Theory of General Minories

[Satan's Goldfish]

Peco small-radius points are roughly 30" radius, yes?

I maybe wrong, but I always thought:

 

Small=24"

Medium=36"

Large=60"

[ejstubbs]

Peco small-radius points are roughly 30" radius, yes?

 

Nope, 24" according to Peco themselves.  Medium radius are 36".  (Those are nominal radii ie for the overall geometry of the turnout.  Certain elements of a turnout may involve tighter radii.)

Edited September 13, 2017 by ejstubbs

[Karhedron]

Thanks for the correction. Don't know where I got 30" from.  

[Satan's Goldfish]

Thanks for the correction. Don't know where I got 30" from.  

In track of the streamline curved points are 30".

[Harlequin]

I've examined the (non-curved) Peco Streamline geometry in depth and it all seems perfectly logical:

 

Small  = 24in = 2ft

Medium = 36in = 3ft

Large  = 48in = 4ft

 

All have 12degree turnout angles, all give 2in spacing between track centres when joined back to back (e.g. as a crossover) and some have short straight sections in them to achieve the 12deg turnout and 2in spacing.

 

I don't know why the internet thinks that Large radius points are 60in (I've seen it repeated many times). I think it's just one of those memes that gets propagated around and taken as fact.

 

PhilM

Edited September 13, 2017 by Harlequin

[Flying Pig]

Better tell Peco, then, as they appear to be completely taken in

 

http://www.peco-uk.com/product.asp?strParents=3309,3322&CAT_ID=3327&P_ID=17449

[martin_wynne]

As best can be determined without access to the actual manufacturing drawings:

 

small radius = 575.4mm radius in the rails = 22.7"

 

large radius = 1137.9mm radius in the rails = 44.8"

 

Sorry I don't have any figures for the medium radius.

 

 

 

 

Martin.

[Harlequin]

Better tell Peco, then, as they appear to be completely taken in

 

http://www.peco-uk.com/product.asp?strParents=3309,3322&CAT_ID=3327&P_ID=17449

 

It depends on what they mean by "Nominal", doesn't it?

 

I've used their downloadable PDF templates (http://www.peco-uk.com/page.asp?id=tempc75) and overlaid exact fixed radius curves at 1:1 scale and this is the result:

 

The large yellow shape is a 12degree sector of a very large circle and it's radius is... 4ft. You can see it's not an exact fit but I believe that's partly due to parallax errors in their photo and maybe manufacturing tolerances. With any larger radius it's not possible to achieve the 12degree turnout angle within the 2in track spacing constraint.

 

Edit: I'm referring to radii to track centres - not to inner or outer rails.

 

Therefore either the Peco product description is incorrect or the word "Nominal" adds some qualification that's not immediately obvious.

 

PhilM

Edited September 13, 2017 by Harlequin

[martin_wynne]

Better tell Peco, then, as they appear to be completely taken in

 

http://www.peco-uk.com/product.asp?strParents=3309,3322&CAT_ID=3327&P_ID=17449

 

They call that "nominal" radius -- i.e. what you get by looking at it from a yard away and guessing, not what you get by actually measuring the rails.

 

School geometry tells you that to achieve 1" offset in a turn of 12 degrees, the absolute maximum radius which is physically possible is 45.76".

 

R =  1 / ( 1 - COS(12)  )

 

Martin.

Edited September 13, 2017 by martin_wynne

[Harlequin]

They call that "nominal" radius -- i.e. what you get by looking at it from a yard away and guessing, not what you get by actually measuring the rails.

 

School geometry tells you that to achieve 1" offset in a turn of 12 degrees, the absolute maximum radius which is physically possible is 45.76".

 

R =  1 / ( 1 - COS(12)  )

 

Martin.

 

Yes, completely agree.

 

In millimetres: Half track spacing is 25.4mm. So 25.4/(1-cos(12)) gives 1162.34mm, which fits their template photo perfectly and is "nominally" 1219mm == 4ft.

 

PhilM

[Chimer]

Fwiw and the hell of it, I just set up a 90 degree turn in in XTrackcad using large radius lefts (OK, 96 degrees altogether) and the centreline "curve" (which of course includes the straight bits at each toe) takes up 49".  I'd always assumed the "nominal" figure Peco quote would relate to this measurement - but obviously not.

 

Chris

[Flying Pig]

Fair enough - the maths does not lie. Thanks for the clarification.

[martin_wynne]

Fwiw and the hell of it, I just set up a 90 degree turn in in XTrackcad using large radius lefts (OK, 96 degrees altogether) and the centreline "curve" (which of course includes the straight bits at each toe) takes up 49".  I'd always assumed the "nominal" figure Peco quote would relate to this measurement - but obviously not.

 

Chris

 

You could get it a fraction bigger by inserting insulated rail joiners between them.

[Pacific231G]

Thanks for the thoughts and the photos.

 

My box design currently only has a sheet of ply at the platform end so it would be very easy to "knock through" and extend the platform lines:

Minories 9 platform hole.png

 

I will take your advice and check the throwover on the route into platform 1 before I go much further. The points in my 2D drawings are exact (scale) geometrical templates of Peco streamline parts and I can print the entire drawing at real world size using Acrobat's "Poster" printing feature. Then I can lay the printout on a table and place real points and track on top. Slight problem: I only have one "coach" at the moment but that can be overcome.

 

I could perhaps model the coaches and track in the computer to check throwover but that would be going too far. You've got to draw the line somewhere!

 

I am really concentrating on the 1989 revision of Minories from "60 plans for small locations", when I guess Peco Streamline points had the same geometry as the present day. That design is labelled, "Minimum radius 3ft 0in", but I'm pretty sure it's not possible to fit the pointwork onto the stated baseboard size with only 3ft radius curves using standard Streamline parts. In the Introduction to that booklet CJF says, "All layouts are designed for 4mm scale, OO gauge, and [snip] use Peco Setrack and Streamline products. [snip] I have chosen 1ft 3in absolute minimum [radius], and have regarded 2ft as the optimum". So I think that "3ft" label on the Minories plan was probably an oversight. (Perhaps you or one of the other RMWeb experts could shed some light on this...?)

 

PhilM

Hi Phil

 

I'm looking at the two plans in the 1989 60 Plans for Small Locations now and the simple four point throat arrangement is shown as fitting into three feet.

 

For buffer locking trials, I've got this laid out on a board with Peco medium radius points, using insulating rail joiners where appropriate. The total length of the actual pointwork is thirty four and a half inches. On the 7ft simple version (SP35 in my copy) the right hand board is 42 inches long and, if you use medium radius points straight out of the box, the point for the kickback siding won't quite fit. However, with a little judicious pruning of stock rails to bring the frogs of that point and the one controlling access to platforms 2 & 3 point closer together. it will fit but only just. So, the minimum three foot radius quoted on the plans is possible . Keeping all the pointwork on one board would also simplify the wiring along with any mechanical way of controlling points from a single "signalbox" such as wire in tube or even working point rodding. The only interboard electrics required would be for traction current.

 

Laying out the Minories throat with small radius points gives a total length of pointwork of 28 inches.

Edited September 14, 2017 by Pacific231G

[tc]

 

I think Tom Cunnington and the other MRC members who built the EM gauge Minoriesfor the 50th anniversary used Templot rather than off the shelf templates for their pointwork. They followed CJF's original plan as closely as possible but slightly lengthened the points to give more generous curves which lengthened the scenic boards to a total of 7ft 10ins.

Indeed I did, I think A5.5 points which was as tight as I dared, but built originally in a single block to ensure the curves flowed. Apologies if this has been covered somewhere previously in the thread, but to quote Cyril himself:

 

Fifty years ago I set out to design a layout based on the plan, after a lot of doodling I decided a three-track terminus was the answer. A little more doodling produced the neat arrangement of crossovers which, quite adventitiously, ensured that any movement only involved one reverse curve. The original scheme was for TT gauge, but most copies have been in 4mm or 7mm scale. I’ve lost count of the number of people who have come up to me at shows and thanked me for the design.

 

 

I believe therefore to be strictly accurate one should use set track points to get it into the length, as the TT points were the same geometry as the OO ones, but I am very ready to be corrected.

 

Tom.

[Harlequin]

Hi Phil

 

I'm looking at the two plans in the 1989 60 plans for small locations now and the simple four point throat arrangement is shown as fitting into three feet.

 

For buffer locking trials, I've got this laid out on a board with Peco medium radius points, using insulating rail joiners where appropriate. The total length of the actual pointwork is thirty four and a half inches. On the 7ft simple version (SP35 in my copy) the right hand board is 42 inches long and, if you use medium radius points straight out of the box, the point for the kickback siding won't quite fit. However, with a little judicious pruning of stock rails to bring the frogs of that point and the one controlling access to platforms 2 & 3 point closer together. it will fit but only just. So, the minimum three foot radius quoted on the plans is possible . Keeping all the pointwork on one board would also simplify the wiring along with any mechanical way of controlling points from a single "signalbox" such as wire in tube or even working point rodding. The only interboard electrics required would be for traction current.

 

Laying out the Minories throat with small radius points gives a total length of pointwork of 28 inches.

 

I found that the 12 degree angle provided by the points back-to-back alone was not enough to swing the lines across the board towards platform 3. So I introduced a ~32mm section of track giving another 3 degree turn. Then, to fit the critical ladder of 5 points and 32mm curve into the 42in half-baseboard requires three medium and two short radius (see my PDF in post #459 http://www.rmweb.co.uk/community/index.php?/topic/60091-00-minories-track-plan-wanted/page-19&do=findComment&comment=2842274).

 

I looked at using the Streamline curved points but I don't think they will help because the outer radius gives 8 degrees turn and the inner radius gives 20 degrees - too much for most positions in the Minories geometry.

 

I'm also looking at clipping the points, as suggested, but that seems a bit like cheating to me when the implication from "60 small plans" is clearly that it can be done using standard parts. Doing that also introduces new problems with track spacing.

 

It all depends on whether the reverse curve into platform 1 will cause running problems for coaches and I'm currently joining points and track together on my dining room table to investigate.

 

Edit: I can see that buffer locking would be an issue if coaches were pushing each other buffer-to-buffer, but not if tension hook couplings are pushing each other. (However I realise that's not necessarily desirable...) Also, swapping one of the Small radius points for a Medium would not prevent buffer locking.

 

PhilM

Edited September 14, 2017 by Harlequin

[Satan's Goldfish]

If buffer locking only occurs when pushed over that location, could an appendix be added to the rule book stating all propelling shunt moves involving platform 1 must be too/from the departure line? Alternatively, allocate the departure line as the only route for shunt moves; the pilot road can only access it anyway and it keeps the arrivals line clear (possibly a little easier from a signalling/signalman's perspective?)

[Harlequin]

Ping! <Lightbulb appears over head>: If buffer locking is considered a problem then the reverse curve needs to be reduced/eliminated and to do that more space is needed between curves. So, smaller radius points (i.e. shorter points) would actually help. Longer, larger radius points make the problem worse...

 

I need to investigate this idea (but I'd better get on with some work first).

 

PhilM

[Clive Mortimore]

If buffer locking only occurs when pushed over that location, could an appendix be added to the rule book stating all propelling shunt moves involving platform 1 must be too/from the departure line? Alternatively, allocate the departure line as the only route for shunt moves; the pilot road can only access it anyway and it keeps the arrivals line clear (possibly a little easier from a signalling/signalman's perspective?)

Hi

 

I made the departure line the only line for shunting on Sheffield Exchange Mk1, not for buffer locking but looking at real locations it was common only to shunt from the departure line to reduce the amount of signalling and in turn the workload of the signalman.

[Pacific231G]

Indeed I did, I think A5.5 points which was as tight as I dared, but built originally in a single block to ensure the curves flowed. Apologies if this has been covered somewhere previously in the thread, but to quote Cyril himself:

 

Fifty years ago I set out to design a layout based on the plan, after a lot of doodling I decided a three-track terminus was the answer. A little more doodling produced the neat arrangement of crossovers which, quite adventitiously, ensured that any movement only involved one reverse curve. The original scheme was for TT gauge, but most copies have been in 4mm or 7mm scale. I’ve lost count of the number of people who have come up to me at shows and thanked me for the design.

 

 

I believe therefore to be strictly accurate one should use set track points to get it into the length, as the TT points were the same geometry as the OO ones, but I am very ready to be corrected.

 

Tom.

Thanks Tom. I've been looking high and low for my notes of a conversation I had with you about Minories GN a few years ago. I couldn't remember what points you'd used  but B5.5 does sound familiar.

 

I don't think you'd need to go down as far as set track to get the original Minories into the length   Setrack didn't exist in 1957 though Tri-ang's own points were probably equivalent but the "scale" track available for  TT-3's launch the month before the Minories article appeared included Gem and Peco spiked track. Gem's  Universal Turnouts for TT-3 (twelve shillings and sixpence) were 15" radius and 4 ¹/₂" long, while Peco's Spiked track included a standard 19" radius point kit (eight shillings & sixpence needs no soldering) which was 5" long.  Sydney Pritchard would have surely wanted a plan that could be built with his new TT-3 track and Minories certainly could.

 

Measuring the two original Minories plans against the scales shows, for  TT-3,  a length for the pointwork of two feet from the end of the board so either Gem or Peco would have fitted comfortably. In 00 the pointwork is about 32 inches long from the scale and you can assemble it with Peco Streamline  small radius (2ft nominal) points in 28 inches so again a comfortable fit. 

I happen to have a few of the then current Pecoway BRMSB 00 points which were nominally three foot radius and with those the Minories throat would have required 34 inches. Since the final set of points is nowhere near the board end,the plan could certainly have been built using any of Peco's OO gauge track (or any other scale track come to that)

Edited September 14, 2017 by Pacific231G

[Pacific231G]

Ping! <Lightbulb appears over head>: If buffer locking is considered a problem then the reverse curve needs to be reduced/eliminated and to do that more space is needed between curves. So, smaller radius points (i.e. shorter points) would actually help. Longer, larger radius points make the problem worse...

 

I need to investigate this idea (but I'd better get on with some work first).

 

PhilM

Hi Phil

Given Peco Streamline's particular geometry it shouldn't make much difference with a crossover occupying the same overall length but I've just tried it with the critical back to back points  1. with two medium radius points connected directly and 2. with two small radius points separated by a short piece of straight track to give the same overall length.

 

What I found with two pairs of identical coaches (Roco Bruhats) just over 10 inches long and fairly close coupled was that the throwover was significantly worse with the separated small radius points than with the back to back mediums .

 

 

This is interesting as NR's track standards do call for a certain amount of straight track between reverse curves with less than a certain radius

 

Using a pair of coaches without their vestibule connections, the relative positions of the  buffers themselves are perhaps a bit clearer. 

 

All these photos show the point of maximum deflection.

 

with the small radius points separated by a short straight the buffers are well out of contact so you would get buffer locking but with the two mediums it's just on the edge.

To do this test really thoroughly I'd need to remove the couplers and push the coaches through the two back to back points.With Kadees (or any other "buffing" coupler)  I wouldn't actually suffer buffer locking but the amount of throwover with almost no overlap between the vestibule connections does look rather silly. I don't know whether this would be quite so obvious with 00 coaches with wider vestibules. 

 

The best results I could get in the same length as four medium radius poiints was by subsituting a medium Y for the right hand of the two back to back points. This definitely produced no buffer locking with these H0 coaches and the vestibules overlapped by at least 60%. This is a slight variation on the classic Minories plan but still meets Cyril Freezer's own description that "the trains only have to wiggle once" .

 

I should say that this isn't an academic exercise. I'm trying to cram a mainline terminus plus fiddle yard quart into a four metre pint pot and the shorter the approach pointwork the longer can be the trains (but I don't want to use slips). I'm looking at a Pacific with  four or five coaches and a four wheel baggage car (and various view blockers to disguise the shortness of trains).  I'd love to base this on Minories but am also looking at a main line development of John Charman's Charford which seems in turn to have been inspired by the old Fort William. There were a number of single track termini that handled main line expresses in France but the busyness of Minories seems more suggestive of a city.

Edited September 14, 2017 by Pacific231G

[Budgie]

Isn't it recommended that you use a straight the length of the longest vehicle to separate the curves of a S bend? So in your example the short straight should be at least 10" long.

[St Enodoc]

Isn't it recommended that you use a straight the length of the longest vehicle to separate the curves of a S bend? So in your example the short straight should be at least 10" long.

The bogie centre distance should be enough.

[Pacific231G]

Isn't it recommended that you use a straight the length of the longest vehicle to separate the curves of a S bend? So in your example the short straight should be at least 10" long.

Hi Jane

That would eliminate the S curve buffer locking problem as each of the back to back turnouts would be effectively separate* and I agree with St.Enodoc that the bogie centre length would be long enough to achieve that*. That's really the secret of Cyril Freezer's Minories for all but one route through it. However, if that were applied to that one route (inbound to platform one) it would rather defeat the aim of creating the shortest possible throat. .For that we're looking for the best compromise.

 

AFAIK any normal crossover will include some straight between the V noses but, except possibly for high speed crossovers, that's not going to be very long. The straights between back to back turnouts can also be fairly short though "as long as possible" is recommended.

 

I used to travel fairly often over the junction at West Ealing between the slow lines of the GW main line and the Greenford Branch. That consists of a facing crossover and a branch turnout and on a down train you got to experience both of them. Looking at it on Google Earth, the distance  between the toes of the two back to turnouts on the up slow line is no more than 14 metres and the straight between the frogs of the crossover is only about10-11 metres, both far shorter than the ~ 23M overall length of  the class 165 DMUs that run the branch service (now normally terminating in its own bay at West Ealing) and Mk 3 coaches in the class 43 HSTs that regularly work ECS over the branch.

 

I'm no expert on PW design but I've been looking at the NR Track Design Handbook, a fascinating document.

 

For a reverse curve (S bend)  in plain track there are rules about the transitions required which should be a mirrror image of one another but for normal transitioned curves there doesn't have to be any actual straight track. You can see this quite clearly on a number of lines, my favourite being parts of the Tattenham Corner  branch south of London, that look rather as if the builders had run out of straights and were using up their spare curved track just like we did with our trainsets as children .(51⁰19'12.89 N 0^o08'48W is a good example)

 

When  a reverse includes a curve of less than 160m (just over 2m in 00 and 1.8m in H0 so almost all curves on our layouts ) then there must be a straight track at least 3m long between the opposite curves. 

 

In the case of two back to back turnouts forming a reverse there are rules about the length of straight track between their toes related to speed (actually based on the time to pass between toes)

For 15MPH it would be an absolute minimum of 3 152mm and for 40MPH 12 875mm. In practice the minimum  would normally be 6 140mm for 15-25MPH and 10 640mm for 30-50MPH. The normal lengths would be about 50% greater than these and the desirable length greater still .

 

The other factor is the need to avoid buffer locking (or exceeding the maximum allowable lateral movement between corridor connections)  The maximum allowable vertical displacement  is normally 300mm (~1 foot) and there are rather complex mathematical formulae that relate the two curves, the length of straight track, the length over buffers of the vehicle and the distance between bogie centres.

 

I reckon the lateral movement between the two halves of the Greenford Branch train sometimes got quite close to a foot!

 

* update. If you tried to use a vehicle length straight and compensated with very sharp points that itself could give buffer locking. I've just tried running a pair of the coaches I used to test the throat across a simple small radius Peco point with straight track off both ends so no reverse curve. The buffers stayed in contact so no buffer locking but the lateral movement was about twenty inches in H0 with the corridor connections about 50% out of alignment ; with Setrack points I think you really would need buffing couplers.

Edited September 15, 2017 by Pacific231G

[Harlequin]

Ping! <Lightbulb appears over head>: If buffer locking is considered a problem then the reverse curve needs to be reduced/eliminated and to do that more space is needed between curves. So, smaller radius points (i.e. shorter points) would actually help. Longer, larger radius points make the problem worse...

 

I need to investigate this idea (but I'd better get on with some work first).

 

PhilM

 

I found that using all small radius points with a straight(ish) section of track between the problem points, with the restriction that it all still has to fit onto the 42in Minories '89 baseboard, doesn't fix the problem unfortunately:

Using small points still doesn't create room for a long enough straight section.

 

If I had more room a straight section (the length of the bogie-centre-to-bogie-centre) would undoubtedly fix the problem.

 

But I want to be faithful to the Minories '89 design and so I have to accept the reverse curve and the limitations that it imposes.

Either:

• Use tension hook couplings (or similar) that can be propelled over that route without problems.

Or:

• Just don't propel any long vehicles with large outswings over that route - and that might not be so bad in practice because:
  • They can be pulled over that route in either direction (so long as corridor connections are not a problem)
  • Shorter vehicles are not a problem
  • Long vehicles must be propelled into or out of platform 1 from the departure line

 

As you say, David, some compromise is inevitable trying to fit any interesting pointwork into such a small area and I guess we've explored the limits and found out where the biggest compromise has to be made! I hope you've got a bit more leeway to fit your terminus and fiddle yard into your design space. Do you have a track plan?

 

PhilM

Edited September 15, 2017 by Harlequin