What Chazza does in his spare time

[Chazza]

For Kiwi John.

Actually only one of the myriad of things, I do in my limited spare time.

I like restoring cars and I also like making things, so when it comes to spare parts time, it is nice to make parts which are unduly expensive to buy, or unobtainable, or just for the fun of it.

One trade I have been practicing for years and I am slowly getting better at, is metal casting. Specifically I cast aluminium parts, which are NOT part of a safety system such as brakes, steering and suspension.

Here are some pics; if anyone is interested, I can show the complete process for making a part.

Photo below; 3 thermostat housings to suit a Series 3, 6 cylinder Land Rover. They come from the same catalogue as rocking-horse poo and hens-teeth, as far as I can work out. This was the most difficult casting I had made up to that time, as I had to devise new processes, not just in moulding, but also in machining. Note the dodgy hole in the bottom left one; it went back into the crucible.


Photo below: Thermostat housing to suit my 1976, Rover P6B saloon. New ones were available for $90 each. This one cost me nothing.



Photo below; inlet-manifold heating-pipe for my 1951, Series 1 land Rover. This was one of the first parts I made for this car, I ended up making most of the parts in the cooling system. I think the previous owner, fed it a diet of salt-water.



Photo below; A customer who was having troubles with his Iveco daily 4x4, contacted me about making a better transfer-box mount. So I made the pattern from his suggested sizes and he did the machining; they have become very popular.



And now things are getting hectic; most of 2019 was spent doing a huge order for a customer based in Britain, still haven't finished it and I want to get on with making panels for the Land Rover, on my new-to-me English Wheel.

Photo below; part of a very big order; everything fits Series 1, Land Rovers from the 1950's.



Cheers Charlie

[cliffrod]

Anything you post about what you do will be of great interest to me, Charley. No rush. No pressure. I've been very interested in doing some foundry/casting work here for a long time. I've been a great fan of the early art bronze work done in Benin and often thought such rustic methods were within my reach. Doing car & bike pieces would be even better. So when you recently mentioned your automotive casting work, it really got my attention.

Not that I need another direction/distraction to water down my time, focus and resources even more, but....

[Chazza]

Thank you Cliffy, you have given me the perfect reason to escape the 43 degrees C in my garage!

So let us begin with a general over-view of the process I use.

• All of my castings are in Aluminium, because my furnace can't get hot enough to melt other metals such as; brass, bronze and cast-iron. Aluminium (Al) melts at about 600 degrees Celsius.
• All of the castings are poured into green-sand moulds, sometimes with resin-sand cores in them.
• My crucible can hold about 1.5 litres of molten metal so I am not only restricted by metal type but also by the size of the casting required.
• In metalworking circles, casting is a very cheap and convenient way to produce parts quickly, accurately and repetitively.

Now some terminology, bearing in mind that these can vary from country to country;

• Pattern-maker - a tradesperson who makes the patterns and core-boxes.
• Moulder - a tradesperson who makes the moulds; pours the metal; strips the mould and liaises with the pattern-maker on results and improvements.
• Casting - a metal shape in an unfinished state after it has solidified.
• Mould - the shape of the object to be cast. Moulds can be made of ceramic, sand, or metal, or any substance that can withstand the heat of the casting, without destruction.
• Pattern - the external shape of the desired object, often with extra material attached, to allow the mould to be completed, or for the casting to be held while it is being machined. The pattern may be in one piece, or in many pieces, depending on the complexity of the mould. It may be able to be used repeatedly, or perhaps only once. It can be made from foam, wax, plastic, metal, wood, or a combination of these materials.
• Core - a mould made usually of sand, which has strength and can be handled by the moulder without breaking. Cores are used to create the internal hollow-shape of a casting.
• Core-box - a special mould made from wood, metal, or plastic into which the core-sand can be poured, rammed and then hardened, to make it handleable.
• Moulding boxes - boxes with no top or bottom, into which sand can be rammed to make the outside shape of a pattern. They can be made of metal, or timber.
• Cope - the top moulding box; it usually has pins to register it exactly with the drag.
• Drag - the bottom moulding box which has holes for the pins of the cope to engage in.
• Forge - the furnace, which is used to melt the metal charge into a liquid. Mine is fired with LPG, but I once used charcoal as a fuel. Industrial furnaces can be electrical, or coke fired as well.
• Crucible - a ceramic pot in which the molten metal is held and poured from, into the mould.
• Binder - a name given to describe whatever agent is used to hold the sand together when a mould is made.
• Green-sand - any fine grained sand will do, as long as it is clean and washed to remove impurities. Coarse sand will work but it will leave a very rough casting. "Green" refers to its fragility, not its colour; I bind my sand with soluble-oil.
• Draught - a slope placed on the relevant part of a pattern, to enable withdrawal from a sand-mould; usually about 3 degrees minimum.

OK, enough theory, let us see how all of these things are used to make one of Peter's stakes, or any of my parts for example.

THE PATTERN:
Photo below, shows a badge that has been taken from a car and attached to a piece of melamine MDF; this is called a pattern-plate. The pattern didn't have to be attached to a plate, I could have still taken a mould off it, simply by laying the badge on a smooth table and placing the drag around it, and ramming sand into the box. There will be photos of the sand later, when the temperature in the shed drops to survivable levels. I will explain more about patterns as we go along, but the fundamental principle with a sand-mould, is that the pattern must be able to be withdrawn easily from the sand, so that the mould is not broken badly. This badge was pressed by Land Rover out of sheet metal and it already has a draught on it, so all I did was make it a bit stronger and fill the back with glue, before screwing it to the plate. The brown square is a print, where I can put a cone, down which the metal will be poured later.



Photo below; here we can see that the pattern-plate has been placed between the cope on top and the drag underneath. These moulding boxes are made from 3mm steel plate and I have made the inside of them rough, by dragging a welding electrode over them. This roughness stops the sand sliding out when I move the boxes.



Photo below; a close up of the end of the core-boxes assembly; the registering-pins can be seen passing through the pattern-plate and engaging the holes in the drag. Neither the pattern-plate nor the drag can move sideways; this is standard practice, so that in the event of there being a moulding cavity in the drag and the cope, they will always be perfectly aligned when they are reassembled, after the pattern has been removed.



Photo below; here is a different type of pattern known as a split-pattern. The sharp-eyed amongst you may recognise it as the P6B thermostat housing, pictured in my first post. In this case, I made the pattern by sawing the original corroded housing in half, after repairing the inside and outside with plastic-filler. The cylindrical bits sticking out of the ends are core-prints, they will leave an impression in the mould, into which I can rest a resin-sand core, before assembling the moulding boxes. I will make one of these castings sometime in the future, so that you can see what I am blathering on about



Photo below: here is the inside of the same pattern showing the registering-dowels and holes, so that the two halves can be connected accurately, before packing the cope with sand. More of that later with photos. Normally I don't have to repair the inside of the pattern, unless I am using it as core-box, which is what I did with this part. More about core-boxes later as well. This is a good example of a composite-materials pattern



Time to go and push some immobile cars out of the shed, to make room for the English Wheel, which arrives on Tuesday!

Cheers Charlie

[neilb]

by all means keep posting! i may even have a job for you

[Chazza]

Anything I can do help you, or any others on the forum, would be a pleasure Neil.

[Chazza]

THE MOULD:
Now comes the fun bit. The mould is made of fine sand; I bought mine years ago when foundry supplies were freely available in this country. Nowadays green-sand moulding is so yester-year, that I can buy very little of anything. Those who live in Australia may have heard of a red sand called pindan; I believe my sand is that. Any sand will do, as long as it is very fine, otherwise the surface finish on the casting will be very poor.

Whatever sand you use needs to have a binding agent in it; for thousands of years people used water, which works perfectly well – think of making a sand castle on the beach, one uses damp sand for best results. There are, however, several problems with water as a binding agent.

Firstly, too much water in the mould can create a steam explosion and droplets of molten metal flying around the foundry, is apt to shorten a person's life.

Secondly, a water-bound mould dries out rather rapidly, so the mould cannot be kept for long, before it becomes useless; or dry sand washes into the mould cavity and wrecks the casting.

Thirdly, even when the amount of dampness is correct, a significant amount of steam is generated in the sand and this can cause porosity in the casting, unless the moulder has vented the mould correctly, which is extra work and therefore additional cost.

Soluble-oil, as used by machinists to cool cutting tools, was very popular in the 1980's and is what I use to bind the sand together. It does not generate explosive gas; it burns very poorly; it makes a bit of smoke but nothing a good respirator can't handle; the mould can be kept indefinitely and will not dry out.

It should be noted that both binding agents allow the sand to be continually re-used, so never throw it away unless it has been contaminated.

Picture above; the drag is placed on the table upside-down and the half of the pattern with no protruding dowels, is placed in the middle.



Picture above; sand is added to the drag and compacted with the rammer. Doesn't have be anything flash; for years I used a piece of 42 x 19mm pine as the rammer.


Picture above; more sand is added and rammed until the drag is full, then the excess sand is removed with a strickle, which is just a straight piece of timber with a crisp edge.



Picture above: The drag is turned over, the cope is placed on top of it, the other half of the pattern is put in place using the registering-dowels and dry sand is sprinkled on the surface of the drag. The red sand is what my sand looked like when it was new; it has gone black due to carbon from the burnt soluble-oil. Dry sand allows the oily sand in the two boxes to separate from each other; I do have some expensive parting-powder but it doesn't work as well. All of the sand on the face of the drag needs to be sprinkled with dry sand.

More to follow.

[Chazza]

Photo above; a tapered sprue-cone is placed at a convenient place on the drag, this is where the metal will get in, so I won't put the cone next to detail I don't want damaged, such as the rim where the hose goes on. Then the cope is filled with sand and rammed.



Picture above; the mould has been separated; in the foreground is the drag with the pattern removed; resting on the drag is the pattern, I used two split pins stuck into holes to give me handles to lift it out; in the background is the cope with the pattern still in it, note the hole where the sprue-cone was; between the two is a resin-sand core that I had in my collection; and at right-middle is a green-sand bush, held together by a baked-beans tin.



Picture above; the core is made in two halves and glued together, I have placed it onto the core prints, being careful not to dislodge any sand. Because the core wants to roll sideways, it is supported underneath by an aluminium chaplet, which will melt during the pour and become part of the casting.



Picture above; the patterns have been removed and the core inserted; the cope has been placed carefully on top and the bush placed over the sprue hole. When the mould was apart I cut a small channel in the sand, so that the metal can get from the sprue to the cavity. Both boxes had loose grains of sand blown out of them by mouth, with eyes shut.

The bush is there to making pouring easier, with the added advantage, that the extra head of liquid metal creates pressure in the mould and forces the aluminium into every nook and cranny.

Making the mould for this part takes 15 minutes – pretty quick metal work!

It is important to understand that much of what I have written so far, is glossing over all sorts of important detail, so if anyone has any questions don't be afraid to ask.

42 degrees today, so I won't be lighting the furnace and my boss told me yesterday, that I will most probably be away tomorrow fighting fires, probably all week.

So the next unexciting episode will be in the unforseeable future,

Cheers Charlie

[Chazza]

THE MELT:
I had a think about this and realised there is not really a way that a one-man-show, can film or photograph the process of melting aluminium; so in lieu of photographs, anyone left who is interested and wants to have a go at casting, may find the following of some use.

As we know from metal-shaping, not all aluminium is the same, as just about every piece of aluminium in today's society, is an alloy of some description. As a general rule-of-thumb if you want to cast aluminium, melt scrap that has been cast before. Window frames, drink cans; car bodies, etc. don't cast well and often produce castings with severe shrinkage. Stick with pistons, cylinder heads, engine blocks, water pumps, lawn mower base, etc. for your melt.

Make sure that the scrap is clean! Remove, where possible, foreign metals such as bronze, brass, steel and cast iron. Wash the aluminium in de-greaser and hose off with clean water. If foreign metal cannot be removed it can be removed from the melt later, but this is best avoided if possible, as these metals often release gas into the melt, which is not good for the casting; steel is one of the worst offenders.

At this point I must digress and talk about how to melt aluminium and what to do it in. Aluminium melts at around about 660°C, which can easily be achieved on a wood fire; in a charcoal hearth; or if you can get it, in a coke hearth. I recently saw a large, spreading-pool of solidified aluminium, propped against a shed wall; I think it was the remains of an aluminium engine caught in a bushfire.

Much more convenient than solid fuels are liquid and gaseous ones; I use bottled LPG with a large burner and an LPG regulator. Be warned though, that in Australia at least, making you own gas appliance is illegal because of the extremely dangerous explosive power of leaking LPG. Don't take this lightly, people have been killed by gas leaks and had their houses and workshops destroyed! If you like the simplicity of LPG, get a registered gas-fitter to approve the design and install it.

Recently I have been taken with the idea of firing a new furnace, on waste engine-oil and an air blower combination. The advantage is cheaper fuel and a much quicker melt-time. Once again though, un-ignited evaporated oil, is an explosive gas and nasty explosions have occurred – get a gas fitter's expert opinion.

So having made or bought, a furnace, or forge you will also need a ceramic crucible to melt the scrap in. I used to use a cast-iron plumber's pot but it does release gas into the melt, so I prefer ceramic. Pots and crucibles need to be warmed up slowly, so as not to crack them from thermal shock; this year I cracked a crucible that I have had since 1984 because I rushed things.


Photo above; my foundry has a sand floor; the furnace was made by myself using refractory fire-bricks; gas regulator visible; dross bin with tools next to it; ceramic crucible with scrap in it; and a falling wedge I made for myself, I plan to anneal the wedge using waste-heat, next time I use the furnace.



Photo above; Furnace lid, the pipe is for lifting it off. This is hard on the back, so a swinging lid is a better idea,

The sand floor is a safety item; any spilt metal will sit there quietly and do no harm. If it was spilt on concrete, moisture in the concrete boils and blows hot metal and concrete at the bystander! When in use, anything flammable is taken away from the foundry, or secured, so that a spill can't burn it.

WATER DOES NOT BELONG IN A FOUNDRY!
No drinks; no puddles; no damp scrap = safe working conditions.

Warm the crucible on top of the furnace, or if like mine the flame can be turned right down, warm it inside the furnace with the lid on. I take 30 minutes to warm it up now, having learned my expensive lesson. At the same time, place your carefully prepared scrap on top of the furnace and keep it there until it is ready to be put into the crucible. This achieves two things; firstly it drives all surface moisture off the casting, which is a safety item; and secondly it gets the scrap temperature up nice and high, which means that the puddle in the crucible stays molten. Adding thick, cool scrap might solidify the melt and delay things interminably.

As the melt increases in volume, replace the scrap you have used on the lid, with more cool pieces so that they can warm up. It is a good idea to start a melt with very small pieces as they will melt quicker and flow together; add the heavier pieces as the melt deepens.

How much scrap to melt, depends on the size and number of your moulds, so there is a bit of guesswork when you get started with a new mould. I keep written notes of how many of a particular type, can be made from one crucible.

By this stage the crucible is looking rather full and the time is drawing near to pour it. But it is time to reverse a bit and think about all of the things that needed doing, while the melt was occurring.

• You have screeded the dry-sand floor smooth and placed your moulds in a sensible place where you can pour metal without hindrance.
• You have placed the tools and equipment you need to clean the dross off the melt; de-gas it and to lift the crucible, in convenient positions.
• You have closed the shed doors so that pets and sticky-beaks can't come and get in the way, or get injured.
• You have donned your personal protective clothing.

So, let us get on with it.



Photo above; home-made crucible tools. L to R; de-gassing plunger; small skimmer; a hook for snaring piston rings; blacksmiths tongs for handling scrap.

When you melt aluminium, a percentage of it oxidises in the atmosphere and forms a crust on the surface called dross. Dross also has impurities in it, which have separated from the scrap, such as; carbon, hard-water scale; burnt paint, etc. This dross needs to be skimmed off the surface and dumped in the foundry rubbish bin; when the surface looks bright and shiny you have done well. However, down in the bottom of the melt there may be some nasties such as; steel reinforcers from pistons, studs, etc. I have two skimmers; one that is about the same circumference-arc as the bottom of the crucible and one about the same as the top. The little one is ideal for feeling for steel scrap and lifting it out.

Last of all, is to remove all of the Hydrogen gas caught in the melt, which was put there by the steel bits. A plunger is used to plunge a very small amount of de-gassing tablet to the bottom of the crucible; the tablet decays very quickly into a large amount of fluxing-gas, which not only removes the Hydrogen but also lifts out any dross in the melt. The top is skimmed again and then the crucible is ready to be removed, once the lid has been removed. I use special lifting tongs I made myself and pouring tongs to tip the crucible, when pouring. I rest the crucible on a fire-brick when I take it out but it is also OK to rest it on the dry-sand floor, if that works for you.

I usually squat down as I lift the crucible from the floor; walk to the first mould and slowly but steadily fill the mould until it appears in the bush, whilst squatting. Repeat until all of the moulds are full, or until you know that you do not have enough metal. If so, lower the crucible into the still running furnace, replace the lid and add more hot scrap. After the last mould has been poured, pour any excess metal into a piece of angle iron, to make a long ingot.

Return the hot empty crucible to the furnace to let it cool slowly; turn off the flame; replace the lid; check that all is safe and nothing is going to burn and go and have a cuppa.



Picture above; L to R; crucible lifting tongs; pouring tongs; bar to poke scrap when it gets stuck in the crucible.



Picture above; excess aluminium is poured into this piece of angle, sheet-metal stops at each end are held in place by sand, the ingot can be sawn smaller, or snapped in a press.

[Chazza]

SAFETY:
I usually put the safety rules first in articles I write, but I didn't want to scare anyone off what is a very useful set of skills, in a very safe trade, if the safety rules are obeyed without question, or exception.

I have been pouring aluminium since 1978 and have had only one near-miss and that was early on, following some very poor instruction and involved water as a binder.

In 1985 I was most fortunate to have had some very useful instruction from a pattern-maker whose father had been a moulder. What that man didn't know about both trades, wasn't worth knowing I suspect.

If you have an interest in casting, then ignore 99% of what you will find on Youtube. Most of the posts there show a complete lack of safe-working practice and in some cases, an appalling ignorance of potential dangers. In any case some of the techniques and practices are not the best ones to use.

I am also a member of a casting forum, where I have been ridiculed for pointing out unsafe practice. One man even suggested it was a point-of-honour, to carry the scars on your arms caused by spalling concrete. If he is more of a man than me, then I am happy to be called something inferior to him!

Before the rules let us consider the potential hazards, which are easily minimised to nothing.

• Hot flying metal and concrete.
• Losing your life.
• Suffering permanent disfigurement and disability.
• Losing your income permanently.
• Losing your house or shed.
• Injuring or killing another person.

All of these consequences are avoidable and when we think about them, are no different to some of the hazards many metal work tasks have.

Hazards in capitals. Safe working practice listed below.

FIRE.

• Melt metal only on a sand floor to prevent spills from contacting flammables.
• No water or dampness EVER in the foundry.
• Remove all flammable materials well away from the pouring area.
• Guard flammables such as gas hoses, with steel screens.
• Have appropriate fire-fighting equipment nearby and ready to use.
• Use only approved and licenced gas equipment.
• If using solid fuel, use spark arrestors.
• Make sure that inside and outside the foundry has all fuel removed.
• Do not use the foundry on dangerous-fire-condition days.
• Only melt inside a dry foundry, never outside.

HAZARDOUS FUMES

• Clean scrap only.
• Teflon coatings are hazardous!
• Wear a respirator capable of dealing with gas and dust.

BURNS

• Wear full PPC; boots; leather spats; leather trousers or chaps; leather boiler-maker's jacket over a leather apron; leather cap with neck protection; goggles or safety spectacles; full face-shield; long leather welding gloves; cotton or woollen clothing under above.
• Treat all metal as hot.
• Use tongs and other suitable tools instead of hands.
• Secure the foundry from people and animals during and after a pour.
• Do not allow others to watch a pour; show them a video instead.
• In the unlikely event of a burn, know where the nearest cold running water is, to begin first aid.

There may be hazards I can't think of but if you work on the principle of;

1. Removing the hazard first.
2. Engineering to minimise the hazard
3. Isolating the hazard
4. And last of all wearing PPC to protect from accidental contact,

it is highly unlikely that anyone will get hurt.

Cheers Charlie

[cliffrod]

Hot damn, Charley- this is fantastic! Maybe someone will post something more interesting (umm, spellcheck changed "interesting" to "erecting", but that's not the kind of exciting I was talking about....) someday. until that happens, you're alone at the end of the ruler. Lots of questions to ask, but want to get them organized. Thank you very much.

Please be carfeful at work. They're talking about the AU fires on the news here in the states.

[Chazza]

... (umm, spellcheck changed "interesting" to "erecting", but that's not the kind of exciting I was talking about....) ...

Ha ha!

Cooler change on Friday; with a bit of luck I will be here still and can do a pour early in the morning on Saturday.

Just a piffling 38C today.

Glad you like the articles Cliffy.

Better go now, English Wheel arriving in 20 minutes!

Cheers C

[cliffrod]

Looking forward to pics of your English wheel someday and maybe some history about its past life, if you happen to have it.

[Peter Tommasini]

I might have a job for you as well,....... keep posting
Cheers
Peter T.

[Chazza]

CORE BOXES:
One of the hardest things for people to get their head around in the pattern making process, is what shape should the core be and how to go about making one. Earlier pages showed a core in the green-sand mould, but no information about making one.

Imagine you are standing in the foundry in 1854 and a customer wants a bronze pipe made, 12" long and 1" in diameter. So the pattern is turned on the lathe and split down the middle – see Fig.1 below. Having established the shape and diameter of the outside of the pipe, you now need to physically make the shape of the inside of the pipe, on a separate pattern called a core-box. Let us say that the pipe was just a straight-forward hollow cylinder. So the wooden blank could be held in a chuck and a drill bit, or boring-bar, passed all the way through the length to make a hollow cylinder and then the core box is split in half, as in the sequence shown below.

This of course is an incredibly inefficient way to make a pipe – although better than machining one from solid bar – which is why no one does it anymore. However, it is essential in casting to know how to make cores, as they are needed to make hollow shapes and to reduce the thickness of castings, where too much metal would be detrimental, or wasteful and expensive.



Fig. 1 Method used to turn identical halves in timber on a lathe.

The method shown in Fig. 1 is very useful, whether or not, the object to be made is cylindrical. When I reverse-engineer a casting, the first thing I look for are lines of symmetry, or lines where a pattern, or part-pattern can be split. Note, that if the piece is to be turned, the wood-to-wood on the ends makes it less lightly to split down the paper line, when the tailstock is tightened.

The second thing I ask myself; is whether the pattern could then be withdrawn from the sand, or not, i.e. does the pattern have a draught, or does draught need to be put onto it? On a new pattern draught can be turned, planed, sanded, or sawn to give the 3° draught required. On existing patterns – e.g. a sawn-in-half metal part – draught can be added with plastic filler.



Fig. 2 Finishing the core-box.

In Fig.2, are some simple core boxes and two suggestions for the type of sand to use. When using either, sand is poured into the core-box and then screeded with a strickle, before hardening. The sheet metal ends can be loosened after the core has hardened, to facilitate removal of the core. Glue the two halves together with PVA wood and paper glue.

Sodium-silicate, also known as water-glass, when mixed in damp sand, can be hardened with carbon dioxide in a matter of seconds. As long as the wooden core-box is smooth and well varnished and a release-agent such as talc, or silicone spray, has been applied to it, then by inverting the box and rapping it on the bench, a core of this shape, will fall out easily. For awkward shaped cores, ejector pins need to be fitted to the core-box to help get the core out. Advantages; cheap sand mix; no hot work; fast; fine sand can be used; strong core. Disadvantage; have to buy CO2 and regulator.

Resin sand requires a metal core-box; the core-box is heated over a gas flame and the resin hardens and so the core is formed. Advantage; makes a strong core. Disadvantages are; expensive sand, if you can get it; time-consuming to make core-box twice; my resin sand is coarse and gives a rough finish on the casting; hot work is tiresome and presents safety hazards.


When I made my first thermostat housing, pictured below, I devised a plan where I would cut the corroded original in half and make some timber ends which were glued to the aluminium. The ends had a hole bored in them, which was the same size as the core prints on the pattern. To bore half a hole, use the cartridge paper method, or even easier, is to clamp the two pieces in a machine vice and then drill them.



Photo 1 above; the old thermostat housing has been cut in half along a line of symmetry; ends have been made to support the core-box, they have draught so that a mould can be taken, to reproduce it in aluminium. Note the use of plastic filler on both halves to repair corrosion damage and to create draught on the base. Note also, the half-holes on the ends; these will form a cylinder, which rests in the core prints on the mould. Top left is the cast part, still needing some finishing where the hose goes on.

Although this part has a line of symmetry in it, the two parts of the core and of course the pattern, will not be identical i.e. they will be left and right-handed. This time both sides of the core-box will be needed, unlike the simple core-box in Fig 2. (1).



Photo 2 above; here is an asymmetrical core for a Land Rover thermostat housing. I cut the old housing in half, to eventually end up with two aluminium core boxes. The sand is resin-sand and as can be seen, it is coarse sand and leaves a rough finish on the aluminium. This is one reason why I am experimenting with sodium-silicate cores, so that I can have smoother castings for the customers.

If I had to make the core box from scratch – perhaps from a drawing – then it becomes much more difficult. In the past pattern makers were experts at making patterns in wood, which is what I wish to be; much like the objectives off this forum to learn traditional shaping methods. Anyway, back to making a core-box from scratch; nowadays we can use plastics, wood, metal or a 3D printer to make patterns. Some of this core could be made on a wood lathe and perhaps parts of it by hand with good chisels.

Soft hardwoods such as Jelutong were used in the past but there isn't enough rainforest left, to countenance that anymore. I use recycled timber from old furniture and a local hardwood called jarrah. Any timber pattern needs to painted, or varnished and kept out of the sun, as this is one disadvantage of using timber; warp and shrinkage.


Photo 3 above; two core-boxes cast in aluminium; the top one is for one-half of the core in photo 2, the heads of the two screws are the ejector-pins. The bottom one is an aluminium box made from the pattern in photo 1, the ejector pins are raised because it is sitting on the bench. When I cook the cores I have to be careful that the pins drop though holes in the mesh, which sits on top of the burner.

On the top core-box, part of a handle can be seen, so that I don't get burnt. For the bottom one I use vice grips, but it is fiddly, hot and not the best way to do things. When I perfect sodium-silicate core sand, I can still use these core-boxes.

More to follow.

[Chazza]

I have never had to make a core-box from scratch, but I have made this pattern from scratch, which involves much the same skills and some of the techniques.
A friend asked me to make an exhaust valve side-cover for a 1948 Land-Rover and he lent me one to copy.



Photo 1. The outside shape made in jelutong, because it is easy to sand out marks.



Photo 2. The inside shape. While this pattern does not require a core, the hollow I created here with gouges and a mallet, is a similar technique used to make some core boxes. Curves were copied with a comb-gauge and templates made out of card – same as we do when copying panels. Plastic filler has been used to correct imperfections and to make concave radii

To begin with, an accurate sketch had to be made of the original part and the dimensions measured with vernier calipers and steel rules. But now comes the risk, which can ruin a good project, if not understood by the person making it.

Metal, as I am sure we all know, expands when it gets hot and contracts when it cools. Aluminium contracts by 1.3% (“The Foseco Foundryman’s Handbook”; 9^th Edition; pp49); therefore when making a pattern it should be 1.3% bigger than the drawing, or the original part. Pattern makers use a contraction-rule to make production quicker but as they are rather expensive, I write the contraction-allowance sizes in red ink, next to the dimensions on my sketch.

Before I take merrily to the sketch with my red pen there are some other things to consider first. Because I took my measurements from a casting, not a drawing, it means that the casting has already shrunk by 1.3 % when it was first made back in ’48. Therefore the original pattern was 1.3 % bigger than the casting in front of me.

But how do I know if this casting is an original, or one that has been made using a casting as a pattern? If that was the case the casting is 2.6% smaller than the original pattern. The answer is that I do not know, so it is wise to measure the component that the casting is to be fitted to, to check that the pattern and casting will fit.

Plus, I also need to consider machining: How will it be held? Would a spigot or two, be useful for clamping, etc.?

Now 1.3% of 6.35mm is 0.08mm; this is such a small size that I can’t reproduce it with hand tools so I can either ignore it, or make the pattern 6.5mm thick to compensate. On long parts such as the side-cover, which is about 420mm long, if I ignored the contraction-allowance, it would come out of the mould about 5.5mm too short! 5.5mm is an easy dimension to correct by making the pattern longer. Similarly all dimensions, which can be measured with a rule, can be enlarged.

Feel free to check my maths; I am notorious for getting it wrong,

Cheers Charlie

[cliffrod]

As a small thank you for all you're posting, maybe I can contribute something worthwhile here. This is how I would figure such contraction calculations for almost free.

In my business, sculptors make the original shape. These are traditionally then given to a carver- professional or apprentice- to measure and copy into stone. The bulk of the stone is removed by lower paid person(s). The project may ship as done by the carver or a master sculptor may finish the job. A carver is generally not as experienced/skilled as the sculptor, so mechanical means are utilized to help them duplicate models very accurately. A Carver may also not be as educated, so handing one a pencil & paper to make endless accurate mathematical calculations or to expect a fluency of geometry and trigonometry are non-starters. So this is how we do it do dirt cheap with almost no calculations-

For 1:1 duplication, an exact model is duplicated in 1:1 scale with a pointing machine as shown in my previous stone carving thread. It can be done with compasses as I will discuss, but unless the job is being carved in reverse, a pointing machine is usually faster.

for other dimensional manipulations, a compass (or caliper- not "proper" terminology, but think of calling a motorcycle vs enginecycle...) is employed. Technically, a compass measures a radius and a caliper measures a thickness. Each measure accurately and quickly without difficulty.

For simple whole factor manipulation like 2x, 4x, the desired measurement from the model is captured with a compass and walked along a straight line (proper terminology is a Ray, but...) from a dedicated origin point. The end point is lmarked gently. A new compass is used to capture the measurement between the origin and this end point. This compass is used to transfer the new calculated measurement to the job.

for all calculations between 0 (model measurement 100%, job measurement 0%) and 200% (model measurement is exactly 1/2 of job measurement), the function of the apex angle of a triangle is employed. For this discussion, I'll use the more simple isosceles triangle method. It's a lot harder to write this down than it is to do it. Enlargement of a scale model also enlarges discrepancies so is best done in several smaller steps instead of one large jump in size.

I've mentioned this before on AMS and have presented at Redneck Roundup. I've been trying to do a concise video of this as well, but have not been happy with multiple attempts. This information is somewhat obscure, so I also have reservations about who knows this & what is done with it. The last USA workshop presenting this info as "lost art" was $700+ per person and completely misrepresented that many of us whom the organizer knows regularly use this technology.

For this example, imagine I have a plaster model that is 3'-0" tall and I want to carve a statue that is 5'-2 1/2" tall. To figure out several hundred accurately factored new dimensions with pencil and paper would be untenable. For Charley's pattern, capture an easily measured dimension from either the model or the job and multiply/divide as needed it by the contraction factor if you like or do as described later producing a generic specific ratio function triangle template. At this point, Using a value that is easily divided by two speeds this process.

See attached. Hope,you can see this ok. If not I'll rework it. I usually do this by using a durable item such as a nail with a small divot struck or drilled into the head to use as the apex point. Fine lines and sharp points are your friends. When using such a triangle I draw it on a sheet of nice plywood for big jobs or even paper for small projects like scaling up pieces fro my Guzzi. I'll post some compass & caliper images after I get back into studio.

A desired job or target measurement is chosen or calculated.



1. Draw a centerline through the apex point.

2. Construct a pair of long parallel lines each equidistant from the apex point, hence the easily divided by two advice. The distance between these parallel lines should be the job measurement. This could be checked with another caliper if desired.

3. Using a compass, capture the corresponding dimension of the same detail on the model. Place one point of the compass into the apex point, swing the other point across each parallel line to strike an intersection.

4. Draw three lines to connect the apex point to the two strikes (model dimension) and the two strikes to each other (job dimension). If the job dimension on triangle is not the largest dimension on the job, draw the two model sides as rays (continuing past base). All you need is the apex angle and these two lines.

Now, take a measurement from the model with compass. Place one point in apex point. Swing and strike across each model leg. Use another compass to capture the distance between those two strikes. Apply this to the job. Repeat until finished.

As mathematical proof of this process, see sidebar of drawing. It's a long time since college, so bear with me-

If model dimension is 100% and job dimension is 0%, this will produce a triangle with superimposed sides but no width. This is a proof.

If the triangle is acute with model dimension greater than job dimension (greater than 0% but less than 100%) the job will be be a reduction of the model.

If model dimension is equal to job dimension, the triangle will be an equilateral. This is a proof.

If job is larger than model but less than twice as large (greater than 100% but less than 200%). The triangle will be obtuse and will allow a model to be enlarged.

If the job is exactly twice as large as the model, the base will be twice as long as each side with apex point located at the midpoint of the segment. This will construct as a flat line with sides superimposed upon the base. This is a proof.

The master who taught me how to use the compasses and triangle was barely literate and hardly fluent in math. That is the brilliance of this method. The proofs and such come from my advanced math instruction in college, which was brought into practical life application by my stone training.

We will then use specific agreement of groupings of three radii, captured from the model & enlarged or reduced with the triangle, to accurately produce and even reverse a job that is replica of a model. That's an entirely other process.... So is using a right triangle instead of an isoceles triangle .

For the purpose of contraction rates, you could simply construct a triangle with a base & sides that are exactly 1:3 to 1 for that specific metal or alloy and use it forever. The specific job & model measurements are irrelevant. Make it easy math because the factor doesn't change. The apex angle will provide the function of 1:3 to 1. I would make one triangle template for each contraction factor, Mark them, hang them on the wall and use as needed.

I have been working to simplify a method of using the most powerful triangle- acute, between 0% (nothing) and 100% (everything)- to facilitate producing work from archive images such as profile photographs and mechanical drawings that lack dimensional information. By reversing the side where a dimension is applied (I.e.- model at base) the calculated dimension can be any factor desired. For example, a profile image of a vehicle with a known wheelbase, wheel diameter, etc could be used to construct triangle which in turn could help produce all visible dimensions. The chance for error is greater, but likely not any worse than a highly pixelated blown up photograph. And it would cost nearly nothing. For these little dimensions described by Charley, I would use a magnifier to see what I was doing, super sharp points on your tools and very fine lines on my triangle.

CAD and such may suit some people better. This is where all of that came from. The calculations done within a computer program are all based upon the mathematical function of the apex angle of a triangle constructed using such corresponding dimensions as described.

Hope this helps. Ask questions if you like and and I'll try to clarify things. This is all second nature to me so I may not be as good as explaining/writing as I am at doing it. Sure takes a lot longer to write it down than to just do it.... Once you do it, it's really fast and easy with no math once you construct the triangle. Drawing two parallel lines equidistant from your apex point is the hardest part..

[cliffrod]

Just in case anyone thinks a sliding divider or proportional caliper is as professionally viable as the proper traditional use of compasses/calipers as I describe above, let's not go there.... There's a reason that none of my masters use those amatuer toys while all of them are fluent with compass and triangle.

TIA.