Latest / Mechanical Engineering Made Simple / (#108) Pressure Vessel Design Codes and Stored Energy

Transcript

(#108) Pressure Vessel Design Codes and Stored Energy · Mechanical Engineering Made Simple

• 0:00
  Welcome back to mechanical engineering made simple.
• 0:02
  It is good to be back. Today we are looking at
• 0:04
  something that honestly, it surrounds us every single day,
• 0:09
  but we mostly take for granted, you know, until something goes
• 0:13
  wrong, right? We're looking at pressure
• 0:14
  equipment, and I think the best place to start isn't with some
• 0:18
  dry technical definition, but with the reality of what these
• 0:22
  things are. Because when you really look at
• 0:24
  a pressure vessel, whether it's a massive boiler or just a
• 0:28
  simple air receiver in a workshop, what you're actually
• 0:31
  looking at is a container of immense stored energy.
• 0:34
  That is exactly right. It's potential energy just
• 0:37
  waiting to get out. And if that energy gets out
• 0:39
  uncontrolled, it's not just a leak, it's it's basically a
• 0:42
  bomb. It is.
• 0:43
  And that's why the engineering behind this is so, so rigorous.
• 0:47
  I mean, it's not just about welding some steel plates
• 0:49
  together. We're talking about a discipline
• 0:51
  that has to manage these huge hazards that, well, they keep
• 0:55
  the modern world running. And what's so fascinating about
• 0:58
  the source material we have for this deep dive is that it
• 1:00
  doesn't just stop at the design. No, it goes way beyond that.
• 1:03
  It goes into the hidden complexity of actually building
• 1:07
  these things, inspecting them, and this is the part people
• 1:10
  always forget moving them. That part just blew my mind, The
• 1:13
  idea of lifting A50 ton steel cylinder without, you know,
• 1:16
  snapping it in half. Exactly.
• 1:18
  The logistics are incredible. So we have a huge stack of
• 1:21
  sources today. We're covering everything from
• 1:23
  the alphabet soup of design codes.
• 1:26
  DSME PD 5500. To the physics of stress and the
• 1:31
  detective work of inspection. It's a lot, but we're going to
• 1:34
  break it all down. Let's dive in.
• 1:36
  OK, let's unpack the basics first.
• 1:37
  Yeah. So Segment 1 design principles
• 1:40
  and the code landscape, the sources define pressure
• 1:44
  equipment and it's it's broader than I thought.
• 1:47
  It's not just the big tanks. No, not at all.
• 1:49
  It covers a huge range. You have the vessels themselves,
• 1:52
  sure, the containers, but you also have simple receivers like
• 1:55
  you said, those air tanks in a garage.
• 1:57
  And it also includes all the piping that connects everything.
• 2:00
  And while the you know the regulations might change
• 2:02
  depending on if you're in the US or Europe, the physics, the
• 2:05
  physics never changes. Pressure is pressure and that
• 2:08
  brings us to the regulations. The text mentions 2 heavy
• 2:12
  hitters that seem to well, dominate the industry.
• 2:15
  We have ASME 8. That's ASME Section 8.
• 2:19
  That's the American standard. It's very widely used all over
• 2:22
  the world. And then there's PD 5500.
• 2:25
  Right, which is the British and European standard.
• 2:27
  Now here's where it gets really interesting.
• 2:29
  For anyone who cares about how engineering actually works in
• 2:32
  the real world, the sources discuss a concept called Code
• 2:36
  Intent. Yes, this caught my eye.
• 2:38
  It sounds a bit like a loophole, but I don't think it is.
• 2:41
  What exactly is code intent versus actual compliance?
• 2:45
  It's a really critical distinction.
• 2:47
  So full code compliance means you followed the rule book to
• 2:50
  the absolute letter. Every calculation, every
• 2:52
  material test, every administrative step.
• 2:53
  And you get the official stamp. Crucially, yes, you've had the
• 2:57
  independent inspections required to get that official use stamp
• 3:00
  or whatever it is on the vessel. But sometimes a client doesn't
• 3:03
  need the stamp. Why?
• 3:04
  Well, maybe they're in a jurisdiction that doesn't
• 3:06
  strictly require it for that specific use, or they're just
• 3:10
  trying to save on the administrative costs so they ask
• 3:14
  for a vessel built to code intent.
• 3:17
  So you build it as if you were getting the stamp, but you don't
• 3:20
  do the final paperwork. Precisely.
• 3:22
  You use the exact same design formulas, the same safety
• 3:25
  factors, the same stress limits. Physically, the vessel is just
• 3:28
  as strong, but you skip some of those administrative hurdles.
• 3:32
  The third party witness the stamping itself.
• 3:35
  It's like buying a supercar without the badge on the hood.
• 3:37
  It drives the same, but you can't prove it to anyone.
• 3:40
  That's a great analogy, but as we'll see later, that little
• 3:43
  decision carries a whole lot of risk when it comes to liability.
• 3:46
  OK, let's get into the math. The sources spend a lot of time
• 3:50
  on the thin cylinder equation. This is the real bread and
• 3:53
  butter of vessel design, isn't it?
• 3:54
  It is if you just imagine a soda can under pressure.
• 3:58
  The metal wants to tear apart in two different directions.
• 4:01
  It wants to split down the side like unzipping a jacket, and it
• 4:05
  wants to pop the top and bottom off.
• 4:07
  And the text makes a huge point that one of these stresses is
• 4:11
  way, way worse than the other. Yes.
• 4:13
  The stress trying to split the cylinder down the side is called
• 4:16
  the circumferential or hoop stress.
• 4:20
  The stress trying to pop the ends off is the longitudinal
• 4:22
  stress, and the math shows us that hoop stress is always twice
• 4:26
  the magnitude of the longitudinal stress.
• 4:29
  So if you're designed for the big one, the hoop stress, the
• 4:32
  other one is automatically taken care of.
• 4:33
  Correct. That's why hoop stress is always
• 4:35
  the controlling factor. I want to look at the formula in
• 4:38
  the source because it really boils it down For circumfrontal
• 4:41
  stress. It reads S sub C = P * D / 2 *
• 4:47
  T. Simple, elegant, and absolutely
• 4:50
  critical. S sub C is your stress, P is the
• 4:53
  internal pressure, D is the diameter, and T is the wall
• 4:57
  thickness. You can see right away.
• 4:58
  If you double the pressure, you double the stress.
• 5:00
  And if you double the thickness, you got the stress in half.
• 5:02
  Exactly. But there's another variable in
• 5:05
  the full version of that equation that I found really
• 5:07
  interesting, the symbol ETA. It represents joint efficiency.
• 5:12
  Right, this is where the human element comes in.
• 5:15
  It is. You're welding steel plates
• 5:17
  together. Is that Weld absolutely perfect?
• 5:20
  I'm. Going to guess no.
• 5:21
  Probably not. So if you don't inspect that
• 5:24
  Weld fully, say if you don't use X-ray or ultrasonic testing on
• 5:28
  the entire link, you can't assume it's 100% strong, so the
• 5:32
  code penalizes you. It makes you use a lower
• 5:35
  efficiency factor, say 0.7 instead of 1.0.
• 5:40
  So if you're ETA is 0.7, the math says you're allowable
• 5:44
  stress goes down, which means which.
• 5:45
  Means your T, your wall thickness has to go up, you have
• 5:48
  to use more steel just to compensate for the fact that you
• 5:51
  didn't inspect the Weld as thoroughly.
• 5:53
  It's a direct trade off between inspection costs and material
• 5:56
  costs. That is a fantastic insight.
• 5:58
  Speaking of thickness, there was one other design factor that I
• 6:01
  found kind of poetic the corrosion allowance noted as C
• 6:05
  Ah. Yes, designing for the Crave.
• 6:07
  That's exactly what it felt like, the source says.
• 6:09
  Designers calculate the thickness based on the end of
• 6:11
  the vessel's life, not the beginning.
• 6:13
  It's the only Safeway to do it. If you know the contents are
• 6:16
  corrosive, you know, maybe it's an air receiver that's going to
• 6:19
  have condensation inside. You have to assume the metal
• 6:22
  will get thinner over 20 years. So if your calculation says you
• 6:26
  need a tube bore of, say, D sub I, you actually design for D sub
• 6:31
  I + 2 * C You're literally adding sacrificial metal that
• 6:35
  you fully expect to just vanish over time.
• 6:38
  So when a brand new vessel shows up, it's technically over built,
• 6:42
  but only because it's preparing for its own decay.
• 6:44
  That's it exactly. We've designed this thing.
• 6:47
  It's thick. It is a corrosion allowance.
• 6:48
  The math works. Now we have to move it and the
• 6:51
  source material on erection was really surprising.
• 6:54
  I kind of think of the engineering ending when the
• 6:56
  thing leaves the factory. And that is a very common
• 7:00
  mistake and a very expensive one.
• 7:02
  The source explicitly mentions that the erection costs alone
• 7:06
  for a major vessel can run into millions.
• 7:08
  Millions. And here's the kicker, if the
• 7:10
  designer didn't put the lifting attachments on the original
• 7:13
  drawing and they have to be added after the purchase order.
• 7:16
  Oh, that's an expensive extra. A massive extra and it's risky
• 7:20
  because now you're out in the field welding onto a finished
• 7:23
  heat treated vessel. You don't want to do that.
• 7:25
  The mechanics of the lift described in the text are just
• 7:29
  intense. I mean taking a vessel that's
• 7:31
  lying flat on a truck and standing it up vertically.
• 7:34
  The transition. That is the single most
• 7:37
  dangerous moment. When the vessel is horizontal,
• 7:40
  gravity is pulling it one way. As you lift it, the loads shift
• 7:44
  completely. They talk about tailing lugs.
• 7:47
  Can you walk me through what's happening there?
• 7:48
  Sure. So imagine you have a big crane
• 7:51
  picking up the top of the vessel.
• 7:52
  That's your main lift. But you can't just drag the
• 7:54
  bottom of it along the ground. You need another crane or some
• 7:57
  kind of tailing device holding the bottom end.
• 8:00
  As the main crane lifts, the vessel pivots and the load on
• 8:03
  that tailing lug, let's call it T, It changes constantly as the
• 8:07
  angle changes. And there's a formula for that
• 8:09
  tailing load in the text. It read T = W sub L * L two
• 8:13
  divided by L1 right? W sub L is the lift weight and
• 8:17
  L1 and L2 are just the distances from the pivot point.
• 8:21
  It's a basic lever arm calculation, but what's critical
• 8:24
  is this. At 0°, when it's flat, that
• 8:28
  tailing lug is carrying a huge part of the weight.
• 8:31
  OK. But at 90° fully vertical, it's
• 8:34
  carrying zero because the vessel is just hanging from the top
• 8:37
  crane? It's the journey between those
• 8:39
  two points that's the problem. That's where things get tricky.
• 8:42
  The text also mentions trunions for lifting.
• 8:45
  Trunions are those those big round ear like things you see
• 8:50
  welded on the side of a vessel. And the source breaks them down
• 8:52
  into different types. Type 1.
• 8:54
  Type 2. And that really matters because
• 8:56
  of the stress. When the vessel is horizontal,
• 8:59
  the weight is pulling down on those trunnions and it's trying
• 9:01
  to twist the shell. It's a circumferential moment.
• 9:04
  But as soon as you stand it up, the weight pulls along the
• 9:07
  vessel's length, causing a longitudinal moment.
• 9:09
  The stress changes direction completely.
• 9:12
  So the steel shell is being twisted one way, and then slowly
• 9:16
  twisted another way, all while it's dangling 100 feet in the
• 9:19
  air. And that is why the source is so
• 9:22
  big. On the worst case scenario
• 9:23
  check, you don't just check 0° and 90°, you have to check all
• 9:27
  the angles in between, because at say 45° you might have a
• 9:32
  combination of those stresses that's actually worse than
• 9:34
  either the start or the finish. It really hammers home that
• 9:37
  static equipment is only static once it's bolted down.
• 9:40
  Until then, it's a dynamic, moving and potentially lethal
• 9:44
  load if you get that math wrong. Moving on to Segment 3,
• 9:48
  manufacturer inspection an the human element.
• 9:51
  We've got the design, we've got the lifting plan.
• 9:53
  Now we're on the shop floor. The first thing the sources
• 9:56
  emphasizes is traceability. This is absolutely huge in our
• 10:02
  industry. You can't just go to a scrap
• 10:03
  yard, find a piece of steel and build a pressure vessel with it.
• 10:06
  And. Chain of traceability, yes.
• 10:07
  Yes, every single plate, every flange, every bit of pipe has to
• 10:11
  have a birth certificate. The source mentions the EN 3:50
• 10:16
  certificates, specifically 3.1 A and three-point 1B.
• 10:19
  What's the difference there? It's about who certifies it.
• 10:22
  A three-point 1B is a certificate from the
• 10:25
  manufacturer themselves. A 3.1 A usually means an
• 10:28
  independent inspector had to be there to witness the tests.
• 10:31
  But the point is, if a vessel fails 10 years from now,
• 10:35
  investigators need to be able to pick up a fragment of steel and
• 10:38
  trace it all the way back to the specific batch at the mill where
• 10:42
  it was first cast. That's incredible
• 10:44
  accountability. Now let's talk visual
• 10:46
  inspection. Source Five had this amazing
• 10:49
  list of quirks that inspectors look for.
• 10:52
  It wasn't just about measuring things, it was looking for signs
• 10:54
  of struggle. Signs of struggle is a great way
• 10:57
  to put it. One of them was bulging.
• 10:58
  Bulging tells a story. It almost always means the
• 11:01
  manufacturer tried to Weld ahead the end cap to the shell and
• 11:05
  they didn't fit up perfectly. Maybe the shell was a little bit
• 11:08
  Oval instead of perfectly round, so they.
• 11:10
  Just forced it they. Forced it.
• 11:11
  They clamped it down hard, tack welded it and literally forced
• 11:15
  the steel to align. But the steel remember numbers.
• 11:18
  It leaves a bulge, and that bulge is a stress concentration
• 11:22
  point just waiting to fail. Another one was nozzle
• 11:25
  orientation checking if nozzles have pulled out of place.
• 11:29
  When you Weld your melting steel, when it cools it shrinks.
• 11:33
  So if you have a big nozzle welded onto the side of a tank,
• 11:37
  that welding process can physically pull the nozzle out
• 11:40
  of alignment. If an inspector sees a flange
• 11:43
  pointing slightly the wrong way, it's a huge red flag that the
• 11:47
  welding or the heat treatment wasn't managed properly.
• 11:50
  And grinding marks. The text was very specific about
• 11:53
  deep grinding marks. Oh this is a pet peeve of every
• 11:56
  single inspector. A fabricator makes a messy Weld,
• 11:59
  so they get a grinder out to make it look pretty, but they
• 12:02
  get a bit too enthusiastic and grind right into the base metal.
• 12:06
  The rule of thumb is if you grind deeper than 10% of the
• 12:09
  plate thickness, you've compromised the vessel.
• 12:12
  You've created a thin spot. It's like trying to erase a
• 12:15
  mistake on paper and you just tear a hole right through.
• 12:16
  It that's a perfect analogy. There was also a note about
• 12:20
  datum lines. I always just assumed you'd
• 12:22
  measure from the Weld seam. No, never measure from the Weld
• 12:25
  seam. A Weld is rough, it's irregular,
• 12:28
  and like we just said, it might have moved because of shrinkage.
• 12:32
  You establish A datum line, a fixed reference line that's
• 12:34
  physically punched into the steel, and you measure
• 12:37
  everything from there. That is your single source of
• 12:40
  truth. OK, the final Test is the big
• 12:42
  moment. The source contrasts hydrostatic
• 12:46
  testing with pneumatic testing. The standard versus the special
• 12:50
  and dangerous case, right? Hydrostatic uses water,
• 12:53
  pneumatic uses air or gas, and the source gives this massive
• 12:57
  warning about. Still just takes right back to
• 12:59
  what you said in the intro. Stored energy water is basically
• 13:03
  incompressible. If you fill a tank with water,
• 13:05
  pump it up to high pressure and the tank cracks, what happens?
• 13:08
  Little squirt of water. A squirt.
• 13:09
  That's it. The pressure drops instantly
• 13:11
  because the water doesn't expand.
• 13:12
  It's relatively safe. Now do that with compressed air.
• 13:15
  You're compressing a gas. You are packing energy into that
• 13:17
  space. So if the tank fails then.
• 13:19
  It doesn't just squirt, it explodes.
• 13:21
  The air expands violently. It can level a building.
• 13:23
  So why would anyone ever do a pneumatic test?
• 13:26
  Sometimes you have to. Maybe the vessel is for a
• 13:29
  chemical process where even a single drop of water residue
• 13:32
  would ruin the product. Or maybe the vessel is so large
• 13:35
  and the structure so light that the sheer weight of the water
• 13:38
  would collapse it. I didn't think of that.
• 13:41
  Water is heavy. Very.
• 13:43
  So if you have to do a pneumatic test, the source says you need a
• 13:46
  much much higher level of NDT non destructive testing
• 13:49
  beforehand. You need a 100% surface crack
• 13:52
  detection. You have to be absolutely sure
• 13:55
  there are no cracks before you turn that vessel into a bomb.
• 13:58
  That makes total sense. OK.
• 13:59
  Let's hit to our last segment and talk about the hardware
• 14:01
  attached to these things, valves and fittings, specifically the
• 14:05
  safety valve. The final line of defense if all
• 14:09
  else fails, if the control system dies, if the operator
• 14:11
  makes a mistake, the safety valve is the only thing stopping
• 14:14
  a catastrophe. The terminology here was really
• 14:17
  fun. I want to ask you about these 3
• 14:18
  words. Simmer chatter and pop.
• 14:21
  OK, simmer, is that sound you hear just before the valve fully
• 14:25
  opens? It's the audible escape of a
• 14:28
  tiny bit of fluid. The pressure is pushing up on
• 14:31
  the disk, just barely overcoming the spring.
• 14:34
  It's the warning shot. Like a tea kettle starting to
• 14:36
  whistle. That's a great way to think of
• 14:38
  it. Then you have O pop is what you
• 14:41
  want. It's the rapid, decisive full
• 14:43
  opening of the valve to release a huge amount of pressure
• 14:46
  quickly. It literally snaps open.
• 14:48
  In Chatter. Chatter is very, very bad.
• 14:51
  Chatter is when the valve opens and closes rapidly.
• 14:53
  Bang Bang, Bang Bang. It's a violent reciprocating
• 14:57
  motion where the disc smashes into the seat over and over.
• 15:00
  That sounds destructive. It destroys the valve.
• 15:02
  See, it can damage the piping. If you hear chatter, something
• 15:05
  is seriously wrong with the sizing or the installation.
• 15:07
  We also touched on flanges. The source mentions ANSIB 16.5.
• 15:12
  That's the Bible for flanges. It tells you that you can't just
• 15:15
  pick any old size. The dimensions, the thickness,
• 15:18
  the bulk pattern are strictly dictated by the pressure and the
• 15:21
  temperature it has to handle. It all comes back to the code
• 15:24
  doesn't. It always, every time.
• 15:26
  So let's bring this all together.
• 15:28
  What's the big take away? We've seen that a pressure
• 15:30
  vessel isn't just a tank, it's a marriage of strict code
• 15:34
  compliance, some really complex stress math, you know, that hoop
• 15:38
  stress being double the longitudinal, and some serious
• 15:41
  logistical planning just to get it standing up.
• 15:44
  And it's a discipline where good enough is never, ever good
• 15:46
  enough. You're fighting physics, and
• 15:48
  physics always wins in the end. The sources really drove home
• 15:52
  that idea of the life cycle. You have to design for the
• 15:55
  corrosion at the end of its life.
• 15:57
  You inspect for the mistakes made during manufacturing.
• 16:01
  You lift it knowing that gravity wants to twist it apart.
• 16:04
  And this brings us right back to that code intent idea we started
• 16:07
  with. I want to leave the listener of
• 16:08
  the final thought here. Go on.
• 16:10
  We talked about building to code intent, but skipping the
• 16:12
  official stamp to save a bit of money or paperwork.
• 16:15
  But just think about the liability if you have a vessel
• 16:18
  full of compressed gas, all that stored energy and it fails.
• 16:22
  Who's responsible? Exactly.
• 16:23
  If you have the stamp, you have the paper trail, you have proof
• 16:26
  of independent verification. If you only have code intent,
• 16:29
  you are essentially just saying, trust me, I did the math right.
• 16:33
  In a true safety culture, that is a really dangerous place to
• 16:36
  be. It raises the question, is the
• 16:39
  money you save on that stamp really worth the liability you
• 16:43
  carry for the entire life of that equipment?
• 16:46
  That is a very sobering thought. It's a question of safety
• 16:48
  culture versus just checking boxes.
• 16:51
  It is. Well, on that note, we're going
• 16:53
  to wrap up this deep dive into pressure equipment.
• 16:56
  It's a heavy topic, quite literally, but an absolutely
• 16:59
  essential one to understand. Absolutely.
• 17:01
  If you haven't already, get over to our YouTube channel
• 17:03
  Mechanical Engineering Made Simple, where we post hands on
• 17:06
  how to videos, visual breakdowns, and practical
• 17:09
  engineering walkthroughs. You can actually use a family
• 17:12
  that solves together, stays together.