For a normal ruler, any accuracy or inaccuracy is just inherited from the machine that made the ruler. For a plastic ruler, that would be the injection molding machine that made it, though in that case the person designing the mold needs to account for the plastic shrinking as it cools. A wooden ruler might also shrink and expand with humidity. Even metal will shrink and expand with temperature, so length standards also specify temperature.
If you’re using a ruler, I’d be confident the measurement is accurate to within 1mm or 1/16″. If you’re using calipers, maybe about 0.1mm or about 4/1000 of an inch. If you need better than that, you use a micrometer. If you need better than THAT you use gauge blocks, a surface plate, and a test indicator.
For more accurate measurement tools, there’s a chain of comparisons called traceability. At the very foundation you look at how length is defined in the first place, and it’s defined by fixing the speed of light. What this means is you can use laser interferometry to measure your physical standards extremely accurately.
Adding to the answers: there used to be a “standard bar” for measurements. If you go to Bern, they have in the street the metal bars every mason used for copy and make their own rulers to build the city (it’s one of the “new cities” by the Zahringen counts from scratch).
Obviously this varied from town to town and country to country, hence why they came up with the metre… whose definition was wrong, …several times. But they had in Paris a “standard bar” of a metal that didn’t expand or contract “much” with temperature. This bar was copied and distributed. factories use stable materials for the copy. The final product wouldn’t be that accurate but “good enough”.
Nowadays these “geniuses” define the metre as something no one will be able to reproduce. That’s why I, as an European architect, still prefer old measurement systems, using them everyday, since I like the flexibility. If a space needs to be 12 feet in length I don’t care if it will translate to 3.82, 3.60 or 3.42 m; your brain doesn’t care and a house it’s not going to explode because of that.
There is an organization in a lot of countries that defines standards and specifies measurement technologies to insure those standards are met. In the US, for example, that organization is the NIST (National Institute of Standards and Technology). You can hire [calibration services](https://www.nist.gov/calibrations) that will check the measurement on a piece of manufacturing or inspection services to insure that your machine that is manufacturing the parts is accurate to within any tolerance you might specify.
Most high volume manufacturing processes are extremely repeatable within a tolerance, and once you’ve inspected it once, you can be assured that future parts will be the same (within the tolerance of the process).
There are standardised measuring for all measurements – weight, length and volume, and everything is based on those. However, as this [pic](https://www.chicagotribune.com/resizer/CJfjkvsfuJMeF0_WiY3QatN5lk8=/415×233/top/arc-anglerfish-arc2-prod-tronc.s3.amazonaws.com/public/OL6BYXWYXZAZDHZSOBD4HO3JSM.jpg) can attest, its all a load codswallop, and none have any idea what the hell they’re doing.
Short answer: certified [gage blocks](https://www.mcmaster.com/2210A85/)
They are traceable, certified accurate, and expensive. But with a full set, you can replicate any measurement required, and create/check a custom gage for that dimension that you can use on your workpiece. You can also calibrate your in-house measuring tools (calipers, micrometers, and the like) using your certified gage blocks.
They come certified to be accurate within a set tolerance. They periodically need to be checked to make sure that they still are within that range, and renew the certification. This is generally done by specialized calibration companies.
The other part is [Handbook 44](https://www.nist.gov/system/files/documents/2017/04/28/hb44-15-web-final.pdf), which gives acceptable tolerance standards for measuring devices. As long as the product falls within the tolerance range, it’s good to go. Linear measures are in Section 5.52, and a one foot ruler has a tolerance of 1/32”.
My time to shine!
There is an international standard called the [International System of Units (SI)](https://en.m.wikipedia.org/wiki/International_System_of_Units) which defines the base unit of measurement for physical quantities. Base SI units are kilogram (mass), meter (distance), second (time), ampere (electrical current), mole (amount of substance i.e. number of molecules), Kelvin (temperature), and candela (light). The definitions of the units are typically based in physical realizations of the unit… a real physical object or phenomena which represents a unit by definition.
An intergovernmental organization called the [International Bureau of Weights and Measures (BIPM)]
(https://en.m.wikipedia.org/wiki/International_Bureau_of_Weights_and_Measures) based in France manages the SI system and hold the prototype realizations of various units (the official measure). For example, BIPM used to house the official kilogram defined as a platinium-iridium cylinder of a standard height and diameter. Governments from around the world duplicate this prototype cylinder and it is compared to the official to ensure calibration. This is used within countries to calibrate all other weight measuring devices. In the US, there is a federal agency and corresponding state agencies where businesses ensure their kilogram measure is the same as the state one, which is the same as the federal one, which is the same as the official prototype in France.
The same kind of definition and realizations exists for the meter and can be standardized and calibrated similarly. The standard has been for some time to do it this way, but in recent years, these definitions are shifting to physical phenomena rather than objects such as a meter now being defined as the distance light travels in 1/299792458 seconds and mass now being defined by some crazy stuff about Planks constant. It s wild.
Latest Answers