Sign my

NS 8600 class

Superstructure: cab, boiler and smokebox

When the build of a locomotive draws to an end it becomes increasingly difficult to spontaneously think of the next job to do. You get the feeling the project "is almost done". I then usually take the manual and read it from start to end and carefully note every single chore I might have omitted or otherwise not yet done. You might be surprised how much work is still left!

For the NS 8600 this is now the remaining work

# Description Remarks Status
1 Add jacks (as per photos)   Done
2 Air pressure pipe from low exit air pump [34] Done
3 Solder rear air hoses (after receiving the parts) [36] Done
4 Grab iron on smokebox door [38] Done
5 Mount valve behind stack [39] Skipped 1)
6 Devise a nut/bolt connection between frame and superstructure [41] Done
7 Trial the Philotrain couplers   Done
8 Reapply rivets to the front bottom of the smokebox   Done
9 Weighting the model   Done
10 Make boiler bands   Skipped 2)
11 Unsolder handrail knobs. I should not have soldered them in the first place because I want to blacken them, and keep the handrails themselves blank.   Done
12 Blacken handrail knobs    

Fill holes

- in boiler

- cab roof

- rear end

- smokebox door

14 Make the roof fit on the cab   Done
15 Another whistle on the roof ?    
16 Mount the new wheel and quarter it.   Done
17 Door handles cab    
18 Restore boiler underside   Done

The list above was originally, 13 March 2018, 11 items long. While working on the model, or walking in the park, other issues sprang to mind, and I duly noted them and ranked them in the list. Most work does not necessarily have to be done in the order of the list


1) Job skipped because the valve is very tiny and connected with wired. It does not correspond to the level of detail to try and mimic this detail.

2) The boiler has cast boiler bands which are too thick and not very crisp. Mind you, for the price kit it is acceptable, but I just want more. So my idea was to scrape them off and replace them with 0.25 mm copper strip bands. The scraping is a lot of work though and I decided it was not worth it.

Grab iron


First the grab iron was added on the smokebox. The few photos I have seen seem to point out that this grab iron was angled under approximately 45 degrees and also that it was attached to the smokebox door on only one side, indicating it may have been some sort of door handle.


I drilled a 0.4 mm hole and inserted a suitably bent 0.33 mm hard brass wire. To keep the distance on the free end of the grab iron I underlaid two sheets of yellow sticky note paper, cut from the sticky part so it would stay in place.

Air pipe

The air pressure pipe was bent and glued. It turned out that the lamp was a nasty obstacle getting this pipe into place, so anyone out there reading this: glue first the pipe than the lamp

Air hoses


The air hoses are soldered in place. I cooled the other already soldered parts by putting a shred of wetted kitchen paper on it to prevent other parts from coming loose. I clamped the loco in my vise and the air hose in a separate clamp and manoeuvred it into position. This way I had my hands free to solder. I used 140C solder which quickly flows. As soon as it flows take the iron out.


Note that the valve levers are both on the inside, facing the coupler. This way the shunter could hang in the coupler, attach the hoses and open the valves with a minimum amount of movement. As a consequence the air hoses are mirrored!

Test fitting the couplers

Philotrain offers a very prototypical European standard coupler. To make it work on model railways it is sprung so that it be extracted by the following car when running through a curve. That is also a reason why I fit this locomotive with sprung buffers. The two combined should, I have never really tried, but should, enable operation with this coupler.

I trial fitted the coupler to see if it would fit in the provided square in the buffer beams and if there would be sufficient space behind the the buffer beams to retract the coupler.


The good news of the trial is that the provided square in the beam was large enough. Also the space behind the beam was ample.

The bad news, sort of, was that the spring and the spigot of the coupler snagged behind the round section of the coupler socket provided by Fleischmann.



I filed the top end of the socket skew to thin it. I also added a 0.3 mm keeper plate that guides the spring and the spigot in the right direction and prevents it from snagging. I took care to choose the length of the plate such that it would not interfere with the superstructure.

The front end of the locomotive was treated in a slightly different way but to the same effect.



The NS/SS jacks were separated, they are supplied in one casting, and glued into place.

Weighting the model

After some consideration I decided to try to salvage the weight from the original Fleischmann T3's superstructure. I could not get it out so I milled the plastic of the underside away. But even then getting it out proved to be difficult and I ended up tearing the superstructure to pieces

At long last I got it out and worked the weight down to the size that approximately would fit in the 8600's boiler.

The metal is soft, clogging the files, millers and saws you are working with on it. Don't use your best tools for it.

I also milled out a corresponding space in the underside of the boiler. The resin is very millable, it does not melt, it is not too hard. But man, this is a dusty affair! Make sure to wear a dust mask and clean up your room afterwards. If your workplace is in the same room as your layout: go somewhere else, for heaven's sake.

After this I continued working on the weight but in the end I could not get the original weight to fit well. Every time I tested the loco with the weight the middle axle, which is connected to the gears, would not turn well.

All in all there was still so much work to make it fit that reverted to my trusted lead supply. I cut six and half strips of lead to fit in the boiler and glued them. Calculations show and estimated 24 gram of lead went in. Not much but every gram counts. The original weight would have brought about 32 grams.


In hindsight I regret having attempted to salvage the original weight. It cost me loads of time and I butchered a perfectly good superstructure that could have been sold or used on a diorama.

After the glue had set I trial fitted the boiler on the chassis and again the middle axle would not turn freely despite my estimates and measurements which proved the opposite. This made me think. One weight did not work, the second did not, but should have.

It dawned on my that there might be another cause.

Once that had occurred to me I started searching systematically for the culprit.

  • Testing the chassis alone it ran freely;
  • with the loco loosely assembled with its four main components, chassis, running board, boiler and cab, the loco would not run freely; it did not really block either but it just was not good enough.
  • I took the cab and boiler off and much to my surprise the loco still did not run freely.

This pointed towards an issue with the running board, so I tested with and without running board alternatively several times and the result was invariably the same: with the running board on the chassis it would not run, and without it ran perfectly.

Taking a close look at the running board I found the culprit in seconds. The lower part of the filler pipe, the component I added last to the running board, was touching the wheel flanges. Just. On both sides in fact. A file made short work of that. And now the loco runs freely! After all the toiling with the weight and getting it to sit without interfering with the moving parts there was an entirely different cause!

Devising a bolted connection between chassis and superstructure

One thing that happened when struggling with the weight is that the original Fleischmann bolt connection between chassis and superstructure of the T3 came out.

I took great care to preserve this bolt plate and it needed only little rework to fit under the now available space under the boiler.


But how on earth could I glue it in the exact position where it should be in order to receive the bolt from underneath? The two photos above demonstrate that there is no way to look underneath the boiler through the running board to see where it should sit.


I came up with a trick which involved my trusted friend Blu-Tack. I first installed the bolt plate in the frame and bolted it completely down (left).

Then I covered it with a thin layer of Blu-Tack. I worked carefully on an evenly spread layer so that the entire patch of Blu-Tack would grip.

Then I assembled all four main loco part, thus positioning the boiler exactly in place. Holding it all firmly together I started unscrewing the bolt from the underside but at the same time pressing the bolt head down, so preventing it from coming out of the chassis. This forced the bolt plate to go up and make a connection with the underside of the boiler. When I felt the boiler wanting to come off, I gave way a bit, turned the bolt two more turns and pressed the boiler down again, thereby pressing the boiler on the bolt plate and squeezing the Blu-Tack to grip. I took the bolt out without further ado and gently disassembled the four main parts

And there it was! The bolt plate nicely stuck to the boiler with Blu-Tack on the exact location were it should be. I put it straight, marked its location (at the arrow), removed it, removed the Blu-Tack and glued the bolt plate with epoxy using the obtained mark as a reference. As soon as the epoxy had set for ten minutes or so I reassembled the four main parts and inserted the bolt to see if the bolt plate was indeed in the correct position. It was. So I took the bolt out again and left the boiler aside to let the glue set.

Based on  the product information's indicated strength of the glue, 300 kg/cm2, and the bolt plate's size, 9.2 by 13.8 mm, this bolt plate should theoretically hold a weight of 380 kilo.

Reapply rivets


When I glued the smokebox on the running board, glue contaminated the row of rivets very close to the bottom. Trying to spare them hampered the clean up so I simply stripped them off, knowing I had to do something about that later.


I got my 3D rivet decals by Archer Fine Transfers out. These are ordinary water decals with 3D printed rivet detailson them. I matched them with the remains of the original rivets to determine the spacing. After that I smoothed the bottom strip clean from any remains of the original rivets.

I cut out the approximate length

There is a lot to say about applying decals. I gave a general description  when I built my MDC Shay.

Archer's rivet decals are particularly delicate so you need time on your side, peace mind inside, and Lady Experience to support you. I must admit I wasted two strips of rivets before getting it right.

It is not difficult though.

Normally I place the decal on the surface where it is to approximately go and then gently coax it into place. The main problem was I could hardly place the strip on the same surface because that is where the smokebox door is with all its details. Decals have a habit to embrace other things they should not and it is very hard to persuade them to let go. It is also difficult to persuade to jump over the small ledge on the lower end of the smokebox. So I ended up putting the strip down on the side of the smokebox and gently millimetre by millimetre push it around the corner onto the ledge. It worked but only after I had worked two strips to waste.

After that I let it dry and treated it with SOL and SET as described at the MDC Shay.

Repairing the roof

What? That chore is not in the list. Correct. I made it fit (chore # 14), which wasdone in minutes. Then I inspected the roof for places that needed filling (chore # 13) because I had seen a minute casting bubble exactly on the edge. When I tried to fill it, the edge crumbled (left). More or less disgruntled I thought of a way to repair this and while handling it I accidentally dropped it. Well, resin may be easy to work on, but it also brittle. When I picked up the roof one corner was gone (right). I spent the next half hour with my nose on the floor searching for the missing corner. Then I decided that enough was enough. The damage was difficult, maybe without the missing parts even impossible, to repair and any repair attempt would never be invisible. I decided to have a go at replicating the part.
Luckily I have a small supply of various thicknesses of styrene sheet and a bag of assorted strips available.

From 0.5 mm sheet I cut the main roof sheet and four supports.

I heated the roof sheet in boiling water and taking it out of the water (first) I rolled with my fingers until it approximately held the curve of the original roof.

Then glued the four supports underneath. Note that I wrote which end is in front ("Voor") and at the rear ("Achter") because the overhang on the front is a little larger.


On the subject of gluing: for styrene I use Revell Contacta Professional. I contains a solvent which dissolves the styrene and effectively welds the parts together. It is a specialised plastic glue that is suited only for joining plastic among each other.

Freshly glued, so while the glue-and-dissolved-styrene mixture, is still viscous, I stuck the roof with tape on the cab so it would take the form of the cab during setting of the glue.
To make the ventilation unit I took four layers of 0.75 mm sheet and glued them together. I also added two 0.5 mm strips on the left and right end on one side and one 0.5 mm strip in the middle on the other side. Before the glue had set I clamped it in my vise for a night and by doing so it obtained a curved form

I sanded the ventilation unit, made an overhanging sheet and glued the whole unit in place,carefully observing back and front orientation of the roof sheet.


Note there is an extremely thin pencil line on the roof. Any normal pencil line would be oversize ands inacurate in this scale. Actually it is a scribed line, but they are very hard to see on the white styrene. So I drew roughly a pencil line along it, wetted my finger and rubbed the lead into the scribed line. Tada, visible!!

Old and new together

And as a finishing touch I added the rivet detail. True rivet counters (I did count them) will note ten rivets on the new roof instead of eight in the old. I chose a spacing as close as I could get it, accepting the small inaccuracy.

Restore boiler underside


My milling in the boiler had left a serious flat in the underside of the boiler. Most of the boiler's underside is well hidden in the running board but the front is is still bit visible.

It can be seen from the see through photo that the boiler is a tad too thin on the underside and a carefull onlooker will se that the boiler is a bit frayed on the milling line. So I decided to restore the boiler's round at the front underside. It does not detail like boiler band and the like, because you you won't be able to see that anyway.

So I glued three layers of 0.75 mm and one layer of 0.4 mm styrene sheet on top of each other and made it fit on top of the lead and the bolt plate.
Once the glue had set, I filed and scraped it round to match the rest of the boiler

Filling holes and smoothing surfaces

Every holes, casting bubble, scratch or dent was now carefully traced and duly filled with Milliput.
24 hours later, after the Milliput had completely hardened, the patches were scraped and sanded to blend in with the surrounding area.


A diagram of a wheelset demonstrating that the left crank is runing ahead of the right crank.

Source and  ©:

Locomotive drivers are quartered: the right and left connecting rod pins are oriented 90 degrees apart. As one crank is on dead center, the other is at the position of maximum torque. This ensures that

  • the power strokes are evenly spaced throughout the driver rotation;
  • an engine at stand-still can always start because at least one cylinder can deliver momentum.
Contrary to real steam locomotives it is not so important that a model locomotive has an exact quartering at 90 degrees. An approximation is enough. But what is really counts is that this apprixomation is the same for all coupled wheelsets. If this is not the case and one of the wheelsets is off-quartering the chassis will not run freely or worse, not at all.

I had to replace my model's right wheel of the leading wheelset. I found it wobbling because it had been damaged and on close inspection I found broken spokes. I tried to correct this but to no avail and I ordered a spare wheel at Fleischmann

Ordering is pretty simple.

  • You look up the loco in the product section (the T3's supplier number is 4011).
  • There you choose the tab Downloads / Spareparts list
  • Look up the part number in the list (in this case 514213)
  • Go to / Service / Spare Parts and look up the spare part with its number in the box at the bottom of the page.
  • Add it to the shopping basket and follow the sales instructions. Done!

The wheel itself cost next to nothing (€3,45) but the order surcharge was no less than €9,50. And delivery times are up to two weeks.

So after a good two weeks a big box arrived from Germany (15x15x15 cm) containing mainly big bubble plastic protective packaging and one diminutive plastic bag containing one wheel.

Yes, it could have been sent in an envelope, but it was well packed and it did arrive albeit at a pace more fitting to the eighties. Well, don't complain, the end result is good and that is what counts.

Getting the wheel on the axle was pretty hard. The size of the bush is of course suited for factory processes where the wheel is mechnically pressed on and quartered. But I could not muster the force needed to do this manually. I needed to ream the plastic bush a little to make it fit. This is a delicate job. Be absolutely careful about this. If you ream too much the wheel will not have a press fit.

Once I made it fit, the quartering was approximated by eye, setting the three left cranks in the up positioen and then adjusting the right front wheel to match the forward position of the other two right hand wheels. Then I mounted the coupling rods on both sides and tried to run it. It was not quite right but very close. I gave a little twist and I had a free running chassis. I have had more difficult quarting jobs!



After all this hard work the loco is as good as completed.

I normally make a series of portrits of the completed but unpainted model, but I was simply to lazy for this one. The are many separate parts that need to be attached after painting, handrails, pipes, buffers etc, so getting them on now with Blu-Tack would a huge effort. I wanted to get on with painting so I skipped that part. Let's go painting!