Tab & Slot Style Mill Frame

[joeavaerage]

Hi Howard,
that you have found a foundry is good news, it may or may not prove to be the best option.

The column of your machine is a more complex molding operation than it could or even should be.

The base is fine.

Look closely at the base, all the '3D shape' is on the upper side whereas the bottom is flat. This means that the cope (might be the drag, I can never remember which is which) can be molded on top of a flat board.
Very simple.

Look now at the column, you have 3D features on both sides, so you will need a split pattern, or a splitter board.

The costs of casting are molding, pouring and fettling. The cost of pouring and fettling are very much a reflection of the size of the casting. Molding on the other hand depends on the complexity of the part.
With some judicious re-design you could make the molding process much simpler and therefore cheaper.

Even a well thought out design that allows the use of a simple drag and cope greensand mold is still going to be expensive, the question is more 'is this option more price/performance
competitive than other alternatives?'

Craig

[peteeng]

Hi - Cope is top and drag is bottom. The drag mould is "dragged" around the floor is one way to look at it... Casting is done for a few reasons 1) if your going to make many parts casting is economical 2) if its a material like CI and you want it for its tribological properties or its damping qualities then casting makes sense 3) maybe its a shape you can't machine... Commercial machine builders cast because they make lots of parts and in the past where ways where machined into parts they use CI as its slippery. These days using rails, that side of the material does not matter. Using ready available steel , welding yourself and machining is traditionally the cheapest solution... Moulds and pouring take energy and that costs a lot for a bespoke part... But explore options. Nows the time to do that... There are machines here that are made from aluminium that have been poured at home... Peter

[joeavaerage]

Hi,
I costed aluminum castings for my axis beds. There is less heat required than for iron and so the pouring is cheaper. The problem is that the decrease in modulus
relative to iron renders any cost benefit moot, and you end up with a larger part for the same stiffness. It is not so much as I ruled it out but made the choice on the grounds
of cost.

I also ruled out lost foam on the basis of cost. Great technique if you have a complicated shape to cast, and means you might be able to avoid sand cores completely,
but nothing beats a simple two piece flask with a drag and cope and no cores for cost.

Craig

[joeavaerage]

Hi,

Even welding and truing surfaces seems expensive so I'm trying to brainstorm ways around that

Every man and his dog has given thought to that.

A number of people have used precision granite, and then effectively bolted it together, relying on the granite being stable and ground true. It has some things to recommend it, ie no welding, no casting,
no stress relief and no precision machining. That has to be weighed against the fact that granite is not hugely stiff, 70GPa, is expensive to buy initially largely due to the precision grinding in its manufacture,
there is very limited things you can do to it, cutting, drilling etc are all diamond grinding operations after all, and that installing threaded inserts is a pain, at least by comparison to drilling and tapping steel or iron.

Industrially granite machine tends to be used in low stress applications. It is highly favoured in CMM machines, cylindrical grinding machines. In both case the cutting forces are absent or low and the stability of granite
outweighs the costs and practical limitations.

Epoxy granite and/or UHPC are also popular. Epoxy granite has a low stiffness, sub 20GPa, by peeteng's samples and UHPC in the region of 50GPa. Both materials are not hugely stiff so will perforce require large sections to gain the
stiffness required. Most people mold their own and that has great appeal. Many have attempted to mould and cast in such a manner that does not require post casting machining. Other techniques like self leveling epoxy have been tried.
All such techniques are only fair at best.

I think it only fair to consider precision machining as part of the process, there is just no avoiding it if you want genuinely flat, true and square surfaces.

Welded steel requires stress relief, while bolted steel not. Cast iron can and would ideally be thermally stress relieved, but vibratory relief is viable also.

I tried to 'thread the needle' with my design. I had the axis beds cast, vibratory relieved and precision machined, and that was expensive, $10,000NZD ($6000USD).
I deemed that the axis beds are the most critical parts and that no expense be spared making them as good as can be. The frame to which they are bolted is plasma cut steel.
To date I've not welded it, and so does not require stress relief. The steel frame was so much cheaper than casting it, not withstanding the the plasma cut steel cost $2000NZD ($1300USD).
In short a steel frame is the stiffest for a given sum of money. When I get around to welding the frame then I'll be up for stress relief.....but that hasn't happened this week!

Now that you are getting a genuine glimpse into the true scale and cost of the project, its time to consider the alternative, buy something ready made and convert it to CNC if that's is
required. Making your own machine is a hugely expensive undertaking.

Craig

[peteeng]

Hi MK - I presume from your first design you have a local laser cutter to make the steel parts. Thats very good. Now you have a heat treater, very good. Now you need to find a machinist that can do all the work and you have solved your manufacturing puzzle. Comes down to cost at that point. Design it, work with your suppliers to get the best results, cost it then make decisions. Peter

If you have an electroplater handy get a price on flashing the parts with 5um zinc. Does the job nicely after final machining not before!! .... and do you weld? if not you have to find a welder... if your really lucky there maybe a laser welder around.

[awerby]

Would brazing or even soldering stick the pieces together well enough without the distortion caused by welding? If not, what about glues? There must be glues by now that rival solder in effectiveness.

[peteeng]

Hi Awerby - I totally support brazing and soldering for this sort of build. But seems there's a bit of pushback on these processes. I'm sure epoxy is strong enough to adhere machines together. But the joint being plastic does not have the stiffness needed for a machine. Brazing and solder being metals are very much stiffer. Solder is particularly easy to do as the heat required is less than 300C. And you can solder/braze aluminium. I have made several saddles for small routers using std lead solder and they have performed very well. I like stainless steel assemblies with silver solder they work very well. There are structural solders that are very good. Been a while since I've researched those. Peter

I think the process to use would be to edge bolt the assemblies together so you can check they are true, clean them prior (soak in vinegar or acid until clean, rinse well) place in oven if you have one and preheat to top oven temp say 250C then remove, flux up and start torching then add solder.....this would be straight forward. You can flux before preheat as well. If too big for oven then use a rose to bring the part up to temp then start soldering....

[joeavaerage]

Hi,
but why bother with soldering or brazing? Just to avoid thermal stress relief......?????

Brazing with a MIG is limited to thin joints, and not particularly strong. 2.4mm Flux Cored CO₂ shielded wire at around 25VDC and 250A will create a weld 10mm-12mm deep
and the weld looks beautiful, no grinding or rework required. Welding 8mm, 10mm, 12mm steel in one pass, strong as you like. Ok, so you have to stress relieve.....so what?
Its done all the time industrially so why would we try to avoid it?

Craig

[peteeng]

Hi Craig - Because there are many people who do not have easy access to TSR. So they need a different solution. Peter

[joeavaerage]

Hi,
I would not call having to ship my parts to Auckland from the South Island particularly convenient, it about 1500k's. If I want parts, large parts, thermally stress relieved
that is the hurdle, and the cost is about $6.70NZD/kg +GST plus shipping, end of story.

I suspect that most hobbyists are unfamiliar with the process and unfamiliar with the companies that do it and therefore react 'Thermal Stress Relief......No, not around here'
when in fact there ARE heat treaters if you are prepared to find them.

The closest alternative to thermal is vibratory stress relief, and I suspect that is less common than thermal ovens.

The real issue is: 'Is there a realistic alternative to stress relief, either thermal or vibratory?'.

I have not seen what I would call a good alternative.....how about you?.

Craig

[joeavaerage]

Hi,
just as a matter of interest, I'm making some small molds for a fuel float. I had some UHMWPE around and I tried it...they came out not too bad. UHMWPE forms long strings
rather than chips which makes it a PITA to machine. I did make a float (filled epoxy), and it sort of worked. This morning armed with what I had learnt from yesterday I went and
got some PTFE. Bloody expensive, but it machines really nicely. The surface still has a little texture under your finger nail, but feels 'as greasy as a butchers prick' under your fingertip.
I think this will work a treat.

Craig

[peteeng]

Hi Craig - Stress relief is mainly required on steel welded machines. As the parent metal and the weldment change from liquid to solid the metal shrinkage creates very large internal strains in the structure. These are an issue for down stream machining and keeping parts straight. SR removes the internal strain so when you machine it, it stays straight. I'm sure you know this, this is just a note for others that may not know it. But you can machine something with no internal strain and it comes out wrong if you clamp it incorrectly for instance. There's lots of traps for machine builders.

So if you bolted/brazed/soldered/glued a machine together you would not need to SR in my experience. I appreciate that hobby builders are not familiar with SR and may say that's too hard or proceed in the hope that it will work out. Another approach is to provide a lot of machining allowance metal. Then lightly machine the parts, release them and see how it's going. Then reset the part and machine a little more. etc until the part is correct. With a little luck this approach works but it's a bit costly due to the multiple set ups, plus there is the risk the part may not achieve stability. Hold your tongue on the left seems to be good.

I had two saddles welded for two machines a while ago. I specified preheat, weld sizes etc etc with the aim of not post machining them. But they came back warped. So when they were corrected they both were different. So then I had to make spacers so they stacked the same heights in the two machines , so drive nuts matched etc... I'm sure if I brazed them they would have been fine...

Hobby makers are cost driven and will try most things (or leave processes out) if it costs less. There are tried and proved manufacturing routes to achieve a required result. So pick your poison and then live with the result. Over the years I have done most things wrong just to see what happens. The "right" way has always resulted in the least or no back tracking...

So the realistic alternative to SR is 1) use a build method that does not need it 2) take the chance and put extra metal on your lands and see what happens. One process I have not tried is laser welding. I think this may be close to the ideal welding process for machines... Someone out there maybe able to comment on this... If I was setting up a commercial welding shop soon I'd be going laser. If I get a windfall I may buy one....Peter

[peteeng]

Hi Craig - If your making moulds I usually use aluminium and use a wax or silicon release. More accurate than cutting plastic. The exotherm can move plastic around as it gets soft? Peter

[joeavaerage]

Hi,

I had two saddles welded for two machines a while ago. I specified preheat, weld sizes etc etc with the aim of not post machining them. But they came back warped. So when they were corrected they both were different. So then I had to make spacers so they stacked the same heights in the two machines , so drive nuts matched etc... I'm sure if I brazed them they would have been fine...

Classic case of trying to save five dollars and yet costing yourself one hundred dollars. Been there, done that.

Over the years I have done most things wrong just to see what happens. The "right" way has always resulted in the least or no back tracking...

I have done likewise, but its finally starting to sink in, there is a reason that its called the 'right way'. My own experience is that the more time, effort and money I pour into something the more
risk averse I become. If stress relief is the right way, or precision machining is the right way, then you'd best budget on them as being required. The alternative is that all that effort and expense turns out very much less
well than the investment should have made. That really pisses me off!

Craig

[joeavaerage]

Hi,
The float is 50mm in diameter and 50mm long, so exotherm is not an issue with a very small quantity of epoxy and a large volume of phenolic micro-balloons. Tolerances are plus/minus 0.5mm
so there is no need for a hard mold, at least from the dimensional standpoint. Epoxy just skids off PTFE, but woe betide you if you miss a step in applying the release agents with a hard mold like aluminum.

Craig

[joeavaerage]

Hi,

If I was setting up a commercial welding shop soon I'd be going laser. If I get a windfall I may buy one....Peter

The problem with laser welding is the absolute mega dollars required. If you are basing your desire for a laser welder on the BS adds coming out of China, then you'll be mightily disappointed.
Those welders are well under 1kW, whereas a 'real welder' will be 5kW and more continuous. Cannot hang on to such a thing, even a minor reflection would cause a beam enough to cut the welders hand
straight off and cauterise on the way through!.

The eyewatering cost of high power lasers relegates them to aerospace and military.

Small and much cheaper laser welders are for thin section welding, at which they excel, not that it stops there being a heat affected zone nor does it stop stress build up. It's the melting and freezing of metals
that cause stress, whether the melting is due to laser, plasma, electron beam or electric arc.

If you've got the money for one high power laser you could have five heat treatment ovens for less!

Craig

[peteeng]

Well Craig I have looked at these things and it seems they are the same relative cost to when I set up my shop 30+ years ago and bought high end TIG welders. The lasers are useful for cleaning, welding and cutting, seems to be a breakthrough tool really. Plus I'd want a metal printer... 5kW is a huge commercial cutter, 1kW is a good size... They do have interlocks so you can't light sabre them. A bandsaw is a messier cut then a laser.... I took a friend to emergency last week when he slipped with an angle grinder and tried to cut his hand off.. Lucky only 3 stitch's. Tools can be nasty.. Everyone out there clamp your jobs down, use two good hands on the tool and if its dicey don't do it!! Peter

[joeavaerage]

Hi peteeng,
for the price of a 1kW laser, and a good one, no Chinese rubbish, I could have a German made ATC spindle with an HSK25 tool interface. Such a spindle I could put to work right now and use daily
for years to come.

Craig

[joeavaerage]

Hi,
popped a fuel float out of the PTFE molds. No release wax or anything and it popped out perfectly.

The fuel float itself is not perfect, it has a couple of surface occlusions but is still vey usable. I may have to consider injecting the epoxy mix, and possibly also
put it under vacuum just to pull any occlusions out. I also need to give some more thought to the alignment spigots, I used 6 mm bolts with one half the mold threaded spigot holes
to align and clamp the two mold halves together, but inevitably you end up with epoxy in the threads. I will have to make the alignment holes go right thru and without threads,
it allows you to push any cured epoxy right out of the hole. Got to love PTFE!

Just tried floating it in water and it looks to have an SG of about 0.7, which will, I hope, still float in fuel with enough excess buoyancy to do what I need of it. Have to hope also that
the filled epoxy does not dissolve or otherwise degrade in fuel.

Still some refinement and research to do yet, but I think my idea will work. I have one customer waiting for me to do it.....and I think I could sell a dozen or more a year if they
work as intended.

Craig