Complex trackwork on the quayside at Eromarbordo.
Narrow gauge railways often had no complex pointwork and the standard gauge avoided it whenever it could. The reasons for this are obvious;~ it was expensive to instal and difficult to maintain. In constricted areas, such as urban or industrial environments it might have been an unavoidable evil necessity but generally the prototype railways had plenty of space allowing them to use simpler track components.

Model railways are often short of space and resort to unrealistic solutions to solve their problems. As I have noted before, it must be said that the AFK has an unlikely abundance of complex trackwork but this was the price that had to be paid for the high intensity operating pattern of the railway that I wanted. Once a few simple points have been constructed and are working properly there is also a quite natural urge to improve and expand one's experience by progressing to more complex formations. It was a combination of space restrictions and this challenge that resulted in the AFK's spaghetti. It is not the intention of this article to give a blow by blow account of construction but simply to examine some of these formations and pass comment upon them.
Photo numbers follow on from the building of a simple point.
37. One reason why the prototype did not like three way points was the difficulty in adjusting the blades. I believe that, in any event, the BoT banned three way points from running lines during the late Victorian period due to a number of derailments. These points, at the entrance to Aspargo depot, similarly proved impossible to adjust so that locos would run down all three legs of the point. It was possible to perm two from three but not get the full set. They were removed and replaced with a tandem point, the construction of which was photographed for the record. One problem with this point is that it was built way below the minimum radius set for the AFK's locomotives. The right hand road barely meets the two foot minimum and the middle road is below this setting. The left hand road (bottom) is one of the tightest on the system and well below the minimum. It has always been restricted to short wheelbase or flexible locomotives.
38. A three way point or a tandem point has three noses which have to be accurately set. The nose closest to the tiebars has been located here and the rough position of the other two can be judged from the marked lines. The usual policies apply. Overlong rails to be cut back and soldering of the rails at four or five sleeper intervals to allow easy adjustment.
39. The other two noses have been set. The one at the top was set first having cut back the rail from the middle nose and setting the second nose with a gauge from the top rail. The alignment with the rail from the middle vee was checked by eye. The second rail of the nose was soldered in and then the third nose was set by using the gauge against this rail and the bottom rail. Two of the knuckle rails have also been set. The one at the bottom is much longer than that on a standard point because the blades will be fitted beyond the central nose. The top one (for the middle road) is elongated so that it acts as the check rail for the bottom nose.
40. Another closure rail has been added but this picture has really been included to show the rail breaks needed on a tandem or a three way point. The central nose must be electrically independent of the two outer roads and the gaps are shown by the red arrows. The straight leg at the top was cut back as far as possible but I prefer to have two soldered joints for each rail at minimum as shown by the green arrow. There are plenty of places on the AFK where another sleeper has had to be slipped in to achieve this which is then hidden with ballast. In fact I could do what I wanted here as the entire point will eventually be put back into its original state of being smothered in ash and coal dust. One problem with complex pointwork is the need to deal with small pieces of metal.
41. The left hand road at the bottom, has had the blades fitted and is now mechanically complete. It has been tested with the bogie. This revealed that the knuckle and crossing nose at the left needed attention (red arrow). The wheels kept taking the wrong route, hardly surprising given the extremely sharp curvature so the vee was filed with a nail file and the knuckle tightened slightly by applying the iron and squeezing with pliers against the bottom rail at the green arrow. The check rail at the bottom was also slightly tightened. As with a simple point there are catches here for the unwary. Tightening the check rail can pull the wheels onto the knuckle if the gauge is not quite correct which will result in a massive lurch and, in the worst scenario, a derailment. The check rail was cut at the blue arrow because there was a fear that it might bridge two electrical sections but later reflections suggest that this was not necessary. The top set of blades will need checking to ensure that they are electrically separate from the nose.
42. The entire point is now mechanically finished, in theory at least, although it awaits gapping. The check rail was cut at the red arrow as a precaution although again retrospective thought suggest that this was unnecessary. An additional sleeper (green) has been added to positively secure the short point blade leading to the right hand road. This is much shorter than its co-acting blade shown by the purple arrow which can induce undue stress into the blades and tiebar. this is one of the unfortunate consequences of building complex pointwork. The blade has had as much metal removed as possible to keep it flexible.

The purple arrow also shows that the blade is wide to gauge and will need correcting. The blue arrow shows that the outside blade has to be soldered to the tiebar behind the blade for the left hand road. This presents an ideal opportunity to solder the whole thing solid! The top blade was attached first and seated as well as possible. The blade into the left hand road has had a notch filed into it to clear the tiebar and soldered joint. This allows it to move freely and reduces the chances of it being soldered to the other blade.

It is also possible, if one is careful, to make minor adjustments to the outer blade by applying the iron inside the inner blade. One other important point highlighted by the orange arrow is that the check rail for the right hand nose is very close to the wing rail of the central nose. There is the potential here for a short circuit as they are attached to different microswitches. There must be no electrical contact between the two which means that careful checking is required.
43. The points have to be gapped with a slitting disc to provide electrical isolation. The red arrow shows the slits cut between the point blades to keep them electrically separate. The blades are connected to the stock rails by the soldered fillets at the green arrow and the unslitted sleepers connecting to the near (return) rail. The central nose needs to change polarity when the blades are thrown and so it is gapped from the blades as shown at the blue arrow. The rail gap at the purple arrow is the end of the blade and is separated from the stock rail by the gap to the left of the arrow. The point of the white arrows shows a trap. This rail could be switched to a different polarity from that under the body of the arrow so there needs to be a gap between the two rails. Note that the check rail in the left foreground has now been gapped.

Now the fun starts! The controller is turned on to the full 12 volts and the polarity of the rails is checked. Needless to say the point is not correctly gapped as is shown by the light which should be out! This stage of construction calls for patience persistence and thoroughness. It is easy to fail to cut an adequate groove into a sleeper. There only needs to be 0.01 mm of continuity to cause a short so this is the first thing to check. By keeping the light on it is easy to tell when the offending piece has been removed as the light goes out. In this instance a point blade had been inadvertently tightened onto an adjacent closure rail causing the short. This was cured by widening the gap with the slitting drill. If all else fails I try to get a knife blade, as noted elsewhere, between the point blade and the closure rail to prove that they are separate. This often knocks off a minute piece of metal which was causing the problem. One other thing to check is that tools on adjacent circuits are not the cause of the short as I have had this happen before! Turn all the circuits off that are not connected to the point and also make sure that metal tools are well clear before starting.
44. The size and complexity of the AFK demand that some records are kept. These are placed into a ring binder and updated as alterations are made. This shows the original microswitch arrangements at Aspargo which were used to change crossing polarity. The new work called for switches A and C to be swapped and switch B was transferred to the right hand rod (X). The switches were left as they were wired as enough slack had intentionally been left to allow for changes in the original installation. The new crossings were connected to the microswitches using chocolate block connectors under the layout to minimise disruption. the record was also amended and dated for future reference.
45. The first locomotive to test the point was 2-6-2 diesel 808, a familiar face for those who follow the AFK. This derailed persistently on the very sharp curve into the siding alongside the coal stack. Each problem was worked through and adjustments made as necessary. It is difficult to explain this process for beginners as it is simply experience and persistence that eventually result in good running. All that I can say is that it is best to proceed gradually and carefully trying to sort out one issue at a time. Often the adjustments to rails are in the region of 0.1 or 0.2 of a millimetre and it is easy to overdo things. (There speaks the voice of experience!) Don't forget to check vertical alignments as well as horizontal ones and remember that one or two passes of a file may be enough to provide smooth running.
46. The Swedish diesel would not run along the middle road because its forward and rear axles are especially prone to derailment. It took a lot of work and time to persuade this loco to go down the centre road. Eventually the princess was finally appeased and deigned to run down the centre road without throwing her toys out of the pram. Her awkward sister, the 2-10-2T could not be persuaded to follow suit and has had to be restricted to the outer road of the shed or the outside siding. This is perhaps to be expected in that the pointwork was pushing the envelope and the AFK prefers to use large locos. In all honesty the depot was an afterthought squeezed into the restricted space inside a two foot radius half circle.
47. A number of three way points remain in service on the AFK without being converted to tandem points. This formation is in the mainline at the northern end of Urteno, at the foot of the Vulpafaŭkangulo, the notorious climb into the Altingablecaŭtoj, the High Reaches area of the layout. The blades have needed some attention to ensure good running. It is easy for the blades to become vertically misaligned causing poor running as a wheel is pushed away from the junction of rails before suddenly springing sideways. The wider radius of these points has eased the problems encountered at the cramped site in Aspargo depot and there have been few derailments once the blades were adjusted. The simplistic approach to wiring, placing it on the baseboard surface as much as possible is evident. The AFK is a large one man layout and this is the middle level of a three layer formation. Things need to be kept as simple as possible.
48a. The original layout at Lacono Cittavecchia with a three way point in the running line.
48b. Some complex pointwork has been removed where possible. The three way point at the entry Lacono seen in the top photo was a constant source of problems which were not resolved by three rebuilds. Eventually it proved possible (just!) to squeeze two standard turnouts into this location and the three way points are no more.  The blades of the nearer point butt against a very abbreviated wing rail of the far point. The three way point in the yard took a lot of persistence to get working correctly but it has not given trouble now for many years. A large number of propelling moves take place over all this pointwork as this is the railway's main mid point marshalling yard. Propelling moves are particularly good at detecting problems in the performance of stock and track!
49a & 49b The double slip at Ithilarak, located in the middle of the level crossing was also not 100% reliable when running onto one of the curved roads. Traffic analysis showed that a simple point in the mainline and another in the loop would provide almost as much operational flexibility so the track layout was modified accordingly.
50. There are places. such as Gasafabrikatastrato yard, at Lacono, where such complications cannot be avoided. Any-one attempting to build a double slip will be well past requiring my advice! This one has seen a couple of rebuilds over the years and, in contrast to commercial ones, has all four blades thrown independently. One difficulty is ensuring that the blades are robustly attached to the tiebars with solder without fouling the adjacent tiebar. The blades can cause a short unless the gap is accurately located when they are in different positions. This, I suppose, is why the commercial offerings (unprototypically) operate from one or two switches on the tiebars and usually have a plastic insert to avoid polarity problems. The other initially disconcerting aspect of double slips is where to fit the microswitch. After a little logical thought it will be apparent that the nearest nose in the photo must be controlled from one of the two operating rods at the far end of the formation. Operation over these points therefore requires that the operator has checked that they are set correctly from the electrical rather than mechanical point of view otherwise the move will stall due to a short.
51. Single slips are similarly prone to the same electrical problems. The polarity of the nearest nose in the photo is selected by the rod on the far set of blades. Unfortunately in this case it sometimes requires moving a blade for electrical purposes even when there is no need to move it mechanically! This one, in the platform at Urteno giving access to the ferry, works perfectly well. It is just that its owner and builder is frequently flummoxed as to how it works! An idiot's guide will one day be produced for his benefit!
52. We have only been concerned with the polarities of point crossings to date but diamond crossings also present problems. Although they have no moving parts they must have their polarities changed if they are not to cause short circuits. The two wide angle rails, such as the one nearest the lorry, will always remain at the same polarity. The two noses must change polarity however depending upon which point is leading into them. This could be solved by including a DPDT switch on the local control board but I prefer to drive microswitches from one of the point rods. Here, at Boursson, three were located in a recess cut into the backscene. They provide one for each crossing nose and one for the point nose. This complexity reflects the fact that the goods loop is on a different traction circuit from the mainline. Obviously if this point is set to the Co-op siding (bottom left, where the van is standing) and a shunt enters the goods loop from via the point at the bottom of the photo a short will occur. This would happen if the DPDT was wrongly set anyhow. Both blades are clearly visible in front of the operator so it is just a case of checking before making the move.
53a. Sometimes, as here in Fenditavalat's Basabazaro, things are more difficult. The three way point at the station throat gives access to the middle road and the siding crosses the goods yard lead by a diamond (at the right). Just to complicate matters the blades are generally hidden from the operator by the gatehouse thereby creating a potential trap when the three way points are set to the siding and the points in the foreground are set to the goods yard, as here. This has been solved by wiring additional microswitches into the circuits to make sure that a warning light, smack bang in the operator's line of sight, comes on at the switch panel when the blades and polarities are in conflict.
53b. You cannot miss that warning light!
53c. Seven switches are operated from the two rods moving Fenditavalat's entry points. The one for the LED is the third on the left on the near side of the track. It is wired through another one on the goods yard point to complete the circuit and illuminate the LED. Incidentally, just to digress, the signal shown in the main photo also has its proceed aspect (two whites) and route indication (white over green) selected by micros attached to point rods, although not these ones in this case! (White indicated proceed in pre war French colour lights and green indicated caution).
54a. Finally we come to the AFK's party piece, shown in a completed state in the heading photo. I do not imagine that too many modellers would require something quite so entangled as this formation and I have to admit it was designed partly to see if it was possible to build a point through a crossing and partly because the traffic patterns on the quayside at Eromarbordo were difficult to reconcile using standard pointwork. It does show, however, that persistence and planning can allow the construction of weird formations using the copper clad sleepers construction method. The only part for concern in this arrangement, having worked up from simpler formations, was the vee in between the crossing and the nose in the centre of the picture. It was necessary because the track was to be embedded in concrete and would otherwise have left a large hole.  It was made over large and cut back with the disc. The heading photo shows the context of the finished layout.
54b & 54c. The top diagram show the track formation and the microswitches required on each rod whist the bottom diagram is the logic table for the polarity of the crossings when points are in different positions. The LED indicators shown on the track diagram were omitted during the installation as the position of the blades is clear to the operator.
55. The pointwork is all part of life for the workers at Eromarbordo but there is always something to talk about, be it women or football!