Excessive noise in tunnels

It only takes a few weeks for the noise to return. TfL themselves notes rail grinding has to take place 6-12 months for it to be effective. If the stretch between STR and LEY E/B is grinded (which I am most certain it will be) I'll use one of those noise measurement apps and see the difference. If I recall correctly it reaches 100DB in that section !

Any thoughts yet on the noise hot spots that have changed as a results of the grinding ?

All of what follows can be better expressed mathematically. However, the forces which cause the conditions we experience as ‘howling’ or ‘screeching’ ensue from unfavourable combinations of track and wheel characteristics which become defects. These are neither limited to rail length and railhead geometry, nor wheel and rail interface, nor acceleration and breaking forces nor propulsion and driving characteristics. Essentially the rail ripples or corrugations combine with wheel-tires passing over them to generate the complex wave forms we find irritating even painful. Now the acceleration and deceleration profiles on the ATO operated lines are likely to be programmed to be much more uniform than on the manually driven trains.

I can hear someone saying, “Well if it was caused mechanically, why can’t it be corrected by reverse mechanical actions?” If you think back, rails used to be bolted in 60 feet lengths, not continuously welded. That meant that many harmonic transmissions along the rails were somewhat naturally attenuated at each joint. Solution: return to using 60 feet lengths of rail. No way! Still thinking back, a nice extreme case is the District Line ‘Q’ Stock which was a potpourri of various generations of rolling stock extending over around 15 years. What’s the relevance here? Well, the distances between truck centres were likely different along the length of the entire train and the distance between the axels also varied somewhat between one car and the next. (I’d add that these cars were not manufactured on today’s sub-millimetre-accurate CAD/CAM rigs and would have had many small manufacturing differences). This means that compared with today’s rolling stock the distances between wheel/rail contact points could be quite irregular. Even the insertion of one of the standard stock ‘58’ trailers into Bakerloo line trains from the end of the 1930s onwards, would have led to irregular distances between the wheel/rail contact points in every train and possibly gone someway to preventing the build-up of distorting harmonics. I travelled often with my mother on the Bakerloo and can never remember the howling experienced in recent times between St John’s Wood and Baker Street. Solution: vary the distances between truck pivot points and axel spacing. That’s not likely to happen either. The last point, uniform (ATO) as opposed to individually driven trains at least facilitating constant rail-head deformation at crucial intervals is not going to convince many to return to personal driver control. There are other factors too.

In the early 1970s when I went to live and work in Germany and Switzerland, I noticed that the insertion of a hard rubber disc between the inner wheel and wheel tire of many rail-borne vehicles led to not only a quieter ride but also led to the unevenness in railhead not being completely transmitted to the vehicle interior in the form of noise. This matter has been investigated in several presentations at the IHRUS meetings in Luzern over the past few years. My feeling is that with continuous rail, uniform truck-pivot distancing, identical inter-axel distances. identical acceleration, deceleration, and speed maintenance profiles possibly ideal conditions are established for rail rippling and corrugations to build up.

It will be interesting to see what the effect of the new 2024 TS for the Piccadilly (and beyond?) with its alternate positioning of power cars and suspended intermediate cars will be on the rails. The whole problem plays out in the micro-metre range, and maybe there is some way to have the wheel ‘hear’ the first signs of deformation while it is still beyond aural acuity and through acoustic data feedback incrementally alter the driving characteristic to ‘hammer-down’ incipient imperfections. Or might it be possible to set up each unit with a microscopically different driving profile so that the uniform conditions which currently engender rail-deformation are countervailed? A lot of words, and not in a language I regularly use for technical matters, so thank you for reading this far.

It'll get grinded, the noise returns months (if not weeks) later and the cycle repeats...

I've had info from one of my (very reliable) sources

There's currently a rail grinder stabled in Woodford sidings where it's been completing further work. One of the sections it's recently grinded is the Leyton tunnel mouth area. So that tells us that the section between Stratford and Leyton tunnel mouth has been fully grinded in the past 3 weeks.

In addition to this, the route from Leytonstone Platform 1 - Leyton is blocked today due to an engineering vehicle (no description of what), so trains leaving platform 1 have to use the crossover just west of platform 1 (rather from the one in the Leyton area)

That could be a tamper? (The rail grinder seems to have 3 windows at the front according to a brief image search)

Here’s an example video of a tube-gauge tamper on YouTube 

I'm sure it's the Schweerbau rail grinder. Been at Woodford sidings for a few days now however will check

Is there only the one left now ?

Were there not 3 at one time, or have I got it wrong that of those 3 only one was tube the other two surface gauge, and the latter two then replced by the current, more recent two ?


One of those two current surface gauge machines was berthed in the Chalfont & Latimer bay road for a few days 2-3 weeks ago (well maybe not continuously; it might have worked from there; I only saw it on long line CCTV). I vaguely remember a thread somewhere, or maybe another forum, suggesting Chalfont bay is completely out of use, and the signalling fuses removed; well I can't say about fuses, but it sure was used this month by engineers trains.

I believed the Plasser tampers were at end of life with much of the control electronics/computers obsolete and hence the 2 new Matisa tampers were bought. Then the Plassers were being phasesd out (possibly cannibalised to keep the others going). I don't know how many are still operational or whether some rejuvenation/upgrade has been undertaken.
The Train details in the EWSA should give some clues as to how many are operational

Interesting, thanks
Rather like similar work in this country, progress/reports/developments have gone rather quiet over the last 10 yrs.
Southampton Uni seems have been heavily involved in the theories of rail noise damping & a similar solution to attach weight to the web/foot of the rail.
www.southampton.ac.uk/engineering/research/impact/sustainable_expansion_of_rail_networks.page.

This seems to have lead to significant fitment on the Thameslink cross London underground line - but can't see any post installation reports or indeed confirmation it is still in place.
www.railwaygazette.com/news/noise-damping-track-to-be-installed-on-thameslink-core/36025.article

A trial of similar seems to have be fitted on the London Underground seemingly involving the specialist you mention - but not much info in public domain
www.wal.hk/#!newsletter16a

Recently there does seem to have been pressure & a proposal for further trials on the Underground - but the wording seems to be confused about whether they are fitted to the rail !!
www.london.gov.uk/who-we-are/what-london-assembly-does/questions-mayor/find-an-answer/tube-noise-29


Interestingly most of the above seem to suggest dampers to address the (higher pitch?) ringing from resonance of the rail, rather the lower frequency rumble/roar from rail head corrugation. That was why I asked the broader question as to other practical ways of hindering resonance.

Also note that on significantly curved tracks, LUL seems to have mainly used lubrication (via on train lubricators or Trackside dispensers) to reduce the squeal and rail head damage - The latest term for this appears to be "Friction modifiers" (rather than greasers) because of the use of a wider range of "potions" being dispensed. The use of this has rightly always been of some concern due to possible significant reduction in emergency braking performance (On LUL, brake blocks apply to the running surface of the wheel)

Stratford-Leyton seems to be a well known problem area (or at least parts of it).
Interesting that the EB and WB noise level recordings are hugely different - is this historically / normally the case?
If so this would perhaps suggest a certain track feature/arrangement that needs some TLC ;-)

Will be interesting if there are different rail types on each road

If correct, that is quite a surprise (as a lot of the discussion re noise is about rail mounts which are specific to flat bottom rail).

Any other contributions regarding the difference between the EB & WB tracks ?

Significant fitment it might be. Significant long term noise reduction no.

The section between the present day St.Pancras and the former KX TL station is particularly bad in this respect.


Speaking as a TL commuter traversing the core, excess noise comes and goes, even today, just like the reports of LU tube lines. TL seems to be on a periodic cycle where it gets excessive then they do something and it all goes quiet for many months.

They do rail grind there, but as you can clearly hear when just been done (700s wheels on new ground rail sound remarkably like a noisy 319 traction motor used to) I know from experience this does not fix it - all you got is total more noise.

We have been in core general noise level now for several months, we must be running up to another excess noise period; if they have now fixed it then that is recent, the RGI item cited is now 13 years old, so it can't really be exactly that if it is fixed now.