Fully Developed Conditions for Microchannel Heat-Sink (MCHS))

mech-engineer · June 2, 2012, 19:40

Hi daer friends
I want apply fully developed conditions (hydrodynamics and temperature conditions) at the end of a micro-channel heat sink (MCHS) that nanofluid flowing it.These conditions at outlet include in: 1)pressure equal to ambient pressure. , 2)Velocity gradient across the length of MCHS equal to zero. , 3)Temperature gradient across the length of MCHS equal to zero.
Please tell me how check these conditions for simulation and apply them via CFX?

ghorrocks · June 3, 2012, 07:02

Don't you mean velocity and temperature gradient is zero in the flow direction?

The simplest way to do these sort of flows is by using a periodic boundary with your defined flow rate and heat addition. Then it will converge on the full developed solution.

mech-engineer · June 3, 2012, 09:48

no,zero gradient only apply on outlet conditions.Can I give these conditions without apply of periodic conditions?

ghorrocks · June 3, 2012, 19:34

I see, you want to apply a zero gradient BC, you are not interested in fully developed flow. A zero gradient BC does not mean the flow in front of it is fully developed by the way. To get fully developed flow you need a duct long enough for it to develop. It has nothing to do with the BC - and this is the same for a physical experiment as well.

CFX does not have zero gradient BCs, it has flow variables convected from the domain (with a few exceptions, see the documentation for those).

mech-engineer · June 3, 2012, 20:07

How I can get enough length for simulation to ensure that flow is fully-developed at outlet.The geometry of my problem is a trapezoidal microchannel haet sink that dimensions of front area include in:400 micron (large base),300 micron (small base),350 micron (height),base angle=54 degree and length of microchannel is 10 millimeter normal to front area.Does these dimensions enough for you to answer me about fully developed conditions?

ghorrocks · June 3, 2012, 21:22

No, you cannot work out fully developed flow from those details as the fluid properties also contribute. And I do not have time to work it out for you anyway.

Read some basic fluid mechanics textbooks on flow development length. Alternately you can do a sensitivity study by testing various lengths and seeing if you can see it converging with increasing length.

mech-engineer · June 4, 2012, 05:16

With use of double precision at CFX-Solver Manager all of the equations (mass and momentum,energy,volume fraction) converge but when I use serial option at CFX-Solver Manager, energy equation can't converge.According to this comment, do you think my length is enough for obtaining a fully developed solution?

ghorrocks · June 4, 2012, 07:06

Whether your simulation achieves your specified residual convergance tolerance or you use double precision numerics has nothing to do with whether fully developed flow has formed.

June 2, 2012, 19:40	Fully Developed Conditions for Microchannel Heat-Sink (MCHS))	#1
mech-engineer New Member Behzad Join Date: Jun 2012 Posts: 26 Rep Power: 13	Hi daer friends I want apply fully developed conditions (hydrodynamics and temperature conditions) at the end of a micro-channel heat sink (MCHS) that nanofluid flowing it.These conditions at outlet include in: 1)pressure equal to ambient pressure. , 2)Velocity gradient across the length of MCHS equal to zero. , 3)Temperature gradient across the length of MCHS equal to zero. Please tell me how check these conditions for simulation and apply them via CFX?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Fully developed temperature profile for laminar/turbulent flow in FLUENT	raghu.tejaswi	FLUENT	10	November 7, 2012 09:57
there is a problem with fully developed Nusselt number	eeman	Fluent UDF and Scheme Programming	4	September 24, 2012 01:36
Fully Developed Flow in Star-cd	SMM	STAR-CD	0	September 5, 2011 22:08
udf for 3D laminar fully developed flow	salman	FLUENT	0	November 9, 2006 21:45
Convective Heat Transfer - Heat Exchanger	Mark	CFX	6	November 15, 2004 15:55

June 3, 2012, 07:02		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Don't you mean velocity and temperature gradient is zero in the flow direction? The simplest way to do these sort of flows is by using a periodic boundary with your defined flow rate and heat addition. Then it will converge on the full developed solution.

June 3, 2012, 09:48		#3
mech-engineer New Member Behzad Join Date: Jun 2012 Posts: 26 Rep Power: 13	no,zero gradient only apply on outlet conditions.Can I give these conditions without apply of periodic conditions?

June 3, 2012, 19:34		#4
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	I see, you want to apply a zero gradient BC, you are not interested in fully developed flow. A zero gradient BC does not mean the flow in front of it is fully developed by the way. To get fully developed flow you need a duct long enough for it to develop. It has nothing to do with the BC - and this is the same for a physical experiment as well. CFX does not have zero gradient BCs, it has flow variables convected from the domain (with a few exceptions, see the documentation for those).

June 3, 2012, 20:07		#5
mech-engineer New Member Behzad Join Date: Jun 2012 Posts: 26 Rep Power: 13	How I can get enough length for simulation to ensure that flow is fully-developed at outlet.The geometry of my problem is a trapezoidal microchannel haet sink that dimensions of front area include in:400 micron (large base),300 micron (small base),350 micron (height),base angle=54 degree and length of microchannel is 10 millimeter normal to front area.Does these dimensions enough for you to answer me about fully developed conditions?

June 3, 2012, 21:22		#6
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	No, you cannot work out fully developed flow from those details as the fluid properties also contribute. And I do not have time to work it out for you anyway. Read some basic fluid mechanics textbooks on flow development length. Alternately you can do a sensitivity study by testing various lengths and seeing if you can see it converging with increasing length.

June 4, 2012, 05:16		#7
mech-engineer New Member Behzad Join Date: Jun 2012 Posts: 26 Rep Power: 13	With use of double precision at CFX-Solver Manager all of the equations (mass and momentum,energy,volume fraction) converge but when I use serial option at CFX-Solver Manager, energy equation can't converge.According to this comment, do you think my length is enough for obtaining a fully developed solution?

June 4, 2012, 07:06		#8
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,700 Rep Power: 143	Whether your simulation achieves your specified residual convergance tolerance or you use double precision numerics has nothing to do with whether fully developed flow has formed.