Modeling Update

[Morgan Bruns]

The following email was just sent out to the MaCFP Condensed Phase emailing list. I've copied it below just in case there's anyone out there who we missed.

Dear MaCFP Contributors,

Thank you for your continued involvement in the IAFSS Measurement and Computation of Fire Phenomena (MaCFP) Condensed Phase Workshop. We were very pleased to have over 40 participants in attendance for a virtual presentation of experimental measurements on October 15. So far, data sets have been submitted by 16 research groups from 10 different countries around the world. You may access all of this data on our GitHub repository. We also would like to encourage you to participate in our Google group by asking or answering questions as they arise.

 

In preparation for the upcoming MaCFP workshop on April 22-23, 2021 (exact dates are tentative), the organizing committee would like to encourage you to participate in the second phase of this workshop by providing model predictions for a set of target scenarios. The stated goal of the Condensed Phase side of MaCFP is "to improve computational predictions of thermal degradation and pyrolysis in fire scenarios." While the experimental phase of the upcoming workshop involved characterizing the pyrolysis behavior of black, cast PMMA, the modeling phase will challenge our ability to make computational predictions of how this material will behave in several scenarios relevant to fire safety science. In short, the experimental phase provides raw data for the modeling phase to extract model input parameters for use in making predictions. Both the experimental and modeling phases will help facilitate communication and data sharing within the community directed towards the development of better computational predictions of fire phenomena.

 

Next Steps: Preparation and Submission of Modeling Results (submit by Feb. 22, 2021)

 

A detailed description of the modeling phase can be found in the Guidelines for Participation in the 2021 MaCFP Condensed Phase Workshop. In the following, we will share a little more about the steps ahead of us, but please note that all submissions should be received by February 22, 2021 to ensure inclusion in the April workshop. Please email your submissions to Morgan Bruns (bru...@vmi.edu), and we will work to have your contributions publicly available on GitHub several weeks prior to the workshop.

 

At the highest level, all modeling submissions should include three components:

1.  a description of your process to determine your model parameters;

2.  a formatted table of all model parameters; and

3.  predictions of the requested target data.

The process description should include the following pieces of information:

 

• a list of the MaCFP experimental data sets used to determine your model parameters;
  
• a brief description of how you obtained parameters from the data (e.g., a particular optimization algorithm, trial and error, an algebraic model, etc.);
  
• a list of any literature data that was used to supplement the MaCFP data (for instance, we do not currently have any data relevant to radiative material properties so you might need to rely on literature values for emissivity, absorptivity, etc.); and
  
• what model you used to predict the target data (e.g., Gpyro, FDS, ThermaKin, or your own code).
  

 

The table of model parameters should be an organized systematically with a format similar to Table 1 on page 11 of the guidelines document. Please make sure to include both (1) the parameter name, (2) the parameter units in SI, and (3) the numerical value of the parameter.

The target data predictions should correspond to the scenarios described in detail on page 12 of the guidelines document. There you will find that we are interested in predictions of two TGA scenarios and six gasification scenarios. Furthermore, please submit your predictions of the requested target data as CSV files formatted according to the specifications on page 13 of the guidelines document.

 

Finally, in order to keep our work aligned with our goals, it is important that participants adhere to the following constraints:

 

• you must use some of the MaCFP data---at least one of the mg-scale data sets (TGA, DSC, MCC) and one g-scale experiment (Cone, FPA, Gasification/CAPA), or at least two of the g-scale experiments; 
  
• you can use some literature data for any parameters that you were unable to obtain from the MaCFP data; and
  
• you cannot use experimental data of the target scenarios since this would no longer make your simulation result a true prediction.

Hopefully, this email will provide you with everything you need to get started making computational predictions of the target scenarios. We are very excited to see the submissions as we continue to believe that MaCFP will be foundational for improving our ability to make reliable predictions of flame spread and fire growth. Please let us know if you have any questions, and thank you again for your willingness to contribute.

 

Sincerely,

 

The MaCFP Condensed Phase Working Group Organizing Committee

Benjamin Batiot (University of Poitiers, France)

Morgan Bruns (Virginia Military Institute, USA)

Simo Hostikka (Aalto University, Finland)

Isaac Leventon (National Institute of Standards and Technology, USA)

Yuji Nakamura (Toyohashi University of Technology, Japan)

Pedro Reszka (Universidad Adolfo Ibáñez, Chile)

Thomas Rogaume (University of Poitiers, France)

Stanislav Stoliarov (University of Maryland, USA)

[Franz Richter]

Dear all,

I would like to ask three questions. These questions came up as I initially thought that the target scenarios would just be simulated and not compared to experimental data, which I am no longer certain is the case.

First:

The target simulations include TGA of a 100 K/min at 5 mg. How can we assure experimental data to be accurate given the evidence (below) in the current literature?

Would it be better to focus on lower heating rates as a target, which we can determine accurately?

Evidence:

Lyon et al. suggest that there will be at least an error of 5 %. Likely more: https://www.sciencedirect.com/science/article/pii/S0040603112003334

Antal et al. suggest even higher error: https://pubs.acs.org/doi/10.1021/ie950368i

My own calculations would also suggest an error: https://www.frontiersin.org/articles/10.3389/fmech.2018.00018/full

Second:

Whose experimental data are we using as a target and can, therefore, not use in the model development?

Third:

Which experimental set-up does not have convection at the free surface? Could I double check that you mean h = 0 with that.

Best wishes,

Franz Richter

-------------------------------------------------------------------------

Franz Richter, Ph.D.

Department of Mechanical Engineering

University of California, Berkeley

Email: franz....@berkeley.edu

--
You received this message because you are subscribed to the Google Groups "MaCFP Condensed Phase Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to macfp-condensed-phase-...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-condensed-phase-discussions/93b543bd-9b5e-41ed-a778-320bc5b6a18bn%40googlegroups.com.

[Stanislav I. Stoliarov]

Hi Franz,

These are idealized conditions that are unambiguous and easy to simulate. They were selected to compare models and determine whether modeling predictions are close or far apart.   As you correctly point out, these conditions do not correspond to any real experiments and the results of these simulations will not be compared to any experiments.

Stas 

--------------------------------------------------------
Dr. Stanislav I. Stoliarov, Professor
University of Maryland
Department of Fire Protection Engineering
3104C J.M. Patterson Bldg.
4356 Stadium Dr.
College Park, MD 20742
Phone: 301.405.0928
Fax: 301.405.9383
Email: sto...@umd.edu

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-condensed-phase-discussions/CANnL_wVn00%3D-aVmCDBSxSKRrWuHNO9JaEmmHZs4jRLzeGqgfvQ%40mail.gmail.com.