Version 4 (modified by md987, 8 years ago) (diff)


2011 IPMU-YITP on Monte Carlo Tools for the LHC, Yukawa Institute for Theoretical Physics, Kyoto, 5-10 Sep 2011


  • Rikkert Frederix (lecturer)
  • Claude Duhr (lecturer)


Exercises 1: Familiarize with MadGraph

  • Logon to the MadGraph web site and register:
  • Register with a valid e-mail address
  • Download the latest version of the MadGraph v5 code from the Dowloads page on the madgraph website.
  • Untar it and start the command line interface by executing =./bin/mg5=
  • Start the tutorial by typing

in the mg5 command prompt and follow the instruction. Discuss the results with your neighbours.

Exercises 2: Implement a new model with FeynRules

  • Read the tutorial sheet Tutorial.pdf and follow the instructions together with your neighbours. If you don't have Mathematica, you can download the model files here and jump directly to corresponding section of the tutorial sheet.

Extra exercises 1: On-line process generation

  • Use the on-line website to generate a few processes (with different QED + QCD couplings) trying to guess which diagrams appear:
    • uu~>tt~
    • gg>tt~
    • gg>tt~h (in SM and HEFT)
    • uu~>tt~bb~
  • Subprocess identification. List all subprocesses contributing to:
    • pp>l+vljj
    • pp>tt~bb~
    • pp>tt~jj
  • Look at the new physics models and check the particle and interactions content.

Extra exercises 2: Calculate cross sections and generate events

  • Generate events for a few selected processes (work in groups, pick one each), for Tevatron and LHC energies. Does the cross section scale as you expect?
  • Compare cross sections with the processes of your neighbors. Do they differ as you would expect?
  • Look at the generated plots. Are the distributions as you would expect? Discuss with your neighbors.

Example processes:

  • ttbar production: pp > tt~, QED=0
  • W production: pp > l+vl
  • Z/gamma* production: pp > e+e-
  • g g > H, H > e+ ve mu- vm~ (in HEFT model)
  • p p > W+ W-, W+>e+ ve, W- > mu- vm~ (important background to previous process)

Extra exercises 3: The full simulation chain and advanced features

Try out the Analysis Tools:

Study SM Higgs production at the LHC. Choose a channel and investigate signal and background, using either of the packages above:

  1. The 2 lepton + missing Et final state: $ pp \to H\to W^+ W^- \to e^- \bar \nu_e \mu^+ \nu_\mu $
    • Signal
Events: Parton Level
Detector Level
Plots: Parton Level
Detector Level
  • Background
Events: Parton Level <span style="color: #000000;">
</span>Detector Level
Plots: Parton Level <span style="color: #000000;">
</span>Detector Level
  1. Top associated production $pp \to t\bar tH$ with $H \to b \bar b$
    • Signal + Background samples:
Events: Parton Level
Detector Level
Plots: Parton Level
Detector Level

Extra exercises 4: Implementation of a new model in MadGraph or FeynRules + MadGraph

  • Implement your favourite Model into MadGraph using FeynRules, specific for studying a given process. Generate events, study relevant distributions.
  • Determine most relevant Standard Model backgrounds with identical final states (&ldquo;non-reducible&rdquo;). If not too complicated, generate backgrounds with cuts determined from signal distributions.
  • Which other (&ldquo;reducible&rdquo;) backgrounds might be important?
  • Can the signal be seen at Tevatron? At the LHC?


Attachments (14)