Knuckle bend radius

[Martin Wynne]

I have spent a damp grey Sunday implementing a feature in Templot which I skipped over when I originally wrote the generator code 38 years ago. It has been lurking on my to-do list ever since, making me feel uncomfortable every time I noticed it there. A kick up the bum on the Templot Club forum has finally got me to do something about it.

The next program update (coming soon) will correctly show the bend radius in the wing rails at the knuckle:



This means that the knuckle gap will now be correctly shown wider than the flangeway gap. There is a table of crossing data showing bullhead knuckle gaps on this page:

http://templot.com/martweb/templot_forum.htm

However, for the normal setting I have instead adopted the GWR practice* of specifying the length and radius of the knuckle bend rather than the resulting knuckle gap. This is more convenient for model use where flangeway gaps are seldom dead-scale (even in P4).

*where the knuckle bend radius in feet matches the crossing angle, or in other words the length of the bend is 1ft. i.e. for a 1:7 crossing the knuckle bend radius is 7ft.

If it is desired to match the knuckle gaps in the table, it will be necessary to do your own calculations and enter the required custom bend radius. This will make sense only if you are using exact-scale flangeways as in S4-X.

Likewise some pre-group companies (e.g. NER) used a more generous knuckle radius, which again will require a custom setting.

The result is rather more noticeable than I expected, and I am now wishing that I had done this years ago. I have added some template markers to show the extent of the knuckle bend, and its original sharp centre:





There is no change to the geometry or alignment of the template. If it is desired to revert to the previous template design showing the geometric knuckle, the sharp option setting can be used. Those templates are not actually wrong, because the templates have always been intended as construction alignment guides and not descriptive drawings. That's why they are called templates. For example both switch blades are always shown closed for that reason.

regards,

Martin.

[Russ Elliott]

I'm not a Templot practitioner, but the implementation of this in real model railway rails is a giant leap toward preventing flanges snatching in the knuckle area.

Well done, Martin. May you have many more damp grey Sundays.

[Julian Roberts]

The result is rather more noticeable than I expected,
regards,

Martin.

I'm not a Templot practitioner, but the implementation of this in real model railway rails is a giant leap toward preventing flanges snatching in the knuckle area.

I hate to be the idiot asking the stupid question.

I can see that the radius has to be known in theory and at full size, but for the modeller who wants trains to run properly but who is not going to examine the knuckle microscopically, does this change imply a different way of making the knuckle bend in practice?

I had been given to understand the right way of making the bend is to hold the rail in some smooth jawed pliers and make a perpendicular bend, without making a notch to create a sharper crisper angle as some espouse.

That the flangeway gauge can sit on each road through the crossing as per the photo below (though perhaps only just through the knuckle on the diverging road), seems to me to be the requirement for smooth running, as far as the knuckle construction itself is concerned - or am I being hopelessly simplistic?

[grovenor-2685]

I can see that the radius has to be known in theory and at full size, but for the modeller who wants trains to run properly but who is not going to examine the knuckle microscopically, does this change imply a different way of making the knuckle bend in practice?

I had been given to understand the right way of making the bend is to hold the rail in some smooth jawed pliers and make a perpendicular bend, without making a notch to create a sharper crisper angle as some espouse.

It only makes a difference to those who mistakenly espouse notching and making sharp bends or perhaps for those sufficiently driven that they create jigs to allow them to bend the exact radius (with the rails inclined correctly each side of the bend.
The rest of us can continue doing just as you describe and see if anyone picks up that our bend radius is a scale inch or so out or the inclination not quite 1:20 .
Regards

[RichardS]

Being in the process of making my first point and learning at every-step of the way I suspect that the status-quo that I have discovered where the theory is fine and vitally necessary but the practical application is a little less precise because after all there are only two decimal places in 18.83 will ensue.

Martin is wise to include this in Templot - which is also applicable to larger gauges don't forget, but track builders will always make things work 'on site' even if 'not exactly not wrong.'

Back to filing this morning.....

[Julian Roberts]

The rest of us can continue doing just as you describe

Pleased to have your agreement Keith!

May you all have many more damp grey Sundays. In Scotland the word for that weather is "dreich" - it's the default setting most days up here in the land of the Caledonian Railway, icebergs and oxcarts -

viewtopic.php?f=19&t=5232

and scroll down