·Home ·Table of Contents ·Standardization

Evaluation of film System by a Radiographic facility Mario Bianchi, Fiat Avio Turin Antonio Liscio, Fiat Avio Turin Umberto Piazza, Alenia Spazio Turin Franco Baratta, Bytest Turin Contact

Introduction

ASTM E1815 in United States and EN 584-1 in Europe give methods for testing a Film System (film, chemicals and processing) by measuring numerical parameters intended for objective standard classification and comparison. Practically these methods show following limitations:
• special instruments are needed that generally are not available in a radiographic facility;
• film samples are produced in specific conditions (220 kV, lead screens, copper filter) that can be very different from standard ones being used in a specific radiographic field;
• generally film manufacturers supply classification for theirs own chemicals only based on regular surveillance in practical conditions ("Certified Film System"= CFS);
• sometimes some film manufacturers give results of measurements of ASTME1815/EN584-1 parameters using chemicals of another manufacturer but he cannot assure the classification in case of changes or out of tolerance of the part of the Film System that he cannot control ("Alternate Film System" = AFS).

Apparently the most practical solution seems CFS exclusive use but, sometimes, AFS use cannot be avoided due mainly to the following technical reasons:
• generally multiple processors with different chemicals are not available (and practical) for a radiographic facility;
• film manufacturing artifacts can be occurred with manufacturer unable to supply same class films for many months (typically where a not-large production film type is used);
• film manufacturers do not produce any size and/or package for any film type.

Obviously economic reasons also can require evaluation of different Film Systems.
This work gives guidelines for CFS/AFS evaluation on the basis of contrast measurement, standard penetrameter equivalent sensitivity and natural cracks detectability. Film samples have been produced using typical conditions for aeronautic/aerospace field where Film System choice can be very important for fine cracks detection. This method can integrate general ASTMEI8I5/EN584-1 classifications in order to enable any Level III radiologist to compare CFS and/or AFS or to approve an AFS when a CFS cannot be used.

Film Systems being tested

42 Film Systems have been tested combining 14 film types (brand a , b , g and d ) with 3 chemicals (brand a , b and g ) being available on Italian market at the end of 1999. Films classified by manufacturers as C5 or C6 according to EN 584-1 were not tested because they are not generally used where high resolution is requested and where low absorption and/or low thickness allow high sensitivity performance. Film size being used is 24x10 cm in package as available on Italian market at the end of 1999 only (special package have been avoided): 13OkV/steel test has been performed using standard lead vacuum pack, if available, or cardboard lead screens in reusable holders (screen thickness is 0.027 mm for vacuum pack or 0.05 mm Front and 0.10 mm Back for reusable holders).

Test Conditions

Equipment:	Seifert Eresco 42MF 200kV/4.5mA, Constant Potential, Beryllium window, focal spot 1.5 mm.
Samples:	Steel welded plate 5 mm thick and Al plate 6 mm thick both with natural crack.
Focal to film distance:	1 meter
kV:	130 kV on steel plate and 50 kV on aluminium plate that are consistent with standard kV range being used in aeronautical field.
Time and density:	from I to 6 minutes in order to obtain film density from 2.75 to 3.25 near the crack and from 2.4 to 2.9 on parent material where contrast is measured.
Penetrameters	for steel/130kV test ASTM 1742 step-hole penetrameters FE.20 and FE2.0 and EN 461-1 wire penetrameter 13 FE EN; for Al/5OkV test ASTM 1742 step-hole penetrameters AL.25 and AL2.5 and EN 461-1 wire penetrameter 13 AL EN;
X-ray direction :	90°without moving samples and X-ray equipment during the test.
Processing:	Automatic Processing a (27°C/130" developing), b (28°C/100" developing), g (26°C/100" developing) as suggested by chemical manufacturers.

X-ray shots for the two film sets and film identification

Exposure time was adjusted in order to obtain density range specified above for each film processing. Then 3 shots are performed for each film system (film & processing) on each material; so 126 films were obtained for each material (=3·[14 film type]·[3 film processing]). Lead identifications were randomly placed on the films from Al to A126 on aluminium sample and from F1 to F126 on steel sample; a traceability list was issued between lead identification and film system. For each material and film system one of three films was selected trying to avoid artifacts and to find the best crack detection.

Percentage Scores

Parameters for evaluation are defined in the following paragraphs (FS, C, EPS, CD, WE).

In order to combine scores Sc for different parameters together, Sc% has been calculated as following:

So results are given as FS%, C%, EPS%, CD%, WE%.

Film Speed (FS)

For each film x, FS% was calculated as following referred to the "speedest" film (SF) being found by all the processing, taking into account difference in film density (D_SF and D_x):
Table 1 shows results as FS%.

film a =	film b =	film g =	film d =
	Film Process a (white)	Film Process b (black)	Film Process g (grey)
50 kV/Al No screen
130 kV/Fe Lead screens
Table 1: Film Speed (FS%)

Contrast (C)

Using FE2.0 and AL2.5 penetrameter images ([penetrameter thickness]=20%[plate thickness]) for each film x, C_x was calculated as following referred to film b3 processed in b (b3b ):
Better C being found is 0.73 for Al and 0.85 for Fe.
Table 2 shows results as C%.

film a =	film b =	film g =	film d =
	Film Process a (white)	Film Process b (black)	Film Process g (grey)
50 kV/Al No screen
130 kV/Fe Lead screens
Table 2: Contrast (C%)

Equivalent Penetrameter Sensitivity (EPS)

EPS is calculated using equations giving in ASTM E1025 Appendix X1.l and ASTM E747 Appendix X1.1 as shown in Table 3.
In order to evaluate EPS each examiner gives number of wires or holes (N) that can be detected. Computation based on N gives EPS. Where necessary, N.25, N.50 and N.75 may be used: for example "4.25 wires" means that 4 wires are detectable very weIl, "4.75 wires" means that the 5^th wire detection is not very good, etc. The best EPS being found by an examiner is 2.38% both for Al and Fe; taking into account average value after 11 examiners the best EPS is 2.84% for Fe and 2.63% for Al. Table 4 shows results as average EPS% after 11 examiners.

Step-Hole penetrameters				Wire penetrameters
X = 6 mm (Aluminium) IQI .25 T = 0.1270 mm		X = 5 mm (Steel) IQI .20 T = 0.1016 mm		X = 6 mm (Aluminium) IQI 13ENAL		X = 5 mm (Steel) IQI 13ENFE
h (mm)	EPS	h (mm)	EPS	d (mm)	EPS	d (mm)	EPS
1.016	4.23%	1.016	4,54%	0.125	3.66%	0.125	4.39%
0.508	2.99%	0.508	3,21%	0.100	3.10%	0.100	3.72%
0.254	2.12%	0.254	2.27%	0.080	2,62%	0.080	3.14%
				0.063	2.19%	0.063	2.63%
				0.050	1.84%	0.050	2.21%
Table 3 : Computation of Equivalent Penetrameter Sensitivity

film a =	film b =	film g =	film d =
	Film Process a (white)	Film Process b (black)	Film Process g (grey)
50 kV/Al No screen
130 kV/Fe Lead screens
Table 4: Equivalent Penetrameter Sensitivity (EPS%)

Crack Detection (CD)

Following Fig. 1, 2 and 3 give some crack images being used for CD evaluation.
Both film sets have been evaluated in accordance to detection level of fine cracks (see black arrows in Fig. 1).

Fig 1a: Al sample

Fig 1b: Fe sample

An higher score is given for a better detection as shown by the following example:
• Score for each of 3 films of the 1^st group 41.0=(42+40)/2
• Score for each of 11 films of the 2^nd group 34.0=(39+29)/2
• Score for each of 15 films of the 3^rd group is 21.0=(28+14)/2
• Score for each of 9 films of the 4^th group is 9.0=(13+5)/2
• Score for each of 4 films of the 5^th group is 2.5=(4+1)/2

Then Percentage Score is calculated as:


42 groups of single films could be done; generally groups are from 4 to 10 depending on examiner.
Table 5 shows results as CD%.

film a =	film b =	film g =	film d =
	Film Process a (white)	Film Process b (black)	Film Process g (grey)
50 kV/Al No screen
130 kV/Fe Lead screens
Table 5: Crack Detection (CD%)

SampIing of different crack detection is given in Fig.2 (Al sample) and Fig.3 (Fe sample).

Fig 2a:

Fig 2b:

Fig 2c:

Fig 2d:

Fig 2e:

Fig 3a:

Fig 3b:

Fig 3c:

Fig 3d:

Fig 3e:

Weighted Evaluation (WE)

In order to obtain a flexible classification taking into account any possible requirement weighted evaluation is given as following:

film a =	film b =	film g =	film d =
	Film Process a (white)	Film Process b (black)	Film Process g (grey)
50 kV/Al No screen
130 kV/Fe Lead screens
Table 6: Weighed Evaluation (WE%) with XCD=50%, XEPS=40%, XC=10%, XFS= 0%.

Ae Classification of Film Systems

Table 7shows a first attempt of classification based on previous WE typical for a field where crack detection and penetrameter sensitivity are much more important than contrast and film speed. Where higher contrast and/or film speed are preferred, classification is different. Ae1 class limitation is obtained as 90% of maximum WE performance of films belonging to EN 584-1 Class 1; same criteria is used for class Ae2 limitation referred to EN 584-1 Class 2, class Ae3 referred to EN 584-1 Class 3 and class Ae4 referred to EN 584-1 Class 4.
Due to class limitation criteria, Ae classification can be considered as complementary to EN 584-1 classification.
To be noted that the dotted line, representing theoretic average values, splits the diagram into two different areas: films on the left side are worse ( in the table) than films on the right side ( in the table) due to lower speed against same or worse WE.

Table 7: Ae Classification being obtained

Review and Remarks

The main purpose of this work is testing a method to compare film systems without using special instruments. So films of different class are compared in order to define their classification. In general terms Ae Classification confirms classification according EN 584-1. It is to be noted that, for a better reliability, a comparison between two film types should be performed using many film samples for any type.

Ae Classification is mainly based on "subjective" detection of cracks and penetrameters.
Fig.4 gives grey profiles on crack images from Fig 2a and 3a (Class Ae1) and from Fig.2e and 3e (Class Ae4); similar diagrams could be obtained from penetrameter images. So "objective" detection could be given using signal to noise ratio.

Fig 4: Grey Profile

In our opinion, however, a so "objective" practice should be avoided due to the following reasons:

• grey profiles shall be obtained by a computerized image processor that generally is too much expensive for a radiographic facility;
• film image is manipulated by data acquisition device (TVcamera or scanner) and by image processing, so image computerized evaluation could disagree with human eye detection;
• at the moment (and, in our opinion, still for many years) an human eye driven by a human brain is much more efficient (and less expensive) for crack detection than a TVcamera or scanner connected to a computer.

Reviewing Crack Detection results against results of Equivalent Penetrameter Sensitivity (EPS), we reminded those radiologists who think that a good penetrameter image rneans certainly a good radiograph.
How can they justify for example a so large Crack Detection range (from ~ 23% to ~ 76%) for EPS%=80% ?
Our answer is that, where cracks have to be detected, each parameter shall be optimized regardless of penetrameter image that is a quality assurance requirement, not the purpose of the radiography!

Fig 5: CD% versus EPS%

Thanks to:

• D.Gattolin (Alenia-Torino), M.Scrimieri and M.Iovacchini (Bytest-Torino), L.Carminati and G.Cesaro (FBM Hudson-Terno d'Isola), S.Baldassarre and L.Randazzo (FiatAvioTorino) for cooperation in film evaluation.
• Semat Italia for Seifert X-ray equipment free lending.
• a , b, g and d film manufacturers and their Italian agents for free supply of films and chemicals not currently used.

© AIPnD , created by NDT.net

|Home|    |Top|