Mid-Cornwall Lines - 1950s Western Region in 00

[Nick C]

On 13/04/2021 at 11:56, St Enodoc said:

Here's the tangent tool:

 

It's just a piece of card with the ends offset from a straight line. The idea is to place it between the curves so that each of the centre bits kisses the curve, then draw along them, remove the tool and join up the gap.

 

Why not just use a straight edge? Because it's far easier to see where a tangent kisses the outside of a curve than the inside.

 

I haven't actually tried it yet but I can't see why it shouldn't work.

That's a clever idea. Just make sure to allow for the thickness of the line, so you don't end up with a slight offset between the two halves.

[St Enodoc]

10 minutes ago, Nick C said:

That's a clever idea. Just make sure to allow for the thickness of the line, so you don't end up with a slight offset between the two halves.

That happens automatically with my standard of marking out.

[Stubby47]

This of course will give you a right curve / straight/ left curve, rather than a right curve / left curve.

 

Is it possible (or desired) to replace the straight tangent with a subtle reverse curve, each half easing off each of the main curves ?

 

Edited April 14, 2021 by Stubby47

[FarrMan]

2 hours ago, Stubby47 said:

This of course will give you a right curve / straight/ left curve, rather than a right curve / left curve.

 

Is it possible (or desired) to replace the straight tangent with a subtle reverse curve, each half easing off each of the main curves ?

 

Strictly it would be transition curves between the two arcs, but the difference between that, a straight arc and a reverse curve would be minimal.

 

Lloyd

[Martin S-C]

On 13/04/2021 at 11:56, St Enodoc said:

Here's the tangent tool:

 

It's just a piece of card with the ends offset from a straight line. The idea is to place it between the curves so that each of the centre bits kisses the curve, then draw along them, remove the tool and join up the gap.

 

Why not just use a straight edge? Because it's far easier to see where a tangent kisses the outside of a curve than the inside.

 

I haven't actually tried it yet but I can't see why it shouldn't work.

This confuses me. It looks mighty simple and mighty clever at the same time, hence why I'm flummoxed. Can you give a diagram showing where the tracks would be your tangentially connecting (or even tan-genitally!) or use some spare track lengths to do ditto please?

[Keith Addenbrooke]

On 14/04/2021 at 16:28, Martin S-C said:

This confuses me. It looks mighty simple and mighty clever at the same time, hence why I'm flummoxed. Can you give a diagram showing where the tracks would be your tangentially connecting (or even tan-genitally!) or use some spare track lengths to do ditto please?

While Australia sleeps, we can ponder.  I’ll take a punt and say I’d use it like this:

 

(Sorry, photos no longer available)

 

You can also match the outer rail (or even sleeper edge) at A to the inner rail at B (and vice versa).  To achieve this with a straight ruler, I think it would need to be exactly Gauge width at the touching points.  I can’t cut card that accurately.

 

Hope this is correct (I’ve made a right wally of myself otherwise of course...)

Edited September 26, 2022 by Keith Addenbrooke
Edited for text only as photos no longer available.

[KeithMacdonald]

On 10/04/2021 at 10:15, LNER4479 said:

I prefer to listen to the wireless or put on a gramophone record. Especially a Peter Handford Argo Transacord one...

 

How about this?

 

 

 

 

 

 

[LNER4479]

Prefer the second one ...

[Dr Gerbil-Fritters]

250 tons... feeble, them black fives I tell you... 

 

Looks in vain for hat and coat

[Clive Mortimore]

1 hour ago, Dr Gerbil-Fritters said:

250 tons... feeble, them black fives I tell you... 

 

Looks in vain for hat and coat

Hi Doc

 

Not sure if we can socially distance on the naughty step.

[Dr Gerbil-Fritters]

3 hours ago, Clive Mortimore said:

Hi Doc

 

Not sure if we can socially distance on the naughty step.

 

They'll just have to build us a bigger step, Clive.

[St Enodoc]

14 hours ago, Stubby47 said:

This of course will give you a right curve / straight/ left curve, rather than a right curve / left curve.

 

Is it possible (or desired) to replace the straight tangent with a subtle reverse curve, each half easing off each of the main curves ?

 

According to the Leeds MRS PW guru the short straight is the important bit here (no, I don't know either)...

 

11 hours ago, FarrMan said:

Strictly it would be transition curves between the two arcs, but the difference between that, a straight arc and a reverse curve would be minimal.

 

Lloyd

...although if there were more room transitions would be the correct answer. By the time I've laid the tracks, though, they won't be quite that precise anyway.

[St Enodoc]

10 hours ago, Keith Addenbrooke said:

While Australia sleeps, we can ponder.  I’ll take a punt and say I’d use it like this:

 

 

 

 

You can also match the outer rail (or even sleeper edge) at A to the inner rail at B (and vice versa).  To achieve this with a straight ruler, I think it would need to be exactly Gauge width at the touching points.  I can’t cut card that accurately.

 

Hope this is correct (I’ve made a right wally of myself otherwise of course...)

Almost spot on, Keith - thank you. The real challenge, though, is that the tangent points A and B are not known - if they were, the exercise is trivial as you can just join them up. The tangent tool will let me draw the tangent without knowing the tangent points. I can then measure the distance between a known offset on each side of the tangent point, halve it and so find the tangent point itself.

 

At the weekend I'll show what I mean with a couple of photos.

[FarrMan]

5 hours ago, St Enodoc said:

.although if there were more room transitions would be the correct answer. By the time I've laid the tracks, though, they won't be quite that precise anyway.

Exactly!

 

Lloyd

[Chamby]

I’ve not used the tangent tool before but however the initial alignment is achieved, you can’t beat getting your eye down to track level and just looking along it.  If the flow looks smooth, that’s the first test passed.  Then run a train of bogie coaches along it, and see how they behave.  Any irregularities will be shown up by the movement between the coaches as they traverse the formation.  

 

If you have the space, transition curves are well worth it.  Using a slightly tighter radius with transition entry and exit will look better than a single radius curve, especially when laid as a reverse curve.  The only time it won’t, is if you are right on the tolerance of the minimum radius that your stock is designed to cope with.  That happens with surprising regularity, especially if you are trying to squeeze as much in the available space as possible!

[FarrMan]

You could calculate it if you prefer. If you know the distance between the curve centres, D, and the radius of each curve, Ra & Rb, then the common tangent will cross the line between the two centres at a point a distance Db from the centre of the curve radius Rb. That distance can be obtained from

 

Db =D/(1+(Ra/Rb))

 

and the distance of this point from the tangent point on curve radius Rb, Tb, will be  obtained from

 

Tb = Sq Rt of ((Db*Db) - (Rb*Rb))

 

Ta = (Ra*Tb)/Rb

 

I have just produced an excel spreadsheet to do the calculations for you. Just change D, Ra and Rb to what you want, and the rest should change automatically.

 

Lloyd

 

 

Setting out tangent between two curves in opposite directions.xlsx

Edited April 15, 2021 by FarrMan
Correct formula for Tb

[Chamby]

On 03/04/2021 at 17:24, Chamby said:

We've had a lovely sunny day here in Cornwall and I decided to visit the local beach for a couple of hours, it was pleasantly warm when in the lee of the chilly winds that we're getting from the Arctic at the moment.

 

I returned to my modelling bench to resume detailing my new Hornby Grange locomotive, to find this!

 

 

I had stupidly left it sitting under the magnifying lens on my workbench, as the sun traversed the sky...  damaged beyond repair, I think - the smokebox has become mis-shapen as well.  

 

 

 

On 03/04/2021 at 19:37, TrevorP1 said:

 

@Chamby You must feel awful. In my junk box are these items. Both have slight damage - clips on boiler broken and fittings missing/broken on the tender but if they help they are yours. They can be in the post on Tuesday.

 

 

 

 

A happy ending, following my stupidly leaving my new 'Grange' on my modelling bench under the magnifying lens on a sunny day.  My profuse thanks to @TrevorP1 of this parish, who promptly rode to the rescue with the (rather large) parts from his 'spares' box, which have now been transformed into this:

 

 

6857 was mostly deployed on the Welsh Borders but was observed at Severn Tunnel Junction, so likely also made occasional further runs down into the West Country.  Tudor Grange was chosen for no other reason than it was the name of the school that I attended as a teenager in Solihull, during the 1970's.  The livery has been modified to BR lined rather than the original GW version, so lining was removed from the splashers, tender frames and the flared fenders on the tender, and a new totem applied.  6857 had a 4,000 gallon riveted tender for at least part of this period, but this 3,500 gallon version with flush rivets was also carried by the class, so may still be correct...  whatever, I am very happy with it as is.

 

Thanks again, Trevor.  She runs very sweetly too!

[Keith Addenbrooke]

8 minutes ago, Chamby said:

 

 

 

A happy ending, following my stupidly leaving my new 'Grange' on my modelling bench under the magnifying lens on a sunny day.  My profuse thanks to @TrevorP1 of this parish, who promptly rode to the rescue with the (rather large) parts from his 'spares' box, which have now been transformed into this:

 

 

6857 was mostly deployed on the Welsh Borders but was observed at Severn Tunnel Junction, so likely also made occasional further runs down into the West Country.  Tudor Grange was chosen for no other reason than it was the name of the school that I attended as a teenager in Solihull, during the 1970's.  The livery has been modified to BR lined rather than the original GW version, so lining was removed from the splashers, tender frames and the flared fenders on the tender, and a new totem applied.  6857 had a 4,000 gallon riveted tender for at least part of this period, but this 3,500 gallon version with flush rivets was also carried by the class, so may still be correct...  whatever, I am very happy with it as is.

 

Thanks again, Trevor.  She runs very sweetly too!

 

Really good to see how this worked out.  I must admit I’d not realised Tudor Grange was one of the Grange class locos - having also grown up in Solihull the name of the School was instantly familiar (although I personally went to a different School and in the 1980s).  I had an earlier Hornby Grange for some years - lovely loco but limited pulling power.  Keith.

Edited April 15, 2021 by Keith Addenbrooke

[Stubby47]

48 minutes ago, FarrMan said:

You could calculate it if you prefer. If you know the distance between the curve centres, D, and the radius of each curve, Ra & Rb, then the common tangent will cross the line between the two centres at a point a distance Db from the centre of the curve radius Rb. That distance can be obtained from

 

Db =D/(1+(Ra/Rb))

 

and the distance of this point from the tangent point on curve radius Rb, Tb, will be  obtained from

 

Tb = Sq Rt of (Rb*Rb)-(Db*Db)

 

Ta = (Ra*Tb)/Rb

 

I have just produced an excel spreadsheet to do the calculations for you. Just change D, Ra and Rb to what you want, and the rest should change automatically.

 

Lloyd

 

 

Setting out tangent between two curves in opposite directions.xlsx 5.05 kB · 1 download

 

This is assuming Ra & Rb are constants...  

[St Enodoc]

1 hour ago, FarrMan said:

You could calculate it if you prefer. If you know the distance between the curve centres, D, and the radius of each curve, Ra & Rb, then the common tangent will cross the line between the two centres at a point a distance Db from the centre of the curve radius Rb. That distance can be obtained from

 

Db =D/(1+(Ra/Rb))

 

and the distance of this point from the tangent point on curve radius Rb, Tb, will be  obtained from

 

Tb = Sq Rt of (Rb*Rb)-(Db*Db)

 

Ta = (Ra*Tb)/Rb

 

I have just produced an excel spreadsheet to do the calculations for you. Just change D, Ra and Rb to what you want, and the rest should change automatically.

 

Lloyd

 

 

Setting out tangent between two curves in opposite directions.xlsx 5.05 kB · 2 downloads

Thanks Lloyd. I'll have a go with the tool and compare the results with your spreadsheet.

39 minutes ago, Stubby47 said:

 

This is assuming Ra & Rb are constants...  

In my case they are, Stu.

Edited April 15, 2021 by St Enodoc

[FarrMan]

1 hour ago, Stubby47 said:

 

This is assuming Ra & Rb are constants...  

I was assuming that you were joining two arc curves together with a straight. If you are trying to join two curves, one or both being transition curves, that is a lot more complicated, and depends on the type of transition curve selected. If i had the time, I could try to work that one out for you for a specific case, but I had rather not try it, today especially, as I am helping my son move house today.

 

Lloyd

Edited April 15, 2021 by FarrMan
To avoid confusion

[FarrMan]

54 minutes ago, St Enodoc said:

Thanks Lloyd. I'll have a go with the tool and compare the results with your spreadsheet.

In my case they are, Stu.

I will be interested to know the results of your comparison.

 

Re transition curves, I might have a spreadsheet that does that already for roads. I produced it for my use in marking Highway Engineering exams. I was a great help to me to identify where they had gone wrong!

 

With regard to Chamby's comment about looking along the curve to check that it is smooth, reminds me of my road setting out days, where we would calculate the appropriate setting out data, and then set out with steel tapes (fibre tapes stretch - not as accurate) and theodolite reading to at least 20 seconds if not 5 or 1 second accuracy, so that it should have been spot on. ( an angle of one second gives an offset of about 1mm at 100 metres!) Then we looked round the curve that we had set out and adjusted it for smoothness by eye! These days they normally use GPS for setting out - supposed to be accurate to within a couple of mm. I'm sure that the old hands especially still eye it in for final adjustments.

 

Lloyd

[St Enodoc]

43 minutes ago, FarrMan said:

Re transition curves, I might have a spreadsheet that does that already for roads. I produced it for my use in marking Highway Engineering exams. I was a great help to me to identify where they had gone wrong!

I use cubic parabolas, as described in my ancient copy of the Scalefour Digest, with a fixed range of radii and shifts, cunningly chosen so that the overall length just fits on the piece of thick card from the centre of bolts of dressmaking material.

 

More detail back on page 6:

 

53 minutes ago, FarrMan said:

I was assuming that you were joining two arc curves together with a straight. If you are trying to join two curves, one or both being transition curves, that is a lot more complicated, and depends on the type of transition curve selected.

When I join two transition curves, I use the templates and set it all out by trial and error. I'll be doing this when I get further round the branch. If I remember I'll take some photos of that.

 

 

Edited April 15, 2021 by St Enodoc

[Keith Addenbrooke]

On 15/04/2021 at 09:58, FarrMan said:

You could calculate it if you prefer. If you know the distance between the curve centres, D, and the radius of each curve, Ra & Rb, then the common tangent will cross the line between the two centres at a point a distance Db from the centre of the curve radius Rb. That distance can be obtained from

 

Db =D/(1+(Ra/Rb))

 

and the distance of this point from the tangent point on curve radius Rb, Tb, will be  obtained from

 

Tb = Sq Rt of (Rb*Rb)-(Db*Db)

 

Ta = (Ra*Tb)/Rb

 

I have just produced an excel spreadsheet to do the calculations for you. Just change D, Ra and Rb to what you want, and the rest should change automatically.

 

Lloyd

 

 

Setting out tangent between two curves in opposite directions.xlsx 5.05 kB · 7 downloads

 

I don't use much maths in my present job, so find it hard to keep up with my kids, so I've taken a lunchtime maths test to see if I could work out for myself how @FarrMan's calculations work.  Starting with 'Ra', 'Rb' and 'D' (which I understand) I think I'm OK with the derivation of 'Db' and 'Ta,' but I've got a reversal in my formula for 'Tb'...? Have I missed something:

 

 

 

Sorry, not much to do with trains.  Just curious, Keith.

Edited September 26, 2022 by Keith Addenbrooke
Reinstating pictures

[TrevorP1]

3 hours ago, Chamby said:

 

 

 

A happy ending, following my stupidly leaving my new 'Grange' on my modelling bench under the magnifying lens on a sunny day.  My profuse thanks to @TrevorP1 of this parish, who promptly rode to the rescue with the (rather large) parts from his 'spares' box, which have now been transformed into this:

 

 

6857 was mostly deployed on the Welsh Borders but was observed at Severn Tunnel Junction, so likely also made occasional further runs down into the West Country.  Tudor Grange was chosen for no other reason than it was the name of the school that I attended as a teenager in Solihull, during the 1970's.  The livery has been modified to BR lined rather than the original GW version, so lining was removed from the splashers, tender frames and the flared fenders on the tender, and a new totem applied.  6857 had a 4,000 gallon riveted tender for at least part of this period, but this 3,500 gallon version with flush rivets was also carried by the class, so may still be correct...  whatever, I am very happy with it as is.

 

Thanks again, Trevor.  She runs very sweetly too!

 

Pleased it's gone well  Phil. Nice work.