Making lathe from scratch and using mild steel plate? Why use cast iron?

[PeterA]

Why are all lathes and mills made from cast iron?

Could I weld up a structure out of mild steel plate to use a a lathe base and frame? Then just face mill or fly cut the surfaces in a mill?

I think you would need to flame stress it right? So you weld it, machine it then flame de stress it. Then will it need to be machined again after the flame treatment?
If that is all that needs to be done, as opposed to using cast iron. I guess it is then possible to make a machine that would hold an okay tolerance then yes?

Any ideas how hot you have to get mild steel to stress relieve it? and for how long?

Peter

 

[steamer]

Welcome Peter,

Lots of good questions Peter. The construction of a lathe can be done many ways. The best for our purposes is to use gray cast iron. It absorbs and dampens vibration the best. Steel would be second. I've designed lathes that have been made of other materials such as granite!. We can wax poetic about the various ways, but for our shops, the first thing you need to answer is ....what is the problem we are trying to solve?
If your project is to make a lathe, by all means have at it.

If your project is to build a miniature engine....well...that's different.

In that case I would just buy one. The hundreds of hours you will spend building a lathe could be better spent learning to use it properly, and you'll get more enjoyment out of it. Additionally, at the cost of the Asian imports...I don't think you could make one as cheap as you could buy one.

OK. So where ya from? What are your interests? We like to know something about the people we bring into our home. If you could Peter, would you please post a quick post in the welcome thread and tell us about yourself. I know the members would very much appreciate it.

And again Welcome! wEc1

Kind Regards,

Steamer

 

[PeterA]

Right, intro post done.

Well I'm thinking that I can get an old mill rather than a new one and it will be cheaper. Then I could use mild steel to weld up a lathe
frame and machine it out. Then use the linear ways for the X and Y axis etc. The idea is to get the most from your dollar.

Still don't know if I will have any funds to purchase anything as yet so I guess I'm in the dreaming/planning stage.

So many people say the chinese units are rubbish, I read this online. I spoke to a few engineering shops today and they
say that the ones from HAFCO www.hareandforbes.com.au here in Australia are not too bad. Good enough to machine items for press fits
etc.

Anyway, then I think why buy new when you can get some good old iron....
Then I think what if it is clapped out...
Then I think get either a good new mill or a good old mill and make a CNC lathe using the mill.

Decisions decisions....

 

[steamer]

The question remains though


What is the problem you are trying to solve? The answer to that will set your course.


Regards,

Dave

 

[Davo J]

Hi Peter,
I bought my 300 x 900mm lathe from Gasweld (toolex brand) and after looking at the H&F's range afterwords, they are not as good in the fit and finish. Saying that plenty of people own them and are happy with them.

You should be able to pick up a used lathe from around $400 up if you keep your eye out and ask around. This would be much better than almost any home made lathe.
Some guys get lathes given to them, but they are the lucky ones, the rest of us have to save up.

If you are around the Newcastle area you are welcome to drop in and talk tools and have a look first hand.

Dave

 

[GeorgeGreek]

Dear Peter,

am afraid that Steamer is right in general terms without going into the specifics, i.e. the make of the main shaft housing, the necessary special conical prestressed bearings,the lining up of it relative to the prism or whatever short of ways you would decide to make, the many lathe errors that make the difference between a lathe and a useless contraption. I believe that a medium quality Chinese lathe would be much better and definitelly much cheaper than anything one could make, unless one aims to a small model without any demand on accuracy and such.

Have a nice time anyway,

George.

 

[rleete]

Where are you located? Plenty of machinists would love to show off their shops, and answer questions in person.

 

[cfellows]

As George and others have said, the chinese lathes out of the box are probably better than anything you can make unless you want to invest hundreds of hours. You can buy a Chinese lathe and with a little work, can make improvements to it and learn something about it at the same time. There are also a number of attachments you'll probably want to make such as quick change tool posts, carriate locks and carriage stops, etc. The popular 7 x 10 lathes have a nice size hole through the spindle which you will quickly come to appreciate. I have a Harbor Freight 7 x 10 and have turned it into a very nice, accurate lathe. I use it more than my Logan 11 x 24 these days. It comes with a variable speed, dual range, reversible drive. The modifications I made were to install tapered gibs on the carriage, tapered roller bearings and metal drive gear in the headstock, cam lever lock on the tailstock, and my own designed quick change tool post. I also made a couple of adjust tru chucks and now have less than .001" runout. I think paid about $350 at Harbor Freight about a year ago on sale and with a 20% off coupon. I might add, that the lathe was very useable out of the box and I was able to make the improvements at my leisure.

While the basic lathe is important, the tooling is of equal or greater importance. You'll likely need a number of different cutting tools such as boring bars of different sizes, thread cutting tools, cutoff and grooving tools, and a variety of facing and turning tools. You'll likely want several types of work holding chucks such as 3-jaw, 4-jaw, perhaps a collet chuck and set of collets, and maybe a faceplate. A vertical, adjustable vice is also handy for holding work and/or odd shaped cutting tools. For the tailstock, you'll need a fixed center, perhaps a ball-bearing center, and a drill chuck. Ideally the tailstock should have a morse taper socket, with #2 being one of the more popular sizes.

Building a lathe has always sounded appealing to me, but in the end, buying a commercial lathe has been the better decision, at least for me.

Chuck

 

[PeterA]

I'm in Australia, Qld, Toowoomba.

Yes I had wondered about the alignment between the bed and the headstock etc. Next question. How accurate is a face mill or fly cut to get
the surfaces square? Would it need to be then surface ground?

The head stock you would weld up, machine the bottom and even surface grind it? Then even get the head stock spindle hole cylindrically ground. Then
you would put in some high precision bearings to reduce the run out etc.

No that it can not be done, but I can see the implications everyone is saying. Here is a machine that was built with mild steel welded plates.
Looks pretty cool.

Video: [ame]http://www.youtube.com/watch?v=Y1IDHIcyHLY[/ame]
Build: http://www.cnczone.com/forums/vertical_mill_lathe_project_log/51688-slant_bed_cnc_lathe_scratch.html

So it can be done. I wonder what the accuracy on it is. I might have to read through the thread again.

 

[John Hill]

There is aYahoo group at http://finance.groups.yahoo.com/group/Multimachine-Concrete-Machine-Tools/?yguid=473405516 related to making concrete machine tools.

I dont know how practical or useful the idea might be but does have some appeal. I presume a lightweight welded steel structure would be more rigid when filled with concrete, I got plenty of advice about stability of concrete when I built my concrete machine bench both dire warnings and other advice that indicated whatever changes there might be would be beyond by ability to even measure them.

As for aligning the spindle to the ways, I expect there is some ancient and recognised procedure for doing this. Just thinking for a moment, maybe you could start with a small diameter spindle set up as accurate as possible then use that to turn a test bar adjusting until the alignment is right then using that same spindle to enlarge the holes to take the bearings, food for thought there?

 

[ShedBoy]

I have a hafco lathe and it is accurate as. I looked at the toolex site and all the parts look the same, gear levers, tailstock headstock ets but back to the question in hand anything is possible but I reckon a lathe built your way will cost a lot more than buying one. You still need to buy alot of parts that will need to be turned by someone on a lathe. I would buy one then make one as a project, this way you won't have to pay for turned parts, unless you can borrow a machine. Make sure you post the progress, I would like to see how it turns out.

http://www.shawmachinery.com.au/UsedEquipment.html#Lathes

Brock

 

[barney_leadhead]

There is a book "Building A Small Lathe' by L.C.Mason (TEE Publishing reprint) which is very useful for ideas on building a lathe and it uses lots of stock materials for the main components; however you still need access to a lathe to machine a lot of the parts that Mr. Mason describes and his design may well be too small for what you are planning.
Despite this the book is a thoroughly good read and is full of hints and techniques that may help guide you should you pursue the home construction route for your machine.

Again whilst small the Peatol/Taig lathe uses an aluminium/aluminum extrusion filled with concrete for stability that carries the ground bed.

Whilst on a very small scale the Fonly lathe uses a mini-drill to produce a lathe capable of high accuracy; given your lateral thinking the idea could be adapted and scaled up then used to produce the components for the lathe you want to build.
Just Google 'fonly lathe' for full details on the 2mm Association site.

 

[steamer]

That slant bed is pretty cool! and I bet a LOT of work.

But....if it solves the problem before you..it's perfect!

Dave

 

[ShedBoy]

I was trying to design a small CNC lathe one day and tried linear bearings for ways. Hardened round shafts and plates tacked together to bore all the important holes (spindle and ways) then seperate them and slide them on the shaft to where they are needed. Linear bearings are quite cheap through ebay from China. Don't now how accurate they are though. Doodling a shaper at the moment using fully supported linear shafts for the sliding bit. I recall seeing a website on the net of a chap in Queensland who had built his own lathe with linear slides, and they are used alot in CNC machines today. They are in use on that machine you posted the video of (that is what reminded me of my previous idea.
Brock

 

[Entropy455]

For marine vessels, it’s a common practice to fabricate machine components/foundations out of plate steel. When close dimensional tolerances are required, the components are typically stress-relieved prior to machining, at 1150 degrees F for several hours, then slow cooled.

If very tight tolerances are desired, the fabricated components are stress relieved, then rough-cut to near final dimensions, then stress relieved again, then machined to final tolerances.

If you weld a lathe bed together from steel plate, and machine it without prior stress relieving, you’ll have significant and unsatisfactory creep during the machining operations. The bed will not hold true. The results will be completely unsatisfactory. However if you properly stress relieve, you can obtain a nice lathe bed.

The next question becomes how to harden the ways. Mild steel cannot be appreciably hardened – there’s not enough carbon. It is too soft, and the ways will eat themselves after a few months of use. I suppose it would be possible to weld in some high-carbon steel to form the ways, stress relieve the assembly, machine it, then induction harden the ways, and then grind them.

It’s a whole lot of work – which is likely why the preferred method for lathe construction is to cast the bed out of a heat-treatable iron.

FWIW – there’s a foundry out in Washington State that does almost daily pours of ASTM-A536 nodular iron. This stuff is tough, strong, heat-treatable, and is reasonably priced. About one dollar per pound, the last time I checked. The pattern making fees “can” be a lot, if you have a complex casting – like an engine block with lots of internal passages. Nonetheless, the cost to have a lathe bed custom cast by a professional foundry is not as expensive as most people would think.

The down side is that it takes VERY large equipment to properly machine, induction harden, and grind the ways, on anything but the smallest of lathes – which is why I purchase my machines instead of making them from scratch.

 

[PeterA]

Okay 1150 F is 621 degrees Celcius.
I reckon I can get that with a few bottles of propane gas, yeah? Make up a bit of an oven and cook it for a few hours. Should be easy enough.

For the ways, I'd just bolt down and shim some hiwin type square block type ways????

 

[Entropy455]

The trick to stress relieving is maintaining the entire work piece at an elevated and uniform temperature, for several hours (12 hours is pretty standard). The heat causes the lattice structure of the steel to slip in a slow and controlled fashion, which is the mode of stress relieving.

It will be very difficult to maintain a uniform 1150 degrees F for 12 hours, with a propane torch. You’ll need an oven – and not the kind you cook with. You’ll need an industrial oven. Trying to maintain 1150 degrees in a home cooking range is dangerous to say the least. I’m pretty sure most ranges have a thermal safety device that prevents sustained operation at high temperatures. Another option is to use a pottery kiln, however most pottery kilns have difficulty maintaining steady heat at lower temperatures.

General stress relieving rules: never place the work piece in an oven hotter than 500 degrees F. Maximum rate of temperature rise is not to exceed 100 degrees F per hour. Cool down is limited to the same rate. Do not remove from oven until work piece is 500 degrees F or less. Always convection cool. Never force cool. Important note: the rate of temperature rise is based on the core temperature of the steel being stress relieved – not the oven temperature.

I’m not trying to discourage you. But you should know that rapidly heating up the piece to a dull red for a few minutes will not stress relieve the component. If anything, the high temperature gradient will further increase stresses. Proper stress relieving is a SLOW and controlled process, requiring precise temperature controls.

If you are serious about it, I recommend finding someone with a pottery kiln large enough to accept your work piece. Or take it to a heat-treating shop, and pay a few bucks to have it done right. Don’t try and use a torch.

 

[PeterA]

Yeah I was thinking of the pottery guys, there is a large mod that have a pretty big kiln they use to make pavers and fire bricks etc.

I will have to phone around and ask them.

I should ask the specialists too how much to properly stress relieve.

I should ask the foundry up town how much they can cast me a block iron for. I have heard they do one offs.

If I got them to make me a solid chunk of cast iron, then I machine it. Will machining it cause to twist and buckle? meaning I'd have
to get it flame treated anyway?

 

[Entropy455]

Hardening the ways only affects the iron near the very surface. The bulk of the casting remains unaffected by the heat treating process – thus distortion is minimal.

If you are having them make you a casting, then have them make you a casting. Simple molds are not that expensive, and there is no need to pay for material that you intend to machine away anyways.

Iron castings are normally fully stress relieved at the foundry, and undergo some sort of minimal inspections – visual inspections at the least. Depending on the contract, castings can also be MT inspected, UT inspected, and/or RT inspected. Normally the extra NDT is reserved for high-reliability castings only. Quality foundries will not give you a junk casting - they’ve got a reputation to maintain.

Even though the castings are stress relieved at the foundry, they can still creep if you hog out a lot of material. For example, if you machine a lathe bed from a sold chunk of iron, you’ll likely want to approach the final dimensions, and then re-stress relieve prior to machining for final dimensions.

If you cast it “close” to final dimensions, you’ll likely be just fine with the foundry stress relief – as is typically done within the industry.

Note: you will NOT be saving money on this project. It will most certainly be cheaper for you to purchase a complete lathe, then to build your own.

 

[PeterA]

Yeah the more I think about it just getting bought lathe seems like the more sensible idea.