# Tikz-Feynman: Putting multiple boson loops into each other

I was trying to draw a Feynman diagram of a quark propagator in rainbow ladder truncation when I encountered a problem.
I need to draw multiple gluon loops within each other. Drawing one gluon loop is no problem, but trying to put another one into that results in something that looks like

The problem are the last two diagrams in that equation. What I want is the diagrams to look somewhat like

Please, excuse my bad drawing. I hope you get, what I mean. I produced the above result with the code

\documentclass[a4paper]{article}
\usepackage{tikz-feynman}
\usepackage{tikz}
\usetikzlibrary{shapes,arrows,positioning,automata,backgrounds,calc,er,patterns}
\tikzfeynmanset{compat=1.0.0}
\begin{document}
$\feynmandiagram[horizontal=a to b, tree layout]{ a-- [fermion] b; }; \quad\raisebox{-0.5ex}{=}\quad \feynmandiagram[horizontal=a to b, tree layout]{ a -- b; }; \raisebox{-0.5ex}{+} \feynmandiagram[horizontal=b to c, tree layout] { b -- c, b -- [gluon, half left] c; }; \raisebox{-0.5ex}{+} \feynmandiagram[horizontal=a to d, tree layout] { a -- b -- c -- d, b -- [gluon, half left] c, a -- [gluon, half left] d; }; \raisebox{-0.5ex}{+} \feynmandiagram[horizontal=a to f, tree layout] { a -- b -- c -- d -- e -- f, c -- [gluon, half left] d, b -- [gluon, half left] e, a -- [gluon, half left] f; }; \raisebox{-0.5ex}{+\dots}$
\end{document}


If I don't use "tree layout" attribute, the diagrams look like

which is still not what I want. Using version 1.1.0 instead of 1.0.0 does not change anything either.
The compiler I am using is LuaLatex.

• Your code gives you precisely what you wan (I think) if you compile with pdflatex. – marmot Mar 6 '18 at 16:58
• It indeed does, thanks! The problem is that I have other diagrams in the same document, that do not compile properly using pdflatex but do fine with LuaLatex. I worked around this by compiling this diagram in pdflatex and then putting it in the document as a graphic. So I guess this topic is solved. – CrustyPeanut Mar 6 '18 at 17:13
• In principle you could work with \diagram* (see end of section 3.1 in the manual) which "uses the most basic algorithm". – marmot Mar 6 '18 at 17:19
• I think it is advantageous to go the \diagram* route since this way global settings of the document (e.g. the line width of the propagators) will automatically apply to the diagram without the need of re-compiling the diagram. – marmot Mar 6 '18 at 17:43

From your comment I understand that you have several diagrams in your document that rely on lualatex. Therefore, I propose to use \diagram* (see end of section 3.1 in the manual) which "uses the most basic algorithm".

\documentclass[a4paper]{article}
\usepackage{tikz-feynman}
\usetikzlibrary{shapes,arrows,positioning,automata,backgrounds,calc,er,patterns}
\tikzfeynmanset{compat=1.0.0}
\begin{document}
$\tikz{\begin{feynman}\diagram*{ a-- [fermion] b; };\end{feynman}} \quad\raisebox{-0.5ex}{=}\quad \tikz{\begin{feynman}\diagram*{ a -- b; };\end{feynman}} \raisebox{-0.5ex}{+} \tikz{\begin{feynman}\diagram*{ b -- c, b -- [gluon, half left] c; };\end{feynman}} \raisebox{-0.5ex}{+} \tikz{\begin{feynman}\diagram*{ a -- b -- c -- d, b -- [gluon, half left] c, a -- [gluon, half left] d; };\end{feynman}} \raisebox{-0.5ex}{+} \tikz{\begin{feynman}\diagram*{ a -- b -- c -- d -- e -- f, c -- [gluon, half left] d, b -- [gluon, half left] e, a -- [gluon, half left] f; };\end{feynman}} \raisebox{-0.5ex}{+\dots}$
\end{document}


I know that you already went another way, but maybe this answer will be helpful for others who encounter similar problems in the future.

• This works just fine, too. I implemented it this way now since it is more elegant. Thanks! – CrustyPeanut Mar 7 '18 at 14:15