LBSC Green Arrow



The inside cylinder casting is a very awkward shape and makes holding it for machining a bit of a challenge.



What makes things more complicated is that the cylinder bore is at an angle of 6° to the valve chest. I did pick up another cylinder casting and steam chest casting off Ebay a couple of weeks ago so if I mess this one up at least I've got a spare!

I decided to get some flat datum surfaces to work from by machining the port face and the bolting face of the mounting lugs. I would only rough machime them at this point just to get them flat.

The port face was very rough so I went over it with a file to get the high spots off and then rubbed it on some emery cloth to get it something like flat. The casting was then clamped in the milling vice with the port face sitting on parallels. The face of the mounting lugs could then be milled flat.



The easiest way to mill the port face seemed to be to clamp the casting to the mill table by the now machined lugs. I had to have the casting at an angle to get the clamps in.



I now had two flat and parallel surfaces to work with as opposed to a knobbly casting. I decided to mill the top and bottom edges of the steamchest so that I could use those to measure the 6° angle of the bore from. I just set the casting up in the vice with the top of the steamchest set parallel to the bed using my Wixey angle gauge.



The casting was turned over with the just machined surface on a parallel and the other edge of the steamchest machined.



I was careful to take the bare minimum off the edges as the casting is undersize in the width before any machining was done so the steamchest will end up slightly narrower than on the drawings.

I thought that it might be useful to make a 6° angle gauge. This should help when I'm setting up the casting in the lathe to machine the cylinder bore although I'm not sure yet how I'm going to do that.

The gauge is just a piece of 1/16" thick steel flat. I clamped it in the vice with the top edge set at 6° to the bed, again using the Wixey angle gauge. That would be plenty accurate enough for this job. The top edge was then milled flat.



The finished angle gauge:



The next puzzle is how to hold the casting for boring the cylinder but that's for another day.


I thought it would be a good idea to set the casting up in the mill first and start the bore with an endmill to get a true surface to use a dial gauge on for centring the casting in the lathe. I clamped it in the mill vice and used the 6° angle gauge with a square to set the casting bore vertical. I then used a 0.375" dowel pin in the chuck to centre the bore in relation to the outside edges of the casting. As it happens the core of the casting was pretty much spot on in the centre which is unusual.



I used a 14mm endmill to 'bore' the cylinder to a depth of 0.125", just to give an edge for the indicator to run on. An endmill was ideal for this as the core of the casting was not perfectly round and a drill would have wandered off. I could have bored it but the endmill was easier and quicker.



Onto the lathe for boring.

I decided to try and mount the cylinder casting on the faceplate using the angle plate as for the outside cylinders. I found that the angle gauge made previously was too thin to sit in the angle plate properly due to the relief slot I had cut in the centre of the Vees so I made another one from 0.1875" steel. This was held into the corner of the angle plate with a small magnet to stop it falling out while I was trying to get the cylinder in place.



With the edge of the port face against the angle gauge and the port face against the other face of the angle plate, the cylinder would be set up with the bore in line with the lathe axis.



I had to be a bit 'creative' with holding the casting in place as I couldn't use the Vee in the normal clamping piece due to the mounting lugs on the casting getting in the way. The mill table clamp holds the edge of the port face firmly against the 6° angle gauge and the lump of steel between the casting and the clamp holds the portface against the face of the angle plate. Hopefully, it would be secure enough.

I fitted the faceplate to the lathe and tapped the angle plate about until the machined register in the bore ran truly. I then went ahead and started to bore the cylinder.



I used one of my carbide insert boring bars again as they seem to give quite a good finish on the gunmetal. I got to about 0.012" of finished size and then called it a night as it was getting late. I'll finish the bore to size tomorrow.

I could do with setting up a power feed to the leadscrew again like I have done before as turning the leadscrew for boring using the handwheel on the end is quite tiring after a bit. One of these days I may have a go at building the electronic leadscrew project as described by Clough42 on YouTube. That would be heaven. No more messing about setting up changewheels! I suppose I ought to get the x axis table feed on the mill done first though. I've had all the bits for that for ages!


Carrying on with the cylinder I finish bored it to size and then took a facing cut across the end to finish the cylinder to length after working out how much needed to come off each end.



At this time I wasn't sure if I needed to face off all the way across the casting so I just faced off to the edge of the face for the cylinder end cover. The casting was then turned around and the other end faced off the same. Later I realised that I should have faced the ends all the way across as I had to put the cylinder back into the lathe and do this.

Now the bore was finished I could put the casting back in the mill and machine the port face and the mounting lugs to the correct distance from the cylinder bore.

I set the casting on parallels in the vice, fitted the length of bar in the bore and then used the DTI on each end of the bar to check that the bore was parallel to the mill table. It was a few thou out so that was corrected before going any further.



Using a depth micrometer to measure the distance from the top of the lug to the bar enabled me to see how much needed to come off the mounting lugs and these were machined to size.



The casting was turned over and the same done to the port face.

I decided to machine the ports next so marked out the centre of the ports on the port face and returned the casting to the vice. You can see the 'ragged' edge left on the end of the port face caused by me not facing the ends of the cylinder bore all the way across the casting and the reason I did this later.



As for the outside cylinders, the ports were firstly spotted, drilled and then milled to final size, all done using the DRO.





There's not a lot of depth for the ports this time as the port face is only about 0.25" from the bore so I played it safe and only made them 0.160" deep.

Whilst the casting was set up like this I also drilled and tapped for the steam chest fixing bolts but I forgot to photograph that.

I thought that I would move on to the steam chest next. This casting is only just wide enough and would finish up undersize by the time I had machined the sides. That shouldn't be a problem as the walls of the steam chest are plenty thick enough.

The casting is a bit unusual in that it has the cover cast in rather than it being a separate piece. This particular casting also has a lug cast on it, presumably to make holding it for machining easier.




I guess that you are meant to hold the casting in the vice by the lug and then machine the bottom face flat. You could also machine the sides at the same setting. However, the lug was far from parallel and wasn't square to the casting so wasn't a lot of use unless you machined it first which would rather defeat the object!

Instead, I rubbed the bottom face of the casting with a file and on some emery paper until it was reasonably flat and clamped it in the mill vice with one edge sitting on a parallel. I then milled off the bare minimum until the face just cleaned up.



The casting was then turned over and the other edge machined, again taking off the bare minimum.

The lug was cut off as it wasn't needed and the top of the casting cleaned up with the bottom sitting on parallels.



It was then turned over and the bottom edge cleaned up.


At this point the casting was 0.1" too thick so I needed to know where to take off the excess material. Looking at the drawings, the bosses are 0.375" in diameter and the centre of the boss is 0.1875" from the bottom edge of the steam chest so the edge of the boss is level with the bottom edge of the steam chest. This meant that most of the excess thickness of the steam chest needed to come off the bottom edge.


I moved onto machining the bosses next by holding the casting in the four jaw chuck as that seemed the easiest way. Firstly, I just chucked the casting roughly centred so that I could face of the end of the front boss. I then marked out the position of the valve spindle using a height gauge on the surface plate and centre popped it.



You'll notice that there are two lines scribed on the boss to mark the position of the valve spindle from the side of the valve chest. The 'non centre popped' one is the correct distance from the edge of the casting but this put the valve spindle way off centre of the boss so I moved it by 1/32" (0.3125") to move it more towards the centre. The valve chest is narrower than the port face on the cylinder so I can compensate for this by moving the valve chest further down the port face.

The casting was then put back into the four jaw chuck and the centre pop on the end of the boss centred in the chuck using a method that I had never used before. I've seen it used many times before and involves using a lathe centre and a dial gauge. The point of the lathe centre is placed in the centre pop and the other end end of the centre held by another centre in the tailstock. Any errors in the position of the centre pop are easily seen on the dial gauge and the casting adjusted in the chuck until the centre runs true. The method works very well.



The boss was machined down smooth and a skim taken across the end of the casting until that cleaned up as well.

As I had started with the front boss of the casting I carried on and drilled and reamed all the way through for the valve spindle. I managed to find a long enough drill to go all the way through the casting and through the boss at the other end. I also machined a recess in the boss to take an O ring seal for the valve rod as on the outside cylinders.

The valve spindle for the inside cylinder is only 0.125" diameter whereas the outside ones are 0.15625" (4mm) diameter. I think that is what confused me earlier on when I thought all the valve spindles were 0.125".

I had intended to tap the end of the boss to take bolts to fasten the O ring retainer but the wall of the boss ended up too thin unless I used 14BA bolts which I didn't fancy! Instead, I tapped the end of the boss 9/32" x 40 to take a threaded retainer.

Once that was done I turned the casting around in the chuck, recentred the boss and machined it and the end of the casting bringing it to the correct length. The last job was to tap the hole in the middle of the boss 4BA to take a brass plug to seal the end of the valve spindle bore. The drawings show the bore in this end of the casting to be a blind hole but it's easier to drill and ream it all the way through and fit a plug.

The final job on the casting was to drill for the steam chest fixing bolts and this was done in the mill using the DRO.



The steam chest is held on by 7BA countersink screws as the top is right up against the frames so the holes were countersunk as well. I managed to put a nice mark on the top of the steamchest by carelessly not raising the drill high enough when I moved from one hole to the next! I didn't notice it at the time as the top was covered in chips.

A trial fit on the cylinder showed that all the holes lined up nicely.



You can see that the end of the cylinder is not flush with the ends of the valve chest due to the 6° angle between the bore and the stean chest. I'm not decided whether to just leave things as they are or try machining the cylinder casting to match the end of the steam chest. The only way it seems to do that accurately is to set the cylinder casting up in the rotary table so that I can just machine the outside of the casting without touching the ends of the cylinder where the covers attach. It probably isn't worth the trouble. We'll see.

I think the next job is to fit the cylinder end covers and drill and tap the holes for their fixing bolts. I can then see where the steam passages need to go to clear the various holes. I've a feeling that it's going to be a bit awkward to position them.


The cylinder end covers needed the same treatment as those for the outside cylinders i.e. the spigots needed reducing in diameter to fit the new bores and generally cleaning up.

I started with the rear cover and held that in the collet chuck using the same threaded bit of brass bar that I used for the outside cylinder covers.



The spigot was reduced to fit the cylinder bore and the bolting face and the outside edge cleaned up.



The cover was then bolted to the cylinder as for the outside cylinders and the holes for the fixing bolts spotted through, drilled and then tapped 7BA. Incidently, the original bolts are actually 3/32" Whitworth which makes the loco fairly old but I decided to 'upgrade' them to the modern 7BA. The whitworth bolts will be saved and used in future restorations where I might want to keep the original fixings.



The front cover was a bit of a mess with two cut outs around the edge which wouldn't have cleaned up. I'm not sure why they were put in but my guess is that the cover may have originally been intended for one of the outside cylinders and someone made a mistake. I mentioned that I had obtained another cylinder casting and steam chest from Ebay and also included was a front cylinder cover casting so I decided to make a new one.



Fortunately, the casting had the spigot on it which made it very easy to hold in the lathe collet chuck and machine the bolting face, spigot and the outside edge. The casting would have done for a much bigger cylinder and I had to remove quite a bit of metal to get it to size. While it was in the collet chuck I removed the chuck from the lathe and mounted it in the mill vice to drill the holes for the fixing bolts using the DRO



It was at this point when I realised that I had machined the outside diameter to 1.25" instead of 1.125" so I had to take some more off. It could have been worse - I could have made it too small!

I machined the front face of the casting using the aluminium bar jig that I had made for the outside cylinder covers, holding it in place with the tailstaock centre against the end of the spigot.



The spigot was then sawn off and the jig drilled and tapped for a couple of bolts to hold the cover in place. The remains of the spigot could then be machined off.



As the fixing holes had been drilled accurately using the DRO I could do the same for drilling and tapping the holes in the end of the cylinder. I didn't need to clamp the cover in place and spot through this time.

Time to drill the steam passages.

I spent hours with the 3D model working out the best way to position and drill these. It really is awkward with the angle between the bore and the steam chest! I must have tried half a dozen different layouts using three 3/32" diameter holes but trying to get one end of the hole at a suitable place on the end of the cylinder and the other end to tie up with the steam ports was a nightmare involving different angles for each passage which would have made the whole thing another nightmare to set up.

I eventually settled on using only two passages but of a bigger diameter to give the same area as the original three. It only meant increasing the diameter to 0.116" (a number 32 drill or near enough 3mm) to get the same area so the passages still fitted in easily. It didn't help that one of the bolt holes fell where the passages needed to go so I had to put one passage either side of that. I could have rotated the front cover to avoid that but the rear cover had to be in a certain position for the slidebar to line up properly.



Using the layout shown and by drilling the passages in line with the cylinder bore the passages came out at the ends of the steam ports which was ideal. It was probably more luck than judgement though!



The problem now was how to hold the cylinder to drill the passages. I had worked out that the passages needed to be drilled at an angle of 7° to the bore so that they exited just above the bottom of the steam ports but couldn't think of an easy way to do this using the ordinary milling vice. Looking through my collection of vices I came across a tilting drill vice which solved the problem. I'd forgotten that I had got it. I think it was possibly an Ebay buy at some time.

By clamping the cylinder in this vice I could bolt it to the mill table and then tilt it to the required 7°. I used the digital angle gauge to set that after squaring up the vice with the table.



Prior to this I had printed out a drawing of each end of the cylinder showing the position of the ends of the passages relative to the centre of the bore so the next job was to centre the bore on the mill spindle using a DTI in the chuck. The bore appears slightly elliptical to the mill spindle due to the 7° angle but I just centred the x and y axis which was accurate enough. Using the DRO I could then find the exact position of the end of the passages ready for drilling.

I first deeply spotted the passages with a 3mm spotting drill. The face of the cylinder is at a slight angle of course but the spotting drill is nice and stiff so wouldn't deflect.



The passages were then drilled with a smaller drill and finally opened out to the correct size. It's always a tense moment when doing something like this in case you have made a cock up and the drill appears somewhere that you don't want it to! Either that or the drill breaks! Fortunately, the drilling went ok with no problems and the passages came out in the steam ports where they were supposed to.



All I've got to do now is mill a slot from the passages to the cylinder bore, turn the cylinder over and do the other end.


The ports are now finished at both ends of the cylinder.



I mentioned before that the valve chest came out too narrow and I dropped it down on the port face so that the bottom edge was flush with the bottom edge of the port face rather than the top edge. This meant that the top of the portface was about 0.008" higher than the steam chest so I decided to put the cylinder block back in the mill and take a skim off the top to bring it flush with the steam chest. Not really necessary but it looks better.



I did some checking of the steam chest next and realised that the cavity for the valve wouldn't be wide enough or deep enough so I put it in the mill and took some off the top and bottom sides of the cavity and increased the depth by 0.060". The top of the steam chest was about 0.125" thick so there was still plenty of metal left.



Time to make the new piston and rod and this was made exactly the same way as for the outside cylinders, finishing the dimensions to size after the piston was fitted to the rod to make sure that everything was concentric. As for the outer pistons, the rod was left long to be trimmed to length later.



That's pretty much everything done for the inside cylinder apart from the valve, valve rod and the glands. It seems to have been a long time coming!



I want to 'rethink' the valve and the method of driving it when I get to that point. The drawings show the valve driven by the valve rod having a flat machined on either side running in a thin slot in the valve which is a method I don't like. There is very little area in contact with the valve to drive it and no means of initially adjusting the valve so I will try and fit the usual rectangular 'nut' running in a cross slot in the valve and threaded onto the valve spindle. The valve can then be adjusted by rotating the valve spindle. This was the main reason for deepening the cavity in the steam chest so that I could fit it all in. The outside valves will need looking at as well as LBSC shows the 'drive nut' locked to the valve spindle by an 8BA grubscrew. The problem is that if you follow the drawings, the 8BA thread in the drive nut will only be 1/64" long! That will be far too weak to be effective.


Happy New Year to everyone! Well, that's the end of 2020 and I can't say as I am sorry. It's been pretty crap! Let's hope 2021 brings something better although I think it will be some time yet.

Time to fit the middle cylinder into the frames and drill the mounting holes for it. The first job was to drill the holes in the mounting lugs of the cylinder casting. This was done in the mill using the DRO but as the frame holes would be spotted through from these they didn't need to be that accurate (but it does save marking out!)



These were drilled tapping size for 5BA to take the countersink screws that would hold the cylinder in place but I didn't tap them just yet.

The valve chest was fitted to the cylinder and then the cylinder clamped into the frames roughly in the right position. Incidently, I did have to take a skim off the top of the valve chest as the cylinder assembly was too tight a fit in the frames.

To get the cylinder in the correct place I bolted the outside cylinders back on and put dummy valve rods through all the valve chests. The valve rod for the inside cylinder needs to line up with the outer ones and the top of the inside valve chest needs to be at the same angle as the top of the outer steam chests. To get this right I used a ruler across the outer valve rods (as I did when positioning the stretcher for the 2 to 1 lever) and used that to position the inside valve rod. The inner valve rod is only 0.125" diameter so is smaller than the outer valve rods which means that there should be a 0.01563" gap betwen the top of the inner valve rod and the bottom face of the ruler. That was fairly easy to set using a feeler gauge.

To make sure the angle of the valve rod was the same as the outer rods, I used the digital angle gauge again to compare the angle of the inner steam chest to the angle of the outer ones.



It was a bit fiddly to set up but with a bit of gentle tapping here and there I managed it. A couple of small machinists clamps were used to firmly hold the cylinder in place as I then had to dismantle the frames so that I could set the RH frame up in the mill, complete with the inside cylinder, and spot the holes in the lugs through to the frame plate.



The holes in the frame plate were opened out to 5BA clearance after the cylinder had been removed and then countersunk on the outside face to take the fixing screws. Finally, the holes in the cylinder lugs were tapped 5BA.

I reassembled the frames and fitted the cylinder in position and the mounting holes all lined up perfectly.



The next job will be to fit the rear cylinder cover along with the piston, crosshead and slidebar and trim the piston rod to length. There is also a small square stretcher that supports the end of the slidebar that will no doubt need to be repositioned and the holes for it in the frames moved to suit.


The piston rod was trimmed to length using the same method as for the outside cylinders i.e. the cylinder was fitted with the piston, crosshead and slidebar and then the middle connecting rod installed and the length of the piston rod adjusted so that there was an equal gap at each end of the cylinder. I did have to fettle the slidebar first though. It was bent for a start and also wasn't parallel to the piston rod as the hole for the threaded stub on the slidebar that screws into the cylinder cover boss hadn't been drilled and tapped straight. I had to resort to a bit of brute force and bent the threaded stub until the slidebar lined up properly. This is another feature of a lot of LBSC designs that I don't particulary like - threading the end of the slidebar into the cylinder end cover. Everything needs to be spot on for the slidebar to line up correctly. I much prefer where the slidebar is bolted to a flat on the boss of the end cover as it's position can be adjusted with shims if necessary.

The other end of the slidebar is bolted to a square stretcher to hold it securely and of course, the holes in the frames did not now line up with the stretcher. The original stretcher was only 0.25" square but I decided that if I made a new one of 0.375" square then I would be able to use the existing holes in the frames. I used a bit of square brass as I had some to hand and machined the ends square in the four jaw chuck in the lathe and trimmed it to a nice fit in the frames. It was then held in the correct position with a machinists clamp, the frames were put in the mill vice and the fixing holes spotted through from the frames.




After fitting the stretcher I found that the top of the connecting rod hit the end of the slidebar when the rod was coming up to it's highest point which didn't really surprise me. I don't think anything surprises me after all this time! The original cylinder was at a shallower angle than the new one and the end of the slidebar was higher so it didn't foul before.



There was a chamfer on the end of the slidebar and I could have made that bigger but it would have reduced the thickness of the slidebar where the fixing bolt is threaded in. Instead, I chose to mill some off the top and bottom of the connecting rod to thin it down. It was thicker than shown on the drawings anyway so that was not going to be a problem. I could have just milled some off the top I suppose but that would have made it a bit asymmetrical, not that you would notice.

I made a simple jig to hold the rod in the mill vice with the centreline of the rod at an angle to the x axis of the mill that matched the existing taper on the rod. The jig was just a length of steel bar that I had used before for a similar job with a 5mm bolt through the little end of the rod and the other end held with a 'top hat' shaped bit of bar to locate the big end of the rod and to clamp it down at the same time. One side was milled and then the rod turned over to do the other.



I ended up taking about 0.020" off each side which now left plenty of clearance between the rod and the end of the slidebar.



I should mention that the axle boxes on the driving axle had been set to running height by putting a spacer (a 5BA nut!) inbetween the bottom of the axlebox and the hornstay before I checked the clearance between the rod and the slidebar. If it does hit when the axle boxes are at the top of the horns (unlikely to happen) then I can increase the chamfer on the end of the slidebar a bit if necessary.


Now that the cylinders are nearly finished it's time to have a look at the 2 to 1 levers that drive the middle cylinder valve from the outside ones. Firstly though, I had to countersink the top two fixing holes for the outside cylinders on the inside of the frames. You can't fit bolts or studs here as they would foul the inside cylinder and valve chest so countersink head screws have to be used.



I had to split the frames again to do this but it didn't take long as I had only put a few of the bolts and screws after the last time I had them apart.

You can't use a standard countersink bit for the screws that I buy as the head has a slight parallel part on it before the countersink starts. Well you could use a standard countersink bit but the countersink would finish up much wider than the screw head by the time you had drilled it deep enough for the head to be flush with the surface. Ideally, you want a drill of the same size as the head but with the grind modified to 90° or a 90° spotting drill with the right diameter. I must look and see what sizes are available. The few that I have are all the wrong size, of course.

Instead I use a normal grind drill the same diameter as the screw head to drill to the depth of the parallel part of the head and then take it to depth with the nearest smaller spotting drill. Not ideal but it works.



The cylinders were then bolted in place along with the steam chests. I fitted the old valve spindles just to check the alignment with the 2 to 1 levers which were also fitted temporarily in place.

To my surprise everything seemed to line up pretty well so I was well pleased. I wasn't expecting that!



I didn't bother cleaning up the levers at this stage in case I had to scrap them and make new ones!

The end of the long lever doesn't quite line up with the RH valve spindle vertically but I can adjust that by bending the levers slightly. They may well be bent anyway to line up with the old position of the valve spindle so may just want straightening out again.

The short end of the long lever is nearly touching the top of the clearance slot in the frames so that needs a bit more filing out.



You may notice that the LH end of the short lever does not line up with the LH valve spindle and the RH end of the short lever does not quite line up with the middle valve spindle. The RH end of the long lever does line up with the RH valve spindle. This is as drawn because the levers are not actually the correct length.The overall length of the lever assembly is 0.0625" wider than the distance between the outside valve spindles whereas really it should be the same, otherwise you will get some inequality introduced into the middle valve events. It probably won't make very much difference and I'm not going to bother altering the original levers. LBSC does mention in his construction notes that the dimensions have been 'rounded up' to make the levers easier to make. He also mentions that the excess width of the levers counteracts the angularity of the connecting links but I think that is just an excuse for not making them the right length.

I may replace the pivot pins for the short connecting links as they are a bit small in diameter at 0.09" and may fit bronze bushes in the links to improve the life. Any wear in these links and pins will have quite an effect on the middle valve travel.


I thought it might be a good idea to check the dimensions of the levers before going any further. To do that I put the levers in the mill vice and measured the distance apart of the pivot holes using the DRO. To find the position of the holes I used a pointed centre that I had made for something before.




I repeated this for the short lever. The dimensions for the long lever were within a few thou but the holes in the short lever were a bit more 'diverse'. This was to be expected as these would have been marked out by hand and I doubt I could have done any better using a ruler and a centre punch. I intended to enlarge the end pivot holes anyway so that would be the opportunity to 'adjust' the hole positions using the DRO.

The original pivot for the long lever was a bit of a sloppy fit in the bore of the lever so I made a new one from silver steel. A nice simple turning job.



I did consider fitting a bronze bush to the pivot hole of the long lever but it only rocks to and fro so I think the combination of mild steel on silver steel will be fine.

The pivot holes for the short lever in both the lever itself and the end of the long lever originally had a 0.125" pin through them which was again a bit of a sloppy fit so I drilled and reamed them out to 4mm which was the next size up in silver steel that I had. I also drilled and reamed the ends of the levers to 0.125" to increase the bearing surface for the link pins.



I made a new pivot pin for the short lever from silver steel and drilled it to take a 7BA bolt to retain it in position (my favourate method of making valve gear pins). I'm halfway through making some 0.125" pins for the links that connect the ends of the levers to the valve spindles.


< Back to Page 4. .....Page 6 >