Latest / Mechanical Engineering Made Simple / (#144) Lesson 2: Laminar Lies vs Turbulent Truths

Transcript

(#144) Lesson 2: Laminar Lies vs Turbulent Truths · Mechanical Engineering Made Simple

• 0:00
  Welcome back to Mechanical Engineering Made Simple, Part 2
• 0:03
  of our five part series on fluid mechanics.
• 0:06
  Textbook theory assumes everything flows in nice, clean
• 0:09
  order. Reality runs on pure chaos, just
• 0:12
  like everything else in life. Laminar is the lie you start
• 0:15
  with, turbulence is the truth you end with.
• 0:18
  So when your pipes start acting up at 3:00 AM like that crazy X,
• 0:21
  just remember school gave you pretty streamlines and polite
• 0:24
  equations. The real world brought
• 0:27
  vibration, rough walls, heat and pressure spikes.
• 0:31
  Fluid mechanics isn't complicated, it's just honest.
• 0:34
  Hit like and follow. Drop a comment if you enjoy the
• 0:36
  episode. I love that intro, it really
• 0:38
  sets the stage for, well, the reality of what we do.
• 0:41
  Thanks. Yeah.
• 0:42
  And that honesty is exactly why we're doing this deep dive
• 0:44
  today. We're on a mission, you know?
• 0:46
  Oh. Absolutely a mission to bridge
• 0:47
  that gap. Right.
• 0:48
  We want to take those pristine, perfect equations you see
• 0:52
  written on a whiteboard and some, you know, university
• 0:54
  classroom, and we're dragging, dragging them kicking and
• 0:56
  screaming into the messy, chaotic reality of the
• 0:59
  industrial. World which is exactly where
• 1:00
  they belong. Yeah.
• 1:02
  So our goal for today's class is to break down the biggest real
• 1:05
  world issues you all keep running into out in the field.
• 1:09
  We're going to call out the common misconceptions that lead
• 1:11
  to honestly catastrophic failures, and we'll answer the
• 1:15
  most popular questions you are asking online about fluid
• 1:18
  behavior. It's a critical transition to
• 1:20
  make, frankly. It's where engineering becomes
• 1:24
  almost an art form. An art form, really.
• 1:27
  Yeah, I mean, because when you're sitting in a classroom,
• 1:29
  you were dealing with this idealized universe, you assume
• 1:33
  the fluid is perfectly behaved. Right, like the math just works
• 1:35
  out neatly. Exactly.
• 1:37
  You assume the pipe is geometrically perfect, and that
• 1:39
  time essentially stands still, so your calculations balance.
• 1:43
  But the real world doesn't care about our neat little math
• 1:45
  problem. Not at all.
• 1:47
  The moment that fluid enters a heavy hydraulic machine or a
• 1:50
  massive cross country pipeline, it stops being polite.
• 1:53
  The physics gets incredibly complex incredibly fast.
• 1:56
  And if we don't adjust for that, things break.
• 1:58
  Big time. We have to rethink our entire
• 2:02
  approach to fluid behavior when it leaves that theoretical
• 2:05
  space. Ignoring the reality of the
• 2:07
  fluid is exactly how you end up with shattered pipes, destroyed
• 2:11
  heat exchangers, and you know, $1,000,000 systems that just
• 2:15
  mysteriously fail to hit their performance specs.
• 2:18
  OK, let's unpack this. I want to start with the
• 2:21
  fundamental way fluids move and interact with themselves.
• 2:25
  Let's look at the first big lie we tell ourselves when we start
• 2:28
  engineering a system. I think I know which law you're
• 2:30
  talking about. Probably.
• 2:32
  But before we get to the lie, let's set the table.
• 2:35
  We all know the basics, right? Liquids are tough to compress.
• 2:37
  They have a free surface. Right, and gases are highly
• 2:40
  compressible and expand to fill whatever container you shove
• 2:43
  them into. Exactly.
• 2:45
  But the real magic, the thing that dictates how a system
• 2:47
  actually performs, comes down to viscosity, the fluid's internal
• 2:51
  resistance to flowing. Yeah, the friction of the fluid
• 2:54
  rubbing against itself. Right.
• 2:55
  And there is a classic visualization for this called
• 2:58
  the CUET Flow, right? Yes, CUET flow.
• 3:01
  It is the perfect mental model to understand viscosity because
• 3:05
  it strips away all the complexity of a round pipe.
• 3:08
  It just looks at the pure shearing action, right?
• 3:10
  Exactly. Imagine 2 completely flat
• 3:13
  parallel metal plates. The space between them, may be a
• 3:17
  gap of just a few millimeters, is filled with a fluid.
• 3:19
  Let's say it's oil. OK, so a thin layer of oil
• 3:23
  between two plates. Right now you bolt the bottom
• 3:25
  plate to the floor so it is perfectly still, and you start
• 3:28
  dragging the top plate horizontally at a constant
• 3:31
  steady speed. So I'm basically sliding one
• 3:33
  plate over the other with the oil acting like a well, like a
• 3:37
  slippery sandwich filling. That's a great way to picture
• 3:39
  it. Now we could zoom into the
• 3:41
  molecular level. We'd see something fascinating,
• 3:43
  the layer of oil molecules immediately touching the bottom
• 3:47
  stationary plate. It doesn't move at all.
• 3:50
  Wait, not. At all, Not even a little.
• 3:51
  It has zero velocity relative to that plate.
• 3:54
  It essentially locks onto the metal surface due to
• 3:56
  intermolecular forces. Oh, right.
• 3:59
  This is a fundamental concept we will come back to later, called
• 4:01
  the no slip condition. OK, so the bottom layer is
• 4:04
  stuck. Yes.
• 4:06
  Conversely, the very top layer of oil molecules locks onto the
• 4:10
  moving top plate, traveling at the exact same speed as that
• 4:13
  plate. So the bottom is stopped, the
• 4:15
  top is moving fast. What is happening to the oil in
• 4:18
  the middle? It creates a velocity gradient.
• 4:20
  It forms this straight diagonal line of increasing speed from
• 4:23
  the bottom to the top. Like a staircase of fluid.
• 4:26
  Sort of. Imagine the fluid is made-up of
• 4:29
  thousands of microscopic horizontally stacked sheets of
• 4:32
  paper. The top sheet moves fast,
• 4:35
  dragging the sheet below it slightly slower.
• 4:36
  Which drags the sheet below that slightly slower still, all the
• 4:40
  way down to 0. Precisely the force it takes to
• 4:42
  drag that top plate to overcome the friction between all those
• 4:45
  sliding sheets of fluid is your shearing stress.
• 4:48
  And how much force is required depends entirely on the fluid's
• 4:51
  absolute viscosity. Exactly.
• 4:53
  And if we map out that relationship between how hard we
• 4:56
  pull and how fast the fluid shears, we get into Newtonian
• 4:59
  versus non Newtonian fluids. Yes, the classic divide.
• 5:03
  For like water, air, or standard lubricating oil, they're
• 5:07
  Newtonian. That means the relationship is
• 5:09
  perfectly linear. Right, the shear stress is
• 5:11
  directly proportional to the velocity gradient.
• 5:14
  So if you push them twice as hard, they move twice as fast.
• 5:18
  They are predictable, reliable and mathematically friendly.
• 5:21
  They follow Newton's law of viscosity to the letter.
• 5:24
  But you know, the industrial world isn't just pumping cure
• 5:28
  water. Far from it.
• 5:29
  Right. We pump liquid pulp and paper
• 5:31
  mills. We pump raw asphalt.
• 5:33
  We pump drilling muds, high molecular weight polymer
• 5:36
  solutions, chemical slurries. And those do not play nice.
• 5:39
  No they don't. These are non Newtonian fluids
• 5:42
  and they do not care about Newton's clean linear equations.
• 5:46
  If you look at a real logical diagram for these fluids, they
• 5:49
  do wild things. I always think of Newtonian
• 5:52
  fluids like a predictable employee.
• 5:54
  Push them twice as hard, they work twice as fast.
• 5:56
  I like that analogy. Yeah, and non Newtonian fluids
• 5:58
  are like a Moody artist. Push them hard and they might
• 6:01
  just turn solid to spite you. Yes, like a sheer thickening
• 6:05
  fluid where the harder you hit it the more viscous it becomes.
• 6:08
  Like cornstarch and water? Oobleck, right?
• 6:10
  Exactly. If you stir it slowly, it acts
• 6:13
  like a thin liquid. If you punch it, it acts like
• 6:17
  solid concrete. That stuff is crazy.
• 6:20
  It is then you have sheer thinning fluids like ketchup or
• 6:24
  certain types of paint. They are incredibly viscous,
• 6:27
  almost solid until you apply a heavy shear force.
• 6:31
  Right, like smacking the bottom of the ketchup bottle.
• 6:34
  Exactly. At that point their viscosity
• 6:36
  collapses and they flow easily. Which makes a mess on your
• 6:39
  fries. Right.
• 6:40
  And then you have Bingham plastics like toothpaste or
• 6:43
  heavy sludge which act entirely as a solid, resisting all
• 6:47
  movement until you apply a massive initial threshold force.
• 6:51
  A yield stress. Right, yes, a yield stress and
• 6:53
  only then do they break down and start flowing like a liquid.
• 6:56
  Man, it's a nightmare to design A pump for something that
• 6:59
  constantly changes its own thickness based on how hard the
• 7:02
  pump is pushing it. It takes a lot of specialized
• 7:04
  engineering. That's for.
• 7:05
  Sure, but that brings me to the massive trap I see engineers
• 7:08
  fall into the lie I mentioned earlier.
• 7:11
  The incompressible lie. Yes, the idea that water is
• 7:15
  completely incompressible. We treat liquids as if their
• 7:19
  volume is a universal constant, completely immune to pressure.
• 7:23
  And that assumption is incredibly dangerous.
• 7:25
  Is it really that bad though? Well, yes and no.
• 7:29
  For everyday municipal plumbing or a low pressure cooling loop,
• 7:33
  assuming water is incompressible is a perfectly acceptable
• 7:36
  mathematical shortcut. Because the change in volume is
• 7:40
  negligible. Right, but textbook shortcuts
• 7:42
  don't survive in high pressure engineering.
• 7:45
  If you are working with heavy hydraulic machinery, deep sea
• 7:48
  submersibles, extreme pressure injection systems, the fluids
• 7:52
  compressibility absolutely must be accounted for.
• 7:55
  So liquids actually do compress. They do compress, they just
• 7:58
  fight back incredibly hard when you try to do it.
• 8:00
  We measure that resistance, that stiffness of the fluid, using a
• 8:04
  property called the bulk modulus.
• 8:06
  OK, I have to push back here on behalf of every mechanical
• 8:09
  engineering student who just spent 3 semesters cancelling out
• 8:12
  the density variable in their equations because liquids don't
• 8:14
  compress. I know it hurts to you.
• 8:16
  Wait, water shrinks? If I have a massive hydraulic
• 8:19
  cylinder, the oil inside is physically getting smaller.
• 8:22
  It physically gets smaller. The molecules are being forced
• 8:26
  closer together, compressing the microscopic void spaces between
• 8:29
  them. That is wild to think about.
• 8:32
  It is, and the consequences of ignoring this are severe.
• 8:36
  Let's look at the mechanical implications.
• 8:38
  Say you are designing a precision hydraulic actuator for
• 8:41
  an automated manufacturing line. OK, so something that needs to
• 8:45
  be super accurate. Exactly.
• 8:47
  It needs to extend a robotic arm exactly 3mm to insert a
• 8:50
  microchip. If you calculate the required
• 8:53
  fluid volume assuming absolute incompressibility, you will be
• 8:57
  wrong. Because under high pressure, the
• 8:59
  volume of your hydraulic oil physically shrinks.
• 9:01
  Yes, your actuator won't extend as far as the math predicted.
• 9:05
  The pressure forces the fluid to compress before it forces the
• 9:08
  piston to move. Oh wow, so you'd missed the
• 9:10
  target completely? It might only be a fraction of a
• 9:12
  percent of shrinkage, but at high industrial tolerances it
• 9:15
  entirely destroys your precision.
• 9:17
  But how much pressure are we actually talking about here?
• 9:20
  To shrink the volume of water by just 1% at standard room
• 9:24
  temperature, you need a massive amount of force.
• 9:27
  Right. You need roughly 20 million
• 9:28
  Pascals, or about 290 atmospheres of pressure, to
• 9:33
  shrink water by 1%. OK, that sounds astronomically
• 9:37
  high to a layperson. It does, but in heavy machinery,
• 9:41
  earth movers, high pressure water jet cutters, or deep well
• 9:45
  oil drilling operations, you hit and exceed those numbers
• 9:48
  routinely. And it's not just volume
• 9:51
  calculations that get thrown off.
• 9:52
  Compressibility stores energy. Like a spring.
• 9:55
  Exactly like a giant, very stiff mechanical spring.
• 9:58
  If a valve suddenly fails, that compressed liquid expands
• 10:02
  violently, releasing all that stored spring energy at once,
• 10:05
  which blows out seals and shatters casings.
• 10:08
  So as an engineer, at what exact point do I have to stop using
• 10:11
  the easy math a good? Question.
• 10:12
  Like when does a liquid cross the line and force me to account
• 10:16
  for its compressibility? It comes down to comparing the
• 10:18
  speed of the fluid to the speed of sound within that fluid.
• 10:22
  This ratio is the Mach number. Mach number right?
• 10:26
  Usually hear that with jets. Right, but it applies to liquids
• 10:29
  too. When the flow velocity starts
• 10:32
  approaching roughly .3 Mach, or 30% fraction of the local speed
• 10:36
  of sound, the compressibility effects become too large to
• 10:40
  sweep under the rug. But for liquids, the speed of
• 10:43
  sound is extremely high, isn't it?
• 10:45
  Very high in water, sound travels at roughly 1480 meters
• 10:50
  per second because the molecules are so tightly packed.
• 10:52
  So you generally only hit these Mach numbers at extreme
• 10:55
  pressures. Exactly.
• 10:57
  Or in highly specialized applications like diesel fuel
• 10:59
  injectors, where fuel is blasted through microscopic nozzles at
• 11:03
  insane velocities. But the threshold exists.
• 11:06
  It does, and assuming a fluid is perfectly rigid just because
• 11:10
  it's a liquid is a great way to design a system that tears
• 11:13
  itself apart. And that leads us perfectly into
• 11:15
  the next massive theoretical illusion we have to shatter.
• 11:18
  Which one? The idea that fluids want to
• 11:20
  flow in polite, orderly lines. Let's talk about the laminar lie
• 11:25
  and the turbulent truth. Yes, this is perhaps the most
• 11:28
  fundamental division in all of fluid dynamics.
• 11:31
  Right, because flow exists in two primary regimes.
• 11:34
  Yes, louner flow is the ideal. It is smooth, it is quiet, and
• 11:39
  the fluid particles move in nice straight parallel streams.
• 11:42
  Like a deck of cards sliding over each other smoothly.
• 11:45
  Perfect analogy. There is absolutely no lateral
• 11:48
  mixing. Fluid at the top of the pipe
• 11:49
  stays at the top, Fluid at the bottom stays at the bottom.
• 11:52
  But turbulent flow is the reality of the universe.
• 11:55
  Always it is chaotic. You have wild lateral motions,
• 11:59
  swirling microscopic whirlpools called Eddie's superimposed on
• 12:02
  the main forward flow, and highly irregular, unpredictable
• 12:06
  fluctuations in velocity and pressure.
• 12:08
  The history of how we mathematically define this is
• 12:11
  absolutely fascinating. Osborne, Reynolds, right.
• 12:13
  Yes, back in the late 1800s, a Victorian era engineer named
• 12:17
  Osborne Reynolds set up this brilliantly elegant experiment
• 12:21
  in Manchester. It really was an elegant setup.
• 12:23
  He took a long straight glass tube, connected it to a water
• 12:27
  tank, and ran Clearwater through it.
• 12:29
  Then he took a tiny, razor thin needle and injected a steady
• 12:32
  stream of highly visible colored dye right into the dead center
• 12:36
  of the flowing water. Such a simple but effective
• 12:39
  visual. Totally.
• 12:40
  When the water was moving slowly, the dye formed a perfect
• 12:44
  unmoving thread like line straight down the center of the
• 12:48
  tube. It looked like a solid glass
• 12:50
  wire suspended in the water. That was pure laminar flow.
• 12:53
  Exactly because the viscous forces of the water, that
• 12:57
  internal friction holding the layers together, were strong
• 13:00
  enough to keep all the water molecules marching in a
• 13:03
  perfectly straight line. The dye had nowhere to go but
• 13:06
  straightforward. But then Reynolds started
• 13:08
  opening the valve. He increased the velocity of the
• 13:10
  water. And this is where things get
• 13:11
  interesting. Right.
• 13:12
  For a while, the die line stayed straight, but suddenly, at a
• 13:16
  very specific speed, that pristine straight line of dye
• 13:20
  just shattered. It just broke apart.
• 13:22
  It began to vibrate, then it whipped back and forth, and then
• 13:25
  it mixed, fixed violently with the surrounding Clearwater,
• 13:29
  instantly turning the whole tube a cloudy, chaotic color.
• 13:32
  It was like a switch had been flipped from order to absolute
• 13:35
  chaos. What Reynolds discovered was
• 13:37
  that this transition wasn't arbitrary.
• 13:40
  It was a battle between two competing forces in the fluid.
• 13:43
  Yes, on one side you have inertial forces.
• 13:46
  This is the momentum of the moving fluid proportional to its
• 13:50
  density and velocity, which wants to push the fluid forward
• 13:53
  but also makes it prone to wandering off course if
• 13:55
  disturbed. And on the other side.
• 13:57
  You have viscous forces. This is the fluid's internal
• 14:00
  friction, which acts like glue, trying to dampen out any
• 14:04
  wandering and hold the layers together.
• 14:05
  And he formalized this battle into the most famous
• 14:08
  dimensionless number in engineering.
• 14:11
  The Reynolds number. Right, it is essentially the
• 14:13
  ratio of those inertial forces divided by the viscous forces.
• 14:17
  Mathematically, it's the fluids velocity multiplied by the
• 14:20
  diameter of the pipe divided by the fluids kinematic viscosity.
• 14:25
  So if the Reynolds number is low, viscosity wins, the fluid
• 14:29
  acts like thick syrup and you get smooth laminar flow.
• 14:32
  Exactly. And if the Reynolds number is
• 14:33
  high, inertia wins, the glue fails, and you get chaotic
• 14:38
  turbulent flow. And this is where the
• 14:39
  engineering heartbreak happens because we absolutely love
• 14:43
  laminar flow. We desperately want our systems
• 14:45
  to operate in the laminar regime.
• 14:46
  We really do. Because the pressure drop is
• 14:48
  drastically lower, right? Yes, the friction is manageable.
• 14:52
  It requires significantly less pumping power, which means
• 14:55
  smaller pumps, less electricity, and millions of dollars saved
• 14:59
  over the life of an industrial plant.
• 15:01
  But in real world industrial pipes, flow is almost always
• 15:04
  turbulent. Almost always.
• 15:06
  Which brings up a massive area of confusion I see online.
• 15:09
  A lot of textbooks and professors will hand you a
• 15:12
  single hard number. They will declare
• 15:15
  authoritatively that if the Reynolds number is below 2320
• 15:20
  the flow is laminar and if it is above that it is turbulent.
• 15:24
  They treat it like a hard border crossing.
• 15:26
  And that is a gross oversimplification. 2320 is
• 15:31
  generally accepted as the lower critical limit.
• 15:33
  Meaning what exactly? Meaning below that number,
• 15:36
  viscous forces are so dominant that even if you intentionally
• 15:39
  violently stir the water before it enters the pipe, the fluid
• 15:42
  will eventually calm itself down and return to a laminar state.
• 15:45
  So the viscosity dampens out the chaos.
• 15:48
  Exactly, but the upper limit, the point where a perfectly
• 15:52
  smooth flow must become turbulent.
• 15:55
  That limit is almost infinitely elastic, depending on your
• 15:58
  laboratory conditions. Hold on.
• 15:59
  So it's not a hard rule. If I have a Reynolds number of
• 16:02
  say 10,000, I could theoretically still have smooth
• 16:06
  glassy laminar flow. Theoretically, yes.
• 16:10
  An experimenter named Ekman actually managed to keep flow
• 16:13
  completely laminar up to a Reynolds number of 50,000.
• 16:16
  Wait. 50,000 Yes. How is that physically possible
• 16:20
  to push the math that far past the breaking point?
• 16:23
  By creating a completely artificial, hyper isolated
• 16:26
  environment, Ekman had to ensure the upstream water in the supply
• 16:29
  tank was perfectly absolutely still.
• 16:31
  Like no vibrations at all. None left to settle for days, so
• 16:35
  not a single convection current remained.
• 16:38
  The entrance to the glass tube had to be geometrically
• 16:40
  flawless, with a perfectly flared bell mouth to prevent the
• 16:43
  slightest separation. That sounds intense.
• 16:46
  He even isolated the entire apparatus from the microscopic
• 16:49
  vibrations of the building itself.
• 16:51
  If you remove absolutely all external disturbances, the
• 16:55
  viscous forces can miraculously hold the laminar layers together
• 16:59
  under immense velocity. But I'm guessing it's a super
• 17:01
  fragile state. Hyper fragile state at a
• 17:03
  Reynolds number of 50,000. The fluid is practically
• 17:06
  screaming to become turbulent the moment a microscopic
• 17:09
  vibration, a truck driving by outside a tiny temperature
• 17:13
  gradient, hits the system. The inertial forces seize upon
• 17:16
  that flaw and the flow flashes into violent turbulence
• 17:19
  instantly. A great everyday visual for this
• 17:22
  fragile transition is a lit cigarette resting in an ashtray
• 17:26
  in a perfectly still room. Oh, that's a brilliant example.
• 17:29
  Yeah, the smoke rising from the cherry is incredibly hot, so it
• 17:32
  accelerates upward for the first few inches.
• 17:34
  It rises in a pristine, perfectly straight blue ribbon
• 17:38
  diameter flow. Because the Reynolds number is
• 17:39
  still low. Right.
• 17:40
  But as it accelerates, the Reynolds number increases,
• 17:43
  inertia builds, and suddenly, for seemingly no reason, that
• 17:46
  straight ribbon buckles, folds and breaks into a WAVY,
• 17:50
  swirling, chaotic cloud of grey smoke.
• 17:53
  The microscopic air currents in the room finally tripped the
• 17:55
  fragile lameter balance into turbulence.
• 17:57
  Exactly. Here's where it gets really
• 18:00
  interesting though. If laminar flow is so incredibly
• 18:03
  efficient and saves us millions on pump sizing, why can't we
• 18:07
  just engineer our way out of this?
• 18:10
  What do you? Mean why can't I just
• 18:11
  manufacture an ultra smooth, perfectly polished titanium
• 18:15
  pipe, isolate my pumps with rubber dampeners, and force the
• 18:19
  fluid to stay laminar even at high velocities?
• 18:21
  It is a brilliant thought experiment, but the universe
• 18:24
  simply won't allow it because of the fundamental nature of
• 18:27
  isotropic turbulence. So they can't beat it.
• 18:29
  No, because turbulence isn't just fluid sloshing around.
• 18:33
  It is a cascade of kinetic energy transferring through a
• 18:36
  mass of hierarchy of swirling eddies.
• 18:39
  OK, walk me through that. In a turbulent flow, you have
• 18:42
  large macro eddies the size of the pipe diameter.
• 18:45
  These are unstable, so they break apart into smaller eddies,
• 18:48
  which break into even smaller microscopic eddies.
• 18:51
  Passing the kinetic energy all the way down.
• 18:52
  Yes, until it is finally destroyed by viscosity and
• 18:55
  turned into heat. This is an inherent property of
• 18:58
  fluids at high momentum. So even with a perfect pipe, the
• 19:01
  fluid itself is the problem. Exactly.
• 19:04
  Even if you manufacture a pipe that is atom perfectly smooth,
• 19:07
  you cannot control the molecular and thermodynamic perturbations
• 19:10
  within the fluid itself. Like tiny heat changes.
• 19:13
  Yeah, tiny fluctuation of the local fluid density, a
• 19:16
  microscopic variance in heat, creates A microscopic lateral
• 19:19
  movement. At high Reynolds numbers, the
• 19:22
  inertial forces are aggressively looking for a reason to break
• 19:25
  the layers. They just want an excuse to go
• 19:27
  crazy. Exactly.
• 19:28
  That tiny thermal perturbation amplifies.
• 19:31
  It trips the adjacent layer, creating a microscopic Eddy.
• 19:35
  That Eddie trips the next layer and suddenly that chaos cascades
• 19:39
  outward in milliseconds until the entire flow is a turbulent
• 19:42
  mess. So you really cannot force a
• 19:44
  high Reynolds number flow to stay laminar in a macroscopic
• 19:47
  industrial setting. The underlying thermodynamics of
• 19:49
  the universe actively fight you. Man that's rough.
• 19:53
  So ultra smooth pipes won't save us from turbulence, but talking
• 19:57
  about the smoothness of the pipe Tees us up perfectly for our
• 20:00
  next massive hurdle. The wall itself.
• 20:03
  Right, because while we've been talking about the fluid in the
• 20:06
  center of the pipe, the fluid that is actually physically
• 20:09
  touching the metal wall is playing by a totally different
• 20:12
  set of rules. We need to explore the hidden
• 20:16
  friction, boundary layers and flow separation.
• 20:19
  This is where things get mechanically punishing.
• 20:22
  To understand pipe friction drag on an airplane wing, or why a
• 20:25
  boat hull needs constant cleaning, we have to define the
• 20:28
  boundary layer. Right whenever a viscous fluid
• 20:31
  flows past a solid boundary, a strange physical phenomenon
• 20:35
  occurs. The fluid velocity directly
• 20:38
  microscopically adjacent to the wall is absolutely zero.
• 20:41
  It stops entirely. This is that no slip condition
• 20:44
  you mentioned earlier. And to be clear, it's not just
• 20:47
  moving very slowly, it is completely, perfectly stationary
• 20:50
  relative to the pipe. Completely stationary.
• 20:52
  The fluid molecules are literally captured by the
• 20:55
  intermolecular forces and microscopic roughness of the
• 20:57
  solid surface. So you have a layer of fluid at
• 21:00
  0 velocity. But just a millimeter away, the
• 21:03
  main body of the fluid is rushing past at, say, 10 meters
• 21:07
  per second. Right, which means there must be
• 21:09
  a zone of transition from that stationary wall moving outward
• 21:13
  toward the center of the pipe. The fluid velocity gradually
• 21:17
  steeply increases layer by sliding layer.
• 21:21
  Until it finally catches up to the free stream speed of the
• 21:24
  main flow. Exactly that entire transition
• 21:27
  region of retarded friction. Heavy flow from zero velocity at
• 21:32
  the wall up to 99% of the free stream velocity is the boundary
• 21:37
  layer. And as the fluid travels further
• 21:39
  down the length of the pipe or further back along the flat hull
• 21:42
  of a ship, that boundary layer doesn't stay the same size,
• 21:45
  right? It grows.
• 21:46
  It does. The friction reaches further and
• 21:48
  further out into the main flow, dragging more and more fluid
• 21:50
  down. We mathematically quantify this
• 21:53
  growth using a couple of distinct measurements.
• 21:55
  There is displacement thickness, which is a brilliant conceptual
• 21:58
  tool. It really is.
• 22:00
  It essentially asks because the boundary layer is slowing the
• 22:02
  fluid down, how much would we have to physically shrink the
• 22:05
  diameter of the pipe to end up with the exact same reduction in
• 22:08
  total flow rate? It represents the mass flow
• 22:10
  deficit. Precisely.
• 22:12
  And then you have momentum thickness, which is similar but
• 22:15
  measures the deficit of physical momentum caused by that dragging
• 22:19
  friction. To actually calculate the
• 22:20
  mechanical energy lost to this dragging, the pressure drop or
• 22:25
  head loss in the pipe, engineers rely on the Darcy Weisbach
• 22:29
  equation. It is the gold standard.
• 22:32
  What goes into that equation? It factors in the length of the
• 22:35
  pipe, the diameter, the fluid velocity, and a crucial variable
• 22:39
  called the friction factor. To find that exact friction
• 22:42
  factor for a turbulent flow, we use the Colebrook white
• 22:46
  equation. Which is an absolute
• 22:48
  mathematical nightmare because it is an implicit equation.
• 22:51
  Yes, it is not fun to solve by hand.
• 22:53
  The friction factor is on both sides of the equal sign, so you
• 22:56
  can't just solve it algebraically.
• 22:58
  You have to guess an answer, plug it in, see how wrong you
• 23:00
  are, and iterate closer and closer to the truth.
• 23:03
  Or you just cheat and look at a Moody diagram like the rest of.
• 23:05
  US crew the Moody diagram is a lifesaver it maps out the
• 23:09
  relationship between the Reynolds number, the relative
• 23:11
  roughness of the pipe wall, and the resulting fiction factor but
• 23:14
  reading. That diagram brings us to a
• 23:16
  massively counterintuitive reality in fluid dynamics.
• 23:20
  It's an assumption that trips up junior engineers constantly.
• 23:23
  Let me guess, a rough pipe always causes more drag than a
• 23:27
  smooth pipe? Exactly, it feels like common
• 23:30
  sense. I mean, if I rub my hand over
• 23:31
  sandpaper it causes way more friction than rubbing my hand
• 23:34
  over a pain of glass. So a corroded rough cast iron
• 23:38
  pipe should obviously cause more fluid drag than a brand new
• 23:43
  polished PVC pipe, right? Sounds perfectly logical, but
• 23:46
  fluid mechanics doesn't care about human intuition.
• 23:49
  So why is that statement false? To understand why it's false
• 23:52
  under certain conditions, we have to pull out our microscopes
• 23:55
  and look deep inside the boundary layer, specifically at
• 23:59
  a sub region called the viscous sub layer.
• 24:00
  OK, what's happening down there? Even if a pipe is experiencing
• 24:03
  violently chaotic turbulent flow in its center, there is a
• 24:07
  microscopic layer of fluid directly against the pipe wall
• 24:10
  that is moving so slowly, so close to 0 velocity that the
• 24:14
  viscous forces dominated. So it remains purely laminar.
• 24:17
  It does this ultra thin smooth sheet of fluid hugging the wall
• 24:23
  is the viscous sub layer. So I have a chaotic turbulent
• 24:26
  storm in the middle of the pipe, but right at the metal surface
• 24:30
  there's a microscopic safe haven of perfectly smooth laminar
• 24:33
  flow. Exactly how does this explain
• 24:35
  the roughness illusion? Because hydrodynamic smoothness
• 24:38
  is not about how the metal feels to your thumb, it's about how
• 24:42
  the metal interacts with that sub layer.
• 24:44
  Meaning the bumps and ridges. Yes, a pipe is mathematically
• 24:48
  considered hydraulically smooth if it's physical roughness
• 24:52
  elements, the microscopic manufacturing bumps, the pits,
• 24:56
  the crystalline grain of the metal are physically shorter
• 24:59
  than the thickness of the viscous sub layer.
• 25:01
  Oh I see. If those bumps are entirely
• 25:02
  submerged inside that laminar sheet of fluid, the turbulent
• 25:05
  main flow never physically sees them.
• 25:07
  That is wild. Right.
• 25:09
  The chaotic flow just glides effortlessly over the laminar
• 25:12
  sub layer as if the pipe were made of frictionless glass.
• 25:15
  The fluid creates its own protective coating over the
• 25:17
  rough metal. Exactly.
• 25:19
  Physical roughness only increases your pressure drug if
• 25:22
  the Reynolds number gets so high and the flow gets so violently
• 25:26
  fast that the viscous sub layer is compressed and thinned out.
• 25:30
  And then what happens? If the sub layer gets thinner
• 25:32
  than the height of the metal bumps, those microscopic jagged
• 25:35
  peaks suddenly poke out of the laminar safe haven and penetrate
• 25:39
  into the chaotic, turbulent region.
• 25:41
  And the turbulence grabs them. Yes, the turbulent eddies crash
• 25:45
  into the jagged metal peaks causing massive drag.
• 25:48
  That is the moment of clarity for many engineers.
• 25:51
  You can have a pipe that feels like gravel to the touch, but at
• 25:54
  low flow velocities it acts as perfectly smooth hydro
• 25:58
  dynamically as polished steel. I love mapping this out
• 26:01
  visually. I like to compare the boundary
• 26:03
  layer to traffic on a massive multi lane highway.
• 26:06
  Go on. The breakdown lane is the solid
• 26:08
  pipe wall. Zero velocity.
• 26:10
  The lane right next to it is crawling at 5 miles an hour.
• 26:13
  That is the viscous sub layer moving slow and steady.
• 26:17
  OK, I follow. The middle lanes are going the
• 26:19
  speed limit and the far left lane is flying at 80 miles an
• 26:22
  hour. Now if every single car stayed
• 26:25
  perfectly centered in their own lane, never merging, that is
• 26:29
  laminar flow. The only friction is just the
• 26:31
  slight aerodynamic drag between the lanes.
• 26:34
  But highways are never like that.
• 26:35
  Never. In reality, a highway is
• 26:38
  turbulent. Cars are constantly aggressively
• 26:41
  changing lanes. A slow semi truck from the right
• 26:44
  lane merges directly into the fast left lane, forcing the
• 26:47
  speeding sports cars to slam on their brakes, instantly
• 26:50
  destroying the forward momentum. Yes, and conversely.
• 26:53
  Meanwhile, a fast car swerves aggressively into the slow right
• 26:57
  lane and rear ends someone. This chaotic lateral lane
• 27:00
  changing transfers momentum back and forth violently.
• 27:03
  It creates massive overall friction and slows the entire
• 27:06
  highway down. And in fluid dynamics, that
• 27:09
  macroscopic lane changing friction caused by turbulent
• 27:12
  eddies smashing into each other is called Reynolds stresses.
• 27:15
  That's an exceptionally accurate way to put it.
• 27:17
  Those Reynolds stresses, the macroscopic exchange of momentum
• 27:21
  between layers, are orders of magnitude larger and more
• 27:24
  destructive to your system's energy than a simple molecular
• 27:28
  viscous friction found in laminar flow.
• 27:30
  But as much energy as boundary layers and Reynolds stresses
• 27:33
  consume, there is a secondary phenomenon associated with the
• 27:37
  wall that is a true system killer.
• 27:40
  Flow separation. Yes, the ultimate manifestation
• 27:43
  of fluid friction. This is when the fluid just
• 27:46
  decides it has had enough of rubbing against the wall, gives
• 27:49
  up, and literally tears itself away from the boundary entirely.
• 27:52
  Exactly, and it happens when the fluid encounters an aerodynamic
• 27:55
  enemy known as an adverse pressure gradient.
• 27:59
  What does that look like in a pipe?
• 28:00
  Let's imagine a fluid flowing smoothly through a pipe.
• 28:03
  Suddenly, the pipe expands into a much larger diameter as the
• 28:07
  cross-sectional area increases. The laws of conservation of mass
• 28:11
  dictate that the fluid's velocity must decrease to
• 28:14
  maintain the same volumetric flow rate.
• 28:16
  Right. A wider pipe means slower fluid.
• 28:18
  Now, according to Bernoulli's principle, which we will dissect
• 28:21
  later, if the velocity decreases, the static pressure
• 28:23
  of the fluid must increase. OK, so the fluid is slowing
• 28:26
  down, but it's pressures going up.
• 28:28
  Right. And that means the fluid is now
• 28:30
  trying to flow forward into a region of higher pressure than
• 28:34
  the region it just left. It is essentially swimming
• 28:37
  upstream against a physical wall of pressure.
• 28:40
  This is the adverse gradient. Yes, the main turbulent flow in
• 28:43
  the center of the pipe has enough kinetic momentum to smash
• 28:47
  through that high pressure wall and keep moving forward.
• 28:49
  But the fluid trapped in the boundary layer.
• 28:52
  It has already been robbed of its momentum by the friction of
• 28:55
  the wall. It is exhausted.
• 28:57
  As it hits that high pressure gradient, it slows down rapidly.
• 29:01
  Eventually the forward velocity of the fluid touching the wall
• 29:05
  drops to absolute 0. And what happens when it hits
• 29:07
  zero? It can't just stay there while
• 29:09
  millions of gallons of fluid are pushing behind it.
• 29:12
  It doesn't because the pressure downstream is now higher than
• 29:15
  the pressure at the stagnation point.
• 29:17
  The fluid actually reverses direction.
• 29:20
  It starts flowing backward along the wall.
• 29:23
  When this backward flowing boundary layer collides with the
• 29:26
  forward flowing boundary layer coming from upstream, the fluid
• 29:29
  has nowhere to go but violently inward away from the wall.
• 29:33
  It physically tears itself off the metal boundary.
• 29:36
  And when it separates, it doesn't leave a vacuum, it
• 29:38
  creates massive energy sucking eddies and swirling vortices in
• 29:42
  the empty space behind the separation point.
• 29:45
  It's extremely destructive to efficiency.
• 29:47
  It's exactly like tailgating a massive semi truck on the
• 29:51
  highway. Back to the highway.
• 29:52
  Always the flat front of the truck punches a blunt hole in
• 29:56
  the air. The air flow simply cannot make
• 29:59
  the sharp 90° turn around the back corners of the trailer, so
• 30:02
  the flow separates. It leaves this massive low
• 30:06
  pressure wake of swirling chaotic air right behind the
• 30:10
  truck doors. And you can feel that, right?
• 30:12
  Yeah. If you are tailgating in your
• 30:14
  car, you actually feel your vehicle get sucked forward into
• 30:17
  that low pressure wake in a pipe system.
• 30:19
  Whether it's a sudden expansion or an elbow joint, those
• 30:22
  separation wakes act like black holes for mechanical energy.
• 30:25
  They completely tank your system's efficiency and force
• 30:29
  your pumps to work twice as hard to maintain flow.
• 30:31
  But here is where we need to shift gears.
• 30:34
  Friction, boundary layers, and separation are all about the
• 30:37
  fluid dragging against the pipe while it is actively moving.
• 30:41
  But what happens when the fluid suddenly stops moving entirely?
• 30:45
  What happens when all that moving mass essentially throws a
• 30:47
  punch at the pipe itself? That brings us to the terrifying
• 30:51
  phenomenon of unsteady flow and water hammer.
• 30:55
  Up until this point in our discussion, we have been
• 30:57
  existing in a state of steady flow.
• 30:59
  Which is the nice polite textbook version.
• 31:01
  Exactly. Steady flow is the bedrock of
• 31:03
  most engineering calculations. It means if you pick a specific
• 31:07
  coordinate point in your pipe system, the conditions at that
• 31:11
  exact point, the velocity of the water passing through it, the
• 31:14
  static pressure, the density do not change as time ticks
• 31:17
  forward. It is a constant stable state of
• 31:20
  operation. Unsteady flow, on the other
• 31:22
  hand, means those parameters are actively dynamically changing
• 31:26
  over time. And the most dramatic,
• 31:29
  destructive and sudden example of unsteady flow is water
• 31:32
  hammer, or as it's known academically, hydraulic shock.
• 31:36
  Let's visualize this. Picture a long, straight steel
• 31:39
  pipe carrying a massive volume of water at a high velocity.
• 31:43
  At the very end of this pipe is a heavy mechanical valve.
• 31:47
  If you take a lever and suddenly slam that valve shut in a
• 31:50
  fraction of a second, the water directly immediately behind the
• 31:53
  valve gate hits a literal brick wall.
• 31:56
  It's velocity abruptly drops from 10 meters per second down
• 31:59
  to absolute zero in an instant. The water has mass.
• 32:03
  Tons of it. And because it was moving, it
• 32:05
  possessed a massive amount of kinetic energy.
• 32:07
  By the fundamental laws of thermodynamics, that massive
• 32:10
  kinetic energy cannot just be deleted from the universe, it
• 32:13
  has to be conserved. And because it can no longer
• 32:15
  exist as the energy of motion, it is instantaneously converted
• 32:18
  into the energy of compression. You get a massive localized
• 32:22
  violent pressure spike right at the face of the closed valve.
• 32:25
  Like the fluid just crumples into itself.
• 32:27
  Exactly. The water molecules physically
• 32:29
  crash into each other, compressing slightly, and the
• 32:31
  steel pipe wall expands out word microscopically under the
• 32:35
  immense strain. And the trap, the reason this
• 32:38
  causes billions of dollars in industrial damage every year, is
• 32:41
  that standard fluid equations like Bernoulli's assume steady
• 32:45
  flow. They don't account for the
• 32:47
  sudden stop. Right engineers will
• 32:49
  painstakingly draft their system, perfectly size their
• 32:52
  pipe diameters, carefully select their wall thicknesses, and
• 32:57
  choose their flanges and fittings based on the steady
• 32:59
  state, continuous operating pressure of the main pump.
• 33:02
  They look at the CAD model and say the pump outputs 100 PSI.
• 33:06
  The pipe can handle A150PSI. We are perfectly safe.
• 33:10
  And they completely forget to account for the dynamic split
• 33:14
  second spikes caused by unsteady flow events like a valve
• 33:17
  closing. Because that pressure spike
• 33:19
  doesn't just sit peacefully at the valve waiting to be
• 33:21
  relieved, It is a kinetic event. It becomes a shockwave.
• 33:25
  Wait, a literal shockwave, like an explosion?
• 33:26
  Exactly like an acoustic explosion, that high pressure
• 33:31
  wave of compressed water and expanded pipe wall travels
• 33:34
  backward through the pipe system, moving away from the
• 33:37
  closed valve and upstream toward the source, and it travels at
• 33:41
  the speed of sound within that specific fluid.
• 33:44
  As we discussed earlier for water, that shockwave is
• 33:47
  blasting backward at roughly 1480 meters per second.
• 33:52
  That is incredibly fast. It acts like a physical bullet
• 33:55
  blasting through the system. It hits elbows and tries to
• 33:58
  straighten out. It reflects off pumps, Tees and
• 34:01
  supply reservoirs, bouncing back and forth down the length of the
• 34:04
  pipe at supersonic speeds until fluid friction and the
• 34:08
  structural yielding of the pipe finally dampen the energy.
• 34:11
  And if the pipe isn't rated for that.
• 34:12
  If the pipe schedule or the threaded fittings are only rated
• 34:15
  for the steady state pump pressure, that supersonic
• 34:18
  pressure wave will literally RIP the steel open along the Weld
• 34:21
  seam or instantly blow the bolts right off a flange.
• 34:24
  OK this explains so much about everyday life.
• 34:27
  Is this exactly why my house pipes violently bang and shutter
• 34:30
  inside the drywall when my washing machine abruptly stops
• 34:34
  filling with water? That is precisely the physics at
• 34:36
  play. Your washing machine and your
• 34:38
  dishwasher use what's called a solenoid valve.
• 34:41
  OK, what does that do? It is an electromagnetically
• 34:44
  actuated valve that doesn't screw shut slowly like a garden
• 34:47
  hose. It snaps shut almost
• 34:49
  instantaneously. When it actuates the water
• 34:52
  flowing vigorously through your entire house, plumbing suddenly
• 34:55
  hits a wall. Creating that same pressure
• 34:57
  spike. Yes, the pressure spikes and the
• 35:00
  acoustic shockwave slams backward into your copper pipes.
• 35:03
  The spike is so forceful it physically causes the pipes to
• 35:06
  jump and bang against the wooden joist inside your walls.
• 35:09
  That's startling. Hammering noise is domestic
• 35:11
  water hammer. It is crazy to think about a
• 35:14
  supersonic shockwave bouncing around inside the walls of my
• 35:17
  suburban house. But scale that up.
• 35:19
  What happens when this hits heavy industry?
• 35:22
  If we connect this to the bigger picture, imagine that exact same
• 35:25
  phenomena but in a hydroelectric dam.
• 35:28
  Oh boy. You have a massive concrete and
• 35:30
  steel pen stock pipe maybe 10 feet in diameter carrying
• 35:34
  thousands of tons of water rushing down a mountainside
• 35:38
  straight into a spinning turbine.
• 35:40
  That is a lot of kinetic energy. If the electrical power grid
• 35:44
  suddenly drops the load, maybe a transmission line trips, the
• 35:47
  turbine loses its magnetic resistance and will instantly
• 35:50
  over speed and destroy itself. So we have to shut off the water
• 35:53
  immediately. Right, massive mechanical valves
• 35:56
  must close to cut off the water supply to the turbine.
• 35:59
  But if you close those massive valves even a second too fast,
• 36:03
  the kinetic energy of a literal river of water is converted into
• 36:07
  a pressure transient. The water hammer would be
• 36:09
  unimaginably violent. It could obliterate the steel
• 36:12
  pen stock, blow apart the powerhouse and threaten the
• 36:15
  structural integrity of the dam itself.
• 36:17
  So how do engineers deal with this?
• 36:18
  We can't just leave valves open forever.
• 36:20
  We obviously still have to shut down turbines and turn off
• 36:23
  washing machines. The primary, most mathematically
• 36:26
  elegant solution is simply altering the timeline.
• 36:29
  You must close the valve slowly. Just buy some time.
• 36:33
  Yes, if you engineer the actuator to extend the closure
• 36:37
  time, a fascinating wave cancellation occurs.
• 36:40
  The initial pressure wave travels up the pipe, hits the
• 36:43
  open reservoir, and reflects backward as a low pressure
• 36:46
  relief wave. Oh, I see where this is going.
• 36:48
  If the valve is still slowly closing by the time that relief
• 36:51
  wave returns, the returning low pressure effectively cancels out
• 36:55
  the building high pressure. The energy is allowed to
• 36:58
  dissipate safely and gradually. But what if system requirements
• 37:02
  demand a fast emergency closure, like a real emergency?
• 37:06
  Then you have to engineer physical pressure relief
• 37:08
  mechanisms into the architecture.
• 37:10
  In hydroelectric plants, they build massive search tanks.
• 37:13
  What are those? They are giant vertical open
• 37:16
  topped concrete towers connected to the penstock.
• 37:19
  When the valve slams shut, the high pressure water has a path
• 37:21
  of least resistance. It simply surges upward into the
• 37:24
  open tower, trading its kinetic pressure energy for potential
• 37:27
  height energy, completely buffering the shock.
• 37:29
  That makes perfect sense. What about smaller systems?
• 37:32
  In smaller industrial settings, we use hydraulic accumulators,
• 37:35
  which are heavy steel tanks containing a rubber bladder
• 37:38
  filled with pressurized nitrogen gas.
• 37:41
  When the water hammer hits, the incompressible water forcefully
• 37:45
  compresses the gas bladder, acting like a giant pneumatic
• 37:48
  shock absorber. OK, so water hammer is a
• 37:51
  violent, destructive vibration caused by the sudden shift of
• 37:55
  fluid momentum inside a pipe. But fluids moving on the outside
• 37:59
  of structures 'cause their own unique, terrifying brand of
• 38:02
  chaos? Oh, definitely.
• 38:04
  This brings us to a massive crossover between fluid dynamics
• 38:07
  and structural engineering. Let's look at when the fluid
• 38:10
  strikes back flow induced vibrations.
• 38:13
  This is where we see some of the most spectacular engineering
• 38:16
  failures in history. When a fluid, whether it is
• 38:18
  water or wind, flows around a solid structure, it
• 38:22
  fundamentally alters the structures physical vibrational
• 38:24
  characteristics. Does sitting water do this too?
• 38:27
  Actually, no. A quiescent fluid, meaning a
• 38:30
  fluid that is just sitting perfectly still around a
• 38:32
  structure, like a bridge piling in a calm, unmoving lake,
• 38:36
  actually helps. The mass of the Stillwater
• 38:39
  couples with the structure, decreasing its natural resonant
• 38:41
  frequencies and increasing its damping.
• 38:44
  It acts like a heavy blanket. But a flowing fluid.
• 38:47
  A flowing fluid introduces active, dynamic, and highly
• 38:49
  destructive mechanical forces. We categorize these aerodynamic
• 38:53
  forces into a few distinct types, added mass effects,
• 38:57
  inertial coupling, and the absolute most notorious system
• 39:00
  killer of them all, vortex shedding.
• 39:02
  We talked briefly about wakes earlier with the semi truck
• 39:05
  tailgating analogy causing flow separation.
• 39:08
  Vortex shedding takes that separation and weaponizes it.
• 39:11
  That's a good way to put it. Let's dig deeper into the
• 39:13
  mechanics. This phenomenon primarily
• 39:16
  happens when a fluid flows past what we call a bluff body.
• 39:19
  A bluff body is an engineering term for any unstreamlined,
• 39:23
  aerodynamically blunt shape. Like a cylinder or a square
• 39:26
  beam? Exactly.
• 39:28
  A perfectly round cylinder, a square concrete beam, a
• 39:32
  hexagonal sensor probe. These are all bluff bodies.
• 39:36
  Because they aren't shaped like an airplane wing, the fluid
• 39:38
  simply cannot follow the sharp curves of the geometry, so the
• 39:42
  boundary layer violently separates.
• 39:44
  But. It doesn't just form a steady,
• 39:46
  continuous pocket of dead air. No, the sheer layers of the
• 39:50
  separated fluid are highly unstable.
• 39:52
  They curl back in on themselves and form vortices, spinning
• 39:55
  whirlpools of fluid that physically peel off the back of
• 39:59
  the object and travel downstream.
• 40:00
  And the critical detail here is that they do not peel off
• 40:03
  randomly. Once the flow reaches a Reynolds
• 40:06
  number greater than about 50, these swirling vortices begin to
• 40:09
  shed alternately from each side of the bluff body in a highly
• 40:13
  regular, perfectly rhythmic, predictable pattern.
• 40:15
  First one side, then the other, yes. 1st a vortex gathers and
• 40:19
  sheds from the left side of the cylinder.
• 40:21
  Then a vortex gathers and sheds from the right side, then the
• 40:23
  left again, then the right. This beautiful alternating trail
• 40:27
  of spinning fluid is called a Von Karman Vortex St.
• 40:30
  Now we have to look at the pressure mechanics of that.
• 40:33
  The fluid inside a spinning vortex is moving extremely fast,
• 40:37
  which means the pressure inside the vortex drops significantly.
• 40:41
  Therefore, every single time a vortex sheds off one side of the
• 40:45
  structure, it creates A localized low pressure zone that
• 40:48
  literally sucks the physical structure toward it.
• 40:51
  This creates an alternating lift force perpendicular to the flow.
• 40:55
  Yeah, left, right, left, right. The solid structure is being
• 40:59
  rhythmically, relentlessly pulled back and forth by the
• 41:02
  fluid. And it's not just a random
• 41:04
  pulsing, it's a measurable mathematical frequency.
• 41:07
  We can actually calculate exactly how fast those
• 41:10
  alternating pulls will happen using a dimensionless parameter
• 41:13
  called the stroll number. Yes, the frequency of the shed
• 41:15
  vortices measured in Hertz is equal to the troll number
• 41:19
  multiplied by the flow velocity of the fluid divided by the
• 41:22
  physical width or diameter of the object facing the flow.
• 41:25
  So if the water flows faster, the alternating pulling gets
• 41:28
  faster. Which brings us to the absolute
• 41:30
  nightmare scenario for chemical and Mechanical Engineers.
• 41:33
  Fluid elastic instability, particularly as it pertains to
• 41:37
  tube and shell heat exchangers. Oh, I've seen the aftermath of
• 41:40
  this. It's ugly.
• 41:42
  It really is. Imagine a massive industrial
• 41:45
  cylinder filled with a dense array of hundreds of thin metal
• 41:48
  tubes. You have a hot fluid flowing
• 41:51
  inside the tubes and a cold fluid flowing across the outside
• 41:54
  of the tubes to transfer heat. Standard heat exchanger setup.
• 41:57
  As the cross flow velocity of that cooling fluid increases,
• 42:01
  you get alternating vortex shedding on every single one of
• 42:03
  those hundreds of tubes. But if you push the pump too
• 42:06
  hard and hit a specific critical velocity threshold, something
• 42:09
  terrifying happens. Yes, the vibration amplitude of
• 42:12
  the tubes doesn't just increase linearly with the flow speed, it
• 42:16
  spikes exponentially, usually proportional to the flow
• 42:19
  velocity raised to the power of 4 or even higher.
• 42:22
  The vibration becomes so insanely violent that the metal
• 42:25
  tubes physically bow outward and clash into each other, denting,
• 42:28
  rupturing and destroying the metal.
• 42:30
  Or they vibrate so hard against their support baffles that they
• 42:34
  literally guillotine themselves and shear right off.
• 42:36
  You noted from the research papers that this fluid elastic
• 42:39
  instability is the second leading cause of catastrophic
• 42:42
  heat exchanger failure globally, right behind long term chemical
• 42:46
  corrosion. It destroys equipment in hours.
• 42:49
  Hours and there is a highly persistent, incredibly costly
• 42:53
  trap that structural engineers fall into.
• 42:55
  Here, what's the trap? When they see a structure
• 42:57
  vibrating violently in the wind or water, their immediate
• 43:01
  intuition is the structure is too weak.
• 43:03
  I just need to make the structure stronger, stiffer, and
• 43:06
  Weld more heavy steel to it to resist the fluid.
• 43:09
  They operate under the assumption that structural
• 43:11
  stiffness implies aerodynamic immunity.
• 43:14
  But the aha moment here, the thing that changes how you view
• 43:18
  structural engineering forever, is realizing it is not a one way
• 43:21
  St., it is a closed feedback loop.
• 43:23
  Precisely if the physical structure is even
• 43:26
  microscopically flexible and all materials are, the alternating
• 43:29
  lift forces from the vortex shedding will cause it to
• 43:32
  physically deflect and move. And that movement matters.
• 43:35
  Massively, that structural movement physically alters the
• 43:38
  boundary layer of the fluid flowing over it.
• 43:41
  The motion of the structure begins to dictate and organize
• 43:44
  the shedding frequency of the vortices.
• 43:46
  Wait, so the structure is now controlling the fluid?
• 43:48
  Yes, if the calculated natural shedding frequency of the
• 43:52
  vortices happens to lock on to the inherent resonant natural
• 43:55
  frequency of the physical steel structure, you achieve
• 43:59
  catastrophic resonance. They sync up.
• 44:01
  The fluid and the structure synchronized perfectly.
• 44:04
  Every push from the fluid happens at the exact perfect
• 44:07
  millisecond to amplify the sway of the building.
• 44:10
  The displacements become violent.
• 44:12
  And making it stiffer doesn't help.
• 44:14
  Building a stiffer, heavier structure might just give it a
• 44:16
  higher natural resonant frequency.
• 44:19
  This means it won't vibrate at low wind speeds, but it will
• 44:22
  still violently resonate and tear itself apart, just at a
• 44:25
  much higher, more dangerous wind speed.
• 44:28
  We have seen this play out in the real world with disastrous
• 44:31
  results. We've seen tall, rigid
• 44:33
  industrial chimneys and high crosswinds swaying back and
• 44:36
  forth until the concrete structurally cracks.
• 44:38
  We've seen deep water, offshore oil terminal risers and marine
• 44:43
  pipelines subjected to steady ocean currents vibrating
• 44:47
  themselves to pieces on the sea floor because of vortex
• 44:50
  shedding. So if we can't just arbitrarily
• 44:52
  throw thicker steel at the problem, how do we actually stop
• 44:55
  this? Without completely redesigning
• 44:57
  the architecture, we can't exactly build a teardrop shaped,
• 45:01
  perfectly aerodynamic smokestack.
• 45:03
  You have to outsmart the fluid. Instead of attacking the
• 45:06
  structures rigidity, you attack the fluids aerodynamics
• 45:09
  directly. You have to intentionally
• 45:11
  sabotage the regularity of the vortex shedding.
• 45:14
  Sabotage it. Because of the vortices cannot
• 45:16
  shed in a synchronized rhythmic pattern, they cannot organize
• 45:19
  into a unified resonant pulling force.
• 45:21
  The poles will be random and cancel each other out.
• 45:24
  So how do engineers do that? Engineers use several clever
• 45:27
  aerodynamic modifications. The most brilliantly simple and
• 45:30
  common is the helical strike. The helical strike, like those
• 45:33
  spiral fins. Exactly.
• 45:35
  If you look at a tall metal smokestack, at a power plant, or
• 45:38
  even something as mundane as the small rigid spiral fin wrapped
• 45:42
  around your car's radio antenna, you'll see ridges spiraling up
• 45:46
  the length of the cylinder. Those strikes are intentional
• 45:49
  speed bumps. They physically force the
• 45:51
  boundary layer to separate at different times at different
• 45:54
  heights along the cylinder. The vortices shed completely out
• 45:57
  of phase with each other, totally breaking up the unified
• 46:00
  rhythmic force. Other brilliant solutions
• 46:02
  include shrouds, which are basically perforated oversized
• 46:06
  metal cylinders placed over the inner pipe, or axial slats
• 46:10
  running straight down the length of it.
• 46:11
  Yes, all of these tools share the exact same mission.
• 46:14
  They are designed to turn organized, rhythmic, resonant
• 46:18
  aerodynamic pulling into disorganized, random chaotic
• 46:21
  turbulence. In the world of flow induced
• 46:23
  vibration, messy chaotic turbulence is vastly safer than
• 46:27
  organized harmonic flow. Exactly right.
• 46:30
  So, having covered the sheer mechanical violence of
• 46:33
  vibrations and supersonic shock waves, we need to take a step
• 46:37
  back and look at the accounting department of fluid mechanics.
• 46:40
  How do engineers mathematically balance the books of all this
• 46:44
  fluid energy? The energy Ledger.
• 46:46
  Let's look at the energy Ledger, Bernoulli and real world losses.
• 46:50
  To understand how energy moves through a system, we have to
• 46:53
  return to Bernoulli's equation, which is arguably the most
• 46:56
  famous heavily utilized equation in all of fluid mechanics.
• 47:00
  It's basically conservation of energy, right?
• 47:02
  It is fundamentally an expression of the first law of
• 47:05
  thermodynamics, the conservation of energy applied directly to a
• 47:09
  flowing fluid. It boldly states that for a
• 47:12
  steady, incompressible, perfectly frictionless flow
• 47:15
  traveling along a single streamline, the total energy per
• 47:19
  unit mass is an absolute constant.
• 47:21
  That total energy is essentially a bank account made-up of three
• 47:24
  distinct currencies, kinetic energy, pressure energy and
• 47:27
  potential energy. Let's translate the math into
• 47:30
  concepts. You have the fluids velocity
• 47:32
  squared. That represents the kinetic
• 47:34
  energy term, the energy of physical motion.
• 47:36
  Then you have the static fluid pressure, that is the pressure
• 47:39
  energy term, the energy of compression.
• 47:41
  Finally, you have the physical elevation or height of the fluid
• 47:45
  relative to the ground. That is the potential energy
• 47:48
  term, the energy of gravity. Bernoulli's equation dictates
• 47:51
  that if you add the balance of those three accounts together at
• 47:55
  point A at the start of a pipe, they must perfectly equal the
• 47:59
  sum of those same three accounts at point B at the end of the
• 48:02
  pipe. The fluid can freely transfer
• 48:04
  funds between velocity, pressure and height, but the total net
• 48:08
  worth of the fluid cannot change.
• 48:10
  And you can see this financial transfer beautifully illustrated
• 48:13
  in Torricelli's theorem. Yes, the water tower example.
• 48:17
  Imagine a massive open topped municipal water tower.
• 48:21
  Near the very bottom of the tank there is a small hole.
• 48:24
  If we look at the very top surface of the water high up in
• 48:27
  the tank, the water is barely moving downwards, so velocity is
• 48:31
  essentially 0. It's open to the air, so the
• 48:33
  pressure is just 0 gauge atmospheric pressure.
• 48:36
  All of its energy, 100% of its bank account, is tied up in
• 48:39
  potential energy, it's height above the ground.
• 48:41
  As that specific particle of water flows down through the
• 48:44
  tank and violently squirts out the small hole at the bottom,
• 48:48
  its physical height is now 0. The pressure is once again
• 48:52
  atmospheric because it's shooting into the open air.
• 48:54
  So where did all that massive potential energy go?
• 48:57
  It transferred entirely into the kinetic energy account.
• 49:00
  The velocity of the water jet shooting out of the hole is
• 49:03
  exactly the same mathematically as if you had simply dropped a
• 49:07
  physical bowling ball from the height of the water surface.
• 49:10
  It's the ultimate roller coaster analogy.
• 49:12
  When the roller coaster car is ratcheting up to the very top of
• 49:15
  the highest hill, it is moving painfully slow.
• 49:19
  Kinetic energy is minimal, but its potential energy from
• 49:22
  gravity is at its absolute peak. As it plunges down the vertical
• 49:26
  drop, it loses that height but gains massive, terrifying speed.
• 49:30
  Potential energy converts directly to kinetic energy,
• 49:33
  which maxes out at the very bottom of the dip where the
• 49:36
  track is lowest. Bernoulli's equation says fluids
• 49:38
  do the exact same thing, just with static pressure thrown into
• 49:42
  the mix as a third highly liquid asset.
• 49:45
  However, there is a fascinating, highly visible quirk when fluid
• 49:50
  exits a hole, like that known as the Vena contracta.
• 49:54
  Yes, the contraction. The streamlines of the fluid
• 49:57
  inside the tank converge from all directions toward the exit
• 50:00
  hole. But fluid has momentum.
• 50:03
  It cannot make a perfect instantaneous 90° geometric turn
• 50:08
  right at the sharp edge of the hole.
• 50:10
  The streamlines literally overshoot the edge and keep
• 50:12
  curving inward toward the center of the jet even after they have
• 50:16
  exited the tank. This means the actual physical
• 50:18
  cross-sectional area of the water jet shooting through the
• 50:21
  air is significantly smaller than the physical area of the
• 50:25
  hole it just passed through. The flow contraction coefficient
• 50:27
  is usually around .65, meaning the jet is only 65% the size the
• 50:31
  whole. But Bernoulli's equation, as
• 50:33
  beautiful and elegant as it is, describes a perfect,
• 50:37
  frictionless utopian universe. And as we have established ad
• 50:41
  nauseam today, the real industrial world is messy.
• 50:44
  The massive pain point for engineers sizing pumps here is
• 50:47
  accounting for what we broadly call head loss.
• 50:49
  The energy Ledger doesn't actually balance perfectly
• 50:52
  between point A and point B because the physical universe
• 50:55
  always, always takes a tax. That unavoidable tax is viscous
• 51:00
  dissipation. As fluid flows through a pipe
• 51:02
  system, mechanical energy is constantly irreversibly being
• 51:06
  destroyed and turned into unrecoverable low grade heat due
• 51:10
  to the internal friction of the fluid and the drag against the
• 51:13
  pipe walls. We divide these system taxes
• 51:15
  into two distinct categories for calculation major losses, which
• 51:20
  is simply the friction accumulated along the long
• 51:22
  straight lengths of pipe we discussed with the Colebrook
• 51:24
  White equation. And minor losses which are
• 51:26
  localized, highly concentrated energy drops caused by the pipe
• 51:30
  fittings themselves. And the trap engineers fall into
• 51:32
  is looking at a massive plumbing schematic and thinking a pipe
• 51:35
  fitting is just a piece of metal and elbow is just an elbow.
• 51:38
  It just changes the direction. It shouldn't consume that much
• 51:41
  energy. But that's a trap.
• 51:42
  That assumption will result in a pump that completely fails to
• 51:45
  deliver the required flow rate. Every single time the fluid is
• 51:49
  forced to change direction in an elbow, squeeze through the
• 51:52
  restricted throat of a valve, or navigate a sudden expansion
• 51:56
  flow, separation occurs. The boundary layer tears off,
• 51:59
  violent turbulent vortices are formed, and immense amounts of
• 52:02
  mechanical energy are burned up simply trying to navigate the
• 52:06
  geometry. Different geometric fittings
• 52:08
  have vastly different experimentally determined loss
• 52:11
  coefficients. For example, if a pipe entrance
• 52:14
  drawing water from a tank is just a sharp square cut edge,
• 52:18
  the fluid separates violently as it tries to turn the corner.
• 52:21
  Creating a massive vena contractor an extremely high
• 52:24
  energy. Loss.
• 52:25
  If the pipe physically protrudes inward into the tank, what we
• 52:28
  call a re entrance shape, the fluid has to execute a full 180°
• 52:33
  turn to get in. The separation is massive and
• 52:36
  the energy loss is even worse. But if the engineer simply
• 52:39
  rounds the pipe entrance, giving it a smooth aerodynamic
• 52:42
  bellmouth curve, the fluid glides in perfectly without
• 52:46
  separating, and the localized energy loss drops to almost
• 52:49
  absolute zero. Geometry dictates efficiency.
• 52:52
  We also rely heavily on Bernoulli's energy principles to
• 52:55
  actually measure flow in the real world.
• 52:58
  We use pedo tubes, particularly on aircraft, to measure local
• 53:02
  velocity by bringing the fluid to a dead stop and measuring the
• 53:05
  difference between the stagnation pressure and the
• 53:07
  static pressure. We use venturi meters and
• 53:10
  orifice plates inside pipes which intentionally physically
• 53:14
  choke the diameter of the pipe down to force the fluid to speed
• 53:17
  up. That speed increase causes a
• 53:20
  measurable pressure drop according to Bernoulli, which we
• 53:23
  can then mathematically correlate directly to the mass
• 53:26
  flow rate. But actually measuring flow in
• 53:28
  an industrial setting presents its own unique set of severe
• 53:32
  challenges, especially when you are dealing with two phase
• 53:34
  flows. Imagine trying to measure an air
• 53:37
  water mixture like the frothy output of a chemical reactor.
• 53:41
  If you use a standard pittotubic equation calibrated for pure
• 53:44
  water, but your water has a 20% concentration of entrained
• 53:48
  microscopic air bubbles, and you naively ignore the
• 53:51
  compressibility of those bubbles, your velocity
• 53:54
  calculations will be massively, dangerously wrong.
• 53:57
  As the fluid hits the stagnation point of the sensor and the
• 54:00
  pressure spikes, the liquid water doesn't compress, but the
• 54:03
  gas bubbles. Compressed dynamically.
• 54:06
  The density of the fluid mixture changes locally right at the
• 54:08
  sensor face, skewing the pressure differential
• 54:11
  completely. Which provides the absolute
• 54:13
  perfect runway to launch into our final theoretical concept.
• 54:17
  We have been mostly talking about liquids today, where we
• 54:20
  only have to worry about compressibility if we hit
• 54:22
  extreme crushing hydraulic pressures or if we are dealing
• 54:26
  with entrained air bubbles. But what happens when the fluid
• 54:30
  itself is inherently highly compressible?
• 54:33
  What happens when we push gases to their absolute physical
• 54:36
  limits? Let's talk about gas dynamics
• 54:39
  and compressibility. When dealing with high speed
• 54:41
  gases, air, steam, natural gas, almost everything we learned
• 54:45
  about the predictable behavior of incompressible liquids go
• 54:48
  straight out the window. The primary governing metric
• 54:51
  here shifts entirely to the Mach number, which as we noted, is
• 54:54
  the ratio of the fluids velocity to the local speed of sound
• 54:57
  within that exact fluid at that exact temperature.
• 55:00
  We strictly categorize flows based on this ratio.
• 55:03
  Subsonic flow is less than Mach 1, transonic flow hovers
• 55:07
  unpredictably right around Mach 1, supersonic flow is undeniably
• 55:11
  above Mach 1, and hypersonic flow is generally anything above
• 55:14
  Mach 5. And the reason everything
• 55:16
  changes is density. Unlike water where a gallon is a
• 55:19
  gallon, the density of a gas changes wildly and rapidly as
• 55:23
  its velocity and pressure change.
• 55:25
  You simply cannot treat density as a constant in your calculus.
• 55:28
  And this intense compressibility leads to wild, violent physical
• 55:32
  phenomena like shock waves. We see normal shocks, which are
• 55:36
  blunt pressure walls that sit detached and perfectly
• 55:39
  perpendicular to the flow. And we see oblique shocks which
• 55:42
  form a sharp angle exactly like the V shaped mock cone trailing
• 55:46
  off the nose of a supersonic fighter jet slicing through the
• 55:48
  atmosphere. But bringing this back to
• 55:50
  terrestrial engineering, the most practical daily hurdle for
• 55:53
  industrial engineers dealing with gas dynamics is the
• 55:56
  frustrating phenomenon of choked flow.
• 55:58
  Let's do another thought experiment.
• 56:00
  Imagine you are designing a massive high pressure natural
• 56:04
  gas cross country pipeline. You have a multi $1,000,000
• 56:07
  industrial scale compressor pushing methane gas down a very
• 56:11
  long, perfectly straight pipe at a steady subsonic speed of say
• 56:16
  15 meters per second. As the gas flows down the miles
• 56:19
  of pipe wall, friction acts continuously upon it.
• 56:23
  This specific scenario, compressible flow in a constant
• 56:26
  area duct with friction, is known thermodynamically as Fano
• 56:30
  flow. Because it is a compressible
• 56:32
  gas, the friction dragging against the pipe naturally
• 56:35
  causes the fluid pressure to steadily drop.
• 56:38
  But critically, because the pressure drops, the gas expands,
• 56:41
  meaning its density also drops. And because the mass flow rate,
• 56:45
  the actual pounds of gas moving through the pipe per minute,
• 56:47
  must remain perfectly constant from 1 into the other, the gas
• 56:51
  must actually physically accelerate to make up for its
• 56:53
  lower density. Wait, hold on, let me get this
• 56:55
  straight. You are telling me that dragging
• 56:57
  friction, the literal force that slows things down, actually
• 57:00
  causes the gas to speed up in the pipe?
• 57:02
  That feels entirely backwards. It is entirely counterintuitive
• 57:05
  to how we interact with solid objects, but it is an absolute
• 57:08
  thermodynamic truth for subsonic gas flow in a constant area pipe
• 57:13
  to maintain the conservation of mass.
• 57:14
  As the gas gets fluffier and less dense due to frictional
• 57:17
  pressure drop, it has to move faster to push the same amount
• 57:21
  of mass through the same sized hole.
• 57:22
  The velocity keeps steadily increasing as it travels further
• 57:25
  and further down the pipe. But it will eventually hit an
• 57:28
  absolute unbreakable physical limit as the velocity
• 57:32
  accelerates and approaches exactly Mach one, either due to
• 57:35
  miles of friction along a straight pipe or because you
• 57:38
  intentionally force the gas through a narrow bottleneck like
• 57:41
  a valve. The flow becomes choked.
• 57:43
  At the exact microscopic point where the gas velocity hits Mach
• 57:46
  One, the laws of physics intervene.
• 57:49
  No matter how much more pressure you crank out of that $1,000,000
• 57:52
  compressor upstream, the actual mass flow rate of the gas
• 57:55
  absolutely cannot and will not increase.
• 57:57
  If I have a massive compressor and I crank it to maximum power,
• 58:02
  dumping mega joules of energy into the system, you're telling
• 58:04
  me physics just says no? Where does the extra pushing go?
• 58:07
  It goes nowhere useful, it just backs up as higher upstream
• 58:11
  pressure. The fundamental reason is
• 58:14
  acoustic. The physical information that
• 58:17
  the compressor is pushing harder upstream communication sell
• 58:20
  through the gas via pressure waves.
• 58:23
  And pressure waves, by definition, travel at exactly
• 58:25
  the speed of sound. But if the gas at the bottleneck
• 58:28
  is already flowing forward at the speed of sound, those
• 58:31
  pressure waves can never travel backward upstream to tell the
• 58:34
  incoming gas to move faster. The communication is cut off.
• 58:38
  It is exactly like a panicked crowd of 100 people trying to
• 58:41
  run out of a single set of double doors during an
• 58:44
  emergency. The people at the very front are
• 58:47
  squeezing through the narrow doors at their absolute maximum
• 58:50
  physical capacity. The people in the back of the
• 58:52
  room panic and start pushing harder and harder against the
• 58:55
  crowd. But pushing violently from the
• 58:57
  back doesn't get people through the front door is any faster.
• 58:59
  The bottleneck is already maxed out.
• 59:01
  The extra pushing just creates A crushing shockwave of bodies in
• 59:05
  the room. The door is mathematically
• 59:07
  choked. A grim but highly accurate
• 59:10
  physical analogy, and the costly assumption here is treating a
• 59:14
  high speed gas pipeline with the exact same math as a municipal
• 59:18
  water pipe. You cannot just plug your
• 59:20
  numbers into the standard Garcia Weisbach liquid friction
• 59:23
  equations and call it a day. Thermodynamic effects completely
• 59:27
  dictate and alter the pressure gradient.
• 59:29
  As an engineer, you have to precisely determine if the flow
• 59:33
  is isothermal, meaning it flows slow enough to constantly
• 59:36
  exchange heat with the surrounding environment and
• 59:39
  maintain a stable temperature. Or if it is adiabatic, meaning
• 59:42
  it flows so fast or is so well insulated that it retains all
• 59:45
  its frictional heat. The calculated limiting lengths
• 59:48
  of the pipe and the exact pressures where choking occurs
• 59:51
  are vastly fundamentally different depending on that
• 59:53
  thermodynamic state. OK, but this raises a massive
• 59:56
  contradiction in my head. If flow is completely
• 59:58
  universally cap at Mach one at the tightest bottleneck, how do
• 1:00:02
  space rockets ever achieve supersonic speeds?
• 1:00:05
  A rocket engine narrows down to a tiny throat to build pressure,
• 1:00:08
  so the exhaust gas must choke at Mach 1.
• 1:00:11
  How does the thrust ever get past the speed of sound?
• 1:00:14
  This raises an important question, and the answer is
• 1:00:16
  arguably one of the most brilliant pieces of mechanical
• 1:00:19
  engineering in human history. The dilaval nozzle or the
• 1:00:22
  converging diverging nozzle? You are correct.
• 1:00:25
  You start with a wide combustion chamber and geometrically
• 1:00:28
  converge the metal down to a narrow throat.
• 1:00:31
  The high pressure subsonic gas accelerates as it is squeezed,
• 1:00:35
  hitting exactly Mach one right at the narrowest point of the
• 1:00:39
  throat. It perfectly chokes.
• 1:00:41
  But then instead of just dumping that Mach one gas into the
• 1:00:44
  atmosphere, you attach a carefully machine diverging bell
• 1:00:47
  section that slowly widens back out.
• 1:00:49
  Because the gas has hit exactly Mach 1, the thermodynamic rules
• 1:00:53
  completely invert. In subsonic flow, a widening
• 1:00:56
  pipe causes the fluid to slow down, but in supersonic flow, a
• 1:01:00
  widening pipe gives the highly compressed gas room to rapidly
• 1:01:03
  expand. The gas density drops incredibly
• 1:01:06
  fast, much faster than the geometric area of the nozzle is
• 1:01:09
  increasing. Because the density is
• 1:01:10
  plummeting, the velocity must violently accelerate to maintain
• 1:01:13
  the mass flow rate. The widening nozzle physically
• 1:01:16
  tricks the choked flow into expanding and blasting out the
• 1:01:20
  back of the rocket at Mach 2, Mach 3, or even Mach 10.
• 1:01:23
  You break the sound barrier not by squeezing harder, but by
• 1:01:27
  giving the Sonic fluid room to breathe.
• 1:01:29
  Which is absolutely incredibly brilliant.
• 1:01:32
  OK, we have covered a massive amount of theoretical and
• 1:01:34
  practical ground today, from the microscopic safe haven of the
• 1:01:38
  viscous sublayer all the way up to the thermodynamics of
• 1:01:41
  supersonic rocket nozzles. Now it is time to bring it all
• 1:01:45
  home to the real world. Let's do it.
• 1:01:47
  Let's finish up the class by answering the biggest, most
• 1:01:50
  frustrating fluid mechanics questions you all are asking on
• 1:01:53
  engineering forums right now. We scoured the Internet to find
• 1:01:56
  the overarching themes and recurring disasters you are
• 1:01:58
  struggling with in the field. Question number one.
• 1:02:01
  And we see variations of this constantly.
• 1:02:03
  I have a new industrial cooling loop.
• 1:02:05
  Why do my system's pipes vibrate violently and bang during the
• 1:02:08
  morning startup and evening shutdown, but run totally fine,
• 1:02:11
  smooth and quiet for the 8 hours in between?
• 1:02:14
  Based on everything we've covered, this is a textbook
• 1:02:17
  classic symptom of unsteady flow and water hammer.
• 1:02:20
  During the 8 hours of steady operation, the flow is stable,
• 1:02:23
  the velocity and pressure aren't changing, so the forces are
• 1:02:26
  balanced. But during morning startup and
• 1:02:28
  evening shutdown, massive centrifugal pumps are spooling
• 1:02:32
  up or heavy isolation valves are opening and closing.
• 1:02:35
  You are creating a sudden massive shift in the momentum of
• 1:02:38
  the fluid mass. That massive kinetic energy of
• 1:02:41
  thousands of gallons of water starting or stopping is
• 1:02:44
  instantly converting into violent pressure transients that
• 1:02:47
  travel as acoustic shockwaves through the piping network,
• 1:02:50
  literally physically shaking the pipes.
• 1:02:52
  And what is the actionable advice for fixing a system
• 1:02:55
  that's already built? First, audit your valve
• 1:02:58
  actuation times immediately. Implement slow acting motorized
• 1:03:02
  valves wherever possible to extend the closure time and
• 1:03:05
  allow the pressure waves to cancel out.
• 1:03:06
  Second, strongly consider retrofitting variable frequency
• 1:03:09
  drives or VF DS onto your pump motors.
• 1:03:12
  Instead of a pump slamming on at 100% speed instantly, a VFD
• 1:03:17
  electrically ramps the motor up slowly over 10 or 20 seconds,
• 1:03:20
  accelerating the massive fluid mass gently.
• 1:03:24
  Finally, inspect your physical pipe supports.
• 1:03:27
  Proper steel anchoring won't stop the internal pressure wave,
• 1:03:30
  but it will physically stop the pipes from thrashing around,
• 1:03:33
  preventing the metal joints from fatiguing and eventually
• 1:03:35
  snapping. Question #2 How do I stop vortex
• 1:03:39
  shedding from destroying my sensor probes in a cross flow?
• 1:03:42
  This engineer has an expensive thermal probe sticking
• 1:03:44
  perpendicular into a high flow municipal water main and the
• 1:03:48
  flow keeps snapping the metal probe clean off after a few
• 1:03:51
  weeks. This goes directly back to flow
• 1:03:53
  induced vibrations. The fluid separating as it flows
• 1:03:56
  around the cylindrical metal probe is shedding alternating
• 1:03:59
  von Karman vortices rhythmically, relentlessly
• 1:04:02
  pulling the probe back and forth.
• 1:04:03
  If the metal keeps fatiguing and snapping, it means the vortex
• 1:04:07
  shedding frequency calculated by the Strowel number based on the
• 1:04:10
  water speed is perfectly matching the natural resonant
• 1:04:13
  frequency of that specific metal probe.
• 1:04:16
  The resonance is amplifying the vibration until the yield
• 1:04:19
  strength of the steel is exceeded.
• 1:04:21
  To fix it, you essentially have two options.
• 1:04:24
  Change the physical structure structure or change the
• 1:04:26
  aerodynamics. You can change the natural
• 1:04:28
  frequency of the probe by swapping it for one that is
• 1:04:31
  physically shorter or significantly thicker.
• 1:04:34
  This shifts its resonant point far away from the shedding
• 1:04:37
  frequency of the water, preventing the destructive
• 1:04:40
  feedback loop. Or, if you absolutely cannot
• 1:04:43
  change the required probe dimensions, you must alter the
• 1:04:46
  aerodynamics. You can intentionally machine a
• 1:04:48
  helical strike, a spiral groove or raised fin onto the outside
• 1:04:52
  casing of the probe. This physically disrupts the
• 1:04:54
  uniform straight line separation of the boundary layer entirely,
• 1:04:58
  breaking up the straw of frequency so the vortices cannot
• 1:05:01
  organize into a synchronized destructive resonant force.
• 1:05:04
  Final question #3 why is my newly installed pump
• 1:05:08
  aggressively underperforming when my CAD math says it should
• 1:05:11
  work perfectly? The engineer calculated the
• 1:05:15
  required head pressure, bought the exact pump from the catalog,
• 1:05:18
  and it's delivering 20% less flow rate than promised.
• 1:05:21
  Assuming the pump isn't defective, 9 times out of 10
• 1:05:24
  this massive headache comes down to an extreme underestimation of
• 1:05:28
  real world energy losses, specifically the boundary layer
• 1:05:32
  friction and the minor losses we covered earlier.
• 1:05:34
  The engineer likely perfectly calculated the major friction
• 1:05:38
  losses of the long straight pipe runs, but they either ignored or
• 1:05:41
  vastly underestimated the minor losses.
• 1:05:44
  Every single elbow, T junction, partially open isolation valve,
• 1:05:48
  and sudden pipe expansion in that schematic destroys
• 1:05:51
  mechanical energy via turbulent viscous dissipation.
• 1:05:54
  It taxes the system heavily. Or tying it back to our earlier
• 1:05:57
  discussion about the viscous sub layer, they assumed A
• 1:06:00
  hydraulically smooth pipe friction factor from the
• 1:06:03
  Colebrook white equation when they shouldn't have.
• 1:06:05
  They looked at a brand new, seemingly smooth steel pipe and
• 1:06:08
  said that smooth friction will be low.
• 1:06:10
  But they completely failed to realize that at their specific
• 1:06:14
  incredibly high operational Reynolds number, the viscous sub
• 1:06:17
  layer is compressed to a microscopic thickness.
• 1:06:20
  The internal manufacturer roughness of that steel pipe is
• 1:06:23
  protruding right through the sub layer causing massive unexpected
• 1:06:27
  turbulent drag. Textbook theory is a starting
• 1:06:30
  line. It is not the finish line.
• 1:06:32
  Exactly. You must aggressively account
• 1:06:34
  for the physical ugly reality of the installation, the exact
• 1:06:39
  fittings, the true Reynolds number, the actual wall
• 1:06:41
  roughness, not just the idealized frictionless straight
• 1:06:45
  line distance on a pristine CAD drawing.
• 1:06:48
  We spend so much time in engineering fighting the
• 1:06:50
  inherent chaos of fluid. We try to chemically suppress
• 1:06:53
  turbulence, try to air, dynamically silence vortex
• 1:06:56
  shedding. We install massive nitrogen
• 1:06:58
  tanks to dampen water hammer. But I want to leave you with a
• 1:07:00
  final provocative thought to Mull over before we close out
• 1:07:03
  this deep dive. What if we stop fighting it?
• 1:07:06
  What if we lean into the chaos? Brilliant researchers are
• 1:07:08
  currently prototyping and designing bladeless wind
• 1:07:11
  turbines. Instead of massive spinning
• 1:07:14
  fiberglass blades that kill birds and require heavy
• 1:07:17
  gearboxes, these turbines are essentially just tall, highly
• 1:07:21
  flexible poles standing vertically in the wind.
• 1:07:24
  They are intentionally designed to harvest energy from vortex
• 1:07:27
  induced vibrations. The engineers mathematically
• 1:07:30
  tune the flexibility of the structure to perfectly
• 1:07:33
  dangerously resonate with the straw hole shedding frequency of
• 1:07:36
  the wind. They take that violent,
• 1:07:38
  oscillating mechanical resonance, the exact same
• 1:07:41
  terrifying phenomenon that destroys bridges and shears off
• 1:07:44
  heat exchangers, and they use linear alternators at the base
• 1:07:48
  to turn that thrashing kinetic energy into a source of
• 1:07:51
  completely frictionless, highly efficient renewable electrical
• 1:07:54
  power. When you truly, deeply
• 1:07:56
  understand the underlying physics of the chaos, you don't
• 1:07:59
  have to fear it. You can harness it.
• 1:08:01
  If you got something useful out of today's episode, hit follow
• 1:08:04
  and drop a like. It helps grow the show and keeps
• 1:08:06
  the engineering knowledge flowing.