Latest / Mechanical Engineering Made Simple / (#55) Fighting Liquid Wood Physics in Pulp Mills

Transcript

(#55) Fighting Liquid Wood Physics in Pulp Mills · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design junkies, to mechanical engineering made
• 0:03
  simple. Today we are taking a deep dive
• 0:05
  into an industry that you know most people probably associate
• 0:08
  with forestry and just basic manufacturing pulp and paper,
• 0:12
  right? But we are skipping the trees
• 0:14
  and the new Sprint to look directly at the hidden hardcore
• 0:17
  mechanical engineering required to process what is essentially
• 0:21
  liquid wood. That is the perfect way to put
• 0:24
  it. I mean you the listener probably
• 0:26
  interact with this industry multiple times a day.
• 0:28
  Think packaging, hygiene products, even special
• 0:31
  industrial filter. Oh, absolutely.
• 0:32
  This is a massive global operation.
• 0:34
  In the year 2000 alone, our sources indicate the industry
• 0:38
  produced 188,000,000 metric tons of pulp.
• 0:42
  And that all has to be mixed. But here's the thing that our
• 0:45
  sources really highlight managing this liquid wood fiber
• 0:48
  is it's just vastly more complicated than mixing water.
• 0:52
  It's a different universe. We're talking about engineers
• 0:54
  constantly fighting against some really strange physics, like
• 0:57
  overcoming yield stress and avoiding catastrophic shaft
• 1:00
  vibration just to keep a mill online.
• 1:03
  OK, that sounds less like a simple mixing problem and more
• 1:06
  like a battle against rheology. Let's let's unpack the physics
• 1:11
  of this wet fiber. To start, you really need the
• 1:14
  chemical context pulp processing.
• 1:17
  It often begins with what's called the craft process.
• 1:19
  And the goal there is to separate the wood fibers from
• 1:21
  the lignan, right? The stuff that holds the wood
• 1:23
  together. Exactly.
• 1:24
  And that separation requires cooking the wood chips in white
• 1:27
  liquor, which is this alkaline solution of sodium sulfide and
• 1:31
  sodium hydroxide. And once the fibers are
• 1:33
  liberated, you're left with this spent cooking liquid black.
• 1:38
  Liquor, you got it. And mixing is critical there,
• 1:40
  too. For environmental and safety
• 1:42
  reasons, you need to oxidize that black liquor before
• 1:44
  chemical recovery, which prevents the release of, you
• 1:47
  know, toxic hydrogen sulfide gas.
• 1:49
  So right from the get go, these systems are operating in some
• 1:52
  seriously harsh chemical environments.
• 1:55
  Very harsh, but the core mechanical headache isn't really
• 1:58
  the chemistry, it's the fiber itself.
• 2:01
  If you look at the raw material, a typical softwood trachite is
• 2:04
  long, maybe 3 to 7mm in the tree.
• 2:08
  And even after the craft process, they're still pretty
• 2:10
  big. They're still sizable, yeah,
• 2:12
  typically 1.9 to 2.4mm long. And when you put those long
• 2:16
  fibers into water, they don't act like sand in a bucket.
• 2:20
  They act like a net. A net is a perfect analogy.
• 2:23
  The individual fibers are suspended, but they form these
• 2:25
  coherent interconnected systems called flocks.
• 2:28
  So they all just tangle. Up instantly.
• 2:30
  Imagine you threw a whole bunch of wet cooked spaghetti into a
• 2:33
  tank and tried to stir it. That tangled network gives the
• 2:36
  entire pulp suspension mechanical strength.
• 2:39
  Before you can do anything with it, pump it, mix it, you have to
• 2:42
  apply enough force to literally tear that network apart.
• 2:45
  That's fascinating. So it behaves like a solid
• 2:48
  despite being mostly water. This brings us to a fundamental
• 2:51
  concept, consistency or mass concentration, So.
• 2:56
  Consistency is everything engineers categorize it.
• 2:59
  Low consistency is under 4%. Fiber medium, which is the sweet
• 3:03
  spot for a lot of processes, is between 8 and 16%, and high
• 3:07
  consistency is anything above 20.
• 3:09
  And here's the counter intuitive aha moment.
• 3:12
  Our sources really emphasized. Let's take medium consistency
• 3:16
  pulp, say at 10%, right. Mathematically, that material is
• 3:19
  90% water by mass, yet it exhibits solid like behavior.
• 3:25
  That is the defining engineering challenge of this entire
• 3:27
  industry because of the strength of those fiber networks, the
• 3:30
  flocks, the suspension develops A yield stress.
• 3:33
  Like ketchup in a bottle. Exactly like ketchup.
• 3:35
  You shake it, you tap it, nothing moves until you hit a
• 3:38
  certain force threshold and then whoosh, it all flows.
• 3:42
  Pulp is similar but way more extreme.
• 3:44
  Motion cannot start until the force applied overcomes that
• 3:48
  internal yield stress. And I'm guessing that yield
• 3:51
  stress climbs dramatically as you add just a little more
• 3:53
  fiber. It increases exponentially with
• 3:55
  concentration, so a mill running at 12% consistency requires
• 3:59
  significantly more robust, higher torque equipment than one
• 4:03
  running at 8:00. You were literally designing
• 4:05
  hardware to stir a semi solid gel.
• 4:08
  And this mechanical complexity isn't just an engineering
• 4:10
  puzzle, it has massive bottom line impacts.
• 4:13
  I saw a great example in the sources about the chemical
• 4:15
  stages. On bleaching you mean?
• 4:17
  Yeah, improving the quality of the mixing can dramatically
• 4:21
  reduce how much chemical you have to use.
• 4:23
  It's a direct trade off. Better mixing means better
• 4:26
  contact, which means better reaction efficiency for a big
• 4:30
  thousand ton per day mill. Our sources indicate that
• 4:32
  optimizing chlorine dioxide use just by mixing better could save
• 4:37
  1.4 million Canadian dollars annually.
• 4:41
  Wow, when you see figures like that, the incentive to get the
• 4:45
  mechanical design absolutely perfect is immense.
• 4:47
  You bet. It is.
• 4:48
  OK, so we've identified the bizarre fluid dynamics.
• 4:51
  Now let's talk about the muscle, the hardware.
• 4:53
  The industry needs a really diverse mixing toolkit, doesn't
• 4:56
  it? It has to be.
• 4:57
  You're handling everything from those viscous yield stress
• 5:00
  fluids to, you know, aggressive high shear dispersions.
• 5:03
  The mixers are always task specific.
• 5:05
  So for that thick non flowing stuff.
• 5:07
  For that you need mixers that move the entire tank.
• 5:10
  This means close clearance impellers like helical ribbon or
• 5:13
  anchor style. They often span 85 to 95% of the
• 5:17
  tank diameter. And they need massive torque at
• 5:19
  really low speeds. Exactly, to prevent the material
• 5:23
  from just spinning around with the shaft like a solid block.
• 5:25
  OK. And what about the flip side?
• 5:27
  What if you need to violently break up aggregates, like for
• 5:30
  preparing coatings? Then you need high shear mixers.
• 5:33
  It's the opposite design philosophy.
• 5:35
  They use tiny impellers, maybe 10 to 20% of the tank diameter,
• 5:39
  running at incredibly high speeds up to 3600 RPM.
• 5:44
  And they use designs like a rotor stator.
• 5:46
  Yep, a spinning inner element passes closely by a static outer
• 5:49
  element, generating these extreme localized shear forces
• 5:53
  that just obliterate lumps. And then there's the quiet hero,
• 5:57
  the static mixer. Ah yes, the static mixer. 0
• 6:02
  moving parts. It looks like a set of fixed
• 6:04
  blades inside a pipe. All the mixing energy comes from
• 6:07
  an external stock pump forcing the fluid.
• 6:09
  Through it, so it just uses pressure drop and turbulence.
• 6:12
  All without a motor or a shaft seal.
• 6:14
  It's pretty clever. The energy dissipation is around
• 6:16
  point O3 megawatts per cubic meter.
• 6:18
  So no matter the type, the prime mover is the electric motor and
• 6:22
  engineers use these NEEMA design codes dash BC or D to specify
• 6:27
  them. Right, and design B is the
• 6:28
  typical choice for mixers. It gives you normal starting
• 6:31
  torque characteristics. What that means practically is
• 6:34
  it can handle that initial heavy load to overcome the yield
• 6:37
  stress without drawing excessive current.
• 6:39
  Now, Speaking of power, the sources stress that impeller
• 6:43
  power is extremely sensitive to speed.
• 6:46
  Why is that relationship so extreme?
• 6:50
  It's a cube function. In turbulent conditions.
• 6:53
  The power required is proportional to the cube of the
• 6:55
  impeller speed. Wait, hold on.
• 6:57
  So if you double the shaft speed, say from 100 to 200 RPM,
• 7:03
  you suddenly need 8 times the power. 8 times.
• 7:05
  That's right. It's due to the exponential
• 7:07
  increase in drag and fluid friction that.
• 7:09
  Explains why the electrical system design is so critical.
• 7:12
  Standard AC motors are locked to the frequency right. 1800 RPM
• 7:16
  for 60 Hertz, 1500 for 50 Hertz. Precisely, and that sensitivity
• 7:20
  is why the industry has moved away from old multi speed motors
• 7:24
  to modern variable frequency drives or VFDS.
• 7:27
  So you can finely tune the speed.
• 7:29
  You can tune it exactly to the state of the pulp.
• 7:31
  Instead of that huge 8 to 1 power jump.
• 7:34
  A modern VFD can offer an incredible 64 to one power
• 7:38
  range. That's amazing control, but you
• 7:40
  still have to get that speed down.
• 7:42
  Motors run around 1800 RPM, but the mixers are what, 30 to 3:50?
• 7:46
  Right, so you need massive speed reducers and this is where the
• 7:49
  engineers focus shifts from speed to torque.
• 7:53
  Because torque is power divided by speed.
• 7:56
  And since you're dropping the speed so dramatically, you are
• 7:59
  massively increasing the torque. That torque is the critical
• 8:02
  factor that dictates the sheer size and strength of every
• 8:05
  single mechanical component downstream.
• 8:08
  The gears, the shaft, everything.
• 8:10
  So the mechanical integrity of that massive shaft is the next
• 8:13
  huge hurdle. It's not just a simple rotating
• 8:16
  rod, is it? Not at all.
• 8:17
  It's in a constant battle, the shaft simultaneously indoors
• 8:20
  torsion, the twisting from the motors torque and bending loads
• 8:24
  from the hydraulic forces of the impellers pushing through that
• 8:27
  thick pulp. And engineers have to calculate
• 8:29
  the minimum shaft diameter to handle those combined loads.
• 8:33
  But they can't just design for smooth operation, they have to
• 8:36
  plan for disaster scenarios. Correct.
• 8:40
  They use something called Hydraulic Service Factors or FH.
• 8:43
  Normal steady state might be a factor of 1, but what happens if
• 8:47
  the mill shuts down for a week and all that fiber settles into
• 8:51
  a dense solid block at the bottom?
• 8:53
  You need immense force to restart it.
• 8:55
  And that's what the service factor is for.
• 8:56
  Our sources show these factors can go as high as 5 or even 7.
• 9:00
  Seven. So the shaft has to be designed
• 9:02
  to withstand 7 times the normal operating torque just for a
• 9:06
  worst case start. Exactly.
• 9:07
  It's a huge, expensive safety margin, but even if the shaft is
• 9:11
  strong enough, there's an even more insidious mechanical killer
• 9:15
  vibration. You're talking about critical
• 9:17
  speed. The critical speed or a natural
• 9:19
  frequency and see it. Every rotating shaft has a
• 9:21
  frequency where it wants to vibrate like a guitar string.
• 9:24
  If your operating speed gets too close, tiny imbalances get
• 9:27
  magnified and the shaft can vibrate itself to pieces.
• 9:30
  I imagine that's a quick way to destroy a multimillion dollar
• 9:34
  piece of machinery. It is so.
• 9:36
  Engineers have a strict non negotiable rule.
• 9:39
  The operating speed N must be kept 20% away from the critical
• 9:44
  speed NC. You never operate between .8 and
• 9:48
  1.2 times NC. And how do they manipulate that
• 9:51
  critical speed? How do they move that boundary?
• 9:53
  By changing the stiffness, they need to increase the stiffness
• 9:56
  of the shaft relative to its mass.
• 9:59
  And sometimes the solution is a little counterintuitive.
• 10:03
  They use hollow shafts, often made from large diameter pipe.
• 10:07
  Although you'd think solid would be stronger.
• 10:09
  It increases stiffness dramatically while keeping the
• 10:12
  mass lower than a solid shaft of the same diameter would be.
• 10:15
  This pushes that critical speed higher, giving them a wider,
• 10:19
  safer operating window. OK, so you've contained the
• 10:21
  vibration. The final mechanical challenge
• 10:24
  is the seal where the shaft passes through the tank wall.
• 10:27
  Assume that's tough. It's notoriously difficult.
• 10:29
  These shafts are long, so they naturally have high deflection
• 10:32
  and run out of wobble. We're talking up the .38mm.
• 10:36
  That's way more than a standard pump seal could handle.
• 10:39
  Way more. You've got this massive shaft
• 10:41
  whipping around slightly inside the seal housing, so you can
• 10:45
  start simple with a stuffing box seal.
• 10:48
  Which is basically just compressed rope, right?
• 10:50
  Pretty much. It's simple and cheap, but it
• 10:52
  will always leak a little, either liquid or vapor.
• 10:55
  And if you're dealing with that hazardous white liquor, any
• 10:58
  leakage is unacceptable. So for real containment you need
• 11:01
  the double mechanical seal. That's the industry standard for
• 11:04
  safety. You have two sets of precision
• 11:07
  seal elements separated by a chamber filled with a
• 11:09
  pressurized barrier fluid. And that fluid does two things.
• 11:13
  It lubricates the seal faces and crucially it acts as a firewall.
• 11:17
  The pressure ensures that if anything leaks, it's the non
• 11:20
  hazardous barrier fluid, not the aggressive process chemical.
• 11:24
  And they're innovating even beyond that.
• 11:25
  I read about seals that don't even touch.
• 11:27
  Absolutely advanced designs like gas barrier seals use nitrogen.
• 11:31
  They machine these tiny micro grooves into the seal faces and
• 11:34
  when the shaft spins it generates a pressure cushion, a
• 11:37
  non contacting gap. So the theoretical life is
• 11:40
  infinite. Or close to it.
• 11:42
  The ultimate solution is a magnetic drive which transmits
• 11:45
  torque through a solid wall, eliminating the rotating seal
• 11:48
  entirely, but they're limited by size and cost.
• 11:51
  That is an incredible progression of ingenuity just to
• 11:54
  keep liquid wood inside a tank. And that brings us to the final
• 11:58
  hurdle, the invisible killer in material science, Corrosion
• 12:02
  fatigue. This is where the harsh
• 12:04
  chemistry meets the mechanical stress of the rotating shaft.
• 12:07
  And it's not just corrosion and it's not just fatigue.
• 12:10
  It's the destructive combination of the two acting together.
• 12:13
  How? Does that work?
• 12:14
  The cyclic stress from rotation opens up microscopic cracks on
• 12:18
  the surface, and the corrosive fluid immediately attacks that
• 12:21
  fresh exposed metal, propagating the crack deeper.
• 12:25
  This combined action is far, far worse than either one acting
• 12:28
  alone. And this failure often starts
• 12:30
  with microscopic pitting, especially in the chloride rich
• 12:33
  environments of a pulp mill. So how do engineers screen
• 12:36
  materials for that? Empirical formulas, mainly the
• 12:39
  pitting resistance index or π. It's a screening tool that
• 12:43
  weighs the chemical composition of the steel.
• 12:45
  It heavily credits elements like chromium and especially
• 12:49
  molybdenum. So a higher π means better
• 12:51
  resistance. Much better, but the core
• 12:53
  difficulty is the duration. These mixers are high cycle
• 12:57
  machines. They might endure 10 to the
• 12:59
  power of 9 to 10 to the power of 11 cycles over a 10 or 15 year
• 13:04
  life. You can't test that in a lab.
• 13:06
  It's just impractical. It's impossible, so engineers
• 13:08
  have to get predictive. They use models based on short
• 13:11
  life data. SN curves stress versus cycles
• 13:15
  to failure, and they mathematically extrapolate that
• 13:17
  out to predict long life performance.
• 13:19
  And since they're predicting performance decades out, I
• 13:22
  assume they need a massive safety cushion on that
• 13:24
  prediction. A huge one.
• 13:26
  They often apply what's called the FLEB or fatigue life
• 13:29
  extrapolation, lower bound correction factor.
• 13:32
  It's a very conservative correction that basically says
• 13:34
  even if our model is statistically wrong, we're
• 13:37
  designing to the absolute safest, lowest end estimate to
• 13:40
  prevent failure. And this drive for multi decade
• 13:43
  durability against both chemicals and stress is what
• 13:46
  pushes them to use those exotic expensive alloys.
• 13:49
  That's exactly why you see a reliance on duplex stainless
• 13:52
  steels like Alloy 2205 and even Super Austin stainless steels
• 13:56
  with up to 6% molybdenum, just to fight off that initial
• 14:00
  pitting that kicks off the whole corrosion fatigue process.
• 14:03
  It is truly astonishing that something is necessary and you
• 14:07
  know, seemingly mundane as paper relies on mechanical systems
• 14:11
  designed to withstand solid like flow, calculate shaft diameters
• 14:14
  to handle worst case scenario seven times the normal load, and
• 14:17
  engineer materials to survive corrosive stress for over a
• 14:21
  decade. We covered a huge amount of
• 14:22
  ground today. I mean, we learned that
• 14:24
  efficient mixing isn't just a small boost, it's a massive cost
• 14:27
  reduction, and that every single component from the gearbox to
• 14:30
  the seal has to compensate for the just plain weird rhinology
• 14:33
  of wet wood fiber. So we've seen the mechanical
• 14:35
  safety requires these huge investments, complex seals, high
• 14:39
  alloy shafts, massive service factors, and this is all
• 14:43
  balanced against the operational savings from better mixing.
• 14:46
  So given that constant trade off between higher mechanical cost
• 14:49
  and higher process efficiency, consider this, where do you
• 14:52
  think future innovation in this industry should focus?
• 14:54
  Maximizing efficiency with existing reliable designs, or
• 14:58
  prioritizing the development of entirely new, cheaper, more
• 15:00
  durable materials to reduce that mechanical cost.
• 15:03
  Sealing.