Backdating a Hornby 2721 — “To saddle tank, or not to saddle tank, that is the question.” (Quotation from William Shakespeare’s Big Book of Great Western Trains, Part the Second, First London Folio, 1591)

[Dana Ashdown]

I have two Hornby 2721 Class Pannier tanks:

 

i)  No. 2744 - Hornby R.059, an original Margate model I’ve had for forty years

 

ii) No. 2764 - Hornby R.2739, a Chinese produced version I bought second-hand a few years back

 

 

As you can see from the picture, the newer 2764 (right) includes separate hand rails and coupling hooks where the original 2744 (left) has moulded ones. The chassis of 2764 is also slightly better so far as detail is concerned. And the centre driving wheels are the same diameter as the others, unlike the original Margate model, which has slightly smaller flanged wheels to accommodate set track curves (I don’t know why, there is plenty of side-play).

 

For its part, No. 2744 has a nicely plated “brass” safety valve cover and a more “coppery” chimney cap.

 

I acquired No. 2764 with the intention of backdating it to pre-Great War condition, as part of my growing interest in the Edwardian Great Western. At the time I was thinking about just altering the coal bunker, but leaving the rest largely as is, following knobhead’s conversion back in November 2012 (<http://www.rmweb.co.uk/community/index.php?/topic/64814-backdating-the-Hornby-pannier/>). However, my impression is that most of the 2721 Class did not receive panniers until after the war, and as I wanted to keep the existing number plates, there would be little point in just altering the bunker.

 

On due consideration, I was wondering if something more drastic might be in order? I looked at the panniers and concluded that it might be possible to convert the engine to a saddle tank by cutting the tanks off at the bottom, just before they turn under, and replacing the upper part with a curved section of plastic. As it happens, I have a couple of prescription pill bottles that are just about the perfect (assuming the plastic isn’t too brittle), being just under a scale 8 feet in diameter. (I should double-check with callipers.)

 

Before I do anything that I might regret, could someone with a copy of the RCTS’ Locomotives of the Great Western Railway, Part Five: Six-Coupled Tank Engines kindly check the entries for the 2721 Class and tell me when Nos. 2744 and 2764 received their pannier tanks.

 

Also, which 2721 acquired the first full-length panniers? I know that No. 2796 was converted in 1904, but it was given short panniers tanks along with a B4 boiler.

 

For various reasons, I would rather not alter No.2744 more than necessary. Still, I can see that it might actually be a better candidate for a saddle tank conversion, in which case, I would re-paint/re-letter 2764 to 1930 condition.

 

Has anyone else turned a Hornby 2721 into a saddle tank? How did you manage building the tank?

 

 

Dana

[Philou]

The man who might be able help you is @The Johnster as he has a 2721 and seems to know a lot.

 

Cheers,

 

Philip

[corneliuslundie]

Pannier tanks fitted to 2744 in 12/21 along with a B4 boiler; pannier tanks fitted to 2764 in 5/10, S2 boiler fitted 4/10, replaced by B4 boiler 12/15.

(RCTS Pt 5)

Jonathan

[Miss Prism]

The first 2721 to be fitted with standard panniers seems to be 2771, in May 1909. Many received panniers before and during WWI  (which is why photos of 2721 saddle tanks are comparatively rare).

 

Your plan to cut the Hornby tanks off at the bottom, just before they turn under, and replacing the upper part is thwarted. Saddle profiles in relation to the cab were completely different to the panniers:

 

 

Btw, what does the Hornby measure over the pannier sides?

 

[JimC]

The S2 boiler with pannier tanks was a rarish beast. The dome is a good way forward of where it is on the Hornby model. Bit smaller too possibly.

[The Johnster]

Thanks, Philipe, with seeming knowledge comes seeming responsibility, and I do not consider myself an expert in 2721ery.  However. here goes...  The history of this class is complex and it is less a class than it is a group of broadly similar locomotives that it was convenient to group together; boilers were interchangeable though and the later Belpaire boilers could be used on 57xx or 8750s

 

Their origin, with the similar 1854 class, was as 'large' saddle tanks for heavy shunting and short distance transfer freight work, with roots going back to Armstrong days at Wolverhampton.  The use of pannier tanks dates from the introduction of Belpaire boilers, as the saddle tanks would not fit over them, and a system had to be found which retained access to the oiling points between the frames, so side tanks were not appropriate.  I believe there were instances of pannier tank 1854/2721s re-entering service after overhauls without Belpaire boilers, and with the saddle tanks refitted. 

 

There were several different bunker profiles as well, and some pannier tank versions were fitted with 57xx style 'full' cabs, in which case the main distinguishing feature between the two classes was the valance beneath the running plate.  TTBOMK, and I would be happy to be corrected/informed, no 1854 or 2721 pannier ever carried a top feed Belpaire boiler, and there must have been a reason for this but I have no idea what it was!  The 57xx can be regarded as a development of these classes using modern standard components, and many parts were interchangeable between the classes.  Boilers in particular got swapped around, because it takes about 3 weeks to fully overhaul a loco in main works, but about 2 weeks longer to overhaul, test, and certify the boiler, so once the work was done on the loco the next available boiler was fitted and the loco returned to revenue service, freeing up works space for the next overhaul job.

 

The Triang Hornby and Hornby models are 'of their time', and were at their introduction a quite brave foray into the world of the pregrouping GWR.  There are fundamental issues that you have to live with, the most intracable with is the incorrect axle spacing derived from the old R52 Rovex Triang Jinty, which was used up to the 90s on Hornby 0-6-0s including the Jinty, for which it is also incorrect, 2721, J86, E2, J52, 'basic' 08.  This means that the splashers are in the wrong place as well. 

 

Mine is worked up to represent 2761 in her final condition as withdrawn from Tondu at the end of March 1950, and based on a photo of her on the recep roads at Swindon later that year awaiting disposal.  I have replaced the chimney, which is tapered the 'wrong' way to enable the body to be released from the plastic mould, with the correct parallel type, in my case from a scrapped 64xx kitbuild, which also provided a better dome and safety valve cover.  I lost a long battle with the 1980s chassis, with the sprung rear axle, gave up, and replaced it with a Bachmann 57xx chassis, which required a good bit of surgery to the body moulding and, regrettably, has a worm drive that protrudes into the cab through the firehole door. I have positioned the crew to hide it as best I can and given the loco a canvas weather sheet; this was strapped to the rear of the half-cab roof when not in use, and attached by eyelets to hooks sticking up from the rear corners of the bunker, giving a sort of 'drophead coupe' look, but not innappropriate for a layout based in South Wales*.  It also has lamp brackets and real coal added, and glazed cab spectacle plate windows.  It has Railtec number plates and is in it's final livery, proper handrails to replace the mouled ones, and is in Caerphilly Works wartime austerity unlined black.  It should have Caerphilly's 'grotesque' austerity script unshaded G W R initials but I have so far failed to source these.

 

The Bachmann chassis for this loco is wrong in the sense that it has plain fishbelly coupling rods, correct for the 57xx it originally came from.  The splashers, as a result of the models Triang Jinty DNA, are not in line with the axles, but this is not immedieately obvious unless you are viewing the loco side on, and can be lived with,  Some 1854/2721s had fishbelly rods in their later lives, but 2761 was withdrawn with fluted parallel rods; plain parallel rods also appeared on some locos.  I also have an old Wills 1854 which Philou kindly gave me, which is a non-runner which I will build a Southeastern chassis for eventually, though I've been saying that for a few years now... 

 

A saddle tank version can be made using the Bachmann chassis with a 3D printed saddle tank body available via Shapeways from Stafford Road Works.  I'd replace my old Hornby 2721, and the Wills 1854, with an RTR version to current standards in a heartbeat, fond as I am of them.

 

 

*If you can see the top of the mountain, it's going to rain.  If you can't, it's raining...

[JimC]

RCTS says top feed was found on 2734/54/61/85/6 and possibly others. Curiously all but 86 were in the minority that were never superheated.

[Steamport Southport]

2 hours ago, The Johnster said:

The Triang Hornby and Hornby models are 'of their time', and were at their introduction a quite brave foray into the world of the pregrouping GWR.  There are fundamental issues that you have to live with, the most intracable with is the incorrect axle spacing derived from the old R52 Rovex Triang Jinty, which was used up to the 90s on Hornby 0-6-0s including the Jinty, for which it is also incorrect, 2721, J86, E2, J52, 'basic' 08.  This means that the splashers are in the wrong place as well. 

 

 

 

 

The J83 had the old early 1970s Triang-Hornby chassis the same as the 8750. Pretty short lived in the catalogue. But did reappear in the 1990s with a new chassis and motor, which the other 0-6-0Ts also got.

 

It was more or less replaced by the excellent for it's time J52.

 

The rest had a brand new chassis that first appeared under the 08 in 1977. That even had an automatic uncoupler which never appeared on the others.

 

http://www.hornbyguide.com/year_details.asp?yearid=14

 

 

As for the 2721, never had one. When they did come out ISTR Mainline had announced the 57XXs. There was no comparison. 

 

I have looked at them as a project but reckon it would be easier just building the kits.

 

 

Jason

[The Johnster]

18 minutes ago, Steamport Southport said:

 

I have looked at them as a project but reckon it would be easier just building the kits

I think I'd agree with that comment, Jason.  The Wills 1854 is a pretty similar model but was also designed orginally for an incorrect chassis, the same R52 Jinty, and the subsequent Hornby 0-6-0 chassis from the 1977 08 up to but not including the J50 were all, despite evolution over time, repeats of the same axle spacing error.  IIRC this was introduce with R52 to use the same coupling rods as the Black Princess, for which they were also incorrect.

 

Ownership of a Triang Hornby or Hornby 2721 confers on the owner the obligation to accept it's irredeemable issues, to live with it's faults.  I like 2761, she has a lot of character and is an appropriate prototype at Cwmdimbath, but scale modelling she ain't!

[DCB]

1 hour ago, Steamport Southport said:

As for the 2721, never had one. When they did come out ISTR Mainline had announced the 57XXs. There was no comparison. 

Jason

Absolutely no comparison. The Manline had an excellent body with an awful motor and even awfuller wheels.  Very few survive.  The early X03 powered 2721 is pretty much bullet proof mechanically.  The Mainline had an upgrade to become the Bachmann 57XX while the Hornby had a downgrade to the small Mk7 motor.

The Hornby chassis has its roots in the Triang  Jinty but the wheelbase is 8ft X 8ft 3" 3" too short at the back for a Jinty, 1ft too long at the front for a 2721.   The pre 2721 Jinty chassis used on  8751 can be corrected by re drilling the chassis 4mm back from the front axle and using the later wheel sets with 6 fixing screws for the coupling rods and Mainline/Bachmann coupling rods it all bolts together. The 2721 is front axle drive with roots in the B12 so this trick doesn't work ,  drilling both centre and rear axles 4mm back would  mean moving the chassis 4mm back which would mean the motor probably fouls the cab.

The Full size 2721 had the same basic boiler as the Dean Goods.  It was developed for the 57XX with revised and closer stays for 200lbs pressure instead of 180 and gave more trouble than the 180lbs version, albeit with more power.  The taper boiler on the 94XX was a 225lbs boiler but fitted with 200lbs safety valves as was the 2251 to save wear on axle boxes and big ends, some 2251s had cranks in line with coupling rods to try to solve the wear issues which were similar to those of the LMS 4F   Too much power for an inside cylinder loco.

 

Harold Holcoft in his book Locomotive adventure explains Pannier tanks were introduced because the shape of the belpaire firebox inside a saddle tank made maintenance and assembly  extremely difficult as someone had to climb inside.  The few suitably modified Saddle tanks could be used with Belpaire fireboxes and when boilers were changed round top fireboxes could be used with pannier tanks,   Just because a round top boiler was refitted don't assume the loco reverted to a saddle tank.

Until the57XX the 2721 were the GWR's latest last word in heavy shunters so unlikely to be found at Little Puddlecombe hauling two four wheel coaches.   That was a job for an 850 or 2021 class light pannier.  Converting the 2721 to one of those is a real challenge. I have destroyed at least 6 X 2721 in trying.

[JimC]

Yes, converting from a large type pannier )1813/1854/2721/5700 to a small one (850/2021)  would mean changing just about everything. A Bachman 64  would be a better candidate for a 2021 conversion (revering history), although the wheels are too large.

[The Johnster]

19 hours ago, JimC said:

RCTS says top feed was found on 2734/54/61/85/6 and possibly others. Curiously all but 86 were in the minority that were never superheated.

There's no button for 'information very useful.  While it may well have had a top feed boiler at some time in it's career and I am not disputing the RCTS information, it did not have one when it was withdrawn from Tondu at the end of March 1950 when it was withdrawn; I would be interested to know what dates RCTS gives, if any, for top feeds on these locos.  2761s last overhaul was at Caerphilly some time during the 1942-5 austerity period, and I would guess she lost her top feed at that time, but it would be nice to confirm.

[JimC]

Unfortunately RCTS doesn't give any dates. The precise wording is "Top feed was fitted at times to..." and I submit its a reasonably safe bet that locos noted gained and lost top feed as boilers were rotated. RCTS mentions a few 1813s, 1661s and 1854s as having had top feed at some stage in their lives too, so one may guess it was a matter of a handful of experimental boilers circulating round the fleet.

[Dana Ashdown]

On 23/10/2021 at 13:32, Miss Prism said:

Btw, what does the Hornby measure over the pannier sides?

 

I measured it at 1-1/4 inches.

 

[Dana Ashdown]

Thanks everyone for your very helpful replies!

 

I can live with 2764 being given panniers in 1910 - that would work for the time from 1910 to 1914 - and would just require a change to the coal bunker to get something reasonable.

 

That just leaves 2744, so I either have to leave it as is, perhaps with the bunker altered, for 1930. Or do something more drastic to adapt it to circa 1907.

 

As Miss Prism says, the profile of the panniers and saddle tanks were different. But I do think that by taking the top off the engine and replacing it with something similar in shape to a saddle tank, it might work. I does mean replacing all or part of the cab front, &c. At the moment, though, I won't be doing anything until I'm sure it will work.

 

Essentially, I can live with the Hornby's many inaccuracies, so long as it looks the part generally.

 

I have seen the Shapeways top, but I it has flaws as well. The tank seems are grooves, rather than overlapping joints. 

this is from the website:

 

 

All in all, I agree with The Johnston in wishing for an up-to-date, accurate Great Western saddle tank. I don't think that's asking too much of the manufacturers

 

In the meantime, I'll consider my next step and let you know how I make out.

 

Dana

 

 

[The Johnster]

I hope you're right; a half-cab pannier must be pretty low fruit by current standards, and either a 2721 or an 1854 would be ideal for my layout.  I would imagine blue box probably think they have saturated the market with panniers as much as is reasonable for one company, so if one is being considered by an RTR producer I would have thought Hornby, Dapol, Oxford, or perhpas even Accurascale would be more likely to come up with the goods.  The Oxford Dean Goods chassis is probably adaptable to a 2721/1854, though the smaller driving wheels will mean that there needs to be packing to get the body to sit at the right height and a new keeper platie tooled.

[JimC]

8 hours ago, Dana Ashdown said:

I can live with 2764 being given panniers in 1910 - that would work for the time from 1910 to 1914 - and would just require a change to the coal bunker to get something reasonable.

And, ideally, the dome and maybe other fittings. It might be an interesting thing to try with a scrap/damaged body. Must be plenty about. 

[chris45lsw]

Surely the best basis for a 2021 conversion, at least dimensionally, would be a Model Rail/Rapido 16xx?

 

Chris KT

[Coach bogie]

Though posted before, I have "updated" several of the Hornby 2721. I only use the later ones due to the separate handrails. I have not gone for the saddle tank version as they would be too late for me. Both run on Bachmann pannier chassis (Bachmann bodies went to Checkrail for his outstanding Stoke layout). Wheel alignment is not spot on but the centre splasher can easily be moved to line up. On one I went for the older bunker made from scrap brass left over from a coach kit. I also went for the all over cab roof, again from scrap brass on one and a tin of something on the other. This also hides the slight intrusion of the Bacmann mech into the cab area. Brake rods were change to the style with drop sections to avoid the crank pins. I still have one set to fit on 2780 They are not going to win any awards but it was a quick fix to get earlier panniers as 57xx were not first on the list for the Westbury area when introduced.

 

 

 

 

 

 

For my open cab with no motor showing, I went for the Wills/SEF

Mike Wiltshire

[Dana Ashdown]

6 hours ago, Coach bogie said:

Though posted before, I have "updated" several of the Hornby 2721.

I haven't seen these before, but they are very nice indeed! I do like what you've done with the new bunker.

Edited October 27, 2021 by Dana Ashdown

[Dana Ashdown]

I’ve made a few notes about the size of the pannier tanks, cab and bunker on the Hornby 2721 Class that might be useful.

 

Referencing all heights from the top of the footplate, I get the following measurements in scale feet and inches:

 

To bottom of tanks: 1 ft. 9 in.

To beginning of turn-under: 2 ft. 9 in.

To top of tanks: 5 ft. 6 in.

Length of tanks: 18 ft. 1 in.

Total extreme breadth of tanks (side to side): 8 ft.

 

To bottom of outside cab roof: 6 ft. 10 in.

To top of cab roof at the centre: 7 ft. 6 in. (approximately)

To bottom of cab windows: 5 ft. 7 in.

To centre of cab windows: 6 ft.

Length of cab side sheet (excluding handrail): 3 ft. 3 in.

Extreme breadth of cab: 8 ft.

 

Distance between cab sheet and coal bunker (doorway): 2 ft. 3 in.

 

To top of coal bunker: 5 ft. 2 in.

Length of coal bunker at bottom: 3 ft. 5 in.

Length of coal bunker at top (including extension): 4 ft.

Breadth of coal bunker: 8 ft.

 

Total length of body along footplate (from front of smokebox to rear of bunker, but does not include smokebox door and bunker extension): 27 ft.

Total length of footplate: 27 ft. 9 in.

 

 

I’ve taken the liberty of attaching a picture of No.2742 from Jim Champ’s “A Beginner’s Guide to Pannier Tanks” at gwr.org, which shows the engine as built. As the photo of No.2755 posted by Miss Prism, above, and this show, the bottom of the saddle tank is set back about six inches from the sides of the cab (unlike panniers, which were flush).

 

 

By my reckoning, it should be possible to fit a saddle tank to No.2744 by cutting the engine/boiler section down between the front of the smokebox and the front of the cab, to the point where the panniers start to turn under. A tube of the proper diameter, cut to a half-circle, glued onto the top of the remaining panniers would essentially make the saddle tank. To get it right, the remaining bottom of the panniers would have to be filed away so they’re slightly set back from the side of the cab.

 

If I’m correct, the top of the saddle tank should come to about 6 ft. 9 in. above the footplate, which would leave about 9 in. to the top of the cab roof.

 

The resulting gaps in the front of the cab would have to be filled-in (or a new cab front made). Similarly, the front of the tank, where it meets the front of the smokebox would also need filling/replacing. The smokebox door would have to be altered as well.

 

To replicate the plates making up the saddle tank, I could glue thin plastic card onto the tank. (Please don’t ask me to do the rivets!)

 

I’m not sure if the coal bunker should just be re-worked, or replaced entirely with plastic card. As the bunker is a separate component to the body of the engine, a plastic card replacement might be simpler to do.

 

The steam dome and valve cover could probably be adapted to fit the top of the saddle tank. I’m not so sure about the chimney — it looks thinner than the one on 2742. The buffers will also need to be reworked/replaced; a new filler cap and steps made; and the handrails replaced. The tool box will, I think, have to stay where it is because it is moulded into the centre splasher, which means removing it successfully would be rather difficult, owing to the overhang of the bottom of the tanks.

 

 

Two days later…

 

I’ve managed to take the body of 2744 apart without breaking anything! Pictures are attached of the individual components, including the die-cast weight.

 

 

As you can see, the cab roof (not shown) does come off, only to reveal another roof underneath. The good news is that the front of the cab is pretty much solid, so no special work would be needed if the panniers were cut down, apart from the usual filing and sanding. The windows might need enlarging, though.

 

 

The back of the firebox is part of the main body moulding, as is the smokebox door. The separate coal bunker moulding includes the cab floor, so the bottom part may/will have to be kept.

 

I could probably reuse the die-cast weight if I reshaped it to clear the saddle tank. There is also room inside the bunker for additional weight.

 

Now I have to check the tube for the saddle tank. The pill bottle I was thinking about using is slightly tapered, so it won’t work. However, some shorter bottles don’t seem to be tapered, and I could glue two of them together if necessary. They are slightly less than a scale 8 ft. in diameter, so that should be fine.

 

With regards to the set-back of the tank from the edge of the cab, there doesn’t appear to be a lot of thickness in the wall of the pannier, so any removal of plastic will have to be modest — just so long as its visible.

 

 

Dana

Edited October 27, 2021 by Dana Ashdown
Removal of duplicate picture.

[Dana Ashdown]

17 hours ago, chris45lsw said:

Surely the best basis for a 2021 conversion, at least dimensionally, would be a Model Rail/Rapido 16xx?

You may well be right, but I want to work with what I have, even if the Hornby chassis is less than ideal.

 

I think a budget conversion would just about say it all in my case.

[JimC]

We're confusing you I think, with a digression onto other types. A 2021 is a considerably smaller Wolverhampton built tank. A 16xx would be a dreadful candidate for a 2721, 1854 or similar - better off with the Hornby chassis!

 

An alternative for the saddle tank would be to fabricate it from plasticard. Multiple layers of thin plasticard glued together hold their shape, and you could to advantage readily simulate the overlapping panels of the original with, well, overlapping pieces. Find a metal or glass jar the right size, tape the plasticard over and soak in hot/near boiling water and it will hold a curve. Do two pieces and cut one up into the actual panels and glue them on and you should be able to get something not a mile away.

[The Johnster]

I don’t think anyone was suggesting that a Hawksworth 16xx would be a viable starting point for a 2721 conversion, Jim, but for a 2021, a smaller beast altogether.  The 16xx provides the 4’1” driving wheels needed for the 2021; the 4’7” drivers on the 2721, 57xx, 8750, and 64xx, which I mention because they are what is available from 2h RTR sources, are too large, as are the smokebox and the smokebox door diameters.  I haven’t looked at it, but the boiler/tank assembly from the 16xx is probably a lot closer to that of a 2021, though that from a 64xx might be suitable as well. 
 

The top of the boiler/tank assembly on the Hornby 2721 comes away on one piece, as shown in Dana’s photos, and this information may be of some use to owners of Bachmann 57xx or 8750s who want to represent a loco running with a non-top feed boiler.  The conversion will need a whistle and whistle guard added. 

Edited October 27, 2021 by The Johnster

[Miss Prism]

The Hornby Collett post-1924 bunker has always seemed to be too long to me. RCTS refers to this as an 'enlarged bunker', but it unclear whether there was an increase of the 6'6" rear overhang, which had been standard from about 1890 up to that point.