MaCFP-2 Condensed Phase Workshop - Developing standard data set formats for experimental data on pyrolysis

[MaCFP Condensed Phase Discussions]

During the Condensed Phase portion of the MaCFP-2 Workshop, we had a lively discussion on a number of topics. In hopes of continuing that progress, I have copied below some of the key questions & comments provided during this meeting.

This thread focuses specifically on what should be our next steps to develop standard data set formats for experimental data on pyrolysis. Please feel free to add new questions to this thread or your thoughts on how to address any of these issues.

1. Is the Information / Formatting of READMEs: sufficient? That is, are the requested metadata ( calibration approach and frequency, test setup, boundary conditions...) sufficient?

Do we need additional info (e.g., backing material physical description and thermophysical properties; back and front surface boundary conditions) and who is responsible for providing this information (timeframe)?

2. How do we identify datasets with missing information, who is responsible for making those needed edits, and should 'insufficient' datasets be removed or just 'flagged'. What is an appropriate timeline and feedback loop for such work.

3. Is the structure of the repository sufficient (especially considering long term goals - potentially more materials, more data types, and flame spread cases)?

4. An extended, parsable format for the README files was proposed- @ T. Hehnen, could you briefly outline that work?

[Dietenberger, Mark -FS]

Thanks to all the presenters for a very nice show. I think I managed to momentarily get my face in there, but not sure how to participate given this is my first virtual conference, and being in MaCFP only a few months.

 

1. When we found out years ago the problems of PMMA of being consistent, price variability, and other problems, I turned to a well-known, consistent, and low cost thermoplastic used in microwaves, and similar electronic devices use, which is Rexolite Polystyrene. It is already optimized for its various uses, not likely to change on us in 10 years. It is very sooty when it burns, where one has to adjust the standards for HRR calculation to account for the non-combustion of the soot. However, we found it to be an ideal “standard” in every other way, ie. testing in TGA, DSC, K-tester, cone calorimeter, mass loss calorimeter, and even in our Modified Schlyter Test (Much like Stan’s vertical parallel plates experiment). By the way we found the pure ethylene glycol as an ideal “standard” material for the cone calorimeter, as it burns as pure as methane (is even better than methanol, as you can use reasonable irradiance on the glycol to pyrolyze it at 200 Celsius) and it saves us in large costs as compared to pure methane tanks…
2. Years ago we did some corner wall tests (like Stan’s) with redwood boards (I have done a lot of tests of redwood in the cone calorimeter) under the HRR hood, but wanted to wait on reporting the experimental data until I was able to achieve the changes required to the source code for either GPYRO or FDS (defining the parameters for the current equations in those codes led to unphysical values, as the late Robert H. White used to call it “Dietenberger’s Fudge Factors” just to get under my skin). There is a lot work in trying to get wood pyrolysis kinetics and also accounting for the phenomena of creeping flame behavior, that I found evident in our larger scale tests. By the way I have developed a Reduced Ignition and Flame Spread (RIFT) for use in the cone calorimeter (it needs the weigh scale design that was used in the Atlas II Cone Calorimeter), for the sole purpose of calibrating the parameters for creeping flame spread (think of faster flame spreading as reducing the time to ignition, just as higher heat fluxes reduces the time to ignition).
3. However, I was distracted by a quite attractive offer to develop pyrolysis and combustion models for live vegetation (with the intention of going back to wood material, by using the same formulations). These formulations also include changes to Cp, K, and density as a result of foliage degradation (as function of temperature, moisture content, & mass losses all the way to ashes), I would have like to have known of the MaCFP, before starting to publish, as I thought I was alone, in terms of defining the nomenclature, methodology, and database requirements, so now have to adjust…
4. I was bemused by the pyrolysis kinetics presentation for the relatively simple PMMA properties, as I had found the GPYRO and FDS useless to calibrate parameters for the live foliage degradation that we obtained in the TGA, using the final rates 3.5, 10, and 60 K/min, live and dried vegetation, and nitrogen and air environments. I was going for simultaneous fitting of 16 different profiles (mass loss and DTG) with a single set of kinetics parameters that also accounted for lipids, protein/sugar, hemicellulose/petin, cellulose/starch, lignin/tannin, and minerals, including their charring behavior. Not possible with the way the GPYRO and FDS are set up, so I developed analytical solutions to the kinetics equation that can go from the beginning and all the way to the end of the degradation process as measured in the TGA with an arbitrary temperature profile, and able to do it with Excel Spreadsheet with its solver (I’m sure I would have better luck with MatLab spreadsheet), but the Excel spreadsheet worked just fine. Babak Shotoban, at my urging, put my kinetics (both equations and parameters values) in the GPYRO, and I believe in MatLab too, and was happy to report on their close agreement with my results in the Excel spreadsheet. Sorry for this lengthy discussion, but if you are looking long term on development/standardization of databases, nomenclature, methodology, accuracies, and so on, our results with live vegetation shows that you have a long ways to go.

 

Best regards, Mark

 

 

Mark Dietenberger, PhD Research General Engineer Forest Service Forest Products Laboratory p: 608-231-9531 f: 608-231-9303 mark.a.di...@usda.gov One Gifford Pinchot Dr. Madison, WI 53726 www.fs.fed.us Caring for the land and serving people

 

 

From: MaCFP Condensed Phase Discussions <macfp-condensed-...@googlegroups.com>
Sent: Friday, April 23, 2021 12:37 PM
To: MaCFP Condensed Phase Discussions <macfp-condensed-...@googlegroups.com>
Subject: [External Email]MaCFP-2 Condensed Phase Workshop - Developing standard data set formats for experimental data on pyrolysis

 

[External Email]
If this message comes from an unexpected sender or references a vague/unexpected topic;
Use caution before clicking links or opening attachments.
Please send any concerns or suspicious messages to: Spam....@usda.gov

--
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/a741652d-5cd5-46e3-8091-92ba9e03c6d7n%40googlegroups.com.

This electronic message contains information generated by the USDA solely for the intended recipients. Any unauthorized interception of this message or the use or disclosure of the information it contains may violate the law and subject the violator to civil or criminal penalties. If you believe you have received this message in error, please notify the sender and delete the email immediately.

[Franz Richter]

Dear Mark,

At Imperial College we recently used with success Gpyro to simulate the burning of wood. Specifically for this work, we have compared the kinetics of different wood models (Bioresource Technology, 2020) and then developed a novel kinetic model for softwood (JAAP, 2019). This model was then used to predict both the Cone Calorimeter test (CNF, 2020) and Fire Resistance test (Fire Tech, 2020) across a range of conditions. 

I believe the papers on kinetics might be of interest to you.

Papers:

Bio Tech: https://www.sciencedirect.com/science/article/abs/pii/S0960852419318498

JAAP: https://www.sciencedirect.com/science/article/abs/pii/S0165237018307514

CNF: https://www.sciencedirect.com/science/article/pii/S0010218020300924

Fire Tech: https://link.springer.com/article/10.1007/s10694-020-01000-1

Happy to provide a free copy of each article if needed .

Franz

[tristan...@gmail.com]

On the README files: Sure, I'll provide an outline. It might be useful to create an individual conversation in this GoogleGroup, for this proposal. I wanted to create one when I would have processed more of the repository, but if you think it would be a good idea to have it open while it is still in a work-in-progess stage I'm happy with that. Might also be helpful such that people are aware and could comment on the progess.

In this proposal efforts are made to unify the README files. The overall goal is to have these files complete, human-readable and parsable by scripts for automatic data processing.

For each experiment a complete list of submitted parameters was compiled and is provided as a template file, such that further submissions can just fill their information in. The formatting was then backpropagated to all submitted REAME files, thus there is no action necessary be the original contributors. For now, not all experiments are available, but a couple are ready (DSC, TGA, Cone and Gasification/CAPA II): https://github.com/TristanHehnen/matl-db/tree/exp_info_dict/Utilities/Templates

Furthermore, some scripts and tools are to be provided with the repository, that allow for automatic processing of the contained information, since in my vision this repository will grow over time. Within these tools, it would be nice to also provide the means of data processing that are agreed upon by the MaCFP group. These tools would all be supplementary, such that everyone can process the data how ever they like, while having direct access to the "standard way". One outcome of this could be to automatically create these reports, like the one that was presented last year by Isacc.

Since I only know/use Python, all the tools are provided using this language. However, if others would like to replicate the functionality in say MatLab, the unified README files would certainly be a useful starting point as well.

Furthermore, we are looking into using GoogleSheets to streamline the data subission process. This would revolve around the aforementioned README templates. People could enter their experiment data in the GoogleSheet and download the files in the correct layout/format and would then submit them to the repository. However, it would only be made available for established experiments. If a new experiment is added to the mix, I would suggest to have people at first submit whatever information they consider relevant, like it whas handeled before. Afterwards, a complete list could be compiled and templates created, as described above.

Best wishes,

Tristan

[Morgan Bruns]

Tristan,

This is all great. I'd suggest maybe starting two separate threads here to discuss (1) suggestions on the content and format of the README files, and (2) suggestions on what we want the processing scripts to do. Of course, I expect this will all evolve over time as we learn more. 

We'll need to do something similar on the modeling side of things as well. I'm pretty busy for the next month or so, but I should be able to start helping out more this summer.

Morgan

[tristan...@gmail.com]

Dear Morgan,

my hope is that we can use the basic formatting strategy of the README files for both, the experiment and the simulation data sets. Then the scripts could look at the context and to group the information accordingly. In principle, both types of README should mostly contain similar information.

Tristan

[Morgan Bruns]

Hi, Tristan. I agree that we should be as consistent as possible in the formatting. I think we will actually end up having three types of data sets:  experiment, model parameters, and model predictions. The specific metadata (the content of the README files) for all three of these data set types will be significantly different, and I think now would be a good time to get input from the community on what metadata needs to be included. 

It looks like you have a very good start on the metadata for the experiments, but it would be helpful if all of the experimentalists submitting results would check and confirm that this metadata is a sufficient description of their experiments. I know that we had some discussion about requiring instrument uncertainty and calibration data so that users could have more confidence in the data set.

The model parameters README file should include at least a list of the underlying experimental data sets, the models used for parameter calibration, and the calibration (fitting) method used.

The simulation prediction README file should include a list of the values of all of the model parameters (there might be some standardization work required here in terms of what we call and how we label variables), the sources of all of the model parameters, the model used (e.g., FDS, Gpyro, ThermaKin, etc.), and as many specifics about the geometry, initial conditions, and boundary conditions as possible. Maybe we will end up requiring modelers to include their input file as well.

Those are initial thoughts, but it would be nice to flesh out the details as publicly as possible.

Morgan

[tristan...@gmail.com]

Hi Morgan,

for the model parameters I'll build a first iteration for the README structure, based on our contribution and incorporate the points you've mentioned above. 

I guess, when the simulation layout has had a few iterations we could relatively easily build a prediction README out of it.

Tristan