ฝัง
- เผยแพร่เมื่อ 25 ก.ค. 2024
- For more information, visit www.airshaper.com or email info@airshaper.com
----------------------------------------------------------------------------------------
In this video we’ll explain flow separation (or boundary layer separation).
In our previous video on boundary layers, we saw that the air close to the surface will stick to it, forming a boundary layer. But in some cases, it can become too difficult for the air to follow the curvature of the surface, causing it to detach. Let’s have a look at why this can happen.
When you look at air flowing across a flat plate, as in our previous video, the free stream velocity remains constant. Because of Bernoulli’s equation, which states that pressure changes with velocity, this means the pressure gradient in the X-direction is more or less to zero. Also, the pressure is imposed onto the boundary layer and so doesn’t really change much in the direction perpendicular to the surface either. That means the pressure gradient in both the x and y direction are close to zero.
But when the air flows around a curved surface, it speeds up and slows down. Take the flow around a cylinder for example: at the front, the pressure is the highest as the air comes to a complete standstill - this is the stagnation point. As the air then curves around the cylinder, it speeds up and this creates a drop in pressure. As the pressure goes down with increasing x, this is a negative pressure gradient, which is good: the air is being pushed downstream, overcoming the friction in the boundary layer, which gradually builds up.
The maximum velocity is reached somewhere around the midpoint of the cylinder, after which the air starts to slow down again. This means the pressure gradient is now reversed: as the air travels across the surface, the pressure goes up. This is what is called an “adverse” pressure gradient and the air is no longer driven but obstructed by the pressure difference: it has to flow into regions with higher pressure.
If the pressure gradient is positive or zero, as in the case of the flat plate, then the velocity gradient perpendicular to the wall is positive: the further you move away from the wall, the higher the velocity. But when the air travels against an adverse pressure gradient, the velocity profile is pushed back and this reduces the velocity gradient at the wall.
At some point, the velocity gradient becomes zero, which is the point where the flow separates or detaches. Beyond this point, the velocity profile is zero both at the wall, because of the no-slip condition, and at the inflection point, which is where the velocity crosses from negative into positive. In the negative velocity region, between the surface and the inflection point, the air flows in a direction opposite to the main flow - this is called a recirculation. The separated flow will now surf on top of this recirculation flow and the boundary layer and wake will continue to grow, which can have a negative effect on drag.
In reality, shapes can be much more complex than cylinders and flows can detach and reattach in various locations. Even on a smooth wing, the flow can detach and reattach. So, this is very much a 3-dimensional challenge and getting it right requires a careful analysis and optimization of local and global pressure gradients.
Also, keep in mind that flow separation can happen in both laminar and turbulent boundary layers. Turbulent boundary layers, however, carry more momentum and will typically stay attached to the surface further downstream compared to a laminar one - just check our video on golf ball dimples or vortex generators to learn more. - วิทยาศาสตร์และเทคโนโลยี
Kudos for your short and clear explanations, thank you
You're very welcome!
Superb explanation and diagrams - thanks a lot!
Thank you very much! Very good explanation!
Thank you!!
brief and informative . great video
Thanks!!
Excellent explanation!
(The video, though, featured very low volume, sometimes hard to listen to without headphones in a noisy environment)
Thank you very much Santiago!! Yes, we will fix the volume for the next recording :) Thanks for letting me know!
I like your explanation. Thanks man
You're very welcome!
Great video!
Thanks!
Very clear video
Thank you very much Raphael!
Best explanation
Amazing explanation. Could be a little louder !
Agree, we've turned up the volume on more recent videos
Hi. Your videos are really good and easy to understand. Can you make a video on how to start a career in aerodynamics/CFD?
Hi Manju,
thank you!
We have a blog exactly on that topic:
airshaper.com/blog/building-your-career-in-aerodynamics
Good luck!
@@AirShaper thank you.
nice video, I just want to ask on flow over plate, what is the meaning dp/dx=0??,. if dp/dx=0, then in flow over plate, separation will not occur even we extend the length of plate ??
Dear Fajar,
dP/dX = 0 relates to the overall flow, where the velocity is constant. Because of the bernoulli equation, this means the pressure is also constant.
However, you are right, this is a simple approximation. Because of friction (which indeed increases as the length of the plate increases) the pressure can indeed vary as you travel along the plate. And, as the boundary layer grows thicker, you can have a transition from laminar to turbulent at a certain point.
This video is also quite interesting
th-cam.com/video/5zLCofDSt_Q/w-d-xo.html
Would it be sensible to make surfaces more complex with sensors like shark skin that can adapt to the given positive or negative flow?
There is definitely a growing interest for "shape morphing" objects that can change shape in function of the aerodynamic goals. Or flexible materials, that deform under aerodynamic loads.
why does the curvature influences on the pressure gradient in streamwise direction?
Dear Josué,
I don't think I fully understand the question - if by streamwise direction you mean the flow lines, the pressure changes as the air slows down & speeds up (linked to the Bernoulli effect).
What do you mean with the streamwise direction?
What would be the best shape if I want to maximize friction
Hi David, we haven't had that question before :) Surface friction goes up with velocity, viscosity, surface area, surface roughness ... so those would be the parameters to work on. Keep in mind that you may trigger separation, creating a turbulent flow which has different friction properties as well.
@AirShaper I am working on a tesla turbine boundary layer turbine but with a cone or bowl shape to maximize friction instead of multiple discs.
@@David_Mash Then indeed the surface roughness is an important parameter :)
In the 2:05, you mentioned that " If the pressure gradient in the horizontal plane is positive or zero, velocity gradient in vertical is positive." but it looks in the image and from the knowledge it must be negative pressure gradient right?
@@denizcakar8285 I was thinking too the same and then came here to check if some one has commented.
Can you try to keep the audio levels the same for all your videos?, some are too high, some too low. Thanks.
Very true - we had this comment before and we switched to a new mic for newer videos. Should be better for more recent ones!
sound volume problem
True - we corrected this on more recent videos
Good! Exp..... But i want ur qualification and affiliation
You mean my educational background, company, ...?
@@AirShaper yes! PhD Msc ... and working company. university or other
@@gtnb8746
I don't feel I owe the community this information - the content needs to be good and speak for itself. Nevertheless:
Education
- Masters in mechanical engineering (with just the basics of fluid mechanics)
- Post-grad degree in product design
- Degree in car mechanics (4 years of after-work evening schooling)
- Self-study on deeper CFD principles
Experience
- 10 years of experience in OpenFOAM (CFD environment)
- 7 years as project manager for FEA/CFD simulations & wind tunnel testing at engineering agency (prior to founding AirShaper)
- 8 years of experience prototyping & developing the online simulation platform AirShaper (I'm the founder / CEO)
- Patented aerodynamics concept (see airshaper.com/cases/aquilo-patented-aerodynamic-concept-car)
- Occasional teaching on aerodynamics at universities / private companies
- ...
Why is the volume so low??
Fixed it for future videos!
That is quite a lot of Mathematics... I'm more of a real-world trial-and-error kind of guy, but this was still incredibly useful, nonetheless!
It's indeed fairly theoretical, but perhaps you can experiment in real life with some tufts on a car? We had fun doing so on this VW Beetle:
airshaper.com/videos/diy-aerodynamics-2-vortex-generators/MC6woj6tsQY
Mistake: If the pressure gradient is NEGATIVE or zero, then the velocity gradient in the vertical direction is a positive one.
low voice
That's true - we corrected this for more recent videos, thanks!
There's no flow hence no separation, pls always keep in mind.
I'm not sure I understand what you mean by "there is no flow"
@@AirShaper the air stays in its place but the aircraft moves on through static airmass.
@@synergy6294
That is true, but flow separation is about the relative velocity between the surface & the surrounding flow. So whether it is the wind moving across a static element, or an element moving through static air, the same principles apply.
@@AirShaper actually not
You can not make a parked plane takeoff no matter how much air the propellers flow over the wings on the tarmac. It will fly only when it moves :) and achieves v1... ok ?
Study the book on new zone theory of lift.
I cant hear anything
Should get that checked by a doctor
@@mcthunderstick2374 I went. He told me to get back to you. Do you know what did he say? You need to get check-in in a mental clinic.
Are u hiring interns?
Not at the moment, we'll post it on airshaper.com/jobs when we do. Best of luck Deepak.
@@AirShaper can't wait :)
Yo volume is low speak louder please
Fully agree - we bought a new microphone and increased the volume for later videos!