Latest / Mechanical Engineering Made Simple / (#150) PV -Engineering and Fabrication Realities

Transcript

(#150) PV -Engineering and Fabrication Realities · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design junkies. This episode of Mechanical
• 0:02
  Engineering made simle comes to you by listener request.
• 0:04
  If you didn't know, we do those O If there's a topic you want us
• 0:07
  to breakdown, explain, or explore, just drop us a comment.
• 0:11
  We rioritize getting the content you want in front of you.
• 0:14
  O Let's dive into today's requested topic.
• 0:17
  Theory prepares you for the test.
• 0:19
  The machine shop prepares you for the moment your calculation
• 0:21
  blows up in your face. Yeah, it really is the ultimate
• 0:25
  reality check. I mean, we spend years in
• 0:27
  academia learning how to, you know, balance equations and run
• 0:31
  finite element analysis and usually we do that in a quiet
• 0:34
  temperature controlled office. Right, where everything is
• 0:37
  perfectly predictable. Exactly.
• 0:39
  The math just kind of assumes a perfect universe.
• 0:41
  It assumes the steel is this flawless, homogeneous isotropic
• 0:45
  material, and it assumes the Weld is just, well, a perfect
• 0:49
  continuation of that steel. Which I'm guessing is not how
• 0:52
  the shop floor actually works. Oh, not even close.
• 0:55
  The shop floor doesn't care about your perfect universe.
• 0:58
  On the floor, heavy steel plate actually springs back when you
• 1:01
  try to roll it and the heat of an arc Weld it fundamentally
• 1:05
  destroys and alters the metallurgical grain structure of
• 1:07
  the base metal. Plus, human beings make
• 1:09
  mistakes. Right.
• 1:10
  Yeah, we are taking the pristine world of CAD models and just
• 1:15
  crashing it head first into the grit, the heat and really the
• 1:20
  terrifying realities of fabricating high pressure
• 1:22
  containment vessels. And the forces we're trying to
• 1:24
  contain here, I mean, they are really difficult to overstate.
• 1:28
  We are talking about vessels holding thousands of gallons of
• 1:31
  boiling, volatile, highly corrosive fluids and in hundreds
• 1:35
  or even thousands of PSI. Yeah, a failure isn't just a
• 1:38
  leak. It's a catastrophic release of
• 1:40
  stored energy. I mean, it's shrapnel.
• 1:42
  It's lethal. So for everyone out there
• 1:44
  actually designing these things, cutting the plate or striking
• 1:46
  the arc, this deep diet is really about bridging that gap.
• 1:50
  How does the choice made in a spreadsheet actually dictate the
• 1:53
  failure mode on the factory floor?
• 1:55
  And that journey? Well, it begins long before
• 1:58
  anyone picks up a welding torch. It starts right at the foundry
• 2:01
  with the chemical makeup of the metal itself.
• 2:04
  Brett. So let's start with the raw
• 2:05
  materials. The foundation of almost every
• 2:07
  industrial pressure vessel is ferrous metal, right?
• 2:10
  Specifically, carbon steel. Yes, but the term steel is
• 2:15
  thrown around very loosely in casual conversation.
• 2:17
  In the realm of high stress containment, it is a high,
• 2:20
  highly specific, rigidly controlled chemical recipe.
• 2:24
  I mean, for an iron alloy to even be legally and structurally
• 2:28
  classified as steel, the carbon content has to live in this
• 2:32
  really narrow band, doesn't it? It does.
• 2:34
  It must be at least 0.02% by weight, but it absolutely cannot
• 2:39
  exceed 2%. 2% just seems like such a tiny threshold.
• 2:43
  It's kind of hard to imagine that adding 3% carbon instead of
• 2:46
  2% completely changes the fundamental nature of the metal.
• 2:49
  What is actually happening at the atomic level if we cross
• 2:51
  that line? Well, to visualize it you have
• 2:53
  to look at the crystalline structure of the iron iron
• 2:57
  atoms. They form a lattice, basically a
• 2:59
  grid, and carbon atoms are much smaller, so they fit right into
• 3:03
  the empty spaces, exactly the interstices between the iron
• 3:06
  atoms. When you keep the carbon under
• 3:09
  2%, you are adding just enough of these smaller atoms to create
• 3:14
  basically roadblocks in the lattice.
• 3:16
  OK, so the roadblocks get in the way.
• 3:18
  Right. They make it harder for the
• 3:19
  layers of iron atoms to slide past each other, and that
• 3:23
  resistance to sliding is what we actually measure as strength or
• 3:26
  hardness. But the layers can still slide
• 3:28
  if they're pushed hard enough, right?
• 3:30
  Which is why why steel can bend instead of breaking.
• 3:33
  It retains its ductility. Precisely the point ductility is
• 3:36
  the metal's ability to deform plastically under localized
• 3:39
  stress without fracturing. But if you push that carbon
• 3:42
  content above 2% while you saturate the grid.
• 3:46
  So it's not just roadblocks anymore.
• 3:47
  No, you're fundamentally altering the compound into
• 3:50
  something called cementite. The material stops being steel
• 3:53
  and it actually becomes cast iron.
• 3:55
  Oh wow, cast iron, which is incredibly hard, but it's
• 3:59
  brittle, right? Yes, very rigid.
• 4:01
  The layers of atoms can no longer slide at all.
• 4:04
  So when a localized stress hits a piece of cast iron, say a
• 4:08
  vibration or a thermal shock, it can't yield to absorb the
• 4:12
  energy, it just cracks. It shatters like glass.
• 4:14
  Exactly. And for a pressure vessel where
• 4:17
  localized bending stresses are totally unavoidable around
• 4:20
  nozzles and supports, a material that can't yield is material
• 4:24
  that will inevitably explode. Man, so we have this strict
• 4:27
  ceiling of 2% for carbon, but the sources point out that even
• 4:31
  within true carbon steels, carbon isn't the only element
• 4:35
  the designer has to aggressively manage.
• 4:37
  Right. Carbon is the main lever for
• 4:39
  strength, but it brings impurities and side effects.
• 4:42
  For instance, manganese is added to improve the hot workability
• 4:45
  of the steel. It binds with sulfur impurities,
• 4:48
  preventing the metal from tearing while it's being rolled
• 4:50
  hot. But there's a cap on that too,
• 4:51
  right? Yep, you have to cap manganese
• 4:53
  at 1.65%. If you go higher, the steel
• 4:56
  becomes way too susceptible to cracking during the rapid
• 4:58
  cooling of a Weld cycle. OK.
• 5:00
  And what about silicon? I saw that mentioned.
• 5:02
  Silicon is used to deoxidize the steel.
• 5:05
  It literally pulls oxygen out of the metal so you don't end up
• 5:08
  with bubbles or porosity in your solid plate.
• 5:12
  But silicon is capped at 0.60%. Why is that?
• 5:17
  What happens if you add too much silicon?
• 5:18
  Too much of it destroys the surface finish and drastically
• 5:21
  reduces the metals toughness. And then there's copper, which
• 5:25
  sometimes sneaks in from recycled scrap that's also
• 5:28
  capped at 0.60% because it causes a phenomenon called hot
• 5:32
  shortness. Hot shortness?
• 5:34
  What does that look like? The metal basically just
• 5:36
  crumbles under stress at high temperatures.
• 5:38
  It's a delicate balancing act. It really is, but it's
• 5:41
  interesting to look backward for a moment.
• 5:42
  I mean, before we had this level of metallurgical control, people
• 5:46
  use wrought iron. You see it on old Victorian
• 5:49
  bridges or antique steam engine boilers.
• 5:52
  Structurally, what is the difference between this highly
• 5:54
  controlled modern carbon steel and historical wrought iron?
• 6:00
  The defining characteristic of wrought iron isn't just its low
• 6:03
  carbon content, it's the physical inclusion of slag.
• 6:07
  Slag like the waste material? Yeah, slag is a glassy byproduct
• 6:11
  of the smelting process created when limestone reacts with the
• 6:15
  impurities in the iron ore. In modern steelmaking, we skim
• 6:19
  that slag off completely to get a pure homogeneous liquid.
• 6:22
  But they couldn't do that back in the day.
• 6:24
  Right in the 18th and 19th centuries, the process involved
• 6:27
  physically stirring or puddling the molten iron.
• 6:31
  Small stringles of this glassy slag actually got folded into
• 6:35
  the metal. So I'd have a grain to it,
• 6:36
  almost like wood. Exactly like wood, the slag
• 6:39
  inclusions create a fibrous structure.
• 6:42
  If you try to bend wrought iron across the grain, it's fairly
• 6:44
  strong, but if you try to pull it apart parallel to the grain,
• 6:48
  it can split right down the slag lines.
• 6:50
  Which is fine for decorative gates because it's soft and
• 6:53
  forgivable under a blacksmith's hammer.
• 6:55
  Sure, but when you are building a modern reactor vessel that has
• 6:58
  to withstand intense multidirectional stresses, you
• 7:02
  absolutely cannot tolerate a material with a built in
• 7:05
  directional weakness. You need a completely isotropic
• 7:08
  material, one that is equally strong in every single
• 7:11
  direction. It's total sense, yeah.
• 7:13
  Which brings us to environments where even perfectly homogeneous
• 7:17
  carbon steel just won't survive. Like when a vessel is holding
• 7:21
  highly corrosive chemicals, carbon steel will just rust away
• 7:24
  into nothing, right? So we upgrade to alloys.
• 7:26
  Yeah, Austinitic stainless steels, duplex stainless steels,
• 7:30
  high nickel alloys. But corrosion is a broad term.
• 7:33
  A designer doesn't just ask, will it rust?
• 7:36
  The sources say they are terrified of a very specific
• 7:39
  localized phenomenon called pitting.
• 7:42
  Why does pitting keep vessel designers up at night?
• 7:45
  Well, general surface corrosion is annoying, but it's
• 7:48
  manageable. It's predictable.
• 7:50
  If you know a mild acid will eat away 1mm of carbon steel every
• 7:53
  10 years, you just make the vessel wall an extra 3mm thick
• 7:56
  to start with. It's just a math problem.
• 7:58
  You just build in a buffer. Exactly.
• 8:00
  But pitting is an ambush. Pitting doesn't attack the whole
• 8:03
  wall evenly. It is a highly localized,
• 8:05
  aggressive electrochemical drill that creates microscopic deep
• 8:09
  holes in the surface of an otherwise perfectly shiny
• 8:12
  uncorroded piece of stainless steel.
• 8:14
  But wait, a microscopic hole in a piece of seal that's like 2
• 8:18
  inches thick doesn't sound like a massive breach risk.
• 8:21
  Would it take decades for a microscopic pit to drill all the
• 8:24
  way through? The danger isn't the pit
• 8:26
  breaching the wall and causing a leak.
• 8:28
  The danger is what the shape of that pit does to the physics of
• 8:31
  the metal around it. A microscopic pit is effectively
• 8:34
  a tiny, incredibly sharp needle pointing directly into the
• 8:38
  thickness of the metal. Oh, I see where this is going.
• 8:40
  Yeah, in mechanical engineering, any sharp change in geometry is
• 8:44
  a stress riser. It takes the normal, safe stress
• 8:47
  flowing through that massive steel wall and concentrates it
• 8:50
  exponentially right at the microscopic tip of that pit.
• 8:53
  So it multiplies the stress locally.
• 8:55
  Massively, and that spike in localized stress will initiate a
• 8:59
  crack. Once the crack starts, the
• 9:02
  pressure inside the vessel rips it open instantly.
• 9:04
  The vessel fails catastrophically.
• 9:07
  It basically unzips long before the pit ever drilled all the way
• 9:10
  through. Man, that is terrifying.
• 9:13
  So how do you know if the expensive stainless steel you
• 9:15
  just bought is going to resist this microscopic drilling?
• 9:18
  The research brings up the pitting index, or Pi.
• 9:22
  It's a mathematical formula used to calculate a metal's
• 9:24
  resistance. Right.
• 9:26
  The pitting index is critical for material selection,
• 9:29
  especially in environments containing chlorides like
• 9:32
  seawater or certain chemical catalysts.
• 9:35
  Chlorides are the ultimate enemy of stainless steel.
• 9:38
  They aggressively attack the passive oxide layer that gives
• 9:42
  stainless it's rust resistance. Let's breakdown how this index
• 9:45
  is calculated. The math isn't just arbitrary
• 9:47
  numbers, right? It maps directly to what the
• 9:49
  alloy elements are actually doing.
• 9:51
  The base of the formula looks at the percentage of chromium in
• 9:54
  the metal. Yeah, Chromium is your frontline
• 9:56
  defense. It is the element that reacts
• 9:58
  with oxygen to form a microscopic invisible self
• 10:02
  healing shield over the steel. So the base of your pitting
• 10:05
  index score is just the pure percentage of chromium.
• 10:08
  More chromium equals a thicker, more robust shield.
• 10:11
  But then the formula adds the percentage of molybdenum and
• 10:14
  multiplies it by a factor of 3.3.
• 10:16
  Why does molybdenum get such a massive multiplier in the fight
• 10:20
  against pitting? Because if chromium is the
• 10:23
  shield, molybdenum is basically the emergency patch kit.
• 10:26
  When a chloride ion manages to breakthrough the chromium shield
• 10:30
  and starts drilling a pit, the environment inside that
• 10:33
  microscopic pit becomes highly acidic very, very quickly.
• 10:37
  And the chromium dissolves in that acid.
• 10:39
  Exactly. The chromium dissolves.
• 10:41
  Molybdenum, however, remains stable in reducing acids.
• 10:45
  It rushes to the site of the attack and repassivates the
• 10:47
  bottom of the pit, effectively plugging the hole and stopping
• 10:50
  the drill. Because it is so highly
• 10:52
  effective at stopping an active pit, it carries a weight 3.3
• 10:56
  times heavier than chromium in the formula.
• 10:58
  Wow. OK.
• 10:59
  And then there's a third component to the equation.
• 11:02
  You take the percentage of nickel in the alloy and you
• 11:04
  multiply it by an X Factor. And this X Factor is a moving
• 11:07
  target depending on the exact family of stainless steel you're
• 11:10
  using, right? Right, Nickel doesn't directly
• 11:13
  fight the chlorides the way chromium or molybdenum do.
• 11:16
  Instead, nickel fundamentally changes the crystalline
• 11:20
  structure of the metal. It transforms the steel into an
• 11:23
  Austinitic structure. What does that mean for the
• 11:25
  metal? It makes the metal incredibly
• 11:27
  tough, ductile, and stable at extreme temperatures.
• 11:31
  So if you are looking at a ferritic stainless steel which
• 11:34
  has no nickel and a different crystal structure, The X Factor
• 11:38
  is 0. Nickel plays no role.
• 11:40
  But if you look at a duplex stainless steel which is a
• 11:43
  hybrid structure. For a duplex stainless, The X
• 11:45
  Factor is 16, and for a fully Austinitic stainless steel The X
• 11:49
  Factor jumps to 30. So the math formula is π equals
• 11:53
  the percentage of chromium plus 3.3 multiplied by the percentage
• 11:57
  of molybdenum plus an X Factor multiplied by the percentage of
• 12:01
  nickel. And that final number is your
• 12:03
  pitting index. Correct.
• 12:05
  Engineers look at the chloride concentration and the
• 12:07
  temperature of the fluid they're trying to contain, and they
• 12:10
  consult tables that dictate the absolute minimum Pi score the
• 12:13
  metal must possess to survive that environment.
• 12:16
  OK, so we've fought off the chemical attack.
• 12:19
  We have a metal that won't rust and won't pit.
• 12:21
  But a pressure vessel isn't just sitting in a vacuum, you know?
• 12:24
  Yeah, it gets hit, it vibrates, pumps surge and cause massive
• 12:28
  dynamic pressure spikes. What are we looking for in the
• 12:31
  Material Beta to ensure the vessel can take a physical
• 12:34
  beating? We are looking at 2 highly
• 12:37
  interrelated properties, the modulus of resilience and notch
• 12:41
  toughness. Let's start with resilience.
• 12:43
  Resilience is the materials capacity to act like a spring.
• 12:47
  Technically, it is the amount of strain energy the metal can
• 12:50
  absorb per unit volume without yielding.
• 12:54
  If a sudden pressure surge hits the wall, the wall needs to
• 12:56
  stretch outward to absorb that energy and then seamlessly
• 12:59
  spring back to its original shape without permanently
• 13:02
  deforming. So a high modulus of resilience
• 13:04
  means the metal has a massive capacity to swallow those
• 13:07
  dynamic impacts. Exactly.
• 13:09
  And what about notch toughness? Notch toughness is the safety
• 13:12
  net for when reality sets in. In a perfect world, your vessel
• 13:16
  has perfectly smooth, polished walls, but in reality there
• 13:20
  might be a scratch from a grinder, a tiny undercut from a
• 13:23
  Weld, or a microscopic slag inclusion that is a notch.
• 13:27
  The built in weak point. Yeah, notch toughness is the
• 13:30
  measure of how much energy the material can absorb before a
• 13:34
  crack initiates at that notch and propagates through the
• 13:36
  metal. I want to pause here and pull
• 13:38
  these threads together because the sources detail a scenario
• 13:42
  that I think is probably the most dangerous trap a young
• 13:44
  engineer can fall into. I call it the Fracture Mechanics
• 13:48
  Paradox. It is the moment where trying to
• 13:51
  make a vessel better actually guarantees it will explode.
• 13:54
  Oh, it is a lethal oversight, and it happens when designers
• 13:57
  look at numbers on a spreadsheet without really understanding the
• 14:00
  physical trade-offs in metology. Let's walk through the exact
• 14:02
  example from the research. We are tasked with designing a
• 14:06
  thin walled cylindrical vessel. It's roughly 3/4 of a meter in
• 14:10
  diameter. It has to hold a serious
• 14:12
  internal pressure about 34.5 megapascals, which is around
• 14:16
  5000 PSI. Right, that's a serious load.
• 14:19
  So we run the standard calculations using a standard
• 14:22
  carbon steel, and the math tells us the wall needs to be 1.26
• 14:27
  centimeters thick, about half an inch.
• 14:29
  Now half and steel plate is heavy, it's expensive to buy,
• 14:32
  it's expensive to ship, and it takes a lot of time and filler
• 14:36
  metal to Weld together. So the designer looks at that
• 14:38
  and thinks how do I optimize this?
• 14:41
  And the most obvious, seemingly logical answer is to use a
• 14:44
  stronger material. I mean, if we upgrade to a high
• 14:47
  yield strength steel alloy, the metal can take more stress
• 14:50
  before it bends, and because it can take more stress, the math
• 14:54
  allows us to make the wall significantly thinner while
• 14:56
  theoretically holding the exact same 5000PSI.
• 15:00
  Right. We save weight, we save money.
• 15:02
  The welds are faster. It looks like a massive win
• 15:04
  across the board. But fracture mechanic steps in
• 15:07
  and reveals the trap. The paradox is this, as you
• 15:11
  increase the yield strength of a steel alloy, you almost always
• 15:14
  decrease his fracture toughness. Yes, you're trading ductility
• 15:18
  for raw strength. The metal becomes much harder,
• 15:21
  but it also becomes stiffer and far more brittle.
• 15:24
  It's like trading a flexible climbing rope for a solid steel
• 15:27
  cable. The cable can technically hold
• 15:30
  more static weight, but if there's a sudden jerk, the rope
• 15:33
  stretches while the cable might just snap.
• 15:36
  That's a great analogy. And now apply that back to our
• 15:39
  vessel. We've built it out of this super
• 15:41
  strong high yield steel and the walls are now very thin.
• 15:45
  The pressure is applied, the stress in the wall is incredibly
• 15:49
  high, but technically below the yield point.
• 15:52
  So the math says we're safe. But remember the reality of the
• 15:55
  shop floor. There is a microscopic flaw in
• 15:58
  that thin wall, maybe a tiny crack in the heat affected zone
• 16:02
  of the longitudinal Weld. If the wall was thick and made
• 16:04
  did the standard ductile steel, what would happen?
• 16:08
  In the thicker, more ductile wall, the immense stress
• 16:10
  concentrating at the tip of that microscopic crack would cause
• 16:13
  the metal immediately surrounding the crack to yield.
• 16:16
  It would stretch plastically on a microscopic level.
• 16:18
  So it absorbs the stress. Exactly that.
• 16:20
  Stretching physically blunts the tip of the crack and absorbs the
• 16:23
  energy, stopping the crack in its tracks.
• 16:25
  The vessel survives. But with our optimized thin high
• 16:28
  yield wall. The high yield steel doesn't
• 16:30
  want to stretch. It's too stiff.
• 16:32
  It lacks the fracture toughness. So when that same stress
• 16:36
  concentrates at the crack tip, the metal can't yield to absorb
• 16:40
  it. Instead, the crack propagates.
• 16:42
  And because the wall is thin. Because the wall is so thin,
• 16:45
  there isn't enough material volume to arrest the running
• 16:48
  crack. The crack shoots across the
• 16:51
  entire length of the vessel at the speed of sound, and the
• 16:53
  vessel detonates. The designer, by choosing a
• 16:56
  stronger material to save weight, actually engineered a
• 16:59
  bomb because they ignored the relationship between yield
• 17:01
  strength and fracture toughness. Man, that is deeply unsettling.
• 17:05
  I mean, if I am the welder on the floor, I'm just following
• 17:08
  the procedure. I look at the blueprint, I see
• 17:10
  the thin plate, I grab the specified filler metal, and I
• 17:13
  lay down what looks like a visually flawless Weld.
• 17:16
  But you were saying that even with a perfect Weld, the vessel
• 17:19
  is fundamentally unsafe before I even strike an arc, purely
• 17:23
  because of a blind spot in the design.
• 17:25
  Yes, unfortunately the welder can do everything right.
• 17:29
  But if the material architecture lacks the inherent toughness to
• 17:32
  tolerate the microscopic imperfections that exist in all
• 17:35
  fabricated structures, the system is doomed.
• 17:38
  You cannot evaluate a material based solely on its ability to
• 17:42
  resist yielding. You have to evaluate how it
• 17:44
  behaves when it inevitably fails locally.
• 17:47
  Does it yield and forgive or does it snap and destroy?
• 17:50
  Right. And that transition from the
• 17:52
  microscopic crystal structure of the metal to the macro level
• 17:55
  physics of the vessel brings us to the next massive hurdle.
• 17:58
  We know what metal we are using. Now we have to figure out how to
• 18:01
  shape it to survive. The architecture of containment
• 18:04
  and the very first dividing line in pressure vessel design
• 18:07
  philosophy is establishing the mathematical boundary between a
• 18:10
  thin walled vessel and a thick walled vessel.
• 18:12
  And this isn't just a visual distinction like looking at a
• 18:15
  soda can versus a scuba tank, right?
• 18:17
  There is a rigid mathematical ratio that changes the entire
• 18:20
  physics of the analysis. It is the radius to thickness
• 18:24
  ratio. You take the internal radius of
• 18:26
  the cylinder and divide it by the thickness of the wall.
• 18:28
  If that ratio is greater than 10, you are operating in the
• 18:32
  thin walled universe. This is where the vast majority
• 18:35
  of industrial storage tanks, distillation columns and
• 18:38
  standard pressure vessels live. Why is the number 10 the magic
• 18:42
  threshold? What changes mechanically at
• 18:44
  that point? It's about how the stress
• 18:47
  distributes through the metal in a thin walled vessel.
• 18:50
  We assume that the membrane stresses which are at the forces
• 18:53
  acting tangentially to the surface, basically pulling the
• 18:56
  metal apart like a stretch balloon.
• 18:58
  We assume those are completely uniform across the entire
• 19:01
  thickness of the wall. So the inside skin of the steel
• 19:03
  and the outside skin of the steel are experiencing the exact
• 19:06
  same amount of tension. Exactly because the wall is
• 19:09
  relatively thin, we can also mathematically ignore the radial
• 19:12
  stress. Radial stress being the direct
• 19:15
  outward push of the fluid against the inside face of the
• 19:18
  steel, right? Correct, It's compressing the
• 19:20
  wall outward. In a thin walled vessel, the
• 19:23
  membrane tension tearing the wall apart is so massively
• 19:26
  dominant that the radial compression is basically
• 19:29
  negligible. But if your radius to thickness
• 19:31
  ratio drops below 10, so you have a high pressure reactor
• 19:35
  with incredibly thick walls, everything changes.
• 19:39
  You are now in thick walled territory.
• 19:41
  And the uniform stress assumption just goes out the
• 19:43
  window. Completely in a thick walled
• 19:45
  vessel, the radial stress becomes a massive factor and
• 19:49
  more importantly, the membrane stress is no longer uniform.
• 19:52
  The tension on the inside skin of the vessel is significantly
• 19:55
  higher than the tension on the outside skin.
• 19:58
  Oh, so it's uneven? Very uneven.
• 20:00
  The stress creates A steeper gradient through the thickness
• 20:02
  of the steel. The math required to ensure the
• 20:05
  inner layers don't tear while the outer layers are barely.
• 20:07
  Doing any work is a vastly more complex.
• 20:10
  But for the sake of our primary focus today, we are staying in
• 20:13
  the thin walled uniform stress universe.
• 20:16
  We have our calculated stress. The next step is to compare that
• 20:20
  stress against the materials limits to see if the vessel will
• 20:23
  survive and the sources present 2 competing schools of thought
• 20:26
  on how to predict failure, the maximum stress theory and the
• 20:30
  maximum shear stress theory. Right.
• 20:32
  These theories are attempts to mathematically model exactly how
• 20:35
  a piece of metal gives up. The maximum stress theory is the
• 20:38
  oldest, simplest, and most historically widespread.
• 20:41
  It is actually the foundational theory behind the incredibly
• 20:44
  ubiquitous ASME Boiler and Pressure Vessel Code, Section 8,
• 20:49
  Division One. OK, so how does the maximum
• 20:51
  stress theory view failure? It looks at the world in one
• 20:54
  dimension at a time. It states that failure occurs
• 20:57
  the moment the maximum principal stress in any single direction
• 21:00
  exceeds the simple tensile yield strength of the material.
• 21:04
  Imagine pulling a steel bar straight apart.
• 21:06
  The theory says I only care about the force pulling the bar.
• 21:09
  If that force exceeds the limit, it breaks.
• 21:11
  I don't care what is happening in the other directions.
• 21:14
  Which sounds incredibly logical, honestly.
• 21:16
  If you pull it too hard, it breaks.
• 21:18
  Why would you need another? Theory, because that logic only
• 21:21
  holds true for brittle materials like cast iron or glass, which
• 21:25
  actually do fail by pulling straight apart.
• 21:27
  But remember our entire discussion about carbon steel
• 21:30
  and ductility? Right, the layers sliding.
• 21:32
  Exactly. Ductile metals rarely fail by
• 21:36
  just snapping in pure tension. They fail by slipping.
• 21:40
  They deform plastically, and that plastic deformation is
• 21:43
  driven by shear forces, not pure tensile forces.
• 21:46
  I want to make sure this is crystal clear for everyone.
• 21:49
  If I am pulling a piece of ductile steel straight apart end
• 21:52
  to end, the tension is horizontal, but you are saying
• 21:56
  the metal isn't actually failing horizontally.
• 21:58
  That's right, it's failing diagonally.
• 22:00
  When you pull that bar, the crystalline lattice we talked
• 22:02
  about earlier starts to slide. And the mathematical maximum
• 22:05
  shear stress, the force trying to slide the atomic layers past
• 22:09
  one another, always occurs on a plane that is exactly 45° angled
• 22:13
  away from the direction you are pulling.
• 22:15
  Think of a deck of cards. If you push the top of the deck
• 22:18
  sideways, the cards slide over each other.
• 22:20
  That is, sheer ductile metals fail by shearing along those 45°
• 22:24
  slip planes long before the pure tensile stress reaches its
• 22:28
  absolute maximum limit. So the maximum shear stress
• 22:31
  theory accounts for this. It looks at the interaction of
• 22:34
  stresses in multiple directions to predict when that diagonal
• 22:37
  slip will actually occur. Yes.
• 22:39
  The maximum shear stress theory states that yielding begins when
• 22:43
  the maximum shear stress at a given point reaches 1/2 of the
• 22:46
  uniaxial yield strength. It is a much more accurate
• 22:50
  representation of reality for the ductile alloys we use in
• 22:53
  pressure vessels. To put this in everyday terms,
• 22:55
  it's like evaluating a rope bridge.
• 22:58
  The maximum stress theory just looks at the total weight on the
• 23:00
  bridge, but the maximum shear stress theory looks at the
• 23:03
  weight, and it also considers the wind blowing the bridge
• 23:06
  sideways and the friction of the ropes grinding against each
• 23:09
  other. It realizes that the combination
• 23:11
  of forces is what actually phrased the rope.
• 23:14
  That is an excellent way to conceptualize it, yeah.
• 23:16
  To get your action of forces is the real killer.
• 23:19
  But if the shear stress theory is more accurate for the metals
• 23:22
  we actually use, why does the most famous engineering code in
• 23:26
  the world, ASME Section 8 Division One, still use the
• 23:30
  older, less accurate maximum stress theory?
• 23:33
  It comes down to the philosophy of design by rules versus design
• 23:37
  by analysis. Doing a full multidirectional
• 23:40
  shear stress analysis on every single square inch of a complex
• 23:44
  vessel, around every nozzle, every support bracket, every
• 23:48
  change in thickness requires immense computational power and
• 23:51
  highly advanced finite element analysis.
• 23:54
  Decades ago, that just wasn't possible, so the ASME code
• 23:57
  adopted A pragmatic approach. The design by rules approach.
• 24:00
  Right, The code basically says we know the maximum stress
• 24:03
  theory is a simpler, slightly less accurate model.
• 24:05
  We know there are localized secondary bending stresses
• 24:08
  around the nozzles that we aren't explicitly calculating.
• 24:11
  To compensate for this mathematical ignorance, we're
• 24:14
  going to impose a massive global safety factor.
• 24:17
  Use the simpler formulas, but the code artificially
• 24:19
  drastically lowers the allowable stress you were permitted to
• 24:22
  use. So the penalty for easier math
• 24:24
  is buying way more steel. Exactly.
• 24:26
  The vessel ends up significantly thicker, heavier, and more
• 24:29
  expensive than it theoretically needs to be, but that massive
• 24:33
  thickness acts as a brute force blanket, absorbing all those
• 24:36
  complex sheer and bending stresses that the simpler math
• 24:39
  ignored. It's an incredibly safe,
• 24:41
  conservative way to build, but it is definitely not optimized.
• 24:45
  OK, so the walls are thick, but what about the overall physical
• 24:49
  shape? If a facility says they need a
• 24:51
  vessel to hold 10,000 gallons of pressurized gas, does the
• 24:55
  designer just arbitrarily pick a length and a width?
• 24:57
  Not at all. Geometry dictates material
• 24:59
  efficiency. You're always trying to maximize
• 25:02
  the internal volume while minimizing the surface area,
• 25:05
  because surface area equals tons of expensive steel.
• 25:08
  If you are building an atmospheric storage tank, say a
• 25:11
  massive diesel tank with 0 internal pressure, the optimum
• 25:15
  link to diameter ratio or L / D is 1.
• 25:18
  Wait, an L / D of 1 is essentially a cylinder that is
• 25:21
  exactly as wide as it is tall, like a giant soup can.
• 25:24
  Exactly. A perfectly spherical shape
• 25:26
  would be the absolute maximum efficiency, but building a giant
• 25:30
  sphere out of flat plate is a nightmare of welding and
• 25:33
  fabrication. A cylinder with an L / D of 1 is
• 25:38
  the most costeffective compromise for atmosheric
• 25:40
  storage. But the moment you introduce
• 25:42
  pressure, the Shae has to stretch out.
• 25:44
  Right. Yes, for pressurized vessels,
• 25:46
  the flat ends, the heads become incredibly thick and expensive
• 25:50
  to handle the outward thrust of the pressure.
• 25:52
  To minimize the cost of those massive heads, you make the
• 25:55
  diameter smaller and stretch the length of the cylinder out.
• 25:58
  So it becomes more tube like. Yeah, for standard pressurized
• 26:02
  vessels with a wall thickness under 2 inches, the optimum L /
• 26:06
  D ratio is generally 6. If the pressure is so extreme
• 26:10
  that the walls are over 2 inches thick, the optimum ratio
• 26:13
  stretches all the way up to 8. You get these long slender
• 26:16
  bullet shaped tanks. But we've only been talking
• 26:19
  about internal pressure trying to blow the tank apart.
• 26:22
  The research dives deeply into the inverse problem, which
• 26:25
  honestly sounds far more terrifying mechanically.
• 26:28
  External pressure. External pressure is a
• 26:30
  completely different physical regime.
• 26:32
  Internal pressure places the steel in tension.
• 26:36
  Steel loves tension. It naturally wants to pull taut
• 26:40
  and stable external pressure places the steel and
• 26:43
  compression. The atmosphere is literally
• 26:45
  trying to crush the vessel inward.
• 26:48
  When would a heavy industrial vessel ever experience external
• 26:52
  crushing pressure? Are we talking about submarines?
• 26:55
  Submarines are the classic example, sure, but in a chemical
• 26:58
  plant it happens constantly. A vessel might be operating
• 27:01
  under a design vacuum to lower the boiling point of a chemical,
• 27:04
  but the most dangerous scenario is accidental vacuum during
• 27:07
  routine maintenance. How does that happen?
• 27:10
  Well, you have a massive vessel that has just been cleaned with
• 27:12
  live super heated steam. The entire interior is full of
• 27:16
  hot steam. The maintenance crew shuts the
• 27:18
  valves and goes to lunch. As the vessel cools, the steam
• 27:21
  condenses back into a tiny puddle of liquid water.
• 27:24
  And because the water takes up a fraction of the volume of the
• 27:26
  steam, it pulls a massive vacuum inside the sealed tank.
• 27:29
  A near perfect vacuum. Suddenly you have the full
• 27:32
  weight of the Earth's atmosphere. 14.7 lbs on every
• 27:36
  single square inch of the vessel's surface, pushing
• 27:38
  inward. I understand the physics of the
• 27:41
  vacuum, but here's what I struggle with.
• 27:44
  We are talking about vessels built with half inch or one inch
• 27:47
  thick solid steel walls. How can near atmospheric
• 27:51
  pressure crush solid steel? The stress can't possibly exceed
• 27:55
  the yield strength of the metal. That is the terrifying beauty of
• 27:58
  external pressure doesn't need to exceed the yield strength of
• 28:02
  the metal. External pressure causes a
• 28:04
  failure of geometric instability and not material strength.
• 28:07
  We call it buckling. How does buckling bypass the
• 28:10
  strength of the steel? Think of an empty aluminum soda
• 28:13
  can. If you stand perfectly straight
• 28:15
  on top of it, the microscopically thin aluminum
• 28:18
  wall can actually support your entire body weight.
• 28:20
  In pure compression, the material is strong enough, but
• 28:24
  if someone walks by and taps the side of the can with a pencil,
• 28:27
  introducing a tiny dent, a geometric imperfection, the
• 28:30
  entire cylinder instantly collapses and crushes flat.
• 28:33
  Because the shape failed. Exactly.
• 28:35
  The material didn't get weaker, the geometry became unstable.
• 28:39
  A massive steel vessel is exactly the same.
• 28:42
  The rolling machines on the shop floor can never create a
• 28:45
  mathematically perfect circle. There's always a slight out of
• 28:48
  roundness. When external pressure hits that
• 28:51
  slightly flattened area, the geometry becomes unstable and
• 28:55
  the massive steel wall instantly buckles inward, folding like
• 28:59
  paper at a stress level far below what the metal could
• 29:02
  theoretically handle. So how do you stop a 50 foot
• 29:06
  long steel cylinder from buckling under vacuum?
• 29:08
  You have to fundamentally change its geometry.
• 29:11
  Designers use stiffening rings. These are heavy steel I beams or
• 29:15
  thick plates welded edge on around the outside circumference
• 29:18
  of the vessel. What are those rings actually
• 29:20
  do? What these rings do is
• 29:21
  effectively break 1 long, mathematically vulnerable
• 29:24
  cylinder into a series of very short, incredibly stiff
• 29:28
  cylinders. The short distance between the
• 29:30
  rings prevents the buckling wave from ever forming.
• 29:33
  But there's a massive red flag on the shop floor regarding
• 29:35
  these rings, right? The placement isn't arbitrary at
• 29:37
  all. Oh no, it is a cardinal rule of
• 29:40
  fabrication. Stiffening rings must never ever
• 29:43
  be welded directly over a circumferential Weld seam
• 29:46
  joining 2 sections of the shell. Because welding the ring over
• 29:49
  the shell seam means you are stacking heat affected zones.
• 29:52
  Exactly. You are crossing a massive
• 29:54
  circumferential Weld with all its residual stresses and
• 29:57
  potential internal flaws. With two more filet welds to
• 30:00
  attach the ring, you are creating a metallurgical
• 30:03
  nightmare of overlapping brittle zones and multidirectional
• 30:06
  stress concentrations. If a crack is going to start
• 30:09
  anywhere, it will start there. The rings must always be offset
• 30:13
  from the shell scenes. OK, so the vessel is designed.
• 30:16
  The walls are thick enough to handle the tension, the rings
• 30:18
  are spaced to prevent the buckling.
• 30:20
  Now we have to actually stand this massive structure up and
• 30:23
  support it. The sources detail several ways
• 30:25
  to do this, starting with skirts.
• 30:27
  Why are skirts the gold standard for vertical vessels?
• 30:30
  A skirt is brilliantly simple. It is just a cylindrical
• 30:33
  continuation of the vessel's shell that extends downward and
• 30:37
  bolts to the concrete foundation.
• 30:39
  The reason it is the gold standard is load transfer.
• 30:42
  If you make the diameter of the skirt exactly match the diameter
• 30:45
  of the vessel, the massive dead weight of the vessel transfers
• 30:48
  straight down the walls directly into the foundation.
• 30:51
  There is no offset, which means there are minimal discontinuity
• 30:55
  stresses. But wait, the vessel might be
• 30:57
  operating at 800°F while the concrete foundation is at
• 31:01
  ambient temperature. Doesn't that extreme temperature
• 31:04
  difference travel down the steel skirt and cause problems?
• 31:07
  Oh, it causes enormous problems. The thermal gradient creates
• 31:11
  massive bending stresses right at the well joint where the hot
• 31:14
  vessel meets the cooler skirt. The metal is trying to expand
• 31:18
  and contract at different rates in the exact same location.
• 31:21
  To mitigate this, engineers use a hot box design.
• 31:25
  How does that work? They Weld the skirt on and then
• 31:27
  they pack the top few feet of the skirt both inside the
• 31:30
  cylinder and outside with heavy insulation.
• 31:33
  What does that insulation do? It prevents the immediate area
• 31:36
  below the Weld from cooling down too quickly.
• 31:38
  It forces the temperature drop from 800° to ambient to occur
• 31:42
  gradually over the length of the skirt, rather than violently
• 31:46
  right at the Weld seam. It spreads the thermal strain
• 31:49
  out over a large area, relaxing the stress.
• 31:52
  That is a very elegant solution. What about legs?
• 31:55
  You see smaller tanks standing on three or four legs all the
• 31:58
  time. Legs are cheap and easy, which
• 32:00
  is why you see them on smaller low pressure tanks, but
• 32:03
  mechanically they are a nightmare.
• 32:05
  Instead of spreading the load evenly around a continuous
• 32:08
  cylinder like a skirt, a leg concentrates the entire weight
• 32:11
  of the vessel. The fluid and the wind loads
• 32:14
  into 3 or 4 tiny localized patches on the thin shell wall.
• 32:19
  They act like hole punches. Exactly right.
• 32:21
  They punch into the shell if the legs are longer than about 7
• 32:24
  feet. The bending moments get so high
• 32:26
  that you have to Weld cross bracing between them to keep
• 32:29
  them from splaying out. But the real danger is dynamic
• 32:32
  loading. If a vessel is mounted on a
• 32:34
  vibrating compressor platform or has an internal agitator
• 32:37
  violently mixing fluids, you should never use legs.
• 32:40
  Why not? The rhythmic vibration will
• 32:42
  cause the localized stress patches where the legs meet the
• 32:45
  shell to fatigue and the legs will literally tear a hole right
• 32:49
  out of the side of the tank. Yikes.
• 32:51
  And for massive horizontal vessels, yeah, we aren't
• 32:54
  standing them on end. They're lying flat like a
• 32:56
  submarine. For horizontal vessels, you'd
• 32:58
  use saddles for the dead weight, but if the vessel is located in
• 33:02
  an earthquake zone or hurricane zone, you have massive lateral
• 33:06
  shear forces trying to RIP the vessel off its foundation.
• 33:09
  For that you use ring girders. How do those differ from
• 33:13
  saddles? A saddle just cradles the bottom
• 33:15
  half of the tank. A ring girder is a massive rigid
• 33:18
  steel hoop that completely encircles the vessel.
• 33:22
  The vessel sits inside the hoop and the hoop is anchored to the
• 33:24
  steel framework of the facility. When a seismic wave hits the
• 33:28
  violent horizontal here, force is transferred from the thin
• 33:31
  skin of the vessel into the massive rigidity of the ring
• 33:34
  girder and down into the foundation, preventing the thin
• 33:37
  shell from twisting and tearing open.
• 33:39
  It is incredible how the entire architecture of the vessel has
• 33:42
  to fight a war on 2 fronts. The chemical and pressure war on
• 33:46
  the inside, and the gravity, wind, and thermal war on the
• 33:49
  outside. But all of this brilliant CAD
• 33:52
  architecture is completely useless if the pieces aren't put
• 33:55
  together perfectly. Which brings us to the next
• 33:59
  arguably most critical phase. Once the geometry is finalized,
• 34:03
  it's time to cut plate and strike an arc.
• 34:04
  The assembly phase, Yeah, This is where the mathematical theory
• 34:07
  meets the molten reality of the Weld pool.
• 34:10
  And the bridge between those two worlds is a concept called joint
• 34:13
  efficiency. Joint efficiency, yeah.
• 34:15
  In the math it is represented by the Greek letter ETA.
• 34:18
  What exactly is ETA measuring? ETA is a blunt mathematical
• 34:22
  ratio. It is the strength of the welded
• 34:24
  joint and divided by the original strength of the solid
• 34:26
  parent plate. It asks the basic question, how
• 34:29
  much damage did we do to the structural integrity of this
• 34:32
  steel cylinder by cutting it in half and melting it back
• 34:35
  together? So a perfect score would be
• 34:36
  1.0100% efficiency. Yes, if you have a fully
• 34:42
  penetrated double sided butt Weld and you perform 100%
• 34:46
  volumetric radio graphic testing, meaning you X-ray every
• 34:50
  single millimeter of the Weld to prove there is not a single
• 34:53
  microscopic pore or slag inclusion inside it, the code
• 34:56
  grants you an ADA of 1.0. The math treats the cylinder as
• 35:00
  if it were a seamless, continuous piece of steel.
• 35:03
  But X-ray miles of Weld seam on a massive vessel is incredibly
• 35:06
  expensive and time consuming. What if a client doesn't want to
• 35:09
  pay for all that testing? The code allows it, but it
• 35:12
  heavily penalizes the design. If you only spot check the
• 35:15
  welds, or if you use a less reliable joint design like a
• 35:18
  single sided Weld with a backing strip, the code will arbitrarily
• 35:21
  drop your giant efficiency factor.
• 35:23
  Your Ada might drop to 0.85 or even 0.70.
• 35:27
  The code assumes that because you didn't prove the Weld is
• 35:29
  perfect, it is inherently flawed.
• 35:31
  And how does that penalty actually manifest in the design?
• 35:34
  Does the designer just accept that the vessel is weaker?
• 35:37
  Oh no, the vessel still has to hold the design pressure.
• 35:40
  The penalty is financial and physical.
• 35:43
  If your Weld is only considered 70% as strong as the base metal,
• 35:47
  the only way to make the weak Weld strong enough to hold the
• 35:49
  pressure is to make the entire vessel thicker.
• 35:52
  You have to specify vastly heavier, thicker, more expensive
• 35:57
  steel plates for the whole project just to compensate for
• 36:00
  the theoretical weakness of the cheaper uninspected welds.
• 36:04
  It's a direct trade off between the cost of non destructive
• 36:07
  testing and the cost of raw steel.
• 36:10
  Let's look closely at those welds, because here's a concept
• 36:13
  from the sources that completely blew my mind.
• 36:15
  When you build a cylindrical vessel out of rolled steel
• 36:18
  plates, you have two types of welds.
• 36:20
  You have longitudinal welds, which runs straight down the
• 36:23
  length of the cylinder from end to end, and you have
• 36:25
  circumferential welds, which wrap around the cylinder like a
• 36:28
  belt joining the ends of the cylinders together.
• 36:30
  I was assumed all the welds were under the same pressure, but
• 36:33
  they absolutely are not. No, they're not.
• 36:36
  The stress distribution in a pressurized cylinder is highly
• 36:39
  asymmetrical, and the math behind it is elegant, but the
• 36:43
  implications are terrifying. Let's break down the two forces.
• 36:46
  We have longitudinal stress and we have hoop stress.
• 36:49
  Right longitudinal stress is the force trying to stretch the
• 36:52
  cylinder lengthwise, basically trying to blow the two end caps
• 36:55
  off. Hoop stress, also known as
• 36:57
  circumferential stress, is the force trying to expand the
• 37:01
  diameter of the cylinder. It is trying to tear the
• 37:04
  cylinder open like a bursting pipe.
• 37:06
  And the staggering fact is that the hoop stress is exactly twice
• 37:09
  as large as the longitudinal stress.
• 37:12
  Why is that? Why is the pressure fighting so
• 37:14
  much harder to expand the diameter than it is to blow the
• 37:18
  ends off? It comes down to the surface
• 37:20
  area the pressure's acting against compared to the amount
• 37:22
  of steel available to hold it back.
• 37:24
  Think about the longitudinal stress first.
• 37:26
  The pressure is pushing against the flat circular area of the
• 37:29
  end cap. The steel fighting that pressure
• 37:31
  is the entire circumference of the cylinder wall.
• 37:34
  OK, that makes sense. Now think about the hoop stress.
• 37:37
  Imagine slicing the cylinder in half lengthwise.
• 37:40
  The pressure is pushing against the massive rectangular
• 37:43
  projected area of the entire length of the tank, but the only
• 37:47
  steel fighting that massive outward push is the 2 thin cross
• 37:51
  sections of the wall you just cut.
• 37:53
  Because the projected area trying to split the tank is so
• 37:56
  much larger relative to the resisting steel, the resulting
• 37:59
  stress is exactly double. So the hoop stress is twice as
• 38:03
  intense. Now apply that back to the
• 38:05
  physical welds on the shop floor.
• 38:07
  This is where it gets critical. Look at the orientation of the
• 38:10
  welds. The circumferential Weld, the
• 38:13
  belt wrapping around the tank is oriented perpendicular to the
• 38:16
  longitudinal stress. It is only fighting the force
• 38:19
  trying to blow the ends off, which is the weaker stress.
• 38:21
  But the longitudinal Weld, the straight seam running end to
• 38:24
  end, is oriented exactly perpendicular to the massive
• 38:28
  hoop stress. Exactly.
• 38:29
  The longitudinal Weld is taking the absolute worst of the
• 38:32
  beating. The force pulling across that
• 38:34
  seam is double the force pulling across the circumferential
• 38:37
  seams. The hoop stress is violently
• 38:39
  trying to RIP that specific Weld apart.
• 38:41
  So that's why vessels fail the way they do.
• 38:43
  Yeah, This is why when a cylindrical vessel fails from
• 38:46
  overpressure, it almost never pops Its end caps off like a
• 38:50
  champagne cork. It splits straight down the
• 38:52
  side, down its length, like a hot dog expanding on a grill.
• 38:56
  The skin of the hot dog splits lengthwise because the hoop
• 38:58
  stress exceeds the strength of the casing before the
• 39:01
  longitudinal stress can grow the ends off.
• 39:03
  Because of this dynamic, the longitudinal welds are by far
• 39:07
  the most critical joints in the entire structure.
• 39:10
  They are the weak links holding back the most terrifying forces.
• 39:13
  They require the highest quality filler metal, the strictest heat
• 39:16
  controls, and the most rigorous X-ray inspections.
• 39:19
  A flaw in a circumferential Weld is bad, but a flaw in a
• 39:24
  longitudinal Weld is catastrophic.
• 39:26
  If the longitudinal Weld is this massive vulnerability holding
• 39:29
  back double the stress, my immediate question is why do we
• 39:33
  just get rid of it? Why don't we just forge seamless
• 39:35
  cylinders for everything? Just cast a solid pipe and cap
• 39:38
  the ends. It's a great question.
• 39:41
  And for extremely high pressure specialized applications like
• 39:45
  deep sea submersibles or high caliber artillery barrels, we
• 39:49
  actually do use thick walled monoblock forgings.
• 39:53
  They take a solid glowing ingot of steel and use a massive
• 39:57
  hydraulic press to Pierce a hole straight through the center,
• 40:00
  creating a completely seamless tube.
• 40:02
  So why not do that for everything?
• 40:04
  Because you cannot do that for a refinery distillation column
• 40:07
  that is 20 feet in diameter and 150 feet tall.
• 40:11
  The scale is impossible and the cost of forging something that
• 40:14
  massive would bankrupt the project.
• 40:17
  For industrial scale, you have to buy flat steel plates, roll
• 40:20
  them into arcs and Weld them together.
• 40:22
  The longitudinal seam is an unavoidable reality of scale and
• 40:25
  economics. And we also have to deal with
• 40:26
  the unavoidable reality of time. I mean, we build this vessel,
• 40:29
  it's perfectly welded, and we put it into service.
• 40:31
  But steel rusts. How does a designer
• 40:34
  mathematically guarantee that a vessel won't burst 20 years from
• 40:37
  now when half the steel is corroded away?
• 40:39
  They use a concept called the corrosion allowance, denoted by
• 40:42
  the letter C The designer does all the complex math we just
• 40:46
  discussed to figure out exactly how thick the steel needs to be
• 40:49
  just to hold the pressure. Let's call that thickness T.
• 40:53
  Then they consult the chemists to determine how fast the
• 40:56
  specific fluid will eat the steel.
• 40:58
  Oh, so they calculate a specific rate of decay?
• 41:01
  Right, if they calculate it will lose 3mm of steel over a 20 year
• 41:05
  lifespan, they add 3mm of sacrificial steel to the design.
• 41:10
  The plate they buy is T + C. But here is the critical rule
• 41:14
  from the sources. When they are running the stress
• 41:16
  calculations, checking for buckling, calculating the joint
• 41:19
  efficiency, which number do they use?
• 41:22
  They must by law use the end of life thickness, which is t -, C
• 41:27
  You completely ignore the sacrificial metal in all your
• 41:29
  strength calculations. You have to mathematically prove
• 41:32
  that the vessel is structurally safe and compliant on the very
• 41:36
  last day of its intended 20 year life, when all that extra metal
• 41:39
  has rusted away into nothing. It's literally engineering for
• 41:43
  the worst case scenario area. And Speaking of worst case
• 41:46
  scenarios, let's look at what happens when the roll plates
• 41:49
  don't actually line up erfectly on the shop floor.
• 41:51
  I mean the CAD model shows erfectly aligned cylinders of
• 41:54
  budding perfectly spun dished heads, but we are bending 2 inch
• 41:59
  thick steel plate. Right, and heavy steel plate has
• 42:02
  a memory. When you run it through the
• 42:03
  massive 3 roll bending machines, you over bend it slightly
• 42:07
  because when it rolls out it springs back.
• 42:10
  Getting a perfectly true, mathematically precise circle is
• 42:13
  virtually impossible. So when the fitters try to butt
• 42:16
  two large rolled cylinders together, the edges don't match
• 42:19
  up perfectly. One lip sits higher than the
• 42:21
  other. This is called misalignment.
• 42:23
  Why is misalignment so dangerous?
• 42:25
  Can't the welder just lay a really thick wide bead of feller
• 42:28
  metal over the gap to smooth it out?
• 42:30
  You can visually smooth it out, sure, but you cannot hide the
• 42:33
  physics. When the vessel pressurizes, the
• 42:37
  enormous internal force tries to force the misaligned slightly
• 42:41
  out of round cylinders into a perfect true circle.
• 42:44
  This violent shift creates massive localized bending
• 42:48
  stresses exactly at the Weld joint, tearing at the root of
• 42:51
  the Weld from the inside. So it's constantly prying at the
• 42:53
  Weld. Yes, that is why the tolerances
• 42:56
  for misalignment are incredibly strict.
• 42:59
  The code might dictate that the offset cannot exceed the
• 43:01
  thickness of the plate divided by 10 up to an absolute maximum
• 43:05
  of maybe 5mm. And if the fit up is worse than
• 43:08
  5mm? The inspector halts the job.
• 43:10
  You have to cut the tack welds, hook the plates back up to the
• 43:13
  cranes, use massive hydraulic jacks and strong backs to
• 43:16
  physically force the steel into alignment, and try again.
• 43:20
  You cannot bridge bad geometry with Weld metal.
• 43:22
  Knowing that the longitudinal Weld is holding back the massive
• 43:25
  hoop stress, and knowing that any misalignment is actively
• 43:28
  trying to pry that joint apart, the actual process of melting
• 43:32
  the steel and laying that bead of filler metal is paramount.
• 43:37
  The theory has taken us as far as it can.
• 43:39
  Now we need the sparks. Exactly.
• 43:41
  This is the metallurgy in action.
• 43:43
  Welding is not just gluing metal together, it is a highly
• 43:46
  localized, rapidly cooling, continuous casting process.
• 43:51
  You are literally melting the base plates and adding a third
• 43:54
  molten alloy into the mix to fuse them.
• 43:57
  The sources outline a few specific welding process is used
• 43:59
  in vessel fabrication. Let's start with the workhorse
• 44:02
  SMAW. SMW stands for shielded metal
• 44:06
  arc welding, universally known on the shop floor as stick
• 44:08
  welding. It is the oldest and most
• 44:10
  versatile process. The welder holds a Stinger with
• 44:13
  a consumable electrode rod. The rod is a solid metal core
• 44:16
  covered in a thick coating of chemical flux.
• 44:18
  You strike an arc and the intense electrical heat upwards
• 44:21
  of 10,000° melts the metal core to provide the filler into the
• 44:25
  joint. But what is the point of the
• 44:26
  flex coating? Why not just use a bare wire?
• 44:28
  If you melt steel in the presence of the open atmosphere,
• 44:31
  the oxygen and nitrogen in the air will instantly dissolve into
• 44:34
  the molten puddle. When the puddle freezes a second
• 44:37
  later, those gases get trapped inside the solid steel, creating
• 44:41
  thousands of tiny bubbles. This is called porosity and it
• 44:45
  turns the Weld into Swiss cheese.
• 44:47
  So the flux stops that. Yes, the flux coating solves
• 44:50
  this. As the heat of the arc melts the
• 44:53
  rod, it also burns the flux. This combustion evolves a cloud
• 44:57
  of shielding gas that physically pushes the atmosphere away from
• 45:00
  the liquid puddle. Stick welding is incredibly
• 45:03
  common for field work, right? Like erecting tanks outdoors.
• 45:06
  Yeah, because that cloud of burn flux is very robust, A slight
• 45:09
  breeze outdoors won't blow it away.
• 45:11
  The downside is that the burn flux leaves a thick, hard layer
• 45:15
  of glassy slag over the top of the finished Weld bead.
• 45:18
  The welder has to manually chip and wire brush every single
• 45:21
  millimeter of that slag away before they can lay down the
• 45:24
  next pass. And if they don't?
• 45:26
  If they miss a spot and Weld over a piece of slag, it gets
• 45:29
  permanently trapped inside the joint.
• 45:31
  That is a slag inclusion, a severe physical defect that acts
• 45:35
  as a crack initiator. To avoid that, slag shops often
• 45:38
  use TIG welding, right? Yeah, gas tungsten arc welding.
• 45:42
  Right, TIG is the surgeon's scalpel.
• 45:44
  It is entirely different. The torch uses a non consumable
• 45:48
  tungsten electrode. The tungsten doesn't melt, it
• 45:51
  just carries the electric arc to heat the base metal.
• 45:54
  The welder holds a separate bare filler rod in their other hand
• 45:57
  and manually dabs it into the molten puddle.
• 46:00
  There is no flux, so. How do they shield the puddle
• 46:02
  from the air? The torch constantly blows a
• 46:05
  steady stream of pure inert gas, usually argon or helium, out of
• 46:08
  a nozzle to shield the puddle. It requires an immense amount of
• 46:11
  manual dexterity. Both hands are doing completely
• 46:14
  different things. It is the most difficult manual
• 46:16
  process to master, but it produces the highest quality,
• 46:19
  most precise Florida State clean welds.
• 46:22
  There is no slag to chip and no spatter.
• 46:25
  It is heavily used for the root pass, the very first, deepest,
• 46:28
  most critical layer of a Weld joint at the bottom of a thick
• 46:31
  plate. And it is the primary process
• 46:34
  for welding exotic, highly reactive alloys like titanium or
• 46:37
  super austenitic stainless steels.
• 46:39
  But it is very slow. And the sources also mentioned
• 46:42
  evolution of TIG called plasma welding, which honestly sounds
• 46:45
  like a sci-fi weapon. It operates on similar
• 46:47
  principles TIG using a tungsten electrode, but the physics of
• 46:50
  the arc are drastically altered. In plasma welding, the arc is
• 46:54
  constricted and forced to pass through a tiny water cooled
• 46:57
  copper nozzle. This constriction radically
• 47:00
  increases the velocity, the density and the temperature of
• 47:03
  the plasma Burk. It creates A tightly focused,
• 47:06
  incredibly intense beam of heat. What does that focused beam
• 47:09
  allow you to do? It allows for something called a
• 47:12
  keyhole effect. The plasma beam is so intense it
• 47:15
  literally punches a tiny hole completely through the entire
• 47:18
  thickness of the steel plate. As the automated torch moves
• 47:21
  forward, the molten steel flows around the edges of the hole and
• 47:24
  seamlessly closes up behind it. It allows for incredibly deep,
• 47:28
  narrow penetration on thick plates in a single pass with
• 47:32
  very minimal distortion. But regardless of whether you
• 47:35
  use stick, TIG, or plasma, the intense heat of the arc does
• 47:38
  something fundamental and permanent to the solid steel
• 47:41
  adjacent to the Weld, the heat affected zone, or HAZ.
• 47:45
  Yeah, the haz is the invisible casualty of welding.
• 47:49
  It is the narrow band of base metal immediately next to the
• 47:52
  molten puddle. This metal didn't get quite hot
• 47:55
  enough to melt, but it got extremely hot, perhaps 2000°,
• 48:00
  and then it was rapidly quenched as the massive volume of
• 48:03
  surrounding cold steel sucked the heat away.
• 48:06
  What does that rapid thermal shock do to the crystalline
• 48:09
  lattice we talked about earlier? It traumatizes it.
• 48:12
  The rapid heating and violent cooling alters the microscopic
• 48:15
  grain structure of the steel in the HAZ.
• 48:18
  It often makes that narrow band of metal significantly harder,
• 48:21
  stiffer, and vastly more brittle than the rest of the plate.
• 48:25
  It locks massive residual tensile stresses into the metal
• 48:28
  when a welded pressure vessel fails.
• 48:30
  The crack almost never starts in the Weld metal itself.
• 48:33
  It almost always originates in the brittle, highly stressed,
• 48:36
  heat affected zone just outside the Weld.
• 48:38
  Aside from the invisible metallurgical damage of the HAZ,
• 48:41
  we also have to worry about the physical visible flaws caused by
• 48:44
  the welders technique. We talked about porosity and
• 48:47
  trap slag. The sources highlight another
• 48:49
  major defect, undercut. Explain what undercut looks like
• 48:53
  and why it's so dangerous. Undercut is a classic visual
• 48:56
  defect. It happens when the welder is
• 48:58
  moving too fast. We're using too much amperage.
• 49:00
  The intense heat melts a small trench or groove into the solid
• 49:04
  base metal right at the outer edge, the toe of the Weld bead.
• 49:08
  But the welder moves on before the molten filler metal can flow
• 49:11
  in and fill that groove back up. Yes.
• 49:13
  So you are basically left with a permanent ditch carved into the
• 49:17
  base metal alongside the Weld. Yes, and it is incredibly
• 49:21
  dangerous for two distinct reasons.
• 49:23
  First, you have literally carved away parent metal.
• 49:27
  If the wall needed to be 1 inch thick to hold the pressure, the
• 49:30
  undercut has gouged it down to 0.9 inches.
• 49:33
  At that exact spot, the wall is physically thinner.
• 49:36
  And the second reason the. Sharp V shaped geometry of the
• 49:39
  undercut groove acts as a severe stress riser.
• 49:41
  It concentrates the hoop stress exactly at the thinnest most
• 49:44
  damaged point of the metal, right next to the brittle heat
• 49:47
  affected zone. It is a perfect storm for
• 49:49
  initiating a crack. So, given how incredibly
• 49:52
  sensitive the metal orgy is, and how easily a welder can
• 49:55
  accidentally carve a stress riser or trap slag, how does a
• 49:58
  manufacturer guarantee that the person holding the torch
• 50:00
  actually knows what they are doing?
• 50:02
  They can't just take a welder's word for it when lives are on
• 50:04
  the line. Absolutely not.
• 50:06
  In pressure vessel fabrication, every single spark is dictated
• 50:09
  by a mountain of highly controlled legal paperwork, and
• 50:13
  that paperwork starts with the WPS, the Weld Procedure
• 50:17
  Specification. The WPS is essentially the
• 50:20
  recipe for the Weld. It is an exhaustive recipe.
• 50:23
  A welding engineer writes the WPS and it dictates every single
• 50:27
  variable. It specifies the exact grade of
• 50:30
  base metal, the exact AWS classification of the filler
• 50:33
  rod, the precise range of amperage and voltage allowed,
• 50:36
  the travel speed, the angle of the joint preparation, the flow
• 50:39
  rate of the shielding gas. The welder is legally bound to
• 50:42
  follow it, but a. Recipe on paper doesn't prove
• 50:45
  that the resulting cake is edible.
• 50:47
  How do you prove that the engineering assumptions in the
• 50:49
  WPS actually result in a Weld that can hold 5000PSI?
• 50:53
  You prove it with violence. You generate APQRA procedure
• 50:57
  qualification record. A master welder takes the newly
• 51:00
  written WPS, goes into the shop and welds 2 pieces of heavy
• 51:04
  scrap plate together, following the recipe exactly.
• 51:08
  Then that test plate is subjected to brutal, destructive
• 51:11
  testing in the laboratory. What do they do it?
• 51:13
  They X-ray it to check for porosity.
• 51:16
  Then they cut it into strips. They put those strips into a
• 51:18
  massive hydraulic tensile tester and literally RIP the steel
• 51:22
  apart. They measure exactly how much
• 51:24
  force it took to snap it, proving it meets the minimum
• 51:27
  tensile strength. Then they take other strips and
• 51:29
  put them in a bending jig, violently folding the Weld
• 51:32
  completely in half into AU shape.
• 51:34
  This proves the Weld and the HAZ have the required ductility and
• 51:38
  won't crack under localized strain.
• 51:40
  Only if the test plate survives all that destruction does the
• 51:43
  recipe get approved. Exactly.
• 51:45
  If the PQR passes, the WPS is legally certified, and only then
• 51:50
  can a welder take a performance test using that WPS to see if
• 51:54
  their hands are steady enough to execute it.
• 51:56
  Welding is essentially localized extreme metallurgy, and
• 52:00
  sometimes controlling that metallurgy means controlling the
• 52:02
  temperature of the entire vessel before and after the welding
• 52:05
  happens. Let's talk about preheating and
• 52:08
  PWHT. Why would you take a blowtorch
• 52:11
  to a piece of steel before you start welding it?
• 52:14
  It's all about controlling the cooling rate.
• 52:16
  If you are welding a very thick 3 inch plate of carbon steel,
• 52:19
  that plate is a massive heat sink.
• 52:22
  When you strike an arc, the tiny molten puddle is surrounded by
• 52:25
  hundreds of pounds of cold steel.
• 52:27
  The moment the arc moves, the cold steel violently sucks the
• 52:30
  heat out of the puddle. That rapid quench is what
• 52:33
  creates the hard, brittle microstructures in the HAZ we
• 52:36
  talked about. So preheating softens the blow.
• 52:39
  Exactly. Using giant rosebud torches or
• 52:42
  electric heating pads, you heat the entire joint area up to say
• 52:46
  3 or 400°F before you strike the arc.
• 52:49
  Now, when the puddle freezes, the surrounding metal is already
• 52:52
  warm. The cooling rate is drastically
• 52:54
  slowed down. The metal cools gently, allowing
• 52:57
  a soft ductile grain structure to form instead of a brittle 1.
• 53:01
  It's exactly like the analogy of the frozen glass.
• 53:04
  If you pour boiling water into a frozen glass, the thermal shock
• 53:07
  shatters it. If you warm the glass in the
• 53:09
  sink first, it survives. Right, perfect analogy.
• 53:12
  And what about post Weld heat treatment?
• 53:15
  PWHT. The well he's done, the vessel
• 53:17
  is built. Why put it in an oven?
• 53:19
  Because the vessel is acting like a coiled spring, even with
• 53:22
  preheating, the localized expansion and contraction of
• 53:25
  thousands of well passes has locked massive residual tensile
• 53:29
  stresses into the joints. The vessel is fighting itself.
• 53:33
  PWHT involves rolling the entire finished vessel into a massive
• 53:37
  gas fired furnace. You slowly bring the entire
• 53:40
  structure up to a dull red heat, maybe 11 or 1200°F, and you let
• 53:45
  it soak at that temperature for hours based on the thickness of
• 53:47
  the steel. And at the temperature, the
• 53:49
  steel is soft enough to relax. Precisely.
• 53:52
  The yield strength of the steel drops so low at that temperature
• 53:55
  that the locked in residual stresses literally force the
• 53:58
  metal to microscopically creep and yield.
• 54:01
  The stresses relieve themselves, then you very slowly cool the
• 54:06
  furnace down over several days. The vessel emerges structurally
• 54:10
  relaxed, dimensionally stable and incredibly tough.
• 54:13
  Before we leave the shop floor, I want to touch on one specific
• 54:16
  detail from the sources regarding external attachments.
• 54:20
  The vessel is designed to hold pressure, but during
• 54:22
  construction you have to physically pick up 100 ton steel
• 54:26
  cylinder with a crane to erect it.
• 54:28
  You do that using heavy lifting lugs or trunnions, but you can't
• 54:31
  just Weld a massive steel hook directly to the side of a thin
• 54:34
  walled vessel, can you? Absolutely not.
• 54:37
  If you Weld a heavy trunnion directly to a one inch shell
• 54:39
  plate and then hook a crane to it and pull with 100 tons of
• 54:42
  force, the localized stress will instantly exceed the sheer
• 54:46
  strength of the shell. The crane will literally tear
• 54:48
  the trunion along with a massive chunk of the vessel wall right
• 54:52
  out of the side of the tank. So how do you distribute that
• 54:54
  massive lifting force safely into the thin skin of the
• 54:57
  vessel? You use reinforcing pads or
• 55:00
  repads. Before you attach the trunion,
• 55:03
  you Weld a large, thick circular or diamond shaped plate of steel
• 55:07
  directly to the shell. This pad is carefully sized by
• 55:11
  the engineer, usually 0.75 to 1.5 times the thickness of the
• 55:15
  shell itself. Then you Weld the massive trend
• 55:19
  into the repad. So the pad acts like a snowshoe.
• 55:22
  Exactly. It takes the massive, highly
• 55:24
  concentrated point load from the crane and spreads it smoothly
• 55:28
  over a vastly larger surface area of the thin shell membrane.
• 55:32
  And when designing that attachment, the engineer isn't
• 55:34
  just looking at the weight. They have to calculate the sheer
• 55:36
  stress on the welds holding the pad, the longitudinal bending
• 55:39
  moments trying to pry the pad off the shell, and the pinhole
• 55:42
  bearing stress inside the trunnion where the crane's
• 55:44
  shackle pin is physically crushing against the steel.
• 55:47
  It is an incredible amount of math just to pick the thing up.
• 55:50
  OK, so the welder claims the beads are flawless.
• 55:52
  The engineer hopes the REPAD math is right.
• 55:54
  The vessel looks complete, but looking at it isn't enough.
• 55:58
  Now the inspector steps in to find the invisible flaws.
• 56:02
  This brings us to the autopsy before the failure.
• 56:04
  Non destructive testing or NDT. This is the validation phase.
• 56:09
  NDT is a suite of physical and chemical interrogations designed
• 56:14
  to prove the structural integrity of the steel without
• 56:17
  actually damaging it, and the first line of defense is always
• 56:20
  visual and dimensional inspection.
• 56:22
  Visual inspection sounds simple, but the tolerances mentioned in
• 56:26
  the research are shockingly tight for something so massive.
• 56:29
  We are talking about vessels the size of a house.
• 56:31
  What is the inspector actually measuring?
• 56:32
  They're looking for geometric deviations that ruin the stress
• 56:35
  math. For instance, they check for
• 56:37
  bow, which is how far the overall vessel deviates from a
• 56:40
  perfectly straight center line. On a vessel 10 meters long over
• 56:44
  30 feet, the maximum allowable bow might only be 6 or 7mm. 6mm
• 56:50
  of sag over 30 feet of welded steel.
• 56:54
  That is incredible precision. What else are they looking at?
• 56:56
  They map the nozzle orientations.
• 56:59
  A massive pipe flange might need to connect to a specific pump on
• 57:02
  the site. The angle of that nozzle
• 57:04
  sticking out of the vessel must be within ±5mm of the CAD
• 57:08
  drawing. They put scrayed edges across
• 57:11
  the flange faces. Why the flange faces?
• 57:14
  Because if the massive steel flange isn't perfectly flat and
• 57:16
  parallel to within ±2°, the bolted gaskets won't crush
• 57:21
  evenly and the volatile chemicals will leak out into the
• 57:24
  plant. And crucially, the inspector
• 57:26
  uses a flashlight to look for heavy grinding marks.
• 57:29
  Wait, why are grinding marks dangerous?
• 57:31
  If a welder laid down a lumpy, ugly Weld, they might try to
• 57:34
  hide it by aggressively using an angle grinder to smooth it out
• 57:37
  and make it look pretty. But if they grind too deep into
• 57:40
  the base metal, they might reduce the wall thickness below
• 57:43
  the minimum T -, C requirement we discussed earlier.
• 57:46
  A pretty Overground Weld is structurally thinner and weaker
• 57:49
  than an ugly thick one. Man, that's sneaky.
• 57:52
  But visual inspection only finds what is on the surface.
• 57:55
  How do we find the microscopic surface breaking cracks that are
• 57:59
  too small for the naked eye? The sources outline a method
• 58:02
  called Dipenetrin testing, or DP.
• 58:05
  How does a chemical dye find a microscopic crack?
• 58:08
  It utilizes the physics of capillary action.
• 58:10
  It is a brilliantly simple three-step chemical process.
• 58:14
  First you thoroughly clean the Weld area with a solvent.
• 58:18
  Then you spray the Weld with a highly visible penetrating dye,
• 58:21
  usually bright neon red. This dye is formulated to have
• 58:25
  extremely low surface tension. You let it sit for 10 or 15
• 58:28
  minutes. During that time, capillary
• 58:31
  action physically pulls the red liquid deep down into any
• 58:34
  microscopic invisible cracks, breaking the surface.
• 58:36
  But right now the whole Weld is just covered in red dye.
• 58:38
  Right, so Step 2 is wiping the surface completely clean with a
• 58:42
  rag and more solvent. You remove all the dye from the
• 58:44
  surface of the metal, but you leave the dye trapped deep
• 58:47
  inside the crack. Step 3 is spraying on a
• 58:50
  developer, which is a fine white powder suspended in a fast
• 58:53
  drying solvent. And the developer acts like a
• 58:55
  sponge. Exactly.
• 58:56
  The developer creates a bright white background over the Weld,
• 58:59
  and its porous nature physically Wicks the trapped red dye back
• 59:03
  up out of the crack. The dye spreads into the white
• 59:06
  powder, painting a bright, glaring red line exactly over
• 59:11
  the location of the microscopic flaw.
• 59:13
  You can't miss it. That is incredibly effective for
• 59:15
  any metal, but if you are working with ferrous metals,
• 59:18
  carbon steel which is highly magnetic, you can use an even
• 59:21
  more advanced technique, magnetic particle inspection or
• 59:24
  MPI. How does magnetism reveal a
• 59:27
  physical crack? MPI is fascinating because it
• 59:30
  doesn't just find cracks on the surface, it can find cracks
• 59:32
  buried a few millimeters below the surface.
• 59:35
  The inspector uses heavy electromagnetic yolks to induce
• 59:38
  a massive magnetic field directly across the Weld jone.
• 59:41
  The magnetic flex lines travel easily through the solid
• 59:44
  contiguous steel. But what happens when those flux
• 59:47
  lines hit a crack? Because the crack is basically
• 59:49
  just a gap of air. Air has a massive magnetic
• 59:52
  resistance compared to steel, so when the flux lines encounter a
• 59:55
  crack they cannot easily cross the gap.
• 59:58
  Instead the magnetic field is forced to jump up and out of the
• 1:00:01
  metal, traveling through the air around the crack before re
• 1:00:04
  entering the steel. This creates what is called a
• 1:00:07
  magnetic leakage field right above the flaw.
• 1:00:10
  How do you see a magnetic leakage field though?
• 1:00:12
  While the electromagnet is energized, the inspector sprays
• 1:00:16
  the Weld with a fine mist of iron dust that has been coated
• 1:00:19
  in a fluorescent dye. The iron dust acts like
• 1:00:22
  thousands of tiny compass needles.
• 1:00:24
  They are drawn instantly and powerfully to the magnetic
• 1:00:27
  leakage field hovering over the crack.
• 1:00:29
  The inspector turns off the overhead lights, turns on an
• 1:00:31
  ultraviolet black light, and the iron dust packed into the crack
• 1:00:34
  glows a brilliant neon green, outlining the flaw perfectly.
• 1:00:38
  That's like making the invisible forces visible, but DP and MPI
• 1:00:42
  only find flaws at or very near the surface.
• 1:00:45
  What about the deep interior of the Weld?
• 1:00:47
  What if there is a massive slag inclusion buried 2 inches deep
• 1:00:50
  inside a multi pass Weld? For that we need volumetric
• 1:00:53
  testing. The sources focus heavily on
• 1:00:56
  ultrasonic testing, or UT. Yeah, UT is basically
• 1:00:59
  echolocation for steel. It uses the A scope pulse echo
• 1:01:03
  method. A highly trained technician
• 1:01:05
  applies a coupling gel to the steel and slides a piezoelectric
• 1:01:09
  transducer over the surface. The transducer fires bursts of
• 1:01:13
  incredibly high frequency sound waves down into the solid metal.
• 1:01:17
  The sound wave travels through the steel until it hits a
• 1:01:20
  boundary. Exactly.
• 1:01:22
  If the steel is solid, the sound wave travels all the way to the
• 1:01:25
  back wall of the plate, bounces off, and returns to the
• 1:01:27
  transducer. The machine displays this as a
• 1:01:30
  large spike on the screen at a specific time interval, but if
• 1:01:33
  there is a lamination hidden inside the plate or deep crack
• 1:01:37
  buried in the Weld, the sound wave hits that flaw and bounces
• 1:01:40
  back early. And because it bounced back
• 1:01:42
  early, the machine knows exactly how deep the flaw is.
• 1:01:44
  Yes, by calibrating the speed of sound and steel.
• 1:01:47
  The technician looks at the screen and says I have a
• 1:01:49
  reflection coming back exactly 1.5 inches deep.
• 1:01:53
  By sweeping the transducer back and forth, they can map the
• 1:01:55
  exact 3D geometry, depth and angle of the buried crack.
• 1:01:59
  I understand the precision of UT, but I want to push back on
• 1:02:02
  this a bit. The sources also talk about
• 1:02:04
  radiography RT, which is essentially taking an X-ray of a
• 1:02:08
  Weld. If you can take a high
• 1:02:09
  resolution permanent photograph of the inside of the Weld, why
• 1:02:13
  bother bouncing invisible sound waves around?
• 1:02:16
  Isn't a picture worth 1000 echoes?
• 1:02:18
  It definitely seems that way, but you have to understand the
• 1:02:21
  physical limitations of radiation.
• 1:02:23
  RT uses an X-ray tube or a radioactive isotope like Iridium
• 1:02:27
  192 to shoot gamma rays through the Weld onto a piece of
• 1:02:31
  photographic film taped to the backside.
• 1:02:34
  RT relies on density differences.
• 1:02:36
  It measures how much radiation gets absorbed by the steel
• 1:02:38
  versus how much passes through. To avoid in the metal, let's
• 1:02:41
  more radiation through. Right, RT is absolutely
• 1:02:43
  phenomenal at finding volumetric defect, things that take up
• 1:02:46
  physical space. A trapped bubble of gas like
• 1:02:49
  porosity or a chunk of glassy slag has much less density than
• 1:02:53
  the surrounding steel. The radiation passes easily
• 1:02:55
  through that void, hits the film, and creates a dark, easily
• 1:02:58
  visible spot on the negative for volumetric flaws, route is the
• 1:03:02
  gold standard. But I hear a butt coming.
• 1:03:04
  But Route is incredibly poor at finding tight Lanar defects like
• 1:03:09
  severe cracks. A crack is a structural failure,
• 1:03:13
  but it doesn't take up any volume.
• 1:03:14
  The two faces of the broken steel are pressed tightly
• 1:03:17
  against each other. I see.
• 1:03:20
  Yeah, if that tight crack is oriented perpendicular to the
• 1:03:23
  incoming radiation beam, the radiation just passes through
• 1:03:26
  the full thickness of the steel as if the crack wasn't there.
• 1:03:29
  There is no missing density, so the crack does not show up on
• 1:03:32
  the film at all. You could have a lethal crack
• 1:03:34
  running the length of the Weld and the X-ray would show a
• 1:03:37
  perfect clear picture. That's genuinely terrifying.
• 1:03:40
  Which is exactly why UT is so critical.
• 1:03:42
  A sound wave doesn't care about density differences, it reflects
• 1:03:45
  off any physical boundary. UT will bounce an echo off a
• 1:03:49
  tight crack regardless of its orientation.
• 1:03:51
  Furthermore, RT involves highly hazardous lethal radiation.
• 1:03:55
  When you X-ray a vessel, you have to clear the entire shop
• 1:03:58
  floor, halt all other production, and set up massive
• 1:04:01
  lead lined exclusion zones. While UT is safe.
• 1:04:03
  Yeah, UT is completely safe. You can do it while the guy next
• 1:04:06
  to you is eating lunch. Our key gives you a permanent
• 1:04:09
  picture of porosity, but UT finds the Lisel cracks and tells
• 1:04:13
  you exactly how deep they are. And the class of the vessel
• 1:04:15
  dictates how much of this expensive, time consuming
• 1:04:18
  testing you actually have to do. Precisely, if you are building a
• 1:04:21
  low pressure water storage tank, the code might only mandate
• 1:04:25
  visual inspection and maybe 10% random UT spot checks.
• 1:04:29
  The risk is low, but if you were building a class I lethal
• 1:04:32
  service vessel, say a reactor containing pressurized hydrogen
• 1:04:36
  fluoride or anhydrous ammonia where a leak means a toxic cloud
• 1:04:39
  over a city, the code will absolutely dictate 100%
• 1:04:43
  volumetric testing. Every single millimeter of every
• 1:04:46
  single pressure retaining Weld must be UT mapped and X-rayed
• 1:04:50
  all right. We've designed the architecture.
• 1:04:52
  We've rolled the steel, welded the seams, X-rayed the joints,
• 1:04:54
  and mapped the interior with ultrasound.
• 1:04:56
  The inspector has signed them out in the paperwork.
• 1:04:58
  The vessel exists in physical reality.
• 1:05:01
  Now we fill it up and try to break it.
• 1:05:02
  This is the moment of truth testing and failure modes.
• 1:05:05
  Let's start with the standard proof test, the hydrostatic
• 1:05:07
  test. The Hydra test is the final exam
• 1:05:10
  before the vessel leaves the shop.
• 1:05:11
  The procedure is conceptually simple.
• 1:05:13
  You cap all the nozzles. You fill the entire massive
• 1:05:17
  vessel completely to the brim with water.
• 1:05:19
  It is absolutely critical that you vent every single pocket of
• 1:05:23
  atmospheric air out of the highest nozzles.
• 1:05:25
  Once it is completely solid with water, you hook up a high
• 1:05:28
  pressure hydrostatic pump and you slowly squeeze more water
• 1:05:32
  into the vessel until the internal pressure reaches 1.25
• 1:05:36
  or sometimes 1.5 times the maximum allowable working
• 1:05:39
  pressure the vessel was designed for.
• 1:05:41
  The obvious question is why water?
• 1:05:44
  Why deal with thousands of gallons of heavy, messy water?
• 1:05:47
  Every machine shop has massive industrial air compressors.
• 1:05:49
  Why not just pump it full of compressed air?
• 1:05:51
  It's lighter, it's faster, it doesn't rust the steel.
• 1:05:53
  It's an incredibly important question, and the answer is pure
• 1:05:57
  physics regarding stored energy. Water is fundamentally an
• 1:06:01
  incompressible fluid. If you have a 10,000 gallon tank
• 1:06:05
  completely full of water, and you use a pump to pressurize
• 1:06:09
  that water to 500 PSI, the actual volume of the water
• 1:06:12
  barely changes at all. You are only adding a few extra
• 1:06:16
  cups of water to spike the pressure.
• 1:06:18
  So the energy isn't stored in the fluid itself.
• 1:06:20
  Exactly. The energy is stored purely in
• 1:06:22
  the slight elastic stretching of the steel walls.
• 1:06:25
  So if the vessel has a critical flaw that the X-rays missed and
• 1:06:28
  it fails during the hydro test, the outcome is dramatic but
• 1:06:31
  localized. The steel cracks, a small jet of
• 1:06:34
  water squirts out, the steel instantly relaxes back to its
• 1:06:37
  original shape, and the internal pressure instantaneously drops
• 1:06:40
  to 0. The event is over in a fraction
• 1:06:43
  of a second. The floor gets wet, but nobody
• 1:06:45
  dies. Now contrast that with the
• 1:06:47
  pneumatic test using compressed air.
• 1:06:49
  Air is a gas. It's highly compressible.
• 1:06:51
  If you fill that same 10,000 gallon tank with compressed air
• 1:06:55
  to 500 PSI, you are taking an astronomically massive volume of
• 1:06:59
  atmospheric air and violently crushing it into a small space.
• 1:07:04
  That highly compressed gas acts exactly like a massive coiled
• 1:07:08
  mechanical spring. It holds a terrifying amount of
• 1:07:11
  potential energy. And if the vessel cracks under
• 1:07:14
  pneumatic pressure? If a crack initiates, the
• 1:07:17
  pressure does not drop instantly.
• 1:07:19
  The highly compressed gas violently expands, trying to
• 1:07:22
  return to its massive atmospheric volume.
• 1:07:25
  That rapidly expanding air acts as an explosive charge.
• 1:07:28
  It catches the edges of the crack, rips the steel wide open,
• 1:07:32
  and literally blows the vessel into fragments, launching multi
• 1:07:35
  ton pieces of jagged steel shrapnel across the shop floor
• 1:07:38
  at supersonic speeds. You have effectively
• 1:07:41
  manufactured a massive bomb. Wow.
• 1:07:43
  Pneumatic testing is incredibly lethal and is avoided at all
• 1:07:46
  costs. It is only permitted if the
• 1:07:48
  vessel simply cannot tolerate water for process reasons like
• 1:07:50
  cryogenic systems, and it requires extreme safety
• 1:07:53
  barricades and 100% perfect NDT beforehand.
• 1:07:56
  But the sources do mention one specific scenario where air is
• 1:08:00
  used to test a vessel safely. Vacuum leak testing.
• 1:08:05
  How does that work without building a bomb?
• 1:08:07
  It works because the pressure differential is reversed and the
• 1:08:09
  maximum energy is capped by the atmosphere.
• 1:08:12
  For equipment like steam condensers that operate under
• 1:08:15
  vacuum, a high pressure hydro test doesn't actually prove the
• 1:08:17
  vessel won't leak air inward during operation.
• 1:08:20
  So you pull a course vacuum on the vessel, essentially sucking
• 1:08:23
  the air out. You seal the valves and you
• 1:08:26
  attach a highly sensitive pressure gauge.
• 1:08:29
  Then you just watch the clock. You measure the leak rate over
• 1:08:31
  time. You calculate the change in
• 1:08:33
  internal pressure, multiply it by the volume of the vessel, and
• 1:08:36
  divide by the time elapsed. If the vacuum holds, the seals
• 1:08:40
  and welds are tight. If the pressure slowly creeps
• 1:08:42
  up, air is leaking in through a microscopic flaw.
• 1:08:45
  OK, so we've run the hydro test, but testing is ultimately about
• 1:08:49
  surviving failure, and the sources breakdown the grim
• 1:08:52
  realities of failure into 8 distinct categories.
• 1:08:56
  I want to walk through these not as a textbook list, but to
• 1:08:59
  understand what they actually look like in reality.
• 1:09:02
  The first one we already touched on elastic deformation.
• 1:09:05
  Right elastic deformation is the geometric instability we
• 1:09:08
  discussed with the soda can and external vacuum pressure.
• 1:09:12
  The critical thing to understand here is that the metal itself
• 1:09:15
  didn't break. The atomic bonds didn't fail.
• 1:09:18
  The geometry simply buckled. The vessel deformed elastically,
• 1:09:22
  meaning if you could magically push it back out, the steel
• 1:09:25
  would still be fine. It is a failure of shape, not
• 1:09:27
  material. The second mode is terrifying
• 1:09:29
  because it happened suddenly. Brittle fracture.
• 1:09:31
  Brittle fracture is the catastrophic explosive failure
• 1:09:34
  we discussed regarding the fracture mechanics paradox.
• 1:09:37
  The crack runs at the speed of sound, and the terrifying part
• 1:09:40
  is that brittle fractures often happen during the hydro test,
• 1:09:43
  right on the shop floor. Wait, if the vessel passes the
• 1:09:46
  X-rays, why would it shatter during the water test?
• 1:09:49
  Temperature. It all comes back to the nil
• 1:09:51
  ductility transition temperature, or NDT.
• 1:09:55
  In the winter, the water pumped into the vessel from the city
• 1:09:57
  mains might be very cold, say 40°F.
• 1:10:01
  If the carbon steel aloe used has an NDT of 50°.
• 1:10:05
  Pouring that cold water in instantly transitions the steel
• 1:10:09
  from a ductile sponge into brittle glass.
• 1:10:12
  So when the pressure rises. When the pump ramps the pressure
• 1:10:15
  up, any microscopic flaw becomes the origin point for a massive,
• 1:10:19
  brittle fracture. The entire vessel shatters
• 1:10:22
  before it even reaches design pressure.
• 1:10:24
  That segways perfectly into a concept the engineers use to
• 1:10:27
  prevent that exact scenario. MDMT the minimum design metal
• 1:10:31
  temperature. How do they design around the
• 1:10:33
  cold? The MDMT is the absolute lowest
• 1:10:36
  temperature of the vessel steel ever experienced, whether from
• 1:10:38
  the fluid inside or the winter weather outside.
• 1:10:41
  To ensure the metal remains ductile at that freezing
• 1:10:43
  temperature, the code typically requires the foundry to perform
• 1:10:46
  expensive sharp EV notch impact testing on the raw steel plates
• 1:10:50
  at that specific temperature. They literally swing a heavy
• 1:10:53
  pendulum at a notch piece of the steel to see how much energy it
• 1:10:56
  takes to snap it in the cold. Exactly.
• 1:10:58
  But impact testing cost money and delays production.
• 1:11:01
  Engineers hate paying for it, so the code offers a mathematical
• 1:11:04
  loophole. You can avoid impact testing if
• 1:11:07
  you dramatically lower the stress in the metal.
• 1:11:10
  And how do you lower the stress? You artificially increase the
• 1:11:13
  thickness of the component. If the wall is vastly thicker
• 1:11:16
  than it needs to be, the operating stress is so low that
• 1:11:19
  the risk of a crack propagating is minimized.
• 1:11:22
  Even if the metal is cold and slightly brittle.
• 1:11:24
  The code allows you to buy your way out of impact testing by
• 1:11:28
  buying vastly thicker steel. The third failure mode is
• 1:11:31
  excessive plastic deformation. This is the classic cartoon
• 1:11:34
  failure. Right.
• 1:11:35
  The internal pressure vastly exceeds the yield strength of
• 1:11:38
  the steel. The metal physically stretches,
• 1:11:40
  yielding plastically. The vessel swells up, pull just
• 1:11:43
  outward like a balloon. The wall thins out and
• 1:11:45
  eventually it just tears. Open No.
• 1:11:47
  The 4th mode sounds insidious. Stress rupture or creep.
• 1:11:51
  What is creep? It sounds slow.
• 1:11:52
  It is agonizingly slow and it only happens at high
• 1:11:55
  temperatures. Think of a vessel operating at
• 1:11:58
  1000°F for 20 years. The internal pressure is well
• 1:12:02
  below the yield strength. The math says it's perfectly
• 1:12:04
  safe, but at those extreme temperatures the atomic lattice
• 1:12:08
  is vibrating violently. So the metal itself is moving.
• 1:12:11
  Over years of constant tension, the microscopic grain boundaries
• 1:12:15
  inside the steel literally begin to slide past each other.
• 1:12:19
  The solid steel, slowly imperceptible, stretches like
• 1:12:23
  warm taffy. The vessel gets permanently
• 1:12:26
  longer and wider over decades until the wall thins out so much
• 1:12:29
  it finally ruptures. That requires constant vigilance
• 1:12:32
  over decades. The 5th mode is plastic
• 1:12:35
  instability. That sounds like a fatigue
• 1:12:37
  issue. It is incremental collapse.
• 1:12:38
  It's not a single massive overpressure event.
• 1:12:41
  It's cyclical. Imagine a vessel that cycles
• 1:12:43
  from 0PSI to Max PSI every single day.
• 1:12:46
  Due to complex geometry around a nozzle, there is a tiny
• 1:12:50
  localized area where the stress exceeds the yield point just a
• 1:12:53
  tiny bit during every cycle. Day one, the metal stretches A
• 1:12:56
  microscopic fraction of a millimeter.
• 1:12:58
  And day 2 day. Two, it stretches another
• 1:13:01
  fraction. It never springs back.
• 1:13:03
  Over thousands of cycles, this localized plastic deformation
• 1:13:07
  accumulates, ratcheting up until that section of the vessel
• 1:13:10
  becomes completely unstable and tears.
• 1:13:12
  Which is closely related to the 6th mode high strain, low cycle
• 1:13:16
  fatigue. Yes, but this applies to highly
• 1:13:19
  ductile material subjected to massive extreme strain
• 1:13:23
  amplitudes. Instead of taking thousands of
• 1:13:26
  small cycles, the material undergoes massive stretching and
• 1:13:29
  compressing. The ductile metal absorbed it
• 1:13:32
  for a while, but the severe strain cares the crystal lattice
• 1:13:35
  apart, leading to massive fatigue cracking after
• 1:13:37
  relatively few cycles. Now we get into the chemistry.
• 1:13:40
  The 7th mode is stress corrosion cracking.
• 1:13:44
  This sounds like an unfair fight.
• 1:13:45
  It is the ultimate ambush. Stress corrosion cracking
• 1:13:48
  requires 3 specific conditions to exist simultaneously.
• 1:13:51
  You need a susceptible material, you need an active tensile
• 1:13:54
  stress pulling on that material, and you need a highly specific
• 1:13:57
  corrosive environment. The terrifying part is that the
• 1:13:59
  corrosive environment might barely affect the metal if there
• 1:14:02
  was no stress applied. Give me an example of that
• 1:14:04
  chemical lock and key mandanism. Chlorides and austinitic
• 1:14:07
  stainless steel. You can drop a piece of three O
• 1:14:10
  4 stainless steel in warm chloride water and it will sit
• 1:14:13
  there relatively happily. But if you take that same piece
• 1:14:17
  of stainless, bend it into AU shape to create massive tensile
• 1:14:21
  stress, and then put it in the chloride water, the chlorides
• 1:14:24
  will aggressively attack the grain boundaries, specifically
• 1:14:27
  where the tension is highest. It zeroes in on the stress.
• 1:14:30
  The chemical attack literally unzips the atomic bonds pulling
• 1:14:34
  against each other. A thick plate of stainless steel
• 1:14:37
  will crack completely in half in a matter of hours.
• 1:14:40
  Caustic environments do the exact same thing to carbon
• 1:14:43
  steels. And the final eighth mode is the
• 1:14:45
  absolute worst of both worlds. You take the mechanical trauma
• 1:14:49
  of. Cyclic fatigue loading the
• 1:14:50
  constant stretching and relaxing, and you drop the
• 1:14:52
  vessel into a corrosive environment.
• 1:14:55
  The two mechanisms accelerate each other logarithmically.
• 1:14:58
  The corrosion eats microscopic pits into the surface, and those
• 1:15:01
  pits become stress risers. Those pits.
• 1:15:03
  Act as massive stress risers. The cyclic fatigue load
• 1:15:07
  concentrates on those pits, rapidly tearing cracked open.
• 1:15:10
  The corrosive fluid immediately rushes deep into the fresh,
• 1:15:13
  unprotected crack, eating away the tip and making it sharper.
• 1:15:16
  It's a self feeding loop of destruction that drastically,
• 1:15:19
  unpredictably slashes the lifespan of the vessel.
• 1:15:22
  I want to pull all of this together and contextualize the.
• 1:15:25
  Hydro test based on these 8 Horrifying ways a Vessel can die
• 1:15:29
  The expert analysis of the Hydra test from the sources is the
• 1:15:33
  most sobering reality check of this entire deep dive.
• 1:15:36
  It is because clients and even young engineers.
• 1:15:39
  Look at a successful Hydra test and they celebrate.
• 1:15:42
  They think the vessel held 1.5 times its design pressure.
• 1:15:45
  It didn't leak. It is perfectly safe, but it's
• 1:15:48
  an illusion of absolute safety. Exactly the.
• 1:15:50
  Hydra test is a very limited. Static proof.
• 1:15:54
  All a successful hydro test proves is that on that specific
• 1:15:57
  Tuesday afternoon, sitting on the shop floor at ambient
• 1:16:00
  temperature, the vessel possessed the gross structural
• 1:16:03
  integrity to resist the principal static membrane
• 1:16:06
  stresses of an over pressure event.
• 1:16:09
  It proves the welder didn't leave a massive hole in a seam.
• 1:16:12
  It proves the plates are generally thick enough.
• 1:16:14
  But it proves absolutely nothing about the future.
• 1:16:17
  Nothing. A successful hydro test does not
• 1:16:19
  prove the. Vessel will survive 30 years of
• 1:16:21
  cyclic pressure fatigue. It does not prove the steel
• 1:16:24
  won't slowly succumb to high temperature creep over a decade.
• 1:16:28
  It certainly does not prove the vessel won't be torn apart by
• 1:16:30
  stress corrosion cracking next month when the chlorides are
• 1:16:33
  introduced. It is a baseline static test.
• 1:16:36
  It is not a guarantee of operational lifespan.
• 1:16:39
  I look at it this way, a hydra test is like a one Rep.
• 1:16:41
  Max deadlift at the gym. Just because you can load the
• 1:16:46
  bar up and pick that massive weight off the floor once
• 1:16:48
  without breaking your back, it absolutely does not mean your
• 1:16:51
  spine won't slowly degenerate and fail if you try to lift a
• 1:16:55
  slightly lighter weight every single day for a decade.
• 1:16:58
  Especially if you're standing in an acidic swamp while you do it.
• 1:17:01
  That is precisely the reality the Hydra test proves you.
• 1:17:04
  Survived day one, the material selection, the pitting index,
• 1:17:07
  the fatigue analysis, the corrosion allowances and the
• 1:17:10
  flawless execution of the WPS by the welder.
• 1:17:13
  That is what gets you to year 30.
• 1:17:15
  Which leaves us with a final, slightly provocative.
• 1:17:18
  Thought to ponder. We spent this entire time
• 1:17:21
  talking about the incredible precision of modern engineering.
• 1:17:24
  Today we have finite element analysis, CAD software that can
• 1:17:27
  model stress gradients down to the millimeter.
• 1:17:29
  We have laser guided plasma welding.
• 1:17:32
  We have ultrasonic phased array mapping.
• 1:17:34
  They can see a microscopic flaw buried inside 3 inches of solid
• 1:17:38
  steel. Yeah, we have unparalleled
• 1:17:39
  visibility into the physics of the material.
• 1:17:42
  But back in the day, the engineers who built the 1st.
• 1:17:44
  Massive industrial boilers and chemical columns didn't have
• 1:17:48
  these tools. They couldn't calculate the
• 1:17:50
  exact localized sheer stress around a nozzle.
• 1:17:54
  So what do they do? They over engineered everything
• 1:17:56
  they built in massive ignorance. So they built incredibly heavy.
• 1:18:00
  They use incredibly thick plates and giant sweeping highly
• 1:18:04
  conservative safety factors. They build battleships to hold
• 1:18:07
  steam. Exactly.
• 1:18:08
  So here's the question for the modern designer listening.
• 1:18:10
  To this are highly optimized, mathematically perfect and
• 1:18:14
  vastly thinner modern designs actually leaving us less margin
• 1:18:18
  for error on the shop floor. If a modern vessel is designed
• 1:18:21
  right to the absolute ragged edge of the yield strength just
• 1:18:24
  to save weight because the computer simulation says it is
• 1:18:27
  technically safe, what happens when reality sets in?
• 1:18:30
  That's the terrifying part. Does a single moment of human.
• 1:18:33
  Error by a tired. Welder or a slight misalignment
• 1:18:36
  on the rollers instantly become a catastrophic event because the
• 1:18:40
  safety net of extra thickness is gone.
• 1:18:43
  Does our advanced technology make us fundamentally safer, or
• 1:18:46
  has it just made us more arrogant, pushing the materials
• 1:18:49
  closer and closer to the edge of disaster?
• 1:18:52
  Something for you to ponder the next time you strike an arc or
• 1:18:55
  finalize a CAD drawing. The margins are undeniably
• 1:18:58
  tighter today. The precision is higher, the
• 1:19:01
  efficiency is beautiful, but the safety net is much, much
• 1:19:04
  thinner. Your GPA won't save you when the
• 1:19:06
  part fails only. Your scars will thank you for
• 1:19:09
  listening. Don't forget, if this helped you
• 1:19:11
  even a little, please tell us. We're putting this out for free.
• 1:19:14
  No paywall, no subscription, nothing locked behind a card
• 1:19:17
  that only works if we know it's actually reaching and teaching.
• 1:19:20
  If you're on a journey trying to figure this stuff out and this
• 1:19:22
  gave you clarity you weren't getting anywhere else, drop a
• 1:19:24
  like, leave a comment. It takes 2 seconds on your end,
• 1:19:26
  but it tells us this is worth continuing.
• 1:19:28
  Also, don't forget to check out our YouTube.