New Zealand A Class 4-6-2

[suctionhose]

Two years ago, I completed a 3” scale Fowler Ploughing Engine which had me thinking: “Is this it?” meaning “Is this my last model? There have been several and at some point, one would expect to declare “I’m over it”. Nothing left to dabble with…

Not so. A couple of months ago, with relish born of abstinence, I started a new locomotive. I don’t propose to do the blow by blow build diary – I did with the plougher on TT – but for general interest, I thought I’d post the progress here now and again.

The prototype is a New Zealand Railways 4-6-2 made famous on the early days of North Island Main Trunk service from Auckland to Wellington in 19 hours or thereabouts. The journey involved many high viaducts, tunnels and the Rimutaka Incline which was essentially a rack railway (technical differences) with grades of 1 in 15.

New Zealand has wild, spectacular topography and the rail connection of two major cities at each end of the North Island was a major achievement in 1908. These loco’s, the A Class, were a response to a desperate need for motive power for the job and as a “balanced 4 cylinder compound with 225lbs WP saturated boilers”, they represented the best technology in the world at that time.

A decade later, superheated, 2 cylinder Ab’s ruled but the A’s still saw 50 years of service all over the Country. A428 still operates in preservation and displays many of the numerous modifications that were implemented over their long service lives, including conversion to superheated 2 cylinder simples. I have chosen to model an early version with HP piston valves prominent, the best-looking chimney barely visible behind the giant oil headlight, 1700 gal tender etc. Best summed up in this photo with the Express enroute to Wellington.




I've only been at it for a couple of months. First the Cowcatcher followed by the main wheels which were manually milled from 170mm diameter slabs of Continuously Cast Iron... At this point, I want to acknowledge the willing assistance provided by various Kiwis whom have freely shared information from their various collections of NZR loco drawings etc! 

[tenor]

I think I am going to enjoy this thread!
Beautiful work as usual.
Martin

[simon6200]

In case anyone missed it, those wheels were “manually milled”. Let that sink in for a while….

[mugbuilder]

If it is built to your usual standard Ross,[and I am sure that it will be], it will be a beautiful model of a handsome machine. Ideal for the heavy West Ryde traffic.

[suctionhose]

To add to the opening remarks above, I will say a few words about “Scope” (of the model).

The A was significant in NZ because it was locally designed and built by the Railway Workshops in 1906. They incorporated a variation on the de Glenn Compound principle, popular in Europe pre superheater days and had wide firebox boilers on a 4-6-2 wheel arrangement– a concept that had been tried only a few years earlier for the first time in the world, in New Zealand in 1901. (Search NZR Q class)

The Compounds had HP’s between the frames driving the leading axle. The first four engines had HP valves driven by Stephenson’s Gear between the frames with eccentrics on the second axle. The LP’s on the outside were conventional with Walschaert's gear ignoring the mass of pipework in the smokebox and interceptor valve for working double high on starting. The 5th engine was simplified by adopting two more sets of Walschaert’s in lieu of the Stephenson’s. The Walschaert’s between the frames derived its motion via rockers from the outside expansion link trunnions. HP & LP valve gears had separate reversers.

I have to tell you I’ve struggled to suppress the urge to do the mixed valve gear version! Wot fun! However, this is where “Scope” enters the equation; more importantly, “Sensible Scope”. So many grand endeavours never make it because the builder cannot resource the dream with adequate time, money or energy. Also overly complex engines can be a bit of pain to own (as the NZR found out...)

Therefore, in view of present circumstances, the Scope of my A Class is this:
1. Build time 2-3 years – and it must get finished.
2. Conventional 2 cylinder superheated model on 5”g at 1.5” to the foot scale (3'6" prototype).
3. External detail and appearance as per A&G Price built Compound (circa 1910)
4. Welded 16mm thick frame for weight, rigidity and simplicity.
5. Minimum complexity for reliability and ease of maintenance.

He’s one from today with the wheels mounted on axles, quartered and ready for tyres:

[runner42]

Hi Ross,

it's going to be a great model locomotive. I am wondering how you are going to fit the tyres, assume heat shrinking them on? If so will this produce any problems on a completed axle arrangement? You have already made the decision on the axle boxes, complete and fitted. The use of 16mm thick material for the frames has me intrigued, I assume that this thickness is occupying the width of the axle box cut out so axle box horns are not fitted.

A more patient obsever will await the subsequent postings to have these questions answered.

Brian    

[David]

I'm really looking forward to this! Even though it goes without saying, those wheels are awesome.

[suctionhose]

Apr 6, 2024 2:04:56 GMT runner42 said:

Hi Ross,

it's going to be a great model locomotive. I am wondering how you are going to fit the tyres, assume heat shrinking them on? If so will this produce any problems on a completed axle arrangement? You have already made the decision on the axle boxes, complete and fitted. The use of 16mm thick material for the frames has me intrigued, I assume that this thickness is occupying the width of the axle box cut out so axle box horns are not fitted.

A more patient obsever will await the subsequent postings to have these questions answered.

Brian    

Hi Brian, yes, no horn blocks. The exact details of the design are still forming in my mind so I'm yet to find all the answers myself. But that's the journey!

Re tyres: 8 thou shrink allowance. They don't need a lot of heat - just barely blue like a temper. Final profile done between centres which is one of the reasons the main crankpin is not yet fitted - gets in the way...

[steamer5]

Hi Ross,
Really looking forward to this build... my dad has a NZR WW, would love to see your setups for your wheel making if you have photos.... they are truly excellent!

Cheers Kerrin

[suctionhose]

Apr 6, 2024 23:28:58 GMT steamer5 said:

Hi Ross,
Really looking forward to this build... my dad has a NZR WW, would love to see your setups for your wheel making if you have photos.... they are truly excellent!

Cheers Kerrin

Hi Kerrin,
Not much to show really unless you consider a big bucket of swarf to be photogenic??

A couple pics showing the basic process of mounting turned blanks on rotary table with spigot to centre for a 5 step process to remove metal by drilling, connecting the holes and gradually getting down to finishing.

I've deliberately shaped them to suit tooling that I have - only an approximation of the actual shape of an A wheel (but then they had a few variations) - but they'll do (see "scope" above)

More telling I suppose would be my 'running sheets' showing the DRO positions and rotary table angles for each step. The finishing cuts were using a tapered end mill, which are a rare legacy from mould making pre cnc days, but can be found online. The taper just gives the appearance of 'draft' like a casting - I was thinking I might double side the draft but decided nothing to be gained by doing so.

The first wheel took a while - maybe 12 hours or so - to settle on the shape and note down the DRO co-ords. With that prep, the other 5 were just under 6 hrs each. More time to bore and key for quartering and final crankpin location (which is on a spigot using the key way as the reference)

[steamer5]

Hi Ross,
Thanks for the pictures & words..... we could of traded buckets of swarf, i ended up with over 30 liters make the stretches for K1.

Wonder how close your wheels are to Dads WW, must have a look when I'm up there this week

Cheers Kerrin

[suctionhose]

I have the 16mm frame welded up and machined now. Not something you could ever do in LBSC's day with hacksaw and file but I have the capacity to remove metal when needed and that has enabled the frame to be cut from 100 x 16 flat bar and welded together in a day or so. Machining a datum surface and thence the horn openings took but a few hours to accomplish.





The rationale here was to keep it simple and heavy for adhesive weight. It weighs 20kg at the moment with another 400mm frame extension still to go on the back to carry the wide firebox, footplate and cab. Interestingly, plate frames are very reliant on running boards and/or side tanks to provide lateral stiffness especially passing the firebox where stretchers are not possible. Curiously, this prototype, though plate framed, had the running boards attached to the boiler. I'm not sure how the frames were held in alignment - possibly with intermediate attachments to the boiler(?) - but in simplifying matters for a model, I felt thick frames were necessary.

There are some other methods to the madness which I will discuss when I have the pictures to help explain my thinking.

[suctionhose]

An update; hoped to be further on but the tyres took longer than expected because they maxed out my little lathe. BTW the lathe is located in a cave way back under the middle of the house these days after doing an extension up above but not moving the original workshop!



The wheels look really good I think! You may notice the positions of the balance weights. The leading wheels (on left) have a small weight on the same side as the outside crankpin though displaced a few degrees around.



Explanation is that the leading axle was the crankshaft for the inside cylinders which was set at 180 degrees plus cylinder inclination to the outside crank pin. Here is a pic of crankshaft turning in the NZR Petone Workshops. (The forging was imported as I understand it).



For reasons already explained, my model will be a 2 cylinder loco carrying the guise of the four cylinder compounds before 1910. Given the necessity of appearances, the external balance weights are located correctly but require counter balancing in lieu of the inside crankshaft, big ends etc. These are the internal weights applying opposite force to the crankpin and external balance weight combined.




An inherent problem with 5"g models is that the frames are necessarily narrower than scale by 10mm or so to accommodate thicker flanges, treads etc. When you push the scale factor a bit as you do for 1 1/8" for standard gauge or, in my case 1.5" for 3'6"g (up from 1.43) the frames end up narrower than actual proportion. This picture shows how the 16mm frames are hidden from view and frames of scale thickness and spacing will be presented to the eye in order to preserve proportion of the HP valve chests which are the dominant feature.




Some other vague evidence of the prototype is starting to show now. Good for morale and good for the many small tasks available now to squeeze in at odd moments. Here's a historical photograph compared to the model as it stands today:

[suctionhose]

Worked on detail that I have been looking forward to making around the HP valve chests:



The bell crank in the centre is the "interceptor valve" to convert to double high for starting. A feature that was very much in the favour of the A Class on the steeply graded North Island Main Trunk line if not so much in favour of the fireman who had make steam for 4 cylinders instead 2 little ones while it was in use.

That said, pictures suggest the actuating cylinder and crank were soon abandoned - an oval plate covered the hole - to which one asks the question, "what then?"

The answer appears to be that the Ad class was rolled out - with the ugliest of all 3 types of funnel used (valiantly defended as being the Capuchon - French for some kind of hat worn in the Caribbean...wot? hmm,,, Okaaay) - anyway the Ad did away with all types of simpling valves, instead supplementing live steam directly into the LP steam pipe from a valve on the backhead.

This practice was common, as I understand it, in those big American compound mallets on the N&W etc. I think (in the US) there was a pressure reducing valve to regulate the supply so as not to choke the HP's with back pressure... or something like that.

Ultimately, with many modifications, there was not much to tell an A and Ad apart so they were all re classified as A's. The Ab was distinctly different having the Vanderbuilt Tender among other defining features of the updated design...

[suctionhose]

Further to the above and to show the interface between "appearances and reality" I can finally show what I was aiming for...

This the front end of the great thick frame...



This is what you will see...



It sits on the step held with two M5 studs at the front and two M6 screws inside the saddle...

[suctionhose]

Lately, just been slaving away on wheels and tyres and over-thinking suspension.

The locomotive frame, without tender, is over 1300mm long (4ft 4ins) which is pretty long for 5”g. I want it to go round 9m curves (30ft Rad) and keep the cowcatcher clear of the track on a 9m Rad vertical curve like coming off the ground level up to a raised steaming bay.

Here’s the challenges:
1. Sufficient side displacement of front bogie (and rear truck when we get to it)
2. Side control / centring device capable of large side displacement
3. Clearance of bogie wheel & cylinders on tight curves
4. Near optimum adhesion without equalised suspension
5. Sufficient vertical travel among axles to keep drivers firmly planted
6. Stability from rocking or pitching when pulling hard

A study of Equalised Suspension systems used by the Americans, reveals ‘grouping of wheels’ in specified ways according to wheel arrangements providing a central point of support in the first group and points of support either side in the second group, making a three-point triangle upon which the sprung proportion of the locomotive sits. Weight is distributed across multiple axles by the equalising beams, which are usually not symmetrical, to equalise axle loads.

Typically, a 4-6-2 has the front bogie as the first group and the 3 driving axles AND trailing truck as the second group. Centre of Mass, excluding wheels, axles, motion (all unsprung weight) is split between the two groups – think of a beam or see-saw with the fulcrum placed to distribute the loads in the required proportion – more on the drivers than the bogie.

Ok. So that being very interesting, I have no wish to make or maintain hundreds of parts for equalised suspension! I need a simplified principle to get the best out of individually, coil sprung axles. And a bogie that will be the equivalent of a “4WD off road version”!

I’ve settled on the principle that my triangle of support will be the centre of the front bogie and the two Trailing Coupled Wheels for reasons that are not obvious until you draw it. The bogie will be unsprung except for where the load is applied at the centre pivot / bolster.

These three points will have very stiff springs and very limited movement to prevent the loco from pitching and rocking. The Main and Leading Coupled Wheels will have long, soft springs – compressed to equal axle loadings – and plenty of vertical travel to meet undulations in track without significant change in axle load.

This sketch shows the effect of humps and hollows at the extremes of a 9m vertical radius. You can see the trailing coupled axle at a fixed height, the front bogie pitched up or down about the swivelling bolster to conform to the track surface. The leading coupled and driving axles rise over the hump or drop down in the hollow courtesy of long travel springs and allowed movement in the horns.



Springs:
Spring Rate refers to the ratio of force vs deflection. A short spring wound from heavy wire has a high Rate meaning a large change of force produces a small deflection. A long spring wound from lighter wire has a low rate and will deflect a lot more for the same change of force.



The short Brown Spring will be used to ‘fix’ the trailing axle within very small limits of movement. The long Yellow springs will be used on the leading and driving axles to provide long travel with very slight changes in load to suit the humps and hollows described above.

Both springs will carry the same axle load however, Yellow will be compressed considerably further than the Brown. Springs are located above the axle boxes in holes drilled up into the16mm thick frame. Adjustment is achieved using spacers above in the blind holes.

Front Bogie:
Reference to the sketch above shows the bogie pitching up or down as required. This is accomplished by a swivelling bolster in the truck which also allows a great deal of freedom for the truck to twist without bottoming out springs or being victim of excessive spring rates. The aim is to balance the loco with 10kgs on the bolster (with a heavy spring for shock absorption) thus ensuring the front of the loco lifts when required to keep the cowcatcher clear of the track.




To prevent the issue of front wheels hitting the cylinder on tight curves, the bogie wheel centres are 10mm longer than scale with the leading wheels pushed out the extra 10mm ahead of the cylinders. It's unnoticeable. Side control or self-centring is yet to materialize and will be further revealed when photo’s are available. Suffice at this point to say that 9m rad curves are very tight for fixed frame of such length so a concept somewhere between Fullsize and Hornby 0 Gauge is required.

Weight Distribution:

Summing the estimated weights of frame, cylinders, boiler and accessories – (excluding the unsprung components i.e. wheels, axles, motion) – I’m guessing 75kgs. I want the distribution front to back to be 10, 20, 20, 20, 5kgs. From the diagram below one can sum the Moments about FB (front bogie centre point) for each axle and resolve them into a single 75kg x 429mm Moment. That is, the CoG needs to be 29mm ahead of the driving axle. Application of counterweights if needed.

[suctionhose]

Been beavering away fairly successfully overall with front bogie, driving wheels and coupling rods completed. Quartering spot on. Great relief!

I bought an expensive piece of hollow bar for the smokebox with the intention to turn and bore leaving a flange at front for the door plate. I want some weight in it to shift the CoG forward but we'll see; it's thicker in the wall than I expected and long, so quite a lot of metal to remove.



Have virtually made all that I have designed at this point so back to drawing now: either the cylinder fabrications or the rear frame extension and back truck (there's another 400mm to go on the back end...)

Below is the centring mechanism for the front bogie. I'll start a thread in General Chat "Bogie Side Control / Centring" to see if there's any interest.



The tee shaped lever with the springs centres the bogie pivot in the slot. The picture below shows one spring extended and other relaxed when the pin moves to the side applying force to centre the in again.



The bogie is not a "model" obviously, rather it is made to function in non scale circumstances. Whether the centring mechanism is helpful for steering, successful or not, will have to wait a while to find out!

[suctionhose]

Splashers:
Some transferable ideas discussed on modeleng.proboards.com/thread/15569/fun-job



Following a lot of little bracket making, I used a plywood jig to locate the splashers concentric with the driving wheels at their running height. Much checking to ensure neither wheel flanges or coupling rods struck the splashers with the axle boxes at the top of the horns.



I took this pic to dispel any thought that model engineers worked in laboratory conditions!



What a lovely feature!

[runner42]

Hi Ross,

excellent interpretation of the prototype design and equally excellent workmanship. How in machining the square section of the main driving pin and its subsequent orientation in the wheel ensure that it is correct to achieve accurate valve timing? Or am I missing something that is fundamental?

Brian 

[suctionhose]

Good question Brian! I've tried different methods on different models. This time the crankpin is bolted into the wheel with csk socket head screw behind.

It was useful to remove the crankpin for tyre turning.I may Loctite it on final assembly once the return crank is made and correctly located.

The other reason for mounting the crankpin this way was a feeling that it was rather long for the amount of engagement in the wheel. Having large dia, low pressure cylinders on the outside may have had something to do with it. Thus I made it with a larger dia shoulder c/bored flush in the wheel that the screw pulls it up against.

Result is a well supported pin and a means to adjust the return crank throw precisely.

Hope that makes sense?

A picture is worth 1 kiloword: