As soon as radiography was discovered, one looked forward on applications to see through objects, that is to say: -either, to see an object behind another one which is considered as a mask. -or, to observe an internal part of an object. But it is surprising to observe that it took a long time to overcome this purely visual attitude and to consider radiography as a tool very well suited to description of a tridimensional object. Two kinds of reasons for that can be stated: i/the first one, purely mental, is due to the difficulty to overcome certain limits more or less formed in the brain coming from human senses. ii/ the second one is due to certain separation, at a certain period of time, between engineers` and physicists' knowledge and the mathematicians' knowledge.

In simultation experiments the consequences of hypothetical reactor accidents are investigated. The materials involved in the tests, as well as the temperatures, require test sections of steel. To follow the materials movement and relocation X-ray cine equipment was developed. The test section is radiated from three axially positioned. X-ray sources, electro optical image amplifiers transform the material density distribution into a visible picture, which is picked up by high speed cameras with, typically, 1000 pictures per second. The resolution of the system is about 1 mm3 of steel with a small dependance on the total thickness of the test section. The technique, its main features and limitations are described and some experimental results are presented as an example. The computer aided analysis of the X-ray high speed films is also explained.

Nuclear reactor safety studies have applied cineradiographic techniques to achieve key information regarding the durability of fuel elements that are subjected to destructive transients in test reactors. Beginning with its development in 1963, the fast-neutron hodoscope has recorded data at the TREAT reactor in the United States of America. Consisting of a collimator instrumented with several hundred parallel channels of detectors and associated instrumentation, the hodoscope measures fuel motion that takes place within thick-walled steel test containers. Fuel movement is determined by detecting the emission of fast neutrons induced in the test capsule by bursts of the test reactor that last from 0.3 to 30 s. The system has been designed so as to achieve under certain typical conditions( horizontal) spatial resolution less than lmm, time resolution close to lms, mass resolution below 0.1 g, with adequate dynamic range and recording duration. A variety of imaging forms have been developed to display the results of processing and analyzing recorded data.*

After reviewing the role of in-pile and out-of-pile experiments in the United States fast-reactor safety testing program, the history of fuel-motion measurements is sketched. The fast neutron hodoscope is described, with particular emphasis on capabilities and principles of operation. Alternative devices for fuel-motion measurement are examined with a discussion of future improvements in material-motion diagnostics. Relationships of fuel-motion diagnostic techniques to other types of cineradiography are also covered.*

The paper is presented in two parts. The first part deals with the investigation of basic parameters of image intensifier performance with X-ray inputs when applied to the study of dynamic events. The use of a cold neutron beam to study oil flows inside a running gas turbine is also discussed. The second part deals with the application of the developed image intensifier systems to specific studies of various problems related to gas turbines, both in the manufacturing process and in development testing.

Essential features in the dynamic behaviours of welding and cutting processes with high energy density beams are reviewed and clarified by the efficient usage of various cineradiographic diagnosises. Formation of a deep beam hole in the weld pool are described and the important effect of the front wall characters in the beam hole is demonstrated on the natures of deep penetration and defect formations such as spiking and porosity. The cutting process is also interpreted in the frame of the same physical viewpoint with the welding. A new and efficient suppression method of spiking, porosity and humping are examined and confirmed using Tandem Electron Beam developed by the author.

L'usinage Laser, realise sans contact d'outil, sans pollution ni contrainte de vide, permet une grande flexibilite de l'atelier et est parfaitement adapte A la robotisation intel-ligente.

Flash Radiography and Cineradiography allow the visualisation of high speed phenomena and the stop motion effect with recording on film of qualitative and quantitative data on the dynamic state of the matter under very intense shock waves. In this paper, we present a set of experimental devices and results obtained with a large range of flash X-ray generators : - small generators made with Marx discharge circuits coupled to void X-ray tubes, working up to 2.5 MV, - a big flash machine, GREC (presented at this conference (ref.1))used with very absor-bing materials. The presented applications illustrate a large field of experiments in the field of shock waves, interaction of 2 shock or detonation waves, flow visualisation of detonation, Taylor instabilities/metal jetting, spalling in iron...

PHERMEX, an acronym for Pulsed High-Energy Radiographic Machine Emitting X Rays, has been used as a diagnostic tool to make quantitative measurements from radiographs of inert materials under dynamic high-pressure conditions and of explosives during the detonation process. In some experiments, radiography is the best method (compared to high-speed optical cameras and contactor pins) to study complicated hydro-dynamic flow occuring in a dynamic experiment. To demonstrate the versatility and uniqueness of PHERMEX and the radiographic method, several experiments on inert solids having high and low atomic numbers will be discussed with some particulars. This includes the observation of the 11.0-GPa-pressure phase transition for antimony and the accompanying two-shock structure and the off-Hugoniot data for lead using regular reflection. Also, by careful design of a radiographic experiment, the Hugoniot state behind a shock front can be completely and precisely specified. Aluminium is an example of a material studies in this manner. PHERMEX is useful in studying some detonation properties of explosives. As an illustration, the discussion will include radiographic results of divergence characteristics of a detonation wave in sensitive and insensitive explosives as it propagates past a corner and the effect of preshocking on the detonation process of insensitive explosives when the detonation wave interacts with a region that has been shock-compressed at a pressure too low to cause detonation. The above are only a few of the hundreds of radiographic studies of shock-wave phenomena of solids and detonation characteristics of explosives. A complete list of PHERMEX experiments is published in LASL PHERMEX Data Volumes, I, II, and III, "University of California Press, Berkeley and Los Angeles, California, 1980, C.L. Mader Editor.

This paper describes the development and use of a flash X-ray system to radiograph the dynamic behaviour of projectiles on exit from gun muzzles. Details of equipment, equipment protection, films, cassettes and system triggering are presented.

This camera was developed specifically for high speed cine radiography and was designed to be used both in laboratory and field trials. The camera is programmable by digital controls and can drive flash x-ray units synchronously with framing up to a rate of 250,000 FPS. An image format of nine frames 24 x 30 mm. A gain of times 10,000 is incorporated to give amplification of low light level, short exposure x-ray flashes.

The interpretation of experiments in pulsed laser-matter interaction requires the dynamic recording of phenomenon involved in the plasma created by the high density of energy concentrated in the target. Since some years, observation and measurements in the X spectrum, and more particularly in the 1-10 keV range at a first time, has became of unquestionable interest either to study radiation self emitted by the plasmas or to X-ray it with an external X-ray source. With these applications in view we have been lead to develop space and time resolved X-ray detectors with time resolution in the picosecond range. In this paper we present a survey of our successive steps in the design and operation of streak cameras with more and more performances in terms of spatial and temperal resolution (3 to 30 ps at 15 1p.mm-1). A large family of image tubes and cameras are described which have been studied and made by collaboration with Thomson-CSF and Laboratoire d'Electronique et de Physique Appliquees (LEP). We also present the range of cameras which are now under development at Limeil, with a spectral response in soft X-rays (.2 to 2 keV). Finally we present the results obtained today, in two dimensional (framing) imagery with the 300ps planar shutter tube from LEP.

In laser implosion studies, the data collected from X-ray radiation of the target itself provide information on density and temperature as well as on the implosion mechanism. The radiography of the imploding target with an external X-ray source is a diagnostic tool who can complete or replace this more conventional technique, particularly to analyze the relatively cold and dense core of a target imploded in the ablative scheme. In the introduction we present the experimental device currently used at Limeil and we recall the main phenomena and mechanisms involved in implosion of spherical target by intense lasers. The main characteristics of the laser bear facilities OCTAL are given : λ = 1.06 μm Eight beams with a total power of : 1 TW - 50 ps (E = 50 J) 500 GW - 600 ps (E = 300 J) The targets are glass spherical microballoons with radius from 30 to 150 pm and thickness from .6 to 3 μm. Targets are pressurized until few tens of bars with an equimolecular compo-sition of D-T. With some results obtained in typical experiments we discuss the different possibilities of imagery in the X-ray domain : . Imagery by emission of the target . Radiography of the target with an external X-Ray source obtained by a laser beam (backliahting) . A device is described who combine these two possibilities used simultaneously in the same experiment. . Imagery by self-emission or backlighting with time resolution. Some results are given which are in agreement with numerical modelization.

A two-frame flash x-radiography system has been developed for studying the implosion hydrodynamics of particle beam driven inertial confinement fusion targets. This system consists of two micro-point flash x-ray (FXR) sources, two micro-channel plate (MCP) x-ray sensitive cameras, and a space frame. The sources and cameras are arranged in pairs on the space frame to secure rigid registration of the sources with the cameras and to provide ease of alignment with the target. The micro-point FXR sources have an output spectrum characteristic of a tungsten bremsstrahlung source with an end point energy of 600 key, a source intensity of 8 mRad at 46 cm, and a spot size of approximately 100 μm. A direct 4x magnification of a 3 mm gold target was achieved by using this source in a backlighting arrangement with an overall system resolution of 4 1p/mm. This system has been used successfully to observe relativistic electron beam driven ablative pusher target implosions. In the inertial confinement approach to controlled thermonuclear fusion, ignition is to be achieved via heating, compression, and confinement of the fuel by an imploding shell. Since the confinement time improves with increasing pusher pR, target design calculationsl achieving gain > 1 require a pusher pR > 1 g/cm2, where p and R are the mass density and shell thickness of the pusher, respectively. For such targets, the pusher can, during its acceleration by the driver and during subsequent deceleration by the compressed fuel, be unstable which can lead to reduced compression, fuel heating, and confinement time. An experimental study of this problem is very difficult because of the pusher thickness and requires a technique such as flash x-radiography (FXR) to directly observe the hydrodynamic behavior of the interfaces between the target ablator and the pusher and between the pusher and the fuel. To make these studies, a two-frame FXR system has been developed. This system consists of two micro-point FXR sources (each having a spot size of 100 pm), two gated microchannel plate x-ray sensitive cameras, and a space frame which holds the FXR source and camera pairs in rigid alignment. A schematic representation of this system is shown in Figure 1. Two 3 ns FXR radiographs of an imploding target can be made with this system with arbitrary interframe separation and with a system spatial resolution of 125 μm.*

High-density, short-duration x-ray pulses are necessary to diagnose the compression of laser fusion targets. Present target designs are such that backlighting sources ranging from a few thousand electron volts to 100 keV will be necessary. The desired source durations range from a few tens of picoseconds for flash radiography to several nanoseconds for streaked backlighting, and the source occurence must be tightly synchronized to that of the target-irradiating laser pulse. For the latter reason, a laser-induced x-ray pulse is preferred. An initial study of the K lines of Ti, Ni, and Zn as possible backlighting sources was conducted. The conversion efficiency of laser light into line radiation was obtained as a function of laser intensity, pulse length, and wavelength. A threshold laser intensity for x-ray line production was identified. Information was obtained on the size and duration of the x-ray emission source, in relation to laser parameters. The experimental results, and their impact on backlighting capability for high-density laser fusion targets, are discussed.*

When a condensor is discharged between two electrodes in a thin Layer of liquid, a light flash is produced. The light emission is detected with a fast silicon detector filtered for red, green and blue light. The spark discharge and the shock waves in the liquid layer are photographed with X-ray flashes. The energy transfer to the gas plasma is much higher for a second or third discharge which follows the first discharge after some microseconds, if it is a discharge into the spark induced gas plasma and not in the liquid. The expansion of the gas plasma has been studied therefore in detail. A simple arrangement of two condensors and a coil makes it possible to produce flash sequences. The spark was produced in different organic liquids and H2O.

A high-speed 16-mm cineradiographic system previously developed at the University of Michigan, Highway Safety Research Institute, for use in biosciences research has been upgraded in capability by 1) increasing the x-ray field to 35-cm (14-in) by 43-cm (17-in), 2) incorporating a 35-mm high-speed motion picture to increase resolution, and 3) acauisition of a second x-ray head for three-dimensional x-ray stereophotogrammetric studies. This system now consists of a 35-mm Photosonics 4B camera, capable of 2500 frames per second, which views a 50-mm (2-in) diameter output phosphor of a highgain, 4-stage, magnetically focused image intensifier tube, gated on and off synchronously with the motion picture camera. A lens optically couples the input photocathode of the image tube to x-ray fluorescent (rare earth) anteelnimages produced by a smoothed d-c x-ray generator of a conventional type. The system is capable of a wide range of magnification ratios.

As a computer technique, the CT scan, recently induced great changes in both medical and radiological practice, photoelectronic detection of X-rays and digital image processing just begin to improve the angio-cardiography. The first generation of intravenous angiography systems is presently available. They allow to visualize the chambers of the heart as well as the great arteries using a fast intravenous injection of contrast medium. In the next future, this technique will probably be suitable for visualizing the entire vascular network, including small arteries and veins, and others cavities such as urinary and digestive tracts. In the long term, a large increase in computer assisted diagnosic techniques is likely to occur, as well as the suppression of the film support.

In the development of devices to represent the three-dimensional structure of radio-activated organs, Nuclear Medicine has been following the progress in Radiology. That parallelism could be observed with all principles used to get three-dimensional data : - analogical systems working by simultaneous displacement of the detector and the object, - coded-aperture imaging devices which consist of special collimators designed to obtain a dependance between the object-to-code distance and the detector response, - multiple-incidence techniques, the 3D reconstruction being extracted from the whole set of projections of the object at different orientations. That last principle was chosen in the studies that we are working on now. It is close to the principle used in Radio-Tomo-densitometry : a detector gets a set of projections as it turns around the object. From these projections, the classical reconstruction algorithms can be used : ART, SIRT, Convolution algorithms... But we have to take into account here some more specific properties : the statistic noise, the self attenuation of the radiation, the distance-dependant resolution. In this paper some correction process will be considered which can be more or less easily implemented depending of the algorithm used. Different compromises can be proposed : they depend strongly not only of the algorithm and the data-processing but also of the detector performances. In this field, this method will greatly take profit of the powerful calculators designed for the Radiographic tomodensitometry.

The discrete formulation of a coded-aperture image is recalled ; the principle of the numerical convolution is given and then we calculate a maximum value for the quadratic error. We show the advantages offered by a post-deconvolution filtering as well as by an iterative restoration of the slices. Finally we present the results we obtained by recons-tructing an object consisting of three slices which is a coarse model of the human heart.

In 1975, nuclear scattering radiography was introduced as a new method giving three dimensional information 1. In a first part we shown the main characteristics and advantages by comparison with the conventional radiography. Various ways of using high energy proton beams for radiography have been considered and tried in the last decade. They relied mostly on the following mechanisms of beam attenuation : absorption by nuclear interaction ; Coulomb scattering. Computerized tomography can be considered with the proton beams treated like X-rays beams : like with X-rays projection, measurements under a considerable number of angles are necessary to obtain tomographic information.

The two following cinematographic methods can be considered : 1. generation of an X-ray pulse train with a single X-ray tube. The deionization time between two discharges limits the upper value of the frame repetition rate ; 2. cineradiography with multiple tubes. The time interval has no lower limit, but errors can result from the parallax and this method is not suitable for the visualization of phenomena lacking axial symetry. We present a novel tube including four flash X-ray sources equally spaced by 20 mm only so that parallax can be neglected in most cases. The electrodes consist of four cathodes and an single anode on which deflectors are mounted in order to prevent the vapors produced in a discharge space from attaining the other discharge spaces. The shortest time interval separating two successive discharges is only limited by the X-ray pulse duration, i.e. 25ns. The novel flash X-ray tube is used in connection with four 500 kV, 60 J, Marx-surge generators. A dose of 1.94 μC/kg per pulse at 1 meter from the anode is reached. The frame recording method uses a fluorescent screen converting X-rays in light photons and a high speed electronic camera ("Imacon 790"). This new flash X-ray cinematographic device has been used for recording phenomena in terminal ballistics. Examples are given of frames showing the penetration of projectiles into a target.

Heating of thermoluminescent materials by pulsed or continuous laser beams is shown to be an attractive means for rapid reading of discrete or continuous two dimensional arrays; applications of this novel laser heating techniques to thermoluminescence imaging of spatial intensity of ionizing radiation with perspectives, in X-ray or y-ray imaaery is considered. The thermoluminescence (TL) phenomenon is based on the presence of electron and/or hole traps which, upon exposure of the TL material to ionizing radiation, become occupied with the respective charge carriers. Thermal energy is provided -- usually in form of a linear heating program -- to release the trapped carriers. During this process recombination takes place and photons may be emitted until the thermal equilibrium distribution of all carriers is reestablished 1. The number of luminous photons emitted during the heating period is usually proportional to the ionizion irradiation of the material. Our experiments have been done with thin layer dosimeters made by powder deposition of CaF2, CaSO4... on glass sheet (22 mm x 22 mm) with a thickness ranae of 5 to 100 μm. The irradiation was achieved with a Cs 137-γ -source (.66 MeV). The typical irradiation was from 10 mR to 10 R. A 10 watts CO2 laser has enabled to heat the layer by small point (.5mm diameter). The interest of such a pinpoint heating is obvious ; it enables a reading of the absorbed irradiation at this point or pixel in a relatively short time (10 ms < t < 103 ms) and in the future to consider the prospects in dynamics X-Ray image detection. Though our study has been left at a prospective and fundamental step its extrapolation a X-ray imagery can be considered ; that point will be approached with some details.

The use of two computer-based, multi-dimensional display methods to efficiently convey three-dimensional (3-D) anatomic information contained within computed tomographic (CT) volume images is described. The term volume image refers to a stack of discrete 2-D CT images and is therefore a 3-D array of volume picture elements (voxels). The first display method is referred to as a direct display approach because it directly represents the 3-D volume image array as a 3-D distribution of brightnesses, i.e., each voxel is a brightened point in space. The second display method is referred to as a transformed display approach meaning that before display, the volume image data is transformed resulting in a new image data set which then forms the basis of the display process. An example of transformed display is shaded surface display, where the transformation is the detection and subsequent representation of the "surface" as a 3-D distribution of reflectors. These display approaches facilitate the direct visualization of 3-D shapes and spatial relationships thereby obviating the necessity of mentally reconstructing shapes from the series of cross sections.*

The investigation of phenomena occuring into micro-plasmas created by laser is mainly based on images obtained from emitted X-rays or α,-particles. To get images, the pinhole camera is widely used but the collection efficiency of radiation or particles, as also the Signal-to-Noise ratio (S/N ratio) decrease as the spatial resolution is increased. To improve the S/N ratio one can use a coding-aperture imaging device, the collecting surface of which is much higher that the pinhole-one. In this study we have chosen a simple coding-aperture made of a circular slit cut in a thin foil which is opaque at the considered radiation. The slit width is δ = 6 μm ; the mean radius is R = 1 mm. The image reconstruction is obtained by numerical processing (Fourier-transform and filtering). The transverse spatial resolution 6t in the object is δ t = δ but can be improved by a factor of 2 or 3 if the edges of the circular slit are sharp. The axial spatial resolution (tomographic effect) δl is δl, = δt l1/R, with ll the code-to-object distance.

Scintillation Techniques used up to now in Nuclear Medicine could only provide, at the detector plane, some projected images of the distribution of the radionuclide concentration in the examined organ or tissue. Now, the use of coded-aperture principle, instead of the conventional collimators in the scintillation cameras allows to get true three-dimensional information and consequently to have some maps of the distribution of the radionuclide along slices of the organ or tissue. The process is realized in two steps. In a first step, an encoding lead mask is placed between the radioactivated object and the detector surface on which the coded-imacTe is formed. From this coded image one can reconstruct, in the second step, a certain number of slices chosen into the object ; these reconstructed slices are obtained by an analogical or digital decoding process of the coded image. In fact, the information corresponding to a given slice, at a given depth determined by its distance to the encoding mask, is obtained by using only some numerical methods ; never-theless these methods are rather complicated and necessitate a large number of operation and calculus. After a quick survey of previous papers on the coded-aperture work in Nuclear Medicine, we present the first results we have obtained with a code originated from the Fresnel zoned array - the. Fresnel zoned linear array - the numerical simulation of the diffraction phenomena being used as the reconstruction method. In the case of the circular Fresnel zoned array the numerical reconstruction necessitates a two-dimensional Fourier-transform. On the contrary the Fresnel linear array necessitates a Fourier-transform in one dimension only ; so it becomes possible to use this method with the processing equipment frequently available in the Nuclear Medicine facilities. The experimental results obtained at the CHR - Toulouse have allowed to verify the focusing-properties of this code (tomographic effect) and also to show that the values obtained by numerical reconstruction agree with those given by the theory of Optics.

We study the 3D reconstruction of simple objects from transmitted radiographs on the basis of computed tomography. We consider a limited number of projections (about ten) and give a comparison for different algorithms (convolution and algebraic reconstruction technique -ART-). Results are presented for computed and experimental well known objects.

Radiographies of moving objects are affected of typical artefacts. The problems encountered in interpreting such pictures are described. Numerical and computing tools provided in order to solve these problems are presented.

In the field of X-Ray Radiography, different three-dimensional reconstruction techniques have been proposed : apart from computer-aided tomodensitometric methods one can reconstruct an object by using the different slices which are obtained by the now classic scanner. Nevertheless, it is also possible to consider that problem more directly, by using a set of planar projections. In the parallel-beam geometry a set of projections obtained by the rotation of the beam in a plane is sufficient toireconstruct the 3 D object. In this case one can generalize the filtering and spreading process, usually carried out on the monodimensional projections, at the case of planar images corresponding to 2 D projections. However, this method meets technological difficulties as it is difficult to have a parallel-beam with a large cross-section. This fact has led us to consider the use of a conical beam. The adaptation of the previous results at this conical geometry which requires a conical spreading and a specific filtering, also makes necessary to get projections which have to be uniformly distributed in all orientations of the space. This causes further technological and practical difficulties due to the generally limited possibilities in the positioning and the displacement of the X-Ray sources. We are thus inevitably lent to use the measurements originating in a system based on the cone-beam geometry whose axis rotates only in one plane. Thus, this axial symmetry will cause a certain number of problems which are in fact, the consequence of,the limited knowledge of the information. The following, then, are the possible solutions : 1) To design on to use the acquisition system in such geometrical conditions that it is possible to disregard the effect of beam divergence. The results will then be identical to those used in parallel geometry 3 D reconstruction. 2) To use, in this particular axial symmetry, the same method of reconstruction as that considered when we have the whole set of conical projections of all directions in the space. In this solution, some errors appear owing to the lack of a large number of directions. 3) To develop algebraic methods of reconstruction adapted to the conical geometry, using some criteria of optimization. The latter solution seems to be the best approach but necessitates strongly in practice, the use of specific computers (like "array processors") in order to reduce the presently too long time taken by the computer process.

The computerized-aided tomodensitometry by X-rays is a powerful, precise and non invasive tool to investigate the human body as well as different objects and structures. The relatively long examination time, from 2 to 10 sec., which the present state of the art necessitates, limits the field of applications of this tool here to the static on quasistatic organs or objects (e.g. the brain or abdomen in medical diagnosis). When one try to reconstruct images of a dynamic object some errors or artefacts appear. In the case of the heart, in particular, the examination time takes several cardiac periods; so, the reconstructed image is highly blurred and for all purposes unusable. One solution, which can be used with a conventional scanner is the "gating" technique. Taking into account of the repetitive property of the cardiac movement, this technique necessitates to catch the projections in synchronization with the electrocardiogram. Thus a sub-program, controlled by the ECG sianal, has to select a set of projections corresponding to a certain time of the cardiac period. The implementation of this sub-program into the reconstruction code allows to obtain more sharp images of the heart at any given instant of the cardiac period. In this paper we discuss the methodology used to acquire the projections, the problems caused by the small number of projections, some mathematical solutions (extensions of signals and data, interpolations, uses of the redundancy in the information aiven by a FAN-BEAM system). These different possibilities are illustrated by some results we have obtained in our experiments on animals with a FAN-BEAM machine.

The laboratory of Image Processing, a part of the Service M.E. at VAUJOURS, C.E.A., was founded in 1975. Its main goal was to contribute to read and process radiographic films. In 1976, this laboratory has been already presented at a previous ANRT Congress in PARIS* Since this time a large number of improvements and extensions have been made ; so it seem useful to describe, in a first part, the present state of the laboratory. In a second part we present some examples of image processing in the field of Non Destructive Testing.

GREC, an acronym for "Generateur de Radiographie EClair" was designed for visualisation, measurements and studies on dynamic behavior of materials under high explosive loading. GREC was built in 1974 by Physics International for the CEA. Since the first operating time, 1975, this facility has been improving, taking advantage of progress in the techniques involved, especialy focusing and transport of intense electron beams. In this paper, we describe the X-ray machine and give some details on the image detection device. The pulse generator, a 80 stages Marx Generator can deliver a peak voltage of 9 MV in 1.2/us by series erection of 2 x 80 capacitors charged at + 60 kV. The Hulse-forming network, a tri-axial Blumlein, delivers a 80 ns-pulse via a 26 ohms impedance transmission line to the transport section. This transport section is a 24 cm diameter, 2 meters long, magnetic-insulated coaxial line. At the end, the X-ray radiation is emitted by bremstrahlung of the electrons impinging the tantalum anode of the dis-charge diode. The main specifications are : - electrical : 7 -MV - 180 kA - 30 ns - X-ray : Diode 1 : 400 R at 1 meter ; spot size : 20 mm Diode 2 : 100 - 300 R at 1 meter ; spot size : 4-8 mm The image acouisition device has been designed for X-raying of objects under loading of un to 50 kg of high explosives. In the paper, we give some details on the exnerimental arrangement used. Despite the severe environmental shock constraints, reasonable measurements of line-integrated density are possible.

The PHERMEX facility used to provide flash radiographs of explosives and explosive-driven metal systems is described. With this facility, precision radiographs of large objects containing materials with high atomic number and high density are attainable. PHERMEX encom-pass the high-current, three-cavity, 30-MeV linear electron accelerator; the 50-MHz radio-frequency power source to drive the cavities; timing, and signal detection system; and a data-acquisition system. Some unique features of PHERMEX are reliability; very intense sub-microsecond bremsstrahlung source rich in 4- to 8-MeV x rays; less than 1.0-mm-diam spot size; precision determination of edges, discontinuities, and areal-mass distribution; and flash radiographs of large explosive systems close to the x-ray target. Some aspects of the PHERMEX -upgrading program are discussed. The program will result(1) in an increased electron-beam energy to about 50 MeV, (2) the use of an electron-gun pulser that is capable of producing three-time-adjustable pulses for obtaining three "radiographic" pictures of a single explosive event, (3) an increased electron injection energy of 1.25 MeV, (4) the capability for recording high-speed signals, and (5) the use of computers to assist the monitoring and control of the data-acquisition system and the PHERMEX accelerator.*

The Artemis Project, a facility for high speed cineradiography at high energy will allow spatial and temporal analysis of physical phenomena in streak or framing operation. The Artemis facility will include two parts : a linear electron accelerator (45 to 55 'mev) and an opto-electronic recording system formed by 1 to 3 cameras depending of the observation mode (streak or framing). The linear electron accelerator built by CGR-MeV, will have to deliver a radiation dose at 1 meter higher than - 10 R/ us during a 15/us pulse, - 20 R for each of three 50 ns short pulses in a 15/us period. In the first part, the main desired characteristics of the X-ray source for our radio-graphic purposes are given. The chosen source, a L-Band linear electron accelerator is described in the second part. The projected image recording system is presented in the third part.

MIRENE is a MIni nuclear REactor for NEutronography. In the first part of this paper MIRENE is described and its characteristics are given. The core uses only 1 kg of enriched uranium in solution state. It works in a self-limited pulse mode. The neutron pulses are collimated in two beams which cross the concrete pro-tection walls surrounding the reactor. The main characteristics are : . peak power : 161 kW . exponential rise time : .87 sec . overall energy in a pulse : 2.9 MJ (6.8 x 1016 fissions) . axial beam : - exposure aera : 30 cm x 30 cm - useful fluence per pulse : thermal neutrons : 9 x 108 n/cm2 ; γ-rays : 22 rads (Cd ratio on gold detector : 2) . lateral beam : - exposure aera : 18 cm x 24 cm - useful fluence per pulse : thermal neutrons : 2.6 x 108 n/cm2 ; γ-rays : . 7 rad (Cd ratio : 9). In the second part of the paper, many applications of MIRENE in much different fields are indicated. The results we have obtained since MIRENE started to operate, in 1977, are shown : - In nuclear engineering :.testing of first neutron reactor fuel-pins .control of "neutrophage screens" used in transport and storage of nuclear-fuel materials to secure the criticity-safety .observation of irradiated-oxyde samples in order to determine the Equation of State of the fuel used in fast-neutron reactors .observation of UO2-H20 mixing conditions in the field of cri-ticity experiments - In engineering, MIRENE has a large field of applications, two examples are given : . the control of the sealing of an electric isolator . the visualization of the bonding layer between two high density metals - Finally we show an original application in agronomy which has given very good results : the observation of the in-situ-growth of a corn-root. All these results prove that MIRENE as well as similar reactors can bring about an important contribution as Non-Destructive-Testing stools in the most large field of applications. Their simplicity of design and working connected to their intrinsic safety enabled them to have a high competitive performance/cost ratio. Users can take advantage of the own performances of nuclear reactors in neutronography while being relieved of a large part of the constraints associated to large-size ordinary reactor installation.

The design and performance characteristics of a set of high intensity flash X-ray systems will be discussed. At present, six (6) systems are available with output voltages in the range of 75 kV to 1200 kV. Peak current is 10 kA and exposure time 20 ns. Demountable X-ray tubes are being used to increase flexibility and reduce operating costs. Typical applications are found in areas such as ballistics, detonics, quality control, etc.

The image recording of fast neutron radiographs can be achieved sufficiently by means of plastic foils and the track etching technique. The optimal etching conditions for cellulose nitrate foils were studied. A mathematical model for the dependence of the optical density on the area and number of the etch pits is given and compared to measurements. Finally, two examples of radiographic applications may underline the possible importance of this method.

In order to detect fast neutrons, the Kodak CA 80-15 plastic detector can be used either as a screen in neutron radiography or as a neutron dosimeter. Especially for 14.9 MeV neutron dosimetry, we describe a helium converter test unit which increases the detection efficiency of cellulose nitrate plastic detectors by a factor of 3.

This camera was developed specifically for high speed cine radiography and was designed to be used both in laboratory and field trials. The camera is programmable by digital controls and can drive flash x-ray units synchrously with framing up to a rate of 250, 000 fps. An image format of nine frames 24x30 mm and a gain of 10,000 is incorporated to give ampli-fication of low light level short exposure x-ray flashes.

We describe the basic structure and the making process of radiation-sensitive image-converter-tubes ; these tubes are self-sensitive for 10 kev to 10 Mev photons or cold and thermal neutrons. In most general use, these tubes have an affective field aperture of 22 or 30 cm but it is also possible to make smaller (15cm) or larger (40cm) tubes. Data and application characteristics of these tubes are given. The conversion factor (the output image luminance corresponding to a given input radiation) is high enough to allow the real-time acquisition of the image, like with a T.V. camera or a high speed camera for example. Typical conversion factor are 100 cd.m-2.mR-1.s for X-rays and 3.10-10 Cd.s.n-1 for neutrons. The spatial resolution is 2.5 to 5 1p.mm-1 depending of the kind of the tube ; density steps of 2 % are detected. Applications of these tubes are illustrated by examples in medicine (cine-angiography and televised-surgery) as well as in ballistics.