4 ms·
> The difference in length between the two rows is 100 times that 1um tolerance. Doesn't tolerance work both ways? One brick is 1um longer than it should, anot
by 988747 4y ago
> The difference in length between the two rows is 100 times that 1um tolerance.
Doesn't tolerance work both ways? One brick is 1um longer than it should, another one is 1um shorter - in the row of 100 bricks those differences will mostly cancel each other out, won't they?
- MagnumOpus 4y agoNot necessarily -- for instance, it could be that the mould for 1x1 bricks is too short and the mould for the 2x1 bricks is too long. Now if someone builds pillars from 1x1 and 2x1s respectively, the differences accumulate.
- rkangel 4y agoSee my sibling comment about tolerance analysis. Also, it's not just about variance between different bricks (1x1 vs 2x1) it's about variance between different parts out of the same tool (the correct term for a mould is a "tool"), and variance between different tools producing the same brick. Some really subtle effects affect the size of parts if you care down to the level of 1um - all sorts of temperature variation, pressure, humidity, colour mix etc.. Lego's skill is the ability to control or compensate for all of these confounding factors and get consistent results, regardless of age of mould tool, manufacturer of machine, ambient conditions etc.
- rkangel 4y agoThere are a number of different approaches to tolerance analysis. I kept it to simple arithmetic analysis for the explanation, but yes there are other approaches too: https://en.wikipedia.org/wiki/Tolerance_analysis https://en.wikipedia.org/wiki/Tolerance_analysis