Modelling loose heeled switches in 4mm

Discuss the prototype and how to model it.
junctionmad
Posts: 247
Joined: Tue Dec 08, 2015 3:38 pm

Modelling loose heeled switches in 4mm

Postby junctionmad » Fri Feb 15, 2019 4:19 pm

I put this question up on Templot Club , but it might be better here as the audience is likely bigger

Ive built a fair bit of semi curved points in various 4mm gauges , but now I've been asked to make some 00-SF GWR loose heeled turnouts.

While I know there are geometry changes and joggles the main issue I have in 4mm is whether the loose heel chair post the fishplate and the " loose" fishplate is actually realisable in 4mm. As I understand it on the prototype the tongue pivots slightly in the "loose" chair , facilitated by a loose fishplate

But in practice in 4mm this is unrealisable in my opinion and the switch blade will react more like the same continuous blade in semi curved , albeit with joggles, straight cut and curved switch rails etc . I know some people suggest a pivot pin ( aka PECO !!!) , but this isnt really prototypically anyway and anyway I dont beleive there is a loose heel chair available anyway in 4mm ( and we dont bolt the tongue upright anyway ). Hence its likely the chair either side of the " loose" fishplate will in effect hold the switch tongue rigid

The other issue of course in OO-SF with loose heel points is maintain wheel clearance across the rail faces because you have much less slide chair movement compared to semi curved . it seems doable in p4 , but in OO-SF ( which has EM clearances , this seems more problematic )

anyone have any comments or suggestions as to how to implement these points in 4mm
dave

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Fri Feb 15, 2019 5:03 pm

Hi Dave,

Loose-heel switches can only be modelled exactly to scale if you are using dead-scale flangeways. Even the overscale flangeways in P4 are border-line.

In 00-SF and EM you must extend the length of the switch rails to the position where the heel offset provides at least 1mm flangeway clearance.

Folks have many preferred ways of replicating the effect of the loose heel. Prototypically it isn't a true "pivot" in the sense of a round pin or shaft rotating in a circular housing on a fixed centre. But that is a common way of modelling them, soldering the end of the switch rail to a pin or screw head which can rotate.

cheers,

Martin.
40 years developing Templot. And counting ...

junctionmad
Posts: 247
Joined: Tue Dec 08, 2015 3:38 pm

Re: Modelling loose heeled switches in 4mm

Postby junctionmad » Fri Feb 15, 2019 5:36 pm

Martin Wynne wrote:Hi Dave,

Loose-heel switches can only be modelled exactly to scale if you are using dead-scale flangeways. Even the overscale flangeways in P4 are border-line.

In 00-SF and EM you must extend the length of the switch rails to the position where the heel offset provides at least 1mm flangeway clearance.

Folks have many preferred ways of replicating the effect of the loose heel. Prototypically it isn't a true "pivot" in the sense of a round pin or shaft rotating in a circular housing on a fixed centre. But that is a common way of modelling them, soldering the end of the switch rail to a pin or screw head which can rotate.

cheers,

Martin.

thanks Martin , I am aware that the switch rail heel position will have to in effect be moved " back " towards the knuckle to provide for better clearance,


can templot do this Martin ?

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Fri Feb 15, 2019 5:55 pm

junctionmad wrote:can Templot do this Martin ?

Hi Dave,

Sure. Click the set custom switch... button on the switch dialog.

More info on all this at:

http://85a.co.uk/forum/view_topic.php?i ... d=6#p19351

cheers,

Martin.
40 years developing Templot. And counting ...

Enigma
Posts: 237
Joined: Tue Aug 13, 2013 2:49 pm

Re: Modelling loose heeled switches in 4mm

Postby Enigma » Fri Feb 15, 2019 10:38 pm

I was told a couple of evenings ago by one of our members that he uses a C&L brass fishplate as the 'pivot' point, fitted loose to both rails. I might try this myself as I have some very tight industrial points to build.

User avatar
steamraiser
Posts: 407
Joined: Thu Dec 31, 2009 4:49 pm

Re: Modelling loose heeled switches in 4mm

Postby steamraiser » Sat Feb 16, 2019 12:49 am

I have used guitar string (10 or 11 thou) as the hinge linking the blade to the closure rail.
I solder it between the head and the foot on the outside of the rail with a small gap.

Gordon A

User avatar
Dave K
Posts: 210
Joined: Sun Jul 20, 2008 6:11 pm

Re: Modelling loose heeled switches in 4mm

Postby Dave K » Sat Feb 16, 2019 11:28 am

Enigma wrote:I was told a couple of evenings ago by one of our members that he uses a C&L brass fishplate as the 'pivot' point, fitted loose to both rails. I might try this myself as I have some very tight industrial points to build.


As ‘Enigma’ said I’ve built some loose heel switches and GW old style curved blades and the Exactoscale brass fishplates for the pivot. Ive taken a couple of photos to show what they look like.

7A72BABF-191E-47F9-A5F6-89CC57C728C3.jpeg

E435FE01-CBAB-4ABD-AE26-9FE77D74EDDC.jpeg
The joggle maker is a Morgan Design I got from the the ‘stores’ for the princely sum of 30p for 5 (although they are not shown on the stores list)


Dave

dal-t
Posts: 623
Joined: Mon Jan 11, 2010 8:06 pm

Re: Modelling loose heeled switches in 4mm

Postby dal-t » Sat Feb 16, 2019 12:06 pm

Oh, so glad those pics right themselves when opened up, my vertigo was having fits (yes, it does anyway, but ...). Now, not being critical, merely curious, but how long do those fishplates last before they snap? My experience of folding photoetch is if I look at it twice it's likely to split, maybe three times if I'm lucky, and don't get me started on those bendy plastic 'hinges' which aren't but simple come apart, so I'd be very concerned about that arrangement for a working turnout - or have they been exhaustively tested and proved OK?
David L-T

davebradwell
Posts: 88
Joined: Fri Jan 18, 2019 3:48 pm

Re: Modelling loose heeled switches in 4mm

Postby davebradwell » Sat Feb 16, 2019 3:30 pm

I've been using Exactoscale brass fishplates as these pivots for years working in the same way as the full-size. There's quite a lot of them and they're used regularly. Fishplate is soldered to the fixed rail and one of the boltheads filed off the moving end. Drill right through both sides at this position with point blade in fishplate. Open out hole in blade slightly and splay fishplate to permit movement. Take a little off this end of the blade to allow free movement. Assemble to point and put wire pin in place last to secure blade, securing with a quick dab of solder to represent missing bolt-head. Rail does not bend.

I did have some trouble with electrical continuity on one of them for a while but somehow that sorted itself out, perhaps with a bit of encouragement. Probably caused by residual flux.

Dave B

seanmcs
Posts: 24
Joined: Wed Dec 10, 2008 12:57 am

Re: Modelling loose heeled switches in 4mm

Postby seanmcs » Sun Feb 17, 2019 12:36 am

This looks wonderfully authentic. The Exactoscale fishplates look really good. I wonder if an alternative might be the relatively recent Peco BH rail connectors, but maybe they might look a bit bulbous at the bottom. Perhaps disguised with a bit of black paint?

Dave, I ask if you have put in a joggle. I can't quite make it out in the photos. Also, are you planning to drill out the rails to accept GWR style stretcher rods that pass through the running rails and are fixed near the end of the planing rails. If so, could you explain the insulation and propulsion techniques. Apologies for all these questions but this thread unravels some of the unclear points in David Morgan's lengthy references/instructions.

Sean 6303

John Palmer
Posts: 566
Joined: Fri Jul 15, 2011 11:09 pm

Re: Modelling loose heeled switches in 4mm

Postby John Palmer » Sun Feb 17, 2019 9:56 am

The switch illustrated by Dave K causes me some concern, given the proximity of the tongue rail to the stock rail in the area I've indicated by a yellow ellipse:
GWR Loose heel switch.jpg
I would expect flange backs to strike the outer face of that tongue rail as they pass through - shades of Grayrigg. Furthermore, with the switch reversed to the opposite position of that shown so that the tongue rail is nestled against the stock rail at the point indicated by the ellipse, I would expect to find a gap between the tip of the tongue and the stock rail, which could also spell disaster.

On the other hand, I do like the appearance of the further of the two fishplates shown, as it appears to have been cranked to accommodate the rotation of the tongue at the joint. It's my understanding that this is how full size loose heel switches were constructed.

My investigations into loose heel switches suggest that heelspread is a critical dimension in getting such switches both to look and function as they should. In Templot you can play around with this dimension to your heart's content when setting up a custom switch reflecting the practice of your chosen prototype.

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Sun Feb 17, 2019 10:47 am

John Palmer wrote:The switch illustrated by Dave K causes me some concern, given the proximity of the tongue rail to the stock rail in the area I've indicated by a yellow ellipse

Hi John, Dave,

The photo illustrates well what happens when there is insufficient (or no) "set" in the diverging stock rail. This problem seems to crop up often, probably because it is so poorly covered in many of the standard trackbuilding books. Even the new trackbuilding book from the 2mm Association barely mentions it.

Here is my 10-year-old topic on the subject:

http://85a.co.uk/forum/view_topic.php?id=491&forum_id=1

Here's a simple idea which costs nothing and is a good way to make precise symmetrical bends in bullhead rail. It needs a bit of practice to know how hard to tap! Image

Image


Image

Hopefully the diagrams are self-explanatory. Just two bits of rail about an inch long laid side by side with a gap between. The smaller the gap, so the harder you need to press/hit, but the more precisely located will be the bend. Make sure the rail is exactly square across them before making the bend -- a sheet of graph paper underneath helps.

Here's a photo showing the "set" bend on an LNER switch:

Image
Thanks to Mick Nicholson for the pic.

cheers,

Martin.
40 years developing Templot. And counting ...

User avatar
Dave K
Posts: 210
Joined: Sun Jul 20, 2008 6:11 pm

Re: Modelling loose heeled switches in 4mm

Postby Dave K » Sun Feb 17, 2019 12:45 pm

dal-t wrote:Now, not being critical, merely curious, but how long do those fishplates last before they snap? My experience of folding photoetch is if I look at it twice it's likely to split, maybe three times if I'm lucky

David,
As Dave Bradwell states these are Exactoscale cast brass fishplates so they don’t spilt and are quite strong, however I don’t know whether they will be produced again.

seanmcs wrote:Dave, I ask if you have put in a joggle. I can't quite make it out in the photos. Also, are you planning to drill out the rails to accept GWR style stretcher rods that pass through the running rails and are fixed near the end of the planing rails. If so, could you explain the insulation and propulsion techniques. Apologies for all these questions but this thread unravels some of the unclear points in David Morgan's lengthy references/instructions.

Sean 6303

Sean,
This is only my 4’ by 1’ test track. The trackwork on Hallatrow was built by my friend Roger who now lives in Somerset and I’ve also moved to South Yorkshire so on any extension I will have to build the track so this was built to enable me to get some practice in.

Yes, there are joggle’s or sets as other’s call them. As you will have noticed I’m thinking of using a C & L tie bar as this should keep the blades in position and they do replicate GWR style stretcher bars.

John,
There is still some work to do before the tie bar is fixed. It was only put in to see if would provide the necessary clearance.

Dave

junctionmad
Posts: 247
Joined: Tue Dec 08, 2015 3:38 pm

Re: Modelling loose heeled switches in 4mm

Postby junctionmad » Sun Feb 17, 2019 9:11 pm

Martin , is the set the same where the stock rails have joggles ? , ie like semi curved


Dave K

I always thought joggles are completely different to sets , semi curved switches have sets , but do gwr loose heeled points have joggles and sets ?


Dave

User avatar
grovenor-2685
Posts: 3167
Joined: Sun Jun 29, 2008 8:02 pm

Re: Modelling loose heeled switches in 4mm

Postby grovenor-2685 » Sun Feb 17, 2019 9:24 pm

You still need a set when you have joggles.
The joggle is to allow for a thicker switch tip.
The set is to make the stockrail sit at the same angle as the planing on the back of the switch so the stock rail and switch fit together properly when the switch is closed and both gauge and flangeway are preserved when the switch is open.
Hence sets are mandatory, joggles are optional depending on the design of the switch tips.
https://www.scalefour.org/resources/switches.html Only a straight cut switch would need a joggle if the tips were on the thick side.
https://www.scalefour.org/resources/trackdetails.html
Regards

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Sun Feb 17, 2019 10:28 pm

junctionmad wrote:Martin , is the set the same where the stock rails have joggles ?

Hi Dave,

Yes, there is still a change in angle in the stock rail with a joggle. It's a slightly smaller angle than for a plain set because of the switch tip thickness. The object in all cases is to end up with the stock rail in the same place at the end of switch planing, at C in the diagram below, where the stock-gauge is measured.

Here's some old stuff from the Templot web site, see: http://templot.com/companion/real_track.php
_________________________________________

The first diagram below shows the left-hand stock rail and switch blade for a left-hand undercut pattern switch with a plain unjoggled stock rail.

A set (bend) is put in the stock rail at the toe of the switch (blade tip) to match the planing angle on the switch blade:

Image

The switch blade is planed (machined) to a sharp tip, and profiled down below the top of the stock rail to fit under the head of the stock rail (undercut). This is in order to have sufficient strength at the tip. Such a blade acts only as a guide to the wheel flanges for the first part of its length and does not actually support the weight of wheels running onto it until they have reached some way along it. Undercut switches can often be identified in photographs because the shiny top of the open blade stops some way short of the tip.

When a more robust switch is needed, the stock rails are joggled outwards by a small amount to create a housing for the switch blade tips and so protect them from wheel damage. This next diagram below shows this alternative straightcut pattern switch with a joggled stock rail:

Image

Instead of a plain set in the stock rail, it is joggled sideways between positions A and C, with the maximum deflection from its previous unjoggled alignment being at the blade tip position B.

In bullhead track the joggled section is always created by bending the rail rather than by machining a notch in it.

The switch blade is planed to a blunt tip (straightcut), to a thickness corresponding to the amount of joggle, which has been exaggerated in this diagram for clarity. Apart from a rounded corner at the tip the blade is the full height of the stock rail and plays its part in supporting the wheels along its full length. A straightcut switch can often be identified in photographs by showing a shiny top on the open blade all the way to the tip.

Image

In defining a joggled switch, two dimensions are needed. The joggle depth (sideways deflection) at B, and the joggle return length in front of the blade tip between positions A and B, in which the rail returns to its normal alignment. (The length between B and C always corresponds to the planing length for the switch blade). These dimensions need to be entered when creating a custom joggled switch in Templot.

Having entered joggle dimensions for your custom switch, you can choose whether to actually use the joggles for an individual template by means of the joggled stock rails option tickbox.
______________________________________

cheers,

Martin.
40 years developing Templot. And counting ...

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Sun Feb 17, 2019 10:54 pm

p.s. Dave,

If it helps to understand what's happening in a joggle, I have added a red line to the diagram:

Image

The red line shows the position of the stock rail for a plain set with no joggle. To make the joggle, imagine the stock rail being held along the red line, and then pushed sideways at B.

You can see that this means that when there is a joggle, there is also very slight bend in the stock rail at C. This angle is so small that it is normally ignored for construction purposes.

(For an A switch with 66" planing, and a tip of 3/8" thickness, that small angle is 1:176 = 0.3 degrees. For longer switches it is even smaller.)

If you like to use over-scale depth joggles, as many do, that angle will increase, and you might want to bear it in mind. It is more evident when there is a joggle in the straight (main-road) stock rail.

Martin.
40 years developing Templot. And counting ...

junctionmad
Posts: 247
Joined: Tue Dec 08, 2015 3:38 pm

Re: Modelling loose heeled switches in 4mm

Postby junctionmad » Mon Feb 18, 2019 2:10 pm

thanks , Martin , as always your a mine of information

Dave

Julian Roberts
Posts: 764
Joined: Sat Jan 09, 2010 11:33 pm

Re: Modelling loose heeled switches in 4mm

Postby Julian Roberts » Tue Feb 19, 2019 7:53 am

Thank goodness Martin has clarified the difference between sets and joggles. It was on 17th Feb last year that Martin put much of the same useful information on Tony's pointmaking thread.

viewtopic.php?f=5&t=5728&start=50

Now we know how much (deep) a joggle should be, and how long it is in each direction before normal track alignment is regained.

I'm not sure where else such real detailed guidance is to be found for making pointwork that uses joggles even though that may be the more frequent approach used by P4 modellers(?)

Julian Roberts
Posts: 764
Joined: Sat Jan 09, 2010 11:33 pm

Re: Modelling loose heeled switches in 4mm

Postby Julian Roberts » Tue Feb 19, 2019 10:59 am

With reference to what I just wrote above, what I would like to see is the information Martin has provided in this thread (and Tony Wilkins thread) being more readily visible to someone looking for guidance on the Scalefour website.

Keith gave links to the website in Tony's thread, but I can't see there the succinct facts Martin has given that are what we really need.



The last of these might be for some of us, could I suggest, a case of "too much information", some difficult to read. (Quite beyond me as a merely amateur modeller!) Whereas the links that show lots of pictures that are useful for getting those visual details correct, but not for explaining what is really going on - a real joggle is very subtle.

The Forum is great for keeping up with such myth-busting clarifications. But easily missed if you don't read it all the time, unless you know how to do a lot of researching. Simply a link from the website to these latest articles would be great. For example, Tony's thread updates and is miles clearer than his article on the website probably written many years ago.

I wonder what others think, and is this suggestion a practical one?

seanmcs
Posts: 24
Joined: Wed Dec 10, 2008 12:57 am

Re: Modelling loose heeled switches in 4mm

Postby seanmcs » Sun Feb 24, 2019 7:30 am

I think that if you are trying to produce GWR style loose heeled turnouts the procedure is different to the above. The switch rails need to be filed in the special loose heeled filing jig. This produces a switch rail that has its flat face to correspond with the straight of the stock rail. The running surface is concave curved at the radius of the switch rail for the selected turnout. These radii are in Smith's GWR Switch and Crossing practice. In the case of the straight switch rail, its filing ends up with a concave curve corresponding to the radius of the curved stock rail. After the theoretical point near the end of the switch rails the radius of the switch rail changes to the radius of the turnout and this continues all the way up to the end of the turnout. Likewise the curved stock rail starts with the switch radius then adopts its turnout radius at its theoretical point, also running all the way to the end of the turnout. This curve starts just at after the joggle.

As all my stuff is in store pending a move, this is as far as I have gone, in theory, though I am going to practice some joggles when the Brian Morgan frets arrive.

Thanks for this thread! Sean

junctionmad
Posts: 247
Joined: Tue Dec 08, 2015 3:38 pm

Re: Modelling loose heeled switches in 4mm

Postby junctionmad » Mon Feb 25, 2019 3:28 am

seanmcs wrote:I think that if you are trying to produce GWR style loose heeled turnouts the procedure is different to the above. The switch rails need to be filed in the special loose heeled filing jig. This produces a switch rail that has its flat face to correspond with the straight of the stock rail. The running surface is concave curved at the radius of the switch rail for the selected turnout. These radii are in Smith's GWR Switch and Crossing practice. In the case of the straight switch rail, its filing ends up with a concave curve corresponding to the radius of the curved stock rail. After the theoretical point near the end of the switch rails the radius of the switch rail changes to the radius of the turnout and this continues all the way up to the end of the turnout. Likewise the curved stock rail starts with the switch radius then adopts its turnout radius at its theoretical point, also running all the way to the end of the turnout. This curve starts just at after the joggle.

As all my stuff is in store pending a move, this is as far as I have gone, in theory, though I am going to practice some joggles when the Brian Morgan frets arrive.

Thanks for this thread! Sean


wow, my head spun reading that piece

seanmcs
Posts: 24
Joined: Wed Dec 10, 2008 12:57 am

Re: Modelling loose heeled switches in 4mm

Postby seanmcs » Wed Feb 27, 2019 12:52 am

I agree - it is very different to the more modern semi-flexible switches.

Some years ago I was in correspondence with the late David Morgan concerning his excellent underframes and other etches. He produced three papers on GWR old style loose heeled turnouts.These are available as downloads from the Society website. In one introduction he wrote:

"It is not widely appreciated that semi-curved switches came into use on the Great Western only in the 1930s and were used only for new construction. The older curved loose heeled switches remained in use into BR times."

"The Great Western regularly ran the turnout curve through the crossing and out on the turnout road, thus not all crossings were straight on straight."

"Sadly great Western layouts in general are not correctly modelled and one can only hope that David Smith's GWR Switch and Crossing Practice published by the Great Western Study Group , together with these instructions will help put that right."

Studying the design and dimensions is a bit of work, but overall there emerges the GWR's desire to provide a smooth running road for the comfort of passengers, with minimal wear and tear. Anyway, that's my take on it.

Exactly when I can try to put this prototypical design into practice, I do not know. Moving house is a bit tricky just now, here in OZ. I will report when I can.

Sean

User avatar
Martin Wynne
Posts: 821
Joined: Mon May 14, 2012 4:27 pm

Re: Modelling loose heeled switches in 4mm

Postby Martin Wynne » Wed Feb 27, 2019 5:09 am

seanmcs wrote:"It is not widely appreciated that semi-curved switches came into use on the Great Western only in the 1930s and were used only for new construction. The older curved loose heeled switches remained in use into BR times."

Hi Sean,

The GWR did not use semi-curved switches. The flexible switches introduced in the 1930s were fully curved switches.

All the GWR switches, old and new types, are available as pre-sets in Templot (and have been for 20 years). It is widely used, so the suggestion that GWR layouts in general are not correctly modelled is wide of the mark.

Templot is free.

"The Great Western regularly ran the turnout curve through the crossing and out on the turnout road, thus not all crossings were straight on straight."

This is called a "curviform" crossing in Templot.

Diagrams of switch and crossing types are at: http://templot.com/companion/real_track.php

cheers,

Martin.
Last edited by Martin Wynne on Wed Feb 27, 2019 5:27 am, edited 1 time in total.
40 years developing Templot. And counting ...

seanmcs
Posts: 24
Joined: Wed Dec 10, 2008 12:57 am

Re: Modelling loose heeled switches in 4mm

Postby seanmcs » Wed Feb 27, 2019 5:27 am

Martin:

Thank you very much. I'd like to explore the GWR turnouts on Templot. Possibly would save me making up my own templates, as the printed ones are unsuitable. Can I look at it with a Mac?

Sean


Return to “Track and Turnouts”

Who is online

Users browsing this forum: No registered users and 2 guests