Stop Valves (February 2015)
There are three screw down stops valves required for the Allchin, one to controller the blower in front of the steam chest and two in the cab area to control the water lifter and the injector. Each of the valves required nine separate parts to be machined, namely a valve body, an inlet and outlet stub, an inlet and outlet cap, a bonnet, a bonnet cap, a spindle and a hand wheel, plus a silver soldering operation to join the inlet and outlet stubs to the body. The valve body is only 3/8” (9.5 mm) in diameter and the hand wheels are secured by a 12 BA thread, so all and all the manufacture of these will test any builders attention to detail and patience!
The first decision was what material to make then from, the drawings stated that all the body parts should be made from Bronze. This was the first time I had really had to identify bronze as the I had only used it once before and that was on the front hubs, where the material came as castings so the choice was essentially done for me. Following some investigation I narrowed the choice down to two types of bronze which appeared to be readily available, namely leaded bronze or phosphor bronze or LG2. I bought a length of ½” leaded bronze following some discussion with the store keeper and on the advice that it would be ideal for the job mainly on the fact that it would machine well. Once I had my new piece of material in hand I then set about sorting my stock of materials at home and as it turned out I actually had a 300mm length of 3/8” dia bronze which was better suited to the job!
There are three screw down stops valves required for the Allchin, one to controller the blower in front of the steam chest and two in the cab area to control the water lifter and the injector. Each of the valves required nine separate parts to be machined, namely a valve body, an inlet and outlet stub, an inlet and outlet cap, a bonnet, a bonnet cap, a spindle and a hand wheel, plus a silver soldering operation to join the inlet and outlet stubs to the body. The valve body is only 3/8” (9.5 mm) in diameter and the hand wheels are secured by a 12 BA thread, so all and all the manufacture of these will test any builders attention to detail and patience!
The first decision was what material to make then from, the drawings stated that all the body parts should be made from Bronze. This was the first time I had really had to identify bronze as the I had only used it once before and that was on the front hubs, where the material came as castings so the choice was essentially done for me. Following some investigation I narrowed the choice down to two types of bronze which appeared to be readily available, namely leaded bronze or phosphor bronze or LG2. I bought a length of ½” leaded bronze following some discussion with the store keeper and on the advice that it would be ideal for the job mainly on the fact that it would machine well. Once I had my new piece of material in hand I then set about sorting my stock of materials at home and as it turned out I actually had a 300mm length of 3/8” dia bronze which was better suited to the job!
Valve body (three off)
The body required a ball to be turned on the end of the rod. This posed a problem as I didn’t have a ball turning attachment for the lathe. I considered profiling it on the mill using a ball nosed bit, but concluded that it was never going to produce a complete finish. Thus, I decided that I would make a 3/8” dia hole gauge and use a file to hand shape the ball on the lathe.
The gauge was made by drilling a scrap piece of gauge plate and reaming to 3/8” then cutting away the material to leave about a 110 deg portion of hole exposed.
The bronze rod was chucked in the lathe and the with right hand knife tool the top cylindrical section was turned to dia to the point where the cylinder transitions into the ball. (Fairly simple trigonometry) can be used off the drawings to find this point. Then the tool was moved so the cutting end was at the centre point of the ball. The tool was then used as a guide to round the lower hemispherical ball section with a file. The gauged was used frequently and to indicate the high spots and the material removed with ever decreasing grades of file. The same process was continued on the upper hemisphere, taking care not to damage the transition in to the cylinder.
Next the inlet and outlet stub recesses were milled into the apposing ‘faces’ of the ball. I found the trick here was to support the rod in a vice and clamp the end of the rod with a tap wrench, upon which could be mounted an angle gauge to ensure that faces were 180 degs apart. I did try to complete this operation on with the rod clamped in the 4th axis on the mill. However, I found the rod was not stiff enough to withstand the tool pressure due to distance it needed to be from the chuck jaws to allow the milling cutter chuck to clear the 4th axis chuck. The centre holes for the steam ways were also completed at this stage.
Next the rod was positioned in the vice 30deg from vertical to complete the steam ways through the body. To be fair a lot of this was done by eye as it was near on impossible to accurately measure the direction or depth of the holes!
The body was then parted off the rod and returned to the mill. 7BA washers were stacked two high and in each of the inlet and outlet recesses to clamp the body on flat surfaces and the body was orientated vertically. The central steam way, bonnet thread and ‘seat’ was completed during this operation.
The body required a ball to be turned on the end of the rod. This posed a problem as I didn’t have a ball turning attachment for the lathe. I considered profiling it on the mill using a ball nosed bit, but concluded that it was never going to produce a complete finish. Thus, I decided that I would make a 3/8” dia hole gauge and use a file to hand shape the ball on the lathe.
The gauge was made by drilling a scrap piece of gauge plate and reaming to 3/8” then cutting away the material to leave about a 110 deg portion of hole exposed.
The bronze rod was chucked in the lathe and the with right hand knife tool the top cylindrical section was turned to dia to the point where the cylinder transitions into the ball. (Fairly simple trigonometry) can be used off the drawings to find this point. Then the tool was moved so the cutting end was at the centre point of the ball. The tool was then used as a guide to round the lower hemispherical ball section with a file. The gauged was used frequently and to indicate the high spots and the material removed with ever decreasing grades of file. The same process was continued on the upper hemisphere, taking care not to damage the transition in to the cylinder.
Next the inlet and outlet stub recesses were milled into the apposing ‘faces’ of the ball. I found the trick here was to support the rod in a vice and clamp the end of the rod with a tap wrench, upon which could be mounted an angle gauge to ensure that faces were 180 degs apart. I did try to complete this operation on with the rod clamped in the 4th axis on the mill. However, I found the rod was not stiff enough to withstand the tool pressure due to distance it needed to be from the chuck jaws to allow the milling cutter chuck to clear the 4th axis chuck. The centre holes for the steam ways were also completed at this stage.
Next the rod was positioned in the vice 30deg from vertical to complete the steam ways through the body. To be fair a lot of this was done by eye as it was near on impossible to accurately measure the direction or depth of the holes!
The body was then parted off the rod and returned to the mill. 7BA washers were stacked two high and in each of the inlet and outlet recesses to clamp the body on flat surfaces and the body was orientated vertically. The central steam way, bonnet thread and ‘seat’ was completed during this operation.
Bonnet and Bonnet Caps (three off)
The bonnet, secures the top chamber of body and provides a threaded hole for the valve spindle to run in and thus retains the thrust necessary to close the valve. The key point which had to be thought about when making this part was concentricity as the valve would not seal if the spindle was not set inline with the seat. Thus, the order in which the part is to be made was important and a jig was necessary to support the part mid manufacture.
The bonnet was also the first part to require a hexagon. Which was easily achieved using the 4th axis on the milling machine. To cut this feature a fairly simple piece of G Code was created, which holds the CNC control point (centre of the cutter) at a fixed point on the X Axis and makes a series of cuts along the Y axis at 0, 60, 120, 180, 270 and 300 degs in turn, at a sequentially lower values of Z.
To initialise this code two flats were cut on to the rod at 0 and 180 degs at a known Z position, which is greater than the finished dimension. Then the actual part was measured across the flats and the finish Z position calculated.
The calculated data points could then be written into the G Code program using a series of values that are called up in the main body of the program, hence the same code could be used again simply by changing the values. Once complete the mill was set in action to result in a Hexagon on the shaft.
The bonnet, secures the top chamber of body and provides a threaded hole for the valve spindle to run in and thus retains the thrust necessary to close the valve. The key point which had to be thought about when making this part was concentricity as the valve would not seal if the spindle was not set inline with the seat. Thus, the order in which the part is to be made was important and a jig was necessary to support the part mid manufacture.
The bonnet was also the first part to require a hexagon. Which was easily achieved using the 4th axis on the milling machine. To cut this feature a fairly simple piece of G Code was created, which holds the CNC control point (centre of the cutter) at a fixed point on the X Axis and makes a series of cuts along the Y axis at 0, 60, 120, 180, 270 and 300 degs in turn, at a sequentially lower values of Z.
To initialise this code two flats were cut on to the rod at 0 and 180 degs at a known Z position, which is greater than the finished dimension. Then the actual part was measured across the flats and the finish Z position calculated.
The calculated data points could then be written into the G Code program using a series of values that are called up in the main body of the program, hence the same code could be used again simply by changing the values. Once complete the mill was set in action to result in a Hexagon on the shaft.
Once the hexagon was cut the shaft was set in the lathe and the lower thread, i.e. the thread that mates with the bonnet was cut and the internal 5 BA thread was drilled and tapped. This way the spindle thread will be as in line as possible to body thread and hence, when fitted, to the central steam way in the body. At this point all three bonnets were part made.
The next operation was to make a jig to support the part made bonnet. This consisted of a brass rod, chucked in the lathe, drilled and taped to suit the bonnet thread and faced off square. The bonnets were then screwed into the jig and the top features finished. The jig remained into the chuck to preserve concentricity throughout.
The caps required another hexagon profile to be milled, which I did individually as I found the small effective diameter of the rod flexed too easily thus preventing machining too far from the 4th axis chuck and thus making it impractical to machine one continuous hexagon and the caps parted from it. As it turned out the mill was sufficiently reliable that it could be set to run and left to cut another hexagon with no requirement for further supervision, allowing me to operate the lathe in parallel.
I did have to make a decision when making the caps as the internal thread was required to be formed ‘right’ to the bottom of a flat bottomed hole. Thus, I had to effectively grind one of the vintage ME taps to a very tight plug tap profile.
Inlet and Outlets (six off)
Four of these were identical turning operations and a bit of fitting to ensure that engaged nicely in the body recesses, plus a cap for each which followed the same lines as the bonnet caps. The two inlets from the steam union, were different as they are were female fittings with a outer hexagon profile.
The next operation was to make a jig to support the part made bonnet. This consisted of a brass rod, chucked in the lathe, drilled and taped to suit the bonnet thread and faced off square. The bonnets were then screwed into the jig and the top features finished. The jig remained into the chuck to preserve concentricity throughout.
The caps required another hexagon profile to be milled, which I did individually as I found the small effective diameter of the rod flexed too easily thus preventing machining too far from the 4th axis chuck and thus making it impractical to machine one continuous hexagon and the caps parted from it. As it turned out the mill was sufficiently reliable that it could be set to run and left to cut another hexagon with no requirement for further supervision, allowing me to operate the lathe in parallel.
I did have to make a decision when making the caps as the internal thread was required to be formed ‘right’ to the bottom of a flat bottomed hole. Thus, I had to effectively grind one of the vintage ME taps to a very tight plug tap profile.
Inlet and Outlets (six off)
Four of these were identical turning operations and a bit of fitting to ensure that engaged nicely in the body recesses, plus a cap for each which followed the same lines as the bonnet caps. The two inlets from the steam union, were different as they are were female fittings with a outer hexagon profile.
Silver Soldering
The assembly of the valve body required the inlet and outlet to be silver soldered into the recesses in body. Beyond a small plug on the Steam Union, this was the first time I had to complete a silver soldered joint and thus I need to learn a new skill! To practice I turned up a test piece out of a couple of pieces of scrap bronze. The test piece had reasonably similar dimensions to the actual job to make the test as realistic as possible.
To complete the joint I firstly cleaned the part with Loctite Super Clean and then dropped a small drop of flux paste into the recess then rolled the part to be inserted in the drop of paste so all internal surfaces were covered. The two parts were then gently pressed together and loosely set in jig to hold the assembled test piece when heating. A piece of 0.5 mm 55% silver solder was then formed into a ring which would sit nicely on top of the joint and the ring held in place with some carefully placed flux paste. The jig was then sat in a hearth of fire bricks and propane torch used to apply the heat. As expected the flux went glassy and just before the part turned red, the solder melted and formed the joint. At this point the heat was maintained and the part gently turned to ensure the solder had completed the ring and once content the heat was removed and the part allowed to cool.
The result was an effective joint which when de-scaled look mechanically sound. At some point in the future I plan to section to the test piece to see how much penetration was achieved.
De-scaling
I didn't have any citric acid lying around, however, I did have some Kilrock Big K de-scaling gel under the sink. So I covered the test piece in the gel and let it for approx 20 mins and found that the scale could easily be removed with a brass wire brush, so that was the technique I used on the completed valve bodies.
The assembly of the valve body required the inlet and outlet to be silver soldered into the recesses in body. Beyond a small plug on the Steam Union, this was the first time I had to complete a silver soldered joint and thus I need to learn a new skill! To practice I turned up a test piece out of a couple of pieces of scrap bronze. The test piece had reasonably similar dimensions to the actual job to make the test as realistic as possible.
To complete the joint I firstly cleaned the part with Loctite Super Clean and then dropped a small drop of flux paste into the recess then rolled the part to be inserted in the drop of paste so all internal surfaces were covered. The two parts were then gently pressed together and loosely set in jig to hold the assembled test piece when heating. A piece of 0.5 mm 55% silver solder was then formed into a ring which would sit nicely on top of the joint and the ring held in place with some carefully placed flux paste. The jig was then sat in a hearth of fire bricks and propane torch used to apply the heat. As expected the flux went glassy and just before the part turned red, the solder melted and formed the joint. At this point the heat was maintained and the part gently turned to ensure the solder had completed the ring and once content the heat was removed and the part allowed to cool.
The result was an effective joint which when de-scaled look mechanically sound. At some point in the future I plan to section to the test piece to see how much penetration was achieved.
De-scaling
I didn't have any citric acid lying around, however, I did have some Kilrock Big K de-scaling gel under the sink. So I covered the test piece in the gel and let it for approx 20 mins and found that the scale could easily be removed with a brass wire brush, so that was the technique I used on the completed valve bodies.
The same process was used do each joint on the valves in turn. There was a concern that when doing the second joint on the valve, the first joint may well fail, however, I either got away with it or capillary action retained the solder in the joint as it appeared to work fine.
Finished Assembly with Steam manifold and union
Finally the machining and soldering operations for the valves were completed and assembly could start. It was found that the second valve bonnet had to be fitted whilst the valve was in position on the union as there was insufficient clearance to screw the completed valve in place.
The next stage will be to make the water gauge fittings and then the steam manifold will essentially be tight, so I'll be able to make up a pressure test connection and test the various valve and connections with an airline.
Finally the machining and soldering operations for the valves were completed and assembly could start. It was found that the second valve bonnet had to be fitted whilst the valve was in position on the union as there was insufficient clearance to screw the completed valve in place.
The next stage will be to make the water gauge fittings and then the steam manifold will essentially be tight, so I'll be able to make up a pressure test connection and test the various valve and connections with an airline.