Member postings for Andrew Johnston

That's because speed control on a single phase motor doesn't work well. You need to change the single phase motor to a 3-phase motor and then fit one of the many single phase to 3-phase VFDs.

Andrew

There's a lot to learn; you can either do it the hard way or the easy way. The choice is yours.

Threads along the lines of "getting a poor finish turning/milling" crop up regularly. In many cases the problem is one, or both, of the following:

Scrap/cheap material of unknown quality and properties

Cheap cutting tools

Andrew

I wonder if my grinder might be different? It simply says Brown and Sharp and No.2 on the plate. Here are the basic table controls:

To start the table feed the knob labelled Start is turned a few degrees clockwise and then let go when it springs back to the original position. This engages a pin held in the horizontal casting between the Start and Stop controls. As the Start knob is rotated the Stop lever also moves slightly but also returns to it's normal position. To stop the table feed the Stop lever is knocked anti-clockwise; it only moves a few degrees before disabling the feed. Once the table feed is active the table direction can be reversed manually by moving the Manual Reverse Lever from side to side. This picture shows the table reverse stops and reverse dog in the manual position:

For auto table reverse the reverse dog is moved to the vertical:

It's important that the reverse dog is between the stops; it can get exciting if it isn't! Standard table speed is 33ft/min and there's a fair old clonk, and the machine shakes, at each reverse.

Andrew

More down to Geomagic. On my old computer I'm still using a 2014 version of Geomagic. In desperation I created a new part on my new computer which which runs Alibre 2019, and imported that into the CAM program. It still had the strange artifacts on the model, but circles and edges could be selected, unlike before.

One of these days I'll move all the Burrell stuff over to the new computer.

Andrew

Edit: after I'd created the new part Alibre offered me an update. Big mistake, after a hiccup in the process it "helpfully" wiped Alibre from my computer. With some help from the UK agent I eventually got it all back again, including my custom text file that lets me explicitiy use English threads, such as BSF, BSP, BSW and BA in models.

Edited By Andrew Johnston on 22/09/2019 11:39:24

The problem with used material of unknown provenance is that you don't know anything about it. Could be easy to machine, or be almost impossible to machine on a hobby lathe. Unless you're a masochist i'd advise against scrap material.

I tend to buy steel up to about an inch or two diameter from commercial stockholders in standard 3m or 6m lengths. Proportionally it's much cheaper than buying from the hobby market and it doesn't take long to build up stock. For aluminium (I mean alloy for the pedants) I used to buy offcuts by the kilo from the bin at the commercial stockholder I was using for work. That's now dried up so I tend to buy what I need from Ebay.

I recycled most of my scrap material years ago, as it was in the way. It's better to buy what you need and of known properties.

Andrew

FInally finished the oilers for my traction engine slidebars:

Both the slidebars, machined from gauge plate, and the oilers have been redesigned to accord with pictures of a fullsize engine. The oilers are intended to be proper wick feed; time will tell if they work.

All external threads (32tpi and 40tpi ME) were screwcut; internal threads used taps. One would have thought that these were all simple parts - don't you believe it! I bought a set of 7/16" x 32 taps for an internal thread. I'm pretty sure the taps are screwcut carbon steel. The bottom tap takes out a few thou more compared to the taper tap. There is a model engineering saw that says this, although I've never seen it stated in the literature. Also never seen it with ground HSS taps. The thread size was off too. For the mating external thread, starting at nominal diameter, the internal thread was a rattling good fit nearly 5 thou before the theoretical thread depth of 20 thou was reached. Never seen that before either, normally have to go full thread depth plus a couple of thou to get a nice fit.

Given how simple it should be I had a lot of trouble with the elliptical base. Easy enough to draw in 3D CAD. But for some reason on import into my CAM program it fudged the model making it difficult to select features and causing my CAM program to fall over ever time I tried to generate a toolpath. I also discovered that the hole pocketing routine in my CAM program doesn't work. It claimed to generate a helical ramp down and then a clean up. But backplotting the G-code just gave a straight plunge, and PathPilot barfed at the code with all sorts of circular plane errors. Once I got working G-code it only took 50 seconds to pocket the central hole and profile the outside.

All I can say is I hope future parts go rather better, or I'm going to working on the engines for a loooooong time.

Andrew

Edited By Andrew Johnston on 21/09/2019 23:43:18

That's why I suggested filing a plate to fit and measuring that. Alternatively do the calculations for the width of the top of the tooth for 14.5 and 20 degres pressure angles. The difference is around 45 thou, that should be easy to see.

It won't be any time soon, but I could lift the table on my B&S No.2 and have a look at the rack in due course.

Andrew

From the link provided by MG it appears that the Welin breech is indeed a screw thread on a helix. Since it is a relatively fine thread and a large diameter the helix angle is small and thus difficult to see. For instance a 1/2" pitch thread of 15" diameter has a helix angle of only 0.6 degrees. I'd been worried about saying the thread was parallel. One of the functions of the breech is to provide a gas tight seal and I didn't see how it could do that without a small axial movement as a thread was tightened.

I've been reading a thread on "practicalmachinist" and I think SoD has it. The consensus was that the thread was single pointed in the conventional way, but with the addition of a cam on the cross slide that stepped it between three pre-determined positions as the work rotated. The slots between threads are to allow the cross slide time to move.

I'm almost tempted to have a go at modelling and making one just for the hell of it. Although I'd use the rotary table on the CNC mill as a sort of vertical lathe rather than mess about with cams.

Andrew

Even better would be to file a piece of sheet material to be a snug fit in the tooth gap. Then measure the included angle. The difference between 29 degrees and 40 degrees should be clear.

Andrew

Edited By Andrew Johnston on 20/09/2019 22:41:58

During my cycle ride this afternoon a couple of piston engine (radial) fighters flew over in close formation. The wings were clipped and had a distinctive curve on the trailing edge. Both aircraft also had invasion stripes. Not sure what they were; Sea Furies or Thunderbolts?

Andrew

Darn it, just wasted 20 minutes looking at Youtube. interestingly, as far as I can see the threads are parallel, not helical. Or at least the helix angle is very small. I'd always assumed they were helical.

Modelling it shouldn't be too difficult, athough everyone has missed out the grooves between the thread segments. A more interesting question is how one would machine the part. Hint: I suspect the grooves are there not for functionality but to aid machining.

Andrew

You don't, but he who dies with most toys wins!

I've got a 3' by 2' cast iron surface plate; £40 on Ebay. While I don't mark out that often the surface plate is useful when I do. The milling machine tables are too small, they've got T-slots, have stuff bolted to them and I'd need to move everything on and off each time. A surface plate is also useful when taking measurements of anything from heights to tapers.

I've never measured, or worried about, the edges of the surface plate being square, or even straight.

Andrew

Although convenient the diagram above is slightly misleading as the pressure angle is not referenced to the horizontal. As correctly described it's the angle between the normal at the contact point and the tangent to the pitch circles at the intersection of the centre to centre line of the gears. The important point to note is that as the teeth roll past each other the pressure angle stays constant and hence the angular velocity is constant.

For a gear with a small number of teeth during the rolling action the tip of one tooth does not clear all the sides of the opposing space. So the bottom of the tooth profile has to be cut away to give clearance. Although a bevel gear these 10 tooth pinions illustrate undercutting:

Look at the top tooth on the left pinion. From the top of the tooth the curve each side comes down in a smooth manner until about half way down the tooth. At this point there is a sudden change of direction and the width of the tooth gets narrower rather than continuing to get wider. This is undercutting. If you think about it the undercut cannot be formed using a rotary cutter cutting one space at a time. Undercutting is not inherent to involute gears, but is a solution to interference problems with small numbers of teeth. The need for it is less with higher pressure angles. As mentioned there are also other ways of reducing the need for it by increasing the OD or tweaking other parameters. If a spur gear is hobbed with a standard hob you will get undercutting for small numbers of teeth, whether you want it or not.

Since the bevel pinions as designed are undercut they were cut on a 4-axis CNC mill to avoid the issues of using involute cutters:

Andrew

Edited By Andrew Johnston on 19/09/2019 18:57:19

In theory the 11 tooth gear will need to be undercut, in which case it cannot be cut one tooth at a time with an involute cutter. The need for undercutting can be overcome by increasing the OD and/or pressure angle. But changing the OD will change the centre to centre distance, which has serious ramifications for the rest of the design. I'd increase the 11 tooth gear to 12 teeth, and reduce the tooth count on the mating gear by one. That way the centre to centre distance will stay the same, with a slight alteration in the ratio.

I think the RDG involute cutters are 14.5PA, which is prototypical, albeit rather out dated. I'd be wary of the RDG cutters. I bought a selection when I was mucking about experimenting with helical gears. They did the job, but eccentricity and wobble weren't always great.

Andrew

I use centre drills for making centres, and for starting drills, on the lathe.

But I never use them on the vertical mill. For general work 4-facet drills start fine, on a flat surface, without the need for spotting or centre pops. If I need better location accuracy, less than 2 thou, I'll spot drill first. The trouble with centre drills is that it's only too easy to break the tip on a vertical mill. And they're not much use for small drills as the tip is bigger than the drill.

Centre pops are only used when I'm drilling by hand (mostly sheet metal) or for rough 'n' ready holes using the drill press.

I expect that in industry spot drills are not held in drill chucks, but collet chucks or similar where the length can be accomodated.

Andrew

How far off is it? Is the starting face perpendicular to the drill? In general:

1. Don't use the drill press, especially if it's a cheap one. Use a lathe or vertical mill and machine vice instead.

2. What is the provenence of the drill, cheap or branded?

Andrew

Screwcutting is simple, but it takes a modeller to over-complicate it.

For normal threads I screwcut using full form inserts and plunge cutting, so that the cross slide reads total DOC directly. A lot of my threads are screwcut to a shoulder and using the topslide at an angle would make it more involved to set the axial finish point. With the plunge method it is simple to set the clearance to a couple of thou once before cutting the thread.

Screwcutting on a CNC lathe often uses the G76 macro. It is untrue to say it cuts down the flank as a default. Annoyingly each manufacturer has a slightly different syntax, but all contain a parameter that controls the infeed angle, a value of zero usually represents radial infeed, ie, plunge cutting. So the angle at which the tool infeeds is set by the operator. I understand that newer controls have an option for infeeding on both flanks, alternately one side and then the other.

I should add a caveat for manual screwcutting imperial threads on my imperial lathe. I have an Ainjest high speed threading unit which automatically trips at the end of each pass. The trip point is pretty consistent, so it makes it easy to screwcut rignt up to a shoulder, internal or external.

Andrew

Although bronze, and gunmetal, can snatch like brass, generally I don't have a problem with drills less than 6mm or so. However, I run industrial machines. Even if the smaller drills do snatch they're not capable of pulling the tailstock along for instance. Larger drills are a different story; they can simply pull the Morse taper out. One can go up in small, say 1mm increments, but that can get tedious. I have used the slot drill trick, sometimes it works well and sometimes not; it depends upon the helix angle of the cutter.

Stoning the cutting edge has never worked for me; possibly I'm too aggresive on feedrates. Instead I've bought a small selection of slow helix drills from about 1/4" to 1/2". Above that I prefer to use a boring bar.

I haven't really noticed bronze closing up after drilling, but have definitely seen it when reaming. After reaming bronze (with a machine reamer) the reamer will not re- fit in the hole by hand. When I was grinding the rams for the water pumps on my traction engines, to fit the existing reamed bore, I had to go about 3 tenths undersize to get them to fit. So either my 17mm reamer is fudged or the gunmetal casting had closed slightly.

Assuming the thread is M3 a 2.5mm drill is slightly on the small size. I'd be using 2.7mm on bronze.

Andrew

Edited By Andrew Johnston on 15/09/2019 21:18:51

Any carbide milling cutter will work fine in HSS. Here's an involute shape milled on the end of a 1/4" HSS toolbit prior to using it to cut an internal gear:

The cutter is a 3 flute uncoated carbide slotdrill. Similar to turning hardened steel use reasonably high surface speeds, shallow DOC and high feedrate. Ideally the shear zone should be red hot as that will soften the steel where it is being cut, but leave the bulk of the material hard.

Andrew

That's interesting; I've never seen spiral flute taps described as anything other than spiral flute. Definitely no options have been listed. I've always tapped both through and blind holes with one tap only. Where are the different types of spiral flute tap advertised?

Andrew