source: ETALON/papers/2019_IPAC/i3d_BPM/FRXXPLS1.tex @ 802

% !TeX spellcheck = en_GB
% !TeX program = lualatex
%
% v 2.3  Feb 2019   Volker RW Schaa
%               # changes in the collaboration therefore updated file "jacow-collaboration.tex"
%               # all References with DOIs have their period/full stop before the DOI (after pp. or year)
%               # in the author/affiliation block all ZIP codes in square brackets removed as it was not %         understood as optional parameter and ZIP codes had bin put in brackets
%       # References to the current IPAC are changed to "IPAC'19, Melbourne, Australia"
%       # font for ‘url’ style changed to ‘newtxtt’ as it is easier to distinguish "O" and "0"
%
\documentclass[a4paper,
               %boxit,        % check whether paper is inside correct margins
               %titlepage,    % separate title page
               %refpage       % separate references
               %biblatex,     % biblatex is used
               keeplastbox,   % flushend option: not to un-indent last line in References
               %nospread,     % flushend option: do not fill with whitespace to balance columns
               %hyphens,      % allow \url to hyphenate at "-" (hyphens)
               %xetex,        % use XeLaTeX to process the file
               %luatex,       % use LuaLaTeX to process the file
               ]{jacow}
%
% ONLY FOR \footnote in table/tabular
%
\usepackage{pdfpages,multirow,ragged2e} %
%
% CHANGE SEQUENCE OF GRAPHICS EXTENSION TO BE EMBEDDED
% ----------------------------------------------------
% test for XeTeX where the sequence is by default eps-> pdf, jpg, png, pdf, ...
%    and the JACoW template provides JACpic2v3.eps and JACpic2v3.jpg which
%    might generates errors, therefore PNG and JPG first
%
\makeatletter%
        \ifboolexpr{bool{xetex}}
         {\renewcommand{\Gin@extensions}{.pdf,%
                            .png,.jpg,.bmp,.pict,.tif,.psd,.mac,.sga,.tga,.gif,%
                            .eps,.ps,%
                            }}{}
\makeatother

% CHECK FOR XeTeX/LuaTeX BEFORE DEFINING AN INPUT ENCODING
% --------------------------------------------------------
%   utf8  is default for XeTeX/LuaTeX
%   utf8  in LaTeX only realises a small portion of codes
%
\ifboolexpr{bool{xetex} or bool{luatex}} % test for XeTeX/LuaTeX
 {}                                      % input encoding is utf8 by default
 {\usepackage[utf8]{inputenc}}           % switch to utf8

\usepackage[USenglish]{babel}

%
% if BibLaTeX is used
%
\ifboolexpr{bool{jacowbiblatex}}%
 {%
  \addbibresource{biblio.bib}
%  \addbibresource{biblatex-examples.bib}
 }{}
\listfiles

%%
%%   Lengths for the spaces in the title
%%   \setlength\titleblockstartskip{..}  %before title, default 3pt
%%   \setlength\titleblockmiddleskip{..} %between title + author, default 1em
%%   \setlength\titleblockendskip{..}    %afterauthor, default 1em

\begin{document}

\title{Tests of a 3d Printed \NoCaseChange{BPM} With a Stretched Wire and With a Particle Beam  \thanks{Work supported by the CNRS/IN2P3 under the 3D-Metal program.}}

\author{N. Delerue\thanks{delerue@lal.in2p3.fr}, D. Auguste, J. Bonis, F. Gauthier, S. Jenzer, and A. Moutardier,\\ LAL, Univ. Paris-Sud, CNRS/IN2P3, Universit\'e Paris-Saclay, Orsay, France.\\
}
        
\maketitle

%
\begin{abstract}
We have successfully printed a beam position monitor using 3D printing. After ultra-high vacuum testing and initial measurements with a network analyser we now reports on tests of this BPM using the stretched wire method. The BPM has been installed on a test stand with a wire going through it and electrical pulses have been sent. The signal measured on the pick-ups was compared to that of two conventional BPMs and shows no anomaly specific to the 3D printed BPMs. Following the success of these tests we have also installed this BPM in a beam line at the PhotoInjector at LAL (PHIL). We show that it can give position measurements with an accuracy comparable to that of other BPMs.
\end{abstract}


\section{Introduction}
During the past few years we have been investigating the potential of metal additive manufacturing (metal 3D printing) for particle accelerator. We have first demonstrated that part produced using Selective Laser Melting (SLM - one of the main additive manufacturing technique for metals) can be compatible with the requirements of Ultra-High Vacuum (see~\cite{Jenzer:IPAC17-WEPVA043}). Additional studies related to the use of additive manufacturing for UHV beam pipes are discussed this year in another paper~\cite{Sattonnay:IPAC19-WEXXPLS3}.
More recently we have produced a beam position monitor and tested it with the Lambertson method~\cite{Delerue:IPAC18-THPAL016}. We have now taken the tests ofthis BPM two steps further, first by testing it on a test bench with the stretched wire method and this is discussed in section~\ref{sec:stretched} and then by installing it in an accelerator to study its behaviour with a real beam and this is discussed in section~\ref{sec:phil}.

The BPM used for the tests described in the paper has been printed using Selective Laser Melting. The powder used to print it was 316L powder. Its design is based on the design of conventionnal BPMs that have produced for another project in our lab. However the design has been optimized using topological optimization as described in~\cite{Delerue:IPAC18-THPAL016}. A CAD drawing of the BPM as well as pictures can be seen on figure~\ref{fig:BPM_view}.

\begin{figure}[!htb]
        \centering
        \includegraphics*[width=90mm]{THPAL016f2.jpg}\\
        \includegraphics*[height=25mm]{THPAL016f4a.jpg}
        \includegraphics*[height=25mm]{THPAL016f4b.jpg}
        \caption{CAD 3D view of the i3D BPM (top) and pictures of the BPM after printing. The shapes have been created by tolopogical optimization. Electrical feedthrough have been added separately.}
        \label{fig:BPM_view}
\end{figure}

After additive manufacturing minor work was still required in the workshop to sharpen the vacuum knife-edges and to solder the electrical feedthrough (which were not made by additive manufacturing) .
This BPM was about 50\% cheaper than a BPM machined in the mechanical workshop and conventionnal one it took two weeks to build against 6 weeks with conventionnal tools. For the same functionnalities the design was more compact as more complex shapes could be done using additive manufacturing.

\section{Stretched wire tests}
\label{sec:stretched}
To measure the resolution of the BPM we used the stretched wire method. 
A short beasmline was assembled using two conventionnal BPMs and this i3D BPM (with beam pipes in between) and a wire was stretched in their middle as shown on image~\ref{fig:test_bench}.


\begin{figure*}[htbp]
        \centering
        \includegraphics*[width=125mm]{stretched_wire_setup.png}
        \caption{The test bench where the three BPMs have been measured. The BPM built using additive manufacturing is in the middle et the two ones built using conventionnal means are on the ends.  Translation stages allow to move the BPMs vertically and horizontally (transverse to the BPM axis). A conductong metallic wire is stretched through the BPMs and electric pulses are sent on it.}
        \label{fig:test_bench}
\end{figure*}


Electric pulses (\SI{5}{V},\SI{10}{$\mu s$}) were sent on the wire. The signal on the four electrodes of the three BPMs were recorded with a Libera Brillance+~\cite{Condamoor:IBIC18-TUPB12} in single pass mode while the BPM triplet was moved vertically using a stepper motor. 


\begin{figure}[htbp]
        \centering
        \includegraphics*[width=65mm]{BPM_tests_setup.png}
        \caption{Measurement setup: a pulse generator is used to send pulses on the stretched wire and trigger signals to the machine clock (\SI{8}{MHz} generator) and to the libera. The libera reads the signal on the four electrodes of each of the three BPMs. The data recorded by the libera are read by a PC using Tango.}
        \label{fig:electrical_setup}
\end{figure}

Example of data acquired are shown on figure~\ref{fig:stretched_wire_data}. The horizontal axis is the estimated position according tp the translation stage settings and the vertical axis is the value calculated by the libera based on  the signal read on the electrodes.
 The figure shows the position for the two conventional BPMs (green and blue dots) and the i3D BPM (red dots). As can be seen the measure for the three BPMs are similar indicating that the i3D BPM has a response similar to that of the conventionnal BPMs.

\begin{figure}[!htb]
        \centering
        \includegraphics*[width=75mm]{IPAC_19_16.png}\\
        \includegraphics*[width=75mm]{IPAC_19_zoom_16.png}
        \caption{Position measured by the three BPMs (after calculation by the Libera). The horizontal axis shows the estimated position according to the number of steps sent to the motor and the vertical axis shows the measured position. To make the figure easier to read an offset has been applied to the value read by the BPMs so that of the three BPM are reading 0 when the motor is at position 0. The red dots correspond  to the i3D  BPM and the blue and  green dots to the conventionnel BPM.}
        \label{fig:stretched_wire_data}
\end{figure}


\section{Beam tests}
\label{sec:phil}

\begin{figure*}[htbp]
        \centering
        \includegraphics*[width=115mm]{BPM_on_PHIL.png}
        \caption{The BPM triplet installed on PHIL. The beam is coming from  the left to the right.  On  the picture one can see (from left to red), first an ICT then one of the conventionnal BPM, the i3D BPM  and the other conventionnal BPM. There is no magnetic elemen t between the BPMs.}
        \label{fig:BPM_on_PHIL}
\end{figure*}

Following these tests we decided to install the BPM triplet on a real electron accelerator, that is the Photo-injector PHIL~\cite{Alves:2012pw} (see image \ref{fig:BPM_on_PHIL}). The BPMs, including the i3D BPM were cleaned and leak tested by the LAL Vacuum group who approved them for installation in the Ultra High Vacuum of the photo-injector.

The BPMs were installed in the accelerator beam line after a focusing solenoid and two steering magnets. The beam energy was \SI{3.5}|{MeV} and its size at the BPMs location about \SI{2.5}{mm}. During  the experiment changes in the machine settings were then made to see the response of each BPM to these changes. We found that the i3D BPM had overall higher readings on all electrodes, indicating  that the electrode pick-up was better. This can be explained by a better impedance matching of the electrodes of this BPM than that of the other BPMs. An example of raw data measured with this BPM triplet is shown on figure~\ref{fig:data_on_PHIL_position}.


\begin{figure}[htbp]
        \centering
        \includegraphics*[width=85mm]{BPM_positions.png}\\
        \caption{Bottom four plots: Reading by each of the four electrodes (SpVa, SpVb, SpVc and SpVd) of the three BPMs (represented by three colors, red, blue and gree, the i3D BPM is in red). Top two plots: X (horizontal) and Y (vertical) position of the beam calculated by the libera. Given that the reading the i3D BPM were higher a factor two higher attenuation was applied to these electrodes. The data were acquired in single pass mode.}
        \label{fig:data_on_PHIL_position}
\end{figure}


As there was no magnetic element between the BPMs, the electron beam follows a ballistic trajectory as it travels through the triplet. It is thus possible to measure the accuracy of the position measurement of the triplet by comparing the reading of the center BPM with the reading expected by drawing a straight line between the two other BPMs. The result of such analysis is shown on figure~\ref{fig:data_on_PHIL_residuals}. As we can see the difference between these two values is less than \SI{200}{\micro m}  in the horizontal  plane ($x$) and less than  \SI{400}{\micro m}  in the vertical  plane ($y$). These difference are similar to those observed using conventionnal BPMs only  and thus using an additively manufactured BPM did not degrade the position resolution of tyhe  triplet.

\begin{figure}[htbp]
        \centering
        \includegraphics*[width=85mm]{BPM_positions_residuals.png}\\
        \caption{Difference between the position calculated by drawing a straight line between the two outer BPMs and the central one. }
        \label{fig:data_on_PHIL_residuals}
\end{figure}


\section{Outlook}

The tests we have done show that the performances of a BPM built using additive manufacturing are  comparable or better than those of a BPM built using conventional methods. Following these tests we decided that future BPMs installed on the ThomX accelerator will be made using  additive mnanufacturing so that we can gain experience over a long period and confirm that the performances remain comparable or better than tthose of a conventionnal BPM.


\section{ACKNOWLEDGEMENTS}
We are grateful to the IN2P3 "3D Metal" network who funded part of the work reported here.

%
% only for "biblatex"
%
\ifboolexpr{bool{jacowbiblatex}}%
        {\printbibliography}%
        {%
        % "biblatex" is not used, go the "manual" way
        
        %\begin{thebibliography}{99}   % Use for  10-99  references
        \begin{thebibliography}{9} % Use for 1-9 references
        
        
        
        \bibitem{Jenzer:IPAC17-WEPVA043}
        S. Jenzer \emph{et al.},
        \textquotedblleft{Study of the Suitability of 3D Printing for Ultra-High Vacuum Applications}\textquotedblright,
        in \emph{Proc. 8th Int. Particle Accelerator Conf. (IPAC'17)}, Copenhagen, Denmark, May 2017, pp. 3356--3358.
        \url{doi:10.18429/JACoW-IPAC2017-WEPVA043}            
        
        \bibitem{Sattonnay:IPAC19-WEXXPLS3}
        G. Sattonnay \emph{et al.},
        \textquotedblleft{Is It Possible to Use Additive Manufacturing for Accelerator UHV Beam Pipes?}\textquotedblright,
        presented at the 10th Int. Particle Accelerator Conf. (IPAC'19), Melbourne, Australia, May 2019, paper WEXXPLS3, this conference. 
        
        \bibitem{Delerue:IPAC18-THPAL016}
N. Delerue \emph{et al.},
\textquotedblleft{Study of the Performances of a 3D Printed BPM}\textquotedblright,
in \emph{Proc. 9th Int. Particle Accelerator Conf. (IPAC'18)}, Vancouver, Canada, Apr.-May 2018, pp. 3656--3659.
\url{doi:10.18429/JACoW-IPAC2018-THPAL016}            

%\cite{Condamoor:IBIC18-TUPB12}
\bibitem{Condamoor:IBIC18-TUPB12}
S. Condamoor \emph{et al.},
\textquotedblleft{Machine Studies with Libera Instruments at the SLAC Spear3 Accelerators}\textquotedblright,
in \emph{Proc. 7th International Beam Instrumentation Conference (IBIC'18)}, Shanghai, China, Sep. 2018, paper TUPB12, pp. 284.       
        
%\cite{Alves:2012pw}
\bibitem{Alves:2012pw}
M.~Alves {\it et al.},
%``PHIL photoinjector test line,''
JINST {\bf 8} (2013) T01001
doi:10.1088/1748-0221/8/01/T01001
[arXiv:1209.6222 [physics.acc-ph]].
%%CITATION = doi:10.1088/1748-0221/8/01/T01001;%%
%15 citations counted in INSPIRE as of 15 May 2019
        
        \end{thebibliography}
} % end \ifboolexpr
%
% for use as JACoW template the inclusion of the ANNEX parts have been commented out
% to generate the complete documentation please remove the "%" of the next two commands
% 
%%%\newpage

%%%\include{annexes-A4}

\end{document}