Project X - Constructional Notes

Valid HTML 4.01 Transitional

The photos below have been reduced to speed up the download time. Clicking on any photo will give you the full-size image.

The Frames

The original design calls for 5/32" thick mild steel frames, but this is getting hard to find. 4mm is very close and was deemed to be good enough. Henry found me two pieces which were very close to the required size and needed only a light skim across the length to tidy the edges and a small strip taking off the end.

Henry is of the 'old school' and marks his work out but I am an enthusiastic supporter of the 'co-ordinate drilling' school. I also hate filing and sawing, so I avoid it as much as possible. As I have a milling machine, this gets used a lot where Hnery would use his hacksaw and files.

I started by clamping a piece of angle to the mill and aligning it along the length to act as a guide and reference. Another block of metal was clamped to one end to act as a second reference. The two frame plates were then clamped down (with thin wood packing beneath) onto the table and as many of the holes as I could reach were spotted with a centre drill. I then drilled through two of the holes to take 4BA screws which clamp the two frames together.

I then turned the frames over, and ran a milling cutter along the length to trim the frame to correct size:

mill to size

I then turned it over again and positioned it so that I could spot the holes in the other end of the frame - I only have about 12" of movement in the mill table (imposed by my digital readout scale) so I had to do the spotting in two halves.

I had decided against using chain-drilling and filing to cut out the profile of the frames, and since the three curves involved were all 3/4" radius, decided to trepan the curves using a fly-cutter. This proved to be very satisfactory, taking cuts of 30 thou at a time. The two curves at the rear of the frame were done at this stage, neccessitating removal of the guide angle for the duration but this didn't prove to be a problem.

cutting curves cutting curves

At this stage I drilled through the remaining holes in the rear half of the frame:

drill holes

Next I decided to mill out the slot for the rear axleboxes. I used a 6mm 'throwaway' slot-mill to cut out the recess, allowing a 10 thou margin for error. This proved to be a good decision as the finish was rather rough. A final pass to remove the remaining 10 thou left a much better finish and required only a very slight clean up with the file afterwards:

axlebox cutouts

Finally, I used the same milling cutter to remove the excess material from the rear. This involved two cuts parallel to the length starting at the top of the curves. Again, I left 10 thou for a finishing cut to avoid any roughness caused by slot-cutting. The second cut should actually be at an angle but I cut it parallel to the length at this stage.

cutting parallel

The only remaining operation at this end was to make a further cut at an angle. Since I was going to unclamp the frame to do the other end I opted to do this cut at this point. This involved a certain amount of fiddling to get the frame lined up correctly. This was the only point at which the marking out procedure was required. The correct line starts at 1 1/4" down from the top of the frame at the rear end, to a point tangential to the rearmost curve. Having marked this out, I lined up the frame so this line was parallel to the length of the table and was then able to mill this piece off.

The frames were then returned to normal alignment and slid along so as to be able to cut the profile at the front corner and the front axle-box slot, and finally to drill the remaining holes.

At this point the frames were unclamped, separated and cleaned up using a fine file to remove any burrs.

Frame Stays

The frames are supported by three frame stays, made from 1/4"x1" mild steel. These were a trivial cutting, cleaning up, drilling and tapping job and didn't justify taking any photos. Again I used the mill to ensure that the ends were square. Henry recommends the technique of 'spotting-through' to ensure that the holes line up correctly, but I relied on the DRO on the mill to provide sufficient accuracy. This was rewarded when I came to assemble the frames and stays using 4BA screws. On assembly the screws fitted perfectly and I was pleased to see that the frames appear to be perfectly aligned.

Buffer Beam mounting brackets

The buffer beams will be mounted on the frame using pieces of 1"x1"x1/8" angle (or in my case 25mm x 25mm x 3mm which is all that seems to be obtainable these days). I cut the angle slightly over length and then trimmed it to length in the mill.

Holes to fix the brackets to the frames need to match those in the frames, and again the co-ordinate drilling technique paid off. There are 8 pieces of angle to be cut and drilled, four go outside of the frames and the other four inside. I have a trick up my sleeve for this job. I have replaced the rear jaw of my milling vice with a mild steel bar. This has been drilled and tapped to attach a small plate which serves as a stop, allowing me to locate pieces accurately in the vice. I could therefore slip a piece of angle into the vice, with a small parallel to ensure it was level, and be sure that it was in the correct position for drilling. I only needed to locate the edge of the first angle, the DRO was reset and I could exchange the pieces with confidence that they were in exactly the same position as the previous piece.

drilling brackets

The outside angles are drilled to be 4BA clearance (9/64"), while the inside angles need to be 4BA tapped. I started by spotting all the holes on all 8 pieces with a centre drill. Then I replace this with the 9/64" drill and drilled through four of the pieces. I then swapped for a 3.1mm drill and did the holes in the remaining 4 pieces. Finally, I put a 4BA tap into the mill, removed the drive belt, and used rotated the drill chuck by hand to tap the holes, thereby ensuring squareness.

tapping holes

On completion of these, the angles were bolted to the frames in readiness for the buffer beams. Note that I have not drilled the holes in the front or rear of the angles at this stage.

Here's a photo of the assembled frames. It's a bit fuzzy as I had problems getting this shot. It gives you the idea though; I'll try to get a better one soon.


Horn Plates

The horn plates are specified to be made from 1/2"x1/2"x1/8" steel angle, but as with the buffer beam brackets, steel angle now appears to be only available in metric sizes, so 12x12x3mm steel was used. Again, these were cut to length and trimmed on the mill. On this job, I used a length of left-over angle to act as a spacer allowing the horn plate to project from the opposite end of the vice for machining. The only point here to note is the profile for the lower end of the horn projects below the frames, so a 3/16 section has to be cut away from one side of the angle, and care must be taken to ensure that you end up with 4 of each hand, not 3 and 5!

trimming horns

First Disaster

After being so pleased with the results up to this point, we now come to the first foul-up of the project. I drilled the holes in the horn plates, all 40 of them, without checking. I then discovered that I'd not allowed for the fact that the angles were 12x12x3mm instead of imperial. Worse - I'd measured the offset to the holes from the wrong side of the angle. If it had been imperial and accurately to size, this would be OK, but it wasn't and all the horns fit beautifully but 1/16" in from the cut-out in the frames. I could make the cut-outs wider to match the plates, but that would require bigger material for the axle-boxes which was already to hand. I decided it was cheaper to redo the horn plates....

Horn Plates Revisited

(2nd September, 2006) I have remade the horn plates using the method described above, but this time I drilled just one to be sure that it would fit OK. Measuring from the 'live' side proved to be much more productive!

Before rivetting the horns in place I met up with Henry and showed him the progress so far. One thing which I should have done was a trial rivetting before countersinking the holes in frame for the horn plate rivets. I had countersunk the holes rather too generously and Henry thought I might have problems get the rivets to form nicely. Here is a close-up photo, rather too close for my camera, but it shows the countersink around the 3/32" rivet holes.

countersunk holes

I then attempted a trial rivetting, making use of one of the discarded horn plates and a piece of scrap 4mm steel. The result can be seen here:

trial rivetting

The rivet bent over to one side and a lot of sideways bashing was required to get anywhere near to a reasonable finish. You can see that the result is lop-sided, with space at one side. I had a quantity of 1/8" rivets so I decided to go up a size. I removed the trial rivet and drilled out the holes to 1/8" and produced the following:

rivet trial 2

This was much better, though not perfect. To get this I trimmed the rivet to approx 3/8" in length (bashing my thumb nail in the process). I decided to go ahead on this basis assuming that I could only improve with practice. I did two rivets into the frame and horns and then took this photo:

first real rivets

As you can see I still need more practice. The first had been too long, while the second was too short. They would do, but were short of the ideal. I then devised a fixture to ensure that my rivets would be of consistent length:

rivet cutting jig

This is a piece of mild steel, 10mm thick from my scrap box. I drilled a 1/8" hole in the corner to take the rivet, mounted it in the vice and proceeded to trim the remaining 30 rivets down to length. This helped enormously as it only took a few seconds to do each rivet. I didn't even need to tidy them up with a file after sawing, as the process of extracting the rivet from the hole bent any burrs so that they would fit into the holes in the work. To secure the rivets in the fixture, I simply applied a little pressure with a piece of wood. I used my thumb for the first couple of rivets, but reckoned it would be dropping off by the time I finished. The wood gripped the rivet better than my thumb anyway!

Finally I pressed ahead with forming the remaining rivets. There was still some variation, as the countersinking had been done by eye. The 10mm length was sufficient for almost all of them, though a couple did show a little gap around the outside. Most of them were a little proud when finished so I had to dress them down with a file.

Once the rivetting was complete I then re-assembled the frames and took more photos, slightly better than the one above:

frames with horns 1

frames with horns 2

frames with horns 3

The last of these shows the frame upside down and you can clearly see the oversized rivet heads within. Fortunately they are on the inside and won't be visible when finished.

I've learnt a few lessons this week, the main one being to do a trial when rivetting unless you have previous experience!

Buffer Beams

The Buffer Beams are fairly straightforward, being simple rectangles of 3/16" mild steel. The material I had was wider than required so I milled off the excess. The two beams are identical, so as with the frames, I bolted and clamped the two pieces togethether for drilling and trimming to length. Again I used a length of angle, clamped to the mill table, as a guide, and another piece at one end. This allowed me to unclamp the material and then re-position it without needing extensive re-alignment.

Here's a photo of the beams being drilled:
drilling buffer beams

Once the beams were completed, they have to be rivetted to the pieces of angle at the ends of the frames. This was done by first clamping the beams to two of the angles, which were bolted to the frames. The crucial things are to ensure that the beams are symmetrical; this was done using a 6" rule, measuring from the frame to the end of the beam. When these two measurements are equal, I then checked that the top of the beams lined up with the top of the frames, using the same rule along the top. Once satisfied on these, the clamps were tightened and the bolts holding the angles were removed. The holes in the beams had already been drilled so it was just as case of drilling through the holes into the angles.

Here is a photo showing the beam clamped to the angles before starting to drill:
drilling angles through buffer beam

There is a small complication in that some of the rivets had to be countersunk and filed flush with the front surface as they come just where the buffer stocks go. After rivetting one piece of angle it dawned on me that I should have done these rivets first, then filed them flat before adding the remaining rivets. This was done for the remaining angles, as it was difficult to achieve with the round-headed rivets being so close.


I have made a start on the buffers, but have not yet completed them. This section is just an interim installment to describe what I have done so far.

In his writeup for the loco, Henry describes the buffers as a straightforward turning job on the lathe; Hmmm, maybe it was straightforward for him, but I found it needed a fair bit of thought beforehand, and care in execution. So far it's gone OK, but a bit more description might have helped...

The buffers on the Sweet Pea that I care for are unfortunate. They are definitely not well-matched. I wanted to ensure that those for this loco are a good match, so I devised a method which should ensure this, we will see...

I decided to rough out the buffer heads in pairs. Since each is just 7/8" long by 1 1/4" diameter, I had to have a decent amount of metal to grip them in the 3-jaw chuck. As my lathe collets only go up to 22mm, I had to use the chuck, at least for the shaping of the stems. Doing them in pairs meant I could retain a central section at 1 1/4" diameter which could be gripped in the chuck.

Here is one of the 2" billets that I cut on my bandsaw:
buffer starting material

The photo is a bit fuzzy, but it's not really important. By the time I discovered this, I had machined the billets so it was too late to try to get another photo.

Mounting the billet in the chuck with about 1" projecting, I machined the stems down to 3/8" diameter by 1/2" long. Then I set the topslide right round to cut what will be the back face of the head at 6 degrees, then continued at this angle for a further 1/8", but leaving a rough fillet where it meets the stem. This was just done by eye, then using a 1/4" diameter rat-tail file, I rounded the fillet out nicely. Finish off with emery paper, then drill and tap the stem for the 2BA of the shaft which will retain the head.

Here's a photo of a billet with one end being machined down:
machining buffer stem

And here's the work after being removed from the chuck:
finished stem

Having down one side, I then reversed the billet and repeated the procedure on the other end. Here's a photo of the outcome of that:
both ends machined down

The next step was to saw the buffer heads apart. As I am not a lover of hacksawing, I used my bandsaw to do this. It's a Warco CY90, and the vice is really much too short. I notice that the latest model of this has much longer jaws which come quite close to the blade. Mine, however has about an inch of clearance between the end of the fixed jaw and the blade. The movable jaw is even worse! An article in Model Engineering Workshop outlined a method of getting around this problem:
cutting into two

The work is clamped between two pieces of angle. The bottom piece is 1.5"x1.5", and extends beyond the bandsaw blade. The upper piece is 1"x1.5" angle and stops just short of the blade. There's another piece of 1.25" diameter steel at the back to keep the upper angle parallel. I was a little concerned as to whether the grip was sufficient, but it worked fine. I was careful to stop the hacksaw immediately the buffer heads separated, to avoid cutting through the bottom angle.

Now for the tricky stuff. I wanted to ensure that the buffer heads were profiled accurately, so that they would all match up. To do this, I used an improvised linkage on the lathe to move the cutting tool in a circular arc. This photo shows a buffer head being machined:
profiling the buffer head

The linkage is as follows: A piece of 10mm steel plate is bolted to the old clamp that held the original South Bend switch box. This was removed when I converted to 3-phase with a VFD. The clamp and plate act as a carriage stop and has been used as such for quite a while. There is a flat area on the saddle which is obviously intended for such a function. I made a larger plate, and also made an extension arm which would screw to the top of the plate. This has a pin screwed in near the end. A small plate, with a slot for adjustment, is clamped to the T-slot at the front of the cross-slide and has another pin. A linkage bar, made from a small piece of steel, was made, 2.5" long with holes to match the two pins at 2.25" centres. The linkage is set up so that when the cutting tool is at the centre of the lathe axis the linkage bar is parallel to the axis. As you wind the cross-slide the saddle is forced to move along so that the tool point follows an arc of 2.25" radius.

Here's another photo, sadly not as clear as I would have liked, taken from above the lathe looking down vertically on the work. You can see the almost finished buffer head, held in a 3/8" collet. vertical view of profiling setup

The method used is to face off the buffer head, to establish the centre of the work. Then set up the linkage so that the link is parallel when the tool is at the centre. Wind the top-slide back, so that the tool can clear the work, then advance the top-slide to that when you wind the cross-slide in, it just clips the edge of the buffer. Wind back the cross-slide, advance the top-slide by 10 thou, then repeat. Eventually, the face of the buffer will be rounded off at the correct curvature. Continue with this until the outside edge of the buffer is reduced to 1/16" (measured with a rule, as I still haven't installed my DRO on the lathe). At this point, remove the link, wind the saddle to the right away from the work and round off the edges with a file, then emery to a good finish. This almost guarantees the resulting heads will have identical profiles.


After a rather long gap, I got round to machining the wheels. This is still 'in progress' but I have one photo that I took while skimming the blanks, which were cut from mild steel bar, down to thickness.

facing off wheel blanks

As it turned out, one of the blanks was cut too thin and by the time I'd machined off the saw-cuts it was about 0.01" thinnner than that shown on Henry's drawing. I decided that this was not too critical and machined all four to the same thickness, or perhaps thinness.

The disks were held in the three-jaw chuck, using reversed jaws, and faced down. After both sides had been tidied up I drilled and bored out the centre hole to be a snug fit on a stub axle which had been drilled and tapped 6M to take a clamping screw.

Using the stub axle mounted in a collet I could then machine the outer edge down and shape the tread.

{more details will be forthcoming when I get a round tuit}.

Page created by Bill Purvis, Last updated 17th October, 2007