Mechanical Engineering - LeFreq/Singularity GitHub Wiki

This document is here because good mechanical engineering principles can be applied to computer science problems. Instead trading more structure to use less energy, you can trade more memory use for less CPU, for example. (These maxims [to] from _? Guide to Mechanical Science, c1925).

FIX: ADD to solut9ions:

• redundancy: either in number or pounds (use another set of ropes for safety in bridge and use 2x thicker lumber rather than stay close to tolerance)
• orthogonality: utilize something that is non related to an existing structure (teeth for a gear mechanism, cross-members for a bridg3, cotter pins, hinges?).
• To achieve more effiency when covering additonal constraints (boundary cases in a sorterm for exmaple), use esxisting inefficiencies in the existing solution to achieve the goals so you don't have crufty solutions.

Since most every good design in mechanical engineering requires a bit of architectural or design finesse. Call it mechanical science (compare to computer science), Good design is more than engineering. Automotive architecture vs. automotive engineering is one example. Mechanical science is a combination of mastering force with statics. Mechanical science is the ability, using only simple materials and forces, to solve arbitrarily-complex problems.

• Maxim #1: No matter what the problem or challenge, unless it involves generating light, there is always a mechanical solution to it. If you step into using electricity, you've stepped away from mechanical science and into electrical engineering. If you've stepped into using chemicals, beyond setting stuff on fire, you've moved to chemical engineering. As for mechanically generating light, the best you can hope for is striking objects together like cavemen.
• Maxim #2: There is always a elegant solution to any mechanical science problem, if you look for it. The elegant solution is all of the following: efficient (work output approx= energy input), precise (getting to the core of the exchange between motion vs. statics), minimal (simple, low material use), and correct (does the specified job).
• Maxim #3: One can trade any energy-use problem into a structural-static problem or vice versa. Each of these should be considered equal constraints/solutions. Further, one can lower the usage of one arbitrarily low (but not 0) by increasing the other. If you have no time constraint on when it gets done, you can often get nature to do the work with 0 energy consumed. (XXExample: if you don't want to change your rolling logs so often transporting that heavy rock, add a larger diameter structure on both ends and it can roll at 1/10th the speed rather than 1/2.)
• Maxim #4: Any mechanical science challenge can be met with one of five solution-paths:
  1. add orthogonality. Add a layer of indirection or an independent member: breaking down something into smaller pieces that connect to one another. (XX? a nail into a screw?) (XXExample: you want to keep the strength of steel but you want the flexibility of leather: ring mail).of the design or turning a linear problem into a rotational one.
  2. Add redundancy: adding redundancy to make things stronger or more reliable.
  3. Better material
  4. Greater precision
  5. As a last resort, using more power or adding more structure, but see Maxim #4.
• Maxim #5: If you've ran out of time or budget (or interest), there's always the hidden sixth option: shift control to the user.
• Maxim #6: One can always get arbitrarily close to 100% efficient. All the time. But no one knows how long or how much money it will take to get there... An engineer should aim for at least one 9 (90% or better), otherwise they're just enthusiasts of the art.
• Maxim #7: Any design that requires lubrication (excess means beyond ~2psi or more than 108^oF) is generally bad engineering, making cheap choices that will effect the whole life of the product. It's poor engineering, because nature is doing the work, not the engineer. Any such designs can be replaced by employing mechanical science options of Rule #2.
• Maxim #8: Design deficiencies are heralded by one of 4 measurable factors:
  1. visual or audible noise (Examples: excess mass/movements, squeaks, clatter/knocks, pings, loud exhaust)
  2. vibrations or turbulence (solutoin #6: symmeteries : counterbalances, rifling of exhaust ports)
  3. heat or light (solution #7: better transformation/interface between energy->structure; i.e. delivery of power source (fuel or human) or less friction of moving surfaces)
  4. odors or waste by-products (smoke, unburned hydrocarbons; solution #8: better direction of energy during use)
• Maxim #9: Finally, master your tools or they will master you. People get hurt by machines when they don't understand them well. An engineer should try to understand everything they use and continually stay on top of their field.
• Most every mechanical contraption can be diagnosed (what is going wrong?) and queried (what is it`s design?) with stethoscope. A machinist should always have one available.
• Stay pure. The test of your skillz is how little you depend on other disciplines to do your bidding: electrical, chemical or otherwise to buffer your lack of mechanical engineering.

---

Mechanical Science, like computer science, is a vast field of study. The primary division is between two independent, yet complementary fields:

• Engineering: This can be divided into two camps: statics (like a bridge) vs. motion (like a flywheel). All solutions ultimately use a combination of these, but there are always two (and only two) elegant solutions that focus on either of these options (except in the boundary case of energy-delivery). An engineer or architect must seek them.
• Architecture: Interfacing with humans (ergonomics) and modularization. The latter deals with how best to design materials/components so that they can be re-used for many applications. A motor module that can be used in a VCR can be re-used in a washing machine. Once a customer knows exactly what they want (generally never), they can customize these modular designs into the perfect, application-specific design, usually saving a little (or a lot) on energy and space, at the trade-off cost.

Along side these is materials science: the science of making materials with the qualities you desire.

Around these are services that provide modification of the standard parts (drilling holes in a metal frame), repair folks that can fix the standard parts where modularization doesn't go(?).

---

A machine is solidity (or structure) + force resulting in some controlled (new) effect. (Note, this would make rubber bands machines.)

Basic machines from which one builds all compound machines:

• inclined plane (trades speed for less force)/(or just counters gravity)
• tooth and mesh: joinery, gear surfaces (counters conductance, adds friction)
• tension and solidity: rope, hose, steel plate (adds order through QM)

A universal ID, 4-4 alphanumeric digits, plus 4 more? at the end to specify variants of the same. A marker of some kind for meta products that support the modular system, and a checkbox that should be checked whenever a modification is made to the unit (does not include repairs with identical parts).

Enclosures need not be part of the inventory or modular, but help provide distinction at local economics and craftslan. Also adapters.

REgarding ergonomics: one should be able to accomplish any kind of work with no more than 4 knobs or levers.