Need Help - Getting error after any small change to input

[Vera-MVD]

Hi.

I am getting this error when I am trying to run a CFAST model:
CFAST: CFAST exited with a non-zero exit code: 408
CFAST: forrtl: severe (408): fort: (10): Subscript #2 of the array DUMPCARRAY has value 2531 which is greater than the upper bound of 2530
For the full error message please see the screenshots attached.

Some background: Ive never had this error before. I have not run CFAST code for the last almost three months. When I open my CFAST model (that was working just fine before) I got this error. First, I thought maybe I did some changes last time I run it and somehow messed it up. So, I opened another model, tried to run - same error. Tried running some of the prebuilt cfast examples (Simple, User Guide Example, and Multifloor); with those first time I opened them, they run fine, but as soon as I tried to make a small change (literally change the simulation time by 1 seconds) - same error at running. Change the time back to what it originally was, saved the model - still same error. Closed the file completely, closed CFAST, open everything again, it took almost a minute for CEDIT to open the Multifloor model now (it was just not responding to anything while opening it), but at the end, still same error when tried to run.

After that I just cloned the CFAST source code again, and started afresh. Still the same. I can run a model once, but after any change (no matter how big or small) I am getting this error. It looks as if that subscript is not refreshing after the run is over.

And just to be clear: I am not touching the source code files when making the changes to the models. I am attaching screenshots of

1 - Multifloor.in model original simulation screen

2 - Multifloor.in model the results screen (right after it run) no changes yet

3. • Multifloor.in model change on simulation screen (note that simulation time was changed from 1000 s to 1001 s - the only change made). The model was saved after that.

4. • Multifloor.in model error after clicking on the "Run" button.

And after that even if I close and open the model again, the error is there:

Anybody has any ideas of why it suddenly decided not to run? There were a few automatic updates to my system over the last few months (they happen periodically, I am not sure what exactly was updated, but CFAST definitely was not touched). I am running Windows 11 Enterprise, and Visual Studio Enterprise 2022 (64-bit).

Thank you in advance for any help,
Vera

[mcgratta]

Could you upload the .in input file.

[Vera-MVD]

Thank you for joining in.
Forgive the question, but how do I run the file? I double clicked on the file, it did open some window for a split second and close it right away, too fast for me to see what was in there. The only other option I see is to right click on the file and run as administrator, and I do not have administrator's rights, I am just a user on this machine.

[rpeacoc]

All it does is copy a few files to their correct locations. I suspect it worked fine for you. If you look in the root of the repository, there should be a file CFAST.sln. If you double click on it, it should open the CFAST solution in Visual Studio. You will likely get errors on CEdit since it requires additional software, bit the CFAST project should be ok.

[Vera-MVD]

I do see the CFAST.sln in the same folder as restore_vs_config.bat. The date both files show is yesterday when I cloned the repo. Will the running .bat file change update the date for .sln file? I tried to open the solution file by just double clicking on it, it opened Visual Studio but with the following message (see the screenshot)

And here is the output window:

[Vera-MVD]

I went into that folder and there is that script file:

[Vera-MVD]

Something is not right. Just noticed that Visual Studio is looking at the wrong place.

That scripts folder is in cfast\source\create_scripts\ with just two files shown in the previous screenshot.

But the Visual Studio is looking for them in cfast\Utilities\Visual_Studio\source\create_scripts.
Should I just copy the whole Source folder to inside the Visual_Studio folder or should I delete everything and just clone the repo again and start afresh?

[Vera-MVD]

It did not ask me for any option. When I open Visual Studio, there is an option to Clone a Repo

I get the path for the repo from this site

I just put it in the Visual Studio. I do specify the path where I want it to be cloned to.

After it cloned, I got a message yesterday:

I don't know if that affects anything. I just open solution file from Utilities/Visual Studio folder

And this solution I use to work with and then generate exe file.

[mcgratta]

Let's stop for a second and rewind. The process of compiling CFAST is complicated. There is little documentation. Is there a reason that you want to compile CFAST yourself? If you just want to run CFAST, you can simply install the package.

[mcgratta]

2525 looks suspiciously like 2531. It appears that there is an array bound being overshot; that is, a limit has been exceeded.

[Vera-MVD]

So that doesn't give me any hints. Is this the distributed version or a built one? Regardless, can you get a bundle at the CFAST website?
That was where I got it from, CFAST website, https://pages.nist.gov/cfast/. Since it was an exe file, I had to ask our IT people to install it on my computer. Then after looking closely at it, I needed to make some changes to the input parameters that were not possible through CEdit GUI (Like interior O2 mass fraction for instance). I did ask question about that here (setting up an inert environment), that was when I was told that in order to do that I would need to modify the source code files. Thus, Visual Studio was installed. For few months everything worked, until recently.

[Vera-MVD]

When I do this, I get an error statement like this

Error on line 2675 of the input file: &OUTP ID = 'Maximum Actual HRR'
***Error exit from CFAST, error 15 from routine read_dump
Error in &OUTP: Too many &OUTP entries. Maximum is2525  

That input file does seem a bit big for CFAST. Are you sure you need all that output? There are only so many quantities per zone.

I do not, I do not even know what they are I did not create this file. I just opened it from cfast\Examples\ folder and make one change (humidity) and the file was automatically turned into this monstrosity.

As for running from command prompt, you would have to explain to me where to find it first and what "cd to where the input file is" means. I do not use command prompt, so I am completely unaware of how to do this.

[preneke]

Error 15 in subroutine read_dump means you have put too many OUTP namelists in the file.

[Vera-MVD]

That's the problem, I did not put any. I don't even know what OUTP namelists are. I literally just changed humidity, hit save button, and then hit Run button. Got the error. And the .in file was like this after that. I did not add any namelists.

[preneke]

I run the case from the command line and it runs fine.

[Vera-MVD]

I run the case from the command line and it runs fine.

You ran the one I sent (the large one) or the original one that from the cfast package?

[preneke]

Original from the CFAST package. Did you make other changes? I missed that, sorry.

[Vera-MVD]

Which one did you run, Large_Building or Simple, I think I sent already both of them to here in form of a txt file? This is Large_Building.txt
Large Building.txt

[preneke]

You have too many OUTP namelists in the input file. The limit is 2525. That is a bit excessive.

[Vera-MVD]

As I said: "That's the problem, I did not put any. I don't even know what OUTP namelists are. I literally just changed humidity, hit save button, and then hit Run button. Got the error. And the .in file was like this after that. I did not add any namelists." How can I take them down from the file (using the CEdit GUI if possible)?

[preneke]

Sorry. However, the file you sent has a massive number of OUTP in the input file. I have no idea how you got them. Just open it in an editor and remove most of the OUTPs at the end file and the file will run.

[Vera-MVD]

Could you please tell me the line number after which I should delete everything, so that I do not mess it up even more? Thank you.

[marcosvanella]

Looks like you want to get rid of all OUTP lines and add:

&OUTP ID = 'Living Room Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Living Room', 'Sensor Activation' /
&OUTP ID = 'Dining Room Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Dining Room', 'Sensor Activation' /
&OUTP ID = 'Kitchen Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Kitchen', 'Sensor Activation' /
&OUTP ID = 'Bedroom 1 Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Bedroom 1', 'Sensor Activation' /
&OUTP ID = 'Bedroom 2 Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Bedroom 2', 'Sensor Activation' /
&OUTP ID = 'Bedroom 3 Alarm'
FILE = 'DEVICES' TYPE = 'TRIGGER_GREATER' CRITERION = 1
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'UL Alarm Bedroom 3', 'Sensor Activation' /
&OUTP ID = 'Time to Upper Layer 600 C'
FILE = 'COMPARTMENTS' TYPE = 'TRIGGER_GREATER' CRITERION = 600
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'Comp 1', 'Upper Layer Temperature' /
&OUTP ID = 'Actual HRR at Upper Layer 600 C'
FILE = 'COMPARTMENTS' TYPE = 'TRIGGER_GREATER' CRITERION = 600
FIRST_FIELD = 'Fire', 'HRR Actual' SECOND_FIELD = 'Comp 1', 'Upper Layer Temperature' /
&OUTP ID = 'Time to Layer Height 1.5 m'
FILE = 'COMPARTMENTS' TYPE = 'TRIGGER_LESSER' CRITERION = 1.5
FIRST_FIELD = 'Time', 'Simulation Time' SECOND_FIELD = 'Comp 1', 'Layer Height' /
&OUTP ID = 'Maximum Upper Layer Temp'
FILE = 'COMPARTMENTS' TYPE = 'MAXIMUM'
FIRST_FIELD = 'Comp 1', 'Upper Layer Temperature' /
&OUTP ID = 'Minimum Layer Height'
FILE = 'COMPARTMENTS' TYPE = 'MINIMUM'
FIRST_FIELD = 'Comp 1', 'Layer Height' /
&OUTP ID = 'Maximum Actual HRR'
FILE = 'COMPARTMENTS' TYPE = 'MAXIMUM'
FIRST_FIELD = 'Fire', 'HRR Actual' /

[mcgratta]

This is getting a bit complicated. Let's step back for a second. I suggest that you save any source code files that you have altered outside of the repository, and then delete and re-clone it. Then copy back in your changed files. Recompile however you were doing it before. But now when you run a case, I would edit the file with a text editor and run cfast from the command line. It is possible that you are somehow adding these extra lines to the input file via the CEdit interface. The CFAST User's Guide describes the format and parameters of the input file.

[Vera-MVD]

IT WORKS AGAIN!!! Thank you all so much for all your patience and advises/suggestions. I did go and edited manually Large_Building.in file deleting all the repeated OUTP entries and tried to run it from CEdit - and it worked. I then changed the humidity, saved and ran again. Fully expected to see the error but it did worked. I made couple more random changes after that, still worked. Then, replaced the CFAST.exe with the one I compiled and ran with it, and it still worked. After that I opened one of my models .in file. Saw that it also had huge number of repeated OUTP entries, deleted the extras. Opened it in CEdit GUI, ran and it also worked. Made few changes just to make sure, still works.

At this point it looks as everything works. I still wish I knew what happened to cause those extra lines in the .in files to appear. I've tried to recompile and clone the repo so many times over the last few days that I don't even remember which exe file version I used when the error happened first time. Lesson learnt.

Anyway, THANK YOU all again for all your help, back to work now!

Vera

[mcgratta]

You're a real hacker now.

[preneke]

I am glad your cases are working. I am wondering if you take a file with one "&outp .../" and do what you were doing with changing things and run several times. Then look at the .in file in a text editor and see if extra &outp entries. I am doubting it but it would be good to test it and best is start where it first might have happened. Thanks.

[Vera-MVD]

Oh, I've been doing exactly that since yesterday when the things finally started to work again. Every time I run now after making a change to the input settings, I am checking the .in file to see if the number of OUTP entries changed, so far no extra entries have appeared. If I ever find out why that did happen in the first place or see it happening again, I will sure let you guys know. As much as I am happy that it was an easy fix at the end, I am still curious what caused it, was it me (most probably) or some internal thing that is triggered at specific settings? Thank you for your help, much appreciated!

[Vera-MVD]

Well, it did not take long to see the OUTP block in the .in file starting multiplying again. Here is one example where I was able to reproduce the automatic duplication. Attached is the .in file (changed to .txt file) for a simple cubic room where I wanted to place a heater (aka electrical heater, no soot, no CO produced) in the middle of the room at 100 seconds into the 3600 seconds long simulation. When I first entered the heater as a fire with given constant HRR, I put it at 10 seconds. The input file (see attached Test_Heater_before.txt) was just 3K. And here is the screenshot of the fire tab:

Then I changed 9s to 99s and 10s to 100s wanting the heater to start at 100 seconds. Here is the fire tab screenshot:

then hit Save button. Checked the .in file. It was now 15K instead of 3K as before the change (see attached Test_Heater_after.txt). I opened it to see why it was so big and saw that 12 more identical blocks were added by the change.
Test_heater_before.txt
Test_heater_after.txt

I closed the model, opened the .in file, manually deleted the extra OUTP blocks, saved it. It went back to 3K size. I then opened it in CEdit, went to the Fire tab. It was showing the latest changes there (99 and 100 seconds). I changed 99 to 9s and 100 to 10s, saved the model. I did not run it. The .in file became 18K, upon opening I saw 15 extra OUTP blocks were added this time.

I have never use the Fire tab before, this is the very first time. So, whatever the issue was before with my .in files getting to 8M in size and stopping running was not because of the fire tab. Over the last couple of days I've been running a lot of cases (without fire) with changes being done on the Wall vent and Mechanical vent tabs, btw the changes were done again in the tables on those tabs. I've noticed the OUTP block being duplicated here and there but it never was more than 1-2 extra blocks appearing at a time. Since I was not able to reproduce the duplication before, I was just deleting the extra blocks when noticed and did not let you guys know.

I am not sure if I have something somewhere checked for this to happen, but I hope this will let you guys to figure out what is causing the OUTP block to multiply.

By the way, adding the heater this way did not work, the T did not change at all after placing the heater. What would be the right way to model a simple electrical heater with constant given output?

Thanks,
Vera