Latest / Mechanical Engineering Made Simple / (#97) From Molecular Bonds to Mechanical Motion

Transcript

(#97) From Molecular Bonds to Mechanical Motion · Mechanical Engineering Made Simple

0:00
Welcome back to mechanical engineering made simple.
0:02
It is good to be back. You know, we spend our entire
0:05
lives surrounded by machines. We drive them, we fly in them.
0:09
We use them to preserve our food and keep our homes livable.
0:13
We are constantly interacting with these devices, pressing
0:15
buttons, turning keys, trusting them with our lives.
0:19
But most of us only ever see the surface, the steel, the plastic,
0:24
the user interface. Right, the shell.
0:26
Exactly. There is a hidden narrative
0:28
underneath all that hardware. I like to call it the invisible
0:31
machine the. Invisible machine.
0:33
I like that you were referring to the energy itself.
0:35
Aren't you? The the fluid reality that
0:37
actually does the work. Precisely.
0:39
We see the piston move or the turbine spin, but we rarely stop
0:43
to think about the journey of the energy driving it.
0:46
It's like watching a puppet show and forgetting about the
0:48
strings. Or, you know, the hands pulling
0:51
the strings. That's a great analogy.
0:53
So today's deep dive is going to be a little different.
0:55
We are going on a walkthrough story.
0:56
We aren't just going to read definitions from a textbook or,
0:59
you know, or site formulas in a vacuum, I hope.
1:01
Not. We are going to follow a single
1:03
unit of energy, a spark if you will, from the moment it is
1:07
locked inside a chemical bond, through the violent explosion of
1:10
combustion, into the mechanical struggle of an engine, and
1:13
finally to the ultimate challenge of getting rid of the
1:17
waste heat. It is a journey from order to
1:19
disorder. Effectively, it is the story of
1:22
the universe in microcosm. It really is.
1:24
And to do this properly, to really do justice to this
1:27
narrative, we have pulled together a stack of heavy
1:31
hitting sources. We have got detailed analysis on
1:34
the thermodynamics of combustion, the rigorous laws of
1:37
equilibrium, engineering cycle analysis from gas turbines to
1:42
steam, and the modern, somewhat philosophical engineering
1:45
concept of exergy. And we're even going to shrink
1:48
down to the micro scale to look at how we cool computer chips.
1:52
So imagine the backdrop for today isn't a recording studio.
1:54
I want you to close your eyes. Imagine we are in a high tech
1:58
engineering lab. On one side you have a massive
2:01
combustion chamber roaring away containing A controlled
2:04
explosion. In the middle, a steam turbine
2:07
spinning at thousands of RPM, and on the far bench, a
2:11
microscope looking at a silicon chip.
2:13
A fitting setting because the physics governing that massive
2:16
turbine are surprisingly similar to the physics governing that
2:19
tiny chip. Really.
2:21
How so? It all comes down to how heat
2:23
moves and how we can or cannot control it.
2:26
Whether you are generating gigawatts for a city or trying
2:29
to keep a processor from melting while playing a video game, the
2:32
rules are the same. So let's light the fuse.
2:35
We are starting with our first section, the spark.
2:37
We have all seen FI, it's elemental.
2:39
But a common question online, one that pops up constantly on
2:42
forearms, is, is fire just hot air or what actually happens to
2:46
the fuel? When we say we are burning gas
2:48
in a car, what is physically going on at the molecular level?
2:51
At a fundamental level, combustion is an oxidation
2:54
process. It is almost always exothermic,
2:56
meaning it releases heat. Right X out.
2:58
It pushes heat out. Exactly.
3:00
But to understand it you have to look at the fuel.
3:03
You have chemical energy stored in the bonds of a fuel, usually
3:06
a hydrocarbon like petrol, diesel or methane.
3:09
Picture these molecules. They are chains of carbon and
3:12
hydrogen, holding hands very tightly.
3:14
They are stable, happy and just full of potential energy.
3:18
So they are. Like a coiled spring just
3:20
waiting. Exactly.
3:22
A perfectly coiled spring and then you introduce an oxidant,
3:25
typically the oxygen in the air. Oxygen is aggressive, wants
3:28
those electrons, but the fuel bonds are strong.
3:31
So you need a catalyst, a kick start.
3:33
That's your spark plug or your compression heat.
3:35
When you introduce that ignition source, you are essentially
3:38
forcing those hands to let go. You are shattering the existing
3:42
structure. You're breaking the bonds apart.
3:43
It's a violent divorce. Precisely.
3:45
You're breaking the bonds of the fuel and the oxygen.
3:48
But here is the magic trick of the universe.
3:50
You then form new bonds in the products.
3:53
The carbon grabs the oxygen to form carbon dioxide.
3:57
The hydrogen grabs the oxygen to form water vapor.
4:00
The key insight the reason fire is hot is that the energy
4:03
required to break the old bonds is less than the energy released
4:08
when the new bonds form. OK, I see.
4:10
So it takes a little energy to break them apart, but you get a
4:13
huge payout when they snap back together into new things.
4:16
Correct. That surplus energy, that is the
4:18
heat you feel, that is the kinetic energy of the molecules
4:23
flying apart, that is the explosion.
4:26
So it's a trade. We trade complex hydrocarbon
4:29
molecules for simpler CO2 and water, and we get a payout in
4:33
thermal energy. That's the transaction.
4:35
But is there a perfect way to do this?
4:37
I see the term stoichiometry come up a lot in the text.
4:39
It sounds like a Greek tragedy. It does, doesn't it?
4:41
Stoichiometry is essentially the perfect recipe.
4:44
Imagine you are baking a cake. You need exactly the right
4:47
amount of flour for the amount of eggs you have.
4:49
If you have too much flour, it's dry.
4:51
Too many eggs, it's a mess. In combustion, the
4:54
stoichiometric condition is when you have exactly enough oxygen
4:58
molecules to react with every single fuel molecule with
5:02
absolutely nothing leftover. No wasted fuel, no wasted air.
5:05
In theory, yes, an equivalence ratio of 1.
5:08
But here is where it gets really interesting, and where the
5:11
sources point out some real world complications that most
5:14
people don't think about. You might think, OK, if I have
5:17
the perfect mixture, I'll get the maximum possible
5:19
temperature. That makes intuitive sense if
5:22
you aren't heating up extra air that isn't participating in the
5:25
reaction, like if you run lean with too much air, shouldn't the
5:29
flame be hottest at that perfect ratio?
5:32
You would think so. It's logical, but reality has a
5:35
speed limit. It's called dissociation.
5:37
Dissociation. When the temperature gets high
5:40
enough, we are talking thousands of degrees Kelvin.
5:44
The products, the carbon dioxide in the water vapor, they don't
5:47
just stay stable. The environment is too chaotic.
5:49
They start to split back apart. Where they unburn.
5:52
In a sense, yes. The violence of the heat is so
5:54
intense that the carbon dioxide splits back into carbon monoxide
5:58
and oxygen. The water splits into hydroxide
6:01
and hydrogen. Now remember what we said
6:03
earlier. Forming bonds releases energy,
6:08
but breaking bonds. Breaking bonds absorbs energy.
6:11
It costs energy. Exactly.
6:13
It is endothermic, so this dissociation process acts like a
6:17
chemical break on the temperature rise.
6:19
It sucks up some of that thermal energy you just released to RIP
6:22
the molecules apart again. So even with a perfect mixture,
6:26
you hit a ceiling, the fire gets so hot it defeats.
6:29
That's it, You've nailed it. That is fascinating.
6:31
It's like a self limiting system.
6:33
Nature says OK, that's hot enough.
6:34
Precisely. And when we try to calculate
6:37
these temperatures in the lab, we have to look at whether the
6:38
process is adiabatic. Adiabatic.
6:41
That's one of those $5 engineering words.
6:43
Let's unpack that for the listener.
6:45
It just describes a process where no heat is lost to the
6:47
surroundings. Imagine a combustion chamber
6:50
wrapped in a perfect magical blanket.
6:52
A perfect insulator. Right.
6:54
In a theoretical adiabatic combustion chamber, all that
6:58
chemical energy stays in the gas, raising its temperature to
7:01
the absolute maximum. But in a real engine, the walls
7:05
are made of metal. They're cold.
7:06
There's coolant flowing through channels in the block.
7:09
Heat is escaping constantly, so the actual peak temperature is
7:13
always lower than the adiabatic flame temperature.
7:16
It's the difference between the theoretical maximum speed of a
7:18
car in a vacuum versus driving it on a road with wind
7:22
resistance and friction. That is a great analogy, a
7:25
perfect one. And Speaking of the gas itself,
7:28
I noticed in the math sections of these sources there are a lot
7:31
of references to two variables C sub B&C, sub P.
7:35
Right, specific heats. I always found this bit
7:38
confusing. Why do we need two different
7:40
variables for specific heat? Isn't heat just heat?
7:43
You would hope so, but gases are tricky.
7:45
Specific heat is a measure of how much energy is needed to
7:48
raise the temperature of a substance by 1°.
7:51
But a gas behaves differently depending on its container.
7:54
C Sub V is specific heat at constant volume.
7:59
Think of a rigid steel box. You heat the gas, the pressure
8:02
rises, but the box doesn't move. All the energy goes into
8:05
temperature. OK, so sealed container and C
8:09
sub. VC sub P is specific heat at
8:12
constant pressure. Think of a cylinder with a
8:14
piston that can slide up and down like in an engine.
8:18
As you heat the gas, it expands. It pushes the piston up.
8:22
So the gas is doing work while it's heating up.
8:23
Exactly. And because it's doing work
8:25
pushing that piston, it is using up some of the energy you're
8:28
heating it. So to get the temperature to
8:30
rise by that same 1° you have to put more energy in.
8:33
Oh I see. You have to feed the temperature
8:35
rise and the work output. That is why C sub P is always
8:38
higher than C sub P. That makes perfect sense. 1 is
8:40
just heating, the other is heating plus lifting weights.
8:43
A very good way to put it. And the ratio of these two,
8:46
often denoted as K or gamma, is vital for calculating engine
8:50
efficiency. It tells us how springy the gas
8:53
is. So we have had our explosion.
8:56
We have generated this massive spike in temperature and
8:58
pressure. The molecules are dissociated,
9:01
then reassociated, vibrating like crazy.
9:04
Now we have to play by the rules.
9:05
The unbreakable rules. Section 2.
9:08
The rules of the road. I'm talking about the laws of
9:10
thermodynamics. The laws that govern the
9:12
universe, or at least the laws that govern how we interact with
9:16
energy. These are the laws that tell us
9:18
what is possible and what is a fantasy.
9:19
And the ones that crush the dreams of every inventor trying
9:23
to build a perpetual motion machine in their garage.
9:26
Let's start with the first law. The first law is the law of
9:29
conservation. It is the accountant of the
9:31
universe, the auditor. It simply states that energy
9:34
cannot be created or destroyed, only change from one form to
9:38
another. Right?
9:39
If you look at a closed system, the change in the internal
9:42
energy denoted as E2 minus E1 must equal the net heat added to
9:46
the system minus the work done by the system.
9:49
So energy change equals heat minus work E 2 minus E 1 = Q -,
9:54
W You can't get something for nothing.
9:56
Precisely, if you want workout, you have to put energy in,
10:00
either as heat or potential energy.
10:03
The bank account must balance. There are no overdrafts in
10:07
physics, none at all. If you see a machine that claims
10:10
to output more energy than it takes in, the first law says
10:13
look for the hidden battery. Right.
10:14
Find the cord plugged into the wall.
10:17
But if the first law is the accountant checking the books,
10:20
the second law feels like the tax man.
10:22
The tax man, or perhaps the traffic cop.
10:24
I like that the first law says you can convert energy.
10:27
It doesn't care about the direction.
10:29
It thinks turning smoke back into a log is just as valid as
10:32
burning a log into smoke, as long as the energy matches.
10:35
The second law tells you the direction that conversion can
10:37
happen. It introduces the concept of
10:40
entropy. Entropy, the measure of
10:42
disorder, but it's more than just a messy room, isn't.
10:45
It it is, yeah. It's a very deep concept.
10:47
Ideally, think of it as unavailability.
10:50
The second law dictates that heat naturally flows from a hot
10:53
body to a cold body, never the other way around unless you
10:56
force it with external work. Like a refrigerator.
10:59
Exactly. And the seemingly simple
11:01
observation that coffee gets cold, it doesn't spontaneously
11:04
boil has massive implications for engineering.
11:07
It means that whenever we try to convert heat into work, we are
11:10
fighting a losing battle against disorder.
11:13
This is where we get into the concept of irreversibility.
11:15
Right. Correct.
11:16
Real processes, friction, turbulence, mixing, heat
11:19
transfer across a temperature gap generate entropy.
11:23
We call this S subgen or entropy generation S Gen.
11:26
Every time you generate entropy, you are permanently destroying
11:29
the potential to do useful work. You are taking organized energy
11:33
and making it disorganized. You are taking a stack of sorted
11:36
playing cards and shuffling them.
11:38
You still have 52 cards. That's the first law.
11:40
Right. Conserved.
11:41
But you can't play the game anymore.
11:43
That's the second law. That is a depressing but
11:46
effective analogy, and this leads us to the ultimate limit,
11:49
the speed limit of the universe, the Carnot limit.
11:54
Saudi Carnot, a French engineer in the 1820s, he was thinking
11:58
about steam engines before we even really understood what heat
12:00
was. That's amazing.
12:02
It is, he realized that the efficiency of a heat engine, any
12:05
engine that turns heat into work, depends entirely on the
12:08
temperatures that operates between.
12:10
No engine can be more efficient than a reversible 1 operating
12:14
between a high temperature reservoir T sub H and a low
12:17
temperature reservoir T sub L. So TH and TL, the hot side and
12:21
the cold side. And the formula for that limit
12:24
is surprisingly simple, considering how powerful it is.
12:27
It is elegant. The maximum theoretical
12:30
efficiency is 1 minus the ratio of the low temperature to the
12:34
high temperature. So efficiency equals 1 -, T sub
12:37
L / T sub H. So to get better efficiency, if
12:40
you look at that math, you really only have two knobs to
12:42
turn. You either make the hot side
12:44
hotter or the cold side colder. Exactly.
12:47
Those are your only options. And think about the cold side T
12:49
sub L. That is usually the environment.
12:51
The air outside your are the river water cooling of power
12:54
plant, right? That is around 300 Kelvin, maybe
12:58
27°C. We can't lower the temperature
13:01
of the planet just to make our engines better.
13:02
We are stuck with T sub L. So the only way is up.
13:05
The only way is up. That is why engineers are
13:08
obsessed with pushing T sub H the combustion temperature
13:11
higher and higher. The hotter we burn, the more
13:13
efficient we can theoretically be.
13:15
That is why we use exotic super alloys in jet engines so they
13:19
don't melt while we chase that high Carnot efficiency.
13:22
OK, so we have the fuel and we have the rules.
13:25
Now let's build the machine. Section 3 turning heat into
13:28
motion. We are talking about engineering
13:31
cycles. This is where the rubber meets
13:33
the road. We use what we call a heat
13:35
engine. As we established, this is a
13:37
device that takes thermal energy, converts some of it to
13:41
work, and this is crucial, rejects the rest to a sink.
13:44
Always. You cannot convert 100% of heat
13:47
into work. The second law forbids it.
13:49
You always have to pay the tax to the cold reservoir.
13:51
Let's walk through the big ones, the workhorses of the modern
13:54
world. Yeah, first step, the steam
13:57
plant, the Rankine cycle. This is what keeps the lights on
14:00
in most of the world. The Rankine cycle is the
14:02
backbone of global electricity. It uses water as the working
14:06
fluid. Let's trace the path.
14:08
You have a boiler that turns liquid water into high pressure
14:11
steam. That steam expands through a
14:13
turbine. As it expands, it hits the
14:16
blades, spinning them. That's your work output.
14:19
Then the steam goes into a condenser where it turns back
14:21
into liquid water, rejecting heat to the environment.
14:24
Finally, a pump sends it back to the boiler to start again.
14:27
Why water? Why not air or mercury?
14:30
Water is magical, but for the Rankine cycle, the secret sauce
14:33
is the phase change. When steam condenses back to
14:36
water, its volume collapses. Liquid water is almost
14:40
incompressible. Right, you can't really squeeze
14:42
it. You can't.
14:43
It takes very little energy to pump liquid water up to high
14:46
pressure. If you tried to compress gas it
14:48
would take a huge amount of energy, but pumping water is
14:50
cheap. In the ranking cycle, the back
14:53
work ratio, the energy cost to run the pump, is tiny, maybe 1
14:56
or 2% of the total output. So you get almost all the
14:59
turbine work is profit. Exactly.
15:01
It's an incredible thermodynamic bargain now.
15:03
Compare that to the engine hanging off the wing of an
15:05
airliner, the gas turbine, the Brayton or Joule cycle.
15:09
Very different beast here. The working fluid is air, and it
15:13
stays a gas the whole time. It never becomes a liquid.
15:16
You suck air in the front, compress it, that's the
15:18
compressor. Then you spray fuel into it and
15:21
burn it. That's the combustor.
15:23
Then that hot high pressure gas roars through the turbine.
15:26
And unlike the steam plant, the compressor here isn't cheap to
15:29
run. Not at all.
15:30
Compressing gas is hard work. Imagine trying to squeeze a
15:33
balloon versus pumping water in a gas turbine.
15:37
The compressor is a glutton. A glutton.
15:39
I like that. It consumes a huge chunk of the
15:42
power generated by the turbine, sometimes up to half or even
15:45
more. The turbine has to drive the
15:47
compressor just to keep the cycle going.
15:49
The rest is what we use for thrust or shaft power.
15:52
That's amazing, half the engines power is just used to keep the
15:55
engine breathing. It is the cost of doing business
15:57
with a gas. It's the price of admission.
16:00
The sources mentioned pressure ratios being a big deal here.
16:03
Huge modern gas turbines operate at pressure ratios of 20 to 1:30
16:07
to 1:00 or even higher. That means the air is squeezed
16:10
to 30 times atmospheric pressure before combustion.
16:14
High pressure ratios give you high power density, a lot of
16:17
power in a small light package. That is why they are perfect for
16:20
aircraft where weight is everything.
16:22
A steam plant is efficient, but it is heavy.
16:25
You can't fly a steam plant. Now here is where it gets really
16:28
cool. The aha moment in power
16:31
generation. The combined cycle.
16:33
This feels like The Avengers team up of engineering.
16:36
It really is. This is the smartest trick in
16:38
the book. You take a gas turbine, which
16:40
runs very hot. Exhaust temperatures can be
16:43
600°C or more. OK, usually that heat would just
16:46
be wasted, blown out the stack. But 600° is plenty hot enough to
16:50
boil water. So in a combined cycle, you use
16:53
that hot exhaust to boil water for a steam turbine loop.
16:57
So the waste of the first engine becomes the fuel for the second
16:59
engine. Exactly.
17:00
You are stacking the cycles. You get the high T sub H
17:03
efficiency of the gas turbine and the low temperature heat
17:06
rejection of the steam cycle. This widens the gap between your
17:09
high and low temperature is pushing the total efficiency way
17:12
up, sometimes over 60%, which is unheard of for a single engine.
17:16
Incredible. And briefly, for everyone
17:19
driving to work today, the internal combustion engine, the
17:24
auto cycle. The auto cycle is what happens
17:27
in your petrol car. Intake, compression, Combustion,
17:30
exhaust. The key distinction here is that
17:33
combustion happens ideally at constant volume.
17:36
What does that mean? Constant volume.
17:38
The piston is at the top of his stroke.
17:40
It stopped for an instant. The smart fires and boom,
17:43
pressure spikes instantly before the piston has a chance to move
17:46
down. It is a violent kick.
17:48
And the efficiency here depends on how much you squeeze the air,
17:51
the compression ratio. Right, the more you compress the
17:54
air before igniting it, the more work you get out.
17:56
But if you compress it too much, the fuel might self ignite
17:59
before the spark due to the heat of compression.
18:01
That's called knocking. The pinging sound.
18:03
That's the one. So there is a limit based on the
18:05
fuel's octane rating. OK, so we are generating power,
18:08
we are spinning shafts, but let's pause and look at the
18:11
quality of what we are doing. Section 4.
18:15
It's not just quantity, it's quality.
18:18
We need to talk about exergy. Exergy is one of the most
18:21
important yet least understood concepts.
18:24
In engineering. We usually talk about energy
18:27
conservation. Energy is never lost, right?
18:30
But if energy is conserved, why do we have an energy crisis?
18:33
Why can't we just reuse the energy from the exhaust pipe
18:36
over and over again? That's the question.
18:38
We didn't destroy the energy, but we destroyed something else.
18:41
We destroyed the quality. Exergy, sometimes called
18:43
availability, is a measure of the maximum useful work that can
18:47
be done by a system as it comes into equilibrium with its
18:50
environment. We compare everything to the
18:53
dead state. The dead state sounds ominous,
18:56
like a post apocalyptic wasteland.
18:58
It's just the environment pressure of 1 bar temperature of
19:01
25°C, usually denoted as T sub zero.
19:04
Once your system reaches that state, it has 0 exergy.
19:07
So explain that. Think about it.
19:09
A bathtub full of lukewarm water might contain a lot of thermal
19:13
energy joules, but you can't run an engine off it because it's
19:16
the same temperature as the air. It has energy but zero exergy.
19:20
It has no potential to do work relative to its surroundings.
19:24
So while energy is conserved, exergy is destroyed.
19:27
Yes, that is the tragedy of thermodynamics.
19:31
Every time you have friction or mix 2 fluids or, and this is the
19:35
big one, transfer heat across a temperature difference, you
19:37
destroy exergy. OK, If you take high temperature
19:40
steam and just let it cool down to warm water without running a
19:43
turbine, you have destroyed that potential.
19:45
You are taking high quality energy and degrading it into low
19:49
quality heat. The sources bring up a theorem
19:51
here that links this back to entropy, the ghost total
19:55
theorem. This is a beautiful piece of
19:57
physics. It puts a price tag on our
19:59
mistakes. It states that the lost work,
20:01
the work you could have had but didn't get, is directly
20:04
proportional to the entropy generation.
20:06
The formula is W sub loss equals T sub 0 * s subgen.
20:12
Work loss equals T Sub zero times is subgen, so the
20:15
temperature of the environment times the mess you made equals
20:18
the opportunity you wasted. Exactly.
20:20
This gives engineers a target. We call it entropy Generation
20:22
minimization, or EGM. Modern design isn't just about
20:26
efficiency in the old sense of burning less fuel.
20:29
It is about identifying exactly where entropy is being generated
20:33
in a component, say a heat exchanger tube, and redesigning
20:36
it to minimize that S subgen. Can you give us a concrete
20:40
example of that trade off? Because usually we think
20:43
friction is bad, heat transfer is good.
20:45
Sure. Think of a heat exchanger tube.
20:47
You want to transfer heat from the fluid inside to the fluid
20:50
outside. To do that, well, you want the
20:52
flow to be turbulent. Turbulence mixes the fluid and
20:55
grabs the heat right? So you might make the tube very
20:57
long or narrow to create that turbulence.
21:00
But turbulence creates friction. Friction causes a pressure drop,
21:04
which means your pump has to work harder.
21:05
So you have a conflict. The thing that helps heat
21:08
transfer also hurts your pump. Exactly, and both generate
21:11
entropy. Friction generates entropy.
21:13
Bad but poor heat transfer also generates entropy because of the
21:16
large temperature difference. Also bad.
21:19
EGM helps you find the optimal diameter where the sum of those
21:21
two evils is minimized. You are balancing the entropy of
21:25
friction against the entropy of heat transfer.
21:27
That is incredibly nuanced. It's not just make it bigger or
21:30
make it smaller, it's find the balance point.
21:33
It is optimization in its purest.
21:35
Form. So we have extracted the work.
21:37
We have minimized our losses using extra gene analysis, but
21:40
we still have heat leftover. And in some cases, like
21:43
electronics, heat isn't a resource we are trying to catch.
21:47
It's the enemy. It's the absolute enemy.
21:49
Section 5. Dealing with the heat.
21:51
This is a huge field. You have a computer chip the
21:54
size of a fingernail generating as much heat flux.
21:57
That's watts per square meter. As a nuclear reactor core, it is
22:01
intensely concentrated. If you don't get that heat out,
22:05
the silicon junction temperature rises, the transistors leak,
22:08
current timing errors happen, and eventually the chip
22:11
physically melts or cracks. I love the analogy.
22:14
The source is used here to explain how we move this heat.
22:16
They compare heat flow to electricity.
22:18
Yes, the thermal resistance network.
22:20
It is exactly like Ohm's law. In electricity, voltage equals
22:23
current times resistance V = I * R In heat transfer we have an
22:28
equivalent The temperature difference delta T equals the
22:32
heat flow rate Q times the thermal resistance R.
22:35
So if you want to keep the chip temperature low, which is like
22:38
keeping the voltage low, but you have a high heat flow, high
22:41
current, you have to lower the thermal resistance.
22:44
Exactly. You want the path of least
22:45
resistance for that heat to escape, and we have 3 modes to
22:49
do that, conduction, convection and radiation.
22:52
Let's break them down. Conduction.
22:54
This is the heat moving through solids.
22:56
That's Fourier's law, heat moving from the silicon dye
23:00
through the thermal paste into the copper base of the heat
23:02
sink. The resistance here depends on
23:04
the thickness of the material and its thermal conductivity.
23:07
K Copper has a high K, which is why we use it.
23:11
Plastic has a low K, which is why it's an insulator.
23:14
You want the heat to Sprint through the solid material.
23:17
Then convection. That's the air moving over the
23:19
heat sink fence. Right, and here we get into the
23:21
dimensionless numbers that engineers love.
23:24
We need to know how effective the air is at grabbing the heat.
23:27
We use the Nusselt number. The Nusselt number.
23:30
The Nusselt number tells us how good the convection is compared
23:33
to just stagnant conduction. A high Nusselt number means the
23:37
air is scrubbing that heat away effectively.
23:39
And to get a high Nusselt number you need flow.
23:41
Yes, and that brings us to the Reynolds number.
23:44
This tells us if the airflow is smooth, laminar, or chaotic
23:49
turbulent. Turbulent flow is much better
23:51
for cooling because it mixes the air, bringing fresh, cool air
23:55
into contact with the fins. But it requires louder, stronger
23:59
fans. And the sources mentioned
24:01
natural convection versus forced convection.
24:04
Forced is using a fan, natural is.
24:07
Doing nothing. Natural convection is passive.
24:10
It is what happens when your computer is silent.
24:12
The air near the fins gets hot, expands, becomes less dense and
24:15
rises. Oh, buoyancy.
24:17
Buoyancy drives the flow. It acts like a weak pump.
24:20
We use the Rayleigh number to describe this flow strength.
24:22
And there's another number for the fins themselves, right?
24:24
The L embossed number. Yes, the L embossed number is
24:27
specific for those vertical fins you see on heat sinks.
24:31
It helped answer the question how close can I pack these fins?
24:34
Because you would think more fins equals more surface area
24:38
equals better cooling. That is, the track air is
24:42
viscous. It's sticky.
24:44
If the fins are too close together, the boundary layers of
24:46
the air, the slow moving air clinging to the surface, merge
24:50
from both sides. They block each other.
24:53
You effectively choke off the flow.
24:55
The air can't get in to grab the heat.
24:58
It's like trying to breathe through a straw that's too thin.
25:00
Exactly. But if they are too far apart,
25:02
you are wasting valuable space where you could have put more
25:05
surface area. The Alenbus number helps you
25:08
find the sweet spot for spacing where you maximize surface area
25:12
without choking the flow. And what about radiation?
25:14
I usually think of that as something the sun does.
25:17
Does my computer really radiate heat?
25:19
In electronics cooling it is often ignored, but it shouldn't
25:22
be. Every surface radiates heat
25:24
based on its temperature and emissivity.
25:26
How black it is in the infrared spectrum?
25:28
It depends on the Stefan Boltzmann constant.
25:32
For a black anodized heat sink in a computer case, radiation
25:36
can account for a significant chunk of the cooling, maybe 10
25:39
to 15%. Especially if the fan fails or
25:42
runs slow. It is the cooling of last
25:44
resort. There was one specific detail in
25:47
the text that I found really interesting.
25:49
Spreading resistance. Yes, this is a geometric
25:53
problem. Imagine you have a tiny chip,
25:57
say one square centimeter, producing 100 watts.
26:00
You clamp a big heat sink on top of it that is 10 centimeters
26:03
wide. OK, you might assume the whole
26:05
heat sink is working, but the heat has to flow from that tiny
26:08
contact patch and spread out to fill the base of the big heat
26:11
sink. So the corners of the heat sink
26:13
might be cool because the heat hasn't reached them yet.
26:15
Exactly. That spreading action creates A
26:17
bottleneck. It's like a crowd of people
26:19
trying to get out of a stadium through 1 exit door.
26:22
The resistance is high at the door.
26:24
So how do you fix it? You need a thick base.
26:26
The math shows that the heat spider needs to be thick enough
26:29
to allow that flow to open up to diverge before it hits the fins.
26:34
A rule of thumb from the text is that the spreader needs to be 3
26:37
to five times thicker than the square root of the heat source
26:40
area to be effective. If it's too thin, the heat just
26:43
shoots straight through and ignores the outer edges of the
26:46
sink. You are carrying dead weight.
26:48
And then there is the interface itself, contact resistance.
26:52
This is microscopic. You put a heat sink on a chip
26:55
and to your eye it looks like they are touching perfectly flat
26:59
on flat. But under a microscope the
27:01
surfaces look like mountain ranges.
27:04
They are rough. They only touch at the peaks,
27:07
the asperities. So most of the gap is actually
27:10
just air. Yes, and air is a terrible
27:13
conductor of heat. It's thermal conductivity is
27:15
tiny. So you have these tiny bridges
27:17
of metal where heat flows well, and huge canyons of air where
27:20
heat flows poorly. And that's where the thermal
27:22
paste comes in. That is why we use thermal
27:24
paste, the Gray goop. The Gray goop.
27:26
It fills in the valleys between the asperities, replacing the
27:29
air with a conductive material. It maximizes the contact area.
27:33
Without that paste, your expensive cooler is practically
27:36
useless because the heat can't bridge the gap.
27:38
OK, we have heated things up, we have cooled them down, but what
27:42
if we want to get really cold? Section 6.
27:46
The cold frontier. Refrigeration and cryogenics.
27:50
Now we are flipping the script. A heat engine takes heat and
27:54
makes work. A refrigerator takes work and
27:57
moves heat from a cold place to a hot place uphill.
27:59
It is pushing heat uphill against the natural flow
28:03
dictated by the second law. Correct.
28:05
We are fighting the tax man and because of that we don't talk
28:08
about efficiency. Efficiency is always less than
28:10
one. But a fridge can move more
28:12
energy than it consumes, right? So we talked about the
28:14
coefficient of performance or COB.
28:17
And the cut can be greater than one, right?
28:19
Which sounds like magic. I put in one Watt and I get 3
28:22
watts of cooling. It does.
28:24
It does sound like magic, but remember you aren't creating the
28:26
cold, you are moving the heat. If a heat pump has a COP of
28:30
three, it means for every one Joule of electrical work you put
28:33
in, you move three joules of heat out of the fridge.
28:35
You are a heat mover, not a heat maker.
28:37
But when we get down to cryogenics, liquefying gases
28:40
like nitrogen or helium, it gets harder.
28:43
The COP drops. Much harder the carnal limit
28:47
bites you here. Remember the efficiency Formula
28:50
One minus TL over TH? Well, for a refrigerator, the
28:55
work required is proportional to the difference between the
28:57
temperatures divided by the cold temperature squared.
28:59
OK. That sounds bad.
29:00
It is. As T sub L gets closer to
29:03
absolute 0, the work required to remove even a tiny amount of
29:07
heat becomes enormous. The price of cooling skyrockets.
29:12
The text mentions the Sterling cycle in this context.
29:14
The Sterling cycle is fascinating.
29:16
It uses a regenerator, a thermal sponge.
29:18
Basically it's a wire mesh inside the engine.
29:20
A sponge. You compress the gas, it gets
29:23
hot, the regenerator absorbs that heat.
29:26
You expand the gas, it gets very cold, and on the way back, the
29:29
regenerator gives the heat back. It recycles the thermal energy
29:32
internally. It's critical for reaching those
29:35
cryogenic temperatures efficiently without needing
29:37
massive compressors. And the design of heat
29:39
exchangers becomes life or death here.
29:41
Absolutely. In a power plant, if your heat
29:44
exchanger is a bit inefficient, maybe you have a five degree
29:46
difference instead of two degrees.
29:48
You lose a percent of fuel efficiency.
29:49
No big deal in cryogenics, because the temperature
29:52
difference delta T represents such a huge loss of exergy.
29:55
At low temperatures, a slightly inefficient heat exchanger can
29:59
completely stop the machine from working.
30:02
If the incoming gas isn't precooled perfectly by the
30:05
outgoing gas, the cycle fails. The exergy loss is magnified at
30:09
low temperatures. We have covered a lot of ground
30:12
equilibrium steady states cycles, but the world isn't
30:16
always stable. Section 7.
30:19
Non equilibrium. This is the frontier of
30:21
thermodynamics. Real life rarely sits in
30:24
equilibrium. Things are flowing, changing.
30:27
The sources mention Onsager's reciprocal relations.
30:29
Reciprocal relations. That sounds complex.
30:31
What does that mean in plain English?
30:33
It means that different forces are coupled.
30:34
We usually think in silos. Temperature difference causes
30:37
heat flow. Voltage difference causes
30:39
current. Right.
30:40
Cause and effect. But Onsager showed that these
30:42
forces can crossover. Crossover.
30:44
Yes, a temperature difference can cause mass to move
30:48
diffusion. That's the sorid effect.
30:50
If you have a mixture of gases and you heat one side, the heavy
30:54
molecules might migrate to the cold side.
30:57
Or a voltage difference can cause a temperature difference,
30:59
the thermoelectric effect or Peltier effect.
31:02
So everything is connected. You push on nature in one way
31:05
and it reacts in multiple ways. Exactly.
31:08
It shows the deep interconnectedness of physical
31:10
phenomena. You cannot treat heat,
31:13
electricity and mass flow as completely separate silos.
31:16
They are all just fluxes of energy and entropy seeking
31:19
balance. So let's bring this all home.
31:22
We started with a chemical bond snapping in a hydrocarbon
31:25
molecule. We released that energy,
31:27
creating a high temperature gas. We fought the dissociation that
31:31
tried to limit our temperature. We understood the violence of
31:33
the molecular divorce. You ran that gas through a
31:35
cycle, maybe a gas turbine, maybe a piston engine, always
31:39
obeying the first law account. Seeing the second law taxman, we
31:42
realized we can't get something for nothing and we can't break
31:45
even. We hit the carno limit,
31:47
extracting as much work as the temperature reservoirs allowed.
31:50
We analyzed our losses using exergy to see where we were
31:53
destroying usefulness. We optimized our tubes and
31:56
flows. And finally, we managed the
31:58
waste heat using fins, fans and thermal paste to coax that
32:02
entropy out into the environment so the machine doesn't melt.
32:07
We navigated the resistance network.
32:08
It is a continuous, elegant struggle against the arrow of
32:11
time. And that leads us to our final
32:13
thought. The sources hint at the design
32:16
philosophy for the future. Yes.
32:19
For a long time, engineering was about energy efficiency, just
32:22
saving fuel. But the future is entropy
32:25
generation minimization. It's a subtle shift.
32:29
We need to design systems from power plants to laptops that
32:32
fight the creation of disorder. We want to guide energy through
32:35
our machines as gently as possible, preserving its quality
32:38
for as long as possible. The engineer of the future isn't
32:41
just an energy manager, they are an entropy architect.
32:44
An entropy architect. I love that it turns engineering
32:47
into an art form. Well, that wraps up our walk
32:49
through of the Engine of the Universe.
32:51
It's a pleasure to walk through it with you.
32:52
If you enjoyed this deep dive and you want to see some of
32:55
these principles in action, maybe watching us try and fail
32:59
to build a Stirling engine or visualizing a heat sink in a
33:02
wind tunnel. If you haven't already, get over
33:05
to our YouTube channel Mechanical Engineering Made
33:07
Simple where we post hands on how to videos, visual
33:10
breakdowns, and practical engineering walkthroughs You can
33:13
actually use. A family that solves together
33:15
stays together.