Latest / Mechanical Engineering Made Simple / (#117) Introduction to Fluid Mechanics - Lesson 2

Transcript

(#117) Introduction to Fluid Mechanics - Lesson 2 · Mechanical Engineering Made Simple

0:00
Welcome back design junkies, to Mechanical Engineering Made
0:02
Simple. If you enjoyed the podcast, show
0:05
your support with a like, a follow or a comment.
0:07
Thank you so much for having me, I'm really excited to dig into
0:10
this today. Yeah, we are thrilled to have
0:12
you here. So for the listener, imagine
0:14
pressing the gas pedals in your car, right?
0:17
But instead of just moving the vehicle, you also have to
0:20
physically drag the entire atmosphere around you.
0:24
Oh man, that would be rough. Right, Suddenly your, you know,
0:27
your two ton car feels like it weighs 50 tons.
0:31
The steering becomes incredibly sluggish, the brakes barely
0:34
work, and every time you turn the wheel, an invisible hand
0:38
violently shoves the back of the car in the opposite direction.
0:41
Which sounds like a just a catastrophic mechanical failure,
0:45
but what you are describing is actually the fundamental
0:48
everyday reality of designing machines that operate in dense
0:51
fluids. It's crazy to think about.
0:53
It really is. I mean, we have a tendency in
0:55
engineering to look at a blueprint point to a steel beam,
0:58
calculate its yield strength, and, you know, treat it as a
1:02
static isolated object in a vacuum.
1:04
And I guess that works perfectly fine if you are building like a
1:09
bridge in a relatively calm environment.
1:11
Civil engineers get to enjoy that concrete precision.
1:14
Exactly. They have it easy in that
1:15
regard. But today, our mission for you,
1:17
the listener, is to take you through what is essentially a
1:20
master class on the dynamic, highly coupled world of fluid
1:24
mechanics and structural design. We are going straight into a
1:28
deep dive on this. Yes, we're going to bridge that
1:30
massive gap between abstract fluid theories and the really
1:35
gritty realities of building things like submarines, nuclear
1:38
reactors, and even refrigeration systems.
1:40
So by the time we wrap up this deep dive, you should understand
1:43
not just how objects move through fluids, but exactly why
1:47
they behave the way they do. And that's based on, you know,
1:50
geometry, mass, and elemental thermohysical properties.
1:53
Right, because we need to shift the paradigm.
1:54
You have to stop looking at the machine as an isolated entity
1:57
and start viewing it as well half of a coupled system.
2:01
OK, so the fluid it sits in is the other half?
2:03
Precisely whether that is ocean water, highly pressurized steam,
2:08
or a complex refrigerant, when you immerse a structure in a
2:11
fluid, the structure fundamentally loses its
2:14
individual identity. Wait, loses its identity?
2:18
That's a strong way to put it. It is, but it's true.
2:20
It adopts a completely new dynamic identity that's dictated
2:24
by the unseen forces of the medium surrounding it.
2:26
Wow. OK, let's start with the most
2:28
pervasive of those unseen forces.
2:30
The sources we're examining describe a phenomenon that
2:33
happens the exact moment you place a structure into a fluid.
2:36
Right. The vibrational characteristics
2:38
are instantly altered. Yeah.
2:39
Even if the fluid is completely still, I think the literature
2:42
calls it a a quiescent fluid. The natural frequencies of the
2:46
structure decrease and the damping increases.
2:48
Exactly. But the real mechanism we need
2:50
to unpack today is what happens when that object accelerates.
2:54
The texts refer to this as added mass or virtual mass.
2:58
OK, So what is physically occurring in the fluid to 'cause
3:00
this added mass? Like my car doesn't literally
3:03
gain mass on the highway. Right, and that's why the term
3:06
added mass can be slightly misleading.
3:08
The object itself isn't, you know, gaining physical atoms of
3:11
steel. That would defy physics.
3:13
It really would. What is happening is a forced
3:16
reallocation of the surrounding environment.
3:19
So to move an object through a fluid, you can't just magically
3:23
occupy a new space. You have to push the stuff out
3:26
of the way. Exactly.
3:27
You must evict the fluid that currently occupies that space.
3:30
As the front of the object pushes forward, it accelerates a
3:34
massive envelope of fluid out of its path.
3:37
And I assume something has to fill the space it leaves behind.
3:40
Right. Simultaneously, fluid must
3:41
violently rush in behind the object to fill the void left in
3:44
its wake. And all of that fluid possesses
3:48
its own mass and its own inertia.
3:50
So it resists being accelerated. You hit the the nail on the
3:52
head. OK, so the energy you're
3:54
expending isn't just to overcome the inertia of the steel or
3:57
titanium that you physically built?
3:58
No, not at all. You are paying a heavy energy
4:01
tax to physically accelerate the fluid around it.
4:04
That makes so much sense. So how do we quantify that tack?
4:07
Well, the mathematics of this energy tax are actually quite
4:10
beautiful, particularly when we look at ideal incompressible and
4:14
irritational fluids. OK, hit me with the math.
4:17
In fluid dynamics, we define this flow using a velocity
4:20
vector, denoted as is V. The math states that this
4:24
velocity vector is equal to the gradient of a potential
4:27
function, so V equals nablify. OK, let's translate that math
4:32
for the listener, because just hearing V equals nablify or the
4:36
gradient of a potential function doesn't give us physical
4:39
picture. Fair enough.
4:40
The gradient essentially represents the rate of change in
4:43
all three spatial directions, like up, down, left, right,
4:46
forward, backward. Yes, exactly.
4:48
That spatial mapping is incredibly critical because it
4:51
leads us directly into Laplace's equation, which is what the
4:54
requirement is that the Laplacian of this potential
4:56
function must equal 0. Mathematically, that's nabla
4:59
squared Phi equals 0. OK.
5:01
And physically, Laplace's equation enforcing that strict 0
5:05
value is essentially A strict rule of real estate in the
5:09
fluid, right? That's a great way to put it, a
5:11
rule of real estate. Yeah, like it means the fluid is
5:14
incompressible. Water cannot be magically
5:17
squished into a smaller volume, and it can't spontaneously
5:20
disappear. Precisely, if a submarine pushes
5:24
into a new cubic meter of space, an exact equal cubic meter of
5:28
water must get out of its way. The math is just, you know,
5:32
meticulously balancing the volumetric checkbook.
5:35
I love that balancing the checkbook and by.
5:37
Balancing that checkbook, we can derive the exact force the fluid
5:41
exerts back onto the accelerating body.
5:43
OK, what does that equation look like?
5:45
The equation reveals that the fluid force F is equal to the
5:48
negative of this virtual mass term little M multiplied by the
5:52
partial derivative of the flow velocity U with respect to time
5:56
T. OK, wait, let me process that.
5:58
The derivative of velocity with respect to time is literally
6:01
just acceleration. So strip down this F = -m times
6:05
partial U over partial T is literally just Newton's second
6:08
law. Force equals vast times
6:10
acceleration. It is.
6:11
It's that fundamental. But the mass in this specific
6:13
equation is not the physical weight of the object.
6:16
It is this dynamic, invisible envelope of fluid.
6:19
You've got it perfectly. But this brings up a really
6:21
critical question of scale for the listener.
6:23
OK, what's that? If every object moving through a
6:26
fluid experiences added mass, why don't automotive engineers
6:31
constantly worry about the added mass of the air when they're
6:34
designing a car? That is a fantastic point.
6:37
Automotive engineers can safely ignore it because of the density
6:41
ratio. Rule density ratio.
6:42
Rule. Yeah.
6:43
So added mass only becomes a significant design altering
6:47
fraction of the total dynamic mass when the density of the
6:51
fluid is comparable to the density of the structure.
6:54
Oh I see. And air is incredibly sparse
6:57
compared to the dense steel and aluminum of a car.
6:59
Exactly, the mass of the air being shoved out of the way is
7:02
literally just a rounding error compared to a 210 chassis.
7:05
But the ocean is a completely different story.
7:07
A vastly different story. Seawater is roughly 800 times
7:11
denser than air as sea level. Wow. 800 times.
7:15
Yeah. So consider a massive military
7:17
submarine. When the captain orders a hard
7:19
acceleration or a rapid dive, the propellers aren't just
7:23
fighting the inertia of the submarine's titanium hull,
7:25
right? They are fighting the inertia of
7:27
thousands of tons of heavy water that must be dragged along with
7:31
the vessel. That's terrifying from a control
7:33
perspective. It really is.
7:35
If a control systems engineer programmed the diving planes and
7:38
rudders based solely on the dry weight of the submarine, the
7:42
vessel would dangerously overshoot every single maneuver.
7:45
Because the momentum of that virtual water envelope would
7:48
just keep pushing the submarine long after the physical engines
7:52
tried to stop it. This means a designer cannot
7:56
just, you know, weigh apart on a scale in a laboratory and call
8:00
it a day. You have to calculate its
8:01
dynamic wet weight. Yes, it's wet weight.
8:04
That's the crucial metric. And if the submarine is dragging
8:07
this massive virtual envelope, I have to assume the specific
8:11
shape of the hull dictates exactly how much water gets
8:14
trapped in that. Envelope.
8:15
Oh, geometry absolutely dictates the physics here.
8:17
Because a blunt brick like object is going to drag a
8:20
completely different volume of water than a sleek teardrop
8:24
shape. Exactly the specific contours of
8:27
the object determine the flow potential, which in turn defines
8:31
the exact magnitude of the added mass.
8:34
So what kind of numbers are we talking about?
8:35
Does the literature give us examples?
8:37
It does. The sources give us some
8:39
pristine geometric examples. Let's start with two dimensions.
8:43
If you take a perfectly infinitely long cylinder of a
8:47
certain radius, let's call it a. OK, an infinitely long cylinder
8:50
radius. A.
8:51
The mathematical derivation shows that the added mass per
8:54
unit length is equal to the fluid density rho multiplied by
8:58
π * a ^2. Wait, hold on rho times π * a ^2
9:03
Yes, that is fascinating because π times the radius squared gives
9:07
you the cross-sectional area right?
9:09
And multiplying that area by the density rho.
9:11
Rho gives you the exact mass of the fluid that the cylinder
9:15
displaced. You see the elegance of it.
9:17
Yeah, for perfect cylinder, the added mass is an elegant 1 to
9:20
one ratio. With the displaced fluid mass,
9:23
it's perfectly equal. It is a rare moment of
9:25
mathematical symmetry and engineering, but as you know,
9:28
mechanical design rarely relies on infinitely long, perfectly
9:33
smooth cylinders. Right, we have brackets and fins
9:36
and all sorts of complex. Shapes, and when you introduce
9:39
sharp geometries, the added mass penalty absolutely skyrockets.
9:43
Really, how much worse does it get?
9:44
Well, if you take a square cross section that shares the exact
9:48
same primary dimensions as that cylinder, the added mass factor
9:52
jumps significantly up to 1.51 times the displaced volume.
9:56
Whoa, 1.51. That jump makes physical sense
10:01
though, when you think about the boundary of the shape.
10:03
How so? Walk the listener through your
10:04
thought process. Well, a smooth cylinder allows
10:07
the flue to glide around its curvature right?
10:10
The water parts at the front and smoothly rejoins at the back.
10:13
But a square has harsh 90° corners.
10:17
Those corners have to disrupt the flow, creating zones of
10:20
stagnation where extra dead fluid gets trapped and dragged
10:23
along. Exactly.
10:24
You aren't just moving the square, you are moving the
10:26
square plus roughly 4 triangular wedges of trapped water on those
10:31
flat faces. That's wild.
10:33
What about in three dimensions? Moving into 3 dimensions, the
10:35
geometry continues to dictate the terms.
10:38
A perfect sphere has an added mass equal to exactly half of
10:41
its displaced fluid mass, or 2/3 rho times π * a ^3.
10:46
OK, so half for a sphere. But a solid cube drops to a
10:49
coefficient of roughly 0.68 times its displaced mass, or 0.7
10:55
times rho times a ^3. Wait, so a sphere is .5 but AQ
11:00
is .68? Yes, the flow field around A3
11:03
dimensional corner is vastly different from A2 dimensional
11:05
corner. OK, but discussing these ideal
11:08
flow fields brings us to one of the most famous and historically
11:11
frustrating concepts in fluid mechanics.
11:14
Yeah, yes, you're talking about the de L'ambert paradox, yes.
11:17
For the listener, this is a concept that historically caused
11:20
a massive rift between theoretical mathematicians and
11:23
practical engineers. It really did.
11:25
They were practically at each other's throats.
11:27
The paradox formulated by Jean Laurent d'alembert in the 18th
11:30
century states that for an ideal invisid fluid, meaning a perfect
11:34
fluid with 0 viscosity and 0 internal friction.
11:37
Right, a mathematically perfect fluid.
11:39
The drag force on a body moving at a steady constant velocity is
11:44
absolutely 0. Zero drag.
11:47
Which sounds completely wrong. Well, from a purely mathematical
11:51
standpoint, in an invisible flow it is completely sound.
11:54
As the object pushes through this frictionless fluid, the
11:57
fluid splits at the front, creating a zone of high
12:00
pressure. OK, it flows around the body and
12:02
perfectly symmetrically reconverges at the rear,
12:05
creating an equal zone of high pressure pushing forward.
12:08
Oh, I see. The pressure at the front
12:10
cancels the pressure at the back.
12:11
Precisely. The net drag force is
12:13
mathematical zero. The force of added mass we've
12:16
been talking about only exists when the object accelerates or
12:19
decelerates, which unbalances that pressure field.
12:22
But the practical engineers building ships and aqueducts in
12:25
the 1700s looked at these mathematicians and thought they
12:28
were totally insane. Of course they did.
12:30
Anyone who has ever rode a boat knows that water exerts A
12:33
continuous drag force even when you are moving at a constant
12:36
speed. Right.
12:37
The disconnect existed for over a century until we truly
12:41
understood boundary layers and viscous friction.
12:44
Yes, real fluids have friction. But the paradox remains A vital
12:48
teaching tool, right? Because it isolates the purely
12:51
inertial component of the fluid from the friction component.
12:54
Exactly. It forces us to separate
12:56
acceleration forces from velocity forces.
12:59
We can visualize the separation through a really simple physical
13:03
analogy. OK, let's use the analogy of
13:05
standing chest deep in a swimming pool holding a flat,
13:09
wide wooden paddle. Erfect.
13:11
If you hold the paddle slat against the water and try to
13:14
rapidly wave it back and forth, accelerating and decelerating it
13:18
rapidly, it feels impossibly heavy.
13:21
Right. It takes immense effort.
13:22
And that massive resistance is the added mass fighting your
13:25
acceleration. Yes, but if you turn that exact
13:28
same paddle sideways, edge 1st, and thrust it forward at a
13:32
constant speed, it slices right through the water with minimal
13:35
effort. The difference is staggering.
13:37
It's night and day, and that paddle analogy reveals another
13:41
layer of complexity that goes beyond simple scalar
13:43
mathematics. What do you mean by scalar
13:45
mathematics? I mean just assigning a single
13:47
number to the added mass. The resistance you feel depends
13:51
entirely on the orientation of the object relative to the
13:54
direction of acceleration. Right, fly versus edge first.
13:57
Exactly. So in engineering, we cannot
13:59
just assign a single number to the added mass of a complex
14:02
shape. We must represent it using an
14:05
added mass tensor. A tensor OK for the listener.
14:08
A tensor in this context is essentially a matrix, right?
14:10
Like a spreadsheet of cross coupling coefficients.
14:13
That's a great visualization. It maps out how accelerating in
14:16
one specific direction generates forces in all the other spatial
14:20
directions. Yes.
14:21
The mathematics denote this as a matrix where the coefficient M
14:25
subscript IJ dictates the force. M Subscript IJ.
14:29
In many ideal cases, this matrix is symmetric, meaning the
14:32
coefficient M subscript IJ equals M subscript Ji.
14:37
The reciprocal coupling is balanced.
14:39
OK, but what is the critical take away for a mechanical
14:41
designer here? The take away is that for highly
14:43
asymmetrical structures, accelerating straightforward can
14:46
induce massive unexpected forces in entirely different
14:49
directions. Wait, let me make sure I'm
14:51
visualizing this correctly. If I design an asymmetrical
14:54
object and I alley A thrust vector perfectly straight along
14:58
the X axis right, the fluid might actively shove the object
15:02
sideways along the Y axis. It will shove it sideways and it
15:06
will very likely induce A rotational moment as well.
15:09
It will try to spin the object. That is insane.
15:11
Let's consider the design of an underwater remotely operated
15:15
vehicle, an ROV like the ones used for deep sea pipeline
15:18
inspection. OK.
15:19
So like a little submarine drum? Exactly.
15:21
These vehicles are often modular.
15:23
An engineer might bolt a heavy, bulky camera payload onto the
15:26
left side of the vehicle and then add a smooth buoyancy block
15:30
to the right side to keep the static weight balanced in the
15:33
water. So it floats perfectly level
15:35
when it is sitting still, yes. But dynamically the shape is
15:38
completely asymmetrical. The bulky camera has a
15:41
drastically different added mass tensor than the smooth buoyancy
15:45
block. Oh, I see where this is going.
15:47
The moment the ROV fires its rear thrusters to accelerate
15:50
straightforward, the fluid will resist the bulky camera much
15:54
harder than the smooth right side.
15:56
The fluid itself generates a massive twisting force.
15:59
So if the control algorithms do not actively calculate and
16:02
counteract that specific added mass tensor, the drone will just
16:07
immediately spiral out of control the second it tries to
16:09
move forward. It would just do doughnuts in
16:11
the water. So the fluid isn't just a
16:13
passive background environment, it acts as an active mechanical
16:17
linkage that manipulates the object based on its geometry.
16:20
Precisely, it is an active participant.
16:23
And this entire discussion so far has only been about a single
16:26
isolated object moving through a fluid.
16:29
The reality of mechanical design is rarely that lonely.
16:33
Very true. We need to transition to what
16:36
happens when you have multiple structures in close proximity
16:39
immersed in the same dense medium.
16:41
Ah. Yes, this brings us to the
16:44
concept of inertial coupling. Inertial coupling.
16:46
When two or more objects are placed close together in a
16:49
fluid, the added mass is no longer just a localized property
16:52
of 1 isolated shape. Because the flows interfere.
16:55
Exactly. The flow field of the first
16:58
object intersects with the flow field of the second object.
17:01
The added mass becomes a highly sensitive function of the
17:04
physical spacing between them. The classic configuration
17:08
analyzed in the text we have is a cylindrical rod centered
17:12
perfectly inside a larger fluid filled outer cylinder.
17:16
Essentially a pipe within a pipe.
17:18
Right, with the annular gap between them completely filled
17:21
with liquid. Now, if you perform this
17:23
experiment in a vacuum chamber with absolutely no fluid,
17:26
shaking the inner rod does absolutely nothing to the outer
17:29
pipe. Right, because they are
17:31
mechanically isolated. There's just empty space between
17:33
them. Exactly.
17:34
But the exact moment you fill that gap with a dense fluid, the
17:38
fluid transforms into an invisible, visible, highly rigid
17:42
mechanical spring. Because of the incompressibility
17:44
rule we talked about earlier. Precisely.
17:46
You remember the real estate rule?
17:47
Yeah. If the inner rod accelerates
17:49
toward the outer wall, the fluid caught between them must be
17:52
violently displaced. And that fluid is forced
17:55
directly against the outer cylinder, transferring all of
17:59
that kinetic energy. The mathematical equations of
18:02
motion become fully coupled. So the physical force exerted on
18:06
the outer pipe is directly dictated by the acceleration of
18:09
the inner rod communicating entirely through that fluid gap.
18:13
Yes, the sources indicate that this coupling effect increases
18:17
dramatically as the fluid gets denser and as the gap gets
18:20
tighter. So if you have a massive inner
18:22
spiel rod and only say a 1mm gap of water separating it from the
18:27
outer shell. Violently shaking the inner rod
18:29
is going to instantly jerk the outer shell as if they were
18:32
physically bolted together. The flu just locks them
18:34
together. It does.
18:35
And while a pipe in a pipe is a neat laboratory example, the
18:39
industrial applications of this phenomenon are monumental.
18:42
Let's give the. Listener a real world example.
18:44
Consider the heart of a pressurized water nuclear
18:47
reactor or a massive petrochemical refinery.
18:50
The primary method of transferring energy in these
18:52
facilities is through shell and tube heat exchangers.
18:55
Right, A commercial nuclear steam generator isn't just one
18:58
pipe, is an array of thousands of U-shaped metal tubes.
19:02
Thousands of them. Packed incredibly tightly inside
19:05
a massive steel vessel with thousands of gallons of high
19:08
pressure water rushing past them every minute.
19:10
Now, if an engineer looks at a single tube in isolation, they
19:14
might calculate its natural resonant frequency and determine
19:17
that it is plenty stiff enough to withstand the flow of the
19:20
water. Furthermore, if the tubes are
19:23
spaced widely apart, the literature generally sites a
19:26
distance of more than two diameters from center to center.
19:29
The fluid effects remain mostly localized.
19:32
So the fluid just adds a bit of static virtual mass to each
19:35
independent tube. No big.
19:36
Deal. No big deal, But what is the
19:38
entire goal of a heat exchanger? Thermal efficiency.
19:41
Right. And which means maximizing
19:44
surface area so engineers pack those thousands of tubes as
19:47
close together as physically possible without them actually
19:50
touching. And by packing them tightly,
19:52
they invite catastrophic inertial coupling.
19:54
They absolutely do. The tiny fluid gaps between the
19:57
tubes act as a massive network of invisible springs, tying the
20:02
entire array together. So when one tube vibrates
20:05
slightly due to the water flowing past it, it pushes the
20:08
fluid into the adjacent tube. Yes, which pushes the fluid into
20:12
the next tube and the next. The fluid ties them into a
20:15
single massive vibrating ecosystem.
20:18
And this coupling drastically lowers the natural frequencies
20:21
of the system as a whole. It creates highly complex group
20:25
vibration modes. Wait, group vibration modes?
20:28
The tubes no longer vibrate independently.
20:30
They lock into synchronized patterns.
20:32
If the velocity of the fluid flowing through the shell
20:35
happens to align with one of these new lower group
20:38
frequencies, the system hits resonance.
20:40
Oh no. Oh yes, the entire array of
20:43
thousands of tubes will begin to oscillate in perfect destructive
20:46
harmony. The fluid amplifies the energy
20:49
until the tubes violently clash into each other.
20:51
Rupturing the metal and causing a catastrophic failure of the
20:54
heat exchanger. A multimillion dollar failure
20:56
just because they didn't account for the fluid acting as a
20:58
spring. That is wild and it isn't just
21:01
the physical tubes violent and hitting each other that the
21:04
designer has to worry about, right?
21:05
No, it does not, because all that chaotic coupled movement,
21:09
the fluid rushing and separating around the tightly packed
21:11
geometries, generates severe pressure fluctuations in the
21:15
fluid itself. That's right.
21:17
And what is a fluctuating pressure wave traveling through
21:20
a fluid medium? It is sound.
21:22
It is sound which brings us perfectly to the acoustics of
21:25
fluid flow. Vibration and acoustics are
21:28
essentially two sides of the same thermodynamic coin, right?
21:32
Sound is, at its core, mechanical vibration propagating
21:35
through a fluid. Absolutely.
21:37
What the study of fluid mechanics reveals is a stunning
21:40
universality in how sound waves are generated by turbulent flow,
21:44
completely regardless of the scale of the machine.
21:46
Or whether it's tiny or huge, the physics are the same.
21:48
Exactly. The moment a fluid stops flowing
21:51
perfectly smoothly along a surface, the moment it
21:53
separates, tumbles and forms swirling vortices, it begins
21:56
exerting random chaotic pressure pulses on the structure.
22:00
And those pulses don't just shake the metal, they radiate
22:03
outward as acoustic energy. The underlying physics of this
22:07
aerodynamic noise generation are governed by the exact same
22:11
mathematical principles whether you are looking at a microscopic
22:14
system or an aerospace leviathan.
22:16
The contrast the sources used to illustrate this is incredible.
22:20
Picture a tiny half inch pressure relief valve on a
22:23
factory compressed airline. When the pressure spikes in that
22:27
valve pox open, a high velocity jet of gas screams out into the
22:31
room. It produces A piercing,
22:34
incredibly loud hiss. We've all heard that noise in a
22:36
factory. Now compare that tiny valve to
22:39
the exhaust nozzle of a commercial turbofan jet engine
22:42
producing 100,000 lbs of thrust at takeoff.
22:45
The. Physical scale difference is
22:46
astronomical. The temperatures are vastly
22:49
different. The thrust is vastly different,
22:52
yet the acoustical characteristics of that
22:54
screaming pressure relief valve share the exact same fundamental
22:58
noise generating mechanisms as the massive jet engine.
23:01
So you're saying the turbulent shear layer, the boundary where
23:04
the fast moving air rips against the slow moving ambient air
23:06
coming off that tiny little valve?
23:08
Yeah, it follows the exact same decay rate and acoustic profile
23:12
as the exhaust of the jet engine.
23:14
It genuinely does. The mathematics governing the
23:17
turbulent eddies, the quadrupole sound sources defined by Sir
23:21
James Light Hills acoustic analogy, and the conversion of
23:23
kinetic flow energy into acoustic pressure waves are
23:26
totally identical. Light Hills Acoustic analogy,
23:29
yes. It is a universal scaling law.
23:31
The literature also draws a Direct Line between the small
23:35
plastic cooling fan inside a desktop computer or household
23:38
air conditioner and a massive multi stage axial compressors
23:42
inside a gas turbine. It's the same mechanism.
23:46
The flow separation that happens over the blade of a 2 inch
23:49
plastic computer fan generates noise using the exact same
23:52
vortex shedding mechanisms as a solid titanium turbine blade.
23:55
That's amazing. Even the aerodynamic knows you
23:57
hear inside your car on the highway.
23:59
You know, the wind rushing past the sleek side mirror that is
24:02
governed by the same flow separation physics that dictate
24:04
the aerodynamic noise of a commercial airplane wing slicing
24:08
through the atmosphere. It is, and for a mechanical
24:10
engineer this universality is a massive, massive advantage.
24:13
Why is that? Because it means you don't have
24:15
to build a full scale jet airliner to understand aero
24:18
acoustics. Oh, I see.
24:20
If you spend your time meticulously solving an annoying
24:22
whistling problem caused by vortex shedding on a small
24:26
automobile side mirror, you are essentially mastering the exact
24:30
same physical laws that quiet down the landing gear of a
24:33
Boeing 777. You are interacting with
24:36
universal principles. Exactly.
24:38
The acoustic power generated by turbulence scales predictably
24:41
with flow velocity, fluid density, and a characteristic
24:45
length dimension. Understand the physics of the
24:48
small system and you hold the key to predicting the behavior
24:51
of the giant system. OK, so to survive these massive
24:55
fluid forces, the added mass, the inertial coupling, the
24:58
violent acoustic vibrations we just talked about, the engineer
25:01
must select the right materials. You can't just build a submarine
25:04
out of tinfoil. No, you definitely cannot, which
25:07
leads us into analyzing the material palette.
25:10
We need to look at the raw thermophysical data of
25:12
structural solids across extreme temperature ranges.
25:16
Material selection in dynamic fluid environments is a really
25:19
complex balancing act. You are not simply looking at a
25:23
materials cost or its static yield strength.
25:26
What else are you looking at? You must evaluate its density,
25:28
its heat capacity, and its thermal conductivity.
25:32
Crucially, you must understand how those specific properties
25:35
behave when the system experiences massive thermal
25:39
loads. So temperature changes
25:41
everything. It really does.
25:42
The extreme ends of the periodic table provide the best
25:45
perspective on this. Let's examine tungsten.
25:48
Tungsten Tungsten is the absolute heavyweight champion of
25:52
structural metals. It really.
25:53
Is the sources list is density at over 19,000 kilograms per
25:57
cubic meter. It is astonishingly dense and
26:00
heavy. But its density is only part of
26:02
the story. What makes tungsten an absolute
26:05
engineering marvel is its extreme thermophysical
26:08
stability. Stability.
26:09
What does that look like? When you expose tungsten to
26:12
immense thermal energy, literally blasting it with
26:15
thousands of degrees of heat, it's a atomic lattice remains
26:18
remarkably rigid. It retains a highly stable heat
26:21
capacity and thermal conductivity profile.
26:24
It does not flinch under thermal duress, providing massive,
26:28
unwavering structural inertia. Let's contrast that dense
26:31
stability with the featherweight of the material world lithium.
26:35
The elemental density of lithium is listed at only 534 kilograms
26:40
per cubic meter. That is incredibly low.
26:43
It is nearly 40 times lighter than tungsten.
26:45
It is so light that solid lithium metal will literally
26:48
float on water. But while it lacks density, its
26:51
thermal behavior is incredibly aggressive.
26:54
The heat capacity of a material, which is denoted as C sub P,
26:57
measures how much thermal energy is required to raise its
27:00
temperature. Right as lithium heats up its
27:02
heat capacity, C sub P absolutely skyrockets, reaching
27:06
massive levels over 5200 joules per kilogram Kelvin.
27:11
To put that into physical terms for the listener, how does the
27:14
atomic structure of lithium actually absorb all that heat
27:17
compared to tungsten? Well, heat at the macroscopic
27:19
level is simply atomic vibration at the microscopic level.
27:22
We track this vibrational energy using quasi particles called
27:26
phonons. Phonons, right?
27:28
Tungsten has a very dense, tightly bound crystal lattice.
27:33
It takes a massive amount of energy to get those heavy
27:35
tungsten atoms to vibrate significantly.
27:38
So it resists heating up. Exactly.
27:40
Lithium, however, has a much looser, lighter atomic
27:43
structure. When you apply heat, the lithium
27:46
atoms vibrate intensely and freely.
27:49
The material acts as an absolute thermal sponge, soaking up
27:53
massive amounts of vibrational energy.
27:55
And looking at the other metals in the sources, we see elements
27:58
like silver and platinum occupying different specialist
28:00
roles. Yes.
28:02
Let's look at silver. Silver boasts an astonishing
28:05
thermal conductivity, represented by the Greek letter
28:07
Lambda, pushing over 400 watts per meter Kelvin.
28:11
It doesn't just absorb heat, it rapidly shuttles it away across
28:14
its crystal lattice. It's a highway for heat.
28:17
And platinum, conversely, remains dense, heavy and
28:20
structurally uncompromising, behaving much more like
28:23
tungsten. Every single material offers a
28:26
distinct physical trade off. So let me play devil's advocate
28:29
here to tie these concepts together for the listener.
28:31
OK, go for it. If lithium is so incredibly
28:33
light and it has the capacity to absorb massive amounts of
28:37
thermal energy, why don't we see it used more in heavy industry?
28:41
Why not use it everywhere? Yeah, why aren't we building the
28:44
internal structures of submarines or the cooling fins
28:47
of marine heat exchangers out of lithium alloys to save weight?
28:50
That question requires us to bridge our knowledge of
28:53
materials back to our very first discussion today, the density
28:58
ratio rule of added mass. Right, the ratio of the
29:01
structures mass to the fluids mass.
29:03
Imagine you build a complex, highly engineered internal
29:06
structure out of a super lightweight lithium alloy.
29:09
OK, I'm imagining it. You then immerse that structure
29:12
in a dense fluid like pressurized water or a heavy
29:15
industrial coolant. Because the solid structure is
29:18
so exceptionally light, the virtual mass of the dense fluid
29:21
surrounding it completely dwarfs the physical mass of the
29:24
lithium. Oh.
29:26
The fluid becomes the dominant inertial force in the system.
29:28
Exactly. If there's any turbulence in the
29:31
flow, any vortex shedding or acoustic pressure waves, the
29:34
heavy fluid will violently throw that lightweight structure
29:36
around. Because it has no mass to
29:38
resist. It right, the inertial coupling
29:41
between the lightweight components will be catastrophic.
29:43
The solid structure simply lacks the inherent mass to anchor
29:47
itself against the invisible springs and forcing functions of
29:50
the fluid. It would be like trying to hold
29:51
a Styrofoam cup completely still underwater while a wave pool
29:56
generator is running. The cup just doesn't have the
29:58
mass to fight the water. That is a perfect physical
30:02
visualization. Sometimes an engineer absolutely
30:06
relies on the massive, uncompromising density of
30:09
tungsten, nickel or platinum. You don't just need those metals
30:13
for their high yield strength, you need them for their dynamic
30:15
inertia. Yes, you need a massive physical
30:18
anchor to resist the powerful coupled fluid dynamics trying to
30:22
tear the entire system apart. That is a profound paradigm
30:26
shift for mechanical design. You aren't just choosing a metal
30:29
because it is strong enough to hold a static load, you are
30:32
choosing it for its dynamic mass ratio against the specific fluid
30:36
it will operate in. It is a totally different way of
30:39
thinking. Which brings us to the final,
30:41
and perhaps most complex piece of the engineering puzzle.
30:44
Today, let's hear it. The structures do not act alone,
30:47
and the fluids rushing through them are rarely just simple
30:50
water. Right.
30:51
In many industrial applications, particularly power generation in
30:54
HVAC, the fluids pumping through the system are highly complex
30:58
chemical refrigerants. Yes, up until now we have mostly
31:01
discussed fluids as if they were consistent, well behaved
31:03
mediums. But when we analyze industrial
31:06
refrigerants like R12R134A or R22, their thermal physical
31:11
properties are anything but consistent.
31:14
The data tables provided in the sources for these refrigerants
31:16
are vast, and they highlight a critical reality.
31:20
These fluids change radically with temperature.
31:22
They do, and when they undergo a phase shift, meaning boiling
31:25
from a liquid into a vapor, the fundamental rules of physics
31:29
governing their behavior practically invert.
31:32
Invert like they go backwards. Pretty much.
31:34
Let's examine viscosity, denoted by the Greek letter Ada.
31:38
Viscosity is the measure of a fluid's internal friction, it's
31:41
resistance to flow. So it dictates how hard a pump
31:44
has to work to push the fluid through a pipe.
31:46
Exactly. Now if you take a liquid
31:48
refrigerant and apply heat to it, the data shows that it's
31:52
viscosity drastically drops. It becomes thinner, much more
31:55
runny. At a molecular level, the liquid
31:57
state is held together by intermolecular bonds, right?
32:00
Yes, when you add thermal energy, you are providing the
32:04
molecules with enough kinetic energy to begin breaking those
32:07
weak bonds. And as the bonds break, the
32:09
molecules can slide past one another with much less
32:12
resistance. So the internal friction
32:15
decreases that. Perfectly aligns with our
32:17
everyday experience. If you put cold, thick honey in
32:20
the microwave and heat it up, it becomes a thin liquid that pours
32:24
easily. Exactly.
32:25
But the data tables reveal a wild contradiction when we look
32:28
at the vapor state. If you take that exact same
32:31
refrigerant, completely vaporize it into a gas, and then continue
32:35
to heat it up, its viscosity actually increases.
32:38
Really. Yes, the gas gets thicker.
32:40
It becomes more resistant to flow.
32:42
How does? Heating a gas create more
32:45
friction that feels completely backwards.
32:47
It is a beautiful demonstration of the kinetic theory of gases.
32:51
In a liquid, the molecules are packed tightly together,
32:54
fighting intermolecular bonds. But in a gas, the molecules are
32:58
already spread far apart. Those bonds are completely
33:02
broken. So adding heat can't break bonds
33:04
that don't exist anymore. Exactly.
33:06
Instead, the thermal energy translates directly into kinetic
33:10
velocity. The gas molecules begin zipping
33:13
around the space at much higher speeds.
33:15
OK if I follow that. Because they are moving so much
33:18
faster, they collide with one another much more frequently.
33:22
Let's use an analogy here for the listener.
33:25
Imagine trying to walk down a long enclosed hallway filled
33:29
with people. A crowded hallway.
33:30
OK. If everyone in the hallway is
33:32
standing perfectly still, which represents a cold gas with low
33:35
kinetic energy, you can easily weave your way through the
33:38
crowd. The resistance is low, right?
33:41
But if you add heat, if suddenly everyone in that hallway starts
33:44
sprinting back and forth in a frantic panic, trying to walk
33:47
through that chaotic high energy crowd is going to result in
33:51
constant violent collisions. You're.
33:52
Going to get bumped around constantly.
33:54
The friction of the crowd is massively higher.
33:56
That hallway analogy perfectly captures the macroscopic effect
34:00
of microscopic momentum transfer.
34:02
The increased rate of high speed molecular collisions across the
34:06
different layers of the flowing gas creates a severe drag force.
34:10
So heating a liquid reduces friction by breaking bonds, but
34:13
heating a gas increases friction by increasing the collision
34:16
rate. You summarize it perfectly.
34:18
That inversion fundamentally alters how a mechanical
34:21
compressor operates depending on the exact temperature of the gas
34:24
it is pushing. It absolutely does, but
34:27
viscosity isn't the only property that drastically
34:30
shifts. We also need to look at the
34:31
speed of sound within the refrigerant represented by the
34:35
variable W. OK, the speed of sound W the
34:38
speed of sound is not a universal constant, right?
34:40
No, it is a mechanical property. The specific medium it travels
34:44
through it is dictated by the fluid's density and it's bulk
34:48
modulus, which is essentially a measure of the fluid's stiffness
34:52
or compressibility. O In the dense liquid state of
34:55
these refrigerants, the speed of sound is incredibly fast, the
34:58
data says, often well over 1000 meters per second.
35:02
Yes, because the liquid is relatively stiff, so the
35:04
pressure wave transfers rapidly from molecule to molecule.
35:07
But as the liquid warms up and expands, becoming physically
35:10
squishier, that acoustic speed plunges.
35:13
And when the fluid undergoes the phase shift into a vapor, the
35:17
acoustic environment changes entirely.
35:19
How so? In the vapor state, the gas is
35:22
highly compressible. The bulk modulus is very low.
35:25
Consequently, the speed of sound drops to a sluggish crawl, often
35:29
hovering around 150 meters per second.
35:32
Wow, from over 1000 down to 150. Furthermore, the data shows that
35:36
in the vapor state, this low speed of sound is stubbornly
35:39
stable. Stubbornly stable.
35:41
Yeah, you can add a significant amount of heat to the vapor, and
35:43
the speed of sound barely twitches compared to the massive
35:46
swings you see in the liquid phase.
35:48
OK, but why does this sluggish acoustic speed in the vapor
35:52
matter so much to a designer? It ties directly back to our
35:55
earlier discussion on acoustics and flow separation, right?
35:58
It does. It is the absolute limiting
36:01
factor for flow velocity. You have to remember the speed
36:04
of sound is the speed at which pressure information travels
36:07
through a fluid, right? If a compressor is pushing a
36:09
vapor refrigerant through a narrow restriction like an
36:12
expansion valve, the flow accelerates because the speed of
36:16
sound in the vapor is so low, only 150 meters per second.
36:21
The accelerating gas can very easily hit Mach one as it
36:24
squeezes through the orifice. And when a fluid hits Mach one
36:28
inside a pipe, you achieve what is called choked.
36:31
Flow. Precisely choked flow.
36:33
Once the flow hits the speed of sound at the restriction, it is
36:36
physically impossible for it to flow any faster, no matter how
36:39
much pressure the compressor applies upstream.
36:41
Why not? Because the pressure wave
36:43
literally cannot travel backward against a flow moving at the
36:46
speed of sound. The fluid creates A
36:48
thermodynamic traffic jam. And a thermodynamic the traffic
36:51
jam isn't just inefficient, it is violently destructive.
36:54
Well. Incredibly destructive.
36:56
The choked flow creates severe shock waves downstream of the
36:59
valve. Those shock waves cause massive
37:02
acoustic vibrations, which couples with the structural
37:04
pipes. Leading directly back to the
37:06
destructive resonance and metal fatigue we discussed in the heat
37:09
exchanger arrays, the valve will literally hammer itself to
37:13
pieces. So if an engineer does not know
37:15
the exact temperature, phase and corresponding speed of sound of
37:20
their specific refrigerant, they have absolutely no idea when
37:25
their system will choke. They are flying completely
37:27
blind, and to make matters even more complex, the distinction
37:31
between liquid and vapor eventually completely disappears
37:34
if you push the fluid hard enough.
37:36
Wait, really disappears entirely?
37:37
Yes. The sources track a property
37:40
called surface tension, denoted by the Greek letter Sigma.
37:43
Surface tension is what creates a meniscus.
37:46
Like the little curve at the edge of a glass of water.
37:48
Exactly. It is the physical boundary
37:50
layer that separates the liquid pool at the bottom of a tank
37:53
from the vapor gas floating above it.
37:55
OK, So what happens as you heat it?
37:56
As you heat a fluid inside a sealed pressurized container,
37:59
the liquid expands and becomes less dense, while the vapor
38:02
above it becomes more pressurized and denser.
38:05
The surface tension Sigma between them progressively
38:08
drops. Because they're becoming more
38:10
similar to each. Other exactly, and eventually
38:13
the fluid reaches its critical point.
38:15
And the critical point, the surface tension hits absolute 0.
38:18
The boundary completely vanishes.
38:20
The liquid and the gas merge into a supercritical fluid.
38:24
A supercritical fluid that sounds like science fiction It.
38:27
Kind of does, but it's very real.
38:29
It is an indistinguishable opaque blur that possesses the
38:33
heavy density of a liquid but the expansive compressibility of
38:37
a gas. The rules of physics for
38:39
supercritical fluids require entirely different engineering
38:43
equations. So an engineer designing a high
38:45
pressure system isn't just dealing with a simple liquid or
38:49
a simple gas. Not at all.
38:50
They're dealing with a thermodynamic chameleon.
38:53
As the operating temperature shifts, this fluid fundamentally
38:57
alters its speed of sound, its capacity to absorb heat, and
39:01
completely reverses its rules of internal friction.
39:04
Which brings us to a final overarching paradigm shift for
39:07
the listener to take away from this entire deep dive today.
39:10
OK, let's bring it all. Together, we began by looking at
39:12
added mass, the reality that a solid structure's physical
39:16
identity, it's very mass in inertia, is instantly altered
39:20
depending on the fluid enveloping it.
39:22
Right, the car in the atmosphere versus the submarine in the
39:24
ocean. We saw how changing a shape from
39:27
a smooth cylinder to a sharp square magnifies that virtual
39:31
mass. And how the D'alembert paradox
39:34
taught us to separate acceleration forces from
39:37
constant velocity drag. We explored how proximity to
39:40
other structures creates a network of invisible mechanical
39:42
springs, locking independent pipes into a single coupled,
39:46
violently vibrating system. We discussed how to anchor
39:49
against those forces using the massive density of tungsten
39:52
rather than the thermal sponge of lithium there.
39:54
Finally, we saw how the fluids themselves are entirely unstable
39:57
entities, flipping their own physical rules.
40:00
Changing viscosity behaviors and creating destructive shock waves
40:04
when shifting from liquid to vapor.
40:06
Therefore, a true engineering design is never just a static
40:09
object sitting on A2 dimensional blueprint.
40:11
It is an active, dynamic ecosystem.
40:13
An ecosystem. I love that.
40:15
The empty space around your parts, the invisible fluid
40:18
rushing through the annular gaps, the ambient temperature.
40:21
These are not just passive environmental conditions, they
40:24
are active physical components of your machine, exerting forces
40:28
just as real as the bolts. The welds and the steel.
40:31
So if you only design the metal and ignore the fluid.
40:34
You are ignoring half of your machine and the fluid will
40:37
inevitably tear the metal apart. That is profound.
40:40
To bring this all the way back to the beginning, that concrete,
40:43
predictable blueprint we expect in mechanical design isn't
40:47
necessarily wrong. We just needed to learn how to
40:50
look at the negative space on the paper.
40:52
The fluid isn't just getting in the way.
40:54
The fluid is the mechanism. If you haven't already, give us
40:57
a like and follow and please leave us a comment if you enjoy
41:00
the podcast. We'd love to hear from the folks
41:02
who make this possible.