·Home ·Table of Contents ·Workshop - Visual testing

Proposal on Calibration Methods of Visual Condition for Penetrant Testing and Magnetic Particle Testing Isao TANAKA Nuclear Systems Division, Hitachi, Ltd., 1-1, Saiwai-cho 3-chome, Hitachi-shi,, Ibaraki-ken, Japan 317-8511 Phone: +81-294-23-5336, Fax: +81-294-23-6612 E-mail : isao_tanaka@cm.hitachi.hitachi.co.jp Tetsuo TAGUCHI and Masahiro HOTTA Nuclear Systems Division, Hitachi, Ltd., Ibaraki, Japan Mineo NOMOTO , Daisuke KATSUTA and Kaoru SAKAI Production Engineering Research Laboratory, Hitachi, Ltd., Yokohama, Japan Takemitsu SATO Japan Power Engineering and Inspection Corporation Tokyo, Japan Mikio TAKAGI Science University of Tokyo Chiba, Japan Contact

ABSTRACT

The eyesight of an inspector and brightness on the inspected surface are important for observations on indications shown by Penetrant Testing (PT) and Magnetic Particle Testing (MT). However, the knack of confirming visual conditions such as brightness and the distance of observation depends on the capability of an inspector in PT and MT. Therefore, the smallest line-shaped indication which inspectors could recognize was investigated. Based on the smallest indication value, a calibration test piece for PT and MT was manufactured. If this calibration test piece is set up on the inspected surface in an actual inspection so that the line pair equivalent to indications shown by PT and MT can be discriminated, brightness and the distance of observation could be easily controlled. Consequently, it could be also expected for PT and MT that calibration on proper visual conditions was easy. So calibration methods of visual conditions for PT and MT are suitably proposed for establishing the judgment standardization of judgment of defects.

1 INTRODUCTION

Penetrant testing (PT) and magnetic particle testing (MT) have been widely applied to the detection of surface defects as a non-destructive testing in the industry. The eyesight of an inspector and brightness of the surface are important to discriminate the defects. The certification system requires that qualified inspectors shall have eyesight, technique and knowledge necessitated for PT and MT. Recommended brightness, which is important in the inspection of the inspected surface is established. However, it is not always effective at the spot because a gauge for brightness is needed. If anything corresponding to the reference test piece could clarify the brightness of the inspected surface and the distance of observation used for PT/MT, the gauge could be more effectively used at the spot and the inspection conditions could be standardized. At radiographic testing (RT), for example, an inspector can easily adjust brightness and the distance of observation by the penetrometer photographed in the film. It is expected that the calibration test piece, in other words, the reference test piece, corresponding to the penetrometer used at RT, is more effectively used for adjustment in brightness of the inspected surface and the distance of observation in PT/MT.
For that reason, the resolution for discrimination of the indication was investigated and calibration test pieces were fabricated for adjustment in brightness and the distance at PT/MT. The calibration test pieces were also used for PT/MT by remote visual test using image processing. In this paper, visual conditions and calibration methods of visual condition for PT and MT are suitably proposed.

2 SMALLEST LINE-SHAPED INDICATION

2.1 CHART FOR JUDGMENT OF LINE-SHAPED INDICATION
Due to penetrant testing regulations by direct visual test, a porosity (a round defect) whose diameter is under 1.0 mm is allowed, but no crack (line-shaped defect) is allowed, and the allowable defect is not clear. Therefore, the smallest line-shaped indication which inspectors could recognize was investigated [1]. A chart for judgment of line-shaped indication where very small red indications were printed was used.

2.2 EXPERIMENTAL PROCEDURE
Fourteen qualified inspectors, all with at least 3 years of experience in penetrant testing, judged the very small red indications, which were printed on a white chart, by direct visual testing. The 36 small red indications whose width and length were different sizes were printed and tested.

2.3 RESULTS OF DIRECT VISUAL TESTING
The results of direct visual testing are shown in Fig. 1. The smallest line-shaped indication size that the qualified inspectors could recognize in direct visual testing was 0.15 mm in width and 0.5 mm in length.

Fig 1: The smallest line-shaped indication

3 LINE PAIR

3.1 RESOLUTION CHART
The relation between the smallest line-shaped indication (0.15 mm × 0.5 mm) and the TV camera resolution was tested. A NIST(National Institute of Standards and Technology) (NBS : National Bureau of Standards) resolution chart was used to find the line pair that is equivalent to human eyesight. Line pair, which means the value of resolution, is defined by a line pair of black and white per mm.

3.2 EXPERIMENTAL PROCEDURE
A 3-CCD camera was used to take pictures of the NIST resolution chart and the reference test piece on which the smallest line-shaped indication (0.15 mm×0.5 mm) is printed, and the pictures were displayed on a 15-inch CRT monitor to 16 qualified inspectors [2,3]. In this experiment, 3 different visual fields were used. They were 0.15 mm/pixel, 0.1 mm/pixel and 0.075 mm/pixel.

3.3 RESULTS OF REMOTE VISUAL TESTING
The results of remote visual testing are shown in Table 1. In the case of 0.15 mm/pixel, the recognition rate of the smallest line-shaped indication (0.15 mm×0.5 mm) was 50 %. With the view field of 0.1 mm/pixel, however, the recognition rate of the smallest indication was 100 % for the inspectors. At that view field, the limit of line pair that the inspectors could discriminate was 3.6. Consequently, it was found that the 3.6 line pair can be discriminated in the view field of a TV camera and is assured to be equivalent to human eyesight by direct visual testing [1,2,3].

Table 1: Results of remote visual testing

4 CALIBRATION TEST PIECE

4.1 ESSENTIAL CONDITIONS
The colors presented on the NIST resolution chart are white or black, which is equivalent to the colors indicated by PT/MT. Essential conditions required for the calibration test piece at PT/MT are shown as follows [4] .

1. Hue
   Hue on the calibration test piece shall be equivalent to colors of indication at PT/MT so that illumination on the inspected surface can be confirmed to be suitable. Red corresponding to penetrant shall be indicated for PT and fluorescent green corresponding to fluorescent magnetic particles shall be indicated for MT.
2. Background color
   If the red penetrant in not sufficiently manually wiped away with a cloth at PT, the penetrant sometimes remains. Line pairs, which correspond to indications, shall be discriminated from the background for PT/MT. If the calibration test piece of clear plate, where line pair indications are printed, is placed near the inspected surface, line pair indications are easily put upon the background and are more effective to calibrate visually.
3. Smallest indication
   If the smallest line-shaped indication (0.15 mm×0.5 mm) is printed on the calibration test piece, the line-shaped indication can be easily discriminated from false indications.
4. Scale
   The size of defect can be easily certified with a scale.
5. Line pair
   If the limit of line pair (3.6 LP/mm) that the inspectors could discriminate is printed on the calibration test piece, the use of the piece can be assured to be equivalent to human eyesight.
6. Tolerance for line pair
   The resolution (3.6 LP/mm), which is necessary in visual testing, means about 0.14 mm in width and interval. It is important how accurate the width and interval of line pair is fabricated. The picture of the smallest indication, 0.15 mm, is displayed on a CRT monitor. That is why the target of tolerance for line pairs is ± 0.025 mm (a quarter of a 0.1 mm/pixel).

4.2 FABRICATION OF CALIBRATION TEST PIECE
The calibration test pieces were fabricated, based on the conditions above [4] . Large calibration test pieces and small calibration test pieces were fabricated for both PT and MT, considering the use at the spot. These calibration test pieces can be used for calibration by not only direct visual testing but also remote visual testing using image processing.

1. Large calibration test piece
   A large calibration test piece fabricated for PT is shown in Fig. 2. The large calibration test piece, made of transparent polycarbonate resin (0.5 mm in thickness), is 65 mm in length and 50 mm in width. The smallest line-shaped indication (0.15 mm×0.5 mm), the smallest round indication (f1.0 mm), and line pairs (maximum 3.6 LP/mm : the visual resolution) are printed on the pieces. The smallest line-shaped indication and line pairs are printed horizontally and vertically. A scale is also printed at 1 mm intervals on the circumference so that the size of indication can be certified. For PT, they are printed with the red (Munsell 1R4/14) corresponding to the hue of PT indication. For MT, they are printed with the fluorescent green (wavelength 540 nm) corresponding to the hue of MT indication.

   Fig 2: Large calibration test piece

2. Small calibration test piece
   A small calibration test piece fabricated for PT is shown in Fig. 3.

   Fig 3: Small calibration test piece

   The small calibration test piece, made of transparent polycarbonate resin (0.5 mm in thickness) like a large calibration test piece, is 10 mm square. Line pairs (3.6 LP/mm : the visual resolution) are printed on the pieces not only horizontally but also vertically and obliquely. A scale is also printed at 1 mm intervals on the circumference so that the size of indication can be certified. They are printed like large calibration test piece.

4.3 EVALUATION OF CALIBRATION TEST PIECE
In remote visual testing using image processing, pictures of PT indication and a small calibration test piece were taken in the visual field of 0.1 mm/pixel by a 3-CCD TV camera. The level on the detection and size of indications, which was investigated, was compared with the results by direct visual testing.
The results are shown in Table 2. It was found that the size of indications by remote visual testing is equivalent to the size by direct visual testing. If this calibration test piece is set up on the inspected surface in an actual inspection so that the line pair equivalent to indications shown by PT and MT can be discriminated, brightness and the distance of observation could be easily controlled. Consequently, it could be also expected for PT and MT that calibration on proper visual conditions was easy and standardization on judgment of defects was established.

5 PROPOSAL ON CALIBRATION METHODS OF Visual condition

It is suitably proposed that proposal on calibration methods of visual conditions for PT and MT based on evalation of test pieces are shown as follows.
1. The light intensity and the distance of observation shall be assured so that the qualified inspectors can discriminate the resolution 3.6 line pairs per mm.
2. In case of remote visual testing, the resolution,3.6 line pares per mm can be discriminated, that is to say, 3 lines of line pairs displayed on a CRT monitor are visible.

6 CONCLUSIONS

The smallest line-shaped indication which inspectors could recognize was investigated. Based on the smallest indication value, a calibration test piece and proposal on calibration methods of visual conditions for PT and MT were suitably proposed. If this calibration test piece is set up on the inspected surface in an actual inspection so that the line pair equivalent to indications shown by PT and MT can be discriminated, brightness and the distance of observation could be easily controlled. Consequently, it could be also expected for PT and MT that calibration on proper visual conditions was easy and standardization on judgment of defects was established.

ACKNOWLEDGMENT

The research was executed as part of the HET (High Efficiency Test) project of Japan Power Engineering and Inspection Corporation (JAPEIC) supported by the Ministry of International Trade and Industry (MITI). The authors would like to express thanks to the HET committee (Chair : Emeritus Professor Fukuhisa Matsuda of Osaka University) as well as to MITI of Japan.

REFERENCES

1. I. Tanaka, T. Taguchi, T. Asano, K. Sakai, T. Sato and M. Takagi, "Guideline for penetrant testing on welds by indirect visual test,"Proc. of the seventh International Conference on Nuclear Engineering (1999), ICONE-7480.
2. T. Asano, K. Sakai, T. Taguchi, I. Tanaka, T. Sato and M. Takagi, "Penetrant testing on welds using image processing," Proc. of the 4th International Conference on Quality Control by Artificial Vision (1998), pp. 497-503.
3. T. Asano, K. Sakai, T. Taguchi, I. Tanaka, T. Sato and M. Takagi,"Application of Image Processing to Penetrant Test on Welds" Proc. of the 2nd Japan-US Symposium on Advances in NDT (1999), pp. 202-207.
4. I. Tanaka, T. Asano, K. Sakai, T. Taguchi, T. Sato and M. Takagi,"Proposal on Calibration Test Piece for Penetrant Testing and Magnetic Particle Testing," Proc. of the 2nd Japan-US Symposium on Advances in NDT (1999), pp. 439-444.

© AIPnD , created by NDT.net

|Home|    |Top|