question about turbulent kinetic energy

junker4236 · December 22, 2010, 20:14

i have a Question about turbulent kinetic energy.

I did a CFD Simulation of Car, i have this Picture of turbulent kinetic

energy

Can somebody explain, what the turbulent kinetic energy mean and the

large turbulent kinetic energy in the Area of the car bring which result??

tks everybody for helping me

Irshad22 · December 23, 2010, 03:55

Turbulence kinetic energy (TKE) is the mean K.E per unit mass associated with eddies in turbulent flow. For the car, it is loss of ke from mean flow in order to continue eddies which are formed due to pressure difference.

Hamidzoka · December 23, 2010, 06:43

Dear Junker;
when flow becomes turbulent, transport mechanisms like diffusion, highly increases beyond the molecular diffusion (which is encountered in laminar flows). it is believed that turbulent flows structure includes parcells with highly transient characteristics called eddies. in fact, eddies movements intensifies flow transportation mechanisms like diffusion. they can be as large as flow integral scale and they can also be very small.
Turbulence kinetic energy (TKE) is the energy content of eddies in turbulent flows. larger the size, the higher is energy content of eddies.
TKE is extracted from mean flow to larger eddies, from larger eddies to smaller ones and finally it is going to be dissipated in very small eddies where viscous effects defeats the kinetic energy.
high TKE points to regions in which high amount of turbulent energy is extracted from mean flow.

i hope it helps
Ragards

junker4236 · December 23, 2010, 12:55

tks for your both Help to explain, what the TKE mean. I understand it a bit

more.

But now my question ist: the Area with highly TKE, just like the Underbody

and the Backside of car in Picture. Do this Area bring some negative

effects of the Aerodynamik because of this eddies

Merry chrismas

Irshad22 · December 23, 2010, 13:15

At the rear end there is low pressure because of flow seperation. Because of low pressure at rear and high pressure at front, presure drag produced which is major in Aerodynamics of road vehicles. In the low pressure region large eddies formed, hence it extracts more k.e from mean flow. similarly, at frond end curvature, flow departs into two ways and because of low ground clearance, high speed flow with low pressure region will form at the underside. That is why TKE dissipation is more.

junker4236 · December 23, 2010, 13:23

can you pls explain, why there is low pressure in Large Eddies? just because of Movement?

Irshad22 · December 23, 2010, 13:49

Whenever the flow seperation occurs, the vaccum space will be created because of unablity to attach. Later flow try to fill the vaccum as flow moves from high preesure to low pressure. In taht process, reversal flow occurs as well eddies form by interaction with mean flow which have sufficient energy. since flow reversal occurs very faster, low pressure will be created and vice versa as per Bernoulli's principle.

bhatiadinesh · December 24, 2010, 00:16

I had an idea of a somewhat different flow seperation phenomenon ocuuring behind 'bluff bodies' and also the logic behind formation of the eddies.
Pls correct me if i am wrong.
1) First of all, there is no vaccum created, (I know that you must have meant a low pressure zone, but just for the heck of mentioning)
2) When the flow across the bluff bodies occur, the total energy of the flow ( I mean relative to the body ), especially total pressure over the body is higher than the zone of stagnant air behind it.
3) Hence when the flow over the body reaches the turn of high gradient in the aft region of the bluff body, the boundary layer can no loger sustain such a pressure difference and if you see the classical boundary layer seperation picture you cn see, due the seperation itself very high vorticity is induced in the flow, Hence whe this flow seperates it 'curls' , leading to formation of eddies.

Hence the answer to Junkers Ques would be that, the low pressure at the core is one of the reasons for formation of eddies

My intention here is to only get a clearer picture by having an open discussion,

ANYBODY HAVING A GOOD IDEA OF THIS PHENOMENON PLEASE SHED SOME LIGHT ON IT AND CORRECT ME IF I AM MISTAKEN.

As far as your question on ill effects of high TKE is concerned, one obvious drawbacks is greater skin friction drag due to increased tubulence because of seperation, but I dont know statistically how much does this matter.

-Dinesh

ignat · December 24, 2010, 03:06

Quote:

Originally Posted by junker4236

Can somebody explain, what the turbulent kinetic energy mean and the

In turbulent flows instantaneous velocity is presented as

where

stands for the mean part of any value and

is the velocity fluctuation part. Mean part is calculated as average within time period T which is large in comparison with typical fluctuations.

Average fluctuations

is equal to zero but

is not.

Value $\sqrt{<u^2>}$ is named turbulent intensity.
Half of sum of square intensities in all 3 directions is named turbulent kinetic energy (k):
$k=\frac{1}{2}(<u_x^2>+<u_y^2>+<u_z^2>)$
see for example Hinze's book "Turbulence" or other.

siamakghh2000 · July 27, 2015, 14:58

I'm measuring the fluctuating component of the velocity in a uniform rectangular open channel flow (imagine at various locations of a section). Does anyone know how I can compute the total kinetic energy of turbulence by having data of only the streamwise velocity fluctuation?

k=1/2(Ux^2+Uy^2+Uz^2)

I only have measured Ux (mean fluctuating component in flow direction),

but perhaps I guess I can somehow calculate k by adding certain constants...
Thanks in advance.

RANSES · January 12, 2016, 08:29

Hello and sorry for using an old thread...it's the first time I write in the forum and I found this conversation suitable for my question. I'm working on a master thesis project, reguarding the simulation of a flow field around a rectangular building. And I'm wondering about which is the theoretical link between the turbulence kinetic energy and an obstacle immersed in the flow field....more specifically, next to the obstacle sides, how does the separation bubble affect (or should affect) the TKE? Is it legit to assume that the presence of an obstacle can improve the value of the speed fluctuations? Is it also legit to assume the higher the mean velocity, the higher the velocity fluctuations?

Hamidzoka · January 12, 2016, 09:01

Dear RANSES;
- one important source of turbulence generation in a flow field is the shear stress. when separation bubble is formed, a shear stress will be present between the recirculating region and the main flow passing by. at their interface first a microscopic momentum transfer happens and then it turns into a macroscopic one very quickly. that is the onset of turbulent flow generation. TKE. in fact, represents the turbulence generation within the flow field. so it will be increased near separation regions.
- Regarding the sound i am not sure, but i think an acoustic simulation should be performed.
- the velocity fluctuations (turbulence intensity) will be more intense if both mean velocity and damping (i.e. dissipation of energy) is higher. in fact turbulent flow is characterized by high level of kinetic energy as well as very high diffusive nature.(so much higher than what we see in laminar flows that is caused by molecular viscosity).

regards

RANSES · January 12, 2016, 11:24

Quote:

Originally Posted by Hamidzoka

Dear RANSES;
- one important source of turbulence generation in a flow field is the shear stress. when separation bubble is formed, a shear stress will be present between the recirculating region and the main flow passing by. at their interface first a microscopic momentum transfer happens and then it turns into a macroscopic one very quickly. that is the onset of turbulent flow generation. TKE. in fact, represents the turbulence generation within the flow field. so it will be increased near separation regions.
- Regarding the sound i am not sure, but i think an acoustic simulation should be performed.
- the velocity fluctuations (turbulence intensity) will be more intense if both mean velocity and damping (i.e. dissipation of energy) is higher. in fact turbulent flow is characterized by high level of kinetic energy as well as very high diffusive nature.(so much higher than what we see in laminar flows that is caused by molecular viscosity).

regards

Thank you very much!

siamakghh2000 · January 12, 2016, 16:01

Dear RANSES,

In addition to Hamidzoka's response, I may add that as Hamid stressed, kinetic energy increasing is an indication of turbulence generation. The disturbance introduced into flow (obstacle) will be the source of turbulence generation where it ignites within the separation bubble and is carried downstream with eddies of different scales forming and gradually dying far downstream as a result of laminar damping at last (or say, when the momentum transfer phase is complete). But, I think for conventional Re numbers an acoustic simulation may not be required.
As an estimate, I think kinetic energy will be always high near the two upstream edges of the obstacle normal to flow and they will be pushed towards the walls as the Re increases and stretched towards the wack.(you may easily investigate the link with a simple fluent simulation by highlighting the TKE intensity field).

Yours, respectfully,
Siamak Gharahjeh

davidwilcox · January 15, 2016, 09:34

Here are my thoughts after reading all the posts.

i think you have to be careful with the words eddy and a fluid structure. They are two different things. A structure like the vortex core has a low pressure region in its core because velocity has to be infinite there. This is in response to the question why an "eddy"( which should be called a fluid structure or vortex/vortices) has a low pressure region.

With simulations, you have to be careful as well because some turbulence models give spurious amounts of tke in stagnation region. The k-epsilon comes to mind and it is well known to over-predict tke in stagnation region. this is known as the stagnation point anomaly.

The car is seen more as an aerodynamic body rather than a bluff body. what's the difference? DRAG. There are two forms of drag namely aerodynamic/viscous drag and pressure drag. If you have more aerodynamic drag than pressure/form drag then it is an aerodynamic body.

Finally the production of turbulence, let's say for simplicity in a 2d flow is u'v'(du/dy). so if you have high velocity gradients, you will have a significant amount of turbulence generation.

siamakghh2000 · January 15, 2016, 09:47

Dear David,
You are right. A vortex has nothing to do with eddy. However, in this case separation (inside which there is vortex) is causing turbulence. Turbulence is defined through eddies of different scales.
Even when flow is as simple as that of a pipe flow, turbulence simulation will be successful with assumption of eddy structures in some cases.
Yours, respectfully,
Siamak Gharahjeh

davidwilcox · January 15, 2016, 09:54

this must be stressed. the word vortex cannot be used loosely. Vorticity doesnt mean vortex. A vortex is a fluid structure which can be identified through mathematical relationships like the Q criterion or Lamba 2 criterion or the more fancy Finite time lyapunov exponent method. Hussein wrote a nice paper on this titled something like identification of a vortex. It's a great read!

RANSES · February 2, 2016, 19:29

Thank you once again for all the fast and useful answers and forgive me for replying only now. I agree with almost all your words. My work with this thesis is going on and, even if a ground mounted buildig could seem an easy case, the simulation of the flow-field in its vicinity is rather challenging. Not only the upwind and the downwind recirculation zones but also (and especially if considering k) the separation bubble on top of the building, in the leading edge of the obstacle are quite problematic. In all the k-epsilon models, including standard, reliazable, quadratic and certain cubic ones, k is somehow always overstimated in the impinging edge. And I too, in the beginning, was expecting high values of turbulent kinetic energy in that position, but from experimental data, it comes out they are extremely low. Probably (the explanation I give to myself) this is due to the low velocity and recirculation values in the small stagnation zone over there, leading also to low sher stress values. The great challenge will be to make the simulations more reliable in those areas!

letsfree · April 18, 2017, 15:31

hi all, i have run a simulation regarding a circular pipe flow and it is heated by a heating coil around the pipe. the contour of turbulent kinetic energy shows that the turbulent kinetic energy (TKE) is change from low to high along the tube. My question is, why at inlet region (left hand side) the TKE is high at near wall region while at the center is relatively low?

Blanco · April 19, 2017, 04:46

The evolution of k in your domain is due to the conditions you used for the inlet boundary. What you obtain here is the development of the boundary layer along the pipe, following the fluid motion. If you want to simulate a fully developed flow in a pipe you have to provide the correct conditions to your inlet boundary, i.e. a velocity profile, etc. etc. I have no experience with Fluent, but I suppose there should be a suitable boundary conditions for this purpose.

December 22, 2010, 20:14	question about turbulent kinetic energy	#1
junker4236 New Member fanwei Join Date: Apr 2009 Posts: 4 Rep Power: 17	i have a Question about turbulent kinetic energy. I did a CFD Simulation of Car, i have this Picture of turbulent kinetic energy Can somebody explain, what the turbulent kinetic energy mean and the large turbulent kinetic energy in the Area of the car bring which result?? tks everybody for helping me

December 23, 2010, 06:43		#3
Hamidzoka Senior Member Hamid Zoka Join Date: Nov 2009 Posts: 282 Rep Power: 18	Dear Junker; when flow becomes turbulent, transport mechanisms like diffusion, highly increases beyond the molecular diffusion (which is encountered in laminar flows). it is believed that turbulent flows structure includes parcells with highly transient characteristics called eddies. in fact, eddies movements intensifies flow transportation mechanisms like diffusion. they can be as large as flow integral scale and they can also be very small. Turbulence kinetic energy (TKE) is the energy content of eddies in turbulent flows. larger the size, the higher is energy content of eddies. TKE is extracted from mean flow to larger eddies, from larger eddies to smaller ones and finally it is going to be dissipated in very small eddies where viscous effects defeats the kinetic energy. high TKE points to regions in which high amount of turbulent energy is extracted from mean flow. i hope it helps Ragards tino_boelke, amolrajan, Kimican and 6 others like this.

December 23, 2010, 13:15		#5
Irshad22 New Member Irshad Join Date: Dec 2009 Location: Bangalore Posts: 19 Rep Power: 16	At the rear end there is low pressure because of flow seperation. Because of low pressure at rear and high pressure at front, presure drag produced which is major in Aerodynamics of road vehicles. In the low pressure region large eddies formed, hence it extracts more k.e from mean flow. similarly, at frond end curvature, flow departs into two ways and because of low ground clearance, high speed flow with low pressure region will form at the underside. That is why TKE dissipation is more. Kimican likes this.

December 23, 2010, 13:49		#7
Irshad22 New Member Irshad Join Date: Dec 2009 Location: Bangalore Posts: 19 Rep Power: 16	Whenever the flow seperation occurs, the vaccum space will be created because of unablity to attach. Later flow try to fill the vaccum as flow moves from high preesure to low pressure. In taht process, reversal flow occurs as well eddies form by interaction with mean flow which have sufficient energy. since flow reversal occurs very faster, low pressure will be created and vice versa as per Bernoulli's principle. Kimican and javadsouri like this.

July 27, 2015, 14:58	Total kinetic energy from unidirectional fluctuating component	#10
siamakghh2000 New Member Siamak Gharahjeh Join Date: Aug 2012 Posts: 27 Rep Power: 13	I'm measuring the fluctuating component of the velocity in a uniform rectangular open channel flow (imagine at various locations of a section). Does anyone know how I can compute the total kinetic energy of turbulence by having data of only the streamwise velocity fluctuation? k=1/2(Ux^2+Uy^2+Uz^2) I only have measured Ux (mean fluctuating component in flow direction), but perhaps I guess I can somehow calculate k by adding certain constants... Thanks in advance.

December 23, 2010, 03:55		#2
Irshad22 New Member Irshad Join Date: Dec 2009 Location: Bangalore Posts: 19 Rep Power: 16	Turbulence kinetic energy (TKE) is the mean K.E per unit mass associated with eddies in turbulent flow. For the car, it is loss of ke from mean flow in order to continue eddies which are formed due to pressure difference.

December 23, 2010, 12:55		#4
junker4236 New Member fanwei Join Date: Apr 2009 Posts: 4 Rep Power: 17	tks for your both Help to explain, what the TKE mean. I understand it a bit more. But now my question ist: the Area with highly TKE, just like the Underbody and the Backside of car in Picture. Do this Area bring some negative effects of the Aerodynamik because of this eddies Merry chrismas

December 23, 2010, 13:23		#6
junker4236 New Member fanwei Join Date: Apr 2009 Posts: 4 Rep Power: 17	can you pls explain, why there is low pressure in Large Eddies? just because of Movement?

December 24, 2010, 00:16		#8
bhatiadinesh Member D Join Date: Nov 2010 Posts: 50 Rep Power: 15	I had an idea of a somewhat different flow seperation phenomenon ocuuring behind 'bluff bodies' and also the logic behind formation of the eddies. Pls correct me if i am wrong. 1) First of all, there is no vaccum created, (I know that you must have meant a low pressure zone, but just for the heck of mentioning) 2) When the flow across the bluff bodies occur, the total energy of the flow ( I mean relative to the body ), especially total pressure over the body is higher than the zone of stagnant air behind it. 3) Hence when the flow over the body reaches the turn of high gradient in the aft region of the bluff body, the boundary layer can no loger sustain such a pressure difference and if you see the classical boundary layer seperation picture you cn see, due the seperation itself very high vorticity is induced in the flow, Hence whe this flow seperates it 'curls' , leading to formation of eddies. Hence the answer to Junkers Ques would be that, the low pressure at the core is one of the reasons for formation of eddies My intention here is to only get a clearer picture by having an open discussion, ANYBODY HAVING A GOOD IDEA OF THIS PHENOMENON PLEASE SHED SOME LIGHT ON IT AND CORRECT ME IF I AM MISTAKEN. As far as your question on ill effects of high TKE is concerned, one obvious drawbacks is greater skin friction drag due to increased tubulence because of seperation, but I dont know statistically how much does this matter. -Dinesh

January 12, 2016, 08:29	Link between TKE and separation bubble	#11
RANSES New Member Riccardo Join Date: Jan 2016 Posts: 16 Rep Power: 10	Hello and sorry for using an old thread...it's the first time I write in the forum and I found this conversation suitable for my question. I'm working on a master thesis project, reguarding the simulation of a flow field around a rectangular building. And I'm wondering about which is the theoretical link between the turbulence kinetic energy and an obstacle immersed in the flow field....more specifically, next to the obstacle sides, how does the separation bubble affect (or should affect) the TKE? Is it legit to assume that the presence of an obstacle can improve the value of the speed fluctuations? Is it also legit to assume the higher the mean velocity, the higher the velocity fluctuations? Last edited by RANSES; January 15, 2016 at 20:23.

January 12, 2016, 09:01		#12
Hamidzoka Senior Member Hamid Zoka Join Date: Nov 2009 Posts: 282 Rep Power: 18	Dear RANSES; - one important source of turbulence generation in a flow field is the shear stress. when separation bubble is formed, a shear stress will be present between the recirculating region and the main flow passing by. at their interface first a microscopic momentum transfer happens and then it turns into a macroscopic one very quickly. that is the onset of turbulent flow generation. TKE. in fact, represents the turbulence generation within the flow field. so it will be increased near separation regions. - Regarding the sound i am not sure, but i think an acoustic simulation should be performed. - the velocity fluctuations (turbulence intensity) will be more intense if both mean velocity and damping (i.e. dissipation of energy) is higher. in fact turbulent flow is characterized by high level of kinetic energy as well as very high diffusive nature.(so much higher than what we see in laminar flows that is caused by molecular viscosity). regards amolrajan and Gerhard like this.

January 12, 2016, 16:01		#14
siamakghh2000 New Member Siamak Gharahjeh Join Date: Aug 2012 Posts: 27 Rep Power: 13	Dear RANSES, In addition to Hamidzoka's response, I may add that as Hamid stressed, kinetic energy increasing is an indication of turbulence generation. The disturbance introduced into flow (obstacle) will be the source of turbulence generation where it ignites within the separation bubble and is carried downstream with eddies of different scales forming and gradually dying far downstream as a result of laminar damping at last (or say, when the momentum transfer phase is complete). But, I think for conventional Re numbers an acoustic simulation may not be required. As an estimate, I think kinetic energy will be always high near the two upstream edges of the obstacle normal to flow and they will be pushed towards the walls as the Re increases and stretched towards the wack.(you may easily investigate the link with a simple fluent simulation by highlighting the TKE intensity field). Yours, respectfully, Siamak Gharahjeh Gerhard likes this.

January 15, 2016, 09:34		#15
davidwilcox Senior Member david Join Date: Oct 2012 Posts: 142 Rep Power: 13	Here are my thoughts after reading all the posts. i think you have to be careful with the words eddy and a fluid structure. They are two different things. A structure like the vortex core has a low pressure region in its core because velocity has to be infinite there. This is in response to the question why an "eddy"( which should be called a fluid structure or vortex/vortices) has a low pressure region. With simulations, you have to be careful as well because some turbulence models give spurious amounts of tke in stagnation region. The k-epsilon comes to mind and it is well known to over-predict tke in stagnation region. this is known as the stagnation point anomaly. The car is seen more as an aerodynamic body rather than a bluff body. what's the difference? DRAG. There are two forms of drag namely aerodynamic/viscous drag and pressure drag. If you have more aerodynamic drag than pressure/form drag then it is an aerodynamic body. Finally the production of turbulence, let's say for simplicity in a 2d flow is u'v'(du/dy). so if you have high velocity gradients, you will have a significant amount of turbulence generation. juliom, amolrajan, Antonio Mota and 2 others like this.

January 15, 2016, 09:47		#16
siamakghh2000 New Member Siamak Gharahjeh Join Date: Aug 2012 Posts: 27 Rep Power: 13	Dear David, You are right. A vortex has nothing to do with eddy. However, in this case separation (inside which there is vortex) is causing turbulence. Turbulence is defined through eddies of different scales. Even when flow is as simple as that of a pipe flow, turbulence simulation will be successful with assumption of eddy structures in some cases. Yours, respectfully, Siamak Gharahjeh Antonio Mota likes this.

January 15, 2016, 09:54		#17
davidwilcox Senior Member david Join Date: Oct 2012 Posts: 142 Rep Power: 13	this must be stressed. the word vortex cannot be used loosely. Vorticity doesnt mean vortex. A vortex is a fluid structure which can be identified through mathematical relationships like the Q criterion or Lamba 2 criterion or the more fancy Finite time lyapunov exponent method. Hussein wrote a nice paper on this titled something like identification of a vortex. It's a great read! juliom, FMDenaro and amolrajan like this.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search
Display Modes
Linear Mode Switch to Hybrid Mode Switch to Threaded Mode

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
How to view subgrid kinetic energy	Craig	FLUENT	3	September 8, 2012 11:34
ATTENTION! Reliability problems in CFX 5.7	Joseph	CFX	14	April 20, 2010 15:45
sgs turbulent kinetic energy	blons	Main CFD Forum	0	February 25, 2010 12:16
the subgrid scale turbulent kinetic energy?	mud2003	Main CFD Forum	3	July 21, 2006 09:29
what's the macro of SGS turbulent kinetic energy?	lcw	FLUENT	1	June 13, 2006 03:08

February 2, 2016, 19:29		#18
RANSES New Member Riccardo Join Date: Jan 2016 Posts: 16 Rep Power: 10	Thank you once again for all the fast and useful answers and forgive me for replying only now. I agree with almost all your words. My work with this thesis is going on and, even if a ground mounted buildig could seem an easy case, the simulation of the flow-field in its vicinity is rather challenging. Not only the upwind and the downwind recirculation zones but also (and especially if considering k) the separation bubble on top of the building, in the leading edge of the obstacle are quite problematic. In all the k-epsilon models, including standard, reliazable, quadratic and certain cubic ones, k is somehow always overstimated in the impinging edge. And I too, in the beginning, was expecting high values of turbulent kinetic energy in that position, but from experimental data, it comes out they are extremely low. Probably (the explanation I give to myself) this is due to the low velocity and recirculation values in the small stagnation zone over there, leading also to low sher stress values. The great challenge will be to make the simulations more reliable in those areas! Last edited by RANSES; February 3, 2016 at 06:58.

April 19, 2017, 04:46		#20
Blanco Senior Member Blanco Join Date: Mar 2009 Location: Torino, Italy Posts: 193 Rep Power: 17	The evolution of k in your domain is due to the conditions you used for the inlet boundary. What you obtain here is the development of the boundary layer along the pipe, following the fluid motion. If you want to simulate a fully developed flow in a pipe you have to provide the correct conditions to your inlet boundary, i.e. a velocity profile, etc. etc. I have no experience with Fluent, but I suppose there should be a suitable boundary conditions for this purpose.