Rail Joints

[Flying Phil]

Whilst travelling in a DMU along the GCR it was very obvious that the rails "dipped" at every bolted rail joint.
I wondered why they didn't place a sleeper under the joints with a flat chair, just to support the ends of the rail where the fishplates are bolted?
It seems such an obvious idea there must be a good reason why it is not done - but I don't know why....

 

[furnessvale]

Whilst travelling in a DMU along the GCR it was very obvious that the rails "dipped" at every bolted rail joint.
I wondered why they didn't place a sleeper under the joints with a flat chair, just to support the ends of the rail where the fishplates are bolted?
It seems such an obvious idea there must be a good reason why it is not done - but I don't know why....

Supported joints, where a sleeper is fully under the joint, and semi supported joints where the end sleepers are moved close to the joint, have been around for years.

They have been out of favour on BR and later for many years but, strangely, my copy of "British Railway Track", the platelayers bible, is ambivalent as to why.

Mention is made of difficulties with tamping the close sleepers of a semi supported joint but that doesn't apply with a fully supported joint. There is some suggestion of increased rail end batter with the rigidity of a fully supported joint.

Basically, there isn't a definitive answer, but modern welded rail has considerably reduced the problem.

 

[Joseph_Locke]

strangely, my copy of "British Railway Track", the platelayers bible, is ambivalent as to why.

It is because even a joint with a baseplate directly beneath it is a point of weakness, as the need to allow the joint to "slide" means it is necessarily "loose". Add to that the issue that each rail end is only supported on half a baseplate and you just get galling on the rail foot and you're back where you started.

As you suggested, the best rail joint is no rail joint...

 

[Flying Phil]

Thanks both for that explanation, in essence - "Clickety click" is better than "Thumpety thump"?

 

[ComUtoR]

Thanks both for that explanation, in essence - "Clickety click" is better than "Thumpety thump"?

Clickety click I'd ignore. Thumpety tump I'd report.

 

[Trog]

A chair that held a rail at the fishplates would probably be a bit problematical as there would be no rail web groove to locate the jaw and key in.

I also wonder if rail gall at the rail end would be a rail break risk, with cracks initiating from stress raisers caused by the gall, and propagating fast as the rail ends are bent/impacted by passing wheels. If the chair at the joint tended to work loose due to a lack of grip on the rail or void due to being right under the rail joint impact, that would then leave a full beds worth of unsupported rail each side of the joint, rather than the normal half bed.

 

[Ploughman]

North Eastern practice used to involve broader chair seats at joints combined with closer spacing of the sleepers either side of the joint. So much so that the actual fishplate was only big enough for 2 holes rather than 4.
These joints survived in the track on the Scarborough line at least until 2009.
Some conversion to 4 hole has taken place.

 

[Railsigns]

So much so that the actual fishplate was only big enough for 2 holes rather than 4.
These joints survived in the track on the Scarborough line at least until 2009.

I photographed some of those at Malton in July 2014.

 

[hwl]

The former LNE practice mentioned above and Southern Railway inter-war experiments on support underneath fishplated joints (or why not to!) is briefly covered in the MoT Hither Green report page 21:
http://www.railwaysarchive.co.uk/documents/MoT_Hither1967.pdf

 

[4973]

On a related issue, would some knowledgeable person explain the reason for 6-bolt fishplates and when they were introduced.

 

[Joseph_Locke]

On a related issue, would some knowledgeable person explain the reason for 6-bolt fishplates and when they were introduced.

Perceived greater mechanical strength, sometime in the late Victorian era.

 

[Railsigns]

Have 6-hole fishplates ever been used for non-insulated joints?

 

[Joseph_Locke]

I'm happy to be found wrong, but 6-hole ordinary joints have never seen widespread use in the UK, to my knowledge.

A contributory factor is probably (as Trog suggests) the UK's persistence with bullhead means that the supporting chairs (and keys) would have to be outside the fishplate limits, so 6-holers had to wait for flatbottom rail anyway.

The UK has also never seen the kind of heavy haul (e.g. >100MGTPA) operations that other countries have, so the drivers to adoption are fewer - there are still a few Burns 4-hole plates in the main lines and some of these may well see both me and Trog out ...

 

[Trog]

My-self I have only seen six hole plates used as IBJ's in 110A/113A/CEN60 flat bottom rail, and for ordinary joints in CEN60 rail. However as CEN60 is generally only used as CWR in the mainlines ordinary fish plated joints in it are some what uncommon. As for 4-hole Burns plates seeing me out I will have Mr Locke know that there is still a bit of life left in this old dog yet.


Until recently there was also some GWR two hole fish plated track at Bearley Junction.

 

[D Foster]

My minor studies of Pway suggest that fully supported joints, including combination chair/fishplate, were tried several times in the Victorian era. I have no specific knowledge of anything later. It seems to me that the various joints tried (and often patented before the patent laws were altered) proved to do more harm than good. The main problem seems to have been that the joints were made far too stiff - creating a hammer blow - with all the attending problems to the head and foot of the rail. Until better metals were produced the blows also tended to destroy the joint chairs as well.

I haven't a clue about foreign or modern high-load practice!

For jointed track in the UK the early lack of success with direct joint support and relative success with plain fishplates seems to have become the accepted practice - with varying degrees of closing up the sleeper spacing and differing practices of using larger sized chairs adjacent to joints.
In the 1890s there was some study of sleeper spacing practice (along with varying chairs). I have no idea where my copies of this are currently - but - I can say that no two late Victorian/early Edwardian companies opted for the same practice. Naturally manual track maintenance made for a completely different situation compared to modern machine maintenance ... Although it's amazing how often a tamper can still whack (and damage) a sleeper.

With conventional fish plated track the key maintenance feature doesn't lay in the joint itself but in the ballast maintenance. When I think about it - this would apply to a supported joint as well.

Another thing that might be commented on is that, where CWR has replaced jointed track (and LWR) there can still be problems with the track where there were previous problems at the site of joints/fishplates. I'm told that the cure for this is heavy work on the ballast and even the sub-formation.

 

[Joseph_Locke]

Another thing that might be commented on is that, where CWR has replaced jointed track (and LWR) there can still be problems with the track where there were previous problems at the site of joints/fishplates. I'm told that the cure for this is heavy work on the ballast and even the sub-formation.

Ah, ballast memory, I remember it well.

 

[Ploughman]

If you need to retain rail joints on a renewal, then if possible make sure that the new joints do not sit in the same bed as the old ones were in.
On the NYMR we try and relocate any new joint at least 4 bays from the old as well as digging out or scarifying of the bed.
That coupled with closer sleeper spacing and 150mm of ballast seems to be working.
Conversion, where possible to CWR, is also playing a part.

 

[4973]

Perceived greater mechanical strength, sometime in the late Victorian era.

Thanks for that.