Recent Flash X-Ray Injector Modeling

Abstract

The injector of the Flash X-Ray (FXR) accelerator has a significantly larger than expected beam emittance. A computer modeling effort involving three different injector design codes was undertaken to characterize the FXR injector and determine the cause of the large emittance. There were some variations between the codes, but in general the simulations were consistent and pointed towards a much smaller normalized, rms emittance (36 cm-mr) than what was measured (193 cm-mr) at the exit of the injector using a pepperpot technique. The simulations also indicated that the present diode design was robust with respect to perturbations to the nominal design. Easily detected mechanical alignment/position errors and magnet errors did not lead to appreciable increase in the simulated emittance. The physics of electron emission was not modeled by any of the codes and could be the source of increased emittance. The nominal simulation assumed uniform Child-Langmuir Law emission from the velvet cathode and no shroud emission. Simulations that looked at extreme non-uniform cathode and shroud emission scenarios resulted in doubling of the emittance. An alternative approach was to question the pepperpot measurement. Simulations of the measurement showed that the pepperpot aperture foil could double the emittance with respect to the non-disturbed beam. This leads to a diplomatic explanation of the discrepancy between predicted and measured emittance where the fault is shared. The measured value is too high due to the effect of the diagnostic on the beam and the simulations are too low because of unaccounted cathode and/or shroud emission physics. Fortunately there is a relatively simple experiment that can resolve the emittance discrepancy. If the large measured emittance value is correct, the beam envelope is emittance dominated at modest values of focusing field and beam radius. Measurements of the beam envelope on an imaging foil at the exit of the injector would lead to an accurate value of the emittance. If the emittance was approximately half of the measured value, the beam envelope is slightly space charge dominated, but envelope measurements would set reasonable bounds on the emittance value. For an emittance much less than 100 cm-mr, the envelope measurements would be insensitive to emittance. The outcome of this envelope experiment determines if a redesigned diode is needed or if more sophisticated emittance measurements should be pursued.

FXR LIA Optimization - Time-resolved OTR Emittance Measurement

“…al. [24] indicate that the normalized emittance is 193 cm-mr at the FXR injector exit, which is both considerably larger than simulations predict (36 cm-mr) and also much larger than measured emittance at other similar facilities (75 cm-mr at DARHT-I, 47 cm-mr at ETA-II, and 110 cm-mr at ARIX). Measurements at the exit of the accelerator yield a normalized emittance of 385 cm-mr on average, indicating emittance growth through the linac and highlighting the need for further diagnosis and modeling.…”

“…A possible solution may be to redesign the injector to increase the electric field gradient on the cathode, leading to more uniform emission; since the details of the current distribution cannot be accurately modeled, empirical emittance measurements are the only option for evaluating the effects of a redesign on injector emittance. Another possible explanation for the higher than expected emittance is the interaction of the beam with the metallic pepper-pot foil -simulations show that the presence of a thin metallic foil at the injector exit doubles the emittance of the beam [24]. Together, these two effects may account for the discrepancy between predicted and measured emittance values.…”

“…Measurements at the exit of the accelerator yield a normalized emittance of 385 cm-mr on average, indicating emittance growth through the linac and highlighting the need for further diagnosis and modeling. Recent modeling efforts have confirmed the robustness of the FXR injector design with regards to mechanical alignment and position perturbations as well as small errors in the magnetic fields and, in general, suggest that the current injector design should produce a much lower emittance than is observed [24]. Three different codes (TRAK, OmniTrak, and LSP) were utilized to simulate the FXR injector and, although they agree well with each other, they clearly do not agree well with reality, suggesting that some feature of the underlying physics is not being correctly modeled.…”

“…Three different codes (TRAK, OmniTrak, and LSP) were utilized to simulate the FXR injector and, although they agree well with each other, they clearly do not agree well with reality, suggesting that some feature of the underlying physics is not being correctly modeled. A likely explanation for this discrepancy is that the emission from the cathode is not uniform -simulations show that a large (10% of total area) non-emitting region at the center of the cathode could double the emittance, and imaging of the plasma light generated at the cathode displays a non-uniform, mottled appearance which is, unfortunately, beyond current modeling capabilities to simulate [24]. A possible solution may be to redesign the injector to increase the electric field gradient on the cathode, leading to more uniform emission; since the details of the current distribution cannot be accurately modeled, empirical emittance measurements are the only option for evaluating the effects of a redesign on injector emittance.…”

FXR Optimization Project: Determining Beam Quality By Cathode Plasma Image Analysis

“…The larger the emittance, the larger the minimum spot size will be. Previous studies 9,10,11 of FXR led to the conclusion that the injector and, more specifically, the cathode were the source of higher than expected beam emittance. Images of the light produced on the cathode during the current pulse showed a very non-uniform light intensity distribution over the velvet emission surface.…”