The University of Manchester, The Cockcroft Institute,
Accelerator Physics, School of Physics and Astronomy,
Schuster Building, Oxford Road, Manchester, M139PL, UK.
Phone: +44-161 2750481
|[This page is still under construction]|
I work as a PostDoc Researcher, with Prof. Roger Jones in the accelerator physics group within School of Physics and Astronomy, at The University of Manchester, which is also a founding member of the The Cockcroft Institute. My research interests lies in the areas of RF acceleration structures, beam instrumentation and application of particle accelerator in medical for radio therapy and in security industry. It also expands into the plasma physics for nuclear fusion and novel acceleration techniques. I work in international collaborations with multiple institutes, such as, DESY (Hamburg, Germany), University of Rostock (Rostock, Germany), CLIC collaboration at CERN (Geneva, Switzerland).
Resent projects at Cockcroft Insitute - University of Manchester:
Wakefield Suppression in a Manifold Damped and Detuned Structures for CLIC Operating at 380 GeV.
Development of high gradient accelerating structure for CLIC 380 GeV staged design. The Compact Linear Collider~(CLIC) aims to collide electrons and positrons at a 3~TeV center of mass. A staged machine development approach has been adopted in which a first stage of 380~GeV machine will require a linac with accelerating gradient of 72~MeV/m, with the surface electromagnetic fields~(EM) and self-induced transverse wakefields bound by electromagnetic breakdown constraints and beam dynamics. We have investigated an alternate to the CLIC baseline linac design, using frequency detuning and moderate manifold damping (Q~4000) to suppress the wakefield. An optimum iris shape was determined by scanning iris dimensions using EM-simulations. Wakefield suppression over short range with different frequency detuning functions is studied using uncoupled model. The effect of moderate manifold damp ing is studied using more accurate spectral function method based on a double chain circuit model coupled to a manifold equivalent circuit. The associated spectral function are fitted using a series of Lorentzian peaks representing modes, to form a wakefield from the associated modal sum. We obtain consistent results with the inverse Fourier transform of the spectral function.
Fig. 1 a) Accelerating electric field along the axis of the linac and maximum electric and magnetic filed on linac surface. b) Wakefield in uncoupled structure damped using Gaussian frequency detuning and coupling out the field.
Fig. 2 Spectral function a) and corresponding wakefiield b) of a manifold coupled structure with 8-fold frequency interleaving.
Characterization of HOM spectrum in a chain of eight third harmonic SC cavities on European XFEL:
The AH1 module in E-XFEL contains a chain of eight superconducting (SC) cavities operating at 3.9~GHz, connected by the beam pipes and bellows. Introducing a charged particle beam into a cavity results in an EM field being excited. This field can be decomposed into a series of multi-poles -both along the axis of the cavity and transverse to it. The beam-excited longitudinal wakefield gives rise to an energy spread along the beam. The transverse component dilutes the beam emittance and can give rise to a beam break up (BBU) instability. Here we focus on the latter component.
The transverse component of the beam-excited wakefield can be decomposed into a series of multi-poles. Provided the beam is close to the axis of the cavity however, the dipole part will dominate, and consequently we restrict our study to this component of the multi-pole. This dipole field is also decomposed into a series of modes, which are in principle infinite. However, we confine our study to those having the largest impact on the transverse momentum of the beam. These higher order dipole modes can be utilised to determine the beam position, and can also be used to align the beam on the cavity axis. Moving the beam to the electrical centre minimises excitation of these damaging HOMs. We also use these HOMs to measure the misalignment of cavities. The frequencies of these HOMs in the third harmonic cavity (3HC) are above the beam pipe cut-off frequency and consequently the majority of the modes in the cavities couple to cavities throughout the chain in the module. This is in contradistinction to those in the 1.3 GHz cavities, which mainly consist of modes trapped within each individual cavity. Thus in these 3.9 GHz cavities it is necessary to model the complete chain as they essentially behave as one cavity "en masse".
As the majority of the dipole modes propagate throughout the AH1 module, it is effectively a 72-cell cavity. We focus on calculating the S-matrix of this module. Modelling this accurately would require the capability of supercomputers equipped with appropriately large memory and computing speed. To obviate these requirements we employ a well-established technique of electromagnetics known as the GSM, which essentially entails breaking up the structure under consideration into small blocks, which are accurately computed, and then the whole set of blocks is cascaded together to evaluate the overall S-matrix.
Fig. 1: Reconstruction of a single cavity by cascading three unit cells made of two end-cells and a mid-cell.
Fig. 2: Reconstruction of complete eight cavity chain by cascading the first and secon cavity unit cells (both cavities are also cascaded as in Fig. 1, and used as unit cells for further calculation.)
Fig. 3: Transmission S-parameters in the a fiirst dipole frequency band from HOM-1 to HOM-2 ports of cavity-1, connected in the eight cavity chain. The S-parameter curve is reconstructed using multiple Lorentzian peaks representing different resonance modes.
Fig. 4 Transmission parameters from HOM-1 port of cavity-1 to HOM-2 ports on cavity-2,4,6,8.
Fig. 5 Comparison of experimentally measured transmission parameters to the same simulated using GSM, from HOM-1 port of cavity-1 to HOM-2 ports on cavity-4.
Cavity Beam Position Monitor (CBPM) for electron accelerator (PhD with JAI at Royal Holloway University of London, 2009-2013):
Beam position monitors are required in all accelerators for the measurement and optimization of the beam parameters. Cavity beam position monitors (CBPM) offer the possibility of measurement of beam centroid positions at the nanometre scale. These devices can be and typically are used at electron accelerator facilities, both existing light sources and test facilities proposed for future linear colliders, such as the International Linear Collider (ILC) and Compact Linear Collider (CLIC).
The requirements for the CLIC main linac are to measure the beam position using approximately 5000 beam position monitors (BPM) with 50 nm resolution, at every 50 ns. The high resolution, enormous scale of the system and the small bunch separation of 0.5 ns present many challenges and demand innovative approaches for the design and operation of the CBPM system. A cylindrical cavity BPM system has been designed in collaboration with the Diamond Light Source, in the C-Band frequency region. The design ideas, such as the deliberate separation of modes coupled to the X and Y position measurements, and the cavity operation without mechanical tuning are tested in the design. The major resonance modes of the cavity are simulated using eigenmode simulation. The coupling and isolation characteristics are simulated using S-parameter simulations, while the beam coupling is studied through time domain simulations.
Four cavities were fabricated and their coupling and isolation were tested through S-parameter measurements. The dipole modes are separated by more than 5 MHz in frequency. The values of the quality factors were measured using the impedance method. The field orientation of the dipole and quadrupole modes were measured using the bead-pull perturbation technique and found to be rotated by 12║ and 3║ from X-axis respectively. The initial beam studies were carried out at the Diamond Light Source and at ATF2 beam line.
The techniques for position determination of temporally closely spaced bunches are studied. A method was developed to remove the errors in the position determination, due to the overlap of the signals from the previous bunches, by subtracting the decayed phasors from the previous bunch. The method is applied to the signals from the CBPM system on the ATF2 beam line, in the two and three bunch mode operation. The overestimation in position determination of the second bunch is reduced from more than 60% to less than 1%. Position resolution of better than 3Ám is demonstrated for the second bunch. The observed phase difference between the consecutive bunches is studied for different bunch spacings. The performance of the code is verified against simulated data.