ACOUSTIC EMISSION MONITERING OF FATIGUE CRACK GROWTH IN BRIDGES

INTRODUCTION

In August of this year I had the opportunity to participate in the acoustic emission monitoring of a fatigue test on cracked sections of a railroad bridge using an AESMART 2000 instrument. These tests were conducted by the Transportation Technology Center, Inc. (TTCI) of the Association of American Railroads on cracked girder specimens from a span of the Mason Creek Bridge from Grande Cache, Alberta. Two spans of this bridge, which is owned by Canadian National Railways (CN) had recently been replaced because of fatigue crack problems in the welded assemblies. In-situ acoustic emission testing to ascertain the condition of fatigue cracks on this bridge had previously been conducted by DNL Infrastructure Technologies Inc. (DNL) The purpose of the fatigue tests at TTCI was to attempt to correlate field test results with laboratory results using acoustic emission testing and accelerated laboratory fatigue tests. A preliminary Summary Report "Fatigue Testing of Specimens from the Mason Creek Bridge," showing details of the test setup and the results of these tests has been prepared by DNL and can be viewed on our web site by clicking on the highlighted name.

DNL utilized both traditional AE instrumentation and DECI AESMART 2000 instrumentation for conducting the acoustic emission tests. In addition TTCI provided strain gage outputs and optical tracking methods for monitoring crack growth. For brevity I will only report on results from the AESMART for one of the tests.

Figure 1 shows a cracked specimen setup for a four point bend test in a large MTS servo hydraulic test machine. Since the load heads of the test machine were not constrained considerable bending of the web took place during the test. Pencil lead breaks were made on the edge and surface of the web in order to determine for this particular specimen where the ratio filter in the AESMART should be set to prevent out-of-plane(OOP) noise from being recorded by the instrument. An integral SE9125-MI dual mode transducer was used as the data transducer. It was placed approximately 10 inches from the known crack in the specimen. An SE150-MI integral transducer was placed on the back side of the specimen at the crack location to provide a trigger signal for any AE emitted from the crack.

With the ratio filter set at 3, one could strike the specimen with a wrench and no signals were received. After starting the test a few high ratio signals were detected. It was determined that these were due to friction created on the top of the web by the load heads of the test machine. We therefore set up another filter based on the trigger delta T in order to eliminate the friction on the web edges from being recorded. With both filters operating no data were received during the initial portion of the fatigue test. Periodic Mag-particle examination confirmed that no crack growth was occurring during this quite period.

Fatigue cycling of the specimen continued with periodic increases in the maximum load. Approximately 1 hour after the start of the test the AESMART 2000 begin to record data from the crack vicinity. This data is shown in figure 2 which shows the summation of counts from the high frequency channel of the instrument as a function of time.

Figure 2- Plot of fatigue crack initiation and growth in Specimen #2.

The MTS machine was allowed to continue cycling for approximately 1/2 hour between +89 and 650 kN (+20 and 145 kip) at 3Hz. At this time it was decided to stop the test and inspect the specimen. A magnetic particle inspection confirmed that the original crack had not grown, but there were two new cracks in the web material. It appeared that there was a single point of origin slightly above the original crack that forked into two cracks. The fairly low ratios recorded from the AESMART 2000 indicated that these cracks were shallow. This was further supported by the fact that the cracks were not visible without the aid of the magnetic particle inspection.

DISCUSSION OF RESULTS

In past reports I have been very critical of laboratory acoustic emission tests conducted on base materials used in bridges and pressure vessels containing artificial fatigue cracks or notches. It is very difficult in such laboratory tests to simulate the conditions encountered by structures fabricated from these materials in the field. It was therefore very exciting for me to participate in the fatigue tests conducted by TTCI on these bridge members containing cracks formed under actual operating conditions. In a welded redundant structure such as a bridge, cracks will usually form in the weld material or heat effected zone. Design errors which produce high stress concentrations, bad welding practices, or residual stresses created by the welding process itself are the primary causes of crack initiation. Once the crack is initiated, environmental conditions may accelerate the crack propagation process. In a redundant structure the stress relief caused by the formation of a crack will cause these stresses to be transferred to other members of the structure. In addition the crack may grow into base material. Both of these conditions could cause the crack to arrest. Therefore the important concern should be the growing crack and not simply that a crack is present.

The fatigue crack present in the specimen shown by figure 1 initiated in the weld material and had propagated into the base material of the web. Through the optical microscope used to observe this crack one could see the opening and closing of this initial crack, but no crack growth occurred during the period of cycling shown by figure 2. New cracks did form in the weld material and were detected by the AESMART 2000 instrumentation.

Precautions were used during the testing to prevent noise from the servo valve of the testing machine from entering the specimen. The load heads of the test machine were unrestrained and therefore bending of the web material occurred. In addition friction noise from the load heads was created due to their contact with the web material. Both of these conditions created extraneous noise within the bandpass of the acoustic emission instrumentation but the special filtering capability of the AESMART 2000 eliminated this extraneous noise in real time and allowed signals from the crack growth to be detected. AE signals from this specimen were also recorded with a traditional AE system. A lot of data were recorded even during periods when no crack growth was occurring. This data is yet to be analyzed by DNL to determine if the crack growth AE signals can be separated from extraneous noise.

One major conclusion derived from this test is that the special filtering capability of the AESMART 2000 will eliminate extraneous noise in a very noisy test in real time and allow AE signals from crack growth to be recorded. Another is the ability of one to estimate the depth of the growing crack from the value of ratios recorded.

NEW PRODUCT

A more economical version of the AESMART 2000, the AESMART Model 302A is now available for researchers interested in the applications of acoustic emission to Tribology, HDI, metal cutting, grinding and leak detection. We and our customers are finding that our innovative application of signal modal analysis which utilizes the ratio of two frequencies contained in the AE signal corresponding to two different plate modes provides much more diagonistic information on the above processes than simply recording a broadband RMS signal from the process. A preliminary data sheet on this new product can be found by clicking on the product name above.