Australian steam locos

[Gumboots]

I probably wont be doing this quickly (a lot to learn) but at some point I'd like to have a go at doing some of the Australian steam locos. Most of them probably wont be too difficult, as they're not too complex (basic 4-6-2 etc).

However, one loco that would be an interesting project is the AD60. These are a 4-8-4+4-8-4 Garrett. Totally ugly suckers, but legendary (over here) for grunt. One of them was claimed to pulled a 1500 ton freight, including the two dead diesels that were supposed to be hauling it, from a standing start up a 2% grade without slipping. Their low axle loads were handy in a lot of places too, especially on branch lines that had timber bridges.

They were even used, in emergencies, to haul express cars as they were capable of over 60 mph. Obviously they were a little lacking in passenger appeal but they got the job done. They were also very reliable, if not exactly economical. One of them (6009) clocked just short of a million miles over a 20 year service life.

So, I reckon this would be a worthy addition to the loco packs, and it is certainly something different. I realise it's going to be a bit of a mission, and will probably require cheating by combining a tender file and a loco file, and seriously mutating the results. Has anyone ever tried to make something like this?

Just so you can drool over its ravishing beauty and sleekness, here's a picture of one.



And a basic profile drawing.

[Gumboots]

There are a few others I'd like to do too. I figure a six pack would be good. Australians are always up for a six pack. So, as well as the AD 60, I'm thinking it should have these.

D50 2-8-0. These were the ones you grabbed when you just wanted a good, basic unit to haul any old thing anywhere. There were stacks of them, and the class had been in service for well over half a century by the end of steam in NSW. I remember these in daily use.

D57 "Lazy Lizzie" 4-8-2. These were a general purpose grunter, restricted to main lines due to their high axle load. They were nicknamed "Lazy Lizzie" because they made hauling anything seem like light work. They were very reliable, and had an unusual sound due to their three cylinder setup. You can hear one near the end of this video: http://www.youtube.com/watch?v=T6gEjZYFs0M

C36 and C38 4-6-2. Express locos. The C38 was rather classy in streamlined version, although most were just bare. Some railwaymen claimed the older C36 was actually the faster of the two, and could top 90 mph on a long flat run. The C38 ran services scheduled to an average 70 mph, so was still not too slow. Both were good, reliable units.

Not sure about the sixth one.

===========================================================================================================

Hey, so then I started thinking about ones I wasn't so familiar with (meaning the ones that were used in other states when I was a lad). Should probably try a few of those too. Would have to include the Victorian S class (used for the Spirit of Progress) and the NM/C17 class (used for The Ghan). I knew about the names of the runs, but had never bothered looking up which locos they ran.

Spirit of Progress is this critter.

Three hundred invited guests joined the train for the inaugural run, ranging from Attorney-General of Australia former Railways Minister Robert Menzies to Mr AO Henty, descendant of Edward Henty, the Victorian pioneer after whom the train's locomotive was named. The train reached 74 mph (119.1 km/h) against a headwind on the Down journey to Geelong, and on the return leg reached a new official Australian rail speed record of 79.5 mph (127.9 km/h) between Werribee and at Laverton before speed was cut to avoid stray livestock on an unprotected level crossing.

The Ghan NM looks like this.

The original Ghan was notorious for washouts and other delays on the line, and the flatcar immediately behind the tender carried spare sleepers and railway tools, so that if a washout was encountered the passengers and crew could work as a railway gang to repair the line and permit the train to continue. This appalling service was tolerated because steam trains needed water, and Stuart's route to Alice Springs was the only one that had available water.

[arop]

Any kind of user made Beyer-Garrat would be most welcome in this game. But I think it really demands graphic skills to "build" it. Some kind of Beyer-Garrat would be useful in scenarios from Australia, South America, Spain and especially Africa. I really hope someday someone can create one

[Altoona+BeachCreek]

Those are great ideas! You know, if we could coordinate the efforts of the big guys like Ned and Arop with your ideas, I bet we could get a really great looking streamliner based on that Henty engine. It reminded me of a lot of engines from the States, like the Hiawatha and the C&O's Chessie Hudsons. Almost any wheel arrangement could probably be made to work once the streamlined shell is finished.

[Gumboots]

Been thinking about this some more. After looking into the .3dp files for wheels, I'm now confident that I can easily code any number of wheels, in any arrangement, to scale. This means that the basic wheels for a 4-8-4+4-8-4 Garratt, or any other locomotive for that matter, are not a problem.

Most Garratts seem to have quite simple body shapes too, so that is handy. It should be possible to use an existing body as a basis, and just add the chassis girder underneath, etc. The detailing could get quite complex if trying for photo-realism though, since the number of pipes and things could burn up a lot of coding and Photoshop time. At some point you'd have to decide on a reasonable balance (unless you were mad).

I took a look at the .3dp files for the Challenger and Big Boy yesterday, just so I could see how the game devs handled modelling Mallets. It doesn't seem too bad (famous last words). It should be possible to model a Garratt by using the same principles. The Mallet models use a "front truck" that holds the cylinders, bogies, etc for the bits forward of the central pivot. Trucks can be added anywhere, and there isn't a limit of the number of them (as far as I can tell from reading WP&P's posts) so it should be possible to effectively model a Garratt as a "Mallet at both ends", if you see what I mean, with the boiler/body just hanging between the two instead of sitting half on top like it does with a Mallet.

The only real problem I can foresee (apart from the sheer amount of time it will take) is making the thing look realistic enough around corners. That could be tricky, given how the game handles pivot points to track and to other components. It's certainly possible to model a Garratt, but it may not be possible to get it looking really good when it's winding its way up your favourite mountain track, because the ends may tend to look derailed around corners. The way to find out is to try it with a very basic model first, then add the eye candy if it can be made to work well enough.

[thietavu]

Garratts... Yum! Wonder how they might handle the Tibet railways in Great China 4 scenario... :D

[Gumboots]

They'll handle it really well, if you give them unlimited free weight and pulling power in the coding.

Just for a laugh, you could make one and code it so it drags off Shinkansens.

Anyway, if the articulation can be sorted so it's realistic enough most of the time, doing two or three shouldn't be a lot more work than doing one. They're all much the same, really, and the bits that will need to be altered will already be documented after the first one.

I'll start with an AD60. It's standard gauge, so will not need any fiddling for width to match the scale of RRT3's other locos. The 4-8-4+4-8-4 is the most complex arrangement that ever ran in real life, and the length of the thing will test the limits of the concept with regard to RRT3's handling of pivot points. If a long one can be made to work, a short one should work better. So if I can get this beast working, anyone else should have a really good starting point for making one they like.

[thietavu]

Can't wait! :)

[Gumboots]

I've changed my mind about the AD60. After doing a fair amount of research into Garratts, I've decided I'm going to start at the other end of the Garratt timeframe.

The first one ever was the 1909 Tasmanian K1, but that wasn't for main line operation. The first Garratts for main line operation were the West Australian Class M, which were built in 1911 but didn't enter service until 1912.

The Class M were 2-6-0+0-6-2. From what I can gather, it seems that Garratts that lacked leading and trailing wheels each side of the drivers were not as successful. Apparently the lack of guidance caused track wear problems, particularly on corners, and could even result in derailing if it got too bad. Also, since one drive unit was always running "in reverse" it tended to lack stability if it had no leading wheels. Because of this, the vast majority of Garratts were built as 2-6-2+2-6-2 or 4-8-2+2-8-4 or whatever. So, those are the types I intend to model. Might as well do the useful ones.

The first ones that had wheels in all the right places where the Tasmanian Government Railway Class L (2-6-2+2-6-2) and Class M (4-4-2+4-4-2). These are the ones I'm going to try first. They share exactly the same central unit, so in terms of modelling work I'll get two for the price of one. The central unit has shapes that will help ease of modelling too. They were also kinda cute and nicely proportioned, which is always a bonus. Even better, I've managed to track down the original 1911 plans from Beyer Peacock.

These are good ones to have as functional units. The Class L 2-6-2+2-6-2 was a freight locomotive, and was basically the same unit that Beyer Peacock kept producing for smallish freight/mixed Garratts right through until the end of steam. Most of the later 2-6-2+2-6-2's were just minor variations on the Class L, for the simple reason that they pretty much got it right the first time. They were good little grunters, easy to look after, good on fuel, and very durable. So, if I make one of those it'll be useful right through the rest of the steam era.

I'll have to re-gauge it from Cape to standard to suit the scale/track in RRT3, but that applies to most Garratts anyway so I might as well get into it. I've taken a preliminary look at it and don't think it will present much of a problem. It can't just be scaled up by a factor of 56.5/42, because that would make it stupid large. Seems like the way to do it is some subtle cheating on the widths of front and rear units, with possibly slightly wider footplates on the central unit. That will allow moving the cylinders out. I may also slightly cheat cylinder bore size if it needs it.

The Class M is an interesting one. That was an express unit, and was the only 8 cylinder locomotive ever built. Four cylinders were outside in the usual places, and the other four were inside in inconvenient places. The inner ones used a rocker arm to actuate the valve gear. This all made it a nightmare to maintain, but as far as I can tell it wasn't unreliable as long as it was maintained. The reports I can find just grumble about the difficulty and cost of servicing, but don't grumble about it breaking down. It seems to have been reliable, with good speed and good haulage ability.

It actually set a world speed record for Garratts in 1912. Basically they just bolted it together, rolled it out onto the tracks, and promptly clocked 55 mph on a test run.

It is understood that as a result of a test run made with the Garratt passenger locomotive, referred to above, a speed of 55 miles per hour has been attained. The trial took place under the supervision of the Chief Mechanical Engineer, Mr. W. R. Deeble, on the section from Launceston to Deloraine, Tasmania, and was highly satisfactory from every point of view. The engine steamed and curved with exceptional steadiness and the speeds were registered by an automatic recorder.

Ok, so we gotta have one of those. Who wouldn't want a 1912 sports model Garratt in the pack? Since a lot of the modelling will be the same for this and for the Class L, I figure start the Garratts off with one good freighter and one good express. Provisional plan is availability from 1912 onwards, with maybe a 1939/40 end date for the Class M and the L class carrying through until the late '50's.

I think the Class M's reliability should be modelled as Above Average, or possibly Good, but with much higher maintenance costs than the Class L. Class L reliability should be at least Very Good. Fuel economy for both should be decent. At a first guess, Average for the Class M and Good for the Class L sounds about right. Free weight and pulling power adjusted to suit, of course. Passenger rating should be Acceptable for the Class L, and just has to be Ultra Cool for the Class M.

I think I'll make top speed for the Class M 60 mph, on the basis that if it did 55 straight off the showroom floor it was probably was capable of a bit more on a good day. There are unverified reports of it topping 60, and it seems quite plausible. Top speed for the Class L should be around 40. That makes sense, based on equal steam capacity for both classses, differences in driver diameter, etc, etc. Acceleration of Above Average or Good for both sounds about right, given that they are basically double heading* a Prairie or an Atlantic, so they aren't going to be slugs when lightly loaded.

Anyway that's the plan.


*By the way, since Garratts don't have separate tenders I cannot see any reason why it wouldn't be possible to do double headed Garratts, just like diesels and electrics are done. Not sure anyone will want to, but it should be possible as long as the Garratts can be modelled via one body and two trucks each (ie: not using up the tender slot).

[thietavu]

Very interesting reading, thank you! :)

If any of you ever play my China 4 map for TM (still wondering whether it could be converted for ordinary RRT3...), one thing becomes evident. The need for reliable, economical mountain engines. I've often wondered whether Garratts of various kinds could have been ideal solutions there, or in neighbouring countries. Thanks to you, we'll possibly find out.

Were those beasts actually good for mountains? Or was their idea more about low axle weights for heavy trains on level ground and light track?

[Gumboots]

The African ones certainly seem to have been good for mountains, and Tasmania isn't exactly flat either. As with any other locomotive, it comes down to tractive effort, adhesion, and steaming capacity. There's nothing magical about Garratts. If you have straight track laid with heavy rail, there wouldn't be much point in building a Garratt. It would make more sense to do what Union Pacific did later and just build a 4-12-2, because it would be simpler and would do the job.

The Garratt design originated to deal with tight radius curves and a requirement for low axle loads, both things that are often found on mountain tracks. They were popular in Africa largely because of the low axle loads, as far as I can tell. This was also the reason the New South Wales railway ordered the AD60. In the 50's, secondary lines in Australia still had a lot of old timber bridges which couldn't cope with high axle loads. The other thing about Garratts was that they avoided having to use two, shorter, rigid locomotives to deal with tight curves.

The boiler and firebox being unrestricted by driving wheels meant that it was easy to get good steaming capacity too, while still keeping a low centre of gravity. That effectively meant the early Garratts were "Superpower" before Lima thought of it. They were smaller than the Berkshires and obviously not as powerful, but still had the free steaming capacity that some other locomotives lacked. The long wheelbase and low centre of gravity was good for stability on uneven or poorly ballasted track, and around tight corners. So, more complicated than normal locomotives, but worth it in some situations, and perfectly reliable if built properly.

There's not a lot of information around on the L and M classes, but in terms of size and tractive effort they are a close match for some of the South African classes. The Class L freighter is virtually identical to the South African Class GK. Tractive effort is within a few pounds. The Class L's drivers are half an inch smaller. Weight is the same.

In service on the NCCR, each Garratt could replace two of its older Class 7 locomotives, with one 95 long tons (97 tonnes) Garratt able to do the work of altogether 166 long tons (169 tonnes) worth of tender locomotives, while saving 5 long tons (5 tonnes) of coal and one crew’s wages in the process.

When the NCCR was amalgamated into the SAR in 1925, the two Garratt locomotives were renumbered 2340 and 2341 on the SAR roster and reclassified to Class GK.

The locomotives were then initially transferred to Cape Town and worked the Overberg line to Caledon for a brief period. They were then sent to Natal, where they remained in service on the Donnybrook to Underberg branch until they were withdrawn from service by 1957.

The comparison with the older Class 7 mastodons isn't quite fair, since they were the old saturated type (no superheating). Still, two of them together had about a third more tractive effort available than the Garratt, so the Garratt was still doing well. In terms of basic grunt the Class L and Class GK are equivalent to the USATC S160 2-8-0, but with much lower axle loads and better stability and cornering.

The Class M was a close match in tractive effort to the SAR Class GC (actually about 5% lower) but had the top speed advantage due to its much larger drivers. So, even a Class M wouldn't be totally useless on grades. The other thing about the Class M is that, with the inner and outer cylinders being 180 degrees out of phase, it would have perfect primary balance. Each driving axle is basically an inline four. Secondary balance wouldn't be perfect, but with the drivers only turning a bit over 300 rpm at top speed this probably wouldn't have been noticeable. "Hammer blow" should have been almost non-existent. If you look at the builder's pictures for the Class M, you can see that the drivers required hardly any balancing.

x00311.jpg (35.43 KiB) Viewed 13431 times

This may partially account for the engineer's comment about "exceptional steadiness". It should have been a very good express locomotive for its day.

Of course in RRT3 you have no limits on axle load, and even a 4-12-2 will go around hairpin corners at silly speeds, so I wouldn't expect a Garratt to do magical things in the game unless it was coded to be magical. I intend to make the coding as realistic as possible, so the Garratts will be mostly for the fun of having Garratts.

Having more useful locomotives in the "world availability" category will be nice. It's a bit lacking at the moment. There's the Connie up until 1912, but after that all the steamers are useless on grades until the QJ becomes available in 1956. Now if we can have nice Garratts fom 1912, that will be cool for African scenarios.

It has also given me an idea for a Tasmanian scenario. I was looking at maps a week or so back and thinking that really I should do one for Tassie, but couldn't think of a good story for it. There are plenty of good scenario stories from the mainland, but nothing much ever happens in Tassie. Now I have an idea to give the poor old Tasmanians a little bit of glory. They need all they can get.

[thietavu]

Fascinating info! :)

[Altoona+BeachCreek]

That is some truly exceptional research you've done there! It's really incredibly exciting to think that the game really could get a Garrat! Or two, if doubleheading works as planned. The real question is: Do we need 4 sets of drivers on an 8 car train?

[Gumboots]

Yes, of course we need it.

Could always try WPandP's double cars mod to make the trains look twice as long. I haven't played around with that yet but I've seen the screenshots he posted.

[Altoona+BeachCreek]

That sounds awesome! I mostly play the game for aesthetics, especially since this new computer has superior graphic handling capacity @ 6gb RAM. Where can I get it?

[Gumboots]

Read this, and the next few posts, if you want to know more. Rolling Stock skinning ideas

I have thought about this sort of thing while thinking about Garratts, since it's all about cheating the game files to make things more bendy in strange places.

Given that there is no apparent limit to the number of "trucks" and that they can effectively have any content, it did occur to me that it may even be possible to extend the concept to create three cars for every one in the consist. It'd require up to nine "trucks" per car, and a lot of playing around with pivot points for them, but I've been pondering the basic geometry and I can see how it might be possible to place all the pivots so that three cars in one could be made visually convincing, even around corners. At some combinations of car length and corner radius it's likely to break down a bit though. That's the big problem.

This screenshot gives an example of what is likely to happen. As you can see, the pivot points are placed to make the wheels and "cars' do what they ought to, but there's a problem when the corner radius gets small. The "length points" that attach one car to the next have to be placed to allow the cars to be spaced correctly on a straight track. These are points that are fixed at the end of each car body (meaning the single "car" the game thinks it has). The centre line of each car body is always tangent to the track. This means that when the corners get tight, there will be problems with the articulated bits overlapping between one car and the next. This is a hard-coded limtiation of the way the game models cars.

Geometry_problems.png (20.62 KiB) Viewed 13087 times

I know WPandP was finding it diffcult to make the double split cars look good around corners. It's not easy to work around, and is bound to look funny sometimes on some parts of the track.

[SteamTrain]

Go to misc. links tab, then other sites of interest to find

http://hawkdawg.com/img/lnk/wp_p.gif

[Hawk]

You do realize you linked to a gif file, don't you?

I think this is the link you were trying to show.

http://wpandp.com/Index2.html

[Altoona+BeachCreek]

Thanks guys! I was going to download it last night, but the site went down or something.

[Hawk]

but the site went down or something.

Apparently some construction workers were digging around outside of the data center and cut the fiber optics that feeds out to the world.
Same thing happened about a week and a half ago.