[joeavaerage]

Hi peteeng,
Ah! This time the formula attached...kool.

Seems basic enough.

[joeavaerage]

Hi,
funnily enough some of my University books do get quite a bit of use. Some of the Electronics and System/Control texts get a work out, but the one that gets the most is my first year Physics
text. I reckon I'd consult it at least once a month and have done for forty years....which must make it one of the best purchases I've ever made.

The thing about CAD and computer simulation is that you can make a mistake and get absolutely the wrong results without knowing it.

The classic example was the Hubble telescope. When it was made they designed and built what amounts to an optical model......and then built the telescope. They checked repeatedly against the model
and it was nanometer perfect. They put the thing is space only to find it would not focus???? The model they had built had one element that had exactly double the curvature that it should have.
They had relied on computer modelling without doing a reality check. If they had done a Foucault test, an optical procedure as simple and old as telescopes themselves and beautifully sensitive and accurate
they would have realised the mistake.

When they realised what they had done they were able to design and build a corrective lens, after all they knew to within a nanometer what they did wrong, flew up there and fitted it. All the magnificent work
done with that telescope over the years was done with that corrective lens in place. Still it was bloody big and expensive cock-up.

The bottom line is that when I'm designing something with a computer I like to have some physical reasoning or measurements that prevent me from going of the rails and having a Hubble moment.
Its worthwhile to compare your solution with other solutions also. For instance if my calculated cross section is 3600mm² and yet Hass made the same thing in a similarly stressed part
of 10000mm².....it starts me wondering if I've genuinely understood the calculation. I am not averse to crosschecking with others.

Craig

[ardenum2]

I haven't seen a data sheet of any material that provided all the necessary specs, material properties are usually disseminated throughout the whole line of products, some grouts have one property defined other have others, its never the full picture.

the only reliable way to define a new material in CAD is to cast pieces of it in the real world, send it out to a lab for testing like Pete did and then enter the date from the tests.

[joeavaerage]

Hi ardenum,

the only reliable way to define a new material in CAD is to cast pieces of it in the real world, send it out to a lab for testing like Pete did and then enter the date from the tests.

Yes, you are correct, although it is a time consuming and expensive business.

I would argue that there is probably enough material data about to make at least some 'broad brush' studies of designs, materials and processes. I would have to agree that to nail down
a specific epoxy granite mix for example would require a sample to be made and tested.

It must be said that Fusion does have a good library of materials, but largely the more commonly known and used engineering materials, and they are my go to materials.

Craig

[joeavaerage]

HI,
I've just heard back from the company that does investment casting.

Due to the size and large flat faces, investment casting is not our recommended option. Large faces have a high chance of shell failure, and you also only get one shot per print and all the shell formation costs.

A sand casting would be our preferred option. We could try casting in SG iron, but it may require a couple of attempts to see if it casts successfully. It poses more challenges than grey iron mainly due to the higher shrinkage rate compared to grey iron.

Would grey iron work for you, as this cast in a sand mould would be our preferred option?

A rough estimate for this option would be about $900.00 plus GST.
This price does not include the pattern cost.

I provided the company with a diagram of a hollow pyramidal shaped part....and that sounds like it would be a problem for an investment cast. Sand casting is viable.
I was not aware that the same company did sand casting....although it would be a slam dunk if you already have the skill to do investment casting, perhaps I should have guessed they had that capability.

The foundry I am most familiar with and is only a few city blocks away from here has yet to come back with a price. I might go and see him, so we can talk through the different options.

Craig

[peteeng]

Hi Craig - Is the cast iron part to be heat treated after casting to stabilise it? annealed, normalised etc? or just a stress relief? Peter

edit - the front flat face would do well to have a couple of ribs at the back of it. The caster will guide on this. The flat surface will cup in or out and this is what the comment is about. This cupping & shrinkage can break a thin part in a fragile metal.

[joeavaerage]

Hi peteeng,
ideally it would be stress relived, but thermal stress relief is expensive, especially now as the local heat treater has closed and de-camped to Auckland.
My axis beds were done with vibratory stress relief. I can only report it seems effective.

The scholarly articles I've read suggest vibratory SR works best with thick sections, and the axis beds certainly fit that description.

For the current part (headstock) I would use the same process. If in grey iron I suspect the wall thickness will need to be about 18mm thick and the flange and nose about 30mm.
I think vibratory SR would be enough and very much more affordable by comparison to sending it to Auckland for thermal SR..

Craig

[peteeng]

Hi Craig - some materials do not respond to VSR very well. Since CI has slippery grain boundaries this maybe the case with CI. So speak to the VSR people about that. I thought the casters would have an oven. The CI out of the mould will have poor grain structure and maybe a bit unstable over time. Normalising or TSR is recommended. Peter

[joeavaerage]

Hi peteeng,

So speak to the VSR people about that. I thought the casters would have an oven.

My axis beds were done that way at the recommendation of the machining company. I have no means of estimating its success or otherwise execept to say my axis beds
have proven to be 'rock solid' as the saying goes, despite being several times more solid than rock!

The foundry does not have a heat treating oven, and even if they did the price of heat treating is something like two thirds of the cost of the casting. At the time thermal SR was $6NZD/kg.
Ideally you'd heat treat everything....in reality you have to choose what you can afford to do.

You made the case earlier that a fabricated steel headstock may not need SR by virtue of its shape and that there is no heavy metal removal required in subsequent machining
for use. I'm hoping that a cast part would be similar. That is to say, the little skim required of the base and the nose is the only machining required and that I would expect any part,
iron or steel, no matter how badly stressed, would not move as a result of such a light operation.


Craig

[peteeng]

The thing with steel is that's its stable. A green casting still has metallurgical aging and internal events happening. But your base proves the pudding. Peter

[joeavaerage]

Hi peteeng,
that may be the case but my axis beds have proven to be superb in every way. If a cast headstock performs similarly then I would be extremely happy.
Cast iron is 'as stiff as an old boot' but has such a nice damped vibratory response....there is a good reason that it has been so widely used and still is. If it were cheap enough
ALL hobbyists would be all over it. Cost is the only barrier.

Craig

[joeavaerage]

Hi,
as a preliminary guess I would say that thermal SR welded steel is marginally ($800 vs $900 casting) cheaper than cast iron, although not a lot in it either way.
If it were acceptable to forgo thermal SR then welded steel is very price attractive, about $350NZD.

If thermal SR is determined to be essential for cast iron then cast iron is effectively twice the price and is just too costly for my consideration.

It really comes down to the cost of thermal heat treatment. From the point of view of asthetics I prefer cast iron as it matches my axis beds, and I can say with some
little justification I am quite proud of, and if there is a small price difference my inclination is to go that way.

Craig

[joeavaerage]

Hi peteeng,
just noticed your edit:

edit - the front flat face would do well to have a couple of ribs at the back of it. The caster will guide on this. The flat surface will cup in or out and this is what the comment is about. This cupping & shrinkage can break a thin part in a fragile metal.

If I cast in grey iron the nose section will be around 30mm thick, and the walls about 18mm thick. I would not expect any great deformation over those thicknesses. I have deliberately avoided
the situation where you need to 'push the boundaries' of the process, that's exactly the situation where process faults can occur.

Craig

[pippin88]

For a part like this I'm not convinced stress relieving is needed.

The major issue with material stress is movement when machined. Data on long term movement in use seems to be lacking / may not be an issue.

Instead of stress relieving, you machine, unclamp, clamp lightly machine again. You might have to do a couple of cycles.

I'm pretty sure an non-stress relieved part with your planned dimensions will perform better and result in more accurate parts from the machine than your current headstock lever.

(I am guilty of chasing perfection also, a common issue of the non-professional / hobbyist. The professional knows when good enough is good enough and to get on with it, or when to spend the extra money).

Another option is glued and bolted and doweled steel plates (use hot rolled steel).

[pippin88]

Uniform or thick section castings normally have low residual stress, while thin or uneven sectioned castings retain higher residual stress


Experienced scrapers will tell you that hanging an iron straight edge and ringing it with a mallet will relieve stress (the casting will move measurably, and these castings are designed not to say / distort).

[joeavaerage]

Hi pippin,

I'm pretty sure an non-stress relieved part with your planned dimensions will perform better and result in more accurate parts from the machine than your current headstock lever.

That is 100% pure common sense. Firstly, like you, I think that this part having moderate wall thickness will harbor little residual stress, and even if it does......so what?. It's not like it's buried
in the machine where it can screw things up can cause everything to be misaligned. I can remove it at any time to send it out for thermal SR if it proves necessary.

You are also 100% about 'perfection being the mortal enemy of good enough'. This is a battle I've had again and again. As I've progressed in this hobby I've become better at sorting 'what I want'
from 'what I need' into two piles ' what I can afford' and the other 'you f****g plonker...you'll never afford that in a million years'.

My thinking is that a welded steel construction with thermal SR is about $800 whereas a cast iron part without thermal SR, but maybe vibratory SR, would be about $1000. Either would be
good enough, and easily better have what I have now, even if they are not perfect. A machined solid Al part is also viable even though its cost is a little higher.

I really need to investigate cast grout and epoxy granite, there may be a cost advantage that I can't ignore.

It seems there hobbyist lead move towards more sophisticated designs than a C frame, and different materials like epoxy granite and grout and others, and machines such as mine attract comments
like 'lacking in imagination' or 'low risk rather than innovation'. What does not seem to be appreciated is that I've actually built my machine, its been in service for over 18 months. Sure, it has shortcomings
such as its current headstock due to budget and/or time constraints when I first built it....but nonetheless its a working proposition. This project (headstock) is about me revisiting one of those areas that I was not able to do
well the first time. Also, I have finished a credible fourth axis and 90% on a trunnion fifth axis. This was always my hope, and while I could not afford it two-three years ago I can now, so I'm adding that
capability.

I take some satisfaction that I've actually taken my ideas and have made some real life hardware, however imperfect it may be, but its real, nothing hypothetical about it.

Craig

[peteeng]

Hi Craig - I think grout is the better bet then EG, so what grouts can you get? Lanko 701 ? any sika products? Peter

sika grout 3200 in NZ E=38GPa model that

SikaGrout®-3200 | Windfarms | Sika New Zealand

[TangentAudio]

It seems there hobbyist lead move towards more sophisticated designs than a C frame, and different materials like epoxy granite and grout and others, and machines such as mine attract comments
like 'lacking in imagination' or 'low risk rather than innovation'. What does not seem to be appreciated is that I've actually built my machine, its been in service for over 18 months. Sure, it has shortcomings
such as its current headstock due to budget and/or time constraints when I first built it....but nonetheless its a working proposition.
Craig

It's easier for people to comment on a forum than it is to build something. It doesn't matter if some armchair expert decides your project is not innovative or takes enough risk for their liking. You made something that works, you've learned many things along the way, and you're at the next project to improve its shortcomings. For a hobby like ours, it sounds like you're doing it right.

Keep at it.

[ardenum2]

It's easier for people to comment on a forum than it is to build something. It doesn't matter if some armchair expert decides your project is not innovative or takes enough risk for their liking. You made something that works, you've learned many things along the way, and you're at the next project to improve its shortcomings. For a hobby like ours, it sounds like you're doing it right.

doing it yourself only makes sense if you pay less doing it than if you bought the machine off the shelf or you build something that you can't buy for a reasonable amount.

you call spending $30,000 a hobby? don't make me laugh son

[TangentAudio]

doing it yourself only makes sense if 1. you pay less doing it than if you bought the machine 2. you build something that can't be bought for a reasonable amount. Craig paid $30 grand for a c frame that he could have otherwise bought off the shelf, there's a ton of machines that fit that price, you can even get second hand deckel mahos for that. $30 grand is anything but hobby, spending that much only makes sense if you build something that would otherwise cost $500 grand to buy. He said it himself he prefers to put innovation into his parts rather than the machine, so building it for $30 grand instead of buying was obviously a mistake.

There is no universal rule about how someone spends their hobby money, nor is your idea of a reasonable budget for a hobby the same as someone else's. People buy vacation homes, boats, race cars, and all sorts of other expensive things for hobbies. Does any of that really make sense from a financial perspective? I'm assuming he's enjoyed his project, so it's his call whether it was worth that much money for that enjoyment, not yours.

Obviously you have made it clear that you would not have spent that much money for the end result he achieved. I would have made different choices about spending $30K too. Everyone is entitled to their opinion, but I don't think anyone is entitled to tell anyone else their choices were a mistake. Obviously the internet is fueled by people telling others that whatever they did is wrong, so clearly I'm in the minority here.