Re: Whistling Billy Replica
Posted by: Steamcarbob
Date: April 13, 2016 07:27AM
I have just finished the job of making flowmotor needles. I am sorry if this is a bit of difficult detail but many White owners I am sure have struggled with these needles. This is my way of making them in my workshop allowing for a 75% failure rate in this first batch for several years.
I started with twelve rods of 3/32inch silver steel which was 0.09364 inch on my electronic micrometer.
The first job after cutting to length was to thread the ends, one long threaded end for the fixing and adjustment and the other end about 6 threads for the head.
I made the first three heads out of silver steel but found that on drilling it, the very small drill was running off center and tapping was difficult with the tap fragile. I then decided to use brass for the heads which made the job much simpler as it drills, threads and gets its final skim more easily on these tiny sizes.
The heads I shaped with the donor rod in lathe collet with the taper and I made them longer than the drawings (attached to last entry). After the rods were screwed into the heads, I heated them and put a tiny blob of silver solder on the top of each.
I then ground an approximate square on the end of the head so that it fitted an old clock key that I have (my improvement on the White design!). This is used when having to rotate the needle 180 degrees for milling the flats to adjust the number of turns and when it is screwed into its mounting rod (one takes up the tension of the spring and adds exactly 5 turns). It makes life easier and one does not need to grip and scratch or bend the shaft with plyers.
Next, I mounted the rod back into the lathe collet and skimmed the angle off again as it is never exactly true at his stage but since it is only mounted on the thin rod, I only want to be taking a very light skim at high speed.
Incidentally, by now my clumsiness had reduced the 12 rods that I started with to 8 as I have bent two and broken the end off two, one when tapping the silver steel thread and one with machining the silver steel head. The silver steel heads had been hardened when silver soldering but of course are now softened using brass making the final machining much easier.
The next job was to reduce the main rod body diameter by 0.0016 inch which is done when held in a collet and I ran the valve head in a well-oiled brass tube held in a chuck in the lathe tailstock just covering blue tape so that I left it alone. This strip of blue insulating tape covered the ¼ inch next to the head that I needed to leave at the original diameter. I took the rod diameter down with fine emery paper and an electronic micrometer which must be clean as working at this level, any dirt registers. It took about 5 minutes per rod to get to the desired diameter ( 0.0920 inch ) using 3 grades of emery.
My next job was to cut the flats which was easiest done for me in the milling machine with the mounting jig looking at me rather than flat as I can use the digital readout on the X and Y axis to get the slots right and use the calibration on the machine vice for the angled face. I was cutting with the side of the mill.
I did several dummy runs on the rods that I had already damaged to get the exact depth of cut, using the digital read out on the milling machine as a guide but my micrometer as the accurate result.
Having cut the first parallel cut (losing two more needles as the cut still went too deep), I then rotated the vice and set the angle for the tapered cut, again not believing the vice angle but using it as a guide. I then cut the tapered flat.
Unfortunately, I got it wrong and messed up three more needles before my quality control picked up the fault. I had cut the flat too long using the measurements on the drawing and not allowing for the width of the milling cutter (1/8th inch). I may be able to fill the gap a little with soft solder and use these last five needles on Billy as it is a bit hit and miss on the amount of fuel that this car needs, one of them could be just right! This car has a 20hp engine and 30hp steam generator and fire and I have been using 30hp needles but have had some overheating although when used in anger the fuel needs to be full on.
The next job was to take a thin skim off the thread, and the end of the rod just by it, with fine emery to allow the bronze bush to pass over it cleanly. The thread of course must be absolutely clean before going through this bush. The bronze bush must be a very good fit up near the head to low no fuel through (or air when shucked on it). It needs to be clean and slightly tight as it is pulled on and moved about when well-oiled to make a free fit.
I finished off by making 6 new copper rods with brass heads which the needles screw into using some 1/8th inch fine thick walled tube that I found in my metal reserves. These rods were changed as service parts with the flowmotor needles presumably as the threads needed to be good and the shafts straight and not scratched going through the small packing glands.
Q.. What is the result of my three days of fine graft?
A.. Three good 20hp needles and possibly five that can be adjusted for Billy. At least the latter have good tops and threads and should shut off well which is essential on Billy to prevent blowbacks.
I found in the past that hand filing the needles took a day for each needle, so scores are about even but I think that if I proceeded to do more the way that I have done these I should now get most of my dozen as good quality needles having worked out a lot of the errors.
I am sure that there is someone somewhere who would make them at a good price but some I purchased from the USA a number of years ago were useless as the threads were not good enough.
Anyway it is very satisfying to tackle a difficult problem. The White Sewing Machine company I am sure would have found them a doddle as this is sewing needle type technology.
My flowmotor kit.jpg (249kB)
Needles mainly ready for the flats.jpg (202kB)
Machining the head. The brass in the tailstock protects the head for thinning the body.jpg (154kB)