In many respects this article covers the more nebulous aspects of model railway operation and I don't doubt that some of the views expressed below, as well as the title, will be controversial.  It must also be observed that many of the issues raised here are not specifically scale or gauge related but could be applicable across the whole model railway spectrum. More experienced modellers will doubtless find plenty to criticise in the section upon reliability but I hope that even they can find something thought provoking in the "steam locomotive motion characteristics" section.

The previous articles have studied the infrastructure underpinning the operation of the typical NG continental railway and have investigated how this can be represented upon our models. The physical plant, such as the rolling stock, locos and buildings, the pattern of train services and simulation of traffic flows have all been examined. In some respects though this is the easy part of creating a model railway because no consideration has been made of how these features interact and actually transfer into operating practices. By their nature operating practices are ephemeral and difficult to discern.

The actual running of trains on our layouts comes down to an attitude of mind. It is not my intention to be prescriptive but simply to provoke some thought as to what we are trying to replicate. At the end of the day model railway operation amounts to little more than playing trains, derogatory as some may find that phrase. It just depends whether you wish to play like Real Madrid or the Dog & Duck on Sunday morning!
A caricature
My only experience of other people's operating techniques is obtained from observation at exhibitions which are, after all, meant to show case the hobby. Whilst accepting that there are extenuating circumstances I often return home disillusioned with what I have observed, wondering whether some exhibitors have actually ever seen a real train move. Unfortunately I have witnessed many examples of the Dog & Duck, and not just at small local shows either. A typical scenario runs along the following lines.

The passenger train leaves the fiddle yard but comes to an abrupt stop at the start of the scenic section as the operator realises that the goods train in the station loop is also in the circuit and that it has reversed against the buffers, catapulting half of the train off the line. Having stopped the incoming train on a sixpence he flicks the switch for the loop into the off position and rerails the wagons.

The arriving train is restarted but it ignores the incorrectly sited entry signal at danger telling it to stop. This is no problem, however, because the driver knows where he is going even if the signals indicate differently! Unfortunately the driver does not know quite as much as he thought and the train enters the occupied loop making another abrupt stop before reversing out again at a scale 60 mph prior to the points being reset. Happily there are no detection bars or track circuits here to prevent this!

The train makes another start like a scalded cat, but jerks to a halt halfway along the platform. After much finger prodding it is found that the relevant section switch is off. Never mind! After a flick of the switch, the train restarts yet again and stops with the loco's coupling over the uncoupler. It immediately pulls forward slightly to prove that it has successfully detached from the train because this is what happens on all model railways. The operator knows this because he has seen it happen so many times that it must be right!

It is now time for the goods train to leave. Perhaps this will present a more realistic spectacle. Once the points have been thrown the engine shudders to life at 60mph, rockets out of the station, ignoring all the signals at danger, takes the sharp 90 degree curve at a gallop and makes the fiddle yard without derailment, although any signalman giving the token to the train would now be in hospital with a broken arm and the wagons' axleboxes would be ablaze!

The loco must run round the passenger train next. The loco sprints into the headshunt, reverses and derails. "That's no surprise", says our hapless exhibitor, "It always derails there!" The engine is replaced, stutters halfway along the loop, and stops. The driver presses upon an adjacent rail until it restarts, "Poor fishplate connection!" he informs us.

It proceeds to a point in the station throat where any locking would definitely not be released but the points are reversed anyway and the engine hurtles onto the train to remove the trailing cattle van. It crashes into this at high speed and shoves the van and the coaching stock down the platform, without stopping, until the coupling between the van and the train is over the uncoupler. It immediately reverses, by flicking the direction switch on the controller, and yanks the van into the station throat.

No worries about coupling onto the van or uncoupling from the coaching stock because tension locks always work, as does the uncoupler! Pity about upended cows in the van but I am sure that the farmer will receive compensation!

A couple more quick changes of direction and then the loco enters the goods shed siding and turns onto its side. "Those @£*[email protected] point blades always do that!" we are informed as a platelayers' bothy hiding a point motor is removed and the solenoid is prodded so that the blades are fully across. The point motors obviously are not as reliable as the couplers! The big hand from the sky rerails the loco to ensure that a van is picked up at the goods shed, despite the "No engines beyond this point" sign.

The signalman, who must have arms like Pop-Eye, miraculously changes the points with the shunt move in the station throat yet again, and the loco and van shoot into the platform. They crash into the coaches and ram them against the buffers without breaking stride before another emergency stop . Never mind the passengers!

A short break and it is time to depart. Given the preceding carnage it is little surprise that no-one seems to have noticed that the van is unfitted and should not be marshalled between the train and the carriages.

Exhibiting brings its own problems, into which I have some insight having occasionally displayed layouts at small shows, but the intention here is to consider operation generally. I would not claim that the AFK is a paragon of virtue and accept that exaggeration has coloured this vignette. I suspect, however, that the commonplace nature of many individual components in the parody suggests that they can be observed quite regularly on layouts away from the public eye.

The fundamental concerns about the Dog & Duck's presentation are amateurish operating practices and the resigned acceptance of equipment malfunction. Indifference and laziness underpin many of the problems, aided and abetted sometimes by ignorance of how the real thing worked. It is down to the individual as to whether such a display is personally acceptable or not in a domestic environment. It is interesting to note that the incompetence exhibited by the Dog & Duck, and generally accepted by bystanders, might not be so well received in other spheres of the 'performing arts'.

As a church chorister, singing cathedral style music on a weekly basis with a choir that aims at a professional standards, I can well imagine the reaction to the Dog & Duck's equivalent rendition of the Messiah. What would be said if the conductor could not read music and had never seen the score, the singers were tone deaf and the orchestra did not know how to play their instruments?!
Operator errors
The problems are easily analysed but perhaps more difficult to resolve. It is obvious that wrongly set section switches or points cause many of the unintended stops or involuntary starts on a layout and that these stem from operator indiscipline.

This reflects an attitude of mind. The people in charge of heavy, potentially lethal, pieces of machinery in the real world do adopt such a cavalier approach because of the danger to themselves and to others. There are no serious life threatening consequences arising from misjudgements on a model railway and so the avoidance of mistakes simply relies upon the operator's concentration. The fact that no-one else is watching can encourage laxity in the domestic environment.

The crux of the matter boils down to a question of how these errors can be avoided. One approach is to replicate the safety equipment provided on the real thing, such as locking frames. Other errors could possibly be eliminated by DCC and it would also be (unprototypically) possible to link DC traction circuits to microswitches operated by signals to ensure that they were obeyed. For many modellers, including myself, these would be complex and expensive overkill solutions. The AFK's approach has been to devise a mnemonic to try to infuse some order into proceedings. This is MESS, and the idea is not to make a mess. It stands for the following:
Mechanical: is the road set to prevent derailments and ensure current is available?

Electrical: are the relevant switches on to energise the route? Just as importantly, are other sections off to prevent conflicting movements occurring?

Signals: are any applicable signals that would authorise the move pulled off?

Start: if the criteria above are fulfilled then move the train.
MESS stickers have been placed on control panels in prominent positions, although I suppose familiarity breeds contempt and sometimes the reminders are overlooked. The AFK’s complex routeing of power, necessary for flexibility, make it easy to run a train onto a section not connected to a controller which sometimes requires observing the switch settings at three separate locations. From the modeller's point of view signalling could be considered as an ancillary activity but as the AFK is, or will be, fully signalled it was included in the mnemonic.
A large potentially dangerous piece of equipment. The driver of loco 929 receives the 'right away' from the guard at Ospicio. The real thing was not quite so casual in its operation as many modellers are and the driver would not start until he had received authority. The person giving that authority would most likely have checked and rechecked that the move was safe.
The site of a potential operating mess. The polarity of the diamond crossing at Fenditavalat in the foreground is one of many on the layout determined by the position of one set of point blades. Unfortunately it is possible to set a conflicting move across the diamond by moving a different set of points thereby causing a potential electrical short circuit. In this instance the conflicting blades are detected by a hard to miss warning light when a short would occur (inset).
Electro mechanical problems
Having theoretically eliminated operator error (!) it is apparent that many of the other failures plaguing the Dog & Duck's day out were due to equipment malfunction. Model railways are complex electro-mechanical devices sensitive to slight variations from strict tolerances which unfortunately only manifest themselves when movement is involved. The interaction between misaligned fixed plant and errant mobile equipment requires tenacity on the part of the layout builder if these problems are to be eradicated. Whilst it is understandable that running repairs are difficult at an exhibition one does have to wonder at the preparation that went into some layouts. There is a temptation to also turn a blind eye to shortcomings in a domestic environment, taking the easy route and subconsciously accepting, if not poor, then mediocre, performance. The problems outlined in the caricature could easily have been resolved with a little adjustment which would have considerably improved reliability.

Assuming that there actually is a desire to improve running, one simple method worth consideration is to place nails into tobacco tins (or similar containers) each time that a problem arises. In itself this achieves nothing but it does act as a shorthand indicator of a layout's 'health'; the more nails that there are the more infirm it is! Every time that there is a derailment a nail is placed into the "mechanical" container. The next time that there is an electrical problem another nail goes into the "electrics" tin and likewise for operator error. In an ideal world the session would finish without any nails in the tins but even if this is not a realistic short term objective there is an incentive to reduce the 'score' as time progresses. I used this system in the layout's earlier days although as the layout rapidly expanded it fell into abeyance. Perhaps now that the layout has had time to settle down the practice should be revived!

Model railways contain two critical components of course: the track and the stock. The plain line on the AFK is a well known commercial brand of 7mm NG track but the points are hand built using the copperclad system. (There is a section on the website that covers their construction (although it has received critical comments from some modellers!)) The locos and rolling stock are hand made although modified proprietary mechanisms power the locos (as outlined in the locos section on the website). Scratchbuilding allows greater latitude than buying ready made components but it can also create more problems unless the mechanical aspects are executed accurately. Commercial items are not immune from problems, however. Mixing products from different rolling stock manufacturers and running them over 'universal' track can also cause mismatches resulting in unforeseen reliability issues. An appendix of likely problems can be found at the bottom of this article, although it does not claim to be an exclusive checklist of problems. In many respects, as the title of the article implies, relaibilty comes down to an attitude of mind. Only the individual can define what level of reliability is personally acceptable. I aim for 100% but I am frequently disappointed!
A number of irritating intermittent derailments were traced to this piece of hidden track behind Aspargo depot. The gauge had been slightly tightened resulting in some stock's wheels jumping the inner rail. This is a potential problem caused by bending flexi track to tight radii, particularly at the rail joints, as here. For once my foresight in inserting copperclad sleepers into the formation to allow easy adjustment paid dividends, although the end result is unsightly. Access to this area requires the removal of scenic items so the remedial work was carried out between sessions.

The original three way point in the running line at Lacono was replaced because it could not be made to work reliably despite a couple of rebuilds. It was just possible to squeeze two conventional points into the space available. The loco stands over the first replacement with the other one alongside it.
The points in the loop at Breco de Glissent consistently gave trouble. It was eventually discovered that the radius had been inadvertently tightened below the two foot minimum which explained why some stock persistently derailed. This necessitated the rebuilding of one point and the resiting of two adjacent points.
Since completion of the rebuilding there have been no problems.
Sometimes something as simple as physical limitations cause problems. It was retrospectively discovered that the 2-10-2T derailed upon leaving the siding at Varden because the adjacent retaining wall was too close to the track. The Rule Book was subsequently amended to prevent this loco entering the siding. Although this was quite a common prescription on the prototype the objective on the AFK was originally that all equipment could run anywhere, with the exception of the rack branch.
One practice almost guaranteed to cause problems (but grudgingly accepted) is the testing of new locos on the two local goods trains that cross the KTT. Usually things run fairly smoothly, as preliminary testing has obviously been carried out, but the 0-8-0 trialled in session 3 suffered many issues, as can be seen by accessing the archived session.
Signalling and ancillary activities
One further problem evident in the Dog & Duck's display was the operator's indifference towards the technical aspects of railway operation. Faced with layouts like the Dog & Duck I usually attempt to find a more convincing portrayal of a railway but, even here, polite attempts to discuss the prototypical hardware or operational practices are often, albeit not always, met with bemused incomprehension by the layout's operator. Even where the problems outlined above have been tackled it is quite apparent that many operators are happily and wilfully ignorant of operational niceties to the extent that I have seen right hand running railways, such as the German and Austrian systems, portrayed as left handed running because that is how it is done in Britain!

Unfortunately exhibition layouts have a disproportionate influence upon the hobby. I have commented elsewhere upon the website about the tendency for modellers to copy their design aspects without necessarily considering their applicability to a domestic situation and the same holds true for operational practices. Modellers often replicate whatever they see on an exhibited layout without necessarily considering the methods of railway operation in the real world.

There is a problem in that it is difficult to understand operating techniques without spending some time observing them. The modern railway, for obvious security reasons, prevents access to operational locations such as the signal box or the footplate whereas it was easier to enter these in earlier eras. Preserved railways can give some insight into steam age operation but modern safety regulations have often changed traditional procedures here as well.

These activities are ephemeral and leave little tangible evidence of how or why certain actions were carried out. Take the running of a train between stations A, B and C for example. The signalman at B operates the equipment to take the train from A, pulls the relevant levers, offers the train to C and once the train passes by puts the levers back into the lever frame. There is now little or no evidence to show how the train was run or why certain actions were carried out. From the driver's point of view there is no record to show where he opened the regulator, applied the brakes or coasted.

In contrast, it is much easier to obtain details about the equipment's appearance and it is this that many modellers concentrate upon. Numerous books have detailed the minute physical differences between the various members of locomotive classes but very little information is available about the services that they worked or the techniques of driving them over the routes that they operated, especially if they were freight engines out of the public eye. The upshot is that many modellers seem to copy practices that have become the accepted norm without really considering their applicability to their chosen subject.
Steam locomotive motion characteristics

Locomotive movement
If you choose to disregard the rest of the article perhaps you might like to ponder the issue of replicating movement, even if you only run trains as the mood takes. The subject has attracted considerable attention in model magazines so any thoughts presented here are hardly original. One of the more pertinent observations is that as most modellers drive a car they assume that the techniques of running a train are similar, but, of course, they are not.

My own experiences of travel on and observation of our preserved railways (it is a long time since I saw steam on BR!) continually remind me of the characteristics of a steam engine’s motion. In simple terms, upon starting the regulator is opened and the engine gradually accelerates until it reaches the desired speed whereupon the regulator is closed and the engine coasts under its inertia, gradually losing speed either until the brakes are applied or the regulator is opened once again. If this motion was plotted against a graph, with speed on the y axis and time on the x, it would show a negative skew.

How many of us are simply content to turn the control knob to a preset position and watch our electrically powered creations run at a constant pace, irrespective of limiting factors such as loads, gradients or curves, with little or no consideration as to what we are trying to emulate? Would it not be better to accelerate relatively swiftly to the desired speed, slowly move the control knob back towards zero and then re-open it? I don’t always get this right myself but I try to adopt this technique whenever I can.

The alternative is to buy a controller that does it for you, although personally I find that this reduces my enjoyment, and none seem to be on the market at the moment. Various manufacturers have produced simulator controllers with ‘brakes’, ‘inertia’ and ‘regulator’ over the years and I do not doubt that these characteristics can be programmed into digital decoders. In reality running a steam powered train is much more complicated than this, but maybe the attempts to replicate such motion are what help to distinguish a model railway from a train set.
There are some stretches of the AFK with a long clear run where driving techniques can be considered. The 2-10-2T is leaving the loop at Glissent with the main express of the day which runs non stop between Relforka and Lacocno, a distance of  60 or 70 feet. There is little coasting however as most of the run is on a steep uphill gradient and the loco is under the collar.
Gravity and sharp curves usually restrict AFK speeds as they did on the prototype. This heavy freight is being piloted up the Vulpafaŭkangulo by the 2-8-2T with the 2-10-2T train engine behind it. There is no need to simulate the effects of gravity on the  gradient, roughly estimated to be the equivalent of 1 in 21 when the curvature is factored in to the calculated equivalent gradient.
Other activities
Once one begins to consider imitating prototypical driving procedures, it becomes almost second nature to begin to question other commonly accepted model railway practices. I do not wish to provide an extensive list of model railway malpractice but shunting is one case in point.

Locomotives approached vehicles cautiously rather than at speed, although rough shunts were not unknown. Generally speaking however the objective was to couple onto the wagons rather than to chase them along a siding as they ran away after initial contact! Once the loco had buffered up some-one, maybe the fireman, the shunter, or the guard had to place the coupling link over the hook and check that the handbrakes had been released on the stock. Even in North America the Janney coupler 'gladhand' had to be operated to make sure that the knuckle was open and the brakes had to be 'pumped up' before the "cars" could be moved. This took time, necessitating a pause in proceedings.

Exhibitors at a shows often use automatic couplers for conveneience but they feel under pressure to ignore these hiatuses because, to the uninitiated, it appears as though 'nothing is happening'. One unintended beneficial side effect of the AFK's use of a home made centre buffer and loop coupling is that a similar procedure has to take place. The brake hoses, if present, might not be coupled up if this was a shunt move, providing that the stock did not contain passengers, but the cut would then be entirely reliant upon the engine brake to stop it. The moves were consequently carried out at low speeds. Loose shunting, where the loco accelerated rapidly before slamming on the brakes to send a raft of wagons forward by their own inertia, was perhaps the only exception to these procedures, and it is difficult to replicate in model form. Once the vehicles were at a stand the hand brakes would be applied. If it was necessary to uncouple the loco at the end of a journey the brake pipes and coupling would be disconnected but the loco would not be moved simply to prove to the fireman that he had done his job!

Automatic model couplers also offer other integral incentives to ignore the real procedures, providing that they work efficiently. The modeller can direct proceedings from a central control panel with all the point switches, uncoupling controls and the loco controls conveniently grouped together within arm's reach. He knows what he intends to do therefore there is no need for delays.

The real activity was carried out by a number of people, rather than an individual, spread across a large area, as was the equipment that they had to operate. Communication between these widely dispersed operatives was difficult in the pre-mobile phone era and often relied upon line of sight and hand signals. One does not have to read too many historical accounts to realise that shunting was an extremely dangerous activity resulting in many railwaymen being killed or sustaining severe injuries. As a result it was time consuming as it was carried out with extreme caution.

Where shunting was carried out on SG running lines the signalman had to operate the points by rodding from the box which meant that stock had to be clear of them before they could be moved. Ancillary safety equipment such as facing point locks, detection bars or track circuits would be present on the SG and the move would be authorised by a fixed signal  These precautions were more likely to be absent upon a NG railway but the points still had to be changed which required that some-one had to walk to them which again caused a pause in operations.

Once the train was ready to leave, assuming that the line had an automatic brake, the hoses had to be coupled together and a couple of minutes would elapse before a vacuum was made or the air pressure was pumped up before it departed. Do you pause to replicate this on your layout? Have you even considered which system your railway uses? (Just for the record the AFK uses the Westinghouse air brake and the brake test is simulated by gently pulling the last vehicle to make sure that all the homemade loops on the wagons are coupled up.) Once it was confirmed that a tail lamp had been added to the last vehicle the train was ready for departure.

The shunting at Lacono is controlled from this hut at the entry to the yard. The points are handthrown and coupling is manual, rather than automatic. It is easy to forget this on a model where all the controls are grouped together for convenience's sake.
Sharp curves, steep gradients and long trains called for care if shunting was to be  carried out safely. The driver of the local goods waits for a signal to indicate that he can push the wagons into the siding at Breĉo de Glissent. the centre buffer coupler with its hoop and pin is visible on the loco. These are made in a male and female configuration, for ease of coupling, with the loop on the upside of the stock.
Signalling                                           (The AFK's system can be acessed by pressing the button)
The AFK's signalling system
Signalling is another weak to non-existent area on many model railways, NG or SG. The arcane complexities of mainline installations were boiled down to their rudiments on the NG. It isn’t difficult to understand a simplified system and many modelling books and magazines contain chapters outlining the basics.

A lot of models omit signals altogether because, as was observed in the caricature, the person running the train knows where it is going and the route that it will take. Even where they are provided and can be made to display the correct aspect modellers still tend to ignore them. On the real thing, of course, running the train was split between the driver and the signalman hence the need for the signalling system, which was made mandatory in Britain by the 1889 Regulation of Railways Act.

NG railways were exempted from the more onerous requirements but systems with more than one engine in steam that carried passengers were obliged to provide at least simplified signalling, particularly at crossing points. This might consist of no more than a fixed distant and a home signal, with the permission to leave the station being conferred by possession of the tablet or staff, a green flag from the guard and a signal placed at the entrance to the next block section. This is usually known as the ‘starter’ by most modellers but is always referred to in the Rule Book as the ‘section signal’.

As usual, things were simpler on the continent. On the more basic systems opposing trains were allowed to pass at stations where the only signals were fixed boards and communication between the trains was carried out by the loco whistle. On more sophisticated systems there was usually a home signal but the authority to start was often given by the station master’s baton.

One accepted model railway practice that should be questioned in a NG context is that every station should have a signal box which is a separate building, usually of distinctive design, at some remove from the station buildings themselves. The particular idiosyncrasies of the design and the colourful levers of course ensure that this becomes a focal point of the model.
The AFK uses rather more signalling than the typical NG railway, which is "justified" by its regional railway status. The equipment is Germanic but typically amounts to an open frame in full view on the platform. This example stands on the platform at Relforka and contains a couple of technical errors that await attention. The German single line block instrument is at the right of the frame, beside the block bell.

In reality the majority of NG railways avoided using signal boxes as much as possible and where they were provided they were often little more than glorified garden sheds. The subject was covered in much more detail in Narrow Lines than there is space for here but perhaps a few observations might be permitted.

The secondary systems in Ireland were provided with boxes and block instruments because sensible operation would not otherwise have been possible. These boxes were reduced to the practicable minimum, because they were expensive to build and operate, and they were therefore restricted to the crossing points. There were only seven intermediate boxes upon the whole length of the Londonderry and Lough Swilly's mainline and the County Donegal's parsimony was typified at Lough Eske where there was an open groundframe, almost in a hedge, despite this being a block post.

Many smaller British NG lines had very few or even no signals or signal boxes, with the points into intermediate sidings being released by the Annet's key on the train staff. In this context it should also be remembered that specialised signalman were rarely provided because NG staff were expected to 'multi-task'. They might accept a train on the block instrument, placed in the station building for convenience, before selling tickets to passengers and then going out to pull the relevant levers for the train to enter the station. They would collect the token, then act as porter and ticket collector before working the block instruments again and returning to the platform to work the signals, having given the new token to the driver along the way. Once the train left they would replace the levers, clear the platform of parcels etc. and return to administrative work such as telephoning consignees to inform them that their loads had arrived, filling out ledgers or even cycling around the district to collect payments due.

This fuzzy view of Boursson's frame has crept into a photo of another subject. The block instrument stands to the left, the two point levers for the loop points are next to it. The keys releasing the points stand in the cabinet and on the right the signal winding handles can be seen.
Their continental counterparts were expected to be similarly flexible in their attitude but less hardware was provided. Trains were offered to the next station by the telephone, although this was not quite as casual a system as it might appear. The person in charge was the only person authorised to make the call and a series of set phrases had to be used with the times being recorded in a train register book. This method is still authorised on Britain's SG lines, even today, in the case of equipment failure between adjacent boxes. Whereas the section signal was in later years electrically interlocked with the block instruments in Britain there was no such provision on the European NG. The station master gave authority for the train to start by showing a baton to the guard, as the person in charge of the train, and the latter informed the driver that he had permission to start. There was usually no fixed signal to indicate that the line was clear beyond the station.

In the majority of cases there was also no signal to indicate to the driver that it was safe to enter the station on the continental minor lines. Aside from the fact that the driver knew the line very well there were usually fixed boards at the approach to stations indicating that the train should slow down. In many cases these amounted to "proceed on sight" authorisations with the driver being expected to approach slowly with the train under control and being able to stop should there be a problem. In some cases a stop board was provided where the train had to halt and await further instructions. On the German lines this took the form of a trapezium and the train's  timetable indicated when the board could be ignored. This halt was usually to allow an opposing train to enter the loop and once this had been done the stopped train's authority to proceed was given by a green flag or light shown from the platform. A sophisticated variation on this theme, used where there were sighting problems, was to include a green light over the trapezium! These procedures were usually time consuming and relied upon the presence of station staff, although in some cases the second train's approach was authorised by the horn or whistle of the train booked to arrive first. Where traffic was more intense it was common to find an entry signal on the German NG, operated from a rudimentary lever frame on the platform.
Continental signals appear to be weird to British eyes but they are a personal interest so the AFK is lavishly equipped, even at minor locations such as here at Ithilarak. The upper part of the line is notable for its pre-WWI examples drawn from an unfeasibly wide range of companies and countries. This is a PO "banjo" signal, the red panes of which were made from silk which folded together to indicate the white "clear" indication. The signal confers the right to enter the Ithilarak to Cadsuiane-Fanhuidol section.
Various systems were used to control the points. At the simplest level, where speeds were low and trains infrequent, weighted point levers were used. When shunting took place these were usually operated by the train crew, frequently jumping from moving vehicles and walking in front of the approaching train in a manner that would be unacceptable today! This method of working was the accepted norm in most countries but the Germans developed a system of key working (schlußelwerke) at busier locations. In simple terms the keys were removed from a cabinet which operated a bar which was interlocked with the approach signals to prevent them from clearing. Each key unlocked an individual point lever from which it was impossible to retrieve it until the point was returned to the normal position. The key had to be replaced in the cabinet before the signals could be cleared. In effect this created what signallers know as 'sequential locking'. This allowed higher authorised speeds  when signals were cleared and was often augmented with the loop's entry points being controlled from the small frame.
The shunter has just unlocked the point in the RT (mainline) at Boursson with a key removed from the central cabinet. This has allowed the point lever behind him to be thrown so that the blades are set into the Co-op siding into which the 0-10-0T is propelling a SG bogie grain hopper. It will be impossible to release the entry signals into the station until this key is replaced into the locking equipment. At the moment the key is trapped into the point which is reversed.

It is relatively easy task to build a functioning layout, especially if time and care are exercised with the hardware, but, as I have hopefully shown, the adoption of a prototypical mindset and plausible operating philosophy possibly requires a little more thought. It is difficult to generalise without referring to specific examples and it is too easy for articles of this nature to degenerate into nebulous treatises, as this probably has. I am sure that some modellers will have dismissed it as being too theoretical for a relaxing hobby and will have failed to read this far. For those that have succeeded in reaching this point I hope that you have found something of interest.
Continue to Section 7
Return to operational articles index

This is a selection of the physical problems that have been observed during the construction and operation of the AFK. I would not claim that it covers every eventuality but it might be useful as a checklist.
Electrical problems
current not turned on at mains

metal tools shorting out circuits

metal wagon wheels shorting out across separate circuits (particularly where the adjacent rail polarities are different due to the point blade positions)

dry joints in soldering

circuit not connected to controller in a 'cab control' system
Mechanical problems

Simple track
track wide to gauge

track narrow to gauge

rail joints horizontally misaligned

rail joints vertically misaligned

sudden kinks in direction

sudden dips in vertical alignment (exacerbated on curves)

track horizontally misaligned (which can be cant where not required or insufficient cant where it is required)

track electrically dead, often very short pieces accidentally overlooked, and often bridged by long wheelbase locos
all of the above!

horizontal misalignment of the vee at the crossing so that the wheels catch it or even go down the wrong route

checkrails incorrectly set so that they do not pull the flange to the correct side of the vee or they are too tight and cause binding

vertical misalignment of the vee where one rail is slightly higher than the other causing the wheel to be pushed away from the higher rail

vertical misalignment between the vee and the closure rails so that the wheel jumps/drops on transferring between the two components

horizontal misalignment of the closure rails so that a wheel passing over them drops into a gap causing lurching or, where the gap is too tight causing binding with the inside of the knuckle

horizontal misalignment between the blade and the stock rail which guarantees derailment! The transition must be smooth as the wheel leaves the fixed rail and enters the moving blade

vertical misalignment between the point blade and the adjacent stock rail which results in the wheel either dropping or being pushed to one side depending upon the direction of travel

inactivation of the microswitch by the point rod resulting in the wrong polarity at a crossing

in more complex formations insufficient clearance between track components which results in an intermittent short circuit as a wheel passes through
Locomotives and stock
wheel back to back gauge too wide

wheel back to back gauge too narrow

wheel treads too wide causing shorting at points as they bridge rails of different polarity

gear train catching track components (a particular problem with older Fleischmann mechanisms where the gears were cast into the back of the wheels)

failure due to unreliable or insufficient electrical pick ups

jerky operation due to tight spots in the mechanism (either in the mechanism itself or because the wheels sets of a steam loco are not correctly quartered)

derailment of pony wheels because they are not correctly centred or there is insufficient downward pressure to hold the track

too long an overhang on curves resulting in the couplings pulling the loco or stock off the rail

axles on wagons not quite being aligned resulting in the 'picking' of points
A download of this list is available by pressing the Troubleshooting button.
Continue to Section 7