The Surprising Problem With Measuring Anything

Main Ideas

• Contact measurement is inherently disruptive. A probe can dent or push a surface, and the warmth of a hand can expand a metal part by microns. A measuring instrument also depends on a reference that must itself be trusted.
• Mass production required repeatable, shared standards. Craftspeople once adjusted parts until they fit their intended counterparts. Interchangeable parts required measurements against independent references rather than relying on each part’s individual fit.
• Measurement depends on more than a fine scale. Resolution is how small a change an instrument can detect; accuracy is how close its reading is to the true value. A sensitive instrument may reveal tiny differences without providing an independently accurate measurement.
• The history of metrology is a search for better references and gentler probes. It progresses from master bars and mechanical gauges to gauge blocks, electronic comparators, automated coordinate machines, and probes that measure at the scale of atoms.

How Measurement Developed

From Fit to Common Standards

• Ancient Egyptians used a granite royal cubit as a master reference. Working cubit sticks were copied from it and checked periodically.
• For centuries, craftspeople generally made each part fit its intended partner. Industrial production created a need for parts that could be selected at random and still assemble correctly.
• In the 1780s, French gunsmith Honoré Blanc advanced this goal for musket locks. Armories used go/no-go gauges: one gauge had to fit a feature, while the other had to be unable to fit.
• These gauges could classify a part as acceptable or not, but could not report how far it missed the target. They also depended on copied masters that could wear down.

Creating References Without an Existing Master

• Henry Maudslay used flat iron surface plates as workshop references and built a bench micrometer, nicknamed the “Lord Chancellor,” that could resolve 1/10,000 of an inch.
• Joseph Whitworth’s three-plate method created a flat reference without relying on a pre-existing flat surface:
  1. Work two rough plates together, and they develop matching, opposing errors.
  2. Add a third plate and scrape high spots while alternating which pairs are worked together.
  3. Over repeated cycles, the opposing errors cancel, bringing all three plates toward true flatness.
• Whitworth argued that length should be defined by contact between flat faces, rather than by the distance between lines viewed through a microscope.
• His 1851 measuring machine used a fine screw and geared wheels to make extremely small adjustments. A thin “feeling piece” in the gap acted as a contact indicator: at the set point it remained suspended by friction, and backing off one division let it fall under gravity.
• The machine could detect changes as small as a millionth of an inch, including thermal expansion. But its screw could not guarantee absolute accuracy, so it functioned as a comparator.

Micrometers: Force and Alignment

• In 1848, Jean Laurent Palmer patented a practical handheld screw gauge. A 1 mm screw and a thimble divided into 20 parts allowed readings down to 1/50 mm.
• The later Brown & Sharpe micrometer used a screw and graduated thimble to make small spindle movements readable. The thimble’s larger movement amplified the spindle movement by roughly 60 to 1.
• That mechanical advantage also multiplied force: pressing too hard could flex the frame and dent the part. Ratchets and friction thimbles were added to limit and standardize measuring force.
• Abbe’s principle, formulated by Ernst Abbe, says the measuring scale should lie in line with the dimension being measured. A micrometer largely follows this principle; a caliper’s scale is offset from the part, so jaw tilt or rocking can create a reading error.

Gauge Blocks and Repeatable Calibration

• In 1894, Carl Edvard Johansson proposed replacing many dedicated gauges with a small set of precision blocks that could be combined to make many lengths.
• His first set, completed in 1896, used 102 blocks to produce about 20,000 sizes.
• Making stacks trustworthy required each block to be exceptionally flat, parallel, and accurate. Johansson finished them on a lapping wheel driven by a converted sewing machine.
• Carefully finished blocks wring together under light pressure, forming a stable stack through molecular attraction and a very thin film of oil or moisture.
• Gauge blocks also highlighted the effect of temperature: steel expands as it warms. The international reference temperature for length was set at 20°C in 1931.

Comparators and Surface Measurement

• Dial indicators use a spring-loaded plunger and gears to magnify small movements. They are first zeroed against gauge blocks; the blocks supply the reference accuracy, while the indicator makes small changes easy to see.
• The linear variable differential transformer (LVDT) replaced mechanical contact within the sensor with a moving iron core and coils. Core movement changes the electrical output, allowing highly sensitive, recordable displacement measurements.
• Surface texture can affect performance even when two parts have the same measured diameter. A stylus instrument converts the stylus’s vertical movement into an electrical signal and a roughness value; the later Talysurf system recorded a greatly magnified surface profile.
• Stylus probing also separated roundness from diameter. A slightly three-lobed pin can measure the same diameter at different angles yet fail to fit a truly round hole.

Coordinate Measuring Machines and Automated Probing

• As machining became faster and parts more complex, manual inspection became a production bottleneck.
• A coordinate measuring machine (CMM) measures points in space using a probe and scales on its axes. Software can fit planes, circles, and cylinders to the points and compare them with a drawing.
• Early CMMs had two key problems:
  • Their scales and probes were often separated, violating Abbe’s principle and making small carriage rotations create significant probe-position errors.
  • Manually pushed hard probes could bend or displace flexible parts, making readings depend on the operator’s touch.
• In 1972, David McMurtry developed the touch-trigger probe to measure flexible engine pipes. Its spring-loaded, six-contact kinematic arrangement returns to a repeatable position. When the stylus touches the part, an electrical change triggers the machine to record its coordinates.
• The probe applies a light, consistent force and acts as a repeatable switch. A computer corrects for the radius of the probe’s spherical tip, commonly made of synthetic ruby.
• Touch-trigger probes made fully automated measurement practical, including on CNC machine tools. They collect one point at a time.
• Scanning probes, introduced by Zeiss, instead maintain continuous contact using flexures and displacement sensors, collecting many points per second along curved surfaces.

Correcting Machine Errors

• A three-axis measuring machine can have 21 significant error sources: six for each axis, plus three errors in the squareness between axes.
• Rather than attempting to eliminate every error through extreme stiffness and precision, Robert Hocken’s group at the U.S. National Bureau of Standards mapped repeatable errors using laser interferometers and electronic levels.
• The machine then applied those error maps in software. This made it possible to use lighter, faster machine structures while maintaining high accuracy.

Measuring Material Properties and Atomic-Scale Surfaces

• Indentation tests measure material hardness by pressing a known shape into a sample and examining the resulting impression. For very thin coatings, the indentation must be too shallow to measure optically.
• Instrumented indentation addresses this by continuously recording the force and depth of a diamond tip. The depth under load indicates hardness; the material’s spring-back during unloading reveals stiffness.
• The Topografiner, developed by Russell Young, used a sharp conductive tip held just above a surface. A feedback system adjusted the tip to maintain a current, and its movement mapped the surface without contact.
• In 1981, Gerd Binnig and Heinrich Rohrer developed the scanning tunneling microscope (STM). It measures the electrical current caused by electrons tunneling across a tiny gap. The current changes sharply with distance, allowing images of individual atoms. The STM requires a conductive surface.
• The atomic force microscope (AFM), developed by Binnig, Calvin Quate, and Christopher Gerber, uses a very soft cantilever and a tiny tip to measure force on a surface. Its extremely light contact allows it to examine nonconductive materials, including polymers and living cells.
• Probe technology can also manipulate matter: in 1989, Don Eigler used an STM tip to arrange xenon atoms.
• The video closes with a return to the idea of a reference: in 2019, the spacing between atomic planes in silicon was formally adopted as a way to realize the meter at very small scales.

Speakers and Sources Featured

• Speaker: One continuous narrator; the subtitles do not identify the narrator by name. No interview or second speaker is indicated.
• People discussed or credited: Honoré Blanc; Henry Maudslay; Joseph Whitworth; Jean Laurent Palmer; Joseph Brown; Lucien Sharpe; Ernst Abbe; Carl Edvard Johansson; John Logan; Ernest Abbott; David McMurtry; John Deer; Robert Hocken; Johan Brinell; Warren Oliver; George Farr; Russell Young; Gerd Binnig; Heinrich Rohrer; Calvin Quate; Christopher Gerber; and Don Eigler.
• Institutions and organizations named: the University of Michigan; Taylor Hobson; Ferranti; DEA; Rolls-Royce; Renishaw Electrical/Renishaw; Zeiss; the U.S. National Bureau of Standards; IBM’s Zurich laboratory; and the Carl Gustaf rifle factory. The video also refers to unnamed researchers in Sweden in 2013.