Furthermore, if you google on “hot water heater” AND “water heater” you get about 2.4m hits, showing that the terms are used interchangeably. Plus if you google “hot water heater” and limit the search just to homedepot.com (ie your chosen cite) you get 137 hits, and not all of them to reviews etc written other than by Home Depot itself.
Mangetout, because it doesn’t boil.
Showing that both terms are in common usage. Yours rather more than mine, signifying little. I must say I get nearly 9m hits for “water heater”.
Hey look I don’t doubt that the term “boiler” is used to mean a domestic water heater. And by my own arguments above, if the term is in common usage that way then feel free.
But you asked why not, and I answered. 
A boiler is for providing heating for buildings.
I much regret interjecting up there in the thread. But no. A boiler (here, at least) is for heating water, for any purpose.
On the Googlefight, note that every page that uses the string “hot water heater” will necessarily also use the string “water heater”, not because the terms are interchangeable, but because the one is a substring of the other. By the numbers above, “hot water heater” is about 2.5 times more common than the unmodified “water heater”.
Don’t think that’s the case. It wouldn’t be the same as a COLD hydrostatic test where you keep pumping water in till the thing pops. That would be pretty boring, as you say!
Thing is, the water was at WAY above the boiling point for ambient pressure, so even without hullage space, when the tank let go a lot of the water would instantly flash to steam from the reduced pressure.
Back of the envelope calcs:
temperature of water-steam equilbrium at 350psi is 432 Farenheit, or 222[sup]o[/sup] C. That is 122[sup]o[/sup] C above boiling point at 1 bar, meaning that there was 122 calories per gram available for turning the water into 100[sup]o[/sup] C, 1 bar steam when the tank let go.
Heat of vaporisation of water is 539 calories/gram, so there was enough heat energy stored in the water for 23% of it to flash to steam when the tank let go. that’s 23% of the water volume expanding ~1600 times.
By comparison, a tank filled only with steam at 350psi and no water is at 24 bar pressure. At most, the steam will expand to 24 times the volume when the tank goes. (I know it won’t be isothermal. It’s an upper bound. Sue me!)
Hey, putting the semantic dancing aside and trying to get back to the actual question, here’s a WAG that is almost certainly not true and is giving the manufacturers of [hot] water heaters more credit than they deserve, but it might be fun to contemplate:
What if the heater was actually designed to fail at the bottom? It could be argued that that would be a last ditch safety measure. Look at the amount of power that an exploding heater wields. If the vessel could fail at any random point, chances are that the vessel would be driven wildly around through the building housing it, increasing the chances of causing injury to a person. But when the vessel fails at the bottom, as we saw, it’s driven straight up. Sure, anyone close by or directly above it is toast, but I’d maintain that from a strictly statistical standpoint, blowing up will cause fewer injuries than blowing sideways.
That’s just about possible. This kind of thinking can be seen in liquified flammable gas cylinders, where a “burst disc” is backed by a low melting point alloy. In a fire, the alloy melts, the disc bursts and you get a nice tornado of flame as the gas comes out. Not great, but preferable to letting the thing store up energy as heat in the liquid and then rupture in an uncontrolled way.
However, I consider it unlikely. The truth is, there’s absolutely no reason why these tanks should be able to hold more than a few bar of pressure, just enough to run a decent shower! Constructing the tanks with a weak spot that pops at 10 bar would be way safer than designing the things to hold 150 bar and let the bottom blow out first. Hell, a rupture disc at the top, backed with Newton’s metal (melts at 97 Celcius, 207 F) would solve all the problems by venting the tank before it even gets to boiling.
Any tank fabricated with an inverted head (as used by M.B.) WILL fail at that seam, if pressure acts equally on all area. The inversion creates a notch (stress riser) that puts the girth weld in shear. Had the head been stretched past its yield point it would exhibit torn metal, instead of the reversed wok look.
Matt,I don’t doubt your calcs.
Just for reader information and straight dope, the T/PR (thermal/pressure relief ) valve is designed to pop at 150 psi and/or 210F, for most domestic heaters in the U.S. There are strict requirements for its discharge piping. It is almost impossible to develop an air pocket at the top of the tank, and the heating system has its own safeguards independent of the T/PR valve.