Gauge blocks are hardened steel blocks lapped so flat and parallel that two pressed face to face stick together, and a small set stacked in combinations builds almost any length. A micrometer cannot certify them to the level they are made to, so their accuracy has to come from comparison: a set with redundant sizes checks itself, because different stacks of the same total must agree and any difference exposes a bad block.
Part of Precision Tool Making
Making, stacking and checking the lapped reference blocks against which micrometers, lead screws and thread work are verified.
A Block Is Only As Good As The Way You Check It
The carrier gives a recipe for making blocks. Cut oversize blocks from high-carbon tool steel, harden and temper to 60 HRC, rough-grind to within 0.05 mm of target, then lap on a cast-iron plate with 5 micrometre aluminium oxide, measuring against a calibrated micrometer, until each block reaches target +-0.002 mm. Finish with 1 micrometre diamond paste. A set of 10-20 blocks in sizes such as 1.0, 1.5, 2.0 and 2.5 mm can be combined into any dimension.
There is a problem in that recipe. Micrometers says a good micrometer reads within +-0.003 mm of a gauge block (Micrometers), and its smallest step is 0.01 mm. An instrument of that accuracy cannot certify a block to +-0.002 mm. The number in the recipe is a target; verifying it takes a method that does not depend on the micrometer.
Wringing Is Both The Use And The Test
Well-lapped blocks wring: pressed face to face with a sliding motion, they hold together by molecular adhesion. The carrier calls this the ultimate test of flatness and finish. A block that no longer wrings is damaged or worn, and a block that wrings with its neighbours across the whole set has faces at least flat and clean to the degree the adhesion needs. Wringing tests the faces; it does not tell you the thickness.
Thickness is established by comparison. Take any two stacks with the same nominal total, such as 1.0 + 2.5 and 1.5 + 2.0, both 3.5 mm. If the set is true, the two stacks agree. If they do not, at least one of the four blocks is off, and by comparing more pairs of stacks you can find which. A set with redundant sizes thus does for length what three plates do for flatness: the errors that cannot be shared across all the comparisons are the ones that show.
What no comparison can supply is absolute scale. All the blocks could be off by the same fraction and every stack would still agree. The absolute size comes from one reference length, the master bar checked against a pendulum (Standard Units), and from the fact that you can remake it.
What You Need
- High-carbon tool steel blocks (Tool Steel) and the means to harden them (Hardening, Tempering)
- A cast-iron lapping plate, flat, and graded abrasive (Lapping for Ultra-Precision)
- A comparator, such as a dial indicator on a stand, resolving finer than your target
- A reference bar of known length, and temperature that is steady
Method: Make And Check A Set
Step 1: Cut, harden and temper the blocks to about 60 HRC.
Step 2: Rough-grind to within 0.05 mm of target.
Step 3: Lap on the cast-iron plate with fine abrasive, finishing with a still finer one.
Step 4: Compare, don’t just measure. Stack blocks to equal totals and compare them with the comparator. Record the differences.
Step 5: Solve for the bad block by comparing more stacks, and lap that block further.
Step 6: Keep a master set separate from the working set. Use the working set daily and check it against the master. A set used every day wears; the master, used rarely, is the thing you compare against, and it is itself checked against the redundancy of its own stacks.
Worked Example
Building a length: with an illustrative set that includes 1.03, 1.2, 5 and 30 mm blocks, a 37.23 mm stack is 1.03 + 1.2 + 5 + 30 = 37.23. Choose the last decimal first, then the next, then the whole millimetres, which keeps the number of blocks to four.
Redundancy: nominal 1.0 + 2.5 and 1.5 + 2.0 are both 3.5 mm. If the comparator shows the second stack 0.004 mm longer, the four blocks together carry at least 0.004 mm of disagreement; comparing other stacks of equal total that share some of the same blocks isolates it.
Temperature: steel grows 0.012 mm per metre per degree (Precision Measurement), so a 25 mm block grows 0.012 x 0.025 = 0.0003 mm per degree. A block warmed by hand 5 degrees above the others is 0.0015 mm longer, which is 75% of the +-0.002 mm target. Let blocks reach a common temperature before comparing, and handle them as little as possible.
Common Mistakes
- Certifying a block with a micrometer to a tolerance the micrometer cannot resolve.
- Using one set for everything. It wears; keep a master.
- Comparing warm blocks.
- Dirty wringing. Grit damages both faces; clean before every wring.
Safety
Hardening, lapping compound and diamond dust
Blocks are hardened by heating to red and quenching: use tongs, an eye shield, and keep oil away from open fire. Lapping compound and diamond paste are fine abrasive that is easy to inhale; lap wet, do not blow the dust away, and wipe with a damp cloth (Dust and Respiratory Protection). Hardened blocks are brittle and chip if dropped.
What’s Next
- Learn next: Lapping for Ultra-Precision - the process that makes them
- Combine with: Master Lead Screw - checking a screw’s travel against stacks