Fire Sprinkler simulation

hup_cheng · March 17, 2011, 02:18

Hi all,
I am very new to this Ansys CFX software. As the title, i want to simulate water distribution for fire sprinkler in a room. I have done through some tutorial in the example but i cant manage to get my simulation right.

i am using multiphase for domain combination of water and air. for the input, the fraction of water is 1 while for the air is 0. output is at the floor.

the streamline result only show velocity at the sprinkler but not reaching the floor of the room.

i hope anybody can enlighten me in this simulation. is there any important setting that i left out? if there is a tutorial close to this simulation, let me know.

thanx.

ghorrocks · March 18, 2011, 07:47

How are you modelling the multiple phases? Particle tracking or eularian?

hup_cheng · March 19, 2011, 02:06

i believe it is particle tracking..

ghorrocks · March 20, 2011, 21:55

Can you post an image of what you are seeing?

hup_cheng · March 21, 2011, 03:25

finally i solved the problem.. it is due to the initial condition of the domain..
but now my problem is the flow of the water is like bended.. its not falling in a straight line.. any idea what causes it?

ghorrocks · March 21, 2011, 20:20

Quote:

its not falling in a straight line.. any idea what causes it?

Umm, isn't that why you are doing a CFD model? Isn't it bending because the air flow bends it?

hup_cheng · March 22, 2011, 02:51

but the condition of the air speed has been set to 0.. plus i only want to study the design of sprinkler distribution cap neglecting the air movement..
im still exploring cfx to get better result presentation..
thanx alot..

ghorrocks · March 22, 2011, 07:55

The water droplet will drive a flow. If you want to disable the flow then use an expert parameter "solve fluids=f" then it will not update the fluid flow field.

nilambar · October 16, 2012, 04:41

Quote:

Originally Posted by hup_cheng

finally i solved the problem.. it is due to the initial condition of the domain..
but now my problem is the flow of the water is like bended.. its not falling in a straight line.. any idea what causes it?

very nice work. will you help me for simulation in tunnel fire. how i start in CFX ?
thanks with best regards

ghorrocks · October 16, 2012, 07:26

Do the tutorials, do some training, read the textbooks and do a literature search. And every once in a while put a post on the forum and we will try to help too.

nilambar · October 18, 2012, 01:05

plz suggest some text books..

ghorrocks · October 18, 2012, 06:49

Any text book with "introduction" and "CFD" is a good start.

Here's a few, there are many more: http://www.cfd-online.com/Books/brow...?category_id=1

nilambar · October 19, 2012, 03:54

thanks for suggestion...

pigei · July 30, 2019, 12:33

I'm new with CFX. Modelling a simple water sprinkler in air. Actually i used Lagrangian particle tracks but my simulation goes in error (floating) every time...
I've tried different solutions to solve this problem but without any effects (mesh, timescale, increase particle injection, decrease particle speed and flow rate, ecc).

I select no turbulence, isothermal and no inlet, so no speed, for air considered as countinous fluid.

Initial condition has been setted to zero for every parameter.

Geometry is actually small (about 3 x 2 x 2 m). Sprinkler inlet is modelled as a cilindrical surface.

Water speed is about 10 m/s and flow rate abou 10 kg/s.

Sprinkler direction is from top to bottom.

I want to start from this basic and easy system in order to study later the behaviour of a water "wall" generated by a sprinkler on a gas release.

I would compare particle track and fluid disperded different results.

Any idea on how to solve this simulation solver error? RMS and imbalance are so oscillating...

Svetlana · July 30, 2019, 21:12

Is this an unsteady simulation? Does it converge if you set the particle number to zero?

ghorrocks · July 30, 2019, 22:17

These are FAQs:
Floating point error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F

Poor convergence:
https://www.cfd-online.com/Wiki/Ansy...gence_criteria

pigei · July 31, 2019, 07:12

Thank you for your suggestion...

This is a steady simulation.

If i put 0 as number of positions for particle solver goes immediately in error.

I noticed that:
particle entered the domain: 10000
particle left domain: 9950
exceeded distance limit:50

I expect that particle will be accumulated on walls (floor). I've tried to reduce also speed and flow rate but nothing changed...

I made a simulation with only 1 particle and RMS still oscillating...less...

It is very strange because it a very simple simulation...no wind, no turbulence, isothermal...i want only to see the water! please help me...

pigei · July 31, 2019, 07:14

mesh is ok and fitted enough in my opinion...

on the sprinker inlet i have about 4mm size against 0.5mm particle diameter...is it ok?

pigei · July 31, 2019, 07:23

another strange things... particle seems not to go out normal to inlet surface but with casual angles...

pigei · August 1, 2019, 04:09

Hi,
in case of interest i found the solution of my problem...

It is related to particle speed.

I noticed that if the speed is too low < 30m/s i have oscillating on RMS and particle seems coming from inlet at casual angle.
If i put a speed > 40 m/s this problems are solved.
Speed are related also with particle diameter. With smaller diameter i have to increase speed.
Can anybody explain this strange behaviour?
Is there any equation for calculating "correct" speed?

I leave flow rate fixed because this is coming form my physics. But basically this should be also for speed...

Any comment?

March 17, 2011, 02:18	Fire Sprinkler simulation	#1
hup_cheng New Member apis cheng Join Date: Mar 2011 Posts: 4 Rep Power: 15	Hi all, I am very new to this Ansys CFX software. As the title, i want to simulate water distribution for fire sprinkler in a room. I have done through some tutorial in the example but i cant manage to get my simulation right. i am using multiphase for domain combination of water and air. for the input, the fraction of water is 1 while for the air is 0. output is at the floor. the streamline result only show velocity at the sprinkler but not reaching the floor of the room. i hope anybody can enlighten me in this simulation. is there any important setting that i left out? if there is a tutorial close to this simulation, let me know. thanx.

March 21, 2011, 03:25		#5
hup_cheng New Member apis cheng Join Date: Mar 2011 Posts: 4 Rep Power: 15	finally i solved the problem.. it is due to the initial condition of the domain.. but now my problem is the flow of the water is like bended.. its not falling in a straight line.. any idea what causes it? Last edited by hup_cheng; March 22, 2011 at 14:29. Reason: edit pics

July 30, 2019, 22:17		#16
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	These are FAQs: Floating point error: https://www.cfd-online.com/Wiki/Ansy...do_about_it.3F Poor convergence: https://www.cfd-online.com/Wiki/Ansy...gence_criteria __________________ Note: I do not answer CFD questions by PM. CFD questions should be posted on the forum.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
AVL fire CI engine simulation neeed HELP !	Mohsen khalafi	AVL FIRE	1	September 10, 2018 11:32
NEED HELP on IC engine simulation with AVL FIRE!	Mohsen khalafi	Main CFD Forum	2	February 6, 2014 08:47
smoke (fire simulation)	matt	Main CFD Forum	0	January 5, 2007 05:45
FIRE simulation	T.H. Ko	Siemens	3	July 25, 2003 00:18
Temperature errors on fire simulation?	raymond	Siemens	0	January 10, 2002 21:32

March 18, 2011, 07:47		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	How are you modelling the multiple phases? Particle tracking or eularian?

March 19, 2011, 02:06		#3
hup_cheng New Member apis cheng Join Date: Mar 2011 Posts: 4 Rep Power: 15	i believe it is particle tracking..

March 20, 2011, 21:55		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Can you post an image of what you are seeing?

March 22, 2011, 02:51		#7
hup_cheng New Member apis cheng Join Date: Mar 2011 Posts: 4 Rep Power: 15	but the condition of the air speed has been set to 0.. plus i only want to study the design of sprinkler distribution cap neglecting the air movement.. im still exploring cfx to get better result presentation.. thanx alot..

March 22, 2011, 07:55		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	The water droplet will drive a flow. If you want to disable the flow then use an expert parameter "solve fluids=f" then it will not update the fluid flow field.

October 16, 2012, 07:26		#10
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Do the tutorials, do some training, read the textbooks and do a literature search. And every once in a while put a post on the forum and we will try to help too.

October 18, 2012, 01:05	hi	#11
nilambar New Member Nilambar Bariha Join Date: Mar 2012 Posts: 14 Rep Power: 14	plz suggest some text books..

October 18, 2012, 06:49		#12
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,854 Rep Power: 144	Any text book with "introduction" and "CFD" is a good start. Here's a few, there are many more: http://www.cfd-online.com/Books/brow...?category_id=1

October 19, 2012, 03:54		#13
nilambar New Member Nilambar Bariha Join Date: Mar 2012 Posts: 14 Rep Power: 14	thanks for suggestion...

July 30, 2019, 12:33		#14
pigei New Member Paolo T. Join Date: Jul 2019 Posts: 15 Rep Power: 7	I'm new with CFX. Modelling a simple water sprinkler in air. Actually i used Lagrangian particle tracks but my simulation goes in error (floating) every time... I've tried different solutions to solve this problem but without any effects (mesh, timescale, increase particle injection, decrease particle speed and flow rate, ecc). I select no turbulence, isothermal and no inlet, so no speed, for air considered as countinous fluid. Initial condition has been setted to zero for every parameter. Geometry is actually small (about 3 x 2 x 2 m). Sprinkler inlet is modelled as a cilindrical surface. Water speed is about 10 m/s and flow rate abou 10 kg/s. Sprinkler direction is from top to bottom. I want to start from this basic and easy system in order to study later the behaviour of a water "wall" generated by a sprinkler on a gas release. I would compare particle track and fluid disperded different results. Any idea on how to solve this simulation solver error? RMS and imbalance are so oscillating...

July 30, 2019, 21:12		#15
Svetlana Senior Member Svetlana Tkachenko Join Date: Oct 2013 Location: Australia, Sydney Posts: 416 Rep Power: 15	Is this an unsteady simulation? Does it converge if you set the particle number to zero?

July 31, 2019, 07:12		#17
pigei New Member Paolo T. Join Date: Jul 2019 Posts: 15 Rep Power: 7	Thank you for your suggestion... This is a steady simulation. If i put 0 as number of positions for particle solver goes immediately in error. I noticed that: particle entered the domain: 10000 particle left domain: 9950 exceeded distance limit:50 I expect that particle will be accumulated on walls (floor). I've tried to reduce also speed and flow rate but nothing changed... I made a simulation with only 1 particle and RMS still oscillating...less... It is very strange because it a very simple simulation...no wind, no turbulence, isothermal...i want only to see the water! please help me...

July 31, 2019, 07:14		#18
pigei New Member Paolo T. Join Date: Jul 2019 Posts: 15 Rep Power: 7	mesh is ok and fitted enough in my opinion... on the sprinker inlet i have about 4mm size against 0.5mm particle diameter...is it ok?

July 31, 2019, 07:23		#19
pigei New Member Paolo T. Join Date: Jul 2019 Posts: 15 Rep Power: 7	another strange things... particle seems not to go out normal to inlet surface but with casual angles...

August 1, 2019, 04:09		#20
pigei New Member Paolo T. Join Date: Jul 2019 Posts: 15 Rep Power: 7	Hi, in case of interest i found the solution of my problem... It is related to particle speed. I noticed that if the speed is too low < 30m/s i have oscillating on RMS and particle seems coming from inlet at casual angle. If i put a speed > 40 m/s this problems are solved. Speed are related also with particle diameter. With smaller diameter i have to increase speed. Can anybody explain this strange behaviour? Is there any equation for calculating "correct" speed? I leave flow rate fixed because this is coming form my physics. But basically this should be also for speed... Any comment?