Member postings for Andrew Johnston

I did things the other way round; I decided which engine to build and then bought a lathe to suit. I am building two 4" scale Burrell SCC engines. I bought a Harrison M300 lathe as it had the between centres length to turn the axles (~30" ) and turn the final drive gears and flywheel (~16" diameter) in the gap, like this:

However, those are the only parts for which I needed the gap. I would agree with Paul, the choice of mill is far more important. I have three; a Bridgeport, a large universal horizontal and CNC. While most of the work "could" be done on the Bridgeport the horizontal is excellent for cutting gears and acting as a poor mans horizontal borer. Even then some ingenuity is needed:

The CNC mill allows me to make parts i can't make on the other mills, like true bevel gears, worms and special cutters:

To summarise, by farming out a few parts the Logan lathe should be capable of being used for a 3", or possibly 4", scale model. The key to the build is selection of a mill, or mills if space is available.

Andrew

Addendum: Logan lathes are quite rare in the UK, but are good lathes. I had one as a kid in the early 1970s, bought for £10 via the local model engineering club. It was in a building in the centre of Bedford that was being knocked down, so either I bought it or it went for scrap. I bought new half nuts from Powermatic Houdaille in the US while still at school. Fortunately i worked at Texas Instruments, up the road from my parents, as a summer job and their accounts department helped me sort out payment in US dollars.

Edited By Andrew Johnston on 12/06/2022 23:39:48

I use angle grinders for fettling castings and cleaning up arc welding. If I need precision (tenths) I have surface and cylindrical grinders. They are for removing a few thou at most, not large amounts of material. I don't have a Dremel style tool, no use to me.

I use hacksaws and files a lot. They are often quicker than setting up a machine tool and it is perfectly possible to work to thous with files. I have just finished making some steps for the tenders on my traction engines. Rather than castings they have been made from hot rolled steel. Most of the work has been done on the manual and CNC mills. But all the fillets were done by hand with files and radius gauges. Much quicker than setting up the horizontal mill and far easier to blend the fillets around curves. For slots in one of the steps i needed to make a woodruff style cutter but with longer reach. The cutting tool was made from gauge plate, turned in the lathe to get the correct diameter. Radii on the cutter edges were done with a file. The teeth were done by eye; the bulk of the material cut out with a hacksaw and the cutting edges cleaned up with a file before hardening. The cutter worked fine and took less than an hour to make, including hardening.

I use good quality files, mostly Vallorbe. I have bought some of the Tome Feteira files from ARC, but haven't got around to using them yet.

Andrew

In theory EN8 comes in a wide range of variants. I have two types in stock, EN8 (080A40) and EN8DM (212A42). The first is straight EN8. The second is an amalgam of D (tweaked composition for improved flame and induction hardening) and M (freecutting). The D variant has a slightly higher manganese content and possibly some molybdenum. When flame or induction hardening it should be possible to achieve a hardness of 50-55Rc. The M variant contains a small amount of sulphur, and in one datasheet I found a small amount of lead.

Both my stockl sizes are around 1" diameter. To clean up the stock I took a 50 thou DOC, 4 thou/rev feed at 1200rpm. The EN8DM felt smoother than the EN8. This was borne ouit by measured Ra values at three points (in microns):

EN8: 3.62 3.74 3.35

EN8DM: 2.55 2.73 2.64

I have a set of hardness files:

I assume that both bars were supplied in the annealed state. In both cases the 40Rc file easily cut the material, so the hardness as supplied is somewhat less than 40Rc.

Each sample was heated for an hour in an electric furnace at 830°C and then quenched in brine. Using the files hardness was as follows:

EN8: ~45Rc

EN8DM: ~50Rc

Each sample was then tempered for an hour at 650°C and quenched in brine. Using the files the hardness in both cases was reduced to less than 40Rc.

I draw the following tentative conclusions; EN8 is not hardenable in the same way as silver steel or gauge plate. Heat treatment does improve hardness. The D version seems to give higher hardness, as expected. Both flame and induction hardening are surface hardening techniques. While EN8 cannot be hardened as would a cutting tool, it does seem possible to increase the Rc value. That should increase the wear resistance, which is an advertised advantage of using EN8.

Andrew

Seems rather odd to go to the trouble of finding, and downloading, a generic picture of the overall table when it would have taken a couple of seconds to step back and take a general picture of the unit while photographing the damage.

It doesn't look like the normal sort of damage one normally sees on a mill table; it's rather uneven?

Andrew

That's the only sensible part of the original post.

Andrew

By all means speak for yourself, but don't invoke the unknown masses to support your 'case'.

I have extensive offcut bins, but I know what the materials are. Over the years I've been given a lot of unknown material. Many of them proved difficult to machine or form, in the end I realised i never used them, so they went for recycling. I value my time and don't wish to waste it on materials that are not suitable for purpose.

Over time there have been quite a few beginner posts about problems with finish when turning, almost always when using unknown material. Moving to a known material often solves the problem, so I stand by my advice.

NB: I have offered some suitable material to the OP, he needs to read his private message(s).

Andrew

Quite right too.

Although I know the Argand diagram I prefer to think of complex numbers (they're not really that complex) as 2-dimensional numbers. I normally visualise them on the s-plane (Laplace transform) as much of my use of them has been related to filter design and manipulation of poles and zeros.

Seeing them as 2-dimensional fits in nicely with the 4-dimensional numbers described by DC31 (quaternions) and 8-dimensional numbers aka Cayley numbers.

Many years ago I had the idea of using quaternions and Cayley numbers to calculate Fourier transforms, as they should simplify, in theory, the computational units needed. But it turns out that quaterions are non-commutative in multiplication and Cayley numbers are non-commutative and non-associative. Another idea bit the dust.

Andrew

The question of whether mathematics is a construct of man, or is inherent the universe, is one which has been debated by mathematicians and philosophers for many centuries, without resolution.

Personally I think mathematics is invented rather than waiting to be discovered. But I also think it is an unanswerable question. If mathematics is an inherent part of the universe we would only know that if it can be observed by a single person with knowledge of mathematics. But it might be that the mathematics exists by the nature of their presence rather than being inherent.

Andrew

First things first; if you don't know what the metal is then bin it and get a known grade of metal. The ideal would be EN1A leaded. If, with a known grade of metal, the finish is still poor then the tool shape and position can be considered.

Andrew

Only partially, Roy has the conversion wrong as well.

Andrew

I'd concur, for steel in particular the supplier matters. I bought this lump of 4" diameter steel on Ebay as it was cheap:

It was to be used as a former for bending sheet metal so quality didn't matter. Just as well, as the finish was poor:

For smaller diameter (less than 1" diameter) I buy a standard 3m lengths from metals4u. For larger diameters, and specialist alloys, I buy from m-machine in Darlington.

Andrew

I suspect that is largely the case. I don't suppose that the mathematicians developing the theory of finite fields in the 19th century had any inkling that they would be central to coding theory in the later 20th century. Coding theory underpins mobile phones among other comms systems.

Andrew

I obviously didn't explain the idea very well. What I meant was related to the statement by PatJ that mathematics is used to describe the universe. Of course it is, but the point I was trying to make is that mathematics does not exist for that reason. It exists in it's own right and is independent of any practical application, at least to pure mathematicians. Mathematics is a useful tool, but is not determined by physical applications.

Andrew

Calum has explained why that is incorrect. I would add the following to the explanation. Mathematics is not a rule based system, it is an axiomatic system. That is one starts with a list of axioms and then one can create rules that can be proved by reduction to the basic axioms. An axiom is a statement that is taken to be true but is unprovable.

It is possible to create a mathematical system that is complete and consistent. What Godel demonstrated is that a mathematical system that has axioms that allow the existence of natural numbers (in other words is useful) to be derived is incomplete and inconsistent.

The same is true of computer programs; they are not deterministic. Much effort has been expended on developing axiomatic bases and formal verification for computer languages to try and ensure that they will do what is intended. I believe it was Tony Hoare who stated that within any large program is a small program trying to get out. As a starter he invented Hoare logic in an attempt to prove program correctness.

One is highly unlikely to come across inconsistencies in mathematics that excite pure mathematicians. The engineer can safely assume that the mathematics we use will work.

Andrew

The question of what is mathematics has stumped mathematicians for centuries, but particularly in the 19th and early 20th centuries. Questions like what makes a mathematical system useful, and what are it's characteristics? Or how many infinities are there and what are their properties. In the 19th century it became clear that there are limits to mathematics, but that there are also alternatives to what most people considered to be mathematics. For instance in this thread many people have referred to trig and geometry. What they actually mean is Euclidian geometry, ie, based on planes. There are a myriad of other geometries constructed on other surfaces. Another useful geometry is spherical geometry (a special case of a Riemannian geometry) which, being based on a spherw, has uses in navigation. Each geometry has different characteristids. In Euclidian geometry parallel lines never meet, in spherical geometry they always meet at a point.

In the 1920s Hilbert proposed a study to show that all mathematics follows from a correctly chosen set of axioms, and that said set can be shown to be consistent. This turned out to be impossible on both counts. In the late 1920s Godel showed that within any mathematical system sufficiently powerful enough to contain natural numbers there are true propositions about natural numbers that cannot be proved, or disproved, from the axoims. In simple terms any useful mathematical system is incomplete and inconsistent as a consequence of it's existence. I'm afraid that those posters who think that mathematics is always right, and consistent, are wrong.

There is nothing in mathematics that bears any relationship to the the real world, so I think PatJ is also wrong in his definition. Of course mathematics is very useful to model the world, but it is only a model, which may, or may not, be a good fit. But there is nothing inherent in mathematics that says it reflects the world. Newtonian mechanics provided a good model for describing the physical world up until the late 19th century, when new experiments and measurements seemed to be inconsistent with the Newtonian model. In the end that led to the development of relativity which extends the validity of the model, but has it's own inconsistencies.

Andrew

Commercially the shapes would be cut on a wire EDM. It is possible to mill HSS with carbide cutters, although some filing would be needed for the sharp internal corner on a curve. Diamond files work well, but they are finishing tools, not for bulk material removal. Heating treating O1 is a darn sight less faff than filing shapes in HSS.

Andrew

I've turned and milled a lot of it, all the bevel gears, pinions and the liners for both my traction engines. It is lovely to machine, soft and consistent with no hard spots or inclusions.

Having said that most of the iron castings for my traction engines have also been soft with no hard spots. I think my supplier uses a small, traditional, foundry.

Andrew

I know the feeling:

Can we assume that aluminium means an alloy? Pure aluminium is horrid to machine, bit like warm fudge. Summarising a few pointers:

* Use 2 or 3 flute cutters

* ideally use flood coolant or an occasional squirt of WD40

* If no coolant or WD40 can be used then a finish cut climb milling gives a much better finish

* Highly polished carbide cutters are available, specifically for aluminium

* Maximise depth of cut and minimise stepover, tends to give long thin swarf that is less likely to clog the flutes - Edit: for slotting one is stuck with full width cutting, flood coolant is the proper answer. Unless the swarf is continuously cleared not only will it tend to clog but will be recut which doesn't do the cutter any good

* On aluminium alloys I run fast, several thousand rpm and high feedrates so one gets thicker swarf rather than fine which is more likely to clog the flutes. For example slotting 22mm deep on 6082 with a 3-flute 6mm carbide cutter parameters were 2500rpm, feedrate 400mm/min and stepdown 5.5mm per pass. Of course I was using flood coolant, primarily to clear the swarf rather than provide cooling.

Andrew

This is my internal grinding setup (I don't have the belt guard casting):

I got one internal spindle arbor with my grinder. Judging by the patina and style I judge it to probably be original Myford manufacture:

My spindle thread is the same, 5/8" 20 tpi Whitworth RH. The key dimensions of my arbor are:

Taper small end: ~0.235"

Taper large end: 0.547"

Length of taper: 1-1/16"

Diameter of plain cylindrical section: 0.471"

Diameter of grinding wheel cylindrical section: Nominally 3/8", but is tapered by ~12 thou along the length

The internal thread for the screw that holds the grinding wheel is 1/4" 26tpi RH, ie, BSF.

I have the original manual for my MG12, with the correct serial number manually typed in red on the page regarding serial numbers. The section on the internal attachment is unhelpful. Mostly about how to fit it, and how to do the initial fitting and alignment if it was bought at a later date rather than with the base machine.

Andrew

I have an older MG12. The internal spindle bolts to the front of the wheel spindle, replacing a plate, as opposed to the later pulldown internal spindle. The dimensions given above do not relate to the internal arbor I have. As far as I am aware the internal spindle didn't change between the MG9 and MG12.

Andrew