Member postings for Andrew Johnston

Use a disc micrometer.

The underlying maths is simple. The disc micrometer will measure the distance across the teeth, but not on the pitch circle diameter. By varying the number of teeth the exact point of measurement will vary. The thickness of a tooth, away from the pitch circle diameter, is determined by the pressure angle. So by comparing the results to the theoretical values the pressure angle can be determined.

Andrew

I have a clone angle gauge - works fine and the battery lasts for ages. I've had to repair it once to fix the 'click' mechanism that angles the display. It's ok for rough work, like bending the spokes for my traction engines or setting the angle plate for basic work. If I need better accuracy I've got a Taylor Hobson clinometer.

I've just finished machining the valve chests on my traction engine cylinders. The cylinder was bolted to an angle plate, but there is no reference surface to set the valve chest face parallel to the table. I used the digital angle gauge to set zero on the face, relative to the table, and then indicated in X and Y with a DTI to eliminate the last few thou of error.

Andrew

Both BSP and BSF threads are Whitworth, so 55 degrees.

Andrew

Acetal machines beautifully; use low speeds and high feedrates or it'll melt. What do you mean by accurate, 1mm, 0,1, 0.01mm?

Andrew

I've never cut an external NPT thread but have tapped them. As you'd expect the tap quickly gets more difficult, followed by impossible, to turn. The tapping drill tends to be slightly larger than one might expect. I expect a die would be similar. I'm not sure if a taper on an external item would work. The die will go on further before it cuts, but I suspect it'll come to a halt more quickly. So possibly not much gain.

This is how I cut external tapered threads:

I was screwcutting BSPT threads but I think they're the same taper as NPT. I thought NPT threads were specified by a nominal imperial dimension representing the pipe size rather than tpi?

Andrew

Osborne's maneuver as detailed in "The Machinist's Second Bedside Reader" by Guy Lautard. Or just pick up on each side of the circle in X and then Y and use the 1/2 function on the DRO at each step to get the centre. Once at the centre a sanity check is to touch off plus and minus X and Y; the readings should be the same. Or, for the lazy, you can use a co-axial indiccator:

Andrew

Glider trailer!

I can get a 3m length in the car without a problem. The 6m lengths of aluminium were always delivered. For 6m lengths of hot rolled steel it depends on size. If I can lift the 6m length then I have it delivered. If I can't then I'll ask for it to be sawn into thirds or quarters so I then have the option of delivery or collection. Same goes for plate. When I ordered a 2m x 1m sheet of 6mm steel I asked the supplier to guillotine it in half.

Andrew

Been working on the cylinder castings. The regulator recesses are done:

Also the valve and regulator packing gland locations and the bores started:

Both bores are undersize by about an 1/8" and only go an inch deep. In due course I'll use the machined part of the bores to pick up the locations on the horizontal mill and use the Wohlhaupter boring head to bring the bores to size. The Bridgeport quill movement isn't big enough to machine the complete bore.

I know the centre to centre distance of the bores from the drawings and my 3D CAD model. As a sanity check I measured the bores as they are now and also measured the width of the web between the bores. Adding half the measured bore diameters to the measured web agrees with the theoretical centre to centre distance to 2 tenths of a thou. I don't believe that for a moment! It's not easy to get an accurate reading with an internal micrometer and likewise the vernier calipers used to measure the web. However, if I'm within a couple of thou I'll be happy.

Andrew

Simples!

Turn the spigot to the wanted diameter and length. If I'm going to be facing I leave the spigot slightly short of the work. Drill and tap the hole. Ideally use a taper tap and only go in a few turns. You only need a couple of full threads and then the thread needs to taper off. So when the screw is inserted it binds and forces the spigot apart. Slit the spigot with a hacksaw. Clean up the saw lines with a needle file. Job done.

Andrew

I use Parkersteel for basic steels. Although they have an outlet near Cambridge the head office is in Canterbury. However, they have got a bit silly on delivery charges recently.

Andrew

If buying from commercial stockholders it's often cheaper to buy the standard length; 3m for steel and 6m for alumiumium. For small sections, less than ~25mm diameter, that's what I do. Over time you build up a good stock of basic steel and alumimium.

Andrew

I bought my first batch of steel from Mackay's after getting my lathe and mill - never again. I had all sorts of problems getting a consistent finish when turning. Eventually I asked a commercial machine shop (that I was using for work) where they bought their metal. They told me and commented they didn't use Mackay's as their steel was carp. I bought from their recommendation and immediately got a lovely turned finish.

I buy from a variety of sources, but mostly Parkersteel and M-Machine.

Andrew

A split bush is easier, and quicker, to make than a tapered arbor for holding work:

The split bush shown was used to face the gland to length, turn the boss to diameter and face the back of the flange. Width of the flange and height of the boss are parallel to better than 0.02mm. A split boss can also provide a high clamping force. Here's one being used to cut a 6DP gear in EN24:

Andrew

RTFP!

Which is precisely the point I was making.

Andrew

Out of idle curiosity I thought I'd measure the variation in width on my two recently squared up cylinders. Taking three points on each, one has a maximum variation of 0.1mm and the other is 0.05mm. Of course the variation doesn't matter. What is critical is that the bore is perpendicular to the face that controls piston alignment. What the other face, which is just a cover, does isn't important.

Andrew

It would be interesting to know if anyone has actually measured an older model to see how accurate it is, as opposed to parts made to fit.

Andrew

Done.

Andrew

Depends on how you hold the cutter. In a Clarkson lookalike chuck on my CNC mill I reckon better than a thou. When the Tormach CNC mill came out there was some work done on this (on CNCZone I think) which came to the same conclusion. In a R8 finger collet I've no idea as I rarely use them. But since they're not dead length collets I suspect that repeatability would be poor.

Andrew

I've bought from them in the past, but no longer do so. Make of that what you will.

Why two DTIs? I only use one and just run it down the back of the fixed jaw tapping the vice until there is no movement as the table moves.

Andrew