4   Timetable considerations
All but the most basic railways ran to timetables. Industrial  'conveyor belt' systems which, by definition, did not carry members of the public, were an exception because they obviously had no need of them. Where a passenger railway operated simply with one engine in steam the timetables could be viewed as advisory and printed primarily for the public's convenience, although how assiduously the individual railway adhered to the published version varied considerably. Once two trains (not necessarily both passenger carrying) were operating upon a railway open to the public they became a necessity. It was obviously in the railway's interest to publish the running times of its passenger trains for commercial purposes but the working timetable was much more important from an operational point of view because it helped to prevent conflicts between trains. This was particularly important on a single line where trains could travel in opposite directions and possibly collide head on with one another. The working timetable itself was simply an advisory document for the operators and indicated which trains should take priority over others. The potentially disastrous consequences of misjudged timings were controlled by some method of regulation, but the exact methods varied between different railways and were dependent upon traffic intensity.

It follows from these observations that any-one attempting to operate a model railway, other than an industrial system, in a realistic manner must use a timetable of some sort. The objective of this section is simply to consider how the AFK's basic timetable document was produced. Section X will consider in more detail how this is works in model form.
Model railway operating: background conditions
Notwithstanding the previous comment, before considering the replication of timetabled running some thought needs to be applied to the manner in which the model railway is likely to be operated. The typical British model railway is a relatively small affair (unless it is aimed at the exhibition circuit) usually intended for operation by one person. This means that many railway operations which would have happened simultaneously on the real thing take place sequentially on the model. American model railways appear to have diverged from this template and the layouts commonly featured in the mainstream modelling magazines present a picture whereby long main line journeys can be undertaken, allowing the use of a 'fast clock'. I suspect that these layouts are atypical and that the US press is skewed towards these gargantuan models in the same way that their British counterparts extol the virtues of large exhibition layouts! Be that as it may, these giant layouts require much more structured operating sessions than the one man band because they rely upon the availability of a large operating team, each of whom undertakes specialised tasks within the session. In these circumstances the physical characteristics of the model actually determine the running times rather than theoretical considerations and many activities are carried out simultaneously as on the prototype. The AFK is run by an aforesaid one man band, although it is larger than the usual individual's layout. As a result the railway is nominally operated in the traditional sequential manner where attention is given to one train at one time. It was therefore necessary to consider notional speeds and running times to devise a framework for the timetable rather than times derived from actual running on the model.
A paper railway
Mention has previously been made of a 'paper railway' and effectively this is what the timetable is. In the example of a terminus to fiddleyard model the rest of the world is represented on paper which is the closest that the railway gets to modelling the stretches for which there is not enough space. The majority of travellers using the railway will be familiar with the concept of a timetable. It is, as its name suggests, a table listing stations vertically (usually, but not always, at one side) and the times at which each train will stop at the named station. As I am sure that most modellers will be aware, these timings are derived from graphs which the planners use to devise the published versions. On the AFK all train running is controlled by reference to the graphs as these give a much clearer visual indication of events than a printed table.

At the most simple level the train graph comprises a line graph with time on the horizontal axis and distance on the vertical axis. To be strictly correct the distances between the stations should be shown to scale and the time should also be shown at a constant scale. If this is done then triangles of card can be used to accurately plot trains running at a constant set speed by drawing a line along the hypotenuse of the triangle. Trains obviously do not run at a set speed so these timings represent the average start to stop speed over a given stretch of line, where the maximum speed is obviously higher than that plotted. It might be desirable to present the finished train graph in this manner but it is not essential and in the earliest stages of planning is probably counter productive. Being realistic, unless the timetable will be very simple there will be a lot of trial versions, with multiple erasures and revisions, and it easiest to do this in a sketched out form. It is, however, necessary to known the point to point timings before beginning this exercise.
Considerations affecting train timings
Before any trains can be placed onto the graph the speeds at which they run needs to be known. This will vary between the classes of train, with goods trains generally being the slowest. The speed will depend upon a number of interacting factors of which some of the more important might be listed as:

the power of the loco
the weight of the train
the gradient of the line

These are the most obvious factors, however, and depending upon how accurate (or obtuse, perhaps, in the case of a freelanced railway) one wishes to be, other variables might be considered such as:

the state of the roadbed
the braking power of the train on descending gradients
the curvature of the line
the number of stops a passenger train will make between timing points
train make up, such as whether SG wagons are in the consist
other limiting factors such as speed restrictions for bridges or level crossings
Whether the roadbed of the AFK would be able to sustain the speeds at which trains run, especially given that SG wagons are included in the consists is open to debate. This is a notorious wet spot in the Kasatritikakamparoj.
It was noted in section 1 that the small local NG lines averaged about 10 to 15 kilometres per hour between stops during the latter part of the nineteenth century but as more powerful locos became available speeds could be well in excess of this. Speed also reflected the choice of gauge. The wider the gauge the more stability it offered, the greater the size of the equipment that could use it and the faster that it could travel. This was, of course, Brunel's justification for the Great Western's 7' 0¼" gauge. The converse in the NG world was that the narrower the line the slower the maximum speed that could be obtained. The 10 to 15 kph quoted was based upon timetables taken from metre gauge lines built for local use in France but which used light rail and had a minimally graded road bed. The latest NG articulated railcars such as the Soulė AMG 800 series on the metre gauge Chemins de fer de Corse (Corsican Railways) are capable of 100 kilometres per hour on the flat and 83kph on a 1in 33 gradient, whilst the modern diesel locos of the 760cm Zillertalbahn are capable of 80kph. These are top speeds of course but Fayle noted that the County Donegal railcars often achieved 35mph and that 40mph was not unknown. The 891mm Tp3500 of Swedish NG railways (of which the AFK has a bastardised model) was also quoted as 80kph maximum.
Although this is a rather tongue in cheek attempt to copy a Tp3500 the prototypes were capable of a 50mph maximum speed. The large bulk of the loco is evident as it towers above the coaches at Relforka. A loco like this implies a good solid road bed and gentle curves.
 The AFK uses OO track which roughly scales out at 2 foot 3 inches or approximately 750mm in 7mm scale. Taking into account the considerations outlined, and especially the track bed, it was decided, somewhat arbitrarily it must be said, to set the following point  to point speeds for the AFK:

20 kph for goods trains
30 kph for passenger trains
35 kph for railcars

Once these speeds were known then the basic point to point running times could begin to be considered. These will, of course, depend upon the distance between stations, ceteris paribus. The previous paper systems had arbitrarily allocated these distances but had not considered timings in any detail because there was no need to. One factor contributing to this laxity was the nature of the railway. Whilst it remained a line of local interest it would be reasonable to assume that there were few trains and plenty of excess infrastructural capacity to accommodate operational problems such as late running.

Once the prospect of an enlarged system became a possibility the nature of the railway changed and these distances and timings gained more prominence. The integrated development of the track plan and the railway's conceived environment had a significant impact upon these distances, timings and the final operating scheme of the AFK. As was outlined in section 3, six stations with a loop were envisaged thereby giving eight sections for the quasi-mainline system. Assuming that these were placed at 10km intervals this would give a distance of 80kms, or 50 miles, which seemed a reasonable distance for such a line. The distances between stations were tentatively allocated for each section, bearing in mind the background geographical template, and the altitudes of each station were simultaneously determined whilst also calculating the ruling gradient that these would create. A series of timing conversion factors, once again arbitrarily decided, was then applied to the intermediate running times between stations to take account of the gradients. Further modifications were made to reflect the number of halts in the section for passenger trains.
The finalised route diagram of the AFK. Sojonno, Varden and Cadsuiane-Fanhuidol have sidings but no passing loop.
The finalised gradient and altitude diagram of the AFK. As in many things the AFK is probably pushing at the boundaries of reality given the intensity of service operated over such a demanding line with so little infrastructure!
The results of these calculations were basically tested to determine the nature of the timings that would ensue. For a freelanced layout this means examining how long a train will take to travel between two stations with regard to the intensity of the service which it is wished to offer. As an example if a train leaves A and it takes 90 minutes to run between station A and station B it will take three hours before a train travelling between B and A can logically appear at A. This would offer a very restricted service and would suggest shortening the distance between stations or speeding up the trains before beginning to write the timetable! On the prototype, where the distance between communities could not be changed, the infrastructure could be improved by inserting another crossing loop. It is of course easy to recast the timetable when the off stage section is part of the imagination and this was done more than once when the layout was a terminus. It became crucial on the later full blown version to get these calculations correct because all the main stations would be modelled. Once these distances and timings have been set they determine the intensity of the service and the nature of the railway. Ignoring these parameters effectively amounts to 'cheating'!  
The AFK basic timetable.

External influences
The AFK's timetable is complex and there is no intention to examine it in great detail. In the standard manner the passenger trains, being the most important services, were put onto the timetable first. These were designed to connect with the FT (Thalnain Railways) services at Relforka and so the times of these SG services had to be known. In the real world these timings were determined by the SG and the NG was obliged to fit around their needs, however inconvenient that might be. The FT service was plotted before any NG train was placed onto the timetable graph and the AFK trains were timed accordingly. It was decided that the FT line into Narnia was a secondary branch that offered four trains per day each way. Only two of these four pairs cross at Relforka which presents further difficulties for the AFK which has to provide ongoing transits with the earliest train whilst ensuring that passengers arriving on the later train can still access the cantonal interior.

Time had to be factored into these calculations for transfer between the standard gauge and the narrow gauge trains. This was set at ten minutes for the SG to NG transfer but this takes into account that the AFK will react to late running by the FT and will endeavour to start its trains as quickly as possible after the late arrival of a SG train. Quite how much leeway was tolerated between a SG system and a late running NG train is open to question. AFK trains are generally timetabled to arrive 20 minutes before the ongoing SG train departs to allow for late running on the NG. SG trains can be held for some time, especially allowing for the lackadaisical ramshackle nature of the FT line into Narnia, but will depart if the AFK trains are too late. A little more latitude is allowed for the last train of the day. The train forming the down NG train is on a tight turn around and uses the loco and stock of a preceding up train which could potentially cause problems if it was running late. The AFK attempts to safeguard against this possibility of late running by having a spare loco and coach set at Fenditavalat. Whenever the up service is running significantly late the spare set departs on time and takes precedence over the late running train which is detained at whichever loop will form the most convenient crossing point. For convenience's sake the up direction on the AFK corresponds to the physical direction rather than the normal practice of trains running towards the capital. FT down trains connect with AFK up trains.
The pattern of passenger services. The Eromarbordo service is supplemented by a branch shuttle and the Jakarutu service is not shown.
As a result of the SG pattern there are four pairs of NG trains between Relforka and Urteno but only three beyond this point. Any connections into the Altingablecaŭtoj/ Magorsoromban (High Reaches) with the earliest and latest SG trains would run at unsociable hours. One of these three is an express which runs non stop between Relforka and Lacono before running all stations across the Altingablecaŭtoj. It is trailed by an all stations sweeper across the KTT and this pattern is reversed later in the day. This arrangement failed to provide any service into Lacono from the Kasatritikakamparoj before mid morning so an additional service, running through from Eromarbordo on the branch provides an early morning service, the balancing working returning late in the evening when there is no ongoing SG service from Relforka. The availability of this railcar provided the opportunity to run a shuttle service between Urteno and Lacono during the middle of the day. This pattern constituted the basic mainline passenger provision for some time until it became apparent that a number of the smaller railcars were idle. To get better use of this equipment schools' trains were introduced over the two sections of the line. I suppose that this is a modeller's reaction to excess stock. In reality these anomalies would have been scrapped and everyone would be thankful that they did not have to ride in such antediluvian veterans!
An anomalistic rod driven antediluvian veteran. One of the smaller railcars for which the school trains were inserted into the timetable.

The service frequency diagram indicates that the upper tier of the layout closely resembles the three trains per day model (especially before the schools' train was added). This was a deliberate decision, taken to emphasise the remoteness of the line across the mountains. The intensity of service between Lacono and Urteno adequately compensates for the lack of a city suburban service that could potentially have been provided for by a tri-level design option and a large city with a suburban service (see section 2). The two branch services connect with the mainline services although there is an imbalance in the Eromarbordo trains as one train splits at Glissent. The Jakarutu line has two scheduled mixed trains each way per day

Once the passenger timings had been determined the goods trains could be added to the graph. The main consideration was that goods train should not delay passenger trains. The goods train timings are regarded as advisory and trains often run later or earlier than scheduled depending upon circumstances.. Regardless of whether the goods trains run on time or not it is important that they are plotted onto the train graph because they indicate potential line usage. These practical aspects of actually running the timetable are examined in section X.

This basic timetable was drawn up alongside the track planning process to determine whether there was sufficient provision for the intended service. Whilst is should be possible to ascertain whether this timetable is viable from the paper version it was felt that a physical rendition of this would improve the design process. A 'cardboard railway' was therefore created by drawing simplified station layouts showing the running lines but not the sidings, onto a piece of corrugated cardboard. Trains were represented by coloured map pins which were shuffled around the diagram to represent the movements demanded by the basic timetable. This might seem to be a little excessive but it did prove that the proposed timetable could be run upon the layout and that construction could proceed without any need for drastic reconsideration or redesign.
Additional documentation
Once the basic timetable has been constructed some consideration must be given to the equipment required to operate the services and how this will be used.  On a small three trains per day line this might only amount to a couple of engines in steam and could be easily worked out. It was not necessary for the same stock to return to the same starting position for the following day, which is a trap that some modellers seem to fall into. Taking the three trains example the loco would work the following pattern.


It can be seen that on day 1 the loco would enter service from A depot whereas on day 2 it would enter service from B depot. This requires a complementary diagram to maintain the service with another loco entering service from depot B on day 1 and working B-A/A-B etc .In reality it is very likely that both locos would be of the same class but given the diversity that modellers prefer this is less likely to be the case on a model. Some earlier paper versions of the AFK had diagrams that were three day arrangements (i.e. locos rotating around depots A,B and C)  and the only provision made was the locos were roughly similar in size and power output. The current version of the AFK does not require such complex diagrams as all the requirements happened to be balanced and locos return to their originating point at the end of the timetable. This is the theory. The practice is that certain trains often run so late that they held over into the following day. The train then passes a locomotive working the same diagram on the following day. One other trap that I attempt to avoid is the consistent rostering of the same loco to the same train for every timetable. This did not happen on the prototype because locos needed temporarily withdrawing from service for routine maintenance. Locos are switched between duties regularly provided that they fall within the power category allocation. Certain locos in these power categories are also subdivided into passenger or freight roles given their suitability for such services although the majority are effectively mixed traffic types. Depending upon anticipated traffic demands the power allocation for each train is also adjusted so that an appropriate loco is placed onto the head of the train.
During the Christmas rush the local passenger was powered by a class I rather than a class II loco to cope with the lengthened train and heavy parcels traffic. The railway bus waits to depart for Eromarbordo.
The I-B-I loco was technically classed as a railcar and is restricted to passenger service. It has sufficient power to handle local freight trains but its ungainly nature mitigates against its choice in this role. It is in power category IIP.
The crocodile is a slow runner and so it is technically classed as a type IF. This is the highest power category but restricted to freight services.
A similar process has to be carried out for carriage sets to ensure that equipment is at the correct starting point to maintain the service. Whilst the AFK's express set consists of half a dozen designated carriages the other sets tend to be made up on an ad hoc basis between sessions. There might not be too much to consider upon a local railway but the AFK needs two other sets on a daily basis and a third during certain seasons. The railcars also add and detach carriages along the route. The AFK has around thirty vehicles to cover these demands. There are also five Postal vehicles that run on set patterns every day with two diagrams being back to back in that one sorting office van runs between Relforka and Krelm (CFS) and its counterpart travels in the opposite direction. The AFK also operates an ambulance which transfers patients to and from the Urteno hospital from outlying cottage hospitals, reflecting the difficulties facing road transport within the province.
Coach sets are made up to meet demand. There was a struggle to find appropriate stock for the mixto during the Christmas session and so the Director's saloon, the last vehicle here, was pressed into service for first class passengers.
Other transport modes
As was observed in section 1, the regional railways were sanctioned by the provincial authorities (or their equivalent) across Europe. When it became apparent that passenger services could be better provided by road than rail the administrative response was to control such services by licensing (unlike Britain). The obvious bodies to receive such authority were the regional railways, who already had the infrastructure in place to sustain such services. In many countries these permits were supplemented by the Postal authorities which operated buses over routes complementing the railway bus provision. The buses ran between the stations connecting with whatever remained of the train service and making good use of expensively provided facilities. The gaps in the AFK's railway services are 'covered' by these bus services, although the models are obviously static. The roll on roll off ferry across the Spegulalaguno also falls under the remit of the AFK, particularly as it is also fitted to transport railway wagons. In a manner similar to the extension of a small layout by documentation these services have been included within the AFK service timetable to suggest that the AFK is integrated into a province wide transport network in the grander scheme of things.
The AFK operates a number of bus routes under license from the cantonal authorities. These services are represented by static models as shown here by number 61 operating the 311 service to Bolzano. The model was bought on eBay and has been repainted but it is difficult to insert passengers without destroying it.
The original Post bus was rather expensive but as few 1:43 continental models were available in pre-eBay days I gritted my teeth and paid out. It was repainted and prised apart to insert homemade passengers. It has now seen around 40 years' service for the Post Office.
Most of this section about timetables has covered the regular passenger services rather than concentrating upon freight trains. The majority of prototypical passenger services are mandatory and run irrespective of demand upon a given day, although passenger loadings are monitored over a longer period to ensure that loadings are at economically viable levels. The running of goods trains tends to be much more flexible than passenger trains which reflects the fluctuating levels of traffic that the railway is offered upon a daily basis. The next section considers this in more detail.

The AFK public timetable is available here: 
Public Timetable
Continue to Section 5
Return to operational articles index