Midterm Review 28-29/05/2015 Progress on wire-based accelerating structure alignment Natalia Galindo Munoz RF-structure development meeting 13/04/2016

Outline 1)Introduction 2)Methodology 3)Experimental Set-Up 4)Results 5)Conclusions, Improvements and Future Work Hélène Mainaud Durand PACMAN Mid-term review 28-29/05/2015 2

Introduction is a study on Particle Accelerator Components’ Metrology and Alignment to the Nanometre scale. Technical aim: to improve the alignment accuracy of the major CLIC components (Beam Position Monitors, Quadrupole Magnets and Accelerating Structures (AS)) by developing new methods and tools. The objective of this particular work is the development of direct measurement techniques for the in-situ internal alignment of the TD24 AS. Pre-alignment requirement for AS: The absolute position of the electrical center should be found with a precision of 7 µm in a laboratory environment using metrology methods. 3

The TD24 Accelerating Structure 4  23 cm long tapered normal-conducting traveling wave AS.  26 cells.  Accelerating Mode at 12 GHz.  Iris mean aperture: 5.5 mm  4 Wakefield Monitors (WFM) in the middle cell with RF pick-ups.  4 radial waveguides in each cell without RF-absorbers with a cut-off frequency of around 16 GHz without distorting the accelerating mode.

Methodology 5 1.The fist dipole mode is excited with a Vector Network Analyzer (VNA) through the WFM in the middle cell. 2.A 0.1 mm diameter stretched Be-Cu wire inside the AS perturbs the electromagnetic field by changing transmitted and reflected power signals between the RF pick- ups. 3.We change the position of the wire to minimize the perturbation. 4.We define this position as the electromagnetic axes in the AS and use the wire as a reference to fiducialise the structure in an accurate environment of a 3D Coordinate Measuring Machine (CMM).

Methodolog y 6 S parameters for different wire displacements: We choose a combinations of S parameters measured in amplitude for maximum sensitivity: S 41 -S 31 and S 42 -S 32, when moving the wire along the X axis. S 14 -S 24 and S 13 -S 23, when moving the wire along the Y axis.

Methodolog y 7 We simulate alternative scans in the X and Y axes with HFSS: Ideal case:  Geometric center = Electromagnetic center.  Accuracy is determined by numerical noise  We do a linear fit of the simulated measurements:  The line crosses zero at the positon of he electromagnetic center of he AS.  The slope represents the sensitivity of our measurements to the displacement of the wire.  Asymmetric coupling do not impact the result.

Methodology: Algorithm for experimental measurement s 8 Random Y position. AS displacements in X Linear fit of the measurements chose X 0 with better R 2 value Fix X at X 0 position. AS displacements in Y S 41 -S 31 S 42 -S 32 X port1 X port2 Y port3 Y port4 S 14 -S 24 S 13 -S 23 N iterations, until both X and Y converge Linear fit of the measurements chose Y 0 with better R 2 value

9 4 ports 10 MHz-50 GHz Resolution: 0.1 µm Accuracy: 4 µm over 100 mm Perpendicularity: 13 µm over 51 mm Accuracy: 0.012º Resolution: º Experimental Set-Up

10 User platform in LabVIEW: The movement of the stages are controlled based on the acquisition from the VNA and the data processing done by the program.

Results 11 Frequency sweep using the VNA. S 21 (dB) with respect to the frequency for different positions of the wire inside the AS  Reference frequency: 17 GHz  The variation of S parameters with the wire displacement is maximized.  Sensitivity: ≈ 7.5 dB/mm in the X axis ≈ 10 dB/mm in the Y axis S 21 (dB)

Results 12 Experimental algorithm with an example: Coarse AS displacements are done at first in the horizontal plane of the AS at a random vertical position while we measure the difference between the transmission coefficients from adjacent ports. Horizontal displacements: S 41 -S 31 S 42 -S 32

13 When we linear-fit each data series we obtain two different values for the zero crossing: X 1 = mm X 2 = mm We can assume this represents two independent measurements of the electronic centre. We choose for the next iteration the value for which the linear fit has a better R 2 and then iterate in the vertical axes. We select only the linear area Resolution: 100 µm 61 points Results

14 Results We select only the linear area Resolution: 100 µm 61 points Zero crossing:Y 1 = mm (better R 2 ) Y 2 = mm For the first two iterations, we roughly locate the central area of the structure measuring: 61 points along 6 mm Resolution: 100 µm For the following two iterations: 21 points along 1 mm Resolution: 50 µm We iterate again in the horizontal axis for this new fixed Y position.

Results 15 For the last scans we measured: 401 points Resolution: 1 µm After the first interactions, the centre as measured by each port converges. The variations of the position in successive measurements can be taken as given by numerical noise and other random causes and will be accounted for in the precision of the measurement. X 1 = mm X 2 = mm Y 1 = mm Y 2 = mm

Results µm

Results 17 5 µm

18 Results Measurement 1 Measurement 2 Measurement 3 Measurement 4 Measurement 5 Resolution: 1 µm Span: 0.4 mm Resolution: 1 µm Span: 0.4 mm

Results 19 Measurement 1 Measurement 2 Measurement 4 Measurement 5 Resolution: 1 µm Span: 0.4 mm

 We can locate the center of the electromagnetic field inside the middle cell of the AS with a stretched wire and a VNA.  We have a sensitivity in the micron range and use two different measurements of the transmission parameters to locate the center with very similar results.  We will use the rotation stage to minimize the tilt of the wire inside the AS.  This might lead in a change of experimental algorithm.  Effect of the relative tilt between the AS and the wire?  Look at other frequencies in order to optimize sensitivity.  Find a compensation between the span and the resolution.  We will compare the results obtained with the WFM with those using tapers. 20 Conclusions, Improvement s and Future Work

 We will make our measurements using different wire manufacturers.  We will implement a wire positioning system (1.5 µm repeatability) and fiducials to precisely locate in space the wire with metrology measurements.  Reproducibility.  Impact on averaging and filtering with VNA.  When do we touch the walls?  Should we choose the mean value of both minimum points found for each axes or are we doing well by choosing the best R 2 of the linear fit? How will this affect in the accuracy of our measurements? 21 Conclusions, Improvement s and Future Work

Midterm Review 28-29/05/2015 Thank you for your attention 22 RF-structure development meeting 13/04/2016