Boundary layer separation has place only in the presence of turbulence (adverse pressure gradient).

Boundary layers are defined as thin regions, around a solid body immerged in a flow at high Reynolds numbers, in which viscous effects are comparable to inertial effects, while they are negligible in the main flow. There exist both laminar and turbulent boundary layers depending on the flow conditions (mainly Reynolds number). Because they are thin, the flow within a boundary layer is quasi-parallel. As a consequence in first order approximation the pressure does not varies in the direction normal to the flow. The pressure in the boundary layer is the pressure imposed by the main flow on the "edge" of the boundary layer. Flow separation in a steady flow occurs when due to the shape of the body is such that a negative pressure gradient in the flow direction occurs in a boundary layer. Flow separation is delayed by viscous entrainment and the flow separation changes drastically the main (frictionless flow) which makes prediction not straight forward. Flow separation implies the formation of a "dead-water region" or "wake" along/behind the body. This low flow speed region is separated from the frictionless main flow stream far from the body by a shear layer. The shear layer is a quasi-parallel flow region in which the flow velocity decreases normal to the flow direction from high (in the main flow) to low (in the wake). Such a shear layer is intrinsically unstable. At sufficiently high Reynolds numbers the instability results in turbulence. When the boundary layer is already turbulent it will already be turbulent from the start. As turbulence promotes momentum transfer normal to the flow direction, turbulence will actually delay flow separation. In some cases surface roughness is sufficient to promote boundary layer turbulence and drastically reduce the width of the wake behind a bluff body. This results in a reduced drag. In general the flow separation can be delayed by injecting momentum in the main flow direction close to the wall or promoting transfer of momentum from the main flow to the boundary layer. There is a HUGE amount of literature on this subject. In the case of a aircraft wing at moderately high Reynolds numbers one can design the shape of the wing to avoid flow separation. At high Reynolds numbers aircrafts use nice tricks to delay flow separation (stall). Examples are the slats (the wing is build up a separate smaller wings, the flow through the slits between the wings reduces flow separation).

Howdy Sanjibonny Buragohain

"Can it be posited that the presence of a severe boundary layer separation, characterized by the detachment of the fluid flow from a solid surface, serves as a significant contributing factor to the augmentation of turbulent flow phenomena?" Yes. The spread of the turbulent wake of a car can be reduced by an inverted wing on the trunk that pulls the flow down and delays or prevents separation from the top of a car. The wake is then smaller and drag on the car is reduced. In a wind tunnel, a trip wire or roughness triggering early turbulence and enhanced momentum transfer can delay separation and produce a smaller wake. For owls, "Comb-like serrations on the leading edge of wing feathers break up the turbulent air that typically creates a swooshing sound." (Wikipedia) which I interpret as another "yes."

"In other words, is there a substantiated relationship between the adverse separation of the boundary layer and the amplification of turbulence in the flow field?" Well, yes, but I prefer intensity to amplification.

"Is it plausible to achieve a state of boundary layer separation characterized by an exceptionally smooth flow transition, resulting in nearly negligible levels of turbulence? No, at least not reliably in any case of Renewable Energy using natural flows. The atmosphere and even tidal currents have too much variability. If there is enough energy in the flow to be worth use, smooth transition is unlikely.

In other words, can the phenomenon of boundary layer separation be effectively controlled and manipulated to minimize the formation and propagation of turbulent flow structures?" Perhaps, to a limited extent. The core of the problem is that the energy is in the flow and turbulence happens, often with separation. Now, there is hope in the controls noted above, and like working athletes up to a second wind before competition, one might impose a small-scale turbulence generation device, say a sequence of grids or an array of columns before a wind turbine to suppress separation and large eddies, but it is unlikely the benefit would justify the effort.

There exist numerous wind farms and extensive instrumentation is in place. Are there any studies of the nature of turbulence in the lee of a number of turbines, like the nature of turbulence leaving an open forest? I've been away since '80 when I helped with siting proposals, so I don't know where to look, but you should be able to find sources and perhaps answers to your question. Good luck!

Happy Trails, Len

I think you should think the other way around. When talking about flow separation, the flow conditions have passed certain Re number, which talks about the turbulence levels in the flow, of course. So, if the flow Re number is small, there is no separation (close to ideal flow) and the behavior is purely viscous. However, when you increase Re number (approaching turbulent regime), more inertial forces are applied, which will cause separation.

Turbulent flow enhanced by increase of flow Reynolds number. And separation is occurring when flow is subjected to severe adverse pressures gradient. But as Dr. Hirschberg said above, separation could be postponed on aerodynamic surfaces by triggering the flow, or to say energising the flow by imposing turbulence artificially. Because turbulence causing higher momentum to the flow and eventually the dead region produced by separation will be diminished resulting in reattachment of boundary layer to the wall. This leads to less drag.

It depends in the Reynolds number, as stated by other people before.

For very low Reynolds, the separation can be laminar (e.g. at 5

Howdy Ricardo Fernández ,

This is great. If I read Sanjibonny Buragohain 's question right, your skills and expertise should enable you to answer it well. Can you supply citations for: "There are also techniques to increase the mixing between the wake and the free flow, to make the wake dissipate quicker (i.e. in a shorter distance behind the body); this is used for example to reduce interaction between wind turbines in a wind farm."? As noted in my reply I have been away from all this since '80, but I expect you are up to date.

I doubt that Sanjibonny Buragohain has benefited much from the general lectures here, since her profile indicates education and experience that would have addressed the topic. I sure she is aware of the Reynolds Number. The lectures are a good review, of course, and maybe I consider persons too important by comparison, but the students that liked my style enjoyed being persons with specific questions. What problem does the person, Sanjibonny Buragohain, need help to solve? It may be how to think about turbulence in a specific case to discover a better solution and a window on focused literature will help. Or, I may be wrong.

Citations of the studies you note might be valuable contributions to the thread.

Happy Trails, Len

The methods that Dr. Hall is talking about to dissipate the wake faster may include exciting the flow artificially behind any body. It can be acoustically which are used behind the trucks, and or suction or blowing into separated boundary layer. All of these techniques reducing the effective area of the wake behind any bluff body by energising the flow.

Howdy Mohsen Jahanmiri ,

Very interesting: of course. As in turbulence initiation by roughness, an acoustic pressure disturbance or just a local adverse pressure gradient emphasized by Oleh Grygorovych Shvydkyi in his post, can be a seed that grows in the shear of the flow. Acoustic energy seeds thus also distribute (not dissipate) the energy into small turbulent structures where they are needed to protect items from unbalanced forces. Large structures and separation will occur beyond the volume that needs that protection.

Happy Trails, Len

P. S. Note: If we "buy" this; if we agree that large turbulence structures grow from small ones; if we recognize in a cumulus cloud or a river that "swirls" of all sizes exist in the flow and interact; if we recognize that the slope of the "inertial subrange" may be derived by a simple geometric calculation of the number of "eddies" that fill the volume at each size: then we may see that L. F. Richardson's little ditty about "swirls," and the confidence in an energy cascade as the source of the inertial subrange slope, need exploration. The natural phenomenon of fluid turbulence is not yet fully known. The activity is called re-search because one must look again, since "they" missed a few details, or could not yet include them. New approaches are needed also, not instead. Knowledge is worth a careful second (or millionth) look. lfh

Dear Leonard,

In common turbulent flows large scale vortices are strained by velocity gradients (shear) which results in a transfer of energy from the main "steady" flow to the vortices. The large vortices strain smaller one in a cascade of the kinetic energy from large to smaller scales, until it reaches scales at which viscous effects are dominant (Kolmogorov length scale) . The kinetic energy is then dissipated by viscous stresses into heat. Those are the flows we are considering in engineering applications. While turbulence is a complex unsteady flow phenomenon, the occurrence of an energy cascade in the inertial region allows beautiful order of magnitude estimates of the rate at which turbulence dissipates energy. For most engineering problems this level of understanding is sufficient.

However, we do agree that nature is not always so simple.

Examples from atmospheric or oceanographic flows should for instance be considered with great care, because the rotation of the earth combined with the relatively limited depth (height) of the flow (10 km) compared to the large scales in such flows (1000 km) implies that these flows are dominated by two dimensional motion. In two dimensional flows one can observe the so-called inverse cascade in which smaller scales (100 km) drive larger scales (1000 km). Similar effects are observed in stars and galaxies. In such cases the energy cascade down to the Kolmogorov length scale is nonsense.

Regards

Mico

Howdy Mico,

All you write is good, fresh information to have for people who do not have a PhD in Meteorology and who have not read additional thousands of pages on the subject, as well as having invested many hours over almost 80 years seeking to comprehend what they see. (Info on RG and Berkeley SelectedWorks site.) When I write something that is not a quotation from the canon of a fluid mechanics religion it is because I disagree with portions of that faith and do not believe in it unquestioningly. Now, I maintain a strong awareness that I may need a "truth fix" but it should be after my thoughts have been considered as a possible alternative and ought not be replaced by the whole story. Your affirmation that "For most engineering problems this level of understanding is sufficient." is very good and "explains" why so many learned academics try to infect rebellious students with the doctrine. -- Your tone has been just fine, but being overwritten time after time is becoming tedious and I guess I'm a bit cranky and not all that civil, sigh. Sorry, but I have expressed my considered opinion.

As for the inertial subrange and shear strain of vortices and vortex stretching, of course they occur, but are you that all-fired certain that no dissipation occurs: ". . . until it reaches scales at which viscous effects are dominant (Kolmogorov length scale)." Further: "While turbulence is a complex unsteady flow phenomenon, the occurrence of an energy cascade in the inertial region allows beautiful order of magnitude estimates of the rate at which turbulence dissipates energy." Oh? For certain? ― Duc de la Rochefoucauld 1613-1680: “One of the tragedies of life is the murder of a beautiful theory by a brutal gang of facts.” Hah!! I had thought it was de la Rochefoucauld since I read it back in the 50's until I checked on my memory this last year and the inaccurate internet gave it to Benjamin Franklin! Bah! A sign on the Hospital Thrift Store wall: Don't believe everything you think. Turbulence is not alone in "not that simple." How did earth scale atmospheric phenomena get into this question? Physics folks say that mountains cause the weather system waves in the northern hemisphere. Bosh! The land-water thermal differences off Asia and North America have very strong affect in locking a north-easterly flow into the wave patterns there. Does that belong here (?), and no, it does not help to tell physicists about it. The 2D convenience for equation and algorithm mechanics may be good, but rapidly growing cumulus turbulence doesn't know about it. (I watch with care and awareness.)

Perhaps thoughtful treatment of Sanjibonny Buragohain's question would be better than all this. Oh, and I am sorta' sorry about the frustration leaks in my response - objective can be overdone.

Happy Trails, Len

Dear Dr. Hirschberg,

I appreciate your fundamental explanation regarding boundary layer,turbulence and separation. But I should remind you separation occurs as positive pressure gradient occurs (so called adverse pressure gradient) and not as you said: "Flow separation in a steady flow occurs when due to the shape of the body is such that a negative pressure gradient in the flow direction occurs". 

In unsteady flows the concept of flow separation is not easy to define. This is why I limited my arguments to steady flows. Of course strickly speaking a turbulent flow is not steady...However, the unsteadiness can be a local effect with a turbulent region.

When gravitational effects (combined with thermal gradients) and other force fields ( electromagnetic) are negligible, an adverse pressure gradient is the result of the shape of the body submerved in the “main” flow. This is the most relevant case when considering aerodynamics, which seens to be the subject of this discussion.

Howdy Avraham Hirschberg,

Thank you for a moderate reply to my enthusiasm. The question was about unsteady flows, and in fact a local effect of unsteadiness. When one reviews Sanjibonny Buragohain's RG profile one finds Investigation of Flow Regimes and Transition in a Lid-Driven Cavity: The Impact of Reynolds Number, and Analysis of Performance and Optimization Strategies for Horizontal Axis Wind Turbine (HAWT), as research reports from this last two months. Her list of skills and expertise indicates that an answer to the question asked would be understood and valued. The only aerodynamics there is in the wind turbine "airfoil" design to maximize lift of the blades. She certainly does not need primer lessons on Re or steady flows. If we answer, should not our emphasis be on the asked question?

While the norm of response in all issues of turbulent fluid dynamics is to read nature from accepted equations, it is not successful when the equations do not cover the nature of Nature there. This includes our century old confidence in the results of valiant efforts accurately to express any detail of this difficult subject. We shall have been pretty funny in 2323.

So, especially: In other words, can the phenomenon of boundary layer separation be effectively controlled and manipulated to minimize the formation and propagation of turbulent flow structures? It is better if the blades are not ripped off a wind turbine, or if the design has not become forbiddingly expensive to withstand untempered turbulence, equations notwithstanding.

Happy Trails, Len

???????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????????

Howdy Mico,

Excellent! Quite witty, in fact. It is that difficult a question.

Happy Trails, Len

Boundary layer serve the purpose of attaching the flow to the solid surface. For the shape of objects in practice laminar flow is more aerodynamically efficient. Yes the flow separation can definately be avoided by modifying the shape geometry. The problem is we are still trying to fit the old designs to new high speed conditions. Instead of innovating new designs to suit the objectives of modern world.

Dear Doctor

"Through our patented complex geometries and by taking advantage of additive manufacturing, we have increased efficiency of  additively manufactured heat exchangers. Our internal geometry encourages swirls and mixing which helps to reduce thickness of boundary layer within our heat exchanger core.  Anyone within the field of heat exchangers understands the concept of a boundary layer. A simple explanation is that when a fluid flows through a channel or on a surface, a thin layer is formed on the interface between the wall and the fluid. In this thin layer, fluid velocity changes from zero (static) at the wall surface to freestream velocity at a certain distance away from the wall surface. A thick boundary layer has a negative effect on heat exchanger performance as it impedes heat transfer. Think of it like a blanket, thicker the blanket, the higher the insulation. This is not ideal for heat exchangers core as the main objective is to conduct heat from fluid to wall surface and vice-versa. Turbulent flow is generally more preferred in heat exchanger design. The swirling and diffusive characteristics of turbulent flow enhances heat transfer. Mixing induced by turbulent flow can also disrupt the growth of boundary layer on heat exchanger core surfaces. However, turbulent flow is often associated with higher pressure drop.

As mentioned, reducing the size of boundary layer and promoting turbulent flow can enhance the performance of a heat exchanger. This can be achieved through the following: -Increase fluid velocity – Speeding up fluid flow through core is an effective way of increasing heat transfer. However, appropriate balance between heat transfer gain and pressure drop penalty needs to be considered. Having closely packed fins and multipass configuration can increase fluid velocity through core. -Mixing enhancement features – Addition of features such as turbulators, baffles, and corrugation can further enhance heat transfer. Conflux Technology has been producing superior heat exchangers with high heat transfer and minimum pressure drop. This is realized through our patented complex geometries, extensive parameters optimization, and by leveraging Additive Manufacturing process and its intrinsic surface roughness."

Hi Leonard Hall,

I am replying to your comment on the 9th of June (sorry for the late reply).

You asked for citations regarding my previous comment that "There are also techniques to increase the mixing between the wake and the free flow, to make the wake dissipate quicker (i.e. in a shorter distance behind the body); this is used for example to reduce interaction between wind turbines in a wind farm.".

I recently came across a short paragraph on this topic in the Wind Energy Handbook [1], in section "9.3.1 Wake control options", subsection "Other possibilities". Here they reference two additional publications [2] and [3]. I think this might be of interest to you and others.

Regarding the second part of your comment, I very much agree and I thank you for your insight.

Best regards, Ricardo

References

[1]: Burton, T. L., Jenkins, N., Bossanyi, E., Sharpe, D., & Graham, M. (2021). Wind energy handbook. John Wiley & Sons.

[2]: Munters, W., & Meyers, J. (2018). Towards practical dynamic induction control of wind farms: analysis of optimally controlled wind-farm boundary layers and sinusoidal induction control of first-row turbines. Wind Energy Science, 3(1), 409-425.

[3]: Frederik, J., Doekemeijer, B., Mulders, S., & van Wingerden, J. W. (2020, September). On wind farm wake mixing strategies using dynamic individual pitch control. In Journal of Physics: Conference Series (Vol. 1618, No. 2, p. 022050). IOP Publishing.

Howdy Ricardo,

I am in mourning at present; I shall likely return. Your post is good info for Sanjibonny. Thank you.

Happy Trails, Len

I am sorry to hear that Leonard. All the best to you and your loved ones.

Thank you. She and I planned for me to continue and I intend to give it a shot.

Howdy Ricardo, on tragedy and recovery of survivors I find "In action, watch the timing" from the Tao Te Ching quite significant. I'm up in the night deep in thought again. My thoughts are even on-thread if one considers how non-linear mind and fluid turbulence turn out to be. Separation of flow enables structure formation in three dimensions greater freedom, and input of another thought path disrupts stagnation when the time is right. This is what we had expected in my life after-her, to embrace living.

So, thank you for the delay because this timing was "right" to reopen living after healing enough from a loved one dying.

Happy Trails, Len