Levels and a level

[Tim+]

Humber bridge towers apparently diverge by 36mm due to curvature.

I wonder what the *smallest* man made object is that has to take account of the earth?s curvature?

Tim

[Andrew]

There are asian offshore airfields built on reclaimed land that have to be jacked up to keep them level. But the length of a runway must make that a tricky job.

[Andrew]

There are no flat walls or completely rectangular door frames in my 1976 semi. The lounge ceiling below the hot tank has deflected by about 20mm because they used pairs of 50mm beams but didn't bolt them together, just laid with a 10 cmm gap between them.

Naturally they sagged as they dried (3 inch cinder block upstairs internal walls with sand'n'cement base coat 'plaster'.)

[Johnny B Good]

I tried to find out if they'd done something similar when building the Severn suspension road bridge just over 50 years ago but the only reference to 'tower divergence' was in relation to 388 millimetres[1] of lean back to compensate for the weight of the suspension cables which has nothing to do with the curvature of the earth over the one mile separation distance.

I'd guess at something about a tenth of the size of the Humber Bridge since the resulting 3.6mm 'curvature of the Earth allowance would likely come within normal tolerances on a structure one tenth the size of the Humber bridge but that's just a guess.

Sadly, I've not been able to find an "Earth's curvature compensation" figure for the slightly shorter Severn bridge (1 mile span versus the Humber bridge's 1.4 mile span), just that 388mm[1] 'lean back' on the towers during construction to compensate for the suspension cable forces. It might have been mentioned in the Timewatch/Timeshift documentary on this record breaking bridge but I've not got the time right now to spin through the recording just to find out whether or not a figure was actually mentioned.

[1] AFAIR, the figure was 388mm BICBW.

[Tim+]

Thinking about it, you could argue that any structure built using plumb bobs automatically compensates for earth curvature. I suppose the question is, at what point does the discrepancy actively need to be compensated for?

Tim

[Rob Morley]

I fitted a mantle shelf and tiled down from there, only to find that the tiles on one side didn't quite meet the hearth because it was "level" with the floor. :-( Just painting the wonky grout gap made it blend in sufficiently that the unevenness isn't noticeable, but that's not a mistake I'll make again.

[Huge]

I'd wager no-one notices apart from you!

[Mike Clarke]

I wonder why they bothered to make them so precisely flat. Airfield runways don't need to be dead level. One end of the runway at Newcastle Airport is about 30 foot higher than the other.

[Dave Plowman (News)]

At least the dog always knows where to find his ball.

[newshound]

I think my worst is about 1 in 20. Otherwise, snap. Rotating head laser is excellent for this, I had a cheapie which has more or less died. Don't need to do much DIY now, trying to pluck up courage to invest £100

• in a better one.

[Johnny B Good]

It had nothing to do with making the road deck "100% flat". The 'lean back' was to compensate for the final cable loading so that the towers would become perfectly aligned to the vertical downthrust forces once construction had been completed. I dare say the same thing applied with the Humber bridge.

[Johnny B Good]

The precision will be a consequence of making sure the towers were precisely aligned to the downthrust forces in both cases. Assuming the Humber bridge's vital stats were a factor of 1.4 scale up of the Severn bridge design, my best guess for the divergence at the top of the Severn bridge's support towers would approximate to a figure of 18mm.

Are you trying to suggest that a mere 30% shortening of bridge length compared to its successor, the Humber Bridge justifies a disregard for the same level of precision that was deemed necessary for the 40% longer bridge?

You have to consider that the contractors working on the shorter bridge weren't just 'anyone'. I'd expect they were aiming to achieve a divergence that matched the theoretical divergence due to the curvature of the Earth as a consequence of taking the 'lean back' measures to make sure the towers were in perfect alignment with the down thrust forces once the cable loading was applied. Whether they were able to measure it to such accuracy and whether the measurement matched expectations is, for the time being, an unknown since I couldn't track down any mention of divergence due to the curvature of the Earth.

[Johnny B Good]

True, you seemed to have an obsession with 'flatness' in spite of my only wondering why I couldn't find a divergence figure to compare with the one for the Humber Bridge which Tim+ had managed to find.

It was not I who was obsessed with bridges needing to be "flat", it was you. I think you are conflating Brian's mention about 'flatness' and curvature of the Earth corrections with my not being able to find a curvature of the earth divergence figure for the Severn bridge to compare against the figure that Tim+ had found for the Humber bridge. I made no mention of 'flatness' until you introduced it into this sub-thread.

[Tim+]

To ensure perpendicular loading of the towers maybe?

Tim

[Rednadnerb]

and I seem to remember that the distance between the two towers of the Humber bridge is greater at the top than at the bottom, by about 1" I think.

[Johnny B Good]

According to Tim+, almost 50% more than that (36mm in fact).

[David]

Bought the Bosch level as recommended.

Looks good so far, except that the tripod will only reach up 4 feet.

Dug out my stock of tripods and I can gain about another foot, but I want a line even higher up the wall.

So now trying to work out what I can position opposite the wall which is tall enough AND broad enough to support the tripod so I can get the line where I want it.

I can tilt the head, but I get the annoying flashing which says "I am not level" and also I need to be sure that it is level from side to side.

Whatever do you do if you are trying to level a picture rail in a 18th Century room with 10 foot (or higher) ceilings?

Ah well, nothing is easy.

Cheers

Dave R

[Tricky Dicky]

The usual process is to use the laser line as a datum line and measure up from it to the height required. If you repeat the measurement in two places then you can use use a chalk line to get a continuous line at the height you need.

Richard

[Robin]

Some have an extending floor-to-ceiling pole which can be used to mount the level at the right height - as eg in

Some of us balance it on blocks a rung of a ladder leaning against the wall ;)

[Robin]

Sorry, that should have been "the opposite wall"