Making a Start - The Peak District Midland / Monsal Dale pre 1903

[Allan Goodwillie]

Hi Tim,

This is all looking very good - the Protocab system allows for a more robust way of fitting the point details and seems a definite advantage although only those with the same system will be able to run locos on your railway, but that is true of the difference between DC/DCC systems although it is possible to switch between the two. I am not worried about others running locos on my layout as there will be two systems prevailing - one on the Wemyss/NCB part of the railway and the other on the Buckhaven branch running alongside. It is all looking great and so tidy. I am busy doing the gradients and earthworks and it all looks a bit messy just at the moment. We are both making good progress. I have started laying track on key areas and test it with uncompensated wagons.

The differences between gauges I noticed years ago, so not a new phenomena and I would not use the triangular gauge with C&L components as they end up tight to gauge - that is covered in another thread. I kept to my original B-to-B one - a P4 Soc one which gave the best result. I think I have about 4/5 back to backs in all - all different! There was much arguing in the early years about all this as the P4 Group wanted to have total control over what was released to the public and the argument was that over time various inaccuracies would creep in. Currently I am bothered about samples of flexi track which we have bought recently which are tight to gauge and not enough attention being paid to standards of production. I will not say any more here about that as I do not want to ruin your very interesting thread.

[Le Corbusier]

Hi Tim,

This is all looking very good - the Protocab system allows for a more robust way of fitting the point details and seems a definite advantage although only those with the same system will be able to run locos on your railway, but that is true of the difference between DC/DCC systems although it is possible to switch between the two. I am not worried about others running locos on my layout as there will be two systems prevailing - one on the Wemyss/NCB part of the railway and the other on the Buckhaven branch running alongside. It is all looking great and so tidy. I am busy doing the gradients and earthworks and it all looks a bit messy just at the moment. We are both making good progress. I have started laying track on key areas and test it with uncompensated wagons.

The differences between gauges I noticed years ago, so not a new phenomena and I would not use the triangular gauge with C&L components as they end up tight to gauge - that is covered in another thread. I kept to my original B-to-B one - a P4 Soc one which gave the best result. I think I have about 4/5 back to backs in all - all different! There was much arguing in the early years about all this as the P4 Group wanted to have total control over what was released to the public and the argument was that over time various inaccuracies would creep in. Currently I am bothered about samples of flexi track which we have bought recently which are tight to gauge and not enough attention being paid to standards of production. I will not say any more here about that as I do not want to ruin your very interesting thread.

Alan ... I suspect it is easy to be tidy when you are laying straight track on a 6" wide shelf with a simple crossover

I have taken the decision that if I am going down the Protocab route it would be sensible to make use of its advantages, even if that means it limits what will be able to run on the system. This is the idea underpinning the idea of trying to sort the rodding out mechanically and preferably above board. It may not work ... but fingers crossed. If it does and I can also work out the signalling there is a chance I can have a layout without power and no below baseboard paraphernalia ... which would be pretty different

I have been enjoying your articles in the snooze ... really helpful

Have you had a chance to read Roger Sander's article on Back to Backs? I would be interested in your thoughts.

Tim

[Allan Goodwillie]

Hi Tim,

Yes I can see the direction you are travelling in terms of having little or nothing working below the baseboards - we did discuss this over Scalefourum. I am doing the same - all main electrics and mechanicals are above board on my new layout and it takes some of the experiments I made about 20 years ago to a new level. Like you I think there is much to recommend this especially if like me - getting older and a not so good back, I want lightweight and easy to work on and maintain these days as well as robust for exhibitions. Will seemed to suggest I might take up boating some time soon, but I have been sailing now for the last twenty years and have covered all the main islands of both inner and outer Hebrides - beautiful! However I have accomplished what I set out to do lightweight boards that are rigid and stable, a bit more complicated than the norm and a bit more costly, but they are working well and easy to work with. Just because there have been several attempts that does not mean it is impossible to build lightweight boards - I did teach design and using materials and do have some experience after all. I have been asked to bring a few boards along to Scalefourum North next year and do some demos - which I am more than happy to do. The layout is going to the Glasgow show as a demo layout for modellers outside our own little bubble. I am sure it will be a good source of discussion and perhaps some inspiration.

They are not purely aluminium or all wood, instead each material has been chosen for its own strengths/ weights and the baseboards are treated more as a piece of machinery rather that a piece of woodwork. Everything is bolted together and uses ideas from a variety of sources - sprung floors, etc. So it is far from classic, but it does work and is adaptable as the aluminium sub frame can be used for several layouts and all parts are recoverable. Most parts have more than one function as well- so QI really.

We have been looking at Protocab recently with the East Group and one of our members has some really nice Highland locos working with it - I had thought of it for the new layout, but I think my locos are too small (some of them) and anyway there are a few other experiments I want to try that may need DCC.

I have read the article and think it is long overdue as a subject for discussion, but will not comment further at the moment as I realise it will probably appear as a separate thread and may add some comments on that if it starts - don't want to get too far away from what you are doing here and time spent on that might be better spent on building railways!

[Le Corbusier]

Hi Tim,

we did discuss this over Scalefourum.

I'd forgotten we had discussed all that. sorry for repeating myself

However I have accomplished what I set out to do lightweight boards that are rigid and stable, a bit more complicated than the norm and a bit more costly, but they are working well and easy to work with.

I have been reading your thread with interest as I have yet to even think about my base boards

I have read the article and think it is long overdue as a subject for discussion, but will not comment further at the moment as I realise it will probably appear as a separate thread and may add some comments on that if it starts

I shall await with interest. What I have observed to date with this hobby is that there seem to be many different ways of doing the same thing .. each with merits, each with problems and of course each with powerful advocates .... so I am experimenting and trying to find what works for me.

[Allan Goodwillie]

Hi Tim,

That's a good attitude to have as it means that you will probably not just end up following one trend after another and enjoy the exploration as there is much to explore. It does mean that you will travel up to a few dead ends and some disappointment, but part of the enjoyment I find is thinking for yourself and improving what you do. Although I have put courses together on various subjects I have been putting them together for people just starting off and who want to have some success straight away, but I ask them not to think of it as being totally prescriptive.

There should always be space for your own ideas to thrive and you are already showing signs of trying to do things which I know are on the difficult side to achieve, however that does not mean you should not go there.I was showing photographs of Howard's Minories to the starters group yesterday at our monthly meeting. Most are only starting to make track for the very first time, but I wanted to show them something I find myself quite inspiring, despite the miles of track I have built for my own projects over the years.

If it turns out that something should not work out despite spending some time on it do not be too downhearted, I have been there quite a few times, just remember the things that have worked and have another go - it all comes down to experience. I used to lecture for the Arts Council and one slide I used was of a painting by Manet of a girl playing the piano and the impression of the fingers moving was done with a real flourish and clearly painted within a few short moments, but caught clearly the movement - a second or two's worth of painting that took about forty years of preparation and effort, but wonderful!

[Le Corbusier]

That's a good attitude to have as it means that you will probably not just end up following one trend after another and enjoy the exploration as there is much to explore. It does mean that you will travel up to a few dead ends and some disappointment, but part of the enjoyment I find is thinking for yourself and improving what you do. Although I have put courses together on various subjects I have been putting them together for people just starting off and who want to have some success straight away, but I ask them not to think of it as being totally prescriptive.

Don't want to sound too arrogant here .... more a case of trying the ' prescriptive' advice for each of the various approaches to see what I find easiest - and perhaps a case of mixing and matching. For example I have built a compensated loco chassis which seems to work ok, I now want to have a go at CSB .... but I am not trying to reinvent the wheel just yet

Howard on the other hand .... as you say, inspirational!

[Allan Goodwillie]

That's good Tim,

The starters Group started with a run past of a range of my locos and I asked them to tell me which system the locos had , compensated, sprung and springy beams. A slow run then one at speed for each loco - they really couldn't tell and I think that tells you something. If it works it works!

[Will L]

The starters Group started with a run past of a range of my locos and I asked them to tell me which system the locos had , compensated, sprung and springy beams. A slow run then one at speed for each loco - they really couldn't tell and I think that tells you something

It tells you that they were competently built and the track was reasonably good. The delights of suspension, whichever style you choose, only really begins to show when the going starts getting difficult. In practice it is only the experience of running loco regularly, as you do at exhibitions, that really lets you appreciate the good points of each loco, as well as what its bad habits are.

I do think a well sprung (e.g. CSB) loco has a greater sense of imperturbable momentum, and that is most apparent when the track is less than perfect, or when crossing complex point work. The loco which did this best for me (my LNER 04 an 2-8-0) did also have a bad habit. You had to be careful when putting it on the track as it would sit apparently level and run quite happily with one driven axle derailed, until, that is, it came to a point.

A fully compensated (or equalised - Andy) loco will run very nearly as well, and can be hard to tell from a CSB fitted example. However, compensated locos with one fixed axle will always give themselves away in the end. The reason for choosing CSB rather than full compensation is more to do with constructional simplicity than anything else.

The final point is that a loco with good suspension is much less likely to have intermittent electrical pick up problems, which is why I got into loco suspension in the first place; so as to ensure locos that would start reliably and often in public. In this the owners of continuous run layouts have a big advantage over anybody with a terminus/fiddleyard set up. That said, If Tim's layout develops according to plan, he shouldn't be bothered by that particular problem.

[Le Corbusier]

That said, If Tim's layout develops according to plan, he shouldn't be bothered by that particular problem.

Initially I suspect I shall have my hands full simply building track work that is negotiable and locos/rolling stock which will negotiate it

[Will L]

That said, If Tim's layout develops according to plan, he shouldn't be bothered by that particular problem.

Initially I suspect I shall have my hands full simply building track work that is negotiable and locos/rolling stock which will negotiate it

But with on board battery power right? And hence no electrical continuity problems from pickups.

[Le Corbusier]

That said, If Tim's layout develops according to plan, he shouldn't be bothered by that particular problem.

Initially I suspect I shall have my hands full simply building track work that is negotiable and locos/rolling stock which will negotiate it

But with on board battery power right? And hence no electrical continuity problems from pickups.

Yep ... thats the plan .... so far so good. Current mission is to finish the track and post a video of a csb 1F tank engine built and successfully negotiating it all with wagons in tow .... hopefully before the spring - by which time the mind tends to turn to getting the VW Camper back on the road and the bike beckons

[Allan Goodwillie]

Hi Tim,

I agree with Will when he said

It tells you that they were competently built and the track was reasonably good. The delights of suspension, whichever style you choose, only really begins to show when the going starts getting difficult. In practice it is only the experience of running loco regularly, as you do at exhibitions, that really lets you appreciate the good points of each loco, as well as what its bad habits are.

When I had the museum back in the 80's the trains in the latter years had to start first time, run around the circuit automatically, stay on the track and come to a stop at the allotted loop, setting the points for the next train out to do the same ad infinitum all day every day, 7 hours a day, 7 days a week all season from April until October then on weekends only during the low season. It had to work even when there was someone looking after the museum who was not a modeller and had no experience as such. I was teaching during term time and the lady looking after the layout when I was not there had other duties as well as run the trains.

Good running comes with good construction and methods as Will says, but I had no springy beam locos at all in these days, (sorry, I had forgotten, I tell a lie I had one a BR Standard 4 tender loco - which I built round about 1976-78 with Studiolith wheels and a smaller Portescap gear box. Still runs, but it also had a slightly strange gait which I have never managed to cure, so from my perspective it was an unsuccessful locomotive, despite staying on the track and pulling well. Didn't try another springy beam loco until about 12 years ago and have built one or two since.)

As to the three large Barclays, which I have put on the back shelf until I can get the layout on further, I am building one of each type to compare systems and see what the improvements in running are. I was wondering Will if you had done any such experiment. My friend David Franks has built many springy examples in S4 and they do run well, no doubt about that, but that then brings us back to your first comment.

[Le Corbusier]

Initially I suspect I shall have my hands full simply building track work that is negotiable and locos/rolling stock which will negotiate it

The reason for this ... and I would be interested in any thoughts ...is that I completed the second turnout on my crossover and a little extra straight track to allow me to run my Barney plus wagon and clayton arc carriage through the combination under its own power (hurrah for battery power!)

I found that depending upon the direction and which end of each of the rolling stock was running through I got occasional climbing of the point blades on both turnouts (intermittent) with the stock running straight on instead of across the turn out. I spent some time fettling the blades, checking the gauge and the slight lift on the blades (which I am going to have to solve) to no effect.

I then rechecked the Back to back of all the wheel sets and eased them out to Roger Sanders prescription 17.75. After doing this all the stock seems to run through absolutely fine without a single climbing of the blades during the whole 10 min test? Some of the wheels were tight on my gauge with its masking tape affixed to both sides, but I would think all were comfortably within the P4 tolerances.

Any thoughts on what might be going on here?

[dal-t]

I found the usual reason for that with my EM stock was the wheels being fine wrt btb, but not centred on the axle, so the vehicle was trying to 'crab' on straight track, and one or other axle took off at a crossing. I haven't found it such a problem with P4, perhaps because the wheels are nearer the ends of standard pinpoint axles, so it is more obvious even to tired eyes like mine if they're lopsided. However, my EM stock was either 'rigid' or compensated at best, whereas my (currently quite limited) P4 stock is all fully sprung, so that may also have altered the equation.

[Will L]

As to the three large Barclays, which I have put on the back shelf until I can get the layout on further, I am building one of each type to compare systems and see what the improvements in running are. I was wondering Will if you had done any such experiment. My friend David Franks has built many springy examples in S4 and they do run well, no doubt about that, but that then brings us back to your first comment.

I'm afraid I was always more of a coach builder who did the odd loco. I don't work at the pace you seem to manage either and until recently I've never build any more loco's than I actually needed for use on our layout. Those I have built tended to get run intensively, (or so I thought but not by your standards!!) so I have only ever been able to compare each new effort with what went before. Even when I was building rigid chassis (in 00) I've always got acceptable performance in the end, tenders got pickups, as did carrying wheels on tanks. 0-6-0 tanks were the biggest problem and second goes at the chassis were not unknown. It wasn't until I built my first compensated chassis for one of these that I began to get the sort of performance from an 0-6-0 tank which satisfied the layouts operating team*. However I was still dissatisfied with visible jolts caused by the single fixed axle. My first attempt to eliminate the fixed axle coincided with the discovery of CSB so I gave that a go and finally the ride reached the same standard as the running.

*The operating teams view of good performance, beyond not falling off, and running at an even pace, was reliable starts and stops with clearly visible acceleration from and deceleration to crawling speeds. Non of this bursting into life on full volts in the fiddle yard. Presumably you now get the same sort of demand on Burntisland.

[Allan Goodwillie]

Yes we do Will, but most of the locos are built using the earlier form of compensation or springing. Only one or two are springy beams as far as I am aware. All engines are capable of a crawl and gradual acceleration and deceleration, if not then they are not allowed to run until they have been checked out. They are all digitally controlled which helps, but I would be disappointed in a loco if it did not perform well as per your description.

In the museum there was a system of gradual take off and slow down using resistors in the circuit feeding different lines in the yard, but all engines were able to crawl as I had built them for what was basically a shunting layout with few trains getting up much speed along its 60+ft length from a terminus start. The layout was rebuilt into a circular layout over the first winter as that was more practical for a museum layout. A lot of the engines did many miles as you can imagine and I even wore completely through one set of N/S wheels! Having the feeds in the loops meant that each train could be made to travel at its own rate around the layout on the same setting of the controller. There would be a way to achieve this via dcc these days.

I am all for you using springy beams, I am sure you have gathered that, but much of what I have put on the forum is to encourage the beginners to our Scale and the more we have, the better the Society will develop and it may last longer into the future. We were all beginners once upon a time - for me in P4 that was about 50 years ago! (I still refer to an electric soldering iron as a soldering bolt!)The beginners group I have been working with have chosen all three basic designs to build their first locos. The ones who have taken longest to get their locos built are the ones who chose to build the locos using springy beams, however I would not say that that means very much as other factors may have come into play. Sharing experiences, has been good for discussion.

Burntisland's engines are one of its better aspects, but given the fact that it is a group of, on average, 20 modellers building away (Contrary to the impression given by a certain magazine) it has not always been possible to get the same level of construction throughout, however without that sort of number it could never have been built at all. So it is as it is. I am going back personally to building layouts for myself in my dotage as it were and am trying to encourage others to do the same. There is something nice about going around to a friends house and running trains over an evening. I am building a new exhibition fro myself which I can take around with a few friends, which is lightweight and will keep me out there, I hope, for a few years more. Dubbieside has done well, but in recent years has become more fragile as it has aged and I cannot be sure of exhibition robustness so have decided to retire her from exhibition use and build her more permanently into the garage railway space below Grayrigg, which I will also continue to develop.

[Armchair Modeller]

Slightly OT but there has been a post about the LD&EC's proposed viaduct that would have crossed the MR at Monsal Dale on their Facebook page here

https://www.facebook.com/groups/5382863 ... 588214621/

One example of the illustrations

[Le Corbusier]

Slightly OT but there has been a post about the LD&EC's proposed viaduct that would have crossed the MR at Monsal Dale on their Facebook page here

https://www.facebook.com/groups/5382863 ... 588214621/

One example of the illustrations

Cripes I thin Ruskin might have had something to say about this as well

Any more image? I can't access the facebook thread.

[Armchair Modeller]

These are probably th best ones

A diagram



and a "photo" from the Monsal head end.



The viaduct design is pure speculation, but the route and height are as planned.

[Tony Wilkins]

Initially I suspect I shall have my hands full simply building track work that is negotiable and locos/rolling stock which will negotiate it

The reason for this ... and I would be interested in any thoughts ...is that I completed the second turnout on my crossover and a little extra straight track to allow me to run my Barney plus wagon and clayton arc carriage through the combination under its own power (hurrah for battery power!)

I found that depending upon the direction and which end of each of the rolling stock was running through I got occasional climbing of the point blades on both turnouts (intermittent) with the stock running straight on instead of across the turn out. I spent some time fettling the blades, checking the gauge and the slight lift on the blades (which I am going to have to solve) to no effect.

I then rechecked the Back to back of all the wheel sets and eased them out to Roger Sanders prescription 17.75. After doing this all the stock seems to run through absolutely fine without a single climbing of the blades during the whole 10 min test? Some of the wheels were tight on my gauge with its masking tape affixed to both sides, but I would think all were comfortably within the P4 tolerances.

Any thoughts on what might be going on here?

Hi Tim.
There are several common errors with switch blades.
The set angle must match the planing angle so that the switch and stock rails remain in contact for the full planing length when closed.
If the angle of the set (just prior to the switch blade tip) is wrong, this can result in the track gauge being over or under gauge through the switch.
If the planing (filing) of the switch blade is not an even taper, or too short then this can result in the middle of the switch being under gauge.
The one most people overlook is that the top edge of the switch blade should also be tapered down over the planing length. On the prototype the amount varies between 1/2" and 3/4" at the tip, so approx 10 thou or 0.25 mm. This is to allow for the curvature of the stock rail head. If not done on ours, there is the possibility of the switch blade standing proud and catching the flange.
Filing the inner top edge of the switch tip at an angle of about 1 in 5 also helps to produce a nice fine edge.
None of the above explains though why increasing the B2B dimension appears to improve the running .

I note earlier in the thread, that you have pivoted your switch blades on a soldered rivet. I tried that dodge many years ago but found that the soldered joint failed with monotonous regularity. I know it can be re-soldered, but each time the wood gets a bit more charred and eventually the rivets pulled out one by one, so a rethink may be required.
Regards
Tony.

[Le Corbusier]

Hi Tim.
There are several common errors with switch blades.
The set angle must match the planing angle so that the switch and stock rails remain in contact for the full planing length when closed.
If the angle of the set (just prior to the switch blade tip) is wrong, this can result in the track gauge being over or under gauge through the switch.
If the planing (filing) of the switch blade is not an even taper, or too short then this can result in the middle of the switch being under gauge.
The one most people overlook is that the top edge of the switch blade should also be tapered down over the planing length. On the prototype the amount varies between 1/2" and 3/4" at the tip, so approx 10 thou or 0.25 mm. This is to allow for the curvature of the stock rail head. If not done on ours, there is the possibility of the switch blade standing proud and catching the flange.
Filing the inner top edge of the switch tip at an angle of about 1 in 5 also helps to produce a nice fine edge.
None of the above explains though why increasing the B2B dimension appears to improve the running .

I note earlier in the thread, that you have pivoted your switch blades on a soldered rivet. I tried that dodge many years ago but found that the soldered joint failed with monotonous regularity. I know it can be re-soldered, but each time the wood gets a bit more charred and eventually the rivets pulled out one by one, so a rethink may be required.
Regards
Tony.

Thanks Tony,

It is all very curious ... I ran everything again last night for upwards of 15mins and everything ran through absolutely fine, so the evidence would suggest that the setting of the back to back to 17.75 has solved the issue. I employed a little gauge widening through the turnouts following advice and checking with the gauges I don't appear to be suffering from things going under gauge. The switch and stock rails may deviate outwards slightly over the planing length on the straight switches, but the curved ones appear pretty spot on and these were the ones where the climbing was occurring. Following Howard's advice I filed the leading edges of the switches at roughly a 60 degree angle from the tip back along the planing length along the top edge such that the leading edge of the blade is quite sharp along the top and tucks down below the top of the stock rail at the leading edge emerging along the planing length.

As I said, I am having a little trouble with the blade lifting slightly unless the actuation rod is held down (which the cranks are doing quite nicely) but with the back to back adjusted the lift on the one without the crank installed doesn't appear to worry the rolling stock.

If it continues to run well at 17.75mm that is fine with me ... but I don't understand why this minor adjustment would make so much difference ... if anything I would have thought it would have made the flange more likely to climb the switch? In his article Roger Sanders suggested that a similar adjustment had improved running on Sidmouth .. but I don't know if the issues were the same or completely unrelated. I am very much a beginner at track construction, so in the absence of an expert eye to look over what I have built, I have no idea how good or not it might be.

Slightly concerned about your observations regarding my soldering of the heal of the blade to the rivet. I had assumed that as there was rotation and therefore limited stress being applied it would be a fairly robust solution.

Tim

[grovenor-2685]

I employed a little gauge widening through the turnouts following advice

My recollection is that you also had a less than standard throw on the points, if the wheels passs behind the open switch correctly then you would normally get a derail;ment if the other wheel picks up on the tip of the closed blade. But the combination of tight back to backs and inadequate switch throw can allow the wheel to passong side of the open switch.
The only way to confirm what is happening is to watch it happening, looking closely. Otherwise we are just hypothesising
Regards

[Le Corbusier]

I employed a little gauge widening through the turnouts following advice

My recollection is that you also had a less than standard throw on the points, if the wheels pass behind the open switch correctly then you would normally get a derailment if the other wheel picks up on the tip of the closed blade. But the combination of tight back to backs and inadequate switch throw can allow the wheel to pass along side of the open switch.
The only way to confirm what is happening is to watch it happening, looking closely. Otherwise we are just hypothesising
Regards

Hi Keith,

Following your previous advice I decided to make the second set of switch blades with the correct throw.

The climbing prior to re-setting the back to backs was happening on both. When I watched it it appeared that the flange rode up over the end of the switch blade and fell on the stock rail side whilst the other wheel fell within the throw gap ... hence the derailment. Now with the re-set back to back this is no longer happening and looking closely there appears to be no sign of it ... I could understand this if the B to B was reduced but don't understand why increasing would help as surely it should push the flange tighter to the blade - unless with the gauge widening through the points it had something to do with the B to Bs being too narrow?

[grovenor-2685]

Ah, i took this

I found that depending upon the direction and which end of each of the rolling stock was running through I got occasional climbing of the point blades on both turnouts (intermittent) with the stock running straight on instead of across the turn out.

to mean the vehicle had run through the wrong route without derailing.

The reason that a small BB and or wide track gauge increases the likelyhood of the first wheel picking up on the points is because it allows the vehicle to slew diagonally to a greater extent when the root radius of the flange can no longer hold the flange away from the point tip.
This effect can also be seen if the axles are slightly off parallel. Always worth running the vehicle on a flat surface to see it it follows a straight path or curves a bit. Another cause is wheelsets where the wheels are not quite in line even with parallel axles, that is another item to be checked.

I don't really agree with widening the gauge at the point toe as it is the same as reducing the BB. If the planing and stock rail set are done properly you can keep accurate gauge through the switches without any unwanted narrowing.
Regards

[Julian Roberts]

Tim I have not read all your thread so I may be missing some things. I think the issue of running through facing point switches is the where P4 is most challenging. The issues are on both stock and track. Replies here have mentioned several possible stock issues. I wonder if you have analysed the full range of possibilities. Is compensation or springing used or are they rigid? Do they go better in one direction than another? What is the failure rate? How heavy are they?

Propelling vehicles through is the real test.

I am very dubious about Roger Sanders article, though it is to be welcomed for posing the questions about running reliability, for several reasons. Chiefly because he makes no mention of the check gauge. I recently found that a friend and highly skilled modeller with much successful train and track building experience did not properly understand the issue so it's possible this is not unique to him. Any BB setting more than 17.75 with compliant P4 flanges gives the possibility of derailment at crossings. Martin Wynne has replied to you about this too. See the P4S4 pros and cons thread for exhaustive discussion. My main point being that if 17.75 is the minimum it is impossible in sheer practicality to make it the maximum too. I have complete reliability touch wood but very variable BBs, several less than 17.67. But my maximum speed is 30mph. I can well believe a scale 90mph might call for tighter control but that is not your issue.

What switch length is it? Tony Wilkins reply confirms what I said to you before about switches though I made mine with a gradient just towards the tip the place from where the width should become less than a knife edge 0.2mm.

Here is a switch at the quayside of the Rhine in Mannheim, so this is not a tram line. Being so bright it's quite difficult to see the end of the switch so I've shown it in a close up. Still very difficult to see! The place where the switch rail becomes full height and begins to take the vehicle weight shows some irregularity. Perhaps if the vehicles weighed 50g this is where they would derail!

It's important to control the blade height. Not to go lower, by using slide chairs at least at the tips, and preventing lift by some means as you have mentioned.

I wonder if you might consider dividing your thread into two: researches etc, and practical issues. I'd be more likely to keep up with the latter, while some people may not be interested in these matters.