Length of straight heel switch

[Julian Roberts]

Hi Martin

I don't know whether this is a fault or something I'm not understanding or doing right.

I have made a track plan for Garlieston station with straight heel switches, for our club. One of the members is making the one on the left. It is an 18ft switch but on the template apparently measures 120mm from blade tip to joint, and is after 14 sleepers. The planing length is the expected 44mm he says.

Playing with the F5 function the joint is only 8 sleepers after the tips on a 15ft switch. On a 24ft switch the joint is in the same place as the 18ft, after the 14th sleeper. A 30ft switch takes the joint one sleeper further.

I've downloaded the latest version of Templot but this file was made in June.

I am sorry to bother you as I expect the obvious thing to do is for him to measure 72mm from the tips and make the joint there. But I thought I should ask as if there is a flaw you'd want to know.

All best
Julian

 

[Martin Wynne]

I have made a track plan for Garlieston station with straight heel switches, for our club. One of the members is making the one on the left. It is an 18ft switch but on the template apparently measures 120mm from blade tip to joint, and is after 14 sleepers. The planing length is the expected 44mm he says.

@Julian Roberts

Hi Julian,

The dimensions for the straight loose-heel switches in Templot are taken from Table 7 on p.81 of BRT3, 1964 (diagram on p.80).

If you don't have a copy of BRT3, here's one:

https://www.abebooks.co.uk/servlet/BookDetailsPL?bi=30918520548

In all those cases the virtual heel of the switch is not at the switch rail joint. In the first 3 cases (9ft, 12ft, 15ft) the switch rail is 2ft-6in longer than the switch size, i.e. for a 12ft switch the switch rail is 14ft-6in long to the joint. The virtual heel at 12ft from the tip is the place on the rail where the offset is 4.5" from the stock rail. So the planing angle is 144/4.5 = 1:32. The rail then continues for a further 2ft-6in to the joint (in which it is curved at the turnout radius).

For the 18ft switch the switch rail continues 12ft beyond the virtual heel to a length of 30ft from the tip, to the loose heel joint. The 24ft and 30ft switches have similar long switch rail lengths.

At the switch heel offset of 4.5" with rail 2.75" wide, the space matches the flangeway of 1.75", so if the heel joint was placed there, there is a risk of wheel-back contact with the joint and possible damage to it. Moving the rail joint further along by lengthening the switch rail means the switch opening is more than the flangeway, and clear of wheel backs.

But most pre-group drawings do show the loose-heel switch rail matching the switch size in length to the joint, for example those in the NER 1912 reprint. But in that case the heel offset is 4.75" giving 2" clearance, and for a 12ft switch a planing angle of 144/4.75 = 1:30.3. If you want to model those switches, you need to create custom switches in Templot.

Sorry I don't know from which prototype company the PWI got the dimensions for Table 7 in BRT3. Except the 30ft switch, which matches the GWR 30ft switches.

cheers,

Martin.

 

[Martin Wynne]

@Julian Roberts

p.s. It turns out that I've written all that stuff before, at:

https://85a.uk/templot/archive/topics/topic_2776.php#p19351

and see the rest of that topic for more info.

I could have saved myself about an hour of typing and referring to books and diagrams if only I had been able to remember that topic. Or thought to do a search first.

Martin.

 

[Julian Roberts]

Hi Martin

Thank you for your reply. I was thinking of making this further question on the S4 website to see if anyone else was interested in replying too, but I'll put it here first.

I'm really sorry to confess I don't understand all of this, and I think the link you gave goes into some quite exotic detail that is more than I need. But I do understand the issue that the joint distance from the stockrail has to be more than the flangeway, which I think is the most important bit(?)

I wonder if there is a mistake in the tables. The reason I say that is that I've taken screenshots of all 6 switches, made from the turnout that Templot provides when first switching it on and using the F5 function. Here they are below. The gaps between stockrail and looseheel joint are all around 1mm (plenty of clearance) except the 18ft one which is nearly 2mm (far more clearance than needed). I've used the ruler for the first time. I'm not claiming I've used it particularly accurately.

To find a similar gap on the 18ft switch it has to be between the 11th and 12th sleeper from the tip, 90mm or so from the tip.

When I used the F5 to lengthen the switch from 18 to 24 ft the ruler (which was still in its 18ft place) happened to line up exactly with the rails as the last shot below. I wonder if that says anything...

My original question was, putting it another way, why is the stockrail to looseheel joint gap so big on the 18ft switch? - much more than with the others.

Please don't waste your time (which we all value so much) in answering if I just need to spend more time studying what you've written already!

Cheers
Julian

 

[Martin Wynne]

I wonder if there is a mistake in the tables.

@Julian Roberts

Hi Julian,

I agree it looks odd, but I would be wary of assuming a mistake in BRT3. It was written in the days when folks were expected to get things right.

Sometimes odd-looking dimensions are explained by other factors, such as the chairing arrangements, or other rail lengths.

For example, the common size with an 18ft switch would be a 1:12 turnout. If you look at an 18ft-12 turnout you will notice that the closure rails between the switch and the V-crossing are just shy of 60ft long. If the switch rails were shorter, the closure rails would have to be longer, and there would be a need for an additional rail joint in the closure rail.

I don't know where the Table 7 dimensions came from, or how to find out, but it's interesting to look at the GWR. Their original old-type 18ft loose-heel straight switch had a 20ft long switch rail, at 1:48 angle. In the 1930s the GWR re-designed the 18ft switches as flexible switches, but using the same geometry as the older switches. This required a longer switch rail of 28ft-9in to provide the fixings for the non-flexing part.

I'm wondering therefore, whether the 18ft and 24ft straight switches in Table 7 are in fact flexible switches, rather than loose-heel? Nowhere in BRT3 does it say one way or the other, but it's noticeable that in both cases the switch rail is significantly longer than the switch size. Clearly the shorter sizes are loose-heel, because the switch rails are not long enough to flex.

The 30ft straight switch is essentially the GWR design anyway.

If you wanted some definite known prototype to copy, and don't want the usual REA flexible D switch, you could look at the NER switches. They had 17ft-6in switches in both traditional loose-heel and flexible form. Full details for both are in the NERA 1912 reprint, from which you could create custom switches in Templot.

cheers,

Martin.

 

[Julian Roberts]

Very many thanks Martin. I apologize for diverting you from the much more interesting task of developing the creation of real model track directly from Templot. As an aside however, the CR drawing that was posted in the interlaced sleepered turnout thread some years ago shows the knuckle rails quite a bit longer than REA standard, so there the closure rails would be that bit shorter This is my interpretation of that drawing - my turnout was a different size so it is far from exactly as per the drawing.

Please don't spend any more time replying!

Cheers
Julian