Excessive noise in tunnels

[zbang]

Air pressure and humidity, however compared to the mass of the tunnel and earth it probably hardly makes a difference.

Also just a theory.

Throwing out wacky ideas- any correlation with the tides and ground water level?

[jonathanburg]

Do the trains run on banded speed settings.

If the timetables are the same every day then necessarily with more loading there will need to be enough extra speed to compensate for longer loading/discharging of passengers.

I wonder if that is being directed into the automated operation after a loading review, which would account for the noticeable rampup and rampdowns (depending on which decision is made)?

[Tom]

I'm not sure how much the Victoria line can vary speeds these days, but it's not inconceivable that there are a range of 'standard' speeds.

[xtmw]

Has the Northern Line speed trial (whatever it's called) started yet?

AFAIK it was something about implementing TSRs to reduce rail corrugation?

[gefw]

Excessive noise initiating from the train wheels moving on rails appears to be rather complicated and not sure if this is fully understood or recorded accurately.
I suggest it can be categorised/explained in several ways.

Method 1 Airborne or Ground transmitted:-
Airbourne is generally higher frequency (in the audio range) mainly heard on the train/platforms and unpleasant even for short spells.
Ground based is generally lower frequency because it travel from track through the ground and building foundations causing "background but persistent annoyance particularly when trying to sleep".

Method 2 source of noise:-
1) Metal on metal abrasion. This screeching occurs- typically on curved track where if one wheel grips the rail, the other wheel/rail grinds on the rail running surface or wheel flange on the inner side of running rail or check rail. Generally controlled by rail head & wheel profile, rail cant and lubrication (more fashionably called friction modification).
2) Resonance/ringing. Squealing caused by movement of the train instigating a resonance of the rail (like rubbing a musical instrument string). This appears to be a function of rail type, chair spacing, rigidity of chair mounting, longitudinal stress in rail and train speed.
3) Rail head corrugation, This rumbling is caused from lengths of roughness that has established on the running rail head. Roughness is often self perpetuating because of consistent train speed (in ATO). Can originate from track features that cause the trains to bounce laterally or vertically affecting the grip or "zig zagging" (hunting) on the rail head. Also possible "mini slides" or "Spins" due to heavy acceleration/braking or frequent/consistent changes in motor demand by ATO driving and/or Traction supply disturbances particularly associated with Traction rail pick up/train shoes.

[Tom]

There was a very good presentation to the PWI London Section on this subject a few years ago.

[zbang]

Of possible interest-
University of Huddersfield Institute of Railway Research
research.hud.ac.uk/institutes-centres/irr/

[gefw]

Dec 30, 2023 22:44:46 GMT xtmw said:

Has the Northern Line speed trial (whatever it's called) started yet?

AFAIK it was something about implementing TSRs to reduce rail corrugation?

This trial is mentioned in a very recent answer from TfL to London Assembly, It confirms the area is Tufnell Park and Kentish Town and says it relates to rail corrugation but does not confirm whether started or timescales/plan but promises a final report.
Does anyone know when it was last ground or next grind planned?

Jan 8, 2024 21:43:53 GMT zbang said:

Of possible interest-
University of Huddersfield Institute of Railway Research
research.hud.ac.uk/institutes-centres/irr/

Not sure of their involvement so far but they may pick up the "open challenge" referred to in the recent London Assembly question/chaser to TfL
www.london.gov.uk/who-we-are/what-london-assembly-does/questions-mayor/find-an-answer/tube-noise-29

[gefw]

Ian visits has done an article which follows on from the recent London Assembly/Mayor question time, where this is a regular issue raised by certain members

www.ianvisits.co.uk/articles/tackling-the-tubes-roar-tfls-battle-against-tube-tunnel-noise-69555

Unfortunately the latest info/plans regarding Grinding, Pandrol rail fixing replacement, current noise levels etc appear not to be publicly available - would be good to know if the mitigation measures (which are expensive) are actually effective.
Interesting comment that the recent Northern line extn also suffers high levels of in train noise ?

[xtmw]

Tube noise hotspot image. No highlighted dot between Stratford and Leyton on the Central Line !

Don't know how to resize an image on here so apologies if it's abit big

[zbang]

Jan 24, 2024 14:30:15 GMT gefw said:

Unfortunately the latest info/plans regarding Grinding, Pandrol rail fixing replacement, current noise levels etc appear not to be publicly available ...

Current noise levels are entirely public , all you need is a sound pressure level (SPL) meter and there are apps for both apple and android phones that work reasonably well. These may not be calibrated, but IME are usually within a couple of db, and that's close enough for most things.

You can purchase a hand-held SPL meter for £20-40 on amazon.

[gefw]

Jan 24, 2024 17:34:27 GMT xtmw said:

Tube noise hotspot image. No highlighted dot between Stratford and Leyton on the Central Line !

Don't know how to resize an image on here so apologies if it's abit big

Thanks - this is the same as in the Ianvisits report - But unfortunately no info what a "Noise Complaint" includes , in the past it included complaints about nightime noise by maintenance staff or contractors (eg those on the Met Wembley Pk to Harrow + on District line).
Perhaps should also exclude those that have been resolved.
As Rail noise is clearly a hot topic that Tfl is spending a lot of money on, you might expect it to have its own data of Noise levels experienced (and the type of noise) so that the control measures can be most effectively targeted. Rather than the traditional who shouts louder method - forgive the pun :-)

[gefw]

The latest quarterly investment committee report says a further 468m of track has been reworked in the deep tube - so hopefully that will have addressed some of these noise hotspots.

Would be good to get some sort of feedback that these rail fixings were worthwhile

[ted672]

Is that miles or metres? if the latter, it's not exactly an outstanding acheivement!!

Feb 27, 2024 8:58:06 GMT gefw said:

The latest quarterly investment committee report says a further 468m of track has been reworked in the deep tube

[rincew1nd]

The standard symbol for mile is "mi".

The standard symbol for metre is "m".

[Chris M]

The whole London Underground network is only 402 km (250 miles) long (Source). Do you really think they could replace all of that (most of it twice!) in three months? Without any major shutdowns? With their current financial position?

[brigham]

In which case it's clearly metres. Not exactly an outstanding acheivement!!

[xtmw]

Most of the track replacement is done during engineering hours...

[Chris M]

Mar 1, 2024 8:17:23 GMT brigham said:

In which case it's clearly metres. Not exactly an outstanding acheivement!!

Like-for-like replacement of 468 metres of above-ground plain line track in three months would indeed not be outstanding. However that is almost certainly not what has been done. Much of the track replacement will have been in tunnels (which takes longer) and carried out during engineering hours (which gives you only a few hours each night). In at least some cases it wont have been a straight like-for-like replacement of the rails (as this seems unlikely to resolve the noise issue) but replacing the track fixings and/or track support too - both adding time and complexity.
This will have been in addition to all the other regular maintenance carried out during the period, e.g. it likely doesn't include any track replaced for reasons other than noise. So while it probably isn't an "outstanding" achievement it almost certainly isn't a bad one either.

[t697]

Apparently the "investment committee report says a further 468m of track has been reworked in the deep tube". Reworked doesn't necessarily mean replaced. So I wouldn't necessarily count all that as replacement of particular rail fastening types. Some will have 'simply' been rail grinding.

[ted672]

Seems I've kicked the hornets'nest! It's just that from a passenger point of view 468m of track is about the length, or part of, one section of noisy track. Looking at the spotty map, it will be many, many years before all the areas are dealt with. What puzzles me is that travelling around in the late 60s, early 70s, there was almost none of the screeching noise we're now subject too. At the risk of kicking off another discussion, is it the change to long welded rail that's a contributory factor?

[brigham]

Long welded flat-bottom rail, perhaps?

[bigvern]

The use of flexible bogie frames where one side of the bogie is attached by torsion rubber links to the other side frame, weight distribution across the bogie wheel set can differ, the bogie flexes to go round curves easier to reduce wheel wear, but the pivot of the bogie is offset and is on a rubber bushed arm, this does restrict the rotation of the bogie on tighter curves, so the bogie can dynamically bounce on curved track this causes rail corrugations on curved track, also wheel wear differs on diagonal corners of the bogie wheel sets, so the Victoria Line and Northern Line uses the same bogie, hence more hotspots on these lines, corrugation only occured after the new stock was introduced. Other lines have noise issues but this is down more to track resilience.

[johnlinford]

Anecdote may not be the plural of data but the rail noise between Canary Wharf and London Bridge seems a lot better than it was

[gefw]

Mar 1, 2024 13:16:14 GMT Chris M said:

One of the purposes of the original post was to try and get some feel as to the benefit of this investment in Deklor fixings (circa £4m for 468m of track).
From the other threads there does seem to be other options for addressing the likely route causes/contributing factors associated with the various types of noise.

Does anyone know if different weights/web profiles of flat bottomed rails have been tried or whether the remit for chair spacing has been considered ?

[d7666]

Feb 29, 2024 20:10:21 GMT rincew1nd said:

The standard symbol for mile is "mi".

The standard symbol for metre is "m".

By what standard ? That is, by what national or international standard ?

mi for mile is a recommended not mandated symbol, and even then is from US systems of standards or any derived from that.

It is true mi exists to avoid confusion with m (where m symbol for metre and for milli multiplier prefix is a standard) - but mi itself is not in those standards.

Furthermore, mph, mpg, etc, continue as recommended symbols for miles per hour, miles per gallon etc.

All over the interweb google results do say mi is a standard, but this is a plagiarised answer, never stating which standard, and repeating incorrect data everywhere does not make it correct; a few google responses do refer to the US recommendation, but these state recommendation, not a standard.

A company or body may recommended use of mi not m, but that still don't make a standard recognised elsewhere.

Anyway, the post that seems to have sparked this referred to 468 m : it must be pretty obvious this is 468 m(etre) not 468 mi(le) - if anyone thinks there is confusion there can they please advise what LU tunnel section is 468 miles long they are confused with, never mind a 468 mile section amid a larger length.

Yet further, seeing as we have drifted into pedantry here, if you want to give advice on symbols for miles, you have to go back one step further and differentiate between statute miles and nautical miles, where the recommended (again not mandated) symbol is often M (i.e. capitalised), so there is the whole gamut of confusion with ISO M for Mega, etc.

Sure, in context here it is pretty obvious statute miles were being talked about, but once once starts invoking standards, one has to be precise.

To be clear, wherever any confusion might arise, mile really ought to be spelled in full.

[zbang]

Dunno, -I- wasn't confused. In context, 468m is most likely 468 meters (or metres, if you prefer), not miles of any sort.

As for "standards", there are those codified by an internationally-recognised body and those "because everyone (here) does it that way". Again, context is important.

[Chris M]

Mar 2, 2024 17:59:37 GMT zbang said:

As for "standards", there are those codified by an internationally-recognised body and those "because everyone (here) does it that way". Again, context is important.

Exactly this.
And that is where we should leave the discussion of units and standards for this thread

[gefw]

Mar 1, 2024 18:16:18 GMT bigvern said:

The use of flexible bogie frames where one side of the bogie is attached by torsion rubber links to the other side frame, weight distribution across the bogie wheel set can differ, the bogie flexes to go round curves easier to reduce wheel wear, but the pivot of the bogie is offset and is on a rubber bushed arm, this does restrict the rotation of the bogie on tighter curves, so the bogie can dynamically bounce on curved track this causes rail corrugations on curved track, also wheel wear differs on diagonal corners of the bogie wheel sets, so the Victoria Line and Northern Line uses the same bogie, hence more hotspots on these lines, corrugation only occured after the new stock was introduced. Other lines have noise issues but this is down more to track resilience.

Would be interesting to see if there is correlation of track geometry (curve, cant etc) with areas where track corrugation occurs.
I also suspect that the bogie dampers may also be relevant to your theory (as they affect wheel unloading + hunting)

[zbang]

I think you would also have to correlate that to the actual wheel profile.

Also, my understanding of surface corrugation is that it comes more from an intermittent longitudinal force (from intermittent braking) then from vertical force; we see that in some road surfaces, usually at the end of motorway ramps (there are some prime examples of that not far from me).