Thunderhead Engineering Forum

Please login or register.

Login with username, password and session length
Advanced search  

News:

Forum moved to https://forum.thunderheadeng.com

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - Dave McGill

Pages: 1 2 [3] 4 5 ... 9
31
PyroSim / Re: Difficulty modelling sprinkler actuation times
« on: June 07, 2015, 05:39:35 am »
Hi,

At the moment, they are positioned at the upper surface of the domain. If you lower them (I suggest at least 1/2 cell dimension) they will appear. As you mentioned, the domain has no open vents, so you won't have a 10 MW fire for very long.

Dave

32
PyroSim / Re: No smoke occurs
« on: May 06, 2015, 07:52:22 pm »
LoL, I didn't even have to break a sweat to solve that one.    :)

33
PyroSim / Re: No smoke occurs
« on: May 05, 2015, 03:12:44 pm »
Hi Max,

Start by checking the _hrr.csv file to ensure you have a fire.

Dave

34
PyroSim / Re: Pyrosim - detector activation
« on: April 08, 2015, 04:43:54 pm »
Hi Michael,

I'm not sure it is. I suspect that it could be duplicated using the custom control feature of PyroSim, but I haven't tried. I was playing with it a few weeks ago, and took the easy route. I put it in the Record section. I also figured out how to have the HRR freeze when one of several sprinklers activate.  I've attached both files.

Dave

35
PyroSim / Re: Pyrosim - detector activation
« on: April 06, 2015, 06:24:35 pm »
try lowering the heat detectors.

36
PyroSim / Re: Pyrosim - detector activation
« on: April 06, 2015, 06:02:28 pm »
Hi Michael,

Read the section on E_COEFFICIENT in the FDS Users Guide.

At the moment, the value of E_COEFFICIENT for your burner is 0, so there is no chance on the sprinkler suppressing the burner. Even with a high value, it won't suppress it completely.

Read this thread. http://www.thunderheadeng.net/forum/index.php/topic,2605.0.html

The other detectors may not be activating due to the cooling effect of the initial sprinkler. Look in the _devc.csv file and check to see what is happening to their temperatures.

Dave

Dave

37
PyroSim / Re: No reaction occuring in smokeview
« on: April 04, 2015, 09:42:38 am »
You're welcome.

I thought the problem was the mis-match between the dimensions as specified in the model, and the grid. Smokeview will build everything according to the grid dimensions, which is why I encourage the use of nice even cells, such as 0.1 m a side, and all dimensions in the model being divisible by 0.1 m.

Dave


38
PyroSim / Re: No reaction occuring in smokeview
« on: April 03, 2015, 12:29:49 pm »
Patience Young Grasshopper  :-)

Let room_fire run to completion. I've attached the input files for both room_fire.fds and couch.fds. I've added comments which should help you understand how each is set up.

For your own model, I suggest you start with,
- a fire vent on the floor of an empty domain with OPEN vents; if that works then
- move the fire onto an obstruction; if that works then,
- add the rest of the obstructions.

Dave

39
PyroSim / Re: No reaction occuring in smokeview
« on: April 03, 2015, 10:49:23 am »
I just wanted to test the burner, so I took out everything, put a burner vent on the bottom of the domain, and changed to polyurethane as a reaction.

40
PyroSim / Re: No reaction occuring in smokeview
« on: April 02, 2015, 07:13:51 pm »
Hi Max,

There are a number of problems.
1. the cells are an odd size, and the obstructions and vents don't line up with the cell divisions
2. there are no OPEN vents, so it is a sealed box
3. something is wrong with your custom reaction (which I haven't sorted out)

Have a look at test2.psm (attached). It is a fire vent on the floor of the domain. This demonstrates that a burner with the default reaction will work. Before trying to fix your model, I suggest you run couch.fds and room_fire.fds from the fds6 examples folder. Open the .fds files and see how they have been constructed.

What I would do to fix your model is,
1. set your cells to 0.1 m a side, and realign the meshes
2. align the edges of all features to the new cell divisions
3. add OPEN vents to the domain boundaries, where appropriate
4. once you get a fire going in that set-up, go back and trouble-shoot your custom reaction

Dave

41
PyroSim / Re: No reaction occuring in smokeview
« on: April 02, 2015, 09:51:03 am »
Hi,

Can you attach your .psm file?

Dave

42
PyroSim / Re: Pyrosim - detector activation
« on: April 02, 2015, 07:52:13 am »
Hi Michael,

The sprinklers were essentially embedded in the ceiling slab. I lowered them all by 0.2 m. The trick is to preview the model in Smokeview and look at the proximity of the sprinklers to adjacent surfaces. Rotate the model in Smokeview and ensure that you can see the entire sprinkler, and that there is a visible gap between the sprinkler and the adjacent surface.

Here are a couple of other suggestions.
  • Change the alpha value of the colour of your glazing to `0.3. That way, it will still look like glass, but you will be able to see the smoke filling the room without having to clip the geometry.
  • I assume you have a quad-core. If you split the apartment mesh in two, the simulation will probably run close to twice as fast.

Dave

43
PyroSim / Re: sprinkler and hss
« on: April 02, 2015, 06:15:46 am »
Hi Tricia,

Short Answer
This feature doesn't work very well, has no predictive value and isn't worth spending time on.

Long Answer
When I was teaching, I had my students do a model using this feature. Over the years, I've looked at several hundred examples. Sometimes the fire is diminished a  lot, sometimes only a little. It was rarely put out completely. There is no predictable pattern. The E_COEEFICIENT has to be determined experimentally, and, to my knowledge, no one has ever done that. So, it is an unvalidated feature, with no predictive value.

I compared your example to the e_coefficient example provided with FDS. There are several differences that cause the fire to not decline in your example.
  • the HRR is 5x greater
  • the flow rate is 1/10th
  • the spray pattern is almost horizontal, and most of the water leaves the domain
I think this combination of factors explains why the fire is not extinguished in your example.

If you want to approximate the effect of sprinklers on a fire, then there are several other options.
  • The hrr_freeze feature will hold the HRR steady when a device, like a sprinnkler activates. This is a common conservative strategy.
  • The old way was to let the fire grow, note when a sprinkler activated, stop the model and adjust the HRR via RAMP to stop growing, or decline at the time of sprinkler activation, and re-run the model

Dave

44
PyroSim / Re: Pyrosim - detector activation
« on: March 31, 2015, 03:44:28 pm »
have you configured them to discharge an appropriate amount of water?

It would help if you attached your file.

45
PyroSim / Re: Pyrosim - detector activation
« on: March 30, 2015, 08:30:37 am »
Hi Michael,

There are two possibilities.

1. You have selected sprinkler links instead of sprinklers. The former just places a link that will indicate if the sprinkler reaches its activation temperature. It does not discharge any water.

2. You have to specify the extinguishing coefficient for the surfaces in your model Search for E_COEFFICIENT  in the FDS Users Guide. This is an unvalidated feature, and you will be forcing the fires to diminish, rather than modelling it based upon the material/surface properties.

Dave

Pages: 1 2 [3] 4 5 ... 9