Table of Contents ECNDT '98
New Developments of Pulsed Holography for Industrial ApplicationsO. Petillon*, Aerospatiale
|TABLE OF CONTENTS|
To get most advantages of the technique in an industrial environment, it was necessary to develop ESPI systems adapted to pulsed laser and dynamic recording. Use of pulsed laser enables indeed to work in a noisy environment and to catch transitory events . This goal is being achieved in the project, with the design of an adapted optical set-up, the development of specific electronic and software for acquisition and reconstruction. A first prototype has been realised, that already proved its performances outside the laboratory on very different types of applications.
The main features of the prototype include : the triggering of a CCD for recording 2 images within a few microseconds, the development of a multi-pulse Ruby laser (up to 4 pulses ) and of the associated recording head. Use of the system in various areas will be illustrated below.
Fig 1: loud speaker with t = 40 µs Ø 30 field viewed on the CCD
Fig 2: loud speaker with t = a few s
Fig 3: thermal loading - honeycomb/skin debonding
Fig 4: delamination in a G/E stiffened panel 640 x 480 CCD image
Fig 5: electrical casing
Fig 6: thermal loading - rubber debonding - 1024 x 1024 CCD
Fig 7: Shock loading (0.060 ms after impact)- shearography
Fig 8: Shock loading (0.030 ms after impact) - holography
The first noticeable aspects of a pulsed holography system are its ability to operate in almost any industrial environment. The results obtained are as good when performed in a laboratory, on a optical table, as when realised on industrial premises, with machines in operation in the neighbourhood.
The tests that have been performed during this first step cover the observing of vibrations and the detection of defects in composite parts. The case chosen for vibration was a loud speaker as it enabled an easy choice of the vibration characteristics. Depending on the existing modes, different vibration figures can be observed : they vary according to the eigen modes, their frequency and amplitude.
The results presented have been obtained with a excitation frequency near 10 kHz. In the first case (see Figure 1), the separation between the two shots is small and the high frequency mode is prevalent. The second case (see Figure 2) has been obtained with a pulse separation of a few seconds. Lower frequency modes, with more energy, become significant; the results are much less reproducible and depend on the acquisition time with respect to the different present modes.
Different composite parts have been tested with holography and shearography for the detection of defects, and with different types of loading. A sandwich sample (aluminium skin, honeycomb core) has been tested for skin-core debonding. This kind of defect could be very easily detected with a thermal loading of the part (see Figure 3).
A graphite/epoxy stiffened panel, representative of a wing part, in which several impact damages are present, has also been tested with thermal loading. Due to the stiffness of the part, it is more difficult to detect all delaminations with such a loading and the contrast is not as large as in the previous case. It is nevertheless possible to visualise some damaged areas (see Figure 4).
A electrical casing has also been tested : this small aluminium part has a rubber coating and it is necessary to assure a good bonding of the rubber. The sample chosen for the tests contained several artificial defects (see Figure 5). With a low thermal loading, it is possible to visualise the larger defects (see Figure 6) : the three of them are detected in the view shown below. To detect the smaller flaws, a vacuum loading would be more adequate.
A sandwich panel with glass fibre skin and honeycomb core has been tested with shock loading. This panel contained one identified debonding. The same area where this defect was localised has been tested with both shearography (see Figure 7) and holography (see Figure 8) systems. The inspected zone is around 300 mm x 300 mm. The identified defect (marked on both figures) is clearly highlighted; there is also other information appearing on the holography image that has yet to be confirmed using another method. The various tests performed have all been realised in an industrial site with no precaution taken with regard to vibrations. The qualitative results that were obtained are as good as previous known results on optical benches. Industrial applications of pulsed holography, in the NDT area mainly, look promising and further tests should allow to validate a large scale of use.
Mechanical test monitoring
In many industrial areas (aeronautics, nuclear casks, automotive), it is necessary to model the behaviour of critical parts and to validate this model through a few mechanical tests, on a restricted scale when possible. The methods used mainly to monitor such mechanical tests are gauges and sensors giving localised data on displacement or strain. The data are used to fit the parameters of the models.
Optical methods present the advantages of providing global information on a part deformation; they allow thus to monitor 100% of the structure behaviour. The most interesting data can then be selected for fitting the model.
Fig 9: sample with inserted flaw submitted to compression test sample size on the CCD : 220 x 350 pixels
The displacement map shows a major deformation along the 135° axis. This is an effect of the positioning of the defect in the lay-off. It is however difficult to compare such data to conventional results as the range of sensitivity is very different. Pulsed holography appears as a method able to give data in the range where conventional sensors are still unusable.
Fig 10: hammer shock (before wave reflection) - 640 x 480 CCD image
Fig11: hammer shock (after wave reflection)
Fig 12: shock wave propagation (23 µs after impact)
Fig 13: shock wave propagation (43 µs after impact)
The results presented hereafter stand for the increasing complexity steps to evaluate the feasibility of pyrotechnic shock wave monitoring. The first phase was the acquisition of mere hammer shock waves on a plate. The aims were to validate the multi-pulse operation and to define an adequate trigger for the acquisition system. The raw results obtained show the propagation of the wave :
This enabled to validate the operation of the multi-pulse head but showed that, due to laser fire delay, a triggering by the event was arriving too late and thus was not sufficient to observe the front wave. A pre-trigger system has then been designed and further tests could be realised on a ball drop. The results obtained on a 40 mm ball show clearly the wave front propagation (see Figure 12 and Figure 13). The contrast of the fringes is however decreased when compared to previous results. This is most probably the consequence of the laser pulse duration ( ~ 30 ns) : during the acquisition, the wave front moves around 5000 m/s and the acquired image gives the mean out-of-plane displacement during the pulse duration.
These first results look promising : it was possible to operate and get results with a multi-pulse holography system in a harsh environment (pyrotechnic shocks), without any special adaptation. The limits that appear are essentially linked to the laser characteristics (pulse duration, pulse repetition) and should be solved with the advances in laser technology.
The last drawbacks of conventional holographic methods have been suppressed with the substitution of photographic plates and thermoplastic cameras by CCD cameras. The first results obtained show that current limitations of the technique are the laser and the CCD size. Although the laser technology is constantly improving, it is still difficult to find commercially available pulsed laser suiting holography and industrial requirements : coherence length, acceptable cost, reduced size. In return, there is a high potential for gains in the CCD domain. It should allow in the future the acquisition of larger deformation amplitudes and the study of wider objects.