Latest / Mechanical Engineering Made Simple / (#76) Complex Stress Analysis The Engineer's Toolkit

Transcript

(#76) Complex Stress Analysis The Engineer's Toolkit · Mechanical Engineering Made Simple

• 0:00
  Welcome back, design junkies, to mechanical engineering made
• 0:02
  simple. Today we are taking a a massive
• 0:05
  leap. We're moving beyond the simple
• 0:08
  push and pull forces we all know.
• 0:10
  Our mission today is a real deep dive into analysis.
• 0:13
  We want to understand how engineers breakdown scenarios
• 0:16
  where forces just collide. You've got bending, twisting,
• 0:21
  and pulling all at once. All happening in the same place.
• 0:24
  We'll look at how they calculate things like stiffness and
• 0:26
  deflection, and maybe most importantly, how those elegant
• 0:29
  classical theories laid the groundwork for the powerful
• 0:32
  computer tools we have today. That's the key connection, isn't
• 0:36
  it? The bridge from calculus to
• 0:37
  computer science. I mean, you think about an
• 0:39
  aircraft wing or offshore rig. The stresses inside are just not
• 0:43
  simple in uniform. Not at all.
• 0:45
  You're dealing with these incredibly complex, non uniform
• 0:48
  stress states, so you need pretty advanced techniques to
• 0:51
  figure out, you know, both the structure's ultimate strength
• 0:53
  and its day-to-day stiffness. OK, so let's start where that
• 0:56
  complexity really begins, with combined loading.
• 1:00
  We know about direct axial stress, but what happens when we
• 1:05
  take a column, say, and the load isn't perfectly centered?
• 1:09
  Ah, that's where it gets interesting, because what
• 1:10
  engineers do is they simplify it and eccentric load P that's a
• 1:14
  load that's just a little off center is statically equivalent
• 1:18
  to two separate effects. Two things at once.
• 1:20
  Yep. It's like having a pure axial
• 1:22
  thrust P right at the center plus a bending moment M, and
• 1:27
  that moment is just the load P times its distance from the
• 1:30
  center. EP times E So you're not just
• 1:33
  compressing the column, you're actually bending it at the exact
• 1:36
  same time. You have two different stress
• 1:38
  profiles just layered on top of each other.
• 1:40
  Precisely. You have the uniform compressive
• 1:42
  stress across the whole section from the thrust, and then you
• 1:45
  have the bending stress compressive on one side, tensile
• 1:48
  on the other. And the final stress is just the
• 1:50
  sum of the two. You just add them up
• 1:52
  algebraically. It's that simple.
• 1:54
  That has, I mean, huge implications, especially for
• 1:57
  construction, right? I'm thinking of materials like
• 1:59
  stone or masonry, things that are just awful intention.
• 2:03
  Absolutely. Think about an old stone arch
• 2:05
  bridge. You have to guarantee that under
• 2:07
  normal loads you never generate any tensile stresses.
• 2:11
  Begins. It'll just crack.
• 2:12
  Instantly, and this is where the famous rule of thumb comes from.
• 2:17
  For a solid rectangular section to keep all the stress
• 2:20
  compressive, the load has to fall within the middle third of
• 2:24
  the section. The middle third rule.
• 2:25
  Right. If that eccentricity is greater
• 2:27
  than 1/6 of the depth, you start pulling the other side into
• 2:31
  tension and that's where you get failures.
• 2:33
  So a simple design rule born from just summing 2 stresses.
• 2:37
  OK, let's shift gears. If direct stress is the push and
• 2:41
  pull, what about the slip and slide?
• 2:44
  Let's talk about shear. A completely different animal
• 2:47
  shear acts tangentially to a surface.
• 2:50
  And the key concept here is something called complementary
• 2:52
  shearing stress, right? It's essential.
• 2:54
  You just can't have shear stress on one plane by itself.
• 2:57
  It's physically impossible if you imagine a tiny little cube
• 3:01
  inside the material if there's a shear stress pushing the top
• 3:05
  face to the right. Then the whole cube would just
• 3:07
  start spinning, wouldn't it? It would start spinning.
• 3:10
  So to stay in rotational equilibrium, the material has to
• 3:13
  generate an equal balancing shear stress on the
• 3:16
  perpendicular planes. Ah, so a shear down one face
• 3:19
  means there has to be a shear sideways on another.
• 3:22
  Exactly. Tau Si must always equal Tau IX.
• 3:26
  It's a fundamental truth that underpins everything from
• 3:29
  torsion and shafts to, well, complex beam loading.
• 3:32
  Which brings us to the really complex stuff.
• 3:34
  What if you have a general 2 dimensional stress system?
• 3:37
  You've got direct stresses in the X&Y directions, plus that
• 3:40
  sheer stress Tau Y the materials being pulled, squashed and
• 3:43
  twisted all at once. How do engineers even begin to
• 3:46
  find the maximum stress in that mess?
• 3:49
  The math seems like it would be a nightmare.
• 3:51
  Oh, it is, and that's why the idea of stress transformation is
• 3:54
  so powerful. We know that if we could just
• 3:56
  rotate our little elemental cube inside the material, there must
• 4:01
  be two specific planes exactly 90° apart where the shearing
• 4:05
  stress is 0. The shearing stress just
• 4:08
  vanishes completely. It vanishes.
• 4:10
  It's the cleanest way to look at the forces.
• 4:12
  We call those the principal planes.
• 4:13
  And the stresses on them are the principal stresses.
• 4:16
  That's right, Sigma one the maximum direct stress and Sigma
• 4:19
  2 the minimum. When you're analyzing for
• 4:22
  failure, these are usually the first numbers you look for.
• 4:25
  They tell you the absolute peak direct stress that the material
• 4:28
  is feeling anywhere. But like I said, calculating
• 4:31
  them with trigonometry all the sines and cosines sounds awful,
• 4:35
  which is where a famous graphical shortcut comes in.
• 4:38
  And thank goodness for it. This brings us to one of the
• 4:41
  most, I think, elegant tools in all of engineering.
• 4:44
  More circle of stress. So instead of pages of equations
• 4:48
  you. Just plot 2 points.
• 4:49
  You take your known stress states, your Sigma X with Tau
• 4:52
  PSI and your Sigma Y with negative Tau PSI.
• 4:54
  Plot them and then you just draw a circle that connects them.
• 4:56
  And suddenly every possible stress state on any plane is
• 5:01
  right there on that circle. It's an incredible visualization
• 5:04
  tool. The coordinates of any point on
• 5:06
  that circle tell you the direct and shear stress on any inclined
• 5:10
  plane you. Want and the radius of the
• 5:11
  circle. Is the radius, is your maximum
• 5:13
  shearing stress Tau Max. It's that simple.
• 5:16
  So you can instantly find the maximum shear, and you know what
• 5:19
  happens on planes that are 45° away from those principal
• 5:22
  planes? Yeah, it's an immediate check
• 5:24
  for shear failure. Exactly.
• 5:26
  Now here's a critical little we tell a caveat that really
• 5:29
  separates a you know, a good analysis from a safe one.
• 5:32
  We always simplify things to 2D, but we live in a 3D World.
• 5:37
  Even if we have our circle, we find our Tau Max.
• 5:39
  We have to remember the third principle, stress Sigma 3, which
• 5:43
  might be 0. And why can't we just ignore
• 5:45
  that? Because failure by shear depends
• 5:48
  on the largest difference between any 2 principal
• 5:50
  stresses, the true maximum shear might actually be between Sigma
• 5:53
  1 and that zero stress from the third dimension.
• 5:56
  That 3D reality can be what actually causes the material to
• 5:59
  yield. That's a huge insight.
• 6:02
  Even in a 2D problem, you have to think in 3D for safety.
• 6:06
  OK, we've been focused on internal forces.
• 6:09
  Let's pivot to the structural response itself.
• 6:11
  Let's talk stiffness and deflection.
• 6:14
  Why is calculating how much a beam bends just as important as
• 6:18
  making sure it doesn't break? It really boils down to
• 6:21
  serviceability and stability too.
• 6:24
  I mean, you can design A floor beam that's strong enough, but
• 6:26
  if it sags too much the ceiling plaster underneath cracks.
• 6:30
  The floor feels bouncy it. Fails its function even if it
• 6:33
  hasn't broken. Precisely.
• 6:34
  And beyond that, you actually can't solve statically
• 6:36
  indeterminate problems without considering deflection.
• 6:39
  Right, those are the ones where you have too many supports, too
• 6:41
  many unknown forces for basic equilibrium equations to handle.
• 6:45
  Exactly. A beam on three supports is a
• 6:47
  classic example. You can't solve it with forces
• 6:50
  alone. You have to use the fact that
• 6:52
  you know the beam's deflection at those supports you bring in
• 6:55
  deformation analysis. So we need a way to calculate
• 6:57
  the exact deflected shape. Where does that begin?
• 7:00
  It starts with the fundamental equation of elastic bending.
• 7:04
  The core relationship is that the bending moment M at any
• 7:08
  point along a beam is proportional to the curvature of
• 7:11
  the beam at that same point. And a constant of
• 7:14
  proportionality is that key property, the flexural stiffness
• 7:18
  E times I. That's the one.
• 7:20
  E is Young's modulus. That's the material property.
• 7:23
  And I is the moment of inertia which is all about the cross
• 7:26
  sections shape. You got it.
• 7:28
  And that gives us the famous differential equation EI times.
• 7:32
  The second derivative of the deflection equals the moment M
• 7:35
  You integrate it once to get the slope, integrated again to get
• 7:38
  the actual deflection. We see those nice clean formulas
• 7:41
  and textbooks for simple cases, but I imagine when the loading
• 7:44
  gets messy or the geometry is complex, integrating those
• 7:47
  equations over and over is. It becomes a huge chore, or even
• 7:51
  impossible. And that brings us to one of the
• 7:52
  most powerful, most elegant tools in all of classical
• 7:57
  mechanics. It's the real bridge to modern
• 7:59
  methods, energy methods. Specifically, the principle of
• 8:03
  virtual work. Now that sounds almost
• 8:05
  philosophical. What is virtual work?
• 8:06
  It is a bit of a mind Bender, but it's a genius concept.
• 8:10
  Imagine a structure is in perfect equilibrium under its
• 8:13
  reel loads. The principle says that if you
• 8:15
  then apply a tiny imaginary virtual displacement to it, the
• 8:20
  total work done by all the forces, internal and external,
• 8:23
  has to be 0. OK, so how does that help us
• 8:27
  find a real deflection? We use it like a trick.
• 8:30
  Let's say you want to find the deflection delta at one specific
• 8:33
  spot on a complicated frame. You apply a unit virtual load, a
• 8:38
  pretend force of one at that exact spot in the direction you
• 8:41
  care about. A fake force.
• 8:43
  A fake force, then you say. The external work done by this
• 8:46
  fake force, which is just one times the unknown deflection
• 8:48
  delta, must be equal to the internal work done by all the
• 8:52
  real bending moments inside the structure.
• 8:54
  So instead of solving a massive equation for the whole beam
• 8:57
  shape, you use an energy balance that's focused only on the one
• 9:00
  point you care about. Exactly.
• 9:02
  It turns these impossible calculus problems into integrals
• 9:05
  you can actually solve. It's an indispensable tool for
• 9:08
  conflicts, frames, trusses, curved beams, things you
• 9:11
  couldn't solve otherwise and. What about complementary energy?
• 9:14
  Is that a similar idea? A very.
• 9:16
  Close relative. It's another powerful shortcut.
• 9:18
  Complementary energy let's you find the deflection by just
• 9:22
  taking the partial derivative of the structure's energy with
• 9:26
  respect to the load causing the deflection.
• 9:28
  It's just another brilliant way to get to the answer.
• 9:31
  It's amazing that these tools could handle such complex
• 9:34
  frames, but even they have limits.
• 9:37
  When you get to, say, a modern skyscraper, this feels like the
• 9:41
  point where we need the digital leap.
• 9:43
  That's exactly right. When you have thousands of
• 9:45
  elements, thousands of degrees of freedom, you can't do but by
• 9:49
  hand, you're solving thousands of simultaneous equations.
• 9:52
  And for that you need the language of computers, matrix
• 9:55
  algebra. How did matrices become the
• 9:57
  language of structural analysis? Well, a matrix is just a really
• 10:00
  organized array of numbers, and the operations you can do with
• 10:03
  them, like multiplication, let computers process huge amounts
• 10:07
  of data incredibly efficiently. This was the backbone for the
• 10:10
  matrix displacement method, which really changed everything
• 10:13
  back in the 1940s. And it started in the aircraft
• 10:15
  industry, right? Why there specifically?
• 10:17
  Speed and weight. Aircraft have to be strong, but
• 10:21
  they have to be ridiculously light.
• 10:23
  They needed a fast, reliable way to analyze these really complex,
• 10:28
  highly indeterminate structures where every single Oz matters.
• 10:31
  And the core idea is to break the whole structure down into
• 10:34
  simpler pieces. Exactly.
• 10:36
  You treat the whole thing as a complex spring system, and then
• 10:39
  you subdivide it into simple elements like individual rods
• 10:42
  and a Truss. The analysis then just focuses
• 10:45
  on the connection points, the nodes.
• 10:47
  And you build a massive system of equations based on how those
• 10:49
  nodes connect. Correct.
• 10:51
  You establish the force displacement relationships at
• 10:53
  every single node, and you assemble it all into one giant
• 10:56
  master equation that's defined by the system stiffness matrix,
• 11:00
  which we just call K. The computer solves that one
• 11:03
  huge equation and poof, you get all your displacements and
• 11:06
  stresses. And that matrix displacement
• 11:08
  method is the direct ancestor of the tool that every engineer
• 11:11
  uses today, the finite element method or FEMME.
• 11:15
  Femme is just the modern commercial application of that
• 11:18
  method. It's the most powerful tool we
• 11:20
  have. It can analyze anything from the
• 11:23
  tiny legs on a circuit board to the huge drilling rigs out in
• 11:27
  the ocean. And it shows that whether you're
• 11:29
  using a pencil and virtual work or a supercomputer, the
• 11:33
  underlying physics of equilibrium and compatibility
• 11:36
  are exactly the same. So we started with a simple bend
• 11:39
  and a column, visualize the stress with a circle, calculated
• 11:43
  the sag with an energy trick, and we've landed at the
• 11:46
  foundation of all modern computer aided engineering.
• 11:50
  What a journey. It is, and it brings us all the
• 11:52
  way back to the material itself. I mean, this whole discussion
• 11:55
  has been about elastic analysis, where stress is proportional to
• 11:58
  strain, but you have to remember how ductile materials like steel
• 12:02
  actually behave. Right.
• 12:03
  They don't just snap at the yield point.
• 12:05
  Not at all. They absorb A tremendous amount
• 12:07
  of work through plastic deformation before they actually
• 12:10
  break. That ability to absorb energy is
• 12:12
  a critical reserve of strength. Which raised the final thought.
• 12:15
  We've been focused on keeping structures within their safe,
• 12:18
  elastic limits. We have.
• 12:20
  But here's the thing, they're same principles of virtual work
• 12:23
  that we use for elastic deflection.
• 12:25
  They can also be used to predict the plastic collapse of a
• 12:28
  structure. So here's the question, why is
• 12:32
  knowing that collapse load often a more realistic and maybe even
• 12:35
  a more reliable estimate of a structures true safety than just
• 12:39
  checking its working stress limits?
• 12:41
  Because even if one part of the structure starts to yield, the
• 12:45
  overall collapse load tells you how much reserve strength the
• 12:47
  entire system has before a catastrophic failure mechanism
• 12:51
  actually forms. It's a clearer picture of real
• 12:54
  world safety than just looking at the first point of yielding.
• 12:56
  Exactly that structural reserve strength is what holds the built
• 13:00
  world together. Something to think about as you
• 13:01
  walk through a world built on stress, strain, and some truly
• 13:04
  beautiful mathematics.