MaCFP-2 Condensed Phase workshop notes

[Randy McDermott]

Isaac, Stas, Morgan,

First, very nice job!  Tough problem and you did an awesome job of organizing everything.  I wanted to send you my thoughts while they are fresh on my mind.

1. Regarding verification (data quality).  Simo and Mark seem to think that the particular material used for this first exercise will not ever be available again.  I wonder if this is actually an opportunity for someone to develop a standard reference material (this is seems like perfect job for NIST).  I'm thinking that something like the NIST "standard cigarette" could be developed as a reference for calibration of TGA, etc.  I'm keen to get beyond the mindset that things are "carefully calibrated".  I realize at this point it is a bit of a "chicken and egg" problem, but we need a good reference it seems.

2. I'm glad there was discuss of, as Morgan put it, the "extremes" of one "true" set of parameters versus optimized, model-specific parameters.  The criticism of the optimization approach usually is to claim that unphysical values might be given by the algorithm.  On the other hand, everyone acknowledges that when there are significant geometry changes, all the parameters change to effective values.  Can't the optimization algorithms be constrained to physically reasonable ranges for the parameters like density, conductivity, etc.?

3. In the gas phase workshop one of the themes emerging is the importance of correct characterization of soot and radiation (no surprise).  Inevitably, this will send us down the road of soot modeling, which will mean we need to know the pyrolyzate gas and what the in-flame soot yields are.  It sounds to me like we need to take the PMMA and do extractive sampling of soot in the flame so that we can calibrate our combustion models.  I'm not sure which subgroup should take responsibility for this.  Anthony Hamins and Ryan Falkenstein-Smith have things set up (at least they did at one point) to measure soot for gas burners.  I wonder if we could so something for PMMA flames.

4. Something else that we seem to agree on in the gas phase side is that we have been going too long without more frequent communication.  To this end, would there be harm in using this discussion forum for both condensed phase and gas phase?  It think it would just be a matter of changing the name.   What do you say?

Thanks,

Randy

[Randy McDermott]

Isaac,

Thanks for your answers.  Anthony mentioned to me that he has already studied MMA pool fires (Hamins et al CST 1994).  I will look at that study and see what else we are missing in the gas phase.  But, beyond PMMA, I think that the pyrolyzate is going to be an issue for general coupling between solid and gas phase, and we need to start thinking about this.

Cheers,

Randy

On Fri, Apr 23, 2021 at 1:12 PM Isaac Leventon <isaac.l...@gmail.com> wrote:

Hi Randy,

Thanks for the salad list of questions. I think I can answer a couple of them.

1. The concern about no longer having this exact material comes from our challenges is in reordering it about 6 months ago. Because of the tariffs that were put in place in the last 4 years, our distributor was no longer able to import it. I searched far and wide and it seemed like these were the last samples we would have. In a more recent email with one of their affiliates over in Tennessee, I may have tracked down what is nominally the same material but with a new trade name. We would have to confirm it is exactly the same, The company said that they had done so, but we should be able to check that with some small scale testing.

As to the creation of a standard reference material - I think that might be something that NIST would actually support and allow us to focus some more of our time on this project. Ignoring the challenge of the SRM process, it's not always that simple to do at a material level:, we would need to agree upon a set of key metrics that we could measure in different types of experiments so that we could say with confidence that this is the exact same material. Maybe that's heat of combustion at this or that scale, onset of degradation, peak mass loss rates under given heating conditions and the temperature that occurs at, measured heat flow... As an example, for MCC calibration, PMMA is actually not a great reference as there is a lot of variability between 'brands' or types of acrylic, but polystyrene was a better fit. If we want to ensure a 10-year supply of this or that material, we may need to be mindful of such challenges.

Even without a new standard reference, there are clearly some simple instrument calibration and baseline corrections that are not being made consistently across groups. At some point we'll need to be able to reject data sets that are submitted that don't meet those criteria. We need to agree on those criteria, and that, may require some healthy discussion. This year, I think we took most anything that we could get just for the sake of encouraging participation.

2. My understanding of most higher order optimization algorithms is they generally do (or at the very least can) constrain values to within certain ranges. What exactly is physically reasonable, that can be up for debate. Whatever the modeling or optimization approach, I think we demonstrated today that (at least some of) the optimization targets that different groups are currently using to say their model " fits "  or is "calibrated" are not good enough because we see a factor of two difference when we are predicting the exact same idealized gasification case (a relatively simple problem). 

Also, However people want to tune their models, I want to make sure we do so only with bench scale data and then separately validate for real burning simulations - cone calorimeter or large-scale flame spread - it seems like there were some suggestions that we need full scale or flame spread data to calibrate these parameters, and if that's the case then we are lost because we will never be able to get all that data for all of our materials. There was a.. vibrant email exchange earlier in the year about needing that validation data before those simple predictions could really be given.

3. I'm happy to bring in Ryan and Anthony if they would like to help the problem and if NIST will support it, I simply don't know enough about how to conduct those measurements myself. For the non-steady cases that we were looking at, even for this PMMA, which doesn't really drip or flow, it is a much more challenging set up to instrument and actually measure than the gaseous wall burner. Even just these flame heat flux measurements, there was a real strong push back against getting the time to take those (let alone calibrate...) and soot measurements would be even more challenging. Just a practical concern to consider.

4. I'm happy to merge that for both groups if we can stimulate better discussion. A simple name change to make it the MaCFP discussion forum would be fine for me. We might want to include a common header or standard format to post titles to help disambiguate which group a post belongs to? Maybe that's a non issue though. If there are no objections from anyone else in the group, I'll go ahead and make that name change.

-Isaac

--
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/86edd2f9-c3e9-4b85-8502-d24133fd78c2n%40googlegroups.com.

[isaac.l...@gmail.com]

Hi Randy,

Thanks for the solid list of questions. I think I can answer a couple of them.

1. The concern about no longer having this exact material comes from our challenges is in reordering it about 6 months ago. Because of the tariffs that were put in place in the last 4 years, our distributor was no longer able to import it. I searched far and wide and it seemed like these were the last samples we would have. In a more recent email with one of their affiliates over in Tennessee, I may have tracked down what is nominally the same material but with a new trade name. We would have to confirm it is exactly the same, The company said that they had done so, but we should be able to check that with some small scale testing.

As to the creation of a standard reference material - I think that might be something that NIST would actually support and allow us to focus some more of our time on this project. Ignoring the challenge of the SRM process, it's not always that simple to do at a material level:, we would need to agree upon a set of key metrics that we could measure in different types of experiments so that we could say with confidence that this is the exact same material. Maybe that's heat of combustion at this or that scale, onset of degradation, peak mass loss rates under given heating conditions and the temperature that occurs at, measured heat flow... As an example, for MCC calibration, PMMA is actually not a great reference as there is a lot of variability between 'brands' or types of acrylic, but polystyrene was a better fit. If we want to ensure a 10-year supply of this or that material, we may need to be mindful of such challenges.

Even without a new standard reference, there are clearly some simple instrument calibration and baseline corrections that are not being made consistently across groups. At some point we'll need to be able to reject data sets that are submitted that don't meet those criteria. We need to agree on those criteria, and that, may require some healthy discussion. This year, I think we took most anything that we could get for the sake of encouraging participation.

2. My understanding of most higher order optimization algorithms is they generally do (or at the very least can) constrain values to within certain ranges. What exactly is physically reasonable, that can be up for debate. Whatever the modeling or optimization approach, I think we demonstrated today that (at least some of) the optimization targets that different groups are currently using to say their model " fits "  or is "calibrated" are not good enough because we see a factor of two difference (in at least some key model outputs; e.g., time to mass loss or peak MLR) when we are predicting the exact same idealized gasification case (a relatively simple problem). 

Also, however people want to tune their models, I want to make sure we do so only with bench scale data and then separately validate for real burning simulations - cone calorimeter or large-scale flame spread - it seems like there were some suggestions that we need full scale or flame spread data to calibrate these parameters, and if that's the case then we are lost because we will never be able to get all that data for all of our materials. There was a.. vibrant email exchange earlier in the year about needing that validation data before those simple predictions could really be given.

3. I'm happy to bring in Ryan and Anthony if they would like to help the problem and if NIST will support it, I simply don't know enough about how to conduct those measurements myself. For the non-steady cases that we were looking at, even for this PMMA, which doesn't really drip or flow, it is a much more challenging set up to instrument and actually measure than the gaseous wall burner. Even just these flame heat flux measurements, there was strong push back against getting the time to take those (let alone calibrate...) and soot measurements would be even more challenging. Just a practical concern to consider.

4. I'm happy to merge that for both groups if we can stimulate better discussion. A simple name change to make it the MaCFP discussion forum would be fine for me. We might want to include a common header or standard format to post titles to help disambiguate which group a post belongs to? Maybe that's a non issue though. If there are no objections from anyone else in the group, I'll go ahead and make that name change.

-Isaac

[Stanislav I. Stoliarov]

Randy,

I like the idea of expanding the discussion forum to include the gas phase.  The 900 nm imaging that we performed on the corner experiments can be used as a target for soot model validation.  We can also try to do extractive soot sampling in a few discrete locations, especially if Anthony teaches us how to do it correctly.

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

--

[Morgan Bruns]

Thanks for the comments, Randy.

1. I'll second (or third) the idea of a NIST standard reference material. There probably needs to be some thought about the characteristics of the material which will depend on which experiments we want it to be a reference for---I don't know this, but one material might be a better reference for TGA, and another might be a better reference for the cone.

2. I would imagine that most people using optimization for fire properties are constraining their parameter space to some extent. We certainly didn't see any clearly "unphysical" properties in this round. I was very happy to see that the BoWFZJ group got so close to the measured density. My concern is more with the optimization algorithms being fooled by compensation effects such that the resultant material parameters don't perform well in scenarios very dissimilar from the calibration experiments. I have no idea whether this concern is valid or not. It's just my intuition. Ultimately, I think exercises like MaCFP will help to settle that question.

3. Yes, I think you're correct about the need to characterize the composition of the pyrolysis gas. Some groups are doing this using FTIR and GC-MS, and it would be great if we could get them involved. As Stas said, for PMMA, people think you mostly get MMA. I imagine that one challenge for other materials will be that the composition will depend pretty heavily on heating rate and oxygen concentration.

4. Yes, we should merge gas and condensed phase discussions. Is there a gas phase Google group?

Morgan

[Arnaud C. Trouve]

To all,

Regarding the first point and the definition of a standard reference material, I remember a workshop in France between fire safety researchers and material science researchers devoted to that topic and that as far as I know went nowhere so this may be a difficult and risky proposition.

Regarding the fourth point, there is no gas phase Google group.

Arnaud

[Randy McDermott]

There is not a gas phase discussion group.  So, it would be a simple matter of changing the name and providing a link from the gas phase repo front page to this forum.  I really think this would be great.  Thanks for the positive response!

You received this message because you are subscribed to a topic in the Google Groups "MaCFP Condensed Phase Discussions" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/macfp-condensed-phase-discussions/LganRxqEjSM/unsubscribe.
To unsubscribe from this group and all its topics, 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/16714944-3a93-44ce-a092-5964f2d4f38an%40googlegroups.com.

[rogaume]

Dear all of you

Thank you for all your very interesting remarks, demonstrating that thermal decomposition study and modelling is a very complex work

Several things please:

-        Thank you for the proposition of those materials Mark. As said by Arnaud, there was in the past two efforts to identify and to define a “perfect and standard” material into the French community, some preliminary tests have been performed and there was no success.

-        The variation of the results can be due to several reasons: the protocol used for the TGA tests (as the sample geometry and size) ; the measurements of the thermos-physical properties remain very difficult and a scientific lock ; the inverse method of optimization of the kinetic parameters, are underlined by Morgan as well as the comparison law used for the optimization…

-        To take into account the gas emitted during the thermal decomposition process remain very important and give very interesting information in order to develop and propose the mechanism of thermal decomposition. However, both considering TGA + gas measurements, it is classically possible to propose several mechanisms… Then pas studies have shown the challenge of considering other parameters at larger scale in order to identify the most appropriate mechanism

-        The consideration of the gas phase is a logical but hard challenge. Some works began in this sense…

Best regards

Thomas

********************************************************

ROGAUME Thomas

Professeur

Institut des Risques Industriels Assurantiels et Financiers

Institut Pprime (UPR 3346 CNRS)

Université de Poitiers

 

 

 

 

 

De : macfp-condensed-...@googlegroups.com [mailto:macfp-condensed-...@googlegroups.com] De la part de Morgan Bruns
Envoyé : vendredi 23 avril 2021 20:52
À : MaCFP Condensed Phase Discussions <macfp-condensed-...@googlegroups.com>
Objet : Re: MaCFP-2 Condensed Phase workshop notes

.
To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-condensed-phase-discussions/16714944-3a93-44ce-a092-5964f2d4f38an%40googlegroups.com.

[Guillermo Rein]

Hello to all,

First of all, I  would like to reiterate my congratulations to MaCFP for a great job organizing and running this very successful workshop and the resulting datasets.

In this message, I wanted to share a few comments regarding the discussion about the *material of choice* for future pyrolysis studies in MaCFP Condensed Phase. There has been comments about previous efforts, in France, to find this material. Can these efforts be summarised or shared somehow? It is difficult to built on previous efforts otherwise. Unfortunately for me, I was not present during those meetings.

I was thinking what would be the top characteristics of such material, or group of materials, that highlight them to be specially good for inter comparisons of measurements and modelling of pyrolysis. I thought of these general characteristics as a preliminary:

1) Relevant: The material shall be of relevance to modern fire safety.

2) Long term: The material shall be available in affordable and repeatable form, in sufficient amount, at the time of the studies and afterwards with a view of decades. 

3) Range of complexities: The group of materials shall include a range of complexities in pyrolysis behaviour, from low to high. The premise is that material of simpler pyrolysis behaviour should be easier to understand, measure and model. Materials of higher complexity should be given priority only after the simpler once have been studied. 

...

n) Many more characteristics.

I wonder if this group is interested in organizing a series of discussions (online for the time being) to agree on a list of characteristics.

Cheers,

Guillermo Rein, London

[Lukas Arnold]

Dear all, dear Guillermo,

I would like to support Guillermo's proposal to reevaluate why the previous efforts on common material(s) failed. Although, I liked the discussion last week, I think we should meet (virtually) in a small round with the active institutions on a regular basis. In my opinion this topic and the MaCFP effort is too pressing to be discussed only once a few years.

[Having said that, I'm happy to offer the organisation of such meetings, if we see benefit in these meetings.]

Best,
Lukas Arnold

> To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-condensed-phase-discussions/2da3fae3-990c-423a-ad9a-fc7fb2044dean%40googlegroups.com.

[Morgan Bruns]

I agree that periodic virtual meetings are a good idea. Some of this discussion points towards smaller groups focusing on particular topics. We have to be careful to not overly complicate things, but I don't think there's anything wrong with some informal smaller efforts focused on specific aspects of the fire modeling problem.

So far the condensed phase MaCFP group has had three main threads:

(1) Small-scale experiments for obtaining model parameters;

(2) Calibration and model parameters; and

(3) Model development, validation, and model predictions.

Would it be worthwhile to organize informal subgroups focused on these topics? 

Several other topics seem to have gathered interest during and after the workshop:

* methods for verifying the quality of calibration data

* characterization of uncertainty

* model sensitivity analysis

* development of a standard reference material or set of reference materials

* specification of appropriate metadata for experiments, model parameters, and model predictions

Are there others? My hope would be to have some informal groups develop within this forum that would attack these various topics. If anyone reading this thread would be willing to step up and take the lead on organizing some collaboration on one of these topics, I'd be happy to support them in whatever ways I can.

Morgan

[Stanislav I. Stoliarov]

I am not sure if we should further divide to different subgroups.  Rather each "in-between" meeting can be dedicated to only one of these specific topics.

Stas 

[rogaume]

Dear all

I have the same opinion that Stas, please do not decompose under subgroups!

Thermal decomposition and the development of models of pyrolysis is a complex challenge which have to be treated in its whole, with a scaling up approach and taking into account the application. It is important that the experts of each scale-bench scale or modelling aspects focused their attention on their part but I think the rule of the condensed phase group is to have the global approach

Best regards

Thomas

 

 

De : macfp-di...@googlegroups.com [mailto:macfp-di...@googlegroups.com] De la part de Stanislav I. Stoliarov
Envoyé : mercredi 28 avril 2021 02:00
À : MaCFP Discussions <macfp-di...@googlegroups.com>

--
You received this message because you are subscribed to the Google Groups "MaCFP Discussions" group.
To unsubscribe from this group and stop receiving emails from it, send an email to macfp-discussi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/e4c76450-8674-4f95-acde-bb8eed308682n%40googlegroups.com.

[Brännström, Fabian]

Dear all,

first of all, thank you all for the effort you put into this! It is very

interesting!

A few comments from us:

1) Virtual Meetings

   - More frequent virtual meetings for gas and/or condensed group (and maybe

     radition group) with special topics would be very helpful. Just like Stas

     mentions a split up might not be needed.

2) Standard Material

   - Aiming for a standard material would be great; the points from Guillermo

     are very good and obviously I see the difficulty to find a specific type 

     for the community.

3) Boundary conditions of system

   - As it looks to me, a better understanding of boundary conditions of the

     systems are needed. So similar like for the standard material a way forward

     might be to set up a standard DSC/TGA/MCC system for which the groups get

     more details about the geometry, known (and unknown ;-) ) boundary

     conditions. It could start with a detailed CAD model and further

     temperature measurements inside the systems.

4) Senstivities/Uncertainties

   - This is in my view essential and should be not only limited to the modeling part

5) Optimization - comparing of algorithmns:

   - In addition to the work done it would be interesting to see the performance

     of the different algorithms and selected (or defined) objective functions

     for a predefined (physical!?) range of the selected parameters; performance e.g. in

     terms of convergence profile for multiple repeated runs (if needed)

Have a nice conference!

Best Regards

Fabian

-- 
Univ.-Prof. Dr.-Ing. Fabian Brännström

Brandtechnologie und Brandschutzingenieurwesen
Fire Technology and Fire Safety Engineering

+49 (0) 202 439-2071

Raum W08.093

https://fire.uni-wuppertal.de/

Bergische Universität Wuppertal
Fakultät für Maschinenbau und Sicherheitstechnik
Gaußstraße 20
42119 Wuppertal

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/701181935.3934536.1619592969607.JavaMail.zimbra%40zimbra-store-8.ldbal.sci.univ-poitiers.fr.

[Hostikka Simo]

Dear all,

 

I also lean on keeping the discussions under the same umbrella.

 

Regarding the benchmark material: I wonder if it could be something that can be manufactured in laboratories. Relying on industrial supplier(s) is a challenge for continuity. I have seen different devices in polymer and composite labs, being used to make samples from simple base ingredients (usually some basic polymer, though) but don’t understand them well enough to know if collaborating with such labs would help here.

 

Simo

 

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/17989835-CF38-4D46-821F-88C4C4D66C40%40uni-wuppertal.de.

[Morgan Bruns]

I agree that we shouldn't start formal subgroups for the reasons Thomas outlined above. I also like Stas' idea of establishing a tradition of holding semi-annual virtual meetings to focus on one of the topics of interest to everyone. I do think it will be necessary for people to step-up to lead or actively contribute to these subtopics (this is kind of what I was getting at by using the phrase "informal subgroup"). In this spirit, I started a separate thread on the development of a standard reference material. I encourage others to do something similar for topics that interest them.

Morgan

[Stanislav I. Stoliarov]

Franz,

It looks like you have posted copyrighted material.  This is not allowed.  You can post a link to your article but not the article itself.

Stas

On Wednesday, April 28, 2021 at 10:56:32 AM UTC-4 franz....@berkeley.edu wrote:

Dear all,

at Imperial College, we recently investigated Point 4 at the microscale. We found that the optimization had little influence on the kinetics for simple fuels, but mattered for complex fuels.

For a fuel like PMMA, we would expect an insignificant variation in the results due to the optimization algorithm and objective function used.

Paper: https://www.sciencedirect.com/science/article/pii/S0016236119316059

The Paper is also attached

Best wishes,

Franz Richter

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

Franz Richter, Ph.D.

Department of Mechanical Engineering

University of California, Berkeley

Email: franz....@berkeley.edu

LinkedIn: www.linkedin.com/in/franz-richter

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/17989835-CF38-4D46-821F-88C4C4D66C40%40uni-wuppertal.de.

[Franz Richter]

Dear all,

at Imperial College, we recently investigated Point 4 at the microscale. We found that the optimization had little influence on the kinetics for simple fuels, but mattered for complex fuels.

For a fuel like PMMA, we would expect an insignificant variation in the results due to the optimization algorithm and objective function used.

Paper: https://www.sciencedirect.com/science/article/pii/S0016236119316059

The Paper is also attached

Best wishes,

Franz Richter

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

Franz Richter, Ph.D.

Department of Mechanical Engineering

University of California, Berkeley

Email: franz....@berkeley.edu

LinkedIn: www.linkedin.com/in/franz-richter

Best wishes,

Franz Richter

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

Franz Richter, Ph.D.

Department of Mechanical Engineering

University of California, Berkeley

Email: franz....@berkeley.edu

LinkedIn: www.linkedin.com/in/franz-richter

On Wed, Apr 28, 2021 at 2:43 AM Brännström, Fabian <braenn...@uni-wuppertal.de> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/17989835-CF38-4D46-821F-88C4C4D66C40%40uni-wuppertal.de.

[Guillermo Rein]

Franz, Morgan,

Stas is right. Copyrighted papers (= not gold OA or government papers) cannot be distributed in email groups like google groups and others. Authors can share pdf copies of their own papers on a one-to-one basis and not for group distribution.

Cheers,

G.

[Franz Richter]

I deleted the message and reposted it without the paper attached. So the group should be okay now.

Best wishes,

Franz Richter

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

Franz Richter, Ph.D.

Department of Mechanical Engineering

University of California, Berkeley

Email: franz....@berkeley.edu

LinkedIn: www.linkedin.com/in/franz-richter

To view this discussion on the web visit https://groups.google.com/d/msgid/macfp-discussions/159242cf-5f88-4c65-8c35-f7f6d838ef6en%40googlegroups.com.