2019 10 15 Task2.2 Coordination Meeting - ibpsa/project1 GitHub Wiki

When

October 15, 5:00pm CET/MEZ (= 17:00 German time = 8:00am Pacific time)

Participants

  • Jérôme Frisch (RWTH E3D)
  • Christian Waluga (liNear GmbH)
  • Christoph Nytsch-Geusen (UdK)
  • David Jansen (RWTH EBC)
  • Andreas Geiger (KIT)
  • Karl-Heinz Häfele (KIT)

Agenda

Discussion and Minutes

Today's discussion continues the face-to-face meetings of the break-out sessions of Rome and will discuss further steps.

  • A modular concept is chosen as a basis for a step-by-step implementation based on available resources of the different contributing partners. Missing modules can be added, altered or adapted to use cases.
  • The first proposed module M1 was discussed and altered according to input from KIT. An IFC file (basically IFC4 or IFC2x3) is checked according to MVDXML rules. A report is generated and delivered to the user. This module is already implemented by KIT in their FZKViewer and is made available to all IBPSA Project 1 participants.
  • Module 2 transforms an IFC file into a BREP geometry. The process should be available for IFC4 and IFC2x3 files. E3D is creating the process for IFC4 and UdK for IFC2x3 files. The code will be provided online and a merge into one process for IFC4 and IFC2x3 should be attempted.
  • The modular approach should also enable a data exchange in between the modules, but a direct transfer in the computer's main memory should also be supported where possible in order to speed up the process.
  • The question arose if a new IFC file could also be the result of M2 which contains now the new BREP geometry plus all other, non-geometric attributes from the original file.
  • For M3 and M4 a way enabling a parametric modification would also be interesting. Currently this is not intended.
  • Nevertheless, M4 contains in the current version potential problems as the question arises which "errors" should be fixed automatically, how to detect them, and how to distinguish in between explicit modelling decisions vs. errors.
  • One further modules could be interesting for creating a decomposition of the model with the purpose of energy simulations. This includes the creation and extraction of space boundaries.
  • The main discussion of this meeting involved the further process and the direction. The group consensus lies in following the process of the defined modules and adding missing ones as soon as possible.

Next Steps

  • Eric provides the code for module 2 on git. Christoph will compare with their implementation and we commonly try to find a way for merging both implementations into one common module.
  • The modules are analysed and an implementation will be discussed in the next meeting.
  • The next meeting is scheduled for end of November. Next proposed meeting will be on Nov. 25 2019, 17:00 CET/MEZ (= 17:00 German time = 8:00am Pacific time)

Thanks for the participation.

(To all participants: please feel free to amend the minutes if positions are missing or need adaptation)