Thanks for the video. This is very useful. The only thing I can say is that is better to say "Dissipation of energy (into heat)" instead of "Destruction"
thanks - good point! I generally talk about this in terms of incompressible isothermal flow, and so try to avoid mention of thermal effects but your explanation is of course more physical.
@@mcji8ar2 Thanks to you for the answer! I'm studying exactly this argument for the Fluids Labs (Fluid Mechanics - Computational Fluid Dynamics) exam, so I'm on the trail and your video is very helpful!
Great video! I have a couple of questions: 1. Since the vortices are shed constantly, where do you actually place your probe to measure the velocity? Even if the vortices were standing in one spot, I still wonder where in the vortex you'd have to place the probe ie at its center or on the edge 2. How long should you take your measurement for? Are there guidelines in this regard maybe in a textbook?
Hi sorry for delay! The diagram is highly simplified - in general many many vortices of all sizes would pass by the location of the probe as time went on - as such you'd get a representative measure of the range of spatial scales as long as your probe was in the turbulent wake region. As to how long - good question! In general it depends on how fast the flow is moving - if you say it takes one unit of time for the largest vortex to completely pass by the location of the probe (this is often referred to as a characteristic time scale) you might need something like 50-100 such time units to obtain decent statistics! in practice the precise duration will vary from flow to flow!
In general this supports a course I give so content follows the syllabus, but if you have a particular question/request let me know and I'll consider it
@@mcji8ar2 I want to know details regarding kolmogorov hypothesis and 4/5 th law. I ll be Very happy if you consider my request. It's my PhD work Thank you
Thanks a lot. So, if I got it correctly, the energy spectrum is nothing less than the energy amount of the "fluctuation" of the flow which depends on the vortices sizes on time. I was struggling to understand this in physical terms. Mostly everywhere I looked talked about the mathematics transformations, but no mention at all to correlating this to the physical world. The drawing was very helpful.
yes that is correct - the analysis is based on the notion that you separate the instantaneous velocity into the time-averaged velocity and the velocity fluctuations, or deviations, away from this average velocity. You then effectively have two velocities and can thus compute two different 'kinetic energies', compared to just one.
great video！i have some questions: 1 the probe is a "point"，the eddy/vortex is a "volume"，how a probe can measure the energy of an eddy having？ 2 giving us a known velocity field v(x y z t)，how to find "large eddy" "small eddy" in the field？drawing the streamlines and using human eyes to capture and identify？3 why eddy's size is usually associated with wavenumber？i mean，i cant feel wave in eddy. how to understand these concepts？
Thanks for your feedback. The eddy is indeed 3D, and it moves with the flow, while the probe is sampling velocity at a single point, i.e. a fixed coordinate. The spectra gives us a measure of the energy at a fixed point only, but if we assume the turbulence is composed of simple rotating, circular eddies, then we can describe the range of the size and strength of these eddies by measuring the velocity at this point. Remember, the spectra measures frequency, and so incorporates the time history of the eddies velocity fluctuations as they move through this point. Remember also that turbulence is the fluctuating part of the velocity, i.e. there is an underlying mean flow which moves the eddies through the flow field and past the probe. While the probe is stationary, the eddies continue to move over them. In the frequency space, large eddies (L>>1) correspond to low wave numbers (1/L
Probably a bit late, but would this be a typical technique for analysing the flow field in say, internal flows, and using the analysis to make modification to the flow? Or would simpler techniques be used first? great videos by the way, thanks for sharing
Thank you, yes, the methods are equally applicable to both internal and external flows. We often start our more complex simulations with more simple ones, both to help inform how we use the complex methods and to provide a decent initial solution. Otherwise the decision whether to use simple or more complex methods generally depends on the balance of accuracy and simulation time you want to strike.
Thanks for the video. So do you mean using a probe situated at one point in the field, u,v and w should be measured and then 0.5(u^2+v^2+w2) is calculated and plotted over time and then somehow correlated to the eddies that pass thru that point? So the graph you mentioned should be constructed using flow visualization data as well as velocity measurement data? Doesn't this approach seem a bit "rough"?
Hi there. The idea is that wherever in the flow you take a measurement of the unsteady instantaneous velocity, you can extract an energy spectrum from that point by considering the range of frequencies in the velocity signal. You don't need any flow visualisation for this - only a single probe gathering velocity data. The explanation in the video is using a sketch to link the passage of a large scale structure in the physical space to a peak in energy in the frequency space. hope that helps!
Thank you very much for sharing your work! I have velocity data (x, y, z) measured in a channel and I want to obtain the "Turbulence Energy Spectrum", how could I do it ?, that is, I know that in literature there are authors who write different equations for that, but How do I do it?
Hi Cesar, thanks for your comment. Basically you need to convert your velocity - time series into a frequency series, i.e. you need to do a (Fast) Fourier Transform or FFT. To do this take any of the components of the velocity as they vary in time and run FFT algorithm, e.g. using MATLAB (but many software alternatives also). For this to work well then 1) you need a fairly larger number of samples or velocity measurements, 2) the timestep or interval between measurements should be sufficiently small (this size dictates the highest resolvable 'frequency' and 3) you ideally need the timestep intervals to be constant. e.g always 1x10^-4 seconds apart.
Everything in CFD can be simplified in terms of convergent and divergent flow. Ignore turbulence, turbulence is a result of viscous forces in air, which is affected by air temperature. Convergence/Divergence, and curl can replace all of this fancy stuff. It's much easier to understand, and much easier to model with computers.
![[CFD] Turbulence Intensity for RANS](http://i.ytimg.com/vi/Xr7BzHImL68/mqdefault.jpg)
Thank you for this great series of lectures on energy cascade Alistair, I always reffer to them to recall my self this topic! cheers, Alexander
Excellent explanation, Dr Revell!
wow, great chanel, keep it up sir!
well done!
Wow sir just what i was looking for
Good video.
These energies in the Eddies, does this explain the jawbreaker scams?
Very nice video. Very didactic your explanation, more videos like this... Greetings 😊
Thank you, I hope to add more soon
Thankyou, very well explained
Can the inverse energy cascade collect energy when growing?
Thanks for the video. This is very useful. The only thing I can say is that is better to say "Dissipation of energy (into heat)" instead of "Destruction"
thanks - good point! I generally talk about this in terms of incompressible isothermal flow, and so try to avoid mention of thermal effects but your explanation is of course more physical.
@@mcji8ar2 Thanks to you for the answer! I'm studying exactly this argument for the Fluids Labs (Fluid Mechanics - Computational Fluid Dynamics) exam, so I'm on the trail and your video is very helpful!
Great video! I have a couple of questions: 1. Since the vortices are shed constantly, where do you actually place your probe to measure the velocity? Even if the vortices were standing in one spot, I still wonder where in the vortex you'd have to place the probe ie at its center or on the edge 2. How long should you take your measurement for? Are there guidelines in this regard maybe in a textbook?
Hi sorry for delay! The diagram is highly simplified - in general many many vortices of all sizes would pass by the location of the probe as time went on - as such you'd get a representative measure of the range of spatial scales as long as your probe was in the turbulent wake region. As to how long - good question! In general it depends on how fast the flow is moving - if you say it takes one unit of time for the largest vortex to completely pass by the location of the probe (this is often referred to as a characteristic time scale) you might need something like 50-100 such time units to obtain decent statistics! in practice the precise duration will vary from flow to flow!
Thanks for the video and will you do some more videos on turbulence and different laws associated with.
In general this supports a course I give so content follows the syllabus, but if you have a particular question/request let me know and I'll consider it
@@mcji8ar2 I want to know details regarding kolmogorov hypothesis and 4/5 th law. I ll be Very happy if you consider my request. It's my PhD work
Thank you
Thanks a lot. So, if I got it correctly, the energy spectrum is nothing less than the energy amount of the "fluctuation" of the flow which depends on the vortices sizes on time. I was struggling to understand this in physical terms. Mostly everywhere I looked talked about the mathematics transformations, but no mention at all to correlating this to the physical world. The drawing was very helpful.
yes that is correct - the analysis is based on the notion that you separate the instantaneous velocity into the time-averaged velocity and the velocity fluctuations, or deviations, away from this average velocity. You then effectively have two velocities and can thus compute two different 'kinetic energies', compared to just one.
Thank youuuuuu🙏🙏🙏🙏
great video！i have some questions: 1 the probe is a "point"，the eddy/vortex is a "volume"，how a probe can measure the energy of an eddy having？ 2 giving us a known velocity field v(x y z t)，how to find "large eddy" "small eddy" in the field？drawing the streamlines and using human eyes to capture and identify？3 why eddy's size is usually associated with wavenumber？i mean，i cant feel wave in eddy. how to understand these concepts？
Thanks for your feedback. The eddy is indeed 3D, and it moves with the flow, while the probe is sampling velocity at a single point, i.e. a fixed coordinate. The spectra gives us a measure of the energy at a fixed point only, but if we assume the turbulence is composed of simple rotating, circular eddies, then we can describe the range of the size and strength of these eddies by measuring the velocity at this point. Remember, the spectra measures frequency, and so incorporates the time history of the eddies velocity fluctuations as they move through this point. Remember also that turbulence is the fluctuating part of the velocity, i.e. there is an underlying mean flow which moves the eddies through the flow field and past the probe. While the probe is stationary, the eddies continue to move over them.
In the frequency space, large eddies (L>>1) correspond to low wave numbers (1/L
Probably a bit late, but would this be a typical technique for analysing the flow field in say, internal flows, and using the analysis to make modification to the flow? Or would simpler techniques be used first?
great videos by the way, thanks for sharing
Thank you, yes, the methods are equally applicable to both internal and external flows. We often start our more complex simulations with more simple ones, both to help inform how we use the complex methods and to provide a decent initial solution. Otherwise the decision whether to use simple or more complex methods generally depends on the balance of accuracy and simulation time you want to strike.
Thanks for the video. So do you mean using a probe situated at one point in the field, u,v and w should be measured and then 0.5(u^2+v^2+w2) is calculated and plotted over time and then somehow correlated to the eddies that pass thru that point? So the graph you mentioned should be constructed using flow visualization data as well as velocity measurement data? Doesn't this approach seem a bit "rough"?
Hi there. The idea is that wherever in the flow you take a measurement of the unsteady instantaneous velocity, you can extract an energy spectrum from that point by considering the range of frequencies in the velocity signal. You don't need any flow visualisation for this - only a single probe gathering velocity data. The explanation in the video is using a sketch to link the passage of a large scale structure in the physical space to a peak in energy in the frequency space. hope that helps!
Thank you very much for sharing your work! I have velocity data (x, y, z) measured in a channel and I want to obtain the "Turbulence Energy Spectrum", how could I do it ?, that is, I know that in literature there are authors who write different equations for that, but How do I do it?
Hi Cesar, thanks for your comment. Basically you need to convert your velocity - time series into a frequency series, i.e. you need to do a (Fast) Fourier Transform or FFT. To do this take any of the components of the velocity as they vary in time and run FFT algorithm, e.g. using MATLAB (but many software alternatives also). For this to work well then 1) you need a fairly larger number of samples or velocity measurements, 2) the timestep or interval between measurements should be sufficiently small (this size dictates the highest resolvable 'frequency' and 3) you ideally need the timestep intervals to be constant. e.g always 1x10^-4 seconds apart.
Everything in CFD can be simplified in terms of convergent and divergent flow. Ignore turbulence, turbulence is a result of viscous forces in air, which is affected by air temperature. Convergence/Divergence, and curl can replace all of this fancy stuff. It's much easier to understand, and much easier to model with computers.