Abstract

Contents
• 1 Introduction and objective
• 2 Defect Conditions
• 3 Testing Methods
  • 3.1 Ultrasonic Testing
    • 3.1.1 Statistical investigation for the detection of shots with Ultrasonics
    • 3.1.2 Defect sizing by use of echo amplitude
  • 3.2 Eddy Current Testing
    • 3.2.1 New techniques for Eddy Current Testing
    • 3.2.2 Determination of position resolution capability
    • 3.2.3 Detection of shots with Eddy Current Testing
    • 3.2.4 Detection of areas free of fibres with Eddy Current Testing
    • 3.2.5 Combination of Ultrasonic Testing and Eddy Current Testing
    • 3.2.6 Investigation by use of higher test speed
  • 4 Summary
  • 5 References

1. Introduction and objective

With the use of fibre reinforced materials with their compound structures, and the need for detection of very small defects up to 100µm, NDT faces a great challenge. The detection of various defects (e.g. shots, inclusions, fibre-free areas) requires the use of a combination of different kinds of NDT methods. Generally, a qualitative rather than only quantitative characterisation is called for, so it is essential to further evaluate the reliability of the results.

The focus of this study is investigation of the probability of inclusion detection. [1, 2]. Because it would be applied in a production environment, a practical design of the test procedure was necessary.

The Eddy Current method needs a small sensor for the detection and determination of the exact position of the defect. With the use of automated C-scan and imaging techniques, it is possible to improve test speed and accuracy, and reduce the subjective aspect of validation.

Another focus is Ultrasonic Testing with the investigation of repeatability, consisting mainly of probe frequency variation and test speed. A technique for testing of industrial parts has been developed which should be applicable to the testing of critical sections in one step, e.g., scanning with multiple probes.

2. Defect Conditions

Depending on the manufacturing process and structure of fibre reinforced light metals various kind of defects can occur: Inclusions: Shots : Ceramic particles Foreign bodies: Inclusions of metals Defects of the fibre orientation fibre-free areas Under certain conditions such defects can cause a fracture of the components.

Fehlerarten Abb 1. Include of iron Abb 2. Shot Abb 3. Pore Abb 4. Delamination Abb 5. fibre-free zone Abb 6. Fibre concentration Abb 7. In-depth deviation of fibres

3 Test methods

3.1 Ultrasonic Testing

High frequency ultrasonic C-scan equipment (model HFUS 2000) was used for application of the ultrasonic method. The best working range of this scan system lies in the 1 MHz to 100 MHz range. It can perform both C- and D-scan (optional B-Scan) images. The system's stepper motors provide a scan raster set in the 0.05 - 2.00 mm range with an accuracy of 10µm. [4]

By using focused probes, and a high dynamic range, it is possible to achieve a precise lateral resolution, necessary for testing fibre reinforced materials . The system provides a high pulser voltage together with a short rise time. The system uses a focused 25 MHz PVDF immersion probe and with the broadband features of this set up, the capability for finding small defects in high absorbing materials is assured [5].

A C-scan image presents the results of an echo amplitude (A) over the planar coordinates x and y, formula C = f (x, y, A). This allows good interpretation of defect sizes. Instead of echo amplitude (A) a so-called D-scan image presents the results of echo transit time (t_z), formula D = f (x, y, t_z) [5]. Interpretation of defect depth can be determined more easily from a colour image.

3.1.1 Detection of defects Determination of position resolution capability by using a reference body A specimen containing 12% Saffil is used for this determination. Artificial flat bottom holes (FBH) between 50 µm to 400 µm are drilled into the specimen. The smallest detectable drill is 100µm. Using a 25 MHz probe, we can calculate a factor of d/wavelength = 0,3, which is a normalized figure. That formula facilitates comparison between reflection (good detection) and scattering (bad or no detection).
Shots The detection of Shots was tested on material reinforced with fiberfrax. We see that with increasing depth the sensitivity of detection increases. Also the factor d/wavelength decreases, improving sensitivity. The reason is that the focused probe gains a better resolution in the area of its focus point. An equivalent circle diameter was defined to get a better determination of the size. That means the real defect plane was used to calculate its equivalent circle diameter. The smallest Shot of 121 µm equivalent circle diameter was detectable (fig. 8), as long as the defect is located in the focus of the probe.	Fig 8. C-image with Shots
Iron Inclusions An artificial foreign body of iron was added to the specimen. Iron wires with a length of 10 mm up to 16 mm and diameters of 100 µm, 200 µm, 300 µm and 500 µm were used. Image 9 shows a result of a 200 µm diameter.	Fig. 9. C-scan image with inclusion of iron.
Fibre concentration The detection of fibre concentration is very difficult. For its detection, size is not the only factor; density can influence results even more. Based on the metallographic method it was determined that detection was possible only if the fibre concentration deviation is 10% of the average. The smallest detectable fibre concentration was 340 µm (fig. 10).	Fig. 10. C-scan image with fibre concentration
Delamination This investigation shows that delaminations are possible to detect for a minimum of 35% fibre density (fig. 11).	Fig. 11. C- scan image with delamination
Fibre-free areas The results show that it is difficult to detect fibre-free areas due to the low reflection coefficient at the defect boundary (water to fibre-free area). The interface echo was evaluated on a part which showed a visible fibre-free area at its surface (fig 12). Fig. 12 C-scan image with fibre-less zones
In-depth deviation of fibres The detection of in-depth deviation of fibres was possible using D-imaging techniques. That measures the time of flight up to the fibre reflection. That could be proven on a sample which showed a higher fibre density in a deeper area of the part. 3.1.2 Influence of the part's roughness on the detection of Shots The geometry and roughness of the specimen's surface has an essential influence on the result of the ultrasonic testing method. That can limit the ability to detect small defects, necessitating a surface treatment. The ultrasonic method needs a constant echo amplitude (constant coupling energy) free from influence of the surface condition. Especially when a small wave length is used, the surface effect is strong. Higher scattering occurs on higher surface roughness - that results in a loss of energy. The capability for detection of small defects decreases. This is intensified when high frequency focused probes are used. The investigation shows that with a roughness of Rm= 18 µm the detection sensitivity decreases significantly. (Test frequency 25 MHz, Filter frequency 25 MHz).

3.1 Statistical investigation for the detection of shots with Ultrasonics

The investigation of a specimen containing 20% Fibrefrax fibres used a focused probe IAP-F 25.3.1 with a nominal frequency of 21 MHz and an equipment filter of 25 MHz. Figure 1 compares the results of detected shots of 8 groups of different provided sizes, determined with metallography. The size tolerance of each group was 30 µm [1, 2]. Since the smallest detectable defect was approx. 120 µm, the statistical investigation used shots bigger than 90 µm. Virtual improvement of detection can be seen in the 121 µm and 150 µm range. Most of these defects were located at a depth of 2 mm. Because the probe focus also lies in this area, the detection of those defect sizes was easier than in other areas. A statistically determined safe detection of 200 µm sizes can be realised.

Fig 1: Comparison of metallography detected defects with the Ultrasonic results .

3.1.4 Defect sizing by use of echo amplitude

For this unusual method for defect sizing was used - the DGS method with its diagrams. The principle is not explained here, since there is a great deal of in-depth literature on the subject. [6, 7]. The investigation continued with the precondition that the DGS method can be applied by considering the conditions of focused probes. The investigation used a focused probe with a 25 MHz filter.

The specimen is fibrefrax reinforced aluminum with a fibre density of 20%. The sample contained some defects. To obtain good repeatability it is necessary to adjust the probe on a reference body. The results show a correlation between echo amplitude and defect size. The result is applicable to defects located at the focus point or not more than 0.6 mm from it. (behind the focus point in the direction of propagation). On defects farther away, that area will appear as interference, and thus influence the results. An exact determination of the defect size will not be possible.

Eddy Current Testing

3.2.1 New techniques for Eddy Current Testing

Based on manual testing experience, eddy current C-scan and imaging techniques were used. [3,4].

The block diagram of the newly built C-scan equipment is shown in figure 2.

The centre of the equipment is the microprocessor controlled eddy current unit "eddyMax" manufactured by TMT. Using new software it was possible to control the eddy current board and the manipulator. The generated data set can be calculated for a C-scan image and plotted, saved and post processed.

Use of the Scan equipment makes essential improvement possible due to its capability for investigation of sensors, repeatability and probability of detection.

The sensitivity was clearly increased by reduction of the effective sensor width. For the effective area of the sensor field - besides the test frequency, the conductivity and the magnetic properties of the material (ferromagnetic or paramagnetic) - sensor construction is especially important. A small defect can easily influence the field of a small sensor. An eddy current probe type AN 05-01 with an 0,5 mm core size, well suited to the high lateral resolution needed for C-scan imaging was used.

3.2.2 Determination of position resolution capability

A test on a reference body was carried out for getting resolution results for the probe AN 05-01.
The drill diameters used were: 50 µm, 100µm, 140µm, 200µm, 280µm, and 400µm.
The fibre ring of the reference body consisted of MMC which was reinforced with 12% Saffilfibre.
The investigation used the following parameters for all tests :

Test frequency:	200 kHz
Test speed:	40 mm/s
Resolution:	50 µm

It was possible to detect all of the 6 drills. The drills were placed in the area between non-reinforced and reinforced material. The C-scan image could show clearly with its grey/colour scale the position of drills in the non-reinforced and reinforced areas (fig. 3). A conclusion of this result is that natural porosities of approx. 50 - 100 µm (simulated by the drills) can be detected as long as they are located on the surface.

Fig 5: Drill (dia 200 µm) in the area between reinforced and non-reinforced material

3.2.3 Detection of shots with Eddy Current Testing

A critical defect for MMC materials are the shots. The reference bodies were reinforced with 20% Fibrefrax. The investigation used the following parameters for all tests :

Probe :	AN 05-01
Test speed :	50 mm/s
Resolution:	50 µm

The detection of a specific shot was possible with all frequencies. The best result was achieved by use of 100 - 200 kHz (Fig. 4). A detected shot measured 400 µm which was determined by metallography.

Fig. 18: Shot on the surface (Test frequency 200 kHz)

3.2.4 Detection of areas free of fibres with Eddy Current Testing Specimens with artificially provided fibre free areas were used. Detection of the fibre free areas was possible up to the minimum of 100 µm in one direction (Fig 19).

fig 5: Fibre-free area on the surface (width approx. 100µm)

3.2.5 Combination of Ultrasonic Testing and Eddy Current Testing

The objective of this investigation is to prove the applicability of a simultaneous test procedure using both the Ultrasonic and Eddy Current methods in one production step.

The investigation was carried out on a specimen reinforced with 20% Fibrefrax with fibre free areas on the surface. Due to the different capabilities of each method it is possible to detect the fibre free areas on the surface with Eddy Current and the shots inside the material with Ultrasonics. (Fig. 20).

Fig. 20 : Comparison of C-scan images of a specimen a). with Ultrasonics b). with Eddy Current

A negative influence of the methods on each other was not shown with the parameters used here. This test could also prove detection of fibre-free areas by Eddy Current up to 100 µm.

3.2.6 Investigation by use of higher test speed

An essential issue for the applicability of NDT methods to production is the need for a short testing time.

By using a hardware divider for the trigger pulses it was possible to increase the test speed from 40 mm/s up to 125 mm/s. To investigate the influence of this speed a specimen with approx. 270 µm fibre-free area was scanned; other parameters were not changed. That was carried out with 10 different speeds in the 25 mm/s up to 125 mm/s range .

The investigation used the following parameters for all tests :

Probe:	AN 05-01
Test frequency	200 kHz
Resolution:	50 µm

The comparison between the C-scan images (fig. 21 and fig. 22), showed that with speed increased three times the fibre-free area was clearly detected. However, related to the scan process (spiraling scanning of the specimen), a directional relation to the direction of the scan axis occurs. This can influence the presentation of the fibre-free area of the C-scan image. The detection of the fibre-free areas as material inhomogeneity is still possible with this test speed.

Fig. 21 u. 22: C-scan image of a fibre-free area test speed 37,5 mm/s (left) test speed 125 mm/s (right)

Since increasing the speed was not possible with the existing equipment,
another solution was sought to achieve the needed speed of 1 m/s.

A turntable was modified for realising the speed up to 1m/s. The Eddy Current unit Defectoscope AF with an AN 05-01 probe was used. An engine piston was used as a specimen , provided with a symmetrical reinforced ring which made a simple adjustment of the system possible. This piston contained two fibre-free areas on the surface.
The eddy current signals were transmitted from the Defectoscope AF and an A/D converter (16 bit resolution) with a sampling rate of 50 kHz to the PC port. Post processing generated the presented images.

Because the investigation chose the x-axis as the influence of the take-off parameter, only the y-part of the Eddy Current signal was used for the evaluation.
Two fibre-free areas (A, B) clearly produce higher voltage, resulting from higher conductivity in the fibre-free area. Also an area of lower conductivity indicative of fibre concentration could be detected.
To provide a clear comparison of different test speeds, the investigated area was reduced to one circumference.
The test results show that with increasing speed detection is more difficult and depends on defect size. Large fibre-free areas ( approx. 750 µm), scanned with a speed of 1 m/s, still generated a clear signal/noise ratio, while an area of 380 µm was already difficult to detect with a speed of 0,5 m/s. Two diagrams (fig. 23) show this result for a speed of 0,1 m/s and 1m/s (maximum).


Fig. 23: Comparison between test speed
0,1 m/s (left) and 1 m/s (right)
A flatness of the curve is visible for a speed of 1 m/s, that means material inhomogeneity can not be detected; the measurement signal flattens.

4. Summary

The results of this study show that complex investigation of fibre reinforced material is possible. The Ultrasonic Testing and Eddy Current testing methods can be performed as one production step. Automation of both methods prevents subjective validation of the test results. The test speed of 500 mm/s is possible, if an imaging technique is not required.

References

1. Wolf, M.; Danzer, W; Petzow, G: Defekt- und Bruchstatistik von SSiC. Fortschritts berichte der DKG; 6 (1991) 1; Festigkeitsseminar Keramische (Verbund) Werkstoffe 990, S.115-125
2. Perdijon, J.: Statistics applied to measurement in ultrasonic testing. Material evaluation 1461, September 1988
3. Beissner, R.E.; Burkhardt, G.L.; Fisher, J.L.: Eddy Current measurment of fiber volume fraction in metall matrix composites. Review of Progress in Quantitative Nondestructive Evaluation, Vol.12, Plenum Press, New York, 1993 S.1321 - 1328
4. Thomas, H -M, Maser, D Entwicklung und Einsatz eines Wirbelstromprüfsystems zur Visualisierung von Korrosionsschäden an Flugzeugtragflächen Berichtsband zur DGZfP-Jahrestagung 1992 in Fulda, S 361-368
5. Betriebsanleitung HFUS 2000.
6. W. Hillger, Ultraschall-Prüftechnik zur Schadensanalyse an CFK-Laminaten, DGZfP. Seminar Ultraschallprüfung neuer Werkstoffe Berlin 6. und 7. November 1989, Tagungsband S. 126-135, DGZfP Berlin 1989.
7. H.-D. Tietz, Ultraschall-Meßtechnik. VEB Verlag Technik Berlin, 2. Auflage, 1974.
8. Das Krautkrämer Taschenbuch. 3 Auflage, Herausgeber Krautkrämer GmhH.
9. Tagungsband DGZfP NDT conference 1996 in Lindau. [Conference Report] , S 287- 294

Authors

Prof. Dr.-lng. habil. Horst - Dieter Tietz Dr. Horst-Dieter Tietz studied material science and material testing in Magdeburg, where he received his doctorate in 1965. He continued there and was tenured in 1970. He has also served as manager of an industrial material institute in the heavy-steel market in Magdeburg. From 1989-1992 he was director of the Institute for Materials and Quality control, Director of the TH Zwickau in 1991-1992, before founding the University for Technique and Economics Zwickau (FH) and serving as its Director until 1996. He is presently the Director of the Westsächsischen University Zwickau (FH), the Technical Ceramic Application Center. Dr. Tietz continues to teach at several Universities. He is the author of many professional books in the fields of materials and nondestructive testing. Email: Horst.Dieter.Tietz@fh-zwickau.de FH Zwickau Homepage http://www.fh-zwickau.de (German) Special expertise in "Prüfung von Kompaktwerkstoffen, Schichtverbunden und Grünlingen".