Latest / Mechanical Engineering Made Simple / (#83) Industrial Mixing Is All Structural Engineering

Transcript

(#83) Industrial Mixing Is All Structural Engineering · Mechanical Engineering Made Simple

• 0:00
  Welcome back to Mechanical Engineering Made Simple.
• 0:02
  You know, if I asked you to mix a huge industrial VAT, you might
• 0:06
  be OK. I'll grab a big propeller and
• 0:08
  just turn it on. Yeah, how hard can it be?
• 0:10
  Exactly. But today we are diving into a
• 0:13
  world where that simple act becomes a a really critical
• 0:17
  precision engineering challenge. It absolutely is.
• 0:20
  We're talking industrial mixing, and this isn't just simple
• 0:23
  stirring. It's a field that demands A
• 0:25
  sophisticated balance of well, of everything, optimizing
• 0:30
  complex process needs, mastering intense fluid dynamics, and
• 0:34
  ensuring total mechanical reliability for equipment that
• 0:36
  might have to run non-stop for 20 years.
• 0:39
  It's the ultimate intersection of process chemistry and heavy
• 0:43
  duty structural mechanics. Our mission in this deep dive is
• 0:46
  pretty straightforward. We're going to distill the
• 0:48
  critical non negotiable role the mixing equipment supplier plays
• 0:52
  and we'll lay out the engineering factors from the
• 0:55
  basic parameters like sizing all the way to managing severe
• 0:58
  dynamic mechanical loads that really dictate whether a system
• 1:02
  succeeds or fails. And we're talking about massive
• 1:05
  applications, sometimes in tanks holding millions of liters.
• 1:08
  So we're going way beyond just stirring.
• 1:10
  We're getting into the structural integrity that makes
• 1:13
  it all possible. That's the plan.
• 1:14
  OK, let's unpack this then. The sources really emphasize
• 1:18
  that a successful, reliable mixing setup begins well long
• 1:22
  before the first piece of steel is even cut.
• 1:24
  It requires an intensely close partnership, and this is the key
• 1:28
  a complete information exchange between the purchaser and the
• 1:31
  supplier. So if a buyer has proprietary
• 1:34
  secrets, why is sharing that process information so
• 1:37
  fundamentally crucial? Because the final mixing
• 1:40
  configuration is sensitive to absolutely everything, the
• 1:43
  supplier has to understand the environment, the constraints,
• 1:46
  and maybe most importantly, the priorities.
• 1:48
  What do you mean by priorities? Well, things like the position
• 1:51
  of the impeller relative to the liquid level, the tank geometry,
• 1:55
  even where you place the baffles.
• 1:57
  All of it dictates the flow pattern and where the shear
• 2:00
  forces are distributed. If the supplier is working
• 2:03
  blind, the machine they provide is.
• 2:07
  It's essentially a guess. A very expensive guess.
• 2:10
  I noticed in the sources they talked a lot about conflicting
• 2:13
  needs. What kind of engineering
• 2:14
  conflict comes up when you're trying to optimize a process
• 2:17
  like this? A classic example is the trade
• 2:20
  off between high shear and high flow.
• 2:22
  It's a constant battle. OK, you might need really high
• 2:24
  shear energy to accomplish, say, a fine dispersion, like creating
• 2:29
  a perfect emulsion where the droplets are microscopic.
• 2:32
  But to make sure that dispersed material actually reaches every
• 2:35
  corner of the tank, you need high bulk flow for blending or
• 2:38
  solid suspension. O you're asking the mixer to do
• 2:41
  2 different things at once. Exactly 2 things that ull
• 2:43
  against each other. Mechanically, if the urchaser
• 2:45
  doesn't specify which task is the absolute most important, the
• 2:48
  process yield severs. The engineer has to make a
• 2:51
  calculated compromise. That makes a lot of sense.
• 2:54
  So when you're submitting a request for proposal, an RFP,
• 2:58
  you can't just say I need a mixer for a 10,000 liter tank.
• 3:01
  You have to get into serious detail.
• 3:03
  Oh, it has to be comprehensive. They need a list of all
• 3:06
  materials, liquids, solids, gases, including both the
• 3:09
  nominal and the peak amounts. You have to define the
• 3:12
  operational mode. Is it batch or is it continuous
• 3:16
  flow? And what about the process
• 3:17
  itself? Here's the critical insight you
• 3:20
  have to identify which step in the overall manufacturing
• 3:23
  process is the rate controlling step that tells the engineer
• 3:27
  where they can, you know, afford to save power and where they
• 3:30
  absolutely must allocate maximum energy.
• 3:33
  And for applications with solids like solid suspension, the
• 3:36
  detail gets almost microscopic. Are we really talking about
• 3:39
  specifying individual particle sizes and densities?
• 3:42
  That sounds like overkill. It's not overkill at all.
• 3:44
  It's pure physics. You absolutely have to specify
• 3:47
  particle size and density, and precisely precisely what level
• 3:50
  of suspension you require. What are the different levels?
• 3:52
  Well, do you just need off bottom suspension, which is just
• 3:56
  enough energy to keep the solids from settling and building up?
• 3:59
  Or do you need full suspension where those particles are
• 4:03
  uniformly distributed throughout the entire liquid volume?
• 4:06
  And the power difference is significant.
• 4:08
  The difference in required power between those two states can be
• 4:12
  monumental. It's huge.
• 4:14
  Wait a second, something in the sources really jumped out at me.
• 4:17
  It's said that the way they pour the raw feed into the tank
• 4:20
  actually changes the dynamic loads on the mixing structure.
• 4:23
  That's that's not just a process variable, that's a structural 1.
• 4:27
  That's where it gets really interesting.
• 4:29
  That's where fluid dynamics connects right back to the
• 4:31
  mechanical design. The purchaser has to specify the
• 4:34
  characteristics of the inlet feed, the velocity, or the flow
• 4:37
  rate. If you have a powerful stream
• 4:40
  aimed directly at the impeller, that stream introduces
• 4:43
  substantial, calculable forces. The equipment supplier can't
• 4:47
  just focus on the loads generated by the impeller
• 4:49
  itself, they have to account for those external dynamic forces
• 4:52
  from the process stream. Which means what for the
• 4:55
  hardware? It means designing A thicker
• 4:57
  shaft and a more robust speed reducer.
• 4:59
  It changes the entire mechanical design from the ground up.
• 5:02
  Wow. OK, so once all that process
• 5:04
  details locked down, the focus shifts to the raw physics of
• 5:07
  sizing the actual machine. Exactly.
• 5:10
  The foundational steps are pretty standard.
• 5:12
  Customer specifies the process goal, the tank geometry like its
• 5:16
  diameter and the volume. Then the supplier uses their own
• 5:20
  proprietary correlations and modeling to calculate the
• 5:23
  required power. So like the best impeller type,
• 5:26
  determine its diameter and then establish a safe operating
• 5:29
  speed. But the sources make it very
• 5:31
  clear this isn't a linear process.
• 5:34
  Let's talk about the D / T ratio, the ratio of the impeller
• 5:37
  diameter to the tank diameter. They call this a very strong
• 5:40
  function of power. Why does that specific ratio
• 5:43
  have such a massive impact on energy needs?
• 5:46
  This is where understanding fluid circulation becomes
• 5:48
  absolutely vital. For an axial flow impeller
• 5:51
  that's designed to suspend solids, there's usually an ideal
• 5:54
  ratio, it's often around 0.35, and at that ratio it creates a
• 5:58
  perfect predictable circulation loop, efficiently lifting
• 6:02
  material off the bottom and sending it up to the surface.
• 6:05
  But if you increase that impeller diameter, pushing the
• 6:08
  ratio up to say 0.5. The power goes up a little bit.
• 6:11
  It doesn't just creep up, it explodes.
• 6:14
  You might suddenly need 16 times the optimum power.
• 6:17
  Why such a massive exponential jump?
• 6:19
  That seems counterintuitive. Because the larger impeller
• 6:22
  physically starts to cripple its own axial circulation loop.
• 6:26
  Instead of efficiently pumping fluid around the tank, the
• 6:29
  impeller is now pushing fluid against itself and against the
• 6:32
  tank wall. It severely disrupts the very
• 6:35
  flow pattern the system needs to live.
• 6:38
  I see. So the efficiency just plummets
• 6:40
  and you have to dump massive amounts of extra energy just to
• 6:42
  force the fluid to move. It's like it's a self-inflicted
• 6:44
  wound caused by going just slightly oversized.
• 6:47
  That inefficiency and mechanical stress that brings us directly
• 6:51
  to the next non negotiable factor avoiding the critical
• 6:55
  speed or ncrit. Exactly.
• 6:57
  Ncrit is the natural frequency of the entire rotating system.
• 7:01
  The shaft, the impeller, any attached components.
• 7:03
  You can think of it like a perfectly tuned tuning fork.
• 7:06
  If the operating speed N hits or even gets close to that natural
• 7:11
  frequency, the whole shaft system enters a state of
• 7:15
  uncontrollable undamped vibration.
• 7:17
  It's called resonance. It's like driving on the highway
• 7:20
  and hitting that one specific speed, say 65 mph, where your
• 7:24
  entire car just starts shaking violently.
• 7:26
  That is the perfect analogy, except in this industrial case,
• 7:29
  the shaking doesn't just annoy you, it breaks the whole
• 7:32
  machine. Catastrophically.
• 7:33
  Catastrophically, that vibration causes rapid mechanical fatigue
• 7:37
  on the shaft, on the seals, on the gearbox.
• 7:40
  It leads to failure. So to protect the system, the
• 7:43
  standard engineering design rule is mandatory and.
• 7:45
  What's the rule? The operating speed N must be
• 7:49
  kept 20% away from the critical speed NC.
• 7:53
  So the operating speed must be less than 0.8 times NC or it has
• 7:58
  to be greater than 1.2 times NC. You just can't be in that middle
• 8:01
  zone. And this applies to the 1st and
• 8:03
  often the second natural frequencies of the system.
• 8:07
  And this critical speed management ties directly into
• 8:10
  the big economic equation, right?
• 8:12
  Capital cost versus operating costs.
• 8:14
  It absolutely does. A larger impeller diameter, for
• 8:17
  example, typically reduces long term operating costs because it
• 8:20
  uses less power. It moves more fluid volume per
• 8:23
  rotation, but there's. Always a but.
• 8:25
  There's always a, but that larger impeller creates much
• 8:28
  higher torque requirements. Higher torque means you need a
• 8:31
  larger, heavier speed reducer, the gearbox, and a significantly
• 8:35
  thicker sticker shaft to keep that operating speed far away
• 8:38
  from encrypt. So the engineer's real challenge
• 8:41
  is finding that precise sweet spot.
• 8:42
  It's an optimization problem you have to solve before the machine
• 8:46
  even starts running. OK, let's shift gears a bit and
• 8:49
  talk about the hardware catalog. The variety of mixing equipment
• 8:53
  is huge, but we can kind of categorize it by what the
• 8:56
  impeller is engineered to. Do you have high flow low energy
• 9:00
  impellers versus high shear, high intensity ones right?
• 9:04
  The high flow, low head impellers, like a high
• 9:06
  efficiency flow impeller, are designed for maximum volume
• 9:09
  movement. Their efficiency is defined by
• 9:11
  something called the power number or NP.
• 9:13
  And what does that tell you? It tells you how much power the
• 9:16
  given impeller needs to generate flow.
• 9:19
  A low NP, typically around 0.3, means you are highly efficient.
• 9:24
  You move a massive volume of liquid for very little energy
• 9:26
  input. These are your workhorses for,
• 9:28
  you know, mild blending or basic suspension.
• 9:31
  And on the other end of the spectrum, you have the
• 9:32
  specialized high shear impellers.
• 9:34
  Yeah, these are built for specific tasks like fine
• 9:37
  dispersion, high intensity blending, or incorporating
• 9:40
  really difficult powders. They run at very high speeds and
• 9:43
  typically have a power number closer to 0.45.
• 9:47
  They basically sacrifice bulk flow for intense local
• 9:51
  turbulence and extreme energy dissipation.
• 9:54
  Looking at the physical designs, we can start small with portable
• 9:57
  mixers. These are usually running from
• 9:59
  1/4 horsepower up to maybe 3 horsepower.
• 10:02
  Right, and they're often just clamp mounted to thin walled
• 10:04
  tanks and use pretty basic axial flow impellers.
• 10:06
  Then you step up to top entering mixers.
• 10:08
  These use more fixed mounting structures, flange or pedestal
• 10:12
  mounts. It might go up to 10 or 15
• 10:14
  horsepower. They have robust mounting to
• 10:16
  stabilize the shaft and a much better mechanical seal system,
• 10:20
  which is crucial for closed vessels.
• 10:22
  And then you get to the heavy duty segment, the massive
• 10:25
  turbine mixers. Now you're talking.
• 10:27
  This is a broad class that covers the most robust designs,
• 10:30
  running from 1 horsepower all the way up to 1000 horsepower or
• 10:34
  even larger. You'll see these in almost every
• 10:36
  massive chemical processing application imaginable.
• 10:39
  We can't overlook the true specialty mixers though, where
• 10:42
  the fluid properties themselves completely dictate the machine
• 10:45
  design, like for high viscosity applications, fluid over 100,000
• 10:50
  centipoise. Right, for those of us who don't
• 10:51
  have a viscometer handy. I was just about to ask, what is
• 10:55
  100,000 centipoise like in the real world?
• 10:58
  That's like trying to make something thicker than peanut
• 11:00
  butter or cold honey or heavy grease.
• 11:04
  You are no longer generating bulk circulation flow, you are
• 11:08
  physically dragging the material through what's called plastic
• 11:11
  flow. So you need a totally different
• 11:13
  strategy. A completely different strategy.
• 11:15
  You shift to a close clearance impeller.
• 11:18
  The impeller diameter has to be 85 to 95% of the tank diameter.
• 11:22
  You see styles like helical ribbon or anchor impellers.
• 11:25
  And the mechanics of that the. Key mechanical outcome is
• 11:27
  extremely low operating speed but enormous torque to overcome
• 11:32
  that viscous drag. And conversely, high shear
• 11:35
  mixers are the other extreme. They use very small impellers,
• 11:39
  sometimes only 10 to 20% of the tank diameter, but they spin at
• 11:42
  incredible speeds up to 3600 RPM.
• 11:45
  The visual is really essential here for the listener.
• 11:48
  Picture a disc impeller with teeth all around the edge, or
• 11:51
  even more complex, a rotor stator design.
• 11:54
  The rotor spins inside a stationary cylinder that has
• 11:57
  slots in it. And the fluid.
• 11:59
  The fluid is physically forced through that narrow gap,
• 12:02
  creating maximum mechanical and localized hydraulic shear.
• 12:06
  It literally rips particles apart or forces immiscible
• 12:09
  liquids to combine. So what this all means for the
• 12:12
  components is that whether it's a tiny portable unit or a
• 12:15
  massive turbine, the core, mechanical parts, the motor,
• 12:18
  speed reducer, seal shaft and impellers, they all have to be
• 12:22
  engineered specifically to withstand those calculated
• 12:25
  severe dynamic loads. That shaft design has to be
• 12:28
  rigid first, strong second. Absolutely, and the suppliers
• 12:31
  job extends well past just the initial design.
• 12:34
  It involves rigorous verification and testing.
• 12:36
  Most pre purchase customer sample testing is done on a
• 12:39
  relatively small scale, maybe in a 0.45 meter diameter tank.
• 12:43
  So. A lab scale test.
• 12:44
  Exactly. And the primary engineering
• 12:46
  challenge then becomes scale up, moving from that small sample to
• 12:50
  a massive industrial reactor. And the sources warn that you
• 12:53
  can't just scale up the size and use the same materials, right?
• 12:55
  The physical properties change with scale, particularly
• 12:58
  viscosity reduction. Precisely, if you use the same
• 13:01
  viscosity material on the small scale and the full scale, the
• 13:04
  flow regime might be, say, transitional or even laminar in
• 13:07
  the sample tank, but at full scale it's highly turbulent.
• 13:11
  So the test isn't valid. You have to model that viscosity
• 13:14
  reduction correctly to make sure the process achieves comparable
• 13:17
  flow conditions at full production capacity.
• 13:20
  If you don't, the full scale unit might underperform
• 13:23
  drastically. Fortunately, engineers have some
• 13:25
  incredibly advanced diagnostic tools now to peek inside the
• 13:30
  flow without actually touching it.
• 13:32
  We see tools like Laser Doppler velocimetry or LDV and article
• 13:36
  image velocimetry. Four, these are non intrusive
• 13:39
  tools. They're used secifically to
• 13:40
  measure the velocity and the entire flow field inside the
• 13:43
  vessel. This is crucial for optimizing
• 13:46
  impeller placement and critically, for identifying
• 13:49
  undesirable low velocity regions.
• 13:51
  Which often pop up in surprisingly key areas, right?
• 13:54
  Like near the top surface where you're adding ingredients, or in
• 13:58
  those little dead spots tucked right underneath a big impeller
• 14:00
  at the bottom of the tank. Identifying and eliminating
• 14:03
  those dead spots is key. It's how you avoid inconsistent
• 14:07
  blending, unwanted heat buildup, or localized reactions.
• 14:11
  And then there's computational fluid Dynamics CFD.
• 14:14
  Right. And this modeling tool is
• 14:16
  becoming excellent and increasingly cost effective.
• 14:19
  We can use CFD to model really complex flow patterns, maybe
• 14:23
  visualize how multiple up pumping impellers interact with
• 14:26
  each other, and it can calculate the power number with an
• 14:29
  accuracy of just a few percent, which validates the design
• 14:32
  before you even fabricate any steel.
• 14:35
  Process optimization is well, it's meaningless if the machine
• 14:39
  tears itself apart after six months.
• 14:41
  Let's get back to mechanical reliability, which hinges
• 14:43
  entirely on knowing the fluid loads.
• 14:45
  The shaft is under constant chaotic abuse.
• 14:48
  The dynamic loads applied are torque bending and thrust.
• 14:51
  What's truly astounding is the randomness of the fluid flow
• 14:54
  even when it's under steady operation.
• 14:56
  How? Random are we talking?
• 14:57
  The blade loading on the impeller can vary dramatically.
• 15:01
  We're talking sometimes by as much as ±40% to ±100% depending
• 15:07
  on where local turbulence is hitting the blade at any given
• 15:10
  millisecond wait. ±100%. That means the load is
• 15:14
  constantly oscillating from zero to double the average load
• 15:18
  hundreds of times a minute. Exactly.
• 15:20
  The entire mixing system has to be engineered not just for the
• 15:23
  average load, but for that extreme chaotic fatigue cycling.
• 15:28
  Which is why we come back to the critical speed rule.
• 15:31
  That's why we come back to it. Stay 20% away from the 1st and
• 15:34
  2nd natural frequencies. Period.
• 15:36
  Your operating speed N cannot be greater than or equal to 0.8
• 15:41
  times NC, and it can't be less than or equal to 1.2 times NC.
• 15:47
  Stay out of that band. And finally, let's address the
• 15:49
  tank support, which is often overlooked.
• 15:51
  The sources are so insistent on this.
• 15:53
  The mounting structure must be stiff, not just strong.
• 15:56
  Yes, strength ensures it won't collapse, but stiffness ensures
• 16:01
  it won't move. A structure designed only for
• 16:03
  strength might experience noticeable deflection or
• 16:06
  movement under a dynamic load, and that flexibility actually
• 16:09
  reduces the overall system's natural frequency.
• 16:12
  And if you have poor mounting, say with those portable mixer
• 16:15
  clamps on a thin walled vessel right, that non rigid support
• 16:20
  effectively lowers the calculated critical speed of the
• 16:23
  shaft. The actual critical speed could
• 16:26
  be much lower than the engineer calculated.
• 16:28
  Potentially dropping the operating speed right back into
• 16:31
  that dangerous resonance zone and that leads to dangerous
• 16:34
  vibration and failure. It's a classic failure mode that
• 16:36
  starts at the foundation. So if we were to summarize this
• 16:39
  entire deep dive. To summarize this entire
• 16:42
  process, effective industrial mixing is fundamentally an
• 16:45
  optimization problem. It requires meticulously
• 16:48
  balancing the desired process yield, your product outcome
• 16:51
  against the long term operating costs which are all about power
• 16:54
  consumption, and against the essential non negotiable
• 16:57
  mechanical integrity which is managed through torque capacity
• 17:00
  and critical speed control. It's a three-way balancing act.
• 17:04
  And thinking about that long term integrity, let's leave you
• 17:06
  with one final provocative thought.
• 17:08
  This one based on material selection mentioned in the
• 17:10
  sources. The actual metallurgy matters a
• 17:13
  lot for decade long reliability. Take 317 L stainless steel.
• 17:19
  It's often chosen specifically for its resistance to chloride
• 17:23
  corrosion. Now the engineering
• 17:25
  specifications for 317 L allow a molybdenum content range of 3 to
• 17:29
  4%. Right.
• 17:31
  Due to market realities and cost pressures, today, new batches of
• 17:34
  that steel often contain only three to 3.25% molybdenum, the
• 17:38
  absolute lower end of that range, which is technically
• 17:41
  still compliant with the specification.
• 17:43
  Whereas 20 years ago, a mixer built of 317 L would have
• 17:46
  averaged closer to, say, 3.5% molybdenum.
• 17:50
  That subtle difference in composition while still meeting
• 17:52
  the spec, could dramatically shorten the lifespan of the
• 17:55
  equipment and a corrosive high chloride environment compared to
• 17:58
  older material carrying the exact same designation.
• 18:01
  So even if the paperwork looks identical.
• 18:02
  Exactly. It shows how subtle market
• 18:05
  pressures can reduce long term reliability even when technical
• 18:08
  specs are met. You have to understand what
• 18:10
  you're buying today to protect your investment tomorrow.
• 18:13
  And if you haven't already, get over to our YouTube channel
• 18:16
  Mechanical Engineering Made Simple, where we post hands on
• 18:19
  how to videos, visual breakdowns, and practical
• 18:21
  engineering walkthroughs you can actually use.
• 18:24
  A family that solves together stays together.