Challenge Problems

The concept of challenge problems for particulate flows can be traced back to the Fluidization VIII Conference in Tours, France (1995), where the first “Challenge Problem Workshop” was organized based on Prof. John Chen’s suggestion. Challenge problems are designed to test the predictability and accuracy of models and their numerical implementations. Ideally, such problems have well-characterized operating conditions to enable accurate representation of the system and involve repeats or replicates for high confidence measurements. As a result, reliable high-quality experimental measurements are generated which may be considered ground truth validation data. This helps in accelerating the development of simulation-based engineering and to identify existing modeling deficiencies.
Typical challenge problem workflows essentially breakdown into four steps:

  1. Formulation: A team is formed to create a new challenge problem. The challenge problem may be designed to test a sparse region of the existing validation phase-space. The problem is defined including the geometry, materials, data acquisition methods, data management plan, etc. Construction and shakedown tests begin.
  2. Announcement: The challenge problem is announced to the community. The geometry, material properties and boundary conditions are released. The problem statement is also provided, i.e., the required system response quantity or quantities of interest.
  3. Modeling: Willing participants model the challenge problem in a blind study, i.e., the experimental results are not known. Questions or missing pieces of information may be asked to the organizers. Modeling results are submitted to the organizers along with any feedback or potential concerns.
  4. Release: At the close of the modeling phase, the results from all respondents are collected by the organizers and released back to the participants and the community along with the experimental data. The release of respondent results may be either open or blind. Modelers may choose to publish and/or reassess their results of the study.

Information regarding NETL sponsored challenge problems can be found below:

• SSCP-II (COMING SOON)

• SSCP-I – https://mfix.netl.doe.gov/experimentation/challenge-problems/small-scale-challenge-problem-i-2013/

• CP-III – https://mfix.netl.doe.gov/experimentation/challenge-problems/challenge-problem-iii-2010/