The future of casting kits

[noel shelleyParticipant @noelshelley55608]

I thought the thread was in the context of kits and there production ! IF it's a one off then yes CNC would possiblly be quicker BUT you have to write the code, I have to make a pattern, BUT once done I can do 6 or even 10 at one time, CNC does them one at a time. There's room for both technologies as each has it's good points ! CNC for pattern making may be a good half way house. Noel

[LukerParticipant @luker]

Sadly I think the argument for sand casting is a little moot if the required skills are not available (at a commercial level). It would be interesting to find the break even in cost between 3D printed patterns for investment casting vs CNC. I have a feeling the more demand there is for a kit the more they would lean towards investment casting.

Incidentally the skills required for investment casting is different to sand casting. With the stiff mould and pouring into a heated shell the riser design is much easier, in fact in my experience the investment foundry-man is unlikely to have ever calculated casting modulus etc. and they generally feed from the cup.

[Phil H1Participant @philh196021]

I think quite a few of the model locomotive castings are being bypassed these days. For example, as HowardT suggests, the cylinders are an incredible price but they can be machined from a block of Iron. There are other parts too like the buffers that appear to be easier to machine from bar and the main horns that can be simple strips of BMS.

When it comes to stationary engines (and this is a personal view), I get the idea of trying to machine or fabricate parts that would be detailed castings on the full sized engine where a modeller wants to replicate the various webs and flanges but I don't understand trying to make a block of metal look like an old fashioned, simple stationary engine casting from say the 1950s. I think the small bar stock versions look better.

[JAParticipant @ja]

I have followed this thread with interest.

In the outside world ordinary casting of metal appears to be dead. I don't think we used much at work during my last days of employment. What was used has highly specialised and outside the ordinary casters' skills.

We used a vast amount of investment casting which gives good finishes etc. However investment casting was adopted because it could produce a very large number of the same part such as turbine guide vanes. The wax patterns could be quickly and accurately produced in their hundreds by low pressure die casting. These would be assembled as a tree, coated in ceramic and the metal poured. Patterns for large cylindrical castings could be build from segments. No one understood it as a small batch casting technology.

There was excitement when 3d printing first arrived but the two technologies never married. Printing did not compete with the speed of die casting wax patterns. The use of lost wax for one off work is best left to jewellers dentists although there will always a demand from model engineers wanting one off castings.

CNC is obviously the way to go but the required investment, knowledge, machinery and time to for the coding, would be very large for our suppliers. I would suspect the number of models on offer to drop considerably and many would never be seen again.

Now to buy that big block of COLPHOS 90 for the cylinders.

JA

Edited By JA on 19/01/2022 17:56:42

[JasonBModerator @jasonb]

Yes the thread was originally about being able to buy a basic set of "castings" produced by other methods than actual casting not to tell people to give up their manual machines, lean to write G-code or build a backyard foundry.

Plenty of posts here and on other forums about the difficulty suppliers are having finding commercial foundries to do the small volumes that the hobby tends to need for various reasons such as not as profitable as batch runs for bigger clients or being priced out due to H&S requirements (not something the home caster tends to be troubled by) Though they may well have to find a CNC shop able to cut their parts or invest in a machine themselves.

As for volume production being better suited to casting yes it has traditionally been done so but how about a couple more photos from the makers of that Whippet all CNC machined from a block and they can sell as many as they make a batch off often several hundread at a time. So if it's viable to the open minded and forward thinking maybe it is something out suppliers could be looking into the next time they loose the foundry they have been using.

 

 

We certainly don't see many completely new design coming onto the market these days just the same old things being produced from patterns that have been about for years is this due to the investment of having to make new ones or just that the market is too small for it to be worth it for a couple of sales a year or they don't want to tie up capital in having a decent size batch cast to get an economical price or a foundry interested in doing the work. can anyone think back to Reeves or Blackgates offering something new? and I'm not counting things like buying up old Clarkson patterns.

 

Edited By JasonB on 19/01/2022 18:35:30

Edited By JasonB on 24/01/2022 07:45:05

[SillyOldDufferModerator @sillyoldduffer]

Posted by noel shelley on 19/01/2022 17:08:39:

… IF it's a one off then yes CNC would possiblly be quicker BUT you have to write the code, I have to make a pattern, BUT once done I can do 6 or even 10 at one time, CNC does them one at a time. There's room for both technologies as each has it's good points ! … Noel

Is there a Production Engineer in the house! She would be able to calculate the break even point.

Where the break point is depends on many factors, but CNC competes well into the middle volume range. Multi-axis CNC machines are extraordinarily flexible and offer special efficiencies. Being general purpose, any machine that's available in the same factory, down the road, or overseas can be put to work. They can all take on a variety of different jobs. If a machine shop happens to have CNC machines idle, they can all be utilised until higher priority work arrives. High-utilisation helps overcome a serious manufacturing problem: any machine-tool that's not cutting metal is wasting loads of money! Labour costs are low because a single minder looks after many CNC machines and it might even be possible to run machines unattended overnight with the lights out. No expensive furnace, specialist staff, moulding floor, cranes, pattern-store or other foundry paraphernalia.

Traditional cast-iron moulding is comparatively labour and energy intensive. Manually machining bar-stock with lathe and mill is probably the cheapest way of making one-offs, and I'd guess moulding is cheapest when tens of thousands are needed. CNC is in the middle somewhere.

Would be interesting to stage a race at an Exhibition between CAD/CNC and Drawing Board/Pattern/Moulding both starting from the same moderately complicated object drawn from a hat. Like this sink overflow:

Dave

[JasonBModerator @jasonb]

An interesting piece Dave but there are a few variables.

A you have already modelled it in CAD do I assume you will send it as an electronic file instantly over the net to one party in a modern format such as a STEP file in which case they would have a head start over the other party who would need a dimensioned drawing or your model of the prototype sent through the post and we know how slow that is these days.

As a CNC part its relatively easy for a machining centre to turn & bore the main shape, quick poke with a parting tool to do the undercut behind the flange before bringing in an active tool for the three slots along the side and the two half circle cut outs. The hole for the chain would be the tricky bit but a 5 axis would get in there fairly well. That s the part made and ready to go off the chrome plate. Probably 1 -2 mins a piece for mass production an 10 mins CAM for an experienced user.

The pattern maker will be at a disadvantage as the part will likely need machining all over externally to get it good enough to chrome, the groves and the undercut. Chain hole could be cored but that adds to initial pattern making. time and moulding time.

In the commercial world things like sink wastes were often a combination of the two with the rough part being cast and then visible surfaces and threads cut. Tradition would still loose out here as manually doing those even on a capstan machine is likely to be slower than second ops on a CNC and labour costs would make it uneconomic. These days they may just CNC the lot.

Edited By JasonB on 19/01/2022 19:10:09

[SillyOldDufferModerator @sillyoldduffer]

Posted by JasonB on 19/01/2022 19:09:05:

An interesting piece Dave but there are a few variables.

A you have already modelled it in CAD do I assume you will send it as an electronic file instantly over the net to one party in a modern format such as a STEP file in which case they would have a head start over the other party who would need a dimensioned drawing or your model of the prototype sent through the post and we know how slow that is these days.

…

No, I meant a physical example to be handed over so both sides could measure it. No giving anyone a head start by providing STEP files! One team would reproduce the example from 3D-CAD to CNC, the other from a drawing board/2D CAD to mould via a pattern.

For extra mayhem, the mould team could challenge the CNC team to reproduce one of their mouldings and vice versa.

I wouldn't expect either CNC or moulding to produce completely finished items, but what was achieved would show what the two processes are capable of, pros and cons. Extra points awarded for making lots of reproductions.

The original sink overflow this thingy replaced was a plastic injection moulding, maybe two parts glued together.

Dave

[JasonBModerator @jasonb]

From a post elsewhere it's interesting to see Stuarts have dropped another cast part in favour of CNC

[HOWARDTParticipant @howardt]

Commercial foundries have to make a profit. Traditional pattern and casting methods are time consuming, more so for small parts. Looking at a small static engine part there may be only pennies worth of material in it but many minutes of labour time, plus all the energy costs, and we know those only too well. When you compare that to continuously cast bar where many metres are produced in the same time. If you look at a simple static engine and take away all the fancy bits the actual bar material cost is far less than a bought set of castings. Some larger parts may need a bit of sideways thinking in order to machine them in the garage workshop, but that just adds to the experience. As an example I have a mini lathe and find it is a little shy of power when turning 5” wheels, so once roughed on one side and bored I mounted them on an arbor and used the mill to turn finish turn them.

[Neil WyattModerator @neilwyatt]

I think you've got the seed of a business model there Jason…

The 'investment' is in building a large enough range of designs to have wide enough appeal for people to buy them.

But I would guess an enterprising designer could get sets of parts CNC'd in small quantities at reasonable cost.

Perhaps you should do a trial?

Neil

[Gerhard NovakParticipant @gerhardnovak66893]

I think it is a bit sad that castings are 'going', for me they are always a bit special, nothing straight, sometimes hard to machine, but they gave the machine the 'real' look. I loved making a few Stuart machines. At the moment I work on a Tubal Cain design (Lady Stephanie). Even most casted parts are brass it has still a fascination for me.

Probably because I have learned the business, on the technical high school I went to we had an aluminium and an iron foundry. The Iron foundry used a cupola furness, about 8m tall. We had 4 casting days per year. (in the time between the furness had to be re-prepared – by the students – and the students made moulds which in case of cast iron had to dry out properly. It was fascinating work, even if it has lost its importance in todays world. I am happy I could have this experience.

[Bill PudneyParticipant @billpudney37759]

Sadly the skills required to produce castings are going the way of many other skills…shipwrights, coopers, fitters, draftsmen etc etc. In the cruel hard world of business, what manager is going to pay to train somebody in a skill which can be easily replaced by a cheaper machine, or by a cheaper item …for instance coopers, with the predominant use of metal (or, shiver, plastic) a true cooper is a disappearingly small niche trade.

SOD suggested that a Production Engineer would be able to determine a break even point. Well I used to be a Production Engineer, so I started to prepare something that would point towards a break even point. It rapidly became obvious that with a hypothetical problem like this there are so many assumptions that have to be made (what is being made, what material, what machinery is available, what machinery is required, what quantity, what delivery rate, what testing, if any, what heat treatment if any and so on) that it rapidly becomes useless. Maybe it's because I'm a bloke, not a lady………….

cheers

Bill

Edited By Bill Pudney on 24/01/2022 05:45:56

[JasonBModerator @jasonb]

I think the "real look" is just down to how much effort the builder wants to put into a part, yes you can leave it looking like a bar stock engine or you can spend time adding things like fillets and rounding corners, texturing the surface etc.

I can also thing of a few casting kits where detail has been specifically left off to simplify the casting process for example leaving out undercuts that would mean the part can't be done from just a simple two part mould.

[JasonBModerator @jasonb]

I wonder if someone at Stuarts did the production engineering calculations?

It can't be because they are having difficulty finding a foundry as they have their own. Though I do see that as Bridport" they offer a CNC machining service, I'm not sure if that is in house or a local shop that they use but could explain the economics if they have a machine that won't be paying for itself if stood idle.

They also do machined kits and ready to run and quite likely it is more cost effective to make these parts from a nice length of easy to hold bar than fettle and try to hold a small casting. It would be a simple matter to stop the program part way through and produce the "raw casting" and leave others to have the final ops of holes added to make them completed parts which is right back to the point in my opening post.

[Martin ConnellyParticipant @martinconnelly55370]

Bill, I started thinking about the costing of the two process options, cast v CNC. I quickly decided I could never get enough information as it really does need insider knowledge from a company that is doing the work. I used to write Applications for Capital Expenditure (ACE) when we wanted to purchase machinery. This had to state the problem that needed solving, what the proposed solutions were, how much they would save, the payback period and there were expected to be at least three options considered in the solutions. If you couldn't make a case on paper then there was no way you would be allowed the funds. In order to write the ACE you needed to know a lot about the company cost rates, how much cost they put on floor space, what services were available, the ongoing consumable, service and maintenance costs etc. All sorts of things you would only get to know if you were heavily involved with the people who could tell you these things. Some of them did get a bit imaginative with the costings but as they had to go in front of a list of people they had to be at the very least plausible. Some of them took well over a year to go from start, thinking of what needed to be done, to finish, first use and validation.

Martin C

[John HaineParticipant @johnhaine32865]

What's all this about writing G Code? All done by the CAD/CAM these days.

[JasonBModerator @jasonb]

I've said it already a couple of times in this thread John. No need to be put off CNC by the fear of G-code, I've never written a single line of it

[John HaineParticipant @johnhaine32865]

It is fun though, though very tedious! There are things that CAM can't do at the moment I think.

[Roderick JenkinsParticipant @roderickjenkins93242]

Posted by JasonB on 23/01/2022 19:13:00:

From a post elsewhere it's interesting to see Stuarts have dropped another cast part in favour of CNC

I believe that the brass components in the ST kits were originally "hot pressed" rather than cast. Possibly these items were outsourced by the the original ST and the Bridport people find it more economical to make them in house using CNC. The brass components that were hot pressed needed very little finishing and the Bridport guys seem to be continuing this tradition.

Rod

[Author]

Posts