G3 Made Simple
Locomotive Project No. 1
A Live Steam 0-4-4 Tank
A step-by-step construction guide provided
by our contributor “Ginger”.
Questions & Answers
Question from Peter D Hingley : Not really a question but just a comment, what an excellent idea your series is ! Also though a slight addition, that the GWR had at least 3 classes of 0-4-4T, though all had gone by 1948; taken over from the Cambrian Railways, the Barry Railway and the Midland and South Western Junction Railway. There may be others and I can check up and locate drawings if liked. The 0-4-4T is a delightful arrangement, often difficult in electric modelling but giving lots of lovely space for the firebox and burner in steam. Working the other weekend at Sheffield Park I was admiring the steadily re-growing Wainwright H class being splendidly bedizened in her gorgeous livery. The other 0-4-4T I can think of are 2 M7s, an 02, and the Caledonian, what's the 5th ?
Keep up the good work
Reply from “Ginger” : Thank you for your comments. I suppose it is a slight slip of the tongue as there are 5 locomotives in 4 classes in existence, however in a few years time there will hopefully be 6 locos in 5 classes. As a former volunteer coach restorer on the Bluebell I know what you mean about the Wainwright H, and I could not agree with you more.
Question from Christopher Cruickshank : I assume that you have talked to the Association (Alan Gibbs now) about wheel (and other)castings - he has something over a hundred wheel patterns.
However if you need a new one the best route is a 3D drawing of the wheel, 3D printed patterns and lost wax steel castings. All you need to produce is the 3D file and give it to Alan.
Reply from “Ginger” : The 0-4-4 tank for James will use existing wheel castings from the Alan Headech range which are now in the care of the 2.5" Assocn. A revised wheel pattern is in preparation for Sam using sound old fashioned methods as 3D CAD is beyond our capabilities.
Reply from “Ginger” : The advantage of an electrically powered version is that will be considerably easier and much quicker to achieve a powered rolling chassis without the need for a lathe to turn the wheels and the crank axle parts etc. The range of Slater wheels may have a suitable size to fit any of the prototypes. However the body work will still require the same amount of effort. Unless parts are available commercially a lathe, or helpful friend, will still be needed to turn the chimney, dome , buffers etc.

p.s. I am starting an electric powered Stirling SER or GNR 0-4-2 tank later next year - the frames are made and I have a power unit and turned wheels already - it just needs TIME.
Question from a member : If I were to build 'Ginger's Loco', would it be possible to power it electrically?
From Peter D Hingley : Here's a bit more gen on the GWR's absorbed 0-4-4Ts;
CAMBRIAN Nos 3/5/7/8/9/23
Wheels 5' 3"  Wheelbase 4'9" 7'8" 7' 2" 5' 3" 2' 0"
Fairly basic MEM Lloyd drawing ( 14 mm / ft !) I think now available from Welsh Railways Res Circ.

M&SWJR No 15
WB 5' 7 1/2" 7' 6" 10' 9" (to bogie centre) 4' 4" Bogie 5' 0"    5' 2" wheels
Very nice drawing in Mike Barnsley's lovely book, pp 72 -3.

Nos 2/3/4/9
WB 6' 1" 7' 5" 10' 0" 5' 0" 2' 2"     Wheels 5' 8"
Diagram only in Russell p 34
[Think there may be an MEM Lloyd dwg of this type also but don't have ; I have copies of all the others mentioned if this helps anybody].
Incidentally the GNOSR had some attractive 0-4-4Ts but think the wheels of these may have been smaller. Also what about the North Staffs engines ? Nice drawing in a recent mag of the NSR study group.
Finally I don't think people need to go pestering the NRM for original GAs, these are very difficult to use and interpret anyway, (don't I know it !!) and there are adequate modeller's drawing of many of the classes listed .
Anyway hope this helps and keep up the good work !!
Reply from “Ginger” : Thank you very much for taking the trouble to research further examples of 0-4-4 tank locos, which I am sure will prove useful to those wishing to construct a model suitable for the GWR.
From Peter D Hingley : JOGGLED FRAMES - just a comment about the latest episode in your heroic saga ! Bending frames in small model sizes is almost impossible to do neatly and accurately without special equipment. What I have done in the past (admittedly only in 0 gauge, on a Rhymney Railway 0-6-2T) is to mark out the frames in two sections with say a 1/4 inch overlap. A spacing piece can be added if liked from an offcut of frame material. I then solder the pieces together, of course being careful to ensure the top endges are dead straight (silver solder might be better in G3 and the like) and then drill and countersink holes for rivets ( say 1/16 in) which are thoroughly bashed in to reinforce the joints.
Also re the comments about getting frame material, I have always found both GLR Distributors, now of Daventry, (01327 878988) and Maidstone Engineering supplies , now of Staplehurst and part of MAXITRACK, extremely helpful and obliging for obtaining materials cut to size as per spec - hard graft with a hacksaw has no terrors for me but one might as well maximise the use of one's scanty modelling time!
Reply from “Ginger” : The Gauge 1 AMRIG team have developed a simple joggling tool for their SECR H class development that ensures an accurate joggle every time. The tool consists of two heavy plates of steel at least 1/8 thick about 1 foot long which clamp together by means of two locating studs made of round stock. Inside one plate at one end and half way along, is riveted a piece of steel plate representing the depth of the joggle, say 2 mm. Inside the other heavy plate is a similar length of 2 mm plate, but riveted at the other end. When the two heavy plates are joined together the inside plates do not quite meet up and are approximately 8 mm apart.

The frame plate to be joggled, is inserted between the two heavy plates, much as a slice of meat in a sandwich, and the whole lot squeezed in a heavy duty vice. The end of the heavy plates and the top surface act as reference surfaces so that each plate to be dealt with always turns out accurately.
From Andy Boothman : Not a question - just some hopefully helpful input.
A great source of information for anyone wishing to build a MR 0-4-4T is 'Midland Engines No.1'.
Authored by Fred James, David Hunt and Bob Essery, published in 1999 by Wild Swan and still available at £9.95.
It has the GA's and lots of other detail drawings, plus photographs and descriptions for the condensing and non-condensing tanks.
Reply from “Ginger” : Thank you for that. I agree and I have recently purchased a copy which is invaluable if you’re making the MR versions.
From Jon Nazareth : I notice that Ginger has already cut the frames for James and he mentions a scale drawing of these. As I am keen to build this engine as an electrical model, where do I get a set of the drawings from?
Reply from “Ginger” : The drawing for the model is an exact derivation of the NRM G.A. for the large wheel version of the Q class. We need to think of how to make this available from the hand drawn version including all of the relevant dimensions. A CAD version would be beyond us at the moment as there are lots of difficult compound curves. Alternatively I would encourage you to obtain from the NRM the G.A. in size A2 for £7.50 which will provide all of the other dimensions that will also be required as you proceed with construction.
From Jon Nazareth : You talk of both large bogie wheels and small bogie wheels for James so, two questions. Which version, large or small bogie wheels, is being constructed or are the frames the same for both?
Reply from “Ginger” : There were 118 Q class built between 1882 and 1897. The first 28 had 3 foot diameter bogie wheels with a short wheelbase bogie of 4' 10". The other 90 all had 3 foot 9 inch diameter wheels on a 5' 4" wheelbase. Only one of the engines with the small diameter bogie made into Southern ownership in 1924. The model is of the larger type.
From Jem Harrison : As one who has made pigs ears when trying to mill hornguide castings, I noted that you recommend only taking off a thou at a time. In doing that, did you mill down in steps, or did you have the milling cutter set at the full depth?
Reply from “Ginger” : The milling is done in three steps for mild steel as it is a pig to mill. First mill 3/16 at the top to clear the horn guide and cut of say 5 thou wider than the final width. Second, mill half way down a thou at a time. Then with the cutter tool full length mill the bottom very gently, not the sides. finally mill the full length sides.
From Jem Harrison : Part 8 is very good. Just one point that I think has been missed, is the how and when to machine the pear shaped boss, and what thickness should it be in relation to the tyre?

Reply from “Ginger” :  A good point, I overlooked to mention this. The width at the hub very much depends on the prototype and varies quite considerably. For example, the MR Tank, the hub is six and five eighths of an inch thick, Stroudley Goods seven and one eighth, Stirling Passenger six and a half and on outside cylinder engines it can be much thicker.

On the wheel castings received from the foundry you will find that quite a lot needs to be removed from the hub. After having turned the back of the wheel, as described in Part 8, you turn the wheel around and get it centred in the chuck so that the rim spins as evenly as possible. Change the lathe speed using the back gear, using a slow speed to start, and with a tungsten tipped tool slice off the waste material on the hub at 10 thou a time. If you try and cut more than this off the lathe will protest. Sometimes you may have to make between six and eight passes at this to get the hub to the right thickness. It is recommended that the hub stands proud of the tyre width by one-sixteenth of an inch.