Latest / Mechanical Engineering Made Simple / Design Basics Geometry Calculus Standards

Transcript

Design Basics Geometry Calculus Standards · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design junkies, to Mechanical Engineering Made
• 0:02
  Simple. We are ripping the cover off the
• 0:06
  foundational knowledge that bridges the mechanical,
• 0:09
  electrical, and electronic worlds.
• 0:11
  We're talking about the core data that, you know, the modern
• 0:13
  electromechanical design engineer relies on every single
• 0:16
  day. Our mission today is to cut
• 0:18
  through the noise, explore the essential shifts in the toolkit,
• 0:21
  the absolutely crucial industry standards that prevent
• 0:25
  catastrophic failure, and the complex hidden calculations
• 0:28
  lurking inside would seem like simple machine elements, things
• 0:32
  like clutches, springs and fasteners.
• 0:34
  And that toolkit shift is, I think, the perfect place to
• 0:37
  start. What's truly fascinating is just
• 0:40
  the evolution of computational power.
• 0:42
  I mean, not that long ago, if you were calculating a complex
• 0:44
  gear train or a dynamic stress analysis, you were relying on
• 0:48
  things like, say, the Texas Instruments 85 or the Hewlett
• 0:52
  Packard 48G. Those were revolutionary at the
• 0:54
  time. They were, but you still had to
• 0:56
  manually manage steps that are just completely automated now.
• 0:58
  Today, powerful software like Mathematica or Mathcad, they
• 1:02
  automatically handle these complex processes like.
• 1:04
  What specifically? Well, logarithms, trigonometric
• 1:08
  functions, even finite element analysis, preprocessing.
• 1:12
  Things that once required you to go look up manual tables and do
• 1:15
  these, you know, laborious steps by hand.
• 1:18
  The result is a design process that is quicker, it's profoundly
• 1:21
  easier, and it is drastically more accurate.
• 1:23
  That speed and accuracy are non negotiable in modern design.
• 1:28
  But here's the thing. Despite all the gigabytes of
• 1:31
  processing power we throw at a design, the foundation is still,
• 1:35
  well, millennia old geometry. Let's start there for you.
• 1:38
  The core rules of plane geometry aren't just academic concepts,
• 1:41
  right? They're the blueprint for
• 1:43
  drafting and mechanism design. Absolutely.
• 1:46
  It's almost too easy, I think, with complex CAD software to
• 1:49
  forget the simple geometric constraints that define
• 1:51
  manufacturing success. It's the difference between just
• 1:53
  drawing a line and defining a relationship.
• 1:56
  Take the simple circle. It defines nearly every rotating
• 1:59
  or oscillating component we design.
• 2:01
  Gears, cams, linkages. Everything.
• 2:03
  Now we assume you know that a line tangent to a circle touches
• 2:07
  at a single point, but the critical take away for drafting
• 2:10
  is that the line connecting the center to that tangent point is
• 2:13
  always perpendicular to the tangent line itself.
• 2:16
  OK. And why is that specific
• 2:18
  perpendicular relationship, that 90° rule, so critical for a
• 2:22
  design junkie? Because that 90° rule is what
• 2:25
  defines the entire basis for, say, tool path definition in CNC
• 2:30
  machining, or for determining pressure angles on critical
• 2:33
  components like your teeth or Cam followers.
• 2:35
  I see. If your CAD model doesn't
• 2:37
  strictly adhere to that geometric axiom, the part won't
• 2:40
  function or it will fail and fail rapidly.
• 2:43
  It's the pure truth that ensures accuracy in manufacturing.
• 2:47
  Sure, truth. That leads us directly into the
• 2:49
  analysis of dynamic change. Which means, well, we have to
• 2:52
  talk about calculus. Now, we aren't going to turn
• 2:54
  this into a math lecture, but we do need to discuss why advanced
• 2:58
  math is so indispensable for analyzing systems that are
• 3:00
  constantly in flux. Absolutely.
• 3:03
  And look, we often need an initial rapid estimate.
• 3:06
  You know, a quick check on the back of a napkin before we fire
• 3:09
  up the full simulation software. This is where differential
• 3:12
  shine. A quick way to find the
• 3:13
  approximate change in a function, let's call it the D,
• 3:16
  is using the differential equation DS equals F prime of
• 3:20
  XXD. X and that approximation is
• 3:23
  often a huge time saver compared to calculating the exact change
• 3:27
  between two points, especially if the change is really really
• 3:30
  small. So for example, if you have a
• 3:32
  nonlinear function like say y = 2 * X to the power three halves,
• 3:36
  and you just want to know the approximate change in Y when X
• 3:39
  moves from 9 to 9.02, that differential calculation is
• 3:43
  nearly instant, and it's extremely close to the exact
• 3:45
  answer. Precisely.
• 3:46
  It gives you immediate feedback during that initial conceptual
• 3:49
  phase, when speed is frankly more important than the last
• 3:52
  decimal point of precision. If the approximation tells you
• 3:55
  the change is too large or too small, you scrap the idea before
• 3:58
  you waste hours on a detailed model.
• 4:00
  But hold on. If I'm relying on that
• 4:02
  differential approximation, how big is the risk of error?
• 4:06
  Is that only safe for a rough estimate or can I rely on it for
• 4:09
  critical sizing? That is an excellent point about
• 4:12
  professional judgment. For safety critical components
• 4:16
  or for final sizing, you always use the exact calculation or the
• 4:19
  simulation. Always the Differential is a
• 4:22
  tool for rapid iteration, for checking basic feasibility and
• 4:26
  vetting boundary conditions early on.
• 4:29
  You use it to make sure you're pointed in the right direction,
• 4:31
  not to sign off on the final design.
• 4:33
  Smart. OK, so we've covered rapid
• 4:35
  checks, but what about when the system you're designing
• 4:37
  involves, you know, complexity? Multiple variables changing at
• 4:41
  the same time? A standard derivative won't cut
• 4:43
  it then. No it won't, because engineering
• 4:45
  is rarely that simple. It almost never involves just
• 4:48
  one changing dimension. When a system involves multiple
• 4:52
  changing dimensions, partial derivatives become absolutely
• 4:55
  necessary to analyze the rate of change.
• 4:58
  Think about something like a rectangular container filling
• 5:00
  with fluid. It's volume V is a function of
• 5:03
  its depth X with West and length L.
• 5:07
  If all those dimensions are changing at the same time, maybe
• 5:10
  due to thermal expansion or a dynamic structural deflection,
• 5:13
  you need to find the rate of volume increase by partially
• 5:15
  differentiating the volume expression V = X * W * L.
• 5:19
  So if I have a scenario where that container's depth, width
• 5:23
  and length are all increasing it, let's say 3 inches per
• 5:25
  minute, you have to use partial derivatives to understand the
• 5:28
  instantaneous rate of volume change.
• 5:31
  It's the only way, the only way to accurately model that complex
• 5:34
  interdependent behavior. It moves you from simple linear
• 5:37
  modeling to the reality of multivariate physical systems.
• 5:40
  OK, we've established the language of forces and changes.
• 5:44
  Now let's talk about the physical components that contain
• 5:47
  and manage those forces. And that brings us to the most
• 5:50
  overlooked and sometimes the most dangerous part of a design
• 5:54
  fasteners. Fasteners are critical
• 5:56
  infrastructure. I mean, they're meant to be
• 5:58
  predictable, reliable and strong.
• 6:01
  Yet we face a serious global crisis.
• 6:05
  Counterfeit bolts. The sources show this is a major
• 6:07
  real world concern. The.
• 6:09
  Importation of them. Yes, counterfeit bolts marked
• 6:11
  with false head markings. They look right, but they are
• 6:14
  made from inferior steels, often without any of the necessary
• 6:17
  heat treatment. And the consequence isn't just
• 6:19
  slow degradation. The failure often happens right
• 6:22
  during installation, which could be disastrous.
• 6:24
  We're talking severe injury, property damage, or immediate
• 6:28
  structural collapse. So how does an engineer looking
• 6:31
  at a huge bin of bolts distinguish A legitimate life
• 6:34
  saving component from a deadly fake?
• 6:36
  That's why standards are our shield.
• 6:38
  The industry combats this risk by rigidly defining strength and
• 6:42
  composition through standards like SAE, the Society of
• 6:45
  Automotive Engineers, and ASTM, the American Society for Testing
• 6:50
  and Materials. A counterfeiter can fake the
• 6:52
  head marking, but they can't fake the underlying metallurgy.
• 6:56
  Can you give us a concrete example of that standard?
• 6:59
  Sure, take the SAE grade 8 fastener.
• 7:01
  You see it all the time in heavy machinery.
• 7:04
  It must be made from medium carbon alloy steel that has been
• 7:06
  quenched and tempered. Q&T That specific processing
• 7:11
  results in a guaranteed high proof load of £120,000 per
• 7:14
  square inch and a minimum tensile strength of 150,000 PSI.
• 7:19
  And that standard holds true across a wide range of common
• 7:22
  diameters, from 1/4 inch all the way up to 1 1/2 inches.
• 7:25
  So the Q&D process, quenching and tempering it isn't just a
• 7:29
  material choice, it's a structural necessity that
• 7:31
  determines how the fastener fails.
• 7:33
  It is. A proper grade 8 bolt will
• 7:35
  typically fail in a ductile manner, meaning it will stretch
• 7:39
  and deform for it ruptures, which gives you a warning sign.
• 7:42
  An inferior or counterfeit bolt lacking that tempering might
• 7:46
  fail brittily. It just snaps without warning,
• 7:49
  which is catastrophic. And for those of us using metric
• 7:52
  components. Same principles apply.
• 7:54
  You use the numerical markings on metric hex cap screws like
• 7:57
  4.68 point 810.9 and 12.9. Those numbers denote the
• 8:01
  strength classification of material grade for structural
• 8:04
  integrity. Adherence to these standards is.
• 8:06
  I mean, it's absolutely non negotiable.
• 8:08
  Moving beyond mechanical fasteners, welding is another
• 8:11
  critical joining method that requires absolute
• 8:13
  standardization. It's a fusion process, right?
• 8:17
  So if it's performed correctly and the preparation is right,
• 8:20
  the resulting junk can theoretically be as strong as
• 8:22
  the original parent metal. And that if performed correctly
• 8:25
  part depends entirely on organizations like the American
• 8:28
  Welding Society or AWS. They provide the specific
• 8:32
  designations for all the various processes.
• 8:34
  For example, knowing the difference between GMAW that's
• 8:38
  gas metal arc welding and TIG welding, or even the various
• 8:41
  metal cutting processes like plasma oxygen cutting POC or oxy
• 8:46
  fuel gas cutting OFCA. And that quantification of
• 8:49
  strength is key for the designer.
• 8:51
  If I specify a Weld, I need to know the force it can actually
• 8:54
  bear. Exactly.
• 8:55
  We can turn a physical joint into calculable forces.
• 8:58
  If you're designing A structural frame and you specify a lay Weld
• 9:01
  with a .25 inch leg, and you use a high strength 80,000 PSI Weld
• 9:06
  rod, you can quantify that it provides an allowable unit force
• 9:10
  of 4125 lbs per linear inch. So you just do the math.
• 9:14
  You do the math, you calculate the force the frame needs to
• 9:16
  bear divided by that unit force, and you know exactly how long
• 9:18
  your Weld bead needs to be. Calculations are one thing, but
• 9:21
  you know shop practice is another.
• 9:23
  We talk about welding massive sections of steel.
• 9:26
  What are the shop floor practices that prevent a design
• 9:28
  failure? Temperature control is
• 9:30
  everything. Large sections of thick metal.
• 9:33
  They require preheating, sometimes up to several 100°.
• 9:37
  And why is that? To prevent thermal shock and
• 9:40
  minimize distortion or cracking as that molten pool cools, the
• 9:44
  preheating minimizes the temperature gradient across the
• 9:47
  material. That makes sense for thick
• 9:49
  material, but what if the material is really thin, say
• 9:52
  less than 1/10 of an inch? For thin materials, you do the
• 9:54
  opposite. You use heat sinks and these can
• 9:57
  be as simple as say, wet burlap bags placed right next to the
• 10:01
  Weld zone or specialized copper blocks.
• 10:04
  To draw the heat away. Exactly.
• 10:06
  They draw heat away quickly to prevent localized buckling or
• 10:09
  warp. And here is a crucial rule that
• 10:11
  gets violated all the time in rush production.
• 10:14
  Welded parts should never be water quenched immediately after
• 10:17
  welding. Rapid cooling causes distortion
• 10:19
  and even worse, harmful changes in the internal grain structure.
• 10:23
  It compromises the integrity of the Weld.
• 10:26
  You have to let them air cool naturally to ambient
• 10:28
  temperature. That attention to thermal
• 10:30
  details brings us neatly to the electromechanical interface,
• 10:33
  starting with the unassuming spring, a component that harbors
• 10:37
  a critical hidden chemical danger.
• 10:39
  Springs often need Environmental Protection, usually by
• 10:42
  electroplating with metals like bright zinc or cadmium.
• 10:46
  And that plating process, while it provides corrosion
• 10:49
  resistance, is a chemical ticking time bomb.
• 10:51
  The process electroplating it unfortunately introduces atomic
• 10:55
  hydrogen into the material structure of the spring steel.
• 10:59
  This causes what we call hydrogen embrittlement.
• 11:02
  Hydrogen embrittlement? That sounds terrible.
• 11:04
  What happens if you skip the preventative step?
• 11:05
  If you skip prevention, failure is not a possibility, it's an
• 11:09
  inevitability. The hydrogen atoms, they migrate
• 11:11
  to points of high stress like sharp corners or grain
• 11:14
  boundaries, and they coalesce into molecular hydrogen.
• 11:18
  This creates immense internal pressure that reduces the
• 11:20
  steel's ductility. The spring will operate just
• 11:23
  fine for a while, maybe hours or days, and then suddenly fracture
• 11:26
  catastrophically under its normal working load.
• 11:29
  Wow, so skipping a single process step creates a time
• 11:32
  delayed brittle failure? What's the fix for this
• 11:35
  dangerous chemistry? The fix is very specialized.
• 11:37
  Plated springs must be post baked immediately after plating.
• 11:41
  This is a dedicated operation. Typically involves oven heating
• 11:44
  the springs at 500°F for a minimum of three hours.
• 11:48
  And that gets the hydrogen out. It bakes out the introduced
• 11:51
  hydrogen before it can cause damage.
• 11:53
  If this post bake is not performed, hydrogen
• 11:56
  embrittlement will invariably occur, leading to failure.
• 11:59
  It's specialized knowledge. You simply cannot overlook an
• 12:02
  electromechanical design. Post baking saves the day.
• 12:06
  OK, now let's move to dynamic power transmission, specifically
• 12:10
  electromagnetic clutches. Their torque rating is vital,
• 12:13
  and their placement in the drivetrain completely dictates
• 12:16
  the capacity you need. That placement is absolutely
• 12:19
  everything because of torque multiplication.
• 12:22
  See, clutches are rated for the torque they can transmit at
• 12:24
  their operating speed. If you have an electromagnetic
• 12:27
  clutch rated for say 280 LB inches of torque at 90 volts DC
• 12:32
  and you install it before a gearbox that has an 8 to 1
• 12:35
  reduction ratio. Then the output of that gearbox
• 12:37
  is dramatically leveraged. Exactly.
• 12:39
  The clutch can now drive a load with a multiplied torque value
• 12:42
  of 8 * 280 or 2240 LB inches. Placing the clutch on the high
• 12:48
  speed side of the gearbox means you can use a much smaller and
• 12:51
  cheaper component That's a. Huge mechanical advantage, but
• 12:54
  when you're sizing that clutch or an equivalent brake, you have
• 12:57
  to account for the kinetic energy of the entire system, not
• 13:00
  just the torque rating of the final load.
• 13:02
  Right, this is where we return to advanced calculation and the
• 13:06
  concept of reflected inertia. Inertia is resistance to change
• 13:09
  in motion. When sizing a clutch or a brake,
• 13:12
  all the rotating parts, gears, pulleys, shafts and all the
• 13:15
  linear moving parts like a long heavy conveyor belt, they must
• 13:19
  have their inertia, which we often represent as WR squared,
• 13:22
  calculated or reflected to the speed of the clutch or brake
• 13:25
  shaft. OK, let's simplify that for the
• 13:27
  designer listening. Why reflect the inertia?
• 13:29
  Imagine you're trying to stop a train with a small handbrake.
• 13:32
  The brake doesn't feel the trains linear weight, it feels
• 13:35
  the rotational inertia of the wheel axle it's clamped to.
• 13:39
  But that axle is moving much slower than the motor shaft.
• 13:42
  You have to convert all the kinetic energy of all the moving
• 13:45
  components, the train, the gears, the motor itself, into a
• 13:49
  single equivalent inertia value at the speed of the clutch
• 13:52
  shaft. If you fail to calculate that
• 13:55
  reflected inertia, the clutch will be undersized, it'll burn
• 13:58
  out prematurely, or the brake just won't stop the system fast
• 14:01
  enough. It ensures the component can
• 14:03
  handle the total kinetic energy of the entire system.
• 14:05
  It is responsible for starting or stopping.
• 14:08
  And tying the whole electromechanical system
• 14:10
  together is, of course, the motor itself.
• 14:13
  Motor name plates give us crucial data, but there's a key
• 14:16
  distinction about speed that we really must address.
• 14:19
  Yes, for induction motors you have to know that the motors
• 14:22
  nominal full load speed is always less than its theoretical
• 14:25
  synchronous speed. This is fundamental to how AC
• 14:28
  induction motors operate. They need what's called slip to
• 14:31
  generate torque. So if I'm doing speed
• 14:33
  calculations for a conveyor belt system, I can't just use the
• 14:36
  theoretical number based on the poles and the frequency.
• 14:38
  Precisely, if you have a standard 4 pole 60 Hertz
• 14:42
  fractional horsepower motor, it's synchronous speed is one
• 14:46
  1000 RPM, but the speed you need to use for accurate pulley or
• 14:50
  gear sizing, the nominal full load speed is typically a 1725
• 14:54
  RPM. That's a big difference.
• 14:55
  It is if you use 1800 RPM, your entire system will run slower
• 14:59
  than you paculated, and that could potentially cause a
• 15:01
  production bottleneck. We have Journey today from, you
• 15:05
  know, pure geometric axioms all the way to the dangers lurking
• 15:09
  inside material chemistry. We've synthesized that fluency
• 15:12
  in advanced mathematics is vital for quick checks and for multi
• 15:16
  variable dynamic analysis. We've seen that structural
• 15:18
  standards for components like bolts and welds must be strictly
• 15:21
  adhered to. They are the designers primary
• 15:24
  shield against unexpected failure.
• 15:26
  And we've highlighted that specialized knowledge, things
• 15:28
  like understanding hydrogen and brittlement in springs and
• 15:30
  accurately calculating reflected inertia for clutches is what
• 15:34
  determines the ultimate success and reliability of an integrated
• 15:37
  electromechanical design. It's a massive landscape
• 15:40
  connecting the theoretical to the physical in every single
• 15:42
  aspect. And let me leave you with a
• 15:44
  final provocative thought that ties back to our very first
• 15:47
  discussion about math versus reality.
• 15:50
  The source material states clearly that even the most
• 15:52
  rigorous mathematical solutions to complex industrial design
• 15:55
  engineering problems are in many cases approximations at best.
• 16:00
  So what does this all mean for you, the designer, when the math
• 16:03
  itself is inherently imperfect? The greatest and most definitive
• 16:07
  proof of a design's adequacy is always the prototype and
• 16:10
  subsequent definitive testing. However, there are monumental
• 16:13
  products and engineering projects.
• 16:15
  Think massive infrastructure or single build space systems we're
• 16:19
  producing. A full scale prototype is
• 16:20
  functionally impossible or just prohibitively expensive.
• 16:24
  In those critical scenarios where testing is not an option,
• 16:26
  you must rely entirely on accepted industry standards, the
• 16:30
  most rigorous calculations possible, and professional
• 16:32
  historical experience alone. This highlights why mastering
• 16:36
  every single fundamental role in your design handbook is not
• 16:39
  optional. It is the only guarantee you
• 16:41
  have when the prototype is beyond reach.