> ...floating forms that nest neatly in the rounded curves of modern devices. These clearly defined shapes feel easy to tap and are designed to relate to the natural geometry of our fingers...
This reminds me for some reason of my preferred answer to the Microsoft interview question "Why are manholes round?" A: Because the average cross-section of a human being is roughly circular.
From the perspective of "design the manhole cover first", the reason is that manhole covers aren't just round — they're (very subtle) domes.
A dome is:
1. the best shape for taking stress from very heavy trucks putting all their weight on them without the manhole itself gradually bowing, and
2. is best at transferring that stress equally into the manhole wall (cast concrete cylinder) itself. (A square manhole + manhole cover would disperse force unfairly, potentially gradually cracking the manhole walls / requiring stronger walls. A flat circular manhole would disperse force upon the center of the manhole equally onto the manhole walls, allowing for lower-material-cost manhole walls. A domed manhole cover additionally disperses force from most points on the dome equally into the manhole walls — important, as vehicles won't necessarily be driving over the exact center of the manhole!)
...but really, this is the wrong direction to work in. The original reason manhole covers are round, is simply that the walls of a manhole are best made round, for the same reason drink cans and barrels are best made round: a closed cylinder is great at taking compressive force from a lid above; passing it through as soft, equal tensile force through its walls without buckling strain; and then turning that force back into an equal compressive force on the floor / subsurface.
Most manholes are generally small closed cylinders acting as maintenance areas for nearby pipes, with the pipes coming in through the sides of the manhole walls, and the concrete bottom floor of the manhole resting upon compacted earth.
In this situation, any shape for the manhole other than a cylinder — if driven over for years/decades by cars — would gradually pound the uneven force acting upon the manhole's floor into the earth below, unevenly accelerating soil subsidence. Eventually, you've created a sinkhole below the road, right outside the manhole wall on one side.
As someone with a mechanical engineering background, all these clever answers to the question (e.g. "manhole covers are round to prevent the cover falling in") are strange. Covers are round because the hole is round, and the hole is round because a cylindrical shape is sturdy and prevents collapse. That's it.
That's not "it". A square cover would not be feasible, without a tether system to fish them out of the literal shit, after they fall in from a few inch misalignment.
The hole came first. They dig a hole, then they have a need to put a cover on it. Making a circular cover to fit over a circular hole is if anything cheaper and easier than making a square cover over an inscribed circular hole, at least when working with metal.
The hole drove the design of the cover, not the other way around.
In this hypothetical, presuming you're putting the manhole (subsurface conduit/plumbing maintenance access point) in after you've already paved the road: it's because the tool you'd use to cut a hole into pavement (i.e. a concrete saw) cuts straight lines — and it's easier to make a square/rectangle out of straight lines than a circle. And sure enough, whenever you see workers hacking up the road, they generally are cutting square holes.
At the end of this process, you have a square hole in the pavement, opening to a square excavation, bottoming out at a square concrete foundation, on which has been set a round concrete cylinder, which is then surrounded out to the edge of the square hole with packed earth.
Given this, you could equally-well finish this job either:
1. by placing a square of metal to fill the entire square packed-earth space you've constructed (as when bridging a pothole with a temporary steel surface plate);
2. or by first paving over the exposed packed-earth part, and then placing a circle of metal to cover only the manhole entrance itself.
...which is why people do justifiably ask why, in practice, we seem to always favor option 2 over option 1.
A manhole is a hole for men to go in, not a near-surface utility access. Those square cut quick access holes often do have square metal covers.
The manhole opens up into a vertical tube, often with a ladder built into the wall, big enough around for a man to descend into a subsurface structure. Hence, manhole.
Manhole covers are always round because the tube they connect to is round, and that tube is round for structural reasons.
You wouldn't for multiple reasons. The fact that the tube you're covering is round is one of those reasons, and the primary one, but not all of the reasons. A round cover would still be used if the tubes were square.
A manhole cover has two purposes: people to traverse, but much more importantly, vehicles to traverse.
You'll find square hole covers all over, they're not rare at all. The strength of the lid and its ability to be supported is not at all an issue on a single human sized hole - it's a thick, relatively cheap slab of metal over a small opening. They're the strongest surface of most roads, by a HUGE margin.
If that was the reason a Reuleaux triangle would have the same property while using (I think) less material for a given size of shaft (as long as the shaft can be triangular). https://en.wikipedia.org/wiki/Reuleaux_triangle
To explain myself better now I am awake and have had coffee: What makes a cover that cannot fall down the hole that it is covering? The width[1] of the closed curve has to be larger than that of the hole. A circle is easy to understand because it is a closed curve of fixed width, but a Reuleaux triangle is also a closed curve of fixed width, meaning if that fixed width is greater than the widest part of the hole, then there is no orientation of the cover in 3D that will allow it to fall down the hole. It’s easy to see if you take an equilateral triangle, circumscribe it with the smallest circle and then (mentally) construct the Reuleaux triangle that the Reuleaux triangle sits inside the circle and so would use less material.
[1] This is defined as the minimum perpendicular distance between parallel lines bounding the shape.
That answer would be akin to someone asking “why is the sun round?” and answering with “it’s roughly the optimal shape for viewing the totality of the sun”.
I guess it varies by country which I should have acknowledged. I was talking UK and EU and Africa which I am most familiar with, although I have also been to the USA and Asia and seen manhole covers there. You certainly get round manhole covers, but there are far more rectangular ones. Eg this kind of thing https://www.drainageshop.co.uk/polydrain-inspection-chamber/...
This reminds me for some reason of my preferred answer to the Microsoft interview question "Why are manholes round?" A: Because the average cross-section of a human being is roughly circular.