Thick-target Bremsstrahlung Research Articles

Runaway electron studies in the ATF torsatron

Runaway electron formation and confinement occur readily in pulsed torsatrons and heliotrons because of the high loop voltages during initiation and termination of the helical and vertical fields (‘‘field ramping’’) and the inherently good containment of the electrons on the flux surfaces in the vacuum fields. This has been confirmed for the Advanced Toroidal Facility (AFT) [Fusion Technol. 10, 179 (1986)] and other stellarators by orbit calculations. Since runaway electrons can cause an unacceptable level of hard x rays near the machine, a runaway electron suppression system was incorporated in ATF. The main component of the system is a movable paddle, which is normally left in the center of the plasma chamber during the field ramps. This device, in conjunction with programmed vertical field ramping, which reduces the volume of good flux surfaces, has proved to be very effective in reducing the runaway electron population. Measurements of hard x rays from ATF have shown that the runaway electrons are produced primarily during the field ramping but that there is usually also a small steady-state runaway electron component during the‘‘flat-top’’ portion of the fields. The paddle is the main source of the hard x radiation (thick-target bremsstrahlung). There is evidence that some of the runaway electrons may be confined to islands. The maximum x-ray energy found by pulse height analysis is ∼12–15 MeV. The mean energy appears to be a few million electron volts. There is a noticeable forward peaking of the bremsstrahlung from the paddle. The limiters do not appear to be major sources of x rays.

Atomic-field bremsstrahlung at zero degrees

Bremsstrahlung spectra have been measured for elements with atomic numbers ranging from 29 to 92 at 0°, 30°, 45° and 90° photon emission angle with a 70-keV electron beam. In an earlier experiment, the electron beam had been deflected away from the detector by a magnet; however, the dispersion of the electron beam in the target created a background that could not be subtracted out easily. Hence, in this work, we introduce an alternate approach to magnetic deflection. Here, the incident or scattered beam is stopped completely by adding a thick carbon backing to the thin-film target. The thick-target bremsstrahlung from the low- Z backing, which decreases linearly to zero at the end point, can be subtracted out, and reliable measurements of the cross section can be made over a wide range of radiated photon energy. Typical results will be presented and compared with theory. Theory is found to reproduce well the shape of the radiated photon energy spectrum at 30°, 45° and 90°, but there is a significant discrepancy between theory and experiment for 0°.

Bremsstrahlung review: An analysis of the Schiff spectrum

A concise approximate formula computed by Schiff for the intensity spectrum of bremsstrahlung photons has been a valuable starting point for many medical physics applications, including the Task Group 21 protocols. This paper provides a brief review of the literature related to determination of the bremsstrahlung spectrum and to the Schiff formula in particular. It describes the approximations Schiff made to obtain this formula, including the Born approximation, and the exponential nuclear screening potential, the infinite-mass nucleus approximation, and the "extreme relativistic" approximation. A derivation of a more exact formula that avoids the last of these approximations is presented. This provides a check on the accuracy of the Schiff spectrum for linear accelerator energies used clinically. Comparisons with the Schiff thin-target result are presented. A thick-target bremsstrahlung spectrum is calculated and compared with the forward spectrum obtained from Monte Carlo simulations of the x-ray production in two linear accelerator treatment heads.

Measurements and simulations of low energy, thick target bremsstrahlung spectra

Abstract Nuclear resonance fluorescence (NRF) experiments on the isotopes 13 C, 23 Na, 27 Al, 31 P, 35 Cl, and 48 Ti with transitions of known strengths have been performed to determine the relative spectral shape of low energy, thick target bremsstrahlung spectra. The experimental results (endpoint energy E 0 = 4.1 MeV) are compared with Monte Carlo simulations. It is concluded that a linear function represents a rather precise approximation of the shape of this bremsstrahlung spectrum near the endpoint energy. Therefore, a linear interpolation of the photon flux between well known data points enables a fast and sufficiently precise analysis of NRF experiments using photon sources of this kind.

Double atomic-field bremsstrahlung from thick targets

Measurements of the two-photon coincidence rate from 84 keV electrons from a 109Cd source on thick targets of Ag, Au, and Pb are compared with a thick-target double atomic-field bremsstrahlung (TBD) model. Each photon energy range was 32–52 keV. The target thickness ranged from 3 to 30 mg/cm 2. Results for the thick-target single-bremsstrahlung (TSB) energy spectrum for 64 keV electrons in coincidence with the 22 keV X-ray from the source will also be compared with the TSB theory.

Alternative interpretation of the double atomic-field bremsstrahlung experiment.

A recent experiment reported a two-order-of-magnitude enhancement of the cross section for double atomic-field bremsstrahlung when compared to various theoretical formulations. In this paper we propose to explain the observed coincidence rate in terms of the coherent emission of two thick-target bremsstrahlung photons from the Mylar vacuum window by electrons elastically scattered in the target. The photon energy dependence of the coincidence rate predicted by this thick-target double bremsstrahlung (TTDB) model is estimated and compared to the previous experimental results for a gold target. Fairly good agreement is obtained with both the order of magnitude of the coincidence rate and its photon energy dependence, even though a crude approximation is used to estimate the TTDB distribution from the better known thick-target single bremsstrahlung spectrum.

High-energy photon spectra from a coaxial gas-puff experiment

An array of plastic scintillator photomultiplier detectors was used to determine the high-energy (hν>35 keV) photon spectra of a 130-kJ, 60-kV gas-injected coaxial gun experiment. The detector array used six different filter material/thickness combinations. High-energy photon signals were readily observed. The spectra were determined by deconvolution. The spectra have three characteristic components—radiative emission from high Z impurities in the plasma, beam-target excited line emission, and thick-target bremsstrahlung from an electron beam generated in the plasma column. The electron beam consistent with the thick-target bremsstrahlung assumption was of the order of 100 A above 10 keV.

Measurement of yield and cross section for the 27Al (γ, 2p) 25Na reaction with a clinical linear accelerator

A clinical electron linear accelerator was adapted for continuous energy selection and used to measure bremsstrahlung yield for the photonuclear reaction 27Al (γ, 2p) 25Na from 24.0 to 32.0-MeV in 0.25 MeV steps. A germanium detector was used to measure the number of photoproduct γ-rays emitted. The yield values were analyzed for photonuclear cross section using PHONUC II, a program especially adapted for use with yield data from thick-target bremsstrahlung. The cross section shows distinct resonance peaks at 26.3±0.3, 27.3±0.4 and 30.3±0.5 MeV. Possible theoretical origins of these resonances and applications of the reaction in accelerator energy calibration are discussed.

Time-integrated x-ray measurements of the very energetic electron end loss profile in TMX-U

The time-integrated 2D profile of the thick-target bremsstrahlung produced by energetic end loss electrons has been measured during electron cyclotron resonance heating (ECRH) operation of TMX-U. Sheets of x-ray film and/or arrays of thermoluminescent dosimeters were placed on the outside of the end tank end wall to measure the relative spatial x-ray profile, with locally added filters of Pb to determine the effective mean x-ray energy. The purpose of this simple survey diagnostic was to allow deduction of the gross features of the ECRH region. The electron source functions needed to fit the x-ray data were modeled for various anchor cell radial distributions mapped along magnetic field lines to the elliptical plasma potential control plates or the Al end walls. The data are generally consistent with (1) major ECR heating in the central 25-cm-diameter core, (2) a mean ECRH electron-loss energy of 420 keV, and (3) and ECRH coupling efficiency to these hot electrons of ≥10%.

Acceleration of Runaway Electrons and Joule Heating in Solar Flares

D.C. electric fields provide the simplest and most direct means of accelerating electrons out of a thermal plasma. Most solar flare models result in the production of D.C. electric fields. On the other hand, microwave and hard X–ray observations of flares provide specific requirements for the number and energy of energetic electrons produced during a flare, and the timescales involved in accelerating them. The microwave emission from flares is understood to be gyrosynchrotron radiation from electrons with energies of 100 keV or greater. The hard X–ray emission (≳ 25 keV) can be interpreted as being either thick–target bremsstrahlung from non–thermal electrons (thin–target radiation may also contribute to the X–ray emission, but the process is less efficient), or thermal bremsstrahlung from hot, impulsively heated plasma. Hence, it is of interest to study the electric field acceleration of “runaway” electrons and the simultaneous Joule heating of the thermal plasma in light of these results from flare observations, without recourse to a specific flare model. Some of the results of such a study are summarized here.

Multiple energetic injections in a strong spike-like solar burst

An intense and fast spike-like solar burst was observed with high sensitivity in microwaves and hard X-rays, on December 18,1980, at 19h21m20s UT. It is shown that the burst was built up of short time scale structures superimposed on an underlying gradual emission, the time evolution of which showed remarkable proportionality between hard X-ray and microwave fluxes. The finer time structures were best defined at mm-microwaves. At the peak of the event the finer structures repeat every 30–60 ms (displaying an equivalent repetition rate of 16–20 s-1). The more slowly varying component with a time scale of about 1 s was identified in microwaves and hard X-rays throughout the burst duration. Similarly to what has been found for mm-microwave burst emission, we suggest that X-ray fluxes might also be proportional to the repetition rate of basic units of energy injection (quasi-quantized). We estimate that one such injection produces a pulse of hard X-ray photons with about 4 × 1021 erg, for ɛ ⪞ 25 keV. We use this figure to estimate the relevant parameters of one primary energy release site both in the case where hard X-rays are produced primarily by thick-target bremsstrahlung, and when they are purely thermal, and also discuss the relation of this figure to global energy considerations. We find, in particular, that a thick-target interpretation only becomes possible if individual pulses have durations larger than 0.2 s.

Thick-target bremsstrahlung interpretation of short time-scale solar hard X-ray features

Steady-state analyses of bremsstrahlung hard X-ray production in solar flares are appropriate only if the lifetime of the high energy electrons in the X-ray source is much shorter than the duration of the observed X-ray burst. For a thick-target nonthermal model, this implies that a full time-dependent analysis is required when the duration of the burst is comparable to the collisional lifetime of the injected electrons, in turn set by the lengths and densities of the flaring region. In this paper we present the results of such a time-dependent analysis, and we point out that the intrinsic temporal signature of the thick-target production mechanism, caused by the finite travel time of the electrons through the target, may indeed rule out such a mechanism for extremely short duration hard X-ray events.

Electron transverse velocity measurements in an intense relativistic electron beam diode

Techniques for making electron transverse velocity measurements on an intense electron beam using thick-target bremsstrahlung emission are presented. The procedure presented here differs from similar previous measurements using thick-target bremsstrahlung in that it allows both time and spatial resolution. Data from the study of a 2.1-MeV, 40-kA, 140-ns-long beam are presented. Measurements on this beam had a transverse velocity resolution of about 0.1 c. The temporal resolution was about 5 ns and the spatial resolution was 6 mm. Techniques used to determine the time-resolved radial current density profile are also presented.

Dynamic spectral characteristics of thermal models for solar hard X-ray bursts

The dynamic spectral characteristics of the thermal model for solar hard X-ray bursts recently proposed by Brown et al. (1979) (BMS) are investigated. It is pointed out that this model, in which a single source is heated impulsively and cooled by anomalous conduction across an ion-acoustic turbulent thermal front, predicts that the total source emission measure should rise as the temperature falls. This prediction, which is common to all conductively cooled single-source models, is contrary to observations of many simple spike bursts. It is proposed, therefore, that the hard X-ray source may consist of a distribution of many small impulsively-heated kernels, each cooled by anomalous conduction, with lifetimes shorter than current burst data temporal resolution. In this case the dynamic spectra of bursts are governed by the dynamic evolution of the kernel production process, such as magnetic-field dissipation in the tearing mode. An integral equation is formulated, the solution of which yields information on this kernel production process, from dynamic burst spectra, for any kernel model. With a BMS-type kernel model in one-dimensional form, the derived instantaneous spectra are limited in hardness to spectral indices γ ≳ 4 for any kernel production process, due to the nature of the conductive cooling. Ion-acoustic conductive cooling in three dimensions, however, increases the limiting spectral hardness to γ ≳ 3. Other forms of anomalous conduction yield similar results but could permit bursts as hard as γ ≳ 2, consistent with the hardest observed. The contribution to the X-ray spectrum from the escaping tail of high-energy kernel electrons in the BMS model is calculated in various limits. If this tail dissipates purely collisionally, for example, its thick-target bremsstrahlung can significantly modify the kernel spectrum at the high-energy end. The energetics of this dynamic dissipation model for thermal hard X-ray bursts also are briefly discussed.

Shielding study of bremsstrahlung in bulk media with electrons part I: Monte Carlo calculation of thick-target bremsstrahlung spectrum

The electron-photon cascade Monte Carlo code, TAURUS, was developed for the calculation of gamma-ray and bremsstrahlung penetration in the matter of cylindrical and plane geometry bombarded by gamma-rays and electrons. This code treats the photon simulation with enough accuracy but the electron simulation with some simplification, because of speeding up the computing time and saving the core memory. The accuracy of the code was checked from the comparison with other experimental and calculated results of thick-target bremsstrahlung and was found to be good enough with small core memory and fast computing time.

Optimization of Bremsstrahlung Energy Deposition

Thick-target bremsstrahlung energy deposition in a thin, gold calorimeter has been measured as a function of converter foil thickness for Mo, Sn, La, Ce, Pr, Nd, and Ta converters at source electron energies of 0.5, 0.75, and 1.0 MeV. The thickness-maximized deposition is not a monotonically increasing function of converter atomic number, but instead exhibits a broad maximum near the type 4f rare-earth region. This maximum is substantially greater than the deposition from the Ta converter--Ta being a material routinely employed in flash x-ray sources. Theoretical models also predict higher deposition from La converters than from Ta converters and are used to show that this enhancement is a consequence of higher characteristic x-ray production in La. Nevertheless, there remain significant discrepancies between theoretical predictions and measured data, even when improved numerical cross sections which have just become available are employed in the calculations.

Radiation-Induced Coupling to a Truncated Cylinder within a Cylinder

Results from an investigation involving the experimentally measured and calculated electromagnetic response of two mathematically tractable geometries exposed to 12 MEV thick-target bremsstrahlung are presented. One geometry involving parallel plates, which can be adequately described by a one-dimensional model, was used to study the interaction of a partially ionized gas with a negatively charged conducting surface. The results indicate the existence of a boundary region, consisting of an electron depletion layer, that can effect the electrical response of systems in an ionized gas environment. A more complicated geometry involving concentric cylinders was used to check the adequacy of our calculational procedure which utilized a finite differencing solution of Maxwell's equations. The experimental results indicate that the electrical response of the more complicated geometry can be approximately described by our calculational procedure; but, the requirement for refinements (such as the inclusion of the boundary layer effects and adequate Monte Carlo transport source descriptions) may be warranted.

Simultaneous investigation of parent electrons and Bremsstrahlung X rays by rocket-borne detectors

Simultaneous measurements of the energy spectrum of precipitating electrons and the resulting bremsstrahlung X-ray spectrum were carried out during an auroral event on March 3, 1971, at the Churchill Research Range, Manitoba, Canada. The electron data were obtained with detectors on a Black Brant VB sounding rocket (275-km apogee), while the X-ray flux was measured by an instrument package that was boosted to 60 km on an Arcas rocket. The X-ray package was deployed on a parachute at apogee to provide a slow descent through the atmosphere. Thick-target bremsstrahlung theory is used to calculate the X-ray flux produced by the incident electrons, and a Monte Carlo method is used to predict the X-ray spectrum at various altitudes appropriate for comparison with the measured X-ray data. Satisfactory agreement between theory and experiment is obtained, and the value of the constant in the thick target theory has been estimated to be about 0.00002.

The effective cross section of the [formula omitted] reaction

The 165 Ho(γ, x n) reaction cross section has been measured with a 100 keV bin in the range from threshold up to 23 MeV. The analysis of the yield curve is based on the least structure solution using the thick-target forward-direction bremsstrahlung spectrum. Finally, taking into account the Nilsson level scheme, the total E1 photoabsorption cross section was computed and compared with the present data and other published experimental data.

A comparison of experimental and theoretical thick-target bremsstrahlung spectra from 1 to 2 MVp

A technique has been developed for measuring high energy bremsstrahlung spectra. The measurements were made using a 5 inch diameter by 5 inch long NaI(Tl) scintillation detector coupled to a pulse pile-up rejector/restorer. The pulse height spectra were recorded in a multichannel analyser. An IBM 370, Model 155, digital computer was used to convert the observed spectra to a photon spectra using a predetermined response matrix which was derived from monoenergetic isotope spectra. Photon spectra were obtained for the Van de Graaff generator and the resonant transformer in the range of 1-2 MVp for direct and scattered beams. The results were compared with calculated spectra using published bremsstrahlung cross-section formulas.