Super elevation

Discuss the prototype and how to model it.
best33
Posts: 41
Joined: Thu May 12, 2011 12:05 pm

Super elevation

Postby best33 » Sun Oct 11, 2020 12:29 pm

I am planning a layout of South Brent and am looking at incorporating the super elevation of the track into the baseboard design.

Brent has a single island platform for the Kingsbridge branch and there is a junction to the East of the station from a curved main line to give access from the yard and branch to the up line. This involves a single slip on the down line and trailing turnout on the up side which is on the inner curve. The transition curve for this begins in the platform area. I can work out the extent of the tilt and the transition from the Great Western Study Group Switch and Crossing book, but I am pondering the following:

The Branch loop and down line share opposite faces of the island platform and are at the same level however as the super elevation develops the outer rail of the down main line at the point of the connection with the branch spur will be higher so there will need to be a slight rise from the branch to the main line followed by a descent as the line traverses the junction onto the up line. Photos show this hump quite well. However to limit the severity of the hump I am wondering where the pivot point to establish the tilt on the main line should be located. Both up and down lines remain on the same plane.

If I were to assume a pivot point in the middle of the 6 foot way so the inside curve rail is lowered the same amount as the outside curve is raised the difference in height between the platform and the trains as they enter the curve would be limited and this seems logical for normal situations.

Ideally the hump for the branch connection should be on the spur before the connection otherwise it will cause no end of trouble. To do this the branch connection spur needs to climb more steeply so that it is over the hump and descending before the connection which means it needs to be higher than the main line at the hump.

However is it possible that the outer rail of the curve is maintained at the same level as the branch in which case the other three rails of the main line have to be lower. The advantage of this is that there is no height difference with the branch at the junction and I can arrange a gentler gradient from the branch to the main line to get over a less severe hump which should cause fewer problems.

Does anybody have a prototypical experience of this type of issue for earlier practice up to 1970? Any comments gratefully received.

Thanks in anticipation for those who have got through this without being totally baffled!

Regards

Mark Humphrys

User avatar
Tim V
Posts: 2456
Joined: Tue Jul 29, 2008 4:40 pm

Re: Super elevation

Postby Tim V » Sun Oct 11, 2020 1:23 pm

You 'may' find this document of some use ...

img249.jpg

img250.jpg
Tim V
Scalefour News Editor

best33
Posts: 41
Joined: Thu May 12, 2011 12:05 pm

Re: Super elevation

Postby best33 » Sun Oct 11, 2020 5:24 pm

Thanks Tim,

Thanks for posting that but it isn't quite what I was asking. The attached PDF should help to make it more clear.
Brent_Superelevation.pdf
(41.18 KiB) Downloaded 65 times


Regards

Mark Humphrys

User avatar
Noel
Posts: 1429
Joined: Wed Jun 23, 2010 1:04 pm

Re: Super elevation

Postby Noel » Sun Oct 11, 2020 6:52 pm

Not sure I've understood either, Mark, but any time I've noticed superelevation on a double track line recently, both outer rails were more or less at the same height, and both inner rails as well, which differs from your diagrams. Was steam era track different?
Regards
Noel

User avatar
grovenor-2685
Forum Team
Posts: 3426
Joined: Sun Jun 29, 2008 8:02 pm

Re: Super elevation

Postby grovenor-2685 » Sun Oct 11, 2020 9:20 pm

Applying superelevation to plain double track would normally be by raising each outer rail. However where there has to be connection between the two tracks then all 4 rails need to be in the same plane which would use a huge amount of ballast if just done by packing in more ballast so usually the underlying formation would be adjusted to suit. Where, as in this case a junction comes off the high side of the double track then the branch track will be heading uphill and need a vertical curve to bring it back level. Vertical curves have a minimum specified radius so that vehicles don't get grounded or have coupling problems, this minimum would likely be in the region of 1000 m. So the engineers would have designed it to achieve the desired result. Whether they used either of your suggested options or something else could be nothing more than a guess.
It was possible to build turnouts with different superelevation in each route using two level chairs, sometimes this helped, but that would not be possible with slips.
You can always have a look through this,
http://www.norgrove.me.uk/permanent_way_notes.htm
although I suspect it won't help much.
Regards
Keith
Grovenor Sidings

ted.stephens
Posts: 9
Joined: Fri Nov 30, 2012 6:04 pm

Re: Super elevation

Postby ted.stephens » Mon Oct 12, 2020 9:46 am

Mark,

As a general rule cant (superelevation) is applied by lifting, simple reason - it's much easier to pack more ballast underneath a sleeper to lift it than to dig it out to lower it. Also the civil engineers were generally aiming to increase ballast depth on the railway for better drainage and stability, so lifting was preferred.

As Keith notes cant is applied by rotating about the inner rail on a curve, lifting the outer rail, thus the railway grade line is measured along the inner rail, generally known as low rail level. Note - practice varies in other countries, for example the Swiss rotate track about the centre line to apply cant.

Based on this line of thinking where two tracks need to be co-planar, as in your case with tracks crossing each other on a canted curve, then the inner rail of your up track would be the grade line for the route and cant would be applied from that across to the other three main line rails. This avoids any twist in the switch and crossing (S&C) work. The down line would in effect have a hump in its longitudinal vertical alignment to accommodate the co-planar cant applied from the up track. This would be accommodated by raising its vertical alignment either side of the S&C area.

Putting some numbers on your site at Brent, the curvature through the junction area on the main line is around 650m, measured from OS maps on http://www.old-maps.co.uk . For the purposes of this calculation it has been assumed that the main line speed is 80 Km/h (50mph). Using the standard formula for calculating equilibrium cant (where all lateral forces are balanced by cant)

Cant Eq. = 11.8 * V^2 / R (Cant - mm. V - Km/h, R - m)
Cant Eq. = 11.8 * 80 * 80 /650 = 116mm (4 1/2")

As a rule of thumb for passenger comfort the applied cant was less than equilibrium cant, commonly around 2/3, so the applied cant would be 2/3 * 116 = 77mm, say 75mm(3" in old money or a scale 1mm). The proportion of equilibrium cant applied varies according to traffic type or particular circumstances. In this case where there is a junction and the civil engineer might choose to under-cant the track, perhaps applying only 50mm (2"). Whilst this makes the ride less comfortable due to high cant deficiency it does mitigate some of the design issues related twist discussed later.

By the time you apply this cant from the inside rail running edge of the up track across to the outside rail running edge of the down track the level difference would be

(1435 + 3405) * 50/1505 =161 mm (c. 6 1/2") (Note 1505 is the distance from rail centre to rail centre over which cant is normally measured, 1435 + 70 mm)

Looking at the photo of the junction area here http://www.cornwallrailwaysociety.org.uk/kingsbridge-branch.html the amount of ballast on the down track side appears to be less than this suggesting that the main line formation had a crossfall applied to it towards the inside of the curve. The other thing to note on this photo is the long timbers between the two crossing noses on the up main line turnout and down single slip, this means that the tracks must be co-planar.

Now lets look at twist. The issue comes for the short piece of track, perhaps no more than 10 sleepers worth (25ft, 7.5m ), between the down line and the refuge siding . Assuming the refuge siding and branch line track has no cant applied and the mainline has only 50mm cant applied then 50mm of cant needs to be reduced to zero over a distance of 7.5m giving a cant gradient of 1:150 which is unacceptable. Desirable cant gradient is 1:800, maximum is 1:400. Any more than this and the twist is likely to cause a derailment as there is insufficient movement in the vehicle suspension to accommodate it. Note your vehicles will need compensation or springing. This could be solved by applying some cant to the turnout in the refuge siding, For the purposes of this design assume 15mm of cant is removed over this 7.5m section of track, giving a twist of 1:500, acceptable over slow speed track like this. However the turnout on the refuge siding will need to be canted at 35mm.

The leg from the refuge turnout going to the branch line has to accommodate a 35mm change in cant over 20 sleepers (50ft, 15m). This gives a twist of 1:428, just below the maximum so OK.

The track between the down single slip and branch line platform area will have a similar issue. Measuring the distance on the OS map is tricky because the branch line track is obscured by a bridge. For working purposes assume it is 20m, the cant to be taken off is 50mm so the cant gradient is 50/20000 = 1:400, on the limit.

What a wonderfully awkward example of a junction when you take into account canted track and it has in it some quite tricky design elements for you to consider.

Now the question is whether it is worth the effort to introduce cant in the model. Some would argue not, however if you don't do it then it will loose the feel of trains sweeping through the curves and rolling into the cant. Depends of what you want to achieve.

Hope this helps, Ted

best33
Posts: 41
Joined: Thu May 12, 2011 12:05 pm

Re: Super elevation

Postby best33 » Mon Oct 12, 2020 3:40 pm

Thanks to everybody for responding to this. The information shared is extremely helpful. To complicate things further the connection from the main line crosses the branch and into the yard via a double slip under the bridge so there is lots of potential to mess this up if I get it wrong!

I have always considered the Kingsbridge junction end of the layout to be critical to its success and so I wanted to sort out any potential problems before I get too far with it. Experience with Chagford Road twenty years ago highlighted the importance of designing the layout before starting to build it!

I am going to design the track bed in 3D using Cadds so I can optimise the baseboard design which I will get CNC cut since my carpentry skills aren't up to much so I can get the parts cut for me and then assemble my own bespoke kit. I intend to model the track in 3D on cadds too so that the template is correct for the cant when building it in the flat.

Regards

Mark Humphrys


Return to “Track and Turnouts”

Who is online

Users browsing this forum: No registered users and 3 guests