Defence Science Journal, Vol49,No 4,August 1999,pp.

339-342
(91a99, DESDOC

SHORT COMMUNICATION

Nondestructive Testing Method for Finding

out the Defects in a Composite Liner

C. Muralidhar and Sheri George

Defence Research & Development Laboratory, Hyderabad - 500 058.

ABSTRACT

A composite liner of carbon phenolic has been inspected by ultrasonic, X-ray radiography and
X-ray computed tomography (CT) to find defects like delaminations, debonds, voids, foreign inclusions, etc.
The geometry, detection of multiple defects and porosity of the liner make ultrasonic testing (pulse-echo and
drycoupling) difficult for inspection. X-ray radiography being a non-contact technique finds multiple defects
but compresses the structural information of 3-D volume into a 2-D image and interferes with overlying
and underlying areas of the object. X-ray CT generates an image of a thin and cross-sectional slice of
an object. R e linear attenuation coefficients in terms of Hounsfield values have been measured, compared
and correlated with CT images at the contrasts observed. 3-D images can be generated by stacking
2-D cross-sectional images of the slices. These 3-D images can be cut at any angle of choice for mapping
the extent of delaminated/debonded areas. This type of information is difficult to obtain with conventional
non-destructive testing techniques.

1. INTRODUCTION X-ray radiography is a non-contact technique

The composite liner under study is a glass which eliminates couplant contamination. The technique
phenolic cylindrical liner of small diameter (approx. is able to detect delarninations, voids, foreign inclusions
30 mm) and having 7.5 mrn thickness. These liners and porosity in the liner. It can detect even multiple
are relatively porous than their epoxy class of composite. defects and debond also if the liner is bonded to the
The liner under study was difficult to be inspected by aluminum casing. However, it compresses the structural
pulse-echo testing (PET) because the couplant seeps information of 3-D volume into a 2-D image, the
through porous structure, thereby causing contamination images interfere with underlying and overlying areas
and inconsistent results. PET probes are difficult to of the object and a single radiographic image does not
sit properly on the small diameter of the liner and also provide sufficient information to localise a feature in
difficult to detect multiple defects. Ultrasonic drycoupling 3-D.
technique eliminates the couplant contamination through X-ray CT generates an image of thin and cross
its rubber-tipped probes. It gives attenuation values sectional slice of an object. CT image represents point-
through transmission mode but it is difficult tc by-point linear attenuation coefficients in the slice. The
interpret the defects. It also suffers from sizing the linear attenuation coefficients in terms of Hounsfield
defects. values(HU) have been measured, compared, and

Received 03 June 1999

 DEF SCI J, VOL 49, NO 4, AUGUST 1999

Figure 1. Tangential radiograph of glass phenolic cylindrical Figure 2. Tangential radiograph of liner and casing showing
liner showing the delaminations and foreign multiple delaminations and debonds.
inclusions.

correlated with CT images of the liner (with and equipment from Richseifert, Germany (320 kV, 13 mA
without casing) at the contrasts observed. The density max). The focal spot of 1.5 rnrn x 1.5 mm was used.
profile of the liner (at the location marked) clearly The liners were inspected using a medical CT system
distinguishes interply density variations from debondsl (Zytec-2000i) from GE, USA. The linear attenuation
delaminations. 3-D images can be generated from the coefficients in terms of HU values were measured,
2-D cross-sectional images of the slices. These 3-D compared and correlated with the CT images at the
images can be cut at any angle of choice for mapping contrasts observed. The voltage, current and time of
the extent of delaminatedldebonded areas. This type exposure were 120 kV, 60 mA and 2.7s, respectively.
of referencing is not possible with conventional NDT A slice thickness of 1 mm was selected. In medical
techniques. CT, the HU value for water and air is 0-1000,
In earlier study1, ultrasonic, radiography and respectively and the values for the remaining materials
thermography were used on a glass phenolic cylindrical are assigned accordingly.
liner to study the delamination. In another study,2 the 3. RESIJLTS & DISCUSSION
debond problem was studied using ultrasonic, radiography
Figure 1 is the tangential radiograph of the liner
and CT on a glass phenolic liner bonded to an
showing multiple delaminations and foreign inclusions.
aluminum casing. In the present study, ultrasonic,
Figure 2 is the tangential radiograph of the casing with
radiography and CT have been used on the liner (with
liner bonded showing multiple delaminations with
and without bonding) to find defects like delaminations,
debond. It is difficult to detect these defects either by
debonds, foreign inclusions and interply density
PET or drycoupling technique. It is difficult to
variations, etc. The relative advantages and limitations
of these techniques in assessing the above defects have distinguish resin richhesin-starved areas from
been brought out. delaminations/debonds in radiography because the
contrast is not appreciable. Figure 3 shows the cross-
2. EXPERIMENTAL PROCEDURE sectional image of the liner. The variation in HU values
The cylindrical composite liners under study were observed in Fig. 3 is due to interply density variations.
of high silica glass phenolic compound (thickness: Figure 4 shows the density profile of the liner at the
7.5 mm, outer diameter: 30 mm) with and without location marked. It is clear from Fig. 4 that HU values
bonding, to aluminium casing (thickness: 2 mm). The measured did not cross zero or negative within the
liners were made by hand layup method and bonded liner, whereas HU values 0- 1000 was observed in the
to a aluminium casing with an epoxy adhesive. hollow portion of the liner(air is present), which reveals
Ultrasonic drycoupling was done using Sonatest UFD- that the contrasts observed are not due to air gaps
200. X-ray radiography was carried out using an but due to interply density variations within the liner.
 MURALIDHAR & GEORGE: NONDESTRUCTIVE TESTING METHOD FOR FINDING OUT DEFECTS IN A COMPOSITE LINEP

Figure 3.' C r o ~ o n a image

l of tbe liner showing interply Figure 4. Density profile of the liner at the location marked
density variations.

These contrasts may be misinterpreted as delaminations (bonded and unbonded). Radiography could detect
in radiography, whereas CT identifies and distinguishes delaminations, debonds, and foreign inclusions. The
them as interply density variations. This type of contrasts are difficult to resolve and may require
referencing is impossible with other NDT techniques. another NDT technique to support the data observed
The interply density variations can be found only by in radiography. CT revealed all the defects, including
CT because it can generate cross-sectional images of
interply density variations. The measurement of HU
the object. Figure 5 shows the areas of debond
at the contrasts could clearly distinguish delaminationsl
between casing and liner, and delaminations within the
debonds. Further, the density profde could clearly
liner. This demonstrates that CT could detect multiple
demark the delaminationddebondsfiom interply density
defects. Figure 6 shows the images of planes cut fiom
variations. This type of self-referencing is not possible
3-D data of aluminum casing with liner indicating
with conventional NDT techniques. Thus, CT has
clearly the debond (shown left) and delamination
(shown right). emerged as a powerful nondestructive evaluation tool
over conventional NDT techniques in assessing and
4. CONCLUSIONS analysing various defects and increasing the confidence
Ultrasonic (PET, drycoupling) techniques were level in the detection of ddect.
not able to detect v d u s defats in glass phenolic liner

Figure 5. Areas of debond between casing and liner, Figure 6. Images of planecut from 3-D data showing debond
delaminations within liner. (left) and delaminations (right).
 DEF SCI J, VOL 49, NO 4, AUGUST 1999

REFERENCES

1. Muralidha, C. Srinivasa Redd~,K. A study 2. Muralidhar, C. Debond study in rn aluminum

of delamination in a high silica glass phenolic casing with composite liner using
I
cylindrical liner by ultrasonic, X-ray radiography ultrasonic, X-ray radiography and X-ray
and thermography techniques. Proceedings of computed tomography. Proceedings of the ASNT
the third International Workshop on Advances Fall Conference and Quality Testing Show, 19-
in Signal Processing for Nondestructive Evaluation 23 October, 1998. Nashville, Tennesse, USA.
of Materials, 5-8 August, 1997, Quebec, Canada,
pp. 269-75.

Contributors

Dr C Muralidhar obtained his PhD (Polymer/Ceramic Composites) from Indian Institute of

Technology, New Delhi. He joined DRDO at the Defence Research & Development Laboratory
(DRDL), Hyderabad, in 1988. He has been working in the field of nondestructive evaluation (NDE)
of composite components since 11 years using various techniques, such as ultrasonic radiography
and thermography. Presently, he is working in a team, which is developing industrial computed
tomography (ICT) for NDE analysis of various composite components. He has published 18 research
papers in national and international conferences. One of his papers was awarded as the best technical
paper at a national conference (NDE-98), held at Thiruvanthapuram in 1998. He has been conferred
with two DRDO Awards for his significant contribution in NDE of composites.

Mr Sheri George obtained his BE (Mechanical Engineering) from Govt College of Engineering,
Thiruvanthapuram, in 1992. He did his Masters in Computer Integrated Manufacturing from PSG
College of Technology, Coimbatore, in 1995. He joined DRDL as Senior Research Fellow in 1995.
.
He is a member of the team which is working on the indigeneous development of industrial computed
tomography. He has published four research papers in international and national conferences. One
of his papers was awarded as the best technical paper at a national conference (NDE-98) at
Thiruvanthapuram, in 1998.