Locking question for MKII Frame

Discussions of the prototypes and how to model them. Show us how you do it.
jgunby

Locking question for MKII Frame

Postby jgunby » Sun May 24, 2020 2:46 pm

Right I'm at a bit of a loss with this one and I think it might just be me having a blank moment or not getting something.
Being a "youth" I have gone for the cheaty locking design option of modelling it in Fusion 360 where I can test all of the locking by turning contact on; now am sure that this is a horribly innefficient way of locking this frame but it is a way that should be fairly easy to make and troubleshoot (feel free to correct me...).

Untitled-1.jpg


My problem...
I have no idea how to have 10 free when 1 is reversed and 6 is normal, but have 10 locked normal when 1 is reversed and 6 is also reversed.

For info, 1 is an outer home, 6 are crossover points and 10 is a shunt signal on the far side of the crossover opposing 1.
If there is something there that doesn't make sense ask and I will do my best to clarify.

Plus just for fun here is the current state of my MKII Frame and Locking kit, not quite there yet and not happy with a few bits but works alright (much to my surprise)!

IMG_4370.jpg


Ta!
Joe
You do not have the required permissions to view the files attached to this post.

User avatar
grovenor-2685
Forum Team
Posts: 3918
Joined: Sun Jun 29, 2008 8:02 pm

Re: Locking question for MKII Frame

Postby grovenor-2685 » Sun May 24, 2020 6:41 pm

jgunby wrote:My problem...
I have no idea how to have 10 free when 1 is reversed and 6 is normal, but have 10 locked normal when 1 is reversed and 6 is also reversed.

For info, 1 is an outer home, 6 are crossover points and 10 is a shunt signal on the far side of the crossover opposing 1.
If there is something there that doesn't make sense ask and I will do my best to clarify.

On the face of it this is simply,
1 locks 6,
10 released by 6.
One simple lock, one simple release.

However, without a signal plan and locking table this is just based on your description. If you post your signal plan and locking table then I can give it the onceover. (Your problem should really have been solved at the locking chart stage before progressing to the dog chart :)
Regards
Keith
Grovenor Sidings

stevemcclary
Posts: 13
Joined: Thu May 03, 2012 3:20 am

Re: Locking question for MKII Frame

Postby stevemcclary » Sun May 24, 2020 11:52 pm

Joe,

From memory, the locking layout looks like Llanfair Caerinion? If so, this is a problem existing on the real frame, which directly locks 1 and 10.

What is needed is some conditional locking. This was achieved on full-sized frames using a pivoted arm on the tappet-blade. See the diagram below.

Image

In your case the locking bar between one and ten would be split into two sections at lever 6. When 6 is reversed, the arm is slid between the two parts of the locking bar (effectively lengthening this bar to a point where it provides direct locking between 1 and 10. When 6 is placed normal, the lever is removed, shortening the locking bar between 1 and 10 allowing both to be reversed.

There is one easier solution. You can provide a double shunt at 10, with a separate lever for each move across 6.
Steve

jgunby

Re: Locking question for MKII Frame

Postby jgunby » Mon May 25, 2020 10:19 am

stevemcclary wrote:Joe,

From memory, the locking layout looks like Llanfair Caerinion? If so, this is a problem existing on the real frame, which directly locks 1 and 10.

What is needed is some conditional locking. This was achieved on full-sized frames using a pivoted arm on the tappet-blade. See the diagram below.

In your case the locking bar between one and ten would be split into two sections at lever 6. When 6 is reversed, the arm is slid between the two parts of the locking bar (effectively lengthening this bar to a point where it provides direct locking between 1 and 10. When 6 is placed normal, the lever is removed, shortening the locking bar between 1 and 10 allowing both to be reversed.

There is one easier solution. You can provide a double shunt at 10, with a separate lever for each move across 6.
Steve


Steve

Thank you for solving my conundrum, I had read about conditional locks but something wasn't quite sticking in my head, all makes more sense now. I was trying to avoid the double shunt as I want to do something interesting with 14 &15 (Any ideas??). As you may have noticed I have re-added the advanced starter and additionally put the crossover locking back to how it was before your simplification efforts, which I imagine will confuse a few crews if I take the final product in!

For the benifit of those wondering what we are talking about (and for Keith) I went for a quick walk into work this morning to take a picture (see below). I would have attached it in my original post but as I have so far done all the locking from "muscle memory" a locking chart and signal plan were not necessary. As Steve says it is an annoyance on the real thing so that was the point where I became very stuck.

IMG_4404.jpg


I hope you are well in upside-down land Steve, I imagine that Philip is very much looking forward to a pint with you when all the madness is over, as am I.


Thanks both,
Joe
You do not have the required permissions to view the files attached to this post.

User avatar
grovenor-2685
Forum Team
Posts: 3918
Joined: Sun Jun 29, 2008 8:02 pm

Re: Locking question for MKII Frame

Postby grovenor-2685 » Tue May 26, 2020 11:43 am

Joe,
Now with Steve's picture everything comes into focus, perhaps?
Your description called 1 an outer home. however 1 on the Llanfair diagram is not an outer home, rather it is a Distant, which makes a considerable difference. So is your locking intended to match the diagram?
Assuming it is, then the discussion about 1 locking 10 suggests you are thinking of all this from the viewpoint of a signalman used to pulling levers and not the viewpoint of a signal engineer designing locking. If the locking is designed in the normal way producing first a locking table then translating that into a gog chart the issue of 1 locking 10 will never arise.
1 would only ever be involved with 10 if 1 can be cleared for trains arriving directly into the loop via 6 reverse. From my very old memory of a visit to Llanfair the approach to the station is up a rather steep hill so I can understand why a clear distant for routes into the loop would be desirable as drivers would apreciate the assurance that they would not have to stop and possibly stall on the hill. Since 4 is a ringed signal implying goods only it is not strictly neccessary to bolt 6 when reverse but given the clear run in as mentioned above bolting 6 for both arrival routes would be safer, so for the purposes of this discussion I will assume 5 bolts 6 in either position.
A locking table is usually started by entering any point to point locking as that can minimise possible conflicts when hand signalling and simplify the locking, here however with only one set of points the only point to point locking is between 5 and 6, so the first entry in the table is 5 locks 6 both ways. Here I am assuming that lever 5 is pulled to bolt the points.(It is sometimes arranged for the bolt to be in with the lever normal).
Then follow up by going through the signal locking.
Signal 1 is a distand and as such can only show green if either 2 or 4 is off, hence we have 1 released by 2 or 4. There is no need of any other locking on 1 as all other requirements are covered by the locking on 2 and 4. In the attached table i have put in both 'released by' and releases columns so the reciprocal is always shown.
Signal 2 has to be released by 5 to prove the bolt in, to lock 6 to prove the correct route and lock 3, 9 and 11. 3 to prevent an illegal aspect, 9 and 11 as directly opposing routes.
Signal 3 as a call on has the same locking as 2 except it does not release 1. So it locks 2, 6, 9 and 11 and is released by 5. A call on may be used by short passenger trains so must include the FPL.
Signal 4 to enter the loop will be released by 5 and 6 then has to lock 10 as that is a directly opposing signal. Note that all the reciprocal locks are entered in the table, ie if 2 locks 6 then 6 locks 2, this is inherent in mechanical locking.
Signal 8, I am assuming here is used only for moves to the loop and can be passed at stop for moves to the workshop since interlocking with hand points is not practical. so it just locks 6 and 10.
Signal 9, has to lock 6 for the route proving and locks 2 and 3 by the reciprocal locking already covered.
This now brings into question the operation of the GF. I assume that the GF will only be released when a train is there so should be locked normal for any moves toward it, hence 9 should lock 13. Whilst considering this it makes sense for the GF to be locked by 2, 3, 4 10 1nd 11 as well to make sure it is always left in the locked condition when not immediately in use. Of course 9 also locks 11 to prevent illegal aspects.
Signal 10 has two routes so has to lock 6 in either position and has to lock 4 and 8 as directly opposing signals. It should lock 13 conditionally when 6 is reverse so as not to block parallel moves.
13 has all the reciprocals of the locking already mentioned so locks 2, 3,4, 9, 11 and 10 when 6 reverse.
I have put this in the attached table, hopefully not to many typos.
Llanfair.xls

To go back to the original question of 1 locking 10 when 6 is reverse. This automatically falls out of the locking as follows. with 6 reverse, 1 is released by 4, 4 is released by 6 and locks 10. So for 1 to be reverse, 4 must also be reverse and hence 10 cannot be pulled. If 6 is normal then 1 is released by 2, 2 locks 6 but 10 can still be pulled as it locks 6 both ways.
The only place that conditional locking is needed is between 10 and 13 given my assumptions on how the GF is worked.
Hope this helps.
You do not have the required permissions to view the files attached to this post.
Regards
Keith
Grovenor Sidings

stevemcclary
Posts: 13
Joined: Thu May 03, 2012 3:20 am

Re: Locking question for MKII Frame

Postby stevemcclary » Wed May 27, 2020 12:38 pm

Keith,

Thanks for the detailed response and your table. As you mention, it is critical the point to point locking is established first, otherwise you end up with significant redundant locking.

It was over 25-years ago when I was a youngish volunteer at Llanfair, and I now live far enough away that I rarely visit (or get to drink a decent pint). However, from memory, there are a few unusual features of this frame, which make locking it a challenge: -

• 1 Signal is a Red/Green home signal, I think the diagram colour has faded. There is a fixed distant, which is missing from the layout.
• A peculiarity of the W&L boxes is that the FPLs are locked when Normal; I believe this was done to match the original lever frames installed on the line.
• Historically, each end of the crossover was provided with a separate FPL, there was also an Advanced Starter (lever 12).
• There are no point detectors, therefore the FPL levers normal are used to prove the points are in the correct position (including the flank)

The ground frame is key released from a local instrument, and electrically locked with the Outer Home, and mechanically with the Advanced Starter.

Looking at Joe’s locking, he is applying it to the historical arrangements mentioned above.

I have updated your spreadsheet to apply the above criteria. This does seem to align with Joe’s dog-chart. Hopefully, no omissions or typos. With purely mechanical locking, one useful check is to verify all the converse conditions exist; for example, If 1 locks 12, then 12 must lock 1.

Llanfair Update.xls


Unfortunately, back to the original question, I cannot offer any suggestion on how the conditional locking could be applied to the Mk-II frame -I’m struggling to work out why a connecting rod is hitting the slidebars on my current project!

Steve
You do not have the required permissions to view the files attached to this post.

jgunby

Re: Locking question for MKII Frame

Postby jgunby » Wed May 27, 2020 12:42 pm

grovenor-2685 wrote:Joe,
... Your description called 1 an outer home. however 1 on the Llanfair diagram is not an outer home, rather it is a Distant, which makes a considerable difference. So is your locking intended to match the diagram? ...

... From my very old memory of a visit to Llanfair the approach to the station is up a rather steep hill so ...

... I will assume 5 bolts 6 in either position... ...It is sometimes arranged for the bolt to be in with the lever normal...

... I assume that the GF will only be released when a train is there so should be locked normal for any moves toward it, hence 9 should lock 13. Whilst considering this it makes sense for the GF to be locked by 2, 3, 4 10 and 11 as well to make sure it is always left in the locked condition when not immediately in use. Of course 9 also locks 11 to prevent illegal aspects...

... Signal 10 has two routes so has to lock 6 in either position and has to lock 4 and 8 as directly opposing signals...

Hope this helps.


Keith

Several things to respond to here so I shall start from the top and work down:

First of all thank you!

Steve's picture cleared up a lot for me, all the other stuff I had been reading was frying my brains for some reason...

The diagram is incorrect in displaying the outer home as a distant, it does operate as an outer home NOT a distant and it is simply a print error; I think it is a carryover from how it used to operate before the colour light signal was installed (as Mcclary installed it in the 90's he can feel free to correct me on that). As such it is not released by 2 or 4. My locking is intended to match the diagram to a certain point but with some "improvements" based on my gripes and inconveriences at work in my day job ie. the 10 conditional lock that doesn't exist in real life, 7 reinstated as the FPL on the loop side of the crossover, 12 as the advanced starter etc...

The approach to Llanfair is slightly uphill to the outer home but then slightly downhill after 13 into the station past 2, 6, 11 etc. so this is not too much of an issue (there is also a PSR of 10mph from 40yds after 1 to add to that).

5 does indeed bolt 6 in either position, it is also arranged to lock in the normal position.

On to the ground frame... for varying (some silly) reasons, the GF key box for the release is on the right in the reflection in the glass above lever 13 in my above photo and not with the GF itself; as a result only 1, 11, 12 are locked by it, allowing shunting moves from the platform and loop (with the key in hand) down to the GF.

The interlocking of 9 and 11 has since been added (if you look closely the locking for 10 against 4 and 8 is already in).

When I make the mentioned adjustments to your locking table it all lines up (from what I can tell) with what I have in my head and on screen in the dog chart which is a nice relief!

Joe

User avatar
grovenor-2685
Forum Team
Posts: 3918
Joined: Sun Jun 29, 2008 8:02 pm

Re: Locking question for MKII Frame

Postby grovenor-2685 » Wed May 27, 2020 1:46 pm

Thanks both of you for the explanations, interesting stuff.
All looks good except one small issue. I have never had to deal with locking on a frame where the FPL stood normally bolted so I have never seen how the inverse of locks both ways would be described in the table. You have chosen to put 5 releases 6 and 6 released by 7 but this implies that with 6 reverse it would not be possible to replace 5 and 7 to put the bolts in. Clearly the 'locks both ways' column does not fit this scenario, you need a 'releases both ways' column which I have never come across, probably just have to cover it with suitable notes.
Incidentally there is one typo in your updated table, you have 6 released by 6.

So far as the conditional locking goes Howard has covered it in his writings on the Mk 2 frame, I think in the instructions for the locking, it can certainly be done. I do understand that some early GWR frames did not have provision for conditional locking, is yours one of those?
All the best.
Regards
Keith
Grovenor Sidings

bécasse
Posts: 377
Joined: Sun Jun 17, 2018 8:26 am

Re: Locking question for MKII Frame

Postby bécasse » Wed May 27, 2020 3:07 pm

grovenor-2685 wrote: I do understand that some early GWR frames did not have provision for conditional locking.


I think, at least as a generalisation, that that applied to ALL early GWR frames as the GWR chose not to use the supplier who held the patent for conditional locking, and that led to the adoption of distinct features in GW signalling layouts that persisted until the early days of BR.

stevemcclary
Posts: 13
Joined: Thu May 03, 2012 3:20 am

Re: Locking question for MKII Frame

Postby stevemcclary » Wed May 27, 2020 11:28 pm

grovenor-2685 wrote:Incidentally there is one typo in your updated table, you have 6 released by 6.


Keith,
Thanks for spotting the typo, 6 should be released by 5 and 7.


The frame at Llanfair is an ex Cambrian Frame manufactured by Dutton & Co. These frames were designed to fit into small spaces, hence the tappet blades are curved. This results in the levers being vertical in their normal position.

The Dutton Frames did have facilities for conditional locking, and I can recall seeing some spare Tappet blades fitted with the additional pivoting lever. No doubt a future project for Joe once he has proven it in minature!

Steve


Return to “Signals and Control Systems”

Who is online

Users browsing this forum: ClaudeBot and 2 guests