Ultrasonic Crack Growth Monitoring
Using the Satellite Pulse Crack Tip Diffraction
Through-Wall Depth Sizing Technique

This month's feature is another example of how nondestructive testing (NDT) can be used in component design and product safety. The author describes the application of ultrasonic shear wave technique in determining the point of initiation and propagation of cracks in truck axles. A procedure was developed whereby ultrasonic data was correlated with the number of loading cycles and the resulting interpolated data allowed prediction of failure very accurately. G.P. Singh Associate Technical Editor

Figures 1-3

A problem on which I recently worked was: How is it possible to determine accurately the depth and length of cracks in a tubular structure (truck axles) using ultrasonic shear wave techniques? A conference with all parties involved determined that it would be feasible using some modified techniques found in the EPRI Intergranular Stress Corrosion Cracking - Crack Detection and Sizing training manual.

The purpose of this study was fourfold. Ultimately, we wanted to predict the number of cycles of loading and unloading necessary to:

1. initiate cracking
2. continue loading with failure of the axle
3. verify the adequacy of a retrofit design assembly
4. verify the adequacy of a new design with greater life cycle potential.

An automated hydraulic test stand loading jig was built with a capacity of four axles. The design of the test stand was intended to simulate the road conditions that a high weight capacity box trailer truck assembly would experience. The loading and unloading was applied with a time interval of 1 s and a peak load of 207 MPa (30 000 psi).

---

We found that the initiation point for the cracking consistently began at the end toe of the fillet welds.

---

With an axle tube wall nominal thickness of 11.8 mm (0.46 in. ) and a tube outer diameter of 96.5 mm (3.8 in.) the procedure used had to provide accurate results. The ultrasonic data and the number of cycles of loading would need to be correlated, and the resulting interpolated data must allow prediction of a failure within a certain degree of accuracy. The test stand was activated and all four stations were loaded with axles. The interior of each axle was sealed and pressurized: a loss in pressure on the gage attached to each would immediately indicate the initiation of a through crack.

We found that the initiation point for the cracking consistently began at the end toe of the fillet welds. These welds were on the outside diameter of the axle tube attaching the plate steel bracketing that was used to bolt the axle to the truck suspension (or to the test stand; see Figure 1).

The most useful and accurate ultrasonic technique we found for this purpose was the second leg absolute arrival time crack tip diffraction technique (satellite pulse).

The test stand was manned by ultrasonic test technicians around the clock. The cycling was halted at each 10 000 cycles and the ultrasonic scanning was repeated and data recorded in the log (See Table 1.) A graphic plot was also developed for easier display of the results to management, and for rapid identification of any trends that were developing.

We used a 6.35 mm (0.25 in.), 5 MHz, 45 degrees miniature angle beam transducer. Sixty degrees and 70 degrees angles were also used occasionally to supplement the inspection. Some of the parties were initially skeptical about the accuracy of the data we could obtain. At the beginning of the project the repeatability from one inspector to the next was poor. The data proved to be very reliable after each inspector had some hours of hands-on application of the techniques, and once results recording could be standardized for each shift and each technician. Failures could be predicted. New assembly designs and welding techniques and bracketing of components were tried to identify the best design approach and to increase the longevity of these truck axles in service, in a cost-effective manner.

For our screen calibration, we used a screen range of two leg lengths for the absolute arrival time (AAT) No. 2 technique. Gain was variable, as needed, as no amplitude evaluations were used with this technique.

Once a tip diffracted signal from a crack was found and the transducer manipulated to maximize the signal evident on the screen, the crack depth could be read directly from the screen as a percentage of the known wall thickness.

In summary, the project was deemed a success. The data collection effort spanned approximately 18 months. Retrofit bracketing assemblies were designed, tested, and implemented as a result of the information obtained from these tests. In addition, modified designs for fabrication of new equipment were also tested in this manner to verify effectiveness. This is an excellent example of how NDT and ultrasonic testing can be a major component of an engineering design study and contribute to public safety on our highways.

 

SIZING PROCEDURE

Depth Sizing of Tubing OD Connected Cracks
The following three key factors were considered in applying the satellite pulse tip diffraction method to determine the depth of an outside surface connected fatigue crack:

• The mechanism of crack initiation and propagation. Observations to date indicate that the fatigue crack initiates at the fillet weld end from the outside diameter surface, and then propagates radially inward and circumferentially around the axle tube. The second half-vee path satellite pulse tip diffraction sizing method is appropriate for this crack configuration. (See Figure 2.)
• The calibration block must have OD notches at 25, 50, 75, and 100 percent of through wall for correct screen calibration. Adjust the screen range and delay controls to show the inside diameter comer reflector of the 100 percent notch at the zero position on the horizontal trace, with the 75 percent notch pulse at one-quarter screen, the 50 percent notch pulse at one-half screen, and the 25 percent notch pulse at three-quarters screen. The distance linearity of the ultrasonic scope is now set to read directly the remaining ligament metal thickness in the tube wall. The crack depth can be determined by subtracting the remaining ligament from the known tube wall thickness.
• The identification of the crack tip satellite pulse. This factor is most critical to an accurate evaluation. To do this, "peak up" or maximize the corner reflector of the previously detected crack (OD connected). Increase the gain until 5 percent to 10 percent noise is seen at the base line. Additional gain is acceptable if needed to detect the tip signal. Slowly move the transducer toward the crack location, carefully observing the area to the left of the screen position of the original corner reflector (crack) signal. A small signal will be seen to come up and down quickly. Focus on this signal and "peak up" or maximize it by moving the transducer forward and back. Once maximized, read the remaining ligament directly from the horizontal scale of the screen. Subtract this value from the known tube wall thickness to find the actual crack depth.
• Example: For sizing crack depths in a tube with a wall thickness of 11.8 mm (0.46 in. ), each major screen division represents 1.18 mm (0.046 in.) in wall thickness. The satellite pulse occurs at three screen divisions, so the remaining ligament is 3.5 mm (0.14 in.). Subtracting this value from 11.8 mm (0.46 in. ), the crack depth is 8 mm (0.32 in.) from the OD surface. (See Figure 3.)

Table 1	Sample Report
Loading cycles (X 103)	Crack depth mm (in.)		Linear length mm (in.)		Seal intact?
10					yes
20					yes
30					yes
40					yes
50					yes
60					yes
70					yes
80					yes
90					yes
100					yes
110					yes
120					yes
130					yes
140					yes
150					yes
160					yes
170					yes
180					yes
190	0.5	(0.02)	1.0	(0.04)	yes
200	1.27	(0.05)	2.54	(0.1)	yes
210	2.54	(0.1)	5.0	(0.2)	yes
220	3.81	(0.15)	7.62	(0.3)	yes
230	5.0	(0.2)	10.0	(0.4)	yes
240	6.35	(0.25)	15.25	(0.6)	yes
250	7.62	(0.3)	17.75	(0.7)	yes
260	8.89	(0.35)	20.0	(0.8)	yes
270	10.0	(0.4)	25.4	(1.0)	yes
280	11.25	(0.45)	35.0	(1.4)	yes
290	50.0	(2.1)			yes
300	66.0	(2.6)			yes
310	86.35	(3.4)			yes
320	122.5	(4.9)			no
330	202.5	(8.1)			no
340	303.0	(11.93)			no

 

* Hellier Pacific, 3710 Artesia Avenue, Fullerton, CA 92833 (714) 562-9227; fax (714) 562-9228

Copyright © 1999 by the American Society for Nondestructive Testing, Inc. All rights reserved.

[ Materials Evaluation ]