Copyright John Dunn Engineering
Welcome to the Barclayblog.
No.8 was the last 16″ 0-4-0 saddle tank built by the Kilmarnock firm of Andrew Barclay and Sons Ltd. Works No. 2369 she was delivered new to the Scottish Region of the National Coal Board in 1955, the last brand new steam locomotive they bought.
I am lucky enough to own a 3/4 share in the locomotive and I thought people might be interested to share the highs and the lows, the trials and the tribulations, which go cap in hand with the day to day maintenance and running of a full sized industrial “pug”.
I will post on a fairly ad-hoc basis depending on just what there is to talk about so keep checking back!
If you have any photographs or information about No.8 when she was in service I would be pleased to hear from you.
I hope you find the site interesting.
Milling the bottom faces of one brass. This is the face which mates with the axlebox keep.
The original brasses were worn but three out of the four were thick enough to stand re-boring, the fourth was too thin and needed replacing.
The problem with the all of the original brasses was the amount of wear on the ends where the thrust faces on the backs of the wheels had worn them away; this needed rectifying if we wanted to ensure a full bearing contact. There was also some wear on the bottom faces which butt against the tops of the keeps. Both of these issues needed attention and there appeared to be several options open to us; in the end, after a great deal of headscratching, it was decided that, whilst we were having one new brass cast, we might as well have four. The amount of work involved in machining a new brass is not that much different to that involved in trying to “sole and heal” the old ones and it would make a better job all round if they were replaced as a set.
Having got the castings we needed to establish a method for machining them. As with many engineering problems, there is more than one answer but, in this case, we are going to use the horizontal borer.
Milling one of the end faces. This will be at 90 degrees to the faces we established in the previous photograph above.
The first job involves establishing some datums which can be used for holding the brasses square. These are not finished faces but are machined in order to provide some flat, square surfaces which can be used to clamp the brasses onto. The photographs show these faces being milled on the borer. The brass will then be clamped onto an angle plate or box cube to allow the outside diameters to be turned, again on the borer.
Milling the second end. When finished we will have two parallel ends which are both at 90 degrees to the faces which mate with the keeps.
I haven’t posted for quite a while now. This is partly due to being extremely busy at work and partly due to the need to make the tooling which is the subject of this post.
The two parts; the boring bar and the socket which is bolted to the facing slide.
Believe it or not, the title is not an archaic form of insult; a boring snout, or snout boring bar, is an attachment which is used on the horizontal boring machine. Unfortunately the one which came with our borer was fairly small and would not have been rigid enough to cope with the work required in turning, boring and facing the brasses for the axleboxes. These things are not readily available off the shelf, at least for our borer, and so I had to set to and make one.
The finished and assembled snout boring bar. There is a key on the rear of the flange to locate it onto the facing slide.
There is a lot of work involved here, particularly when it is being fitted in at weekends and during evenings around regular “paying” work. The next stage of the axleboxes, the actual bearing brasses, could not proceed without it, hence the large gap between posts.
I think the pictures are fairly self-explanatory so will leave them to speak for themselves.
As an aside; we now have a Facebook page
so please feel free to have a browse….you do not have to be a member of Facebook to visit the page.
Once the thrust face castings had all been fastened into position they needed to be machined away at the sides in order to tidy them up and also to ensure that the various parts (keeps, ‘boxes and bearing brasses) would fit together properly. The heads of the countersunk screws were also milled flush with the front faces in order to reduce the chances of them snagging on hands or clothing.
The actual thrust faces will not be machined back to finished thickness until the bearing brasses are bored, this will ensure that the axis of the bearing is exactly at 90 degrees to the thrust face.
The keeps were attended to first on the milling machine. The sides and bottom were milled back and the screwheads reduced as described. They required a bit of judicious filing to ensure a proper fit into the axlebox.
Boring the crown of the thrust face to match the counterbore in the axlebox. (Copyright John Dunn Engineering).
The axlebox faces were slightly more complicated, not only did the outsides and top need attention but the inside edges needed boring to fit the radius in the crown of the ‘box and the straight portions needed milling level with the inside of the aperture which receives the keep. These also had the screwheads reduced in length.
The left leading axlebox after completion of the work described here. (Copyright John Dunn Engineering).
The 1/2″ B.S.W. countersunk screws made to hold the thrust faces onto the front of the ‘boxes. (Copyright John Dunn Engineering).
As I mentioned in a previous post the drawing I found in the Barclay Archive at Glasgow University specified that the thrust faces should be held on with nine 1/2″ Whitworth high-tensile brass countersunk socket screws. I was unable to find any countersunk brass screws of this size so our only option was to make some. The material chosen was CZ114, a high tensile brass.
The keeps needed new faces on the outside because these were also badly worn and I had obtained some suitable castings. These had been machined up in a similar manner to that described for the thrust faces and would also need some means of attachment. It seemed logical to use a similar style of screw to that specified for the thrust faces but, for practical reasons, these were reduced in size to 3/8″ Whit. I decided to use four to a keep plus a couple of “off the shelf” 3/16″ Whit. brass countersunk screws on each side at the top of the radius for a bit of added support in what might be a fairly “flimsy” area – the photographs should make this clear I hope.
The screws were turned to the standard Whitworth countersunk specifications but the heads were left extremely long, the idea being that they would be tightened down with a pair of Stilsons and then the heads milled off flush with the surface. Hopefully this should provide a neat finish and maintain the integrity of the bearing surface.
Countersunk holes being drilled in the thrust-faces on the borer. (Copyright John Dunn Engineering).
The axleboxes were set up on the borer and, once a datum had been established, the various holes were drilled and tapped. The keeps were smaller and easier to manage. In this case the holes were drilled into the plates on the milling machine and these were spotted through onto the actual keeps on the drill; they were then drilled and tapped to size.
Attaching the various plates was a simple matter of tightening the various screws down evenly. All of the holes were thoroughly cleaned first and a thread locking compound was used.
The thrust faces bolted in position on the front of the left leading ‘box and keep. The screws on the keep have been machined back to make them level with the face, there is still some machining work to do on the ‘box. (Copyright John Dunn Engineering).
Having got the rear face of the thrust faces flat, it was now necessary to machine the outside to the correct diameter to allow the face to enter the recess on the axlebox.
There are several ways this job could have been done but I chose to use a rotary table on the milling machine. I felt this made for the easiest and quickest “set-up”.
The thrust faces were milled down to a diameter that was a snug fit in the recess. There is not a great deal more that can be said about this operation and the picture is pretty much self-explanatory I think.
Milling a thrust face to diameter on the rotary table. (Copyright John Dunn Engineering).
To be certain of a decent fit all of the castings for the thrust faces needed to be absolutely flat on the back face where they would mate with the newly machined face on the axlebox or keep. Furthermore, the faces which would fit into the recesses on the ‘boxes needed to be machined to the correct diameter to ensure a good fit.
The first face of one of the thrust-face castings after the cutter has been “touched-on” and before actual machining has started. (Copyright John Dunn Engineering).
One face was machined flat and then this was used as the datum to machine the second face flat and parallel. I was more concerned with getting the face flat than anything else because the outer face will require further machining to finish it once the face has been bolted to the axlebox. There was a certain amount of distortion, which was expected, and this meant that the casting required shimming in order to allow the second face to be machined flat. You can see the shims in one of the photographs.
Fly-cutting the second face of a casting. The shims which are used to support the casting to helsp remove distortion can be seen in the foreground. (Copyright John Dunn Engineering).
The castings are relatively thin and therefore I was forced to use low profile clamps to hold them down. I was conscious of how secure these might be and therefore decided to take the job more cautiously than usual; I avoided a TCT face-mill and used a fly-cutter to reduce the number of passes required. This is a home made affair which you can see in one of the photographs but it’s certainly effective enough.
The completed face side which will butt up to the face on the axlebox. Flat to less than 0.001″. (Copyright John Dunn Engineering).
The last few posts have dealt with the horn cheeks. We managed to get these finished just before the winter set in, since then work has been concentrated inside the workshop where attention has been focussed on the thrust faces of both the axleboxes and keeps.
One of my previous posts dealt with the way in which the outside of the’boxes and the keeps had been machined back to get rid of the wear which was present. We now needed to come up with a method of repair which would provide an effective bearing for the thrust face as well as making up the gap to the point where the clearances between the backs of the wheels and the axleboxes were correct. There are several possible solutions to this problem.
This is the recess that has been milled into the outside face of the axlebox and keep to accept the new thrust faces. (Copyrght John Dunn Engineering).
After a great deal of consideration I decided to get new faces cast from gunmetal; this method seemed to offer the best compromise under the circumstances. The theory was that the faces would be machined to fit the recesses and then held in place with screws. I placed an order for patterns and eight castings, four for the keeps and four for the axleboxes.
The method by which the new faces would be held onto the axleboxes took a fair bit of headscratching and research. I looked at numerous drawings for various locomotives and it appears that it wasn’t unusual to hold similar faces on with brass countersunk screws; it was just fine details of design that seemed to vary.
Some of the new castings and patterns as received from the foundry. The rectangular parts at the bottom right are the crosshead slippers which are being refurbished and will be dealt with in a later post. (Copyright John Dunn Engineering).
When I visited the Barclay Archive recently I was lucky enough to come across a drawing in a bundle which relates to AB2295, the drawing had been done in 1953 and shows a design for thrust faces which was more or less exactly the same as the one I had done; what was even better was that it had the hole positions, sizes and material (high tensile brass) for all the screws. Cracked it! I have worked to this drawing with minor detail modifications which were necessary to take into account the position of some of the wear which had occurred.
Looking back over the last few posts on the hornfaces I realised that I had not explained one of the problems we encountered and the steps taken to overcome it. I thought the following might be of interest.
Each of the replacement horn cheeks is held on with eight 1/2″ Whitworth countersunk Allen screws. It didn’t take long to work out that there was no uniformity in the setting out of the hole centres; each face was different, sometimes by quite a large amount, and it was important that all of the holes lined up accurately if the screwheads were to sit in their countersinks neatly. I wasn’t too happy with the accuracy of the holes in the existing plates and didn’t really want to use these as “jigs” to drill the new plates from.
There are several methods by which the hole centres could be transferred from the horn faces to the new plates but I chose to go with the method that follows.
First of all I turned up three screwed reference plugs. These were threaded 1/2″ Whit up to a shoulder, there was then a portion that was accurately turned to exactly 1.000″ diameter. The head is immaterial, I used some hexagon bar for convenience but anything would do.
Measuring “Buttons”. (Copyright John Dunn Engineering).
Measurements are taken by triangulation. The three reference plugs are placed into three holes in the horn face and measurements taken between them using an inside micrometer. One inch is added on to each measurement obtained in order to arrive at the centres of the tapped holes. Once the first set of three measurements has been taken one of the plugs is moved to a new hole and a fresh set of measurements made to the two existing plugs; again, once complete a plug is moved, measurements taken and so on until all of the centres have been mapped out.
Measurements were taken from the previously ground flange face of the horn to the three nearest holes with a depth micrometer in order to set the position of all of the threads from a known datum. Remember to add only 1/2″ to the micrometer reading in this case!
Once all of the dimensions had been obtained it was a simple matter to draw each face out on CAD and then convert the hole centres to a series of X and Y co-ordinates which could then be used to drill the holes on the milling machine using the DRO.
Drilling & countersinking the holes for the screws in the horn cheeks (Copyright John Dunn Engineering).
In practice it proved to be very effective and the screw heads fitted neatly into the countersinks with no attention needed.
Once one cheek in each horn gap had been ground flat it was used as the datum for setting the opposite one. Exactly the same procedure was followed, a series of six datum points were set but this time the reference was taken as the six points already used on the initial face. The same procedure was followed for shimming, checking for flatness, bluing, grinding etc. The gap between the two faces was set to coincide with the width across the axlebox horn faces plus a certain amount of clearance. No pictures of this because it’s basically the same as what was shown in the previous post.
I was slightly concerned about the possibility of voids existing between the various thicknesses of shim which had been used to true up the horn faces. The shim stock is steel (brass was far too expensive!) and I felt that any voids might tend to attract and hold water by capillary action which could cause rusting of the shim. If this were to occur it is possible the rust would expand the shim and affect the accuracy of the job. I contacted the technical department at Loctite for some advice and they recommended a product called Loctite 290, this is extremely runny and it was pulled into the gaps by capillary action where it set hard. Hopefully this will alleviate the problem, particularly when everything is well caked in oil, paint etc.
I forgot to mention earlier but all of the original countersunk Allen screws were used during the setting up and shimming procedure. They were all replaced with brand new ones once everything had been finalised but before finish grinding took place. These screws were Loctited in place after the holes had been thoroughly cleaned and degreased.
As I said in the previous post, the initial aim was to produce a face that was accurately placed in relation to the vertical wires in six datum positions. A straight edge was used to ascertain any low spots and these were adjusted by shimming so that they were within 0.002″ of the datum points; measurements between the face and the straight edge were taken with feeler gauges. Any high spots were ignored for now.
The next stage was to make the complete face as flat as possible but it also had to remain accurate to the datum points we had already established. In the absence of any sophisticated equipment I decided the best way to do this was by grinding off any high-spots using an angle grinder fitted with an abrasive disc rather than a grinding wheel. This might sound brutal but, with a bit of experimentation, it was possible to work accurately with this method and I have been more than happy with the results – you do need to be very careful though!
This shows the high spots after the initial contact with the surface plate. (Copyright John Dunn Engineering).
To start with we needed a surface-plate to use as a reference. Now, bearing in mind this plate needs to be larger than the surface we are working on but small enough to be lifted into position numerous times, I decided that a piece of gauge plate (ground flat stock) would fit the bill. Gauge plate is guaranteed to be parallel to within something under 0.0005″ (which is fine) but it is only guaranteed to be flat to within about 0.004″ to 0.005″ which wasn’t accurate enough for the job in hand. I contacted a company called Coventry Grinders and they were good enough to supply me with a piece of 18mm x 150mm Gauge Plate which was specially ground to be flat to less than 0.001″ – brilliant, and very reasonably priced too.
Progress has now been made. The shiny spots show where the grinder has been used and the blue is slowly creeping over the surface of the horn-cheek as it becomes flatter. (Copyright John Dunn Engineering).
The plate was coated with micrometer blue and then rubbed on the hornface; high spots received a coating of blue followed up by a lick over with the grinder. This was a painstaking and slow process but gradually thearea covered by the blue gets larger and the low spots get smaller. The aim is to end up with all of the datum points and as much as the face as possible coated in blue; you don’t want to go any lower than the datum points because the gap between the axle centre-line and the horn-face would then be too large. I aimed for about 75% to 80% blued surface when compared to the overall size of the full face.
Still a bit to go yet but you can see the blue gradually covering even more of the face. The closer we get to finishing, the thinner the coating of blue applied to the surface plate which makes for more accuracy. (Copyright John Dunn Engineering).