2.5 in.) to 76.2 mm (3.0 in.) and the head height varies from 38.1 mm (1.5 in.) to 52.4 mm (2.062 in. ) for these rails. The rail height is of utmost significance, since this greatly affects the section modulus and, hence, the strength in bending. This is something that is considered during the designing of rails.

 

During the manufacturing phase of rails, avoiding shatter cracks by controlling the cooling process and the residual stresses generated (Heindlhofer, 1948; Johnson, 1942) helps to produce quality rail that can stay in service longer. Once the rail is in service, initial strength and physical characteristics assume only a part in the overall durability of the rail. In-service conditions and the physical changes it undergoes assume a large responsibility towards the stability of the rail in service. It has been reported that increased load increases the risk of accidents (Bray, 1976). Also, increased loads cause more cold working which, along with extreme plastic flow in the upper layer of the railroad rail, causes shelling. Earlier works (Egle and Bray, 1975) have shown the ill effects of shelling. It has been reported (AISI, 1967) that a typical static force of 117.4 kN (26,400 lb.) acts on the contact area between the wheel and the rail. The average contact stress is around 870 MN/m² (126,000 lb/in.²). The compressive yield strength for a typical rail steel is 503.32 MN/m² (73,000 lb/in.²) (Backofen, 1972). This results in excessive material movement. Hence, it can be easily established that this excessive movement results in the generation of a cold worked layer on the rail, with a texture different from the rest of the rail. The depth of cold work has been estimated to be 3-6 mm (0.11-0.23 in.) (OREIUR, 1970). Transverse cracks in rail have been observed in the cold worked zone of the rail. These transverse cracks may not be apparent at the surface because since the neutral axis during elastic action of the rail is typically below the flaw growth area, higher stresses and, hence, higher flaw growth in the area above the origin cause the crack growth to arrest as it progresses towards the top of the layer. A broken rail section was observed to have fatigue cracks at a depth of 3.5 mm (0.13 in.) from the top (Bray, 1978). These cracks can undoubtedly be dangerous as they can lead to catastrophic failure of the rail. One of the ways to detect such cracks, while the rail is still in use, is by means of ultrasonic testing using M₂₁ waves.

 

Ultrasonic Testing
Testing the top layer of railroad rail using ultrasound is, perhaps, one of the most efficient ways of nondestructively testing railroad rail. Material deformation in railroad rail has been evaluated by Bray and Najm (1983) using ultrasonic critically refracted shear waves. Here detection of flaws using ultrasound is described. Among the ultrasonic techniques available are angled pitch-catch and normal incident ultrasonic testing methods, as addressed by the ASME Boiler and Pressure Vessel Code, Section V, T-543 (1989), and guided surface waves. The normal incident wave methods are not very useful for detecting rail head defects. The angled pitch-catch technique is widely used (Birks et al., 1991) but still is very labor intensive. Since the use of normal beam probes for the inspection of layered specimens is limited, due to the near field effect, angled beam probes are preferred. Angled-beam shear probes may be used to test the material by reflecting the ultrasonic pulse off of the opposite surface. However, this method becomes ineffective due to near field effects as the probe approaches the defect. Angle beam techniques devoid of such problems include testing with surface guided waves such as Rayleigh waves. These techniques take advantage of the critical angle refraction of the angle-beam to send in and receive a strong ultrasonic signal. The fundamental Rayleigh mode, M₁₁, has been used extensively to test materials for surface or near surface flaws. However, since the transverse flaws are not open at the top surface and may exist 2-4 mm (0.07-0.15 in.) below the top surface, detection of these flaws using the M₁₁ mode may not be suitable. The first higher order mode M₂₁ wave offers a more versatile and stable means of testing the top layer of used railroad rail for flaws over the M₁₁ mode, as well as the other methods described above.

 

M₂₁ Waves
The Rayleigh wave equation is a sixth order polynomial. One of the roots to this polynomial describes the fundamental Rayleigh wave, or the M₁₁ mode. Among the other roots, an important one is the first higher order shear mode, or the M₂₁ mode. The M₂₁ mode was first introduced by Sezawa (1927) in the early part of this century. These waves were subsequently explored by various researchers in context of seismological activity. Work by Ewing et al. (1957), Bolt and Butcher (1960), and other has been of great importance in establishing the basic rules for the propagation of these waves. In this paper numerical analysis of the formulation presented by Tiersten (1969) is carried out to present the behavior of the M₂₁ waves in cold worked layer on a railroad rail.

Figure 1 illustrates the path taken by the fundamental and first higher order mode of the Rayleigh wave. As can be seen, the fundamental mode travels along the surface of the top layer and is extremely sensitive towards attenuation by surface imperfections. This mode also is not very useful when it comes to detecting subsurface flaws. The M₂₁, however, travels below the surface and can be used to detect subsurface defects that may not extend to the surface. This mode also is less susceptible to surface roughness or imperfections.

 

Figure 1 - Two modes of the Rayleigh wave in layered media.

 

Using the Tiersten formulation, phase velocities were calculated for various thickness and excitation frequencies. These phase velocities were then converted into group speeds so that a relationship with experimental values could be established. Figure 2 illustrates the derived group velocities and the experimentally observed wave speeds for the M₂₁ wave in the cold worked top layer of a used railroad rail specimen. A 64 degree clear acrylic wedge was used in conjunction with commercially available transducers to excite these waves. The signal received is a very clean signal with an amplitude far greater than the accompanying noise. Since the M₂₁ is faster than the M₁₁, it is the first arrival. Distinction can be made between the M₁₁ and M₂₁ by tapping on the surface of the rail with a greased finger. The M₁₁ damps out, whereas no tangible change occurs in the M₂₁ signal. Using through transmission, distances up to 355 mm (14 in.) could easily be tested using the M₂₁ waves.

 

Figure 2 - Simulated and experimental group speeds for the first higher order. Rayleigh mode in railroad rail.

Summary
Typical transverse flaws in cold worked layer of railroad rail are not open to the top surface, hence may be missed by conventional nondestructive testing techniques. However, these microcracks may extend to cause catastrophic failures of rail while in service. The Rayleigh wave modes are useful in testing the cold worked layer for such flaws. The first higher order mode, M₂₁, has a distinct advantage over the fundamental, M₁₁ mode, in that it ignores the surface condition, travels below the surface, where most of the transverse flaws are located, is the very first arrival, hence is extremely easy to discern, has a good concentration of energy, hence, is not lost in the noise, and can travel long distances, allowing testing of up to 250 mm (10 in.) of rail in one pass.

 

References
AAR, Railroad Facts, 1993 Ed., Aug. 1993. Economics and Finance Department, Association of American Railroads, Washington, DC.

ASI, Steel Products Manual: Wrought Steel Wheels and Forged Railway Axles, 1967. American Iron and Steel Institute.

ASME Boiler and Pressure Vessel Code, Section V, Nondestructive Examination, Section V Topic T-543, 1989. The American Society of Mechanical Engineers, New York, NY.

Backofen, W.A., Deformation Processing, 1972. Addison-Wesley, Reading, MA.

Birks, A.S., and R.E. Green, Jr., eds., Nondestructive Testing Handbook, Vol. 7, Ultrasonic Testing, 2nd ed., pp 612-617. American Society for Nondestructive Testing, Columbus, OH.

Bolt, B.A., and J.C. Butcher, "Rayleigh Wave Dispersion for a Single Layer on an Elastic Half Space," Australian Journal of Physics, Vol. 13, No. 3, Sep. 1960, pp 498-504.

Bray, D.E., "Railroad Accidents and Nondestructive Inspection," 1975 Proceedings of the Rail Transportation Division, paper No. 74-WA/RT-4, 1976. American Society of Mechanical Engineers, New York, NY.

Bray, Don E., "Ultrasonic Pulse Propagation in the Cold-Worked Layer of Railroad Rail," US Department of Transportation Report No. FRA/ORD-77/34.11, Jan. 1978, p 13.

Bray, Don E., and M. Najm, "The Effect of Material Deformation on the Velocity of Critically Refracted Shear Waves in Railroad Rail," Proceedings of the 1983 Ultrasonics International Conference, Halifax, Nova Scotia, Jul. 11-14, 1983, pp 13-18.

Egle, D.M., and D.E. Bray, "Nondestructive Measurement of Longitudinal Rail Stresses," Federal Railroad Administration, Report FRA-75-40091, Jun. 1975.

Ewing, M., W. Jardetsky, and F. Press, Elastic Waves in Layered Media, 1957. McGraw-Hill, New York, NY.

Heindlhofer, K., Evaluation of Residual Stress, 1948. McGraw-Hill Book Co., New York, NY.

Johnson, J., "Time as a Factor in the Making and Treating of Steel," Transactions of AIMME, Iron and Steel Division, Vol. 150, 1942, pp 13-29.

OREIUR, "Behavior of the Metal of Rails and Wheels in the Contact Zone, Residual Stresses in the Rail (Continued) Study of the Work Hardened Zone," Report No. 6, C53/RP6/E, Oct. 1970. Office for Research and Experiments of the International Union of Railways, Utrecht.

Sezawa, K., "Dispersion of Elastic Waves propagating on the Surface of Stratified Bodies and on Curved Surfaces," Bulletin, Earthquake Research Institute, Tokyo University, Vol. 3, 1927, pp 1-18.

Tiersten, H.F., "Elastic Surface Waves Guided by Thin Films," Journal of Applied Physics, Vol. 13, Feb. 1969, pp 770-780.

 

* Texas A&M University College Station, TX 77841; (409) 862-1897; e-mail grewal@tamu.edu

Copyright © 1996 by the American Society for Nondestructive Testing, Inc. All rights reserved.  

[ Materials Evaluation ]

IRRSP, NDT Handbook, The NDT Technician and www.asnt.org are trademarks of the American Society for Nondestructive Testing, Inc. ACCP, ASNT, Level III Study Guide, Materials Evaluation, Nondestructive Testing Handbook, Research in Nondestructive Evaluation and RNDE are registered trademarks of the American Society for Nondestructive Testing, Inc. ASNT exists to create a safer world by promoting the profession and technologies of nondestructive testing.