Steam, 35028 Clan Line.
5/22/16:Click here to See it pulling a train.
I have indulged a desire long postponed, and that is to have a high quality working model steam locomotive. Though I do have the skills, I have realised that I am not willing to commit the time (several thousand hours) or the money for machine tools to build from scratch. Having got this clear, I decided to purchase a kit and assemble it. Though I would prefer 5″ gauge, such machines are extremely costly so having been watching the gauge 1 scene (1:32 scale, width between tracks 45mm) for several years, I decided to purchase a rebuilt Merchant Navy (RMN) kit for the princely sum of $5600 from Aster Hobbies who have been producing gauge 1 live steam models for the last 40 years or so. The models are produced in limited runs and I got lucky in locating the kit for this model (in England). Well, here she is, 35028, Clan Line.
I was born in 1957 and steam ended on the Southern Region in England in July of 1967 with 35023 Holland Africa Line hauling the last down steam train from Waterloo to Weymouth. She had “THE END, THE LAST ONE” written in chalk on the smokebox door. We lived in Farnborough, about 1 mile away from the main line from Waterloo. I can remember Bulleid Pacifics tearing through, 100mph speeds were known, a couple of times officially too! I once met a man who used to fire on Merchant Navies and he told me that though it was not sanctioned, 100mph speeds were not unusual. He quit the railway when disiesel came, too boring for an active, “up and at it” type. (Almost all British steam locomotives, being relatively small, were hand fired.) While I appreciate the sheer audacity of Oliver Bulleid in his design for the original “Spam Can” Pacifics in both Merchant Navy and West Country/Battle of Britain guises, (the later being referred to as light Pacifics having greater route availability), I think that the rebuilt locomotives were and still are, the finest looking loco’s ever, bar none. These locomotives have three cylinders each having a set of Walschaerts valve gear (unlike the Gresley Pacifics that had outside gear only, the motion for the the middle cylinder being derived from the outside gears using a system of levers.) The unrebuilt loco had outside admission cylinders and the outside cylinders were carried over into the rebuilt loco’s. The use of slide valves for the model naturally leads to outside admission and so the appearance of the valve gear having the radius rod connected to the combination lever below the valve spindle is correct. Forgive me but if you find all this baffling, I do give a basic explanation of the valve gear further into this narrative.
(The use of three or four cylinders was common British locomotive practice to overcome the cylinder diameter limitations imposed by our restricted loading gauge. Other than power, a big advantage of multiple cylinders, especially three, is that smaller balance weights for the reciprocating masses are required resulting in less forceful hammer blow on the rails* and thus greater route availability. The downside is horrible access for maintenance and oiling. The driver would have to get underneath the loco between the frames to oil the inside motion. I have never read of an accident occurring….)
*As a balance weight – located inside the rim of each driving wheel – comes toward the rail and then retreats the vertical component of the motion momentarily ceases, imparting an impulse to the rail just like a hammer blow. Damage to the tracks from this effect was known with large 2 cylinder American passenger locos when pressed to the very high speeds of which they were capable. The superb N&W J class 4-8-4 when tested on the Pennsy reached 110mph, and track damage resulted. (Note: No 611 is preserved and currently in steam. I was blithely unaware of her overhaul at Spencer shops NC and living in Greensboro NC on the route back to Roanoke, could have seen her in steam. Drat. I have been to Roanoke twice over the years just to gaze at her.)
A good reference source for all things Bulleid Pacific is “Bulleid Pacifics At Work” by Col. H.C.B. Rogers OBE.
Here is the box!
Here is the end of the box showing contents and serial number (141 200). A plate with the serial number is affixed to the underside of the right hand (fireman’s side) running board:
This kit requires much more work than simply screwing the parts together (using a multiplicity of tiny metric screws of various sizes). Having read the experiences of others I chose to start with the tender to get used to the instructions and diagrams:
Water tank and hand pump:
Top view with fuel tank removed showing fuel sump:
Complex brake rigging (faux):
The tender is functional and contains both the alcohol fuel and water. There is a hand pump under the cover on the rear deck to fill the boiler. The loco has an axle pump to keep the boiler level up as she runs. The front portion of the tender contains the fuel tank which feeds a triple wick burner, the wicks being around 12mm diameter, each having some 30 wick strands. The burner is force draughted in proper locomotive fashion by the hot gases being drawn through the two boiler flues (that contain superheater elements) by a miniaturised Lemaître style multiple jet exhaust ejector, albeit having 4 pinholes instead of the five 2.5 inch diameter jets that the full size loco is equipped with. This is not a toy, it is truly a miniature locomotive.
Moving on to the loco, the starting point is with the cylinders. To lap the machine marks off the valve and corresponding cylinder faces. A sheet of 1000 grit wet and dry is provided:
The 80 lb surface plate lives on top of my woofer under the CD player! It stays there too, I now avoid lifting such things alone though the CD player is also 80 lb! The amp next to it weighs around 250 lb, I am glad I made it so that the power supply and amp can be separated, it can also be rolled over forwards on the lexan side plates to permit access if needed.
OK, getting back on track as it were, here are the components for the middle cylinder with a completed outside cylinder. A slide valve can be see at top left with it’s steam chest at the centre bottom:
Completed outside cylinders and middle cylinder components with small bottle of steam oil:
Complete middle cylinder with crosshead and slidebars:
Herewith the crank axle (in this case, the middle axle) and eccentric for the inside cylinder valve:
I forgot to photograph the assembly of the frames that are held together and apart by members called stretchers. The instructions admonish the builder to take great care to ensure that the frames are aligned to each other properly, I used the surface plate shown in the cylinder/valve lapping process above to do this.
Here are the frames and wheels:
Steam valve gear is designed to allow the relative phase of the piston and valve motion to be shifted by 180° to permit reversing. The mechanism that does this also allows the percentage of the piston stroke at which steam admission is cut off to be varied so that as the loco gains speed, it can be “notched up” resulting in the steam being used more expansively. Various forms of this mechanism have been around since 1841, the earliest being called Stephenson’s valve gear but of course, it was not invented by him (a precedent for Edison I suppose). Belgium Engineer Egide Walschaerts came along with his mechanism in 1844, and it became the most common locomotive valve gear up until the end of steam on railways. The major component of both gears is the expansion link that oscillates back and forth with the rotation of the wheels. The link carries a die block that may be slid under driver control from one end of the link to the other to obtain variable cut off and reversing. In the case of Walschaerts, the expansion link provides one component of the valve motion via the die block and radius rod, the other component being derived directly from the piston rod, the two motions being combined at the valve spindle by amazingly, a combination lever! Whodathunkit? Stephenson’s gear also combines two motion components but differently, both being derived directly by eccentrics on the axle. (Don’t you just hate it when you get eccentrics on the axle?) Interestingly, the two motions sum to a somewhat square characteristic so that the valve opens quickly, slows while open and then closes quickly which is quite clever, especially when you consider just how long ago all this was thought out. It makes all the variable valve timing hoopla for cars seem somewhat less extraordinary.
Here is one expansion link with radius rod and die blot in the slot. The two plates sandwich the link/radius rod/die block and have trunnions on which the expansion link swings. The elbow connects to the rod that provides motion from a flycrank mounted on the end of the driving wheel axle. The slot in the radius rod carries a bearing on the end of the reverser lifting arm that allows the rod to move back and forth while being held at the required position in the expansion link.
Here it is assembled:
Here is a picture of the left side motion assembled. The flycrank on the rear driving wheel is where the speedometer drive is located, of course faux at this scale!
Here are the three cylinders installed in the frames complete with the middle valve gear, the pipes fitted so far are for exhaust. All three cylinders drive the middle axle. The shaft, sometimes referred to as a weigh shaft, will carry the three reverser lifting arms so that as the reverser is operated, all three die blocks and radius rods are moved together to the required setting. I made a point of drilling a small indent into the shaft where the locking screw for the middle reverser lifting arm engages since it will be hard to access later if it moves. The outside ones can be adjusted relative to the middle one during valve setting. I locked all the motion screws using 222 Loctite, that can be undone if needed. In particular, it is essential to ensure that the cross heads are securely screwed home on the pistons shafts, ditto the valve couplings.
Talking about valve setting, the point is to first align all the reverser lifting arms at the centre, then set the valve slides on the valve rods such that the port openings are equal at both ends of the cylinders and to do this in both forward and reverse gear. (Having the centre axle solid mounted as it is, must help this procedure.) The instructions suggest that if perfect settings cannot be obtained in both forward and reverse gear, to make the forward setting “perfect”. I did have to do that though the reverse settings are still very close to equal. Quite good.
Here, I have fitted the plumbing ready for an air test of the motion. The superheater and four jet exhaust nozzle are clearly visible:
This picture shows the drive system on the rolling road ready for an air test (that was successful). The stainless steel tubes are the super heaters that will later live inside each boiler flue. A fitting is attached where the steam outlet from the boiler would be to allow an air line to be connected. The two U shaped tubes at the front are connected to a displacement oiler. The oiler works (hopefully) by allowing steam to pass down one tube into the oiler where it condenses forming water that pushes the oil up into the other tube. If the top of the smoke box isn’t oily after running, there is reason to take a close look to ensure no blockages!
Superheater and manifold:
While we are on the motion, here is the underside much further along also showing the combustion chamber. The bogie with it’s side control spring and rear pony truck are fitted too. The axle boiler feed pump is between the centre and rear drivers, driven from an eccentric (the buggers get everywhere don’t they) on the front axle.
Now to the boiler, here we have the two superheater elements partially inserted into the boiler flues. The inclusion of a superheater is good news. Heating the steam out of contact with the water makes a lot more work available, furthermore, it saves wasting a lot of the fuel in getting the cylinders hot enough to prevent stalling on condensing saturated steam. Having some experience of small locos that do not have superheaters and having done three steam tests on the rolling road, I can attest that this scheme is much more than cosmetic!
Here is the combustion chamber showing two water tubes that sit in the rather fierce flame from the three large wicks. The flame is sucked around the back of the boiler and into the flues. The chamber is fully lined with ceramic insulation cloth. When the loco is getting up steam using it’s own blower, (steam nozzles located in the exhaust ejector) the flame almost howls! It raises steam VERY quickly. The steam blower is effective from roughly 2 bar (29 psi), half full operating pressure. Below that, it is necessary to use a heat resistant suction fan inserted in the chimney. I used a hand vacuum cleaner run from a variac with a couple of feet of copper tube from the chimney. I also drilled large holes into the tube to permit cooling air to be drawn in, also I used a box fan to blow the fumes outside. It works very well but I do not recommend it, the vacuum still got very hot and smelly, the fumes are most likely toxic too.
Here is the completed boiler, note the water level gauge glass. The long lever is the steam regulator and the small lever behind the regulator is for the steam blower:
Here, I am trying the boiler in it’s shroud, onto the frames. (It may be helpful to cut up a detergent bottle and make a thin plastic sleeve the guide the boiler into the shroud.) It took many tries to get everything aligned properly! If you look carefully, you can just see the water feed to the boiler between the rearmost driving wheel and the boiler. Getting this bend right was tricky, there is barely enough room to sneak between the corner of the combustion chamber and the wheel. The pipe as made had the correct curve shape but it was in the wrong place! I annealed the copper to get the bend out and reformed it where it needed to be. I have attempted to illustrate this in the problems area at the end of this narrative.
Here is a view of the multiple jet exhaust nozzle and the end of the boiler flues. The two longer jets are for the steam blower that is used to force the draught when the loco is stationary, the pipe emerging from the boiler feeds the blower jets and is controlled by a valve on the back-head. The top spigot is the steam delivery union from the regulator.
I forgot to take pictures of the smokebox construction however, I did find it necessary to open the smokebox after some runs. You can see the outside of the exhaust ejector venturi. The small pipes are for the lubricator.
Words to the wise on steam testing. Watch the water glass constantly to ensure that the axle pump is working. If water does not come out when you open the bypass valve, the pump is not working. For the first test at least, leave the return water connection off so you can see the return water flow. WATCH THE LEVEL IN THE TENDER! It gets through water surprisingly fast.
I did something VERY stupid. The fuel tank dispenses the fuel into an open well under the tender. There is a tube with a slant cut end that dips about half way down into the well and a needle valve opening in the bottom of the tank that lets the fuel run into the well. It maintains a constant level because when the fuel covers the end of the cut tube, no air is admitted to the tank, cutting the flow off IF YOU REMEMBER TO FIT THE FUEL CAP! I forgot on the second run and wound up with a nasty fire. I can’t smell wood alcohol* and it is hard to see alcohol burning until it has set something else alight! In mitigation, I did have a soaking wet towel ready and that saved the day, snuffing the fire out instantly with no mess. As with a frying pan fire, there is NOTHING to beat a wet towel. I am so glad that I knew this and maybe these words will save somebody else’s pride and joy.
*The British use methylated spirit that has a strong smell and is coloured purple. Very sensible.
Many components required significant filing to make them fit without forcing, certainly more than simply removing paint here and there.
One expansion link hanger was bent such that it was unusable, here is a picture showing the other side hanger for comparison. I held my breath and carefully bent it straight. Fortunately, the (lost wax) cast parts are brass, both tough and malleable!
Regarding the incorrectly formed water feed pipe, here is a picture showing the bend. You may spot the slight crinkles where the bend was and the pen mark I made to relocate the bend to sneak between the frame and the corner of the combustion chamber. The front corners of the combustion chamber are very close to where the frames rise over the rear horns and it’s width almost occupies the space between the frames making the fit very right. I studied things for quite a while before seeing how to relocate the bend because it is hard to see what’s so with the boiler in place, yet that is what must be accommodated!
All three reverser lifting cranks have forks that fit around the slot in each radius rod, these were all distorted such that I had to carefully tweak them open.
A much more serious problem had to do with the driving wheels. The centre wheels have no suspension by design. The front and rear driving wheels do have horn blocks and springs. I found that the centre wheels were barely touching the rails with the front and rear horn blocks topped out. I checked this by removing the springs and put engineer’s blue at the top of each horn. Sure enough, the centre wheels still barely made contact and there were two blue witness marks left on the top of each horn block. I carefully filed 20 mil (I trained as a toolmaker so precision filing is a skill that I have) off the top of each horn block. By the way, “mil” is NOT an abbreviation for millimeter, it is a correct term for 1 thousandth of an inch!
I found it necessary to work on the expansion link assemblies quite a bit, beyond the lapping of the inside surfaces per the instructions. I found it necessary to shim one pair of expansion link plates further apart, I use paper which provided a small by necessary increase of 2.5 mil. Here is the link with paper shims:
I also found it necessary to reduce the length of the die block pin on one radius rod, the pin was slightly proud of the die block causing the reversing action to balk, here it is after filing the pin down slightly:
I found that one connecting rod was fouling the underside of a slide bar. The lower slide bars are unfinished lost wax castings and looking carefully, I could see that one end was quite a bit thicker than the other causing the screw that secures it to project too far. Simply reversing the bar solved the problem.
Try to test the action of the axle pump when you do the air test, it is imperative that this works properly. I had a bit of a panic when I realised that the water had disappeared below the bottom of the gauge glass during the first steam test! The action of the inlet check ball must be correct and the clearance for it can be adjusted using the fibre washers. For the most part, I don’t use fibre washers, preferring metal to metal plus a smear of sealant so I had things too tight. David Stick speaks to this issue in his expert narrative:
I’m OK with these issues, I enjoyed the fitting challenge but the price of the kit at $5600 led me to expect better.
The vendor seems to think that no significant filing is necessary. I’m not clumsy or stupid, it was necessary. Interestingly, one area of fit that seems to have troubled many, the running boards, was not a problem for me. However, one of the lost wax cast steps up to the running board did need significant filing to fit. Oh, there is a hole on the frame side of both steps that does not align with the corresponding hole in the frames at all, they are only visible on very close scrutiny and contribute little if anything to the structure so I simply omitted the screws that are intended to be fitted. More than one experienced kit builder has suggested that Aster did not have such quality issues until they came out with the rebuilt Merchant Navy kit. The diagrams that come with the instructions are superb. The instructions need careful review. In many places, the order of assembly simply doesn’t work. I did not keep notes and I am not going to try to recall the details. Suffice it to say that trial assembly at every step is essential and it is well worth the effort! A handsome model indeed.
Update, 12/28/15. Since the boiler is of the forced draught type, it needs a device to create a draught until the steam pressure is sufficient to force the draught by blowing steam up the chimney venturi. These devices do not seem to be readily available. Aster Models UK sell them but they are quite expensive so I decided to see what I could cobble together. I have a 4 inch computer fan and I was concerned that the heat would be too much for it. So I fixed it to a plastic funnel, stuck it in the chimney and tried it. (Serendipitously, the funnel interfaces perfectly with the fan.) It worked just fine, even though aerodynamically speaking it is all wrong. I “should” be using a centrifugal blower. As you may expect, there was some melting of the pointy end of the funnel. It seemed to me that if I could fit a coupling onto the funnel that would sit snugly around the chimney rim rather than inside it, all would be OK. So a trip to Home Despot saw me with a PVC coupler and some copper fittings.
Here is a cut down funnel, a PVC coupler and a 1″ copper coupler. The copper coupler fits nicely around the rim of the chimney. I soldered a short length of 1″ copper tube into it to form a transition from the chimney rim to the funnel. The disc was cut from copper clad circuit board (using hole saws with suitable clamping) to form a centering device for the transition piece. The scollops in the PVC coupler were also formed using a hole saw, again with suitable clamping. I had no trouble with this operation, even though the hole saw centering drill was in fresh air, it went perfectly.
The copper transition piece was pushed into the funnel up to the abutment formed by the coupler. The PVC coupler fits the smokebox such that the flats rest on the smoke deflectors, this would ensure that the somewhat top-heavy contraption would sit solidly and not tend to fall off!
Here’s the resulting interface to the smokebox / chimney:
And here it is in place. The top of the funnel is just visible. I cut out a plywood flange to fit behind the rim of the funnel which allowed me to clamp the fan to the funnel. All the joints were made using silicone sealant hence the messy appearance. I took advantage of the annular chamber formed by the pvc coupler around the copper transition piece to fill it with water and made a tiny pressure relief hole into the funnel. This way, the highest temperature that can communicate to the funnel is the boiling point of water. Having tested the complete device, this precaution seems to be unnecessary. It works very nicely, providing just enough draught to make lighting up easy and steam is raised in about 3 to 3 1/2 minutes. The loco blower can take over at 50% pressure (2 bar). I will finish it by fitting a Ni-mH battery pack.