Filename
stringlengths 22
64
| Paragraph
stringlengths 8
5.57k
|
|---|---|
Processed_A_Note_on_the_Overall_Magnification_of_a_Gravitati.txt | the primary image, as seen in Equation (4). This differs from the result implied by the use of the moderate an- gle formula of Wucknitz (2008), where the magnification tends to a finite amount always less than one. |
Processed_A_Note_on_the_Overall_Magnification_of_a_Gravitati.txt | The authors gratefully acknowledge the financial sup- port for this work provided by Australian Research Council Discovery Grant DP140100094. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | We present the first four-flavour lattice calculation of the leading-order hadronic vacuum- polarisation contribution to the anomalous magnetic moment of the muon, ahvp µ , and the hadronic running of the QED coupling constant, ∆α hvp QED(Q2). In the heavy sector a mixed-action setup is employed. The bare quark masses are determined from matching the K- and D-meson masses to their physical values. Several light quark masses are used in order to yield a controlled extrapola- tion to the physical pion mass by utilising a recently proposed improved method. We demonstrate that this method also works in the four-flavour case. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | c(cid:13) Copyright owned by the author(s) under the terms of the Creative Commons Attribution-NonCommercial-ShareAlike Licence. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | The anomalous magnetic moment of the muon, aµ , serves as a benchmark test of the standard model (SM). It has been measured very accurately [1, 2] and can be computed precisely within the SM. A comparison between the experimental result for aµ and the SM prediction reveals a discrepancy of more than three standard deviations which has persisted for many years now and has been confirmed by computations of a number of groups, see, for example, the review [3]. The question is whether this discrepancy originates from some effect missing in the experimental or theoretical determination of aµ or whether it points to physics beyond the SM. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | A key ingredient in the calculation of aµ is the leading-order hadronic vacuum-polarisation contribution, ahvp µ , which presently is the largest source of uncertainty in the theoretical compu- tation of aµ , since the QED and electroweak contributions have been computed very accurately employing perturbation theory, see [4, 3] and references therein. As ahvp is intrinsically nonpertur- bative, a lattice QCD computation of this observable is highly desirable. The currently accepted SM values for this quantity are obtained mainly from the investigation of e+e− scattering and τ decay data. In a recent study [5, 6], using two flavours of mass-degenerate quarks, a modified method to compute ahvp µ has been introduced resulting in a determination of ahvp µ,light with a precision of a few percent. Besides ahvp µ , the leading-order QCD contribution to the running of the QED coupling constant, ∆α hvp QED, also requires the hadronic vacuum-polarisation function and can likewise be investigated once this function is known. As an important input parameter to SM calculations, the QED coupling constant needs to be known very precisely in order to facilitate high precision tests of the SM at any future linear collider [7]. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | Our calculations are based on the configurations with four dynamical quark flavours generated by the European Twisted Mass Collaboration (ETMC) [11, 12]. These sets of configurations are obtained at different values of the lattice spacing and several lattice volumes, thus enabling us to estimate discretisation and finite size effects as systematic uncertainties of our lattice calculation. In addition, at each value of the lattice spacing, configurations exist at several values of the pion mass, ranging from 230 MeV (cid:46) mπ (cid:46) 450 MeV. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | The first term is the contribution of a narrow-width vector meson with mass mV and electromagnetic coupling gV, which are determined directly in our calculation. The remaining terms parametrise any deviations from this form. The results reported below are obtained using uncorrelated fits to determine gV and mV from the zero-momentum current correlators. In these proceedings we do not perform alternative fits such as the Padé approximants suggested in [17]. We also do not provide an estimate of systematic effects by allowing additional terms in eq. (2.5). These aspects will be addressed in a forthcoming publication. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | There are several possible choices for H. Below we give results for H = Hphys ≡ 1 (the standard choice) and H = mV(mPS), Hphys = mρ (improved choice). Here mV(mPS) is the light vector meson mass as measured on the lattice at unphysical pseudoscalar masses mPS whereas mρ is the physical value of the ρ-meson mass. See refs. [5, 6] for a more detailed discussion. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | Figure 1: Light quark contribution to ahvp as a function of the squared light pseudoscalar mass from our N f = 2 + 1 + 1 computations. The lower set of data points corresponds to using the standard definition, H = Hphys = 1 in eq. (2.8). The upper data points are obtained using the improved method, setting H = mV(mPS) and Hphys = mρ . The open square represents the two-flavour result of refs. [5, 6]. The dark grey error band corresponds to the quadratic fit (solid green line) and the light grey one belongs to the linear fit (dashed black line). |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | definition of ahvp proved definition, the pion mass dependence of ahvp the extrapolation to the physical pion mass. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | Figure 2: Comparison of the results for ahvp having the light up and down quarks as well as the strange quark in the valence sector for different collaborations. Only the upper set of data points for twisted mass fermions are obtained from the modified definition, H = mV(mPS) and Hphys = mρ in eq. (2.8). Data are from [19] (clover-improved) and [18] (domain wall). The open square represents the standard model value obtained from the dispersive analysis of ref. [8]. Concerning the error bands of the fits, the same comments as in fig. 1 apply. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | Fig. 4 shows our results for ∆α hvp QED at a typical hadronic scale of Q2 = 1 GeV2 obtained from µ . Similar to ahvp the same ensembles used to determine ahvp µ , we find reasonable agreement with the dispersive result [7] to within the statistical uncertainties. A full determination of the systematic uncertainty for the extrapolated value will also be presented in a later publication. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | performed a comprehensive analysis of the systematic uncertainties. In particular, the disconnected contributions originating from the strange quark might be non-negligible. Additionally, it might be necessary to take isospin breaking effects into account as the precision of our computation improves. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | We conclude by emphasising that the strategy followed here can also be applied to other ob- servables such as the Adler function, the corrections to the energy levels of muonic and ordinary hydrogen, and the weak mixing angle, as discussed in [6]. |
Processed_Leading-order_hadronic_contributions_to_a_μ_and_α_.txt | We thank Andreas Nube for performing the matching of the K- and D-meson masses in the mixed-action setup with their physical values. This work has been supported in part by the DFG Corroborative Research Center SFB/TR9. G.H. gratefully acknowledges the support of the German Academic National Foundation (Studienstiftung des deutschen Volkes e.V.) and of the DFG-funded Graduate School GK 1504. This manuscript has been coauthored by Jefferson Science Associates, LLC under Contract No. DE-AC05-06OR23177 with the U.S. Department of Energy. The numer- ical computations have been performed on the SGI system HLRN-II at the HLRN Supercomputing Service Berlin-Hannover, FZJ/GCS and BG/P at FZ-Jülich. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | Abstract: The chemical compositions of stars from the Asymptotic Giant Branch are still poorly known due to the low temperatures of their atmospheres and therefore the presence of many molecular transitions hampering the analysis of atomic lines. One way to overcome this difficulty is by the study of lines in regions free from molecular contamination. We have chosen some of those regions to study the chemical abundance of the S-type star GZ Peg. Stellar parameters are derived from spectroscopic analysis and a metallicity of -0.77 dex is found. Chemical abundances of 8 elements are reported and an enhancement of s-process elements is inferred, typical to that of an S-type star. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | The Asymptotic Giant Branch (AGB) is one of the most important phases of stellar evolution due to the rich nucleosynthesis that occurs during this stage, and therefore AGB stars are crucial objects for studies of Galactic chemical evolution. During this epoch stars produce important amounts of carbons and s-process elements which are brought to their atmospheres by deep dredged-up events and then ejected to the inter- stellar medium by stellar winds. Although fundamen- tal stones for the understanding of the final phases of low-mass stellar evolution and the stellar populations abundances, the chemical composition of such stars are still poorly known due to their low temperatures and to the blanketing of the atomic lines by molecu- lar transitions. Some authors have tried to overcome this problem by selecting spectral regions free from molecular contaminations or the so called molecular windows (e.g. Smith & Lambert 1985, 1990, Lambert et al. 1995, Vanture & Wallerstein 2002a,b, 2003), and the use of the near-infrared and infrared regions (e.g. Origlia & Rich 2005, Cunha & Smith 2006, Woolf & Wallerstein 2005a,b). Although working with a smaller number of lines, the chemical picture of these stars can still be delineated, provided that a careful inspection of the continuum is made and an appropriate model at- mosphere is adopted. The knowledge of the elemental abundances of such stars is an important tool for the understanding of the many steps followed by red giants along the M-MS-S-C spectral sequence, and their role in the chemical evolution of galaxies. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | Observations were carried out at the 1.52m telescope of ESO (European Southern Observatory), La Silla, in September 1999. The spectra were obtained using the FEROS spectrograph (Fiber-fed Extended Range Optical Spectrograph) with wavelength range 356 to 920 nm and a resolution of R = 48,000. Reductions were performed using the DRS (online data reduc- tion system of FEROS). A subsequent reduction was performed using the CONTINUUM, RVIDLINES and DOPCOR tasks of the IRAF package. The mean sig- nal to noisy ratio of the spectrum is S/N = 61.97. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | pected. We have adopted the bolometric luminosity of Mbol = -5.02 from Guandalini & Busso (2008). The temperature was inferred by the IRFM using the em- pirical calibrations of Alonso et al. (1999) and 2MASS colors. The resulting values are T(V-K) = 3502 K, T(J-H) = 3621 K and T(J-K) = 3454 K. These tem- peratures have been checked by the excitation equilib- rium of the Fe I lines (Fig. 1), and a best fit has been found for Teff = 3600K. Although S stars are predicted to have masses in a range of 1.5 - 2 M⊙ (Jorissen et al. 1998), GZ Peg is an extrinsic S star, therefore a series of tests have been made to better constrain the mass of this star. Tests requiring simultaneously (i) the excitation equilibrium of the Fe I lines and (ii) the curve of growth fitting, were made with masses of 3 M⊙ to 1.5 M⊙ and the iteration procedure converged for a model with 2 M⊙. This value has been adopted to derive the trigonometric surface gravity also using the Hipparcos paralaxe. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | The stellar parameters have been inferred by re- quiring the ionization equilibrium of the Fe I and Fe II lines to derive the spectroscopic gravity and metallic- ity, and zero slope in the [Fe/H] vs. EW (equivalent width) plot to infer the microturbulence velocity. The resulting stellar parameters are the following: Teff = 3600 K, log g = 0.06, [Fe/H] = -0.77 dex and ξ = 1.88 kms−1. The [Fe I/H] vs. EW relation is depicted in Fig. 2 and the curve of growth fitting is given in Fig. 3. The model atmospheres adopted here is a new version of the MARCS models atmospheres (Gustafsson et al. 2008). We have selected 2 M⊙ spherical models and interpolated between grids to obtain the best result. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | A change of 50K in the temperature yields a change in gravity of 0.025 dex and a change of 0.4 mag in the bolometric magnitude yields a change in gravity of 0.16 dex. We have determined the errors in stellar abun- dances due to stellar parameters uncertainties taking into account a change of ∆ log g = +0.2 dex in gravity, ∆ Teff = +50 K in temperature, ∆ [Fe/H] = +0.15 dex in metallicity and ∆ ξ = 0.10 kms−1 in microturbu- lence velocity. The resulting errors are given in Table 1. The uncertainties have been added in quadrature and the total errors are given in the last column. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | Table 2 we give the final abundances with the number of lines used in the analysis and the statistical errors (σ/√N ). |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | In order to better understand the origin of the abundance pattern of GZ Peg, we have compared in Fig. 5 the resulting abundances with those predicted by Goriely & Mowlavi (2000) for an AGB star with metallicity [Fe/H] = -0.70 and mass of 2.5 M⊙ which has undergone 10, 30 and 50 dredge-ups. Surface abun- dances predicted for 10 dredge-ups are depicted by the solid line, for 30 dredge-ups by the dotted line, and for 50 dredge-ups by the dashed line. Except for the Ba abundance, which has been derived from only one line, GZ Peg data are better described by the 30 dredge- up model. Such analysis may indicate that this star has probably accreted matter from an AGB star which has passed through an important number of dredge-up events. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | A detailed analysis of the extrinsic S star GZ Peg has been performed. The spectroscopic stellar parameters derived for this star are: Teff = 3600 K, log g = 0.06, [Fe/H] = -0.77 dex and ξ = 1.88 kms−1. The derived s-process abundances are in agreement with those of S-type stars or other stars with enhanced s-process abundances. Comparing the abundance level of GZ Peg with theoretical surface abundances of AGB stars from Goriely & Mowlavi (2000) we have found a good agreement with the 30 dredge-up model. Such anal- ysis indicates that the pristine AGB star which con- taminated GZ Peg has passed through a considerable number of dredge-up events. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | Figure 5: [X/Fe] vs. Z for GZ Peg compared to theoretical surface abundances predicted for AGB stars of Goriely & Mowlavi 2000. Our data are compared to models with [Fe/H] = -0.70 and M=2.5M⊙. The solid line represents surface abun- dances predicted for 10 dredge-ups, the dotted line for 30 dredge-ups, and the dashed line for 50 dredge-ups. Error bars are the same as in Fig. 4. |
Processed_Chemical_Abundances_of_the_S_star_GZ_Peg.txt | Table 1: Changes in chemical abundances due to stellar parameters uncertainties. In the last column we give the total error. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Cancer therapy using protons and heavier ions such as carbon has demonstrated advantages over other radiotherapy treatments. To bring about the next generation of clinical facilities, the requirements are likely to reduce the footprint, obtain beam intensities above 1 × 1010 particles per spill, and achieve faster extraction for more rapid, flexible treat- ment. This review follows the technical development of ion therapy, discussing how machine parameters have evolved, as well as trends emerging in technologies for novel treat- ments such as FLASH. To conclude, the future prospects of ion therapy accelerators are evaluated. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | INTRODUCTION In recent years, using hadrons for radiotherapy has be- come more widely recognised for significant benefits with the sparing of healthy tissue [1]. Carbon ions can be more beneficial than light ions when treating radio-resistant and deep-seated tumours, however the adoption of carbon ion therapy is hampered by the footprint of the accelerator and gantry, which are presently much larger than conventional radiotherapy devices [2]. This has been the focus for im- provement over the past 20+ years [3]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | CURRENT AND DEVELOPING BEAM DELIVERY SYSTEMS There have been many accelerators built for hadron ther- apy. A diagram of the progress of representative machines is shown in Fig. 1. The general trend is that the size of machines has decreased, whereas the number of particles per beam spill rises as is required for new treatments. In addition, details of specific designs are given in the main text and briefly summarised in Table 1 (with further detail in the Appendix). |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The design choice for most facilities is based on a syn- chrotron over a cyclotron, as it is capable of acceleration of particles with higher magnetic rigidity than protons, and the beam can be extracted over a wide range of energies, avoiding losses due to energy modulation [3]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The first hadron therapy treatments were delivered at the Bevelac [4]. Hadron therapy was also delivered at GSI from 1997 to 2007; treatment there was taken over by HIT [5]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Figure 1: Progress in the design of ion therapy accelerators over time. Accelerators with relatively few particles per spill rely on technologies that allow for a higher repetition (cycle) rate to produce the same overall current, such as the use of static magnetic fields for PAMELA. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The first accelerator purpose-built for heavy ion therapy was the HIMAC at NIRS in Japan, which began treatment in 1994 [6]. The synchrotron has two separate, identical rings, each able to deliver ions from 100−800 MeV/u. Several similar synchrotrons were then developed to achieve the same clinical requirements with a smaller machine and lower costs. One option was the PATRO project, resulting in the HIBMC where carbon ion treatment started in 2003 [7]. Both machines deliver to smaller energy ranges than HIMAC, reducing the circumference. A more compact alternative at NIRS led to GHMC (12C only), which used the same focusing structure but increased the number of dipoles to three per cell [8]. The GHMC design has been the basis for the most recent generation of Japanese ion therapy centres. In Europe, HIT has offered full 3D raster scanning since treatments began in 2009, delivering protons, carbon and heavier ions [9]. This was supplemented by the Proton-Ion Medical Machine Study (PIMMS), a collaboration to provide a strong baseline for future cancer therapy synchrotrons [10]. The lattice uses triplet focusing, where the dispersion-free straights are used for injection, extraction, and acceleration by RF; the extraction method uses a betatron core. Two machines were subsequently built from the PIMMS design at CNAO (Italy) and MedAustron (Austria). |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Table 1: Summary of parameters of ion therapy machines. Some also offer proton therapy, detail not given here. Proposals that were never constructed are given in italics. For a full version with extended detail including citations, see Table 2. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | IST ion therapy centre proposal [11, 12]. The aim is to offer a range of ions with an order of magnitude more particles per spill, while also being smaller than its predecessors. It is expected to offer slow extraction with multiple energies per spill, and fast extraction at microsecond timescales. This requires extensive R&D, with three main proposals: a com- pact synchrotron using normal-conducting (NC) magnets; an even smaller synchrotron using superconducting (SC) magnets; and a ‘full linac’ design. Of these, the linac option requires the most R&D [12]. The NC synchrotron has two options: 1) a modification of the PIMMS design with im- proved injection and extraction, or 2) a novel double bend achromat (DBA) design with a smaller circumference [13]. The SC proposal [14] follows ideas evolved at NIRS [15], using large-angle magnets with canted cosine theta (CCT) geometry [16] and nested alternating gradient (AG) coils [17]. Although this option is more compact, development is likely to take considerable time. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | There has also been some interest in an ion Rapid Cycling Medical Synchrotron (iRCMS), which allows for more rapid energy variation assuming single energy extraction [18]. Other than the higher magnet ramp rate, the iRCMS would have similar parameters to other synchrotron options. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | For example, a therapeutic heavy ion cyclotron has been proposed at ARCHADE in France [19, 20], potentially be- ginning treatments in 2023. Developments in high gradi- ent cavities have made linacs more promising. As well as NIMMS, designs have also been presented by AVO-ADAM [21](p+) and ANL (p+, 12C); the latter is still in its design phase [22]. A ‘bent’ full-linac for carbon ions has also been designed [23]. CABOTO, an NC ‘cyclinac’, is a fast-cycling linac design, but has not yet been built due to the requirement of multiple high frequency klystron systems [24]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The time-independent magnets of Fixed Field Accelera- tors (FFAs) allow for a higher cycle rate, which could in- crease the number of particles delivered without requiring more particles per spill. However, FFAs tend to be larger than equivalent synchrotron counterparts, and magnet designs can become complicated. Multiple proposals for hadron ther- apy FFAs exist: the most-developed of these was PAMELA [25], which required two rings to accelerate protons and carbon ions over their full energy ranges (see Table 1), but had a much higher cycle rate (up to 1 kHz) than equivalent synchrotrons (<1 Hz). Although designs continue to ad- vance, none have been constructed for treatment, and R&D is required in areas such as beam stability and extraction. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Though synchrotrons are the current workhorse for heavy ion therapy, there are alternatives that could be implemented. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | stage. In the first stage, a Petawatt laser is incident on a thin target, producing an ion beam with low mean energy but broad energy spread [27]. LhARA is a similar proposal [28]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The gantry is one of the key components of therapeutic beam delivery systems. It bends and focuses the beam in the plane perpendicular to a patient to deliver a precise dose to the treatment volume [2, 29], with a typical momentum ac- ceptance of (± 0.5 − 1 %)Δ𝑝/𝑝 [30]. The majority of proton gantries use NC magnets. As carbon ions have three times the rigidity of protons, a suitable gantry’s size and magnet weight increases dramatically; the first carbon ion gantry (at HIT) weighs 600 t and measures 25 m [31], making instal- lation and integration a challenge. Though it can transport fully-stripped carbon ions in the range 48-430 MeV/u [32], an alternative solution is required to shrink the gantry and increase momentum acceptance to deliver flexible, efficient treatment. One option is to use SC magnets: HIMAC em- ploys six 2.88 T combined-function SC magnets, reducing gantry mass to < 300 t and the length to 13 m [15]; a develop- ing design will incorporate 5 T magnets, reducing the gantry radius to < 5 m [15]. A recent design by TERA/CERN pro- poses a 35 t gantry[14], utilising 90°, 4 T AG-CCT dipoles [33] to reduce the radius to 5 m. These magnets are similar to those used for the NIMMS SC synchrotron. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Another path is to modify the gantry configuration. An example is the ‘GaToroid’ (in its R&D phase), which re- places the rotating beam transfer line with a large toroidal field, combined with a ‘vector’ (steering) magnet to bend the beams from several directions towards the patient isocentre [34, 35]. One could also use high momentum acceptance FFA-style magnets [36], though no such gantries have yet been constructed. A final solution may be a mounted gantry, similar to one constructed by Mevion (p+) that rotates around the patient isocentre [37]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | RESEARCH DIRECTIONS The main targets of hadron therapy R&D are: cost and size reduction [3]; improved reliability; new treatment capability. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | To reduce the size of accelerators and gantries, the options are either to change the structure and/or use SC technology. An example of the former is the NIMMS baseline lattice, which could be shrunk by using a DBA. Further reductions are limited by the necessity of including long drifts for ex- traction, while also providing enough space for the required bending elements. Conversely, the adoption of SC mag- nets uses similar beam optics, but the larger fields allow for tighter bending radii, significantly reducing the overall size. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The use of SC magnets in accelerator and gantry design can increase the maximum field strength. Using combined- function magnets also reduces the total number of magnets. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | This can lead to cost reductions for the mechanical support structures and construction, but must be evaluated against the price of the cooling system and materials used for the magnets [3]. To provide the most benefit, the system’s mag- netic composition (SC or hybrid NC/SC), configuration (e.g. CCT) and material (e.g. Nb-Ti) must be considered. CCT Nb-Ti magnets are being developed for future machines [12]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Most ion therapy facilities now achieve slow extraction by excitation of a third-order resonance in combination with RF- knockout (RF-KO), as it can be used to extract multiple beam energies over the course of a single cycle [38]. This allows different energy layers to be delivered rapidly, however the energy switching time is too long for treatments such as FLASH [39]. Some also use a kicker for fast extraction, delivering the entire stored beam in a single turn. The higher dose rates this provides may be useful for FLASH, but will be insufficient if rapid energy variation is also required. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | One consideration for future extraction methods is the possibility to deliver multiple ions in a single treatment. This may be desirable to achieve a more conformal dose than with a single species [40], or for novel imaging methods such as particle tomography and range verification [41]. For multi-ion treatments, the issue is similar to that of rapid energy switching, and it is not yet clear what technologies will provide the required advances. A proposed SC gantry may be capable of performing proton tomography, so long as it has sufficiently large momentum acceptance [42]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Almost all current ion therapy accelerators use a linac to accelerate ions before injection, although some facilities (in China) opt for a cyclotron [43]. Future pre-accelerators may need to transmit larger currents and reach higher energies to avoid space-charge effects in high intensity beams, such as those required for FLASH therapy [39]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | The main objectives for new pre-accelerator systems are to reduce cost and size, while maintaining or increasing beam current. These depend on the ion source used; many facilities use the Supernanogan [44], an ECR ion source, although higher current options such as AISHa [45] or TwinEBIS [46] may come into greater use. An EBIS can produce small emittance beams, allowing for higher current through multi- turn injection, but cannot yet match the ECR in reliability and intensity. The linac itself will likely move to higher gradients, but the low duty cycle of injectors for medical synchrotrons makes SC options less favourable [12]. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Novel ion production and pre-acceleration methods may eventually take over from linacs. laser- accelerator methods discussed earlier may provide beams of sufficient energy and intensity for clinical use. However, these technologies require further development; improve- ments on traditional methods will likely be used for the next generation of ion therapy machines. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | REFERENCES [1] W. D. Newhauser et al., “The physics of proton therapy,” Physics in Medicine & Biology, vol. 60, no. 8, 2015. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Field Alternating Gradient Accelerator for Protons and Car- bon Ions for Charged Particle Therapy,” PRAB, vol. 16, 2013. [26] T. Shirai et al., “Design of quantum scalpel for the new heavy ion radiotherapy,” 15th Annual Meeting of Particle Accelera- tor Society of Japan, 2018. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | APPENDIX In Table 2 below, we give a full version of the ion therapy machine summary. This includes additional information on each accelerator, as well as some other facilities based on the same base machine design, and full citations for all the information. |
Processed_Review_of_Technologies_for_Ion_Therapy_Accelerator.txt | Table 2: Parameters of ion therapy machines. Entries that are proposals but never constructed are given in italics. As well as the sources given in the table, additional information is available from [47] and [48]. |
Processed_Kinematics_of_a_hot_massive_accretion_disk_candida.txt | H. Beuther1 & A. Walsh2 [email protected],[email protected] Draft version from November 26, 2018. Accepted for publication in Astrophysical Journal Letters. |
Processed_Kinematics_of_a_hot_massive_accretion_disk_candida.txt | Characterizing rotation, infall and accretion disks around high-mass protostars is an important topic in massive star formation research. With the Australia Telescope Compact Array and the Very Large Array we studied a massive disk candidate at high angular resolution in ammonia (NH3(4,4) & (5,5)) tracing the warm disk but not the envelope. The observations resolved at ∼0.4′′ resolution (corresponding to ∼1400 AU) a velocity gradient indicative of rotation perpendicular to the molecular outflow. Assuming a Keplerian accretion disk, the estimated protostar-disk mass would be high, similar to the protostellar mass. Furthermore, the position-velocity diagram exhibits additional deviation from a Keplerian rotation profile which may be caused by infalling gas and/or a self-gravitating disk. Moreover, a large fraction of the rotating gas is at temperatures >100 K, markedly different to typical low-mass accretion disks. In addition, we resolve a central double-lobe cm continuum structure perpendicular to the rotation. We identify this with an ionized, optically thick jet. |
Processed_Kinematics_of_a_hot_massive_accretion_disk_candida.txt | Outflow and jet studies in high-mass star-forming re- gions have revealed ample indirect evidence that accre- tion disks should be common in these regions. However direct observational characterization of such disks has so far been difficult (e.g., Beuther et al. 2007; Cesaroni et al. 2007). While this is partly caused by the typically large distances and the clustered mode of massive star forma- tion, an additional problem arises because these regions are still deeply embedded within their natal cores, they have visual extinctions of the order 1000, making it diffi- cult to disentangle the dense core emission from the gen- uine accretion disks. In the past, different molecular line tracers were employed for different massive star-forming regions (e.g., Cesaroni et al. 2007), however, it is interest- ing to note that for many sources only one or two molecular lines exclusively indicated rotating motions whereas many other promising spectral lines showed no such signatures (e.g., Beuther 2007). The problem of identifying a reliable massive disk tracer prohibited statistical studies of larger disk samples, until today. Here we present new observa- tions toward the massive disk candidate IRAS 18089-1732 in the high-excitation ammonia lines NH3(4,4) and (5,5), with lower energy levels of 200 and 295 K, respectively. While the high excitation temperatures ensure that we are not tracing the cold gas envelope but only the warm central regions (e.g., Cesaroni et al. 1998), recent radiative trans- fer calculations of 3D hydrodynamic massive core simu- lations indicated that the wavelength regime < 100 GHz should be particularly advantageous to studying the inner disk regions (Krumholz et al. 2007a). This work predicts high dust optical depths at frequencies >100 GHz poten- tially limiting spectral line studies of the central disk in the (sub)mm regime. |
Processed_Kinematics_of_a_hot_massive_accretion_disk_candida.txt | ied at (sub)mm wavelengths with the Submillimeter Ar- ray, and a velocity gradient was identified from east to west, perpendicular to the molecular outflow emanat- ing from the core approximately in north-south direc- tion (Beuther et al. 2004, 2005). However, these data did not resolve the velocity pattern of this structure and hence did not allow to study its kinematic proper- ties. The region is located at ∼3.6 kpc, it has a bolo- metric luminosity of about 104.5 L⊙, and it exhibits sev- eral other signs of early massive star formation (e.g., Walsh et al. 1998; Sridharan et al. 2002; Beuther et al. 2002a,b; Williams et al. 2004, 2005; Fuller et al. 2005; Edris et al. 2007). |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Abstract In light of recent work of the third author, we revisit a classic example given by Fritz John of a semi-linear wave equation which exhibits finite in time blow up for all compactly supported data. We present the construction of future global solutions from asymptotic data given in [12] for this specific example, and clarify the relation of this result of Yu to John’s theorem in [6]. Furthermore we present a novel blow up result for finite energy solutions satisfying a sign condition due to the first author, and invoke this result to show that the constructed backwards in time solutions blow up in the past. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | At the MATRIX workshop on Hyperbolic PDEs and Nonlinear Evolution Prob- lems, in September 2023, Dongxiao Yu presented several global existence results from asymptotic data for a class of quasi-linear wave equations [12]. These back- wards in time results are compelling because they apply in particular to wave equa- tions for which forward in time global existence is not known, or worse, are known to admit blow up solutions. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | This is a semi-linear wave equation for which Fritz John showed that forward in time solutions arising from data of compact support blow up in finite time [6]. We recall the statement of this well-known fact, and his argument briefly in Section 2; excellent expositions of John’s results in [6] can be found in the lecture notes [1, 2]. The forward in time results use the initial value formulation for (1), and rely on the method of spherical means. The backward in time solutions constructed by Yu use Hörmander’s asymptotic system [3, 4, 5] to find an approximate solution, and rely on energy estimates for the remainder. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | where r = |x|, q = r − t, and ω = x/|x|. In particular, scattering solutions can be constructed from knowledge of the radiation field U(q,ω). |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | The situation is very different for (1). As we will discuss in this note, the exis- tence results for (1) that go beyond almost global existence for small data, depend on carefully chosen data, either asymptotically or initially. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Following [12], we first prepare a solution to (5) which does not blow up, and then find a corresponding solution to (1). This construction is a special case of the results in [12] which apply more generally to quasi-linear wave equations. In contrast to (3), these solutions do not have a radiation field. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | exist Ck solutions of (1) for t ≥ 0. We will call these solutions future global. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | This also shows that the blow-up results of John [6] do not trivially extend from compactly supported data to general finite energy solutions. As we have seen, there are some finite energy solutions, namely the solutions of Theorem 3 which are future global. However, this raises the question if these solutions are also past global. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | In Section 4, we show with an argument that is due to the first author, that the solutions of Theorem 3 are not global: While they are future global, they blow up in the past. Theorem 11 gives a sufficient sign condition for blow up of finite energy solutions to (1), and Corollary 12 shows that the solutions constructed in [12] satisfy that sign condition. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | The occurence of a sign condition is expected in this setting: Already for the asympototic equation (5) the sign of the initial data for Uq at s = 0 is relevant for existence, for s > 0. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Thus this note clarifies the relation between the global existence results of the third author, and the blow up results of John, in the special case of the equation (1). We leave open the question if there are any global finite energy solutions to (1). |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | with smooth data of compact support u0, u1 ∈ C∞ 0 . If u ∈ C2(R3 × [0, ∞), then u = 0 identically. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | This ODE should be compared to (5). In both cases the sign of the initial data is crucial: In (8) it is chosen to ensure blow-up, and for (5) it is chosen to prevent blow- up. In view of these simple examples it is also not surprising that a sign condition is relevant in the general setting of Section 4. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | for (t, r) ∈ ΣR = {(t, r) : R < R + r < t < 2r}. For a detailed derivation of this in- equality we refer the reader to [1, 2]. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | from the blow-up argument in Section 4, which yields blow up at the level of ∂t u. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | We have seen in the previous section, that according to the result of John [6], any nontrivial C3 solution to (1) with compactly supported initial data blows up in finite time. In this section, however, we discuss a recent result in [12, Corollary 8.3] on the construction of nontrivial future global solutions to (1). |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Theorem 3 (Yu ’23) There exist a large family of nontrivial future global solutions to (1) for t ≥ 0. Moreover, for each of these future global solutions, the solution and its time derivative cannot be both compactly supported at t = 0. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Remark 4 For each constant C > 0, the function − ln(t + C) is a future global so- lution to (1); cf. Remark 2. However, the future global solutions constructed in [12] decay as |x| → ∞ and thus are different from these constant-in-x solutions. See, e.g., the estimates (23) below. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Remark 5 To be more precise, in this paper, we prove that for each nonzero func- c (R× S2) with A ≥ 0, there exists a nontrivial future global solution tion A(q,ω) ∈ C∞ u to (1). The solution u is associated to the function A in the following way. At infi- nite time, the solution u matches a future global solution U(s, q,ω) to the asymptotic equation (17). Meanwhile, this U is uniquely determined by the data Uq|s=0 = A and the boundary condition limq→−∞ U(s, q,ω) = 0. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Remark 6 Note that the second half of Theorem 3 follows directly Theorem 1. Meanwhile, we can also prove that the data of these nontrivial future global solutions are not compactly supported without applying John’s finite time blowup result. See Section 3.3. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Remark 7 Theorem 3 can be generalized. First, it was mentioned in Remark 5 that we start with some function A and construct a corresponding future global solution to (1). The assumptions of A can be relaxed. Instead of assuming A ∈ C∞ c , we only need to assume that A along with its derivatives decays faster than certain negative powers of hqi as |q| → ∞. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Besides, the construction used in the proof of Theorem 3 also applies to a large family of quasilinear wave equations in R1+3. We refer our readers to [12] for more details. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | We may compare this observation to Remark 2. Indeed the asymptotic equation (18) is precisely of the form (8). In both cases the blow-up is sensitive to the sign of the initial data. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Here ZI denotes a product of |I| commuting vector fields: translations ∂α for α = 0, 1, 2, 3, scaling S = t∂t +r∂r, rotations Ωi j = xi∂j −x j∂i for i, j = 1, 2, 3 and Lorentz boosts Ω0i = xi∂t + t∂i for i = 1, 2, 3. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | However, if the data of u is compactly supported, we have u ≡ 0 whenever r − t ≥ R0 for some constant R0 > 0. Note that the intersection of {r − t ≥ R0} and {t > 1, r − t > q0, |r − t| < tγ} is not empty for all sufficiently large t, so here is a contradiction. We conclude that the solution u constructed above cannot have compactly supported initial data. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | conditions for solutions to (25) to blow up in finite positive time or finite negative time respectively. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | ulin(2r − q, x0) = ∂r(rMx0 [u0])(2r − q) + (2r − q)Mx0[u1](2r − q). |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | and so by the same argument the second sign condition (29) is satisfied if there exists some x0 ∈ R3, q ≥ 0 such that ulin(q, x0) < 0. It follows that if neither of (28),(29) hold, then the third sign condition (30) is necessarily satisfied. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Theorem 11 (Bernhardt ’24) Suppose u is a nonzero solution of the initial value problem (6) with initial data u(0, x) = u0(x), ∂t u(0, x) = u1(x). Suppose u0, u1 ∈ C2(R3) satisfy the decay assumption (27), and let ulin be the solution to the homo- geneous wave equation (cid:3)ulin = 0 with matching initial data u0, u1. If ulin satisfies the sign condition (28), then ∂t u → ∞ in finite positive time. Alternatively, if ulin satisfies (29), then ∂t u → −∞ in finite negative time. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | By (34), this implies that the spherical mean Mx0 [∂t u(r − q)](r) blows up, and hence ∂t u → ∞ at some point (t, x) ∈ (0, ∞) × R3. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Let vlin(t) denote the solution to the homogeneous wave equation with the same initial data as v. By uniqueness, we have for all t ∈ R that vlin(t) = ulin(−t). As ∂t vlin(t) = −∂tulin(−t), it is clear that if ulin satisfies the second sign condition, then vlin satisfies the first sign condition. From the previous argument, ∂t v → ∞ in finite positive time , and so since ∂t u(t) = −∂tv(−t), it follows that ∂t u → −∞ in finite negative time. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | Corollary 12 Let u be a C2 solution to (25) constructed in Theorem 3, and let ulin be the solution to the homogeneous wave equation (cid:3)ulin = 0 with the same initial data as u. Then ulin does not satisfy the third sign condition (30). That is, there exists some t ∈ R, x ∈ R3 such that ulin(t, x) < 0. Moreoever, ∂t u blows up in finite negative time. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | These upper and lower bounds cannot both be true, hence ulin(t, x) < 0 for some (t, x) ∈ R1+3. |
Processed_John's_blow_up_examples_and_scattering_solutions_f.txt | As γ > 1, u still decays faster than O(r−1), which is incompatible with the lower bound (42). Hence the solutions constructed in [12] also blow up in finite negative time. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | This paper proposes a generalised and robust multi-factor Gated Recurrent Unit (GRU) based Deep Learning (DL) model to forecast electricity load in distribution networks during wildfire seasons. The flexible modelling methods consider data input structure, calendar effects and correlation-based lead- ing temperature conditions. Compared to the regular use of instantaneous temperature, the Mean Ab- solute Percentage Error (MAPE) is decreased by 30.73% by using the proposed input feature selection and leading temperature relationships. Our model is generalised and applied to eight real distribution networks in Victoria, Australia, during the wildfire seasons of 2015-2020. We demonstrate that the GRU-based model consistently outperforms another DL model, Long Short-Term Memory (LSTM), at every step, giving average improvements in Mean Squared Error (MSE) and MAPE of 10.06% and 12.86%, respectively. The sensitivity to large-scale climate variability in training data sets, e.g. El Niño or La Niña years, is considered to understand the possible consequences for load forecast- ing performance stability, showing minimal impact. Other factors such as regional poverty rate and large-scale off-peak electricity use are potential factors to further improve forecast performance. The proposed method achieves an average forecast MAPE of around 3%, giving a potential annual energy saving of AU$80.46 million for the state of Victoria. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | To achieve the goal as set out in the Paris Agreement of minimising global mean temperature increase to under 2°C and ideally not more than 1.5 °C goal, a growing coalition of countries, businesses, and institutions have committed to implementing a low carbon energy transition thereby con- tributing to a net-zero emission future [1]. In the era of the energy transition, the diversification of the generation mix and electrification of household demand both raise great un- certainty in the power system. Meanwhile, extreme weather events are becoming more frequent and severe due to Cli- mate Change. Climate-related disasters increased drastically from 3,656 events during the last 20 years of the 20𝑡ℎ century to 6,681 events in the first 20 years of the 21𝑠𝑡 century [2]. The doubled risks of floods, 40% more risk of storms, and a higher risk of wildfires, droughts, and heatwaves result in significant economic and social losses. The increasing cli- mate variability also challenges the reliability and resilience of the energy system operation, e.g., more and more uninten- tional power blackouts and emergent power constraints are triggered due to extreme weather events. Between them, the emerging energy transition and the extreme weather events simultaneously bring uncertainty to the energy system, gen- erating significant extra demands for power system opera- tors. In recent years, power grid damage attributed to wildfire risks has become more and more common, such as decreased line capacity, conductor line sag, unintentional power cuts, and a number of less apparent but no less impactful effects. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | techniques, such as DL, are worthy of study to address un- certainty, multi-factor and nonlinear issues in load forecast for realistic and extreme use cases. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | forecast using SVM, considering the past seven-day load, av- erage daily temperature and electricity price as the model input. From the beginning of the 21𝑠𝑡 century, some ANN- based basic forecast models have been designed and improved to capture the complex non-linear behaviour of energy use [18], [19]. In the past decade, ANN techniques have been further developed in two main directions: Recurrent Neural Network (RNN), with a specialisation in handling sequen- tial data, and Convolution Neural network (CNN), which has the strength to capture the spatial features of data (im- age recognition) [7]. In [15], the basic RNN model and its derivative techniques: Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) models were studied to forecast household load in the next few hours. As the LSTM and GRU are techniques good at understanding the sequen- tial data, LSTM and GRU are selected as the main model structure for the research reported in this paper, and the mem- ory cell structures are explained in detail in the Methodol- ogy. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | Our paper proposes a multi-factor DL-based load fore- cast model via a case study of 8 real DNs in Victoria, Aus- tralia, achieving the average MAPE of around 3% during wildfire seasons. In this subsection, each paragraph discusses one research gap and how we address the issue, reviewing the main contributions of our paper. |
Processed_A_generalised_multi-factor_deep_learning_electrici.txt | So far, studies have yet to be implemented to assess the im- pact of adjusting input structures for the RNN-based load forecast model. In [15] and [20], the 24-hour ahead load data was default selected as the input for the DL forecast models since it is thought to be more appropriate in learn- ing the characteristics exhibiting daily periodicity. How- ever, the performance accuracy may already reach a stable level even with the load data input shorter than 24 hours. If we can find the shortest length of input that can achieve acceptable accuracy, the forecast model can be equipped with a higher computational efficiency. - The input matrix structure of the multi-factor LSTM- based and GRU-based forecast model is adjusted to find the appropriate input data length with both good accuracy and less computational burden. |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.