TECHNOLOGY DEVELOPMENT ARTICLE

BARC NEWSLETTER

Radiography of 140 mm Thick Weld- Multiple Film

Technique
S.P. Srivastava, S.P. Pandarkar, K.B. Santhosh, G.P. Sahu and S.B. Jawale
Centre for Design and Manufacture

Abstract
Radiography of 140 mm thick weld joints of Low Alloy Carbon Steel Cruciform Test Specimen was a
challenging task due to radiation hazard attributed to large exposure time, high energy and high strength
radiation source. High thickness of steel also adds to internal scatter within the specimen, reducing the
quality of radiographs. Multiple film technique, which uses more than one film of same or different speed
in a single cassette, was developed for panoramic exposure of four weld joints using cobalt 60 source. The
technique was able to improve the quality of the radiographs besides reducing the exposure time by one
third. This paper presents the detail of multiple film technique and its role in reducing the effect of thickness
variation on the image quality.

Introduction
Under the Component Integrity Test Programme, achieved by machining the weld edge to compound
Low Alloy Carbon Steel Cruciform Test Specimen bevel angles, thus making the groove narrower than
are manufactured at Centre for Design and the conventional groove. Ultrasonic testing (UT) was
Manufacture (CDM) for biaxial load testing as a part not feasible because of number of notches adjacent
of studies pertaining to safety aspects of Indian to the weld joint. Since radiographic testing (RT)
Pressurised Heavy Water Reactor. Specimens are was the only suitable NDE method for the
made out of ASTM 508 Low Alloy Carbon Steel (20 volumetric examination of the weld, it was employed
MoNi55) in the shape of cruciform having unequal to detect lack of fusion (LOF) in the side wall, lack
arms each of length 3300 mm and 1900 mm. Four of penetration (LOP) at the root, porosity, crack and
slabs, each 135 mm thick x 300 mm wide, are any other discontinuities existing in the weld. RT of
welded to a symmetrical cross having two equal thick specimen has inherent problems such as
arms of length 175 mm (Fig. 1). Manufacturing requirement of high strength source, very high
process includes: Cutting of centre block and arms, internal scatter, high radiation hazard due to longer
Machining of centre block and arms for edge exposure time, low sensitivity for the portion away
preparation, Machining of notches and slots in the from the film. Any attempt to align the source with
centre block, Welding of arms with the centre block, weld edge in order to detect lack of side wall fusion
Radiographic testing of weld joints and Post weld gives rise to large thickness variation demanding
heat treatment. different exposure time for different shots. At CDM,
As per the standard practice, symmetrical double V multiple film technique for panoramic exposure was
groove was adopted for butt weld joints between developed to radiograph all the four weld joints
arm and centre block, considering very high thickness simultaneously. On account of reduced exposure
of 135 mm of the base metal. Further, in order to time and less number of shots, this technique offers
limit the distortion, heat input was minimized by advantages of less radiation hazard, reduced scatter,
reducing the volume of weld deposition. This was high sensitivity and high productivity.

38 I ISSUE NO. 338 I MAY - JUNE 2014

 BARC NEWSLETTER
TECHNOLOGY DEVELOPMENT ARTICLE

Fig. 1: Cruciform Test Specimen

Selection of NDT Method zero degree angle of incidence for better detectability
Welds, with thickness less than 50 mm are examined of LOF. In this case, due to presence of slots in the
by RT due to ease of interpretation and availability scanning area UT was difficult, hence RT was
of image record. However, in case of RT of thick considered suitable NDT method for volumetric
weld, radiation hazards are more due to high examination of the weld.
exposure time and high strength of radiation source Theoretical Considerations for RT of Thick
required for radiography. Therefore, UT with no Weld
radiation hazard is considered a better choice for In RT, direction of radiation beam with respect to
thickness 50 mm and above. Further, UT as the orientation of the flaw plays an important role
compared to RT, gives better sensitivity for planar for the flaw detectability. If the central beam is
defects such as lack of fusion and crack. In order to aligned with the major dimension of the flaw, the
detect lack of side wall fusion, welds are subjected probability of flaw detection is very high. In groove
to ultrasonic testing using angle beam technique. weld, side wall and root are more vulnerable as far
Particularly for thick weld, obstacle free large as defect is concerned. Radiographic shooting sketch
scanning area on both sides of the weld shall be (RSS) for butt weld with double V groove is shown
available so that sound beam hits the side wall at in the Fig. 2. For single V groove weld having

ISSUE NO. 338 I MAY - JUNE 2014 I 39

 TECHNOLOGY DEVELOPMENT ARTICLE
BARC NEWSLETTER

Fast film records thicker portion whereas slow film

gives the detail of thinner portion.
In the second case, in order to reduce the exposure
time, normally applicable for very thick job, two
films of the same speed are exposed together which
reduces the exposure time by one half. Combination
of these films, when viewed separately will not meet
density range of 1.0 3.5 individually and therefore
will not reveal any detail. Such exposed films are
Fig. 2: RSS of double V butt weld in plate having viewed together i.e. by super imposing one on
thickness 80 mm or more another in front of the viewing illuminator. The
thickness less than 10 mm, even source at S1, combined density of two films shall meet the density
aligned with root position can detect LOP in the range of 1.3-4.0. In the technique developed at
root and LOF in the side wall as well. However, as CDM in total two pairs of film were used, as
the thickness increases, three different shots with explained in the subsequent paragraph, to get the
source positions S1, S2a and S3a are required in advantage of both improved latitude and reduced
order to detect LOP at root and LOF at each face. exposure time.
Object away from the film produces large Radiogr ap hic Testing of Weld Joints of
geometrical unsharpness and therefore image of the Cruciform
weld groove which is away from the film will not
Weld groove for butt joint in 300x135 mm cross
be discernable, if thickness of the weld is more than
section consists of symmetrical double V with 2
80 mm. Hence, beyond 80 mm thickness it is
mm root land. Each V groove, includes two bevel
required to take shot from both sides of the weld, if
angles; first one 75 starting from root up to 20
accessible, to limit the unsharpness. Source positions
mm height and second one 24 starting from 20
S2a, S3a, S2b, and S3b are suitable for detecting
mm up to 47.5 mm height (Fig. 3). Qualified welding
LOF in walls GH, CD, EF and AB respectively. In
procedure specification (WPS) and qualified welders
case of cruciform, since the weld metal thickness
were used for welding. All weld edges were checked
including reinforcement was 140 mm similar
for presence of any lamination by liquid penetrant
shooting scheme was adopted for full volumetric
test (PT). Root was fused by GTAW, and remaining
examination of the weld.
weld was deposited by SMAW process. Pre-heating,
Multiple Film Technique- Theoretical Aspects with inter pass weld temperature control was done
In multiple film technique of radiographic testing, to avoid the formation of martensite and to prevent
two or more films kept in close contact, are exposed hydrogen induced cracks. Post weld heat treatment
together either to cover the varying thicknesses of
the component or to reduce the exposure time. Films
in combination may be either of same speed or
different speed depending upon the requirement.
In case of component of varying thicknesses, to
improve the latitude i.e. range of thicknesses covered
on the radiograph within acceptable density range
of 1.0 to 3.5, combination of fast and slow film is
exposed together and each film is viewed separately. Fig. 3: Weld groove configuration

40 I ISSUE NO. 338 I MAY - JUNE 2014

 BARC NEWSLETTER
TECHNOLOGY DEVELOPMENT ARTICLE

was adopted basically for stress relieving. Root pass shots with source at an angle with respect to weld
welding and inter stage welding were subjected to axis WW. Considering the symetricity of all the
RT and PT to detect any gross defect whose welds in each arm with respect to central axis XX,
elimination after final pass requires lot of material it was decided to use panoramic shot (PS) to
removal by grinding process. radiograph all the four weld joints simultaneously
In order to detect lack of fusion in the side wall in one shot, in order to reduce the total exposure
each weld joint in each arm requires at least five time. In this set up, source is required to be
shots- one straight and four inclined shoots, two positioned either at PS1 or PS2 on the central axis
from each face. Though thickness penetrated in XX of the cruciform such that the source to film
straight shot is 140 mm, interpretation is carried distance (SFD) remains same for all the four weld
out only for half the thickness i.e. 70 mm of the joints with beam parallel to the walls inclined at 37o
weld which is closure to the film. Straight shot (SS) with the vertical.
in which source and film are kept at SS1 and F1 Use of Multiple Film Technique for RT of
respectively (Fig. 4), the total volume covered, is Cruciform
ABCDEx300 mm (Fig. 3). In this shot, because of
3
Straight Shot: For straight shot with source at
its diverging nature, penetrating radiation is more SS1 and film at F1 (Fig. 4), multiple film technique
or less parallel to faces AB and CD which are inclined was used in which two films of same speed loaded
at 12 with the vertical. Therefore, these shots are in the same cassette were exposed to half the
suitable for detecting any LOF in faces AB & CD exposure time required for single film. Such exposed
and LOP at the root CO. Similarly by keeping source films were viewed together i.e. by super imposing
at SS2 volume MNOPQ can be examined with high one on another in front of the viewing illuminator.
degree of confidence for detecting LOP in faces MN Since penetrameter was kept on full thickness,
and PQ. minimum SFD obtained by Equation 1 (Table 1) for
Since faces BC (or CD) and NO (or OP) are inclined T = 140 mm, ug = 1.7 mm and = 7.5 mm was
at 37o with the vertical and job thickness is more 950 mm. During radiography, actual SFD was
than 80 mm, to increase the probability for detecting increased to 1100mm to minimize the divergence
LOF, it is necessary to take another radiograph with of the beam with respect to the face AB or DE.
source in offset position, such that radiation beam Exposure time for D7 single film using 40 curie
is parallel to these walls. This requires two inclined cobalt source was estimated to 6 hour.
Inclined and Panoramic Exposure: Thickness
across the weld cross section, seen by the radiation
beam is different, because of the inclination and
therefore it becomes essential to use multiple film
to accommodate variation of optical density within
acceptable limit. Radiographic setup for the
panoramic exposure is shown in the Fig. 5. With its
entire arm in horizontal plane, cruciform was kept
above the source to take advantage of the shielding
during panoramic exposure.
Ray diagram, using AutoCAD was prepared to know
the source position, source to film distance and
Fig. 4: RSS for panoramic shot for all four arms
thickness variation for a set up which will give

ISSUE NO. 338 I MAY - JUNE 2014 I 41

 TECHNOLOGY DEVELOPMENT ARTICLE
BARC NEWSLETTER

parallel to the face BC within 6 and geometrical

unsharpness was within the limit.
For source at 600 mm below the top surface of the
cruciform on vertical axis XX, thickness, angle, and
SFD with exposure time for each combination, are
tabulated in Table 2. Exposure time were calculated
using Equation 2, which is basically applicable to
narrow beam geometry and is meant only as a guide
for knowing approximate exposure time. Actual
exposure time is established by trials rather than
Fig. 5: Set up for panoramic exposure formula as it depends on the accuracies of many
variables such as source size and its strength, SFD,
adequate coverage of the weld with central beam film processing time/condition, scatter radiation etc.
parallel to face BC. Values of geometrical For very thick specimen actual exposure time is less
unsharpness were calculated using Equation 1 for than the calculated one, due to internal scatter
three SFDs and thicknesses as shown in the Fig. 6. contributed by the specimen itself. During trial shot,
From the results of several iterations, with varying using fast film Agfa D7, it was observed that even
SFD and thickness, it was observed that for source with low exposure time, film density for the thickness
at 600 mm, gamma ray from a point source was

Fig. 6: Ray diagram for panoramic exposure using multiple film

42 I ISSUE NO. 338 I MAY - JUNE 2014

 BARC NEWSLETTER
TECHNOLOGY DEVELOPMENT ARTICLE

Fig. 7: Single film with 1.63 optical density Fig. 8: Double film with 3.32 optical density
Ta and SFDa, was on higher side and therefore to Interpretation of Multiple Films
offset this, slow film Agfa D4 was considered a
During interpretation of radiographs, two D7 films
better option as compared to Agfa D7 film.
were viewed together for thicknesses Tb and Tc.
To reduce the exposure time, combination of fast For thickness Ta, because of adequate density only
(D7) and slow (D4) films was found suitable for one film either D7 or D4 can be viewed individually,
lower thickness region Ta whereas combination of however, if D7 and D4 are viewed together, better
two fast films (D7) were adequate for higher perception of the IQI image is obtained. As per the
thickness regions Tb and Tc. To take care of the ASME requirements, image of the 4T hole in 60
continuous variation in the thickness and for the number plate type IQI was seen when both the films
continuity of the weld image, one large size D7 were viewed together. In addition to plate type IQI,
film, overlapping the total thickness range was used. wire type IQI No.1 ISO 7 (DIN) was also used, for
All the three films with their relative positions, were which 5th wire image was seen. Weld at intermediate
loaded in a single cassette. Film were exposed for 3 stage having thickness of 110mm was radiographed
hours 30 minutes, which is approximately one half using two films of same speed at half the exposure
the maximum exposure time Tb = 6 hours 12 time required for single film. Fig. 7 and Fig. 8 are
minutes and greater than the half of the minimum radiographic images of single film and superimposed
exposure time i.e. 3.9/2 hours. double film respectively. On single film, with optical
Tungsten arrow markers were placed on the film as density 1.63, even image of the fourth wire of a
well source side to show the coverage of the weld wire type penetrameter is not seen clearly, whereas
width and projection of the source side weld image two films superimposed and viewed together is able
on the film. In total three image quality indicators to reveal well defined image of the fifth wire and
(IQI), two for source side and one for film side were slightly faint indication of the sixth wire. Besides
placed to check the adequacy of the image quality. wire, images of letter B and J4 are clearly seen in
Fig. 8.

Table 3: Comparison between single and multiple film technique

S. No. Exposure Type Source Position Shots Time Shots for Total Time (hour)
per arm per shot (hour) four arms Single Film Multiple Film
1. Directional SS1 & SS2 2 6 8 48 24
2. Directional PS1 & PS2 2 6.2 8 49.6 24.5
3. Panoramicl PS1 & PS2 2 6.2 2 12.4 6.2

ISSUE NO. 338 I MAY - JUNE 2014 I 43

 TECHNOLOGY DEVELOPMENT ARTICLE
BARC NEWSLETTER

Conclusions Multiple film technique with panoramic exposure

Radiography of thick weld requires proper planning not only reduced the exposure time but also
to take care of flaw detectability, scatter radiation, minimized the density variation along the composite
high exposure time, large value of unsharpness, film from one end to another end.
shielding to reduce the radiation hazard, IQI References
sensitivity etc. At CDM, multiple film technique was 1. Richard A. Quinn and Claire C. Sigl,
developed to address these problems. Review of Radiography in Modern Industry, Fourth edition.
exposure time (Table 3) indicates that total exposure EASTMAN KODAK COMPANY, Rochester, New
time for four welds of a cruciform is 97 hour when York 14650
single film is used with directional exposure. Whereas
2. Nondestructive Inspection Methods Manual,
time for panoramic exposure combined with multiple
Section VI, Special Radiographic Techniques,
film technique is only 30.2 hour, which is
TM-1-1-1500-335-23, Integrated Publishing,
approximately one third (31%) of the total time.
[Link]

44 I ISSUE NO. 338 I MAY - JUNE 2014