[Apples]

What are the capabilities of a CNC large bench mill?

What kind of accuracy are we able to manage with a converted RF45 size machine?

I think a CNC bench mill would be better than a manual knee mill?

[recordingwhiz]

FAirly broad question, It depends on your project . can you elaborate?

G

[jid2]

A CNC RF45 is SOO much better than a manual knee mill, if you need to do anything beside "square" machining.

You just have to be willing to invest the time to do the conversion. Holding +/- .001 is very doable, better than that is possible, but you need to know what you are doing.

[TiagoSantos]

What he said ^^

My machine is accurate to within a couple thousandths.. Maybe 0.002", maybe 0.004".. If I were to do it again, I could do much better than that, but the outcome of these conversions is mostly a reflection of your skill and care as a machinist/mechanic/engineer. I'm better now than I was when I started, but still lack the patience and some of the skills to really turn out a perfect machine.

If I knew how much money and time I'd be spending on the conversion, I probably would have worked some overtime and bought a turn key mill

[LongRat]

Getting better than 0.001"/0.025mm accuracy on a cheap benchtop retrofit is difficult. You need skill in the conversion and cash to buy quality parts such as screws and bearings. For the more accurate stuff it becomes hard to justify when you start from such cheap Chinese iron - usually better to buy a higher quality machine from the start. Plus, accuracy higher than this isn't needed for most peoples' hobby projects, so these reasons are why you'll see few machines of this type more accurate than this.

[Apples]

Well with the bench mill if going that route I would get HIWIN type ways installed and rolled ballscrews. These parts will not be of any high tolerance, as I would just be purchasing them of ebay. One seller is that linearmotionbearing2008 (or a similar name to this), I do not know what this sellers products are like. Maybe these products will hold a tighter tolerance than the standard dovetail ways, I don't know?

Yes my projects are not for any production runs, but rather to just tinker around in the shed etc. Still having a cnc mill I think would make it so much quicker and easier even for simple tasks. Using Mach3 and some simple wizard programs, how easy would it be to drill 4 evenly spaced holes around a circle, and all that kind of stuff.


But the bench mills are small, considering I am used to seeing larger machines. But I figure you take a heap more smaller cuts with a bench mill under cnc control than compared to a manual knee mill, and you will still finish quicker with the smaller mill....Well that is my thinking.

But then I think hell, why don't I just weld up a big structure out of mild steel, then encase the whole lot with that epoxy granite / mineral casting. I don't know if I would still then need to post heat treat the welded structure. The linear rails would simply be bolted down and then packed up with shims to get alignment etc. I think the issue with welded structures is movement when you start to machine material off.

So it's like well if you use a bench mill, you need to machine the base to fit linear rails + more machining work for other areas = $. And the whole structure is to that big/strong/rigid. But if possible to DIY cast at home and for that same price as buying a bench mill you could make a much bigger stronger machine.

[LongRat]

If you only want to do easy things like basic hole patterns and wizard stuff, all you need is to retrofit the chinese machine with screws and cnc gear. Forget the linear ways. Way more effort for no benefit if you don't need it. And might actually be LESS rigid than dovetails.
I would not suggest building your own welded machine, it would be a much longer project than you might at first think, and accuracy and rigidity would probably come nowhere near a cast iron machine. If you don't have access to a much bigger machine it will be basically impossible to make anything decent.
Lots of people have used the Linearmotionbearings Ebay parts, myself included. The general consensus is that the parts are good quality, and you really can't argue with the prices.

[D.L]

" why don't I just weld up a big structure out of mild steel?", I started down this path and wound up with a chinese mill when common sense prevailed.
Steel is funny stuff, it weighs more than Iron, hit it with a hammer and it rings like a bell.
When you spend a good deal of time on a machine tool, you realise these vibrations are a penalty not worth the headache, it becomes a nightmare and you just end up pushing blocks of wood and aluminium against the frame to keep it all from building a resonant buzzing.

Trust me, Iron is far wiser, look into 'continuously extruded cast iron', the answer lays there, but you need a mill in the first place to true it all - catch 22 !

[Apples]

Yeah that stupid catch 22 thing.

Have you not read up a bit on the epoxy granite material and its damping properties that are something like 6-10 times more than cast iron? It's one thing to make a machine just out of mild steel. But I think it's totally different if you cast that steel frame in the epoxy concrete.........

[D.L]

I agree concrete will dampen, but i am yet to see a class act machine from fabricated steel utilising it, maybe it's been done and i have not seen it on my radar.
If you start a build log, i will be watching with interest.

[109jb]

Bench mills like the RF45 and similar can be very capable, but the envelope will never reach what a large knee mill can do. For example, I have had V8 engine blocks on a knee mill standing vertically. I have rotated the ram around and extended it out on many occasions for different jobs. Most bench mills can only have an x-y envelope directly over the table, but extending the ram or swivelling it on a knee mill allows the tool to be way off the confines of the table. Maybe you don't need all the capabilities of a knee mill, but there are still things it can do that a no bench mill can.

That said, I have had few occasions that I needed more that what a bench mill can do and I am a tinkerer that works on many different type projects and am looking toward upgrading from my G0704 to an RF45 type machine myself.

After my experiences with the G0704, I don't see a need for linear rails for hobby use unless that hobby REQUIRES way out of the box precision and accuracy. The dovetails are fine in my book.

[Bob La Londe]

Yes my projects are not for any production runs, but rather to just tinker around in the shed etc. Still having a cnc mill I think would make it so much quicker and easier even for simple tasks. Using Mach3 and some simple wizard programs, how easy would it be to drill 4 evenly spaced holes around a circle, and all that kind of stuff.

YES! and NO!

An example.... I could make a simple distribution manifold for something like coolant and coolant lines by whacking off a piece of aluminum on the saw, drilling some holes on the drill press and hand tapping the holes. No big deal. If I took some care to get things aligned roughly It might take me an hour or so. If I do it on the CNC mill it still takes me an hour or so, but I still have to tap those tapered pipe threads by hand afterwards. Before hand I have to draw the part, and then do CAM to generate the code to cut the part.

My CNC part will be prettier, everything will line up "pretty good", and I can add cool features like a clamp type mounting bracket that would not be so easy (but not impossible) by hand or with other tools.

Faster and easier can be a real toss up. If I needed to tap a hole quick and dirty its much faster to throw a flat part on the drill press and punch through with a tapping head. For an irregular part or a part I can't move easily to put it on something, I've been know to put a machine tap in my cordless drill. If I had a bunch of identical pieces with identical size holes it could be faster mount the tapping head on the mill, and hit run. It might be worth the time to setup mounting that was fast and repeatable (after its setup).

It really depends on what you plan to do with it.

I can't imagine how many little gizmos and widgets I have made over the years by hammering and grinding.

(Yes I know that technically widget is an economic abstract to represent for presentation purposes a real tangible product.)

[handlewanker]

Hi, anyone who thinks just making something out of cast iron automatically gives you a superior machine build does not know cast iron.

If you want to see what a steel origin machine build can achieve, Google the UMC-10 build by David Dcaussin on UTUBE.

Steel is a practical material for machine building and cast iron has it's place too.......the design determines what the outcome is.

Perhaps it would be better if a proposed purely welded steel design was posted in simple sketch form so that the various features can be examined and challenged for viability to resist cutter deflection etc.

That way anyone anticipating using steel sections without the knowledge of FEA or any other design analysis can find out what will work and what won't.

Just for starters......supposing you wanted to build a column mill with an XYZ work envelope of 500mm X 200mm X 300mm and the column was going to be made from 150mm X 150mm X 8mm wall square steel tubing, 1 metre long, welded to a 20mm thick steel base plate that was intended to be bolted down to the Y axis base.

What factors would come into play to make the column design viable or a non starter?
Ian.

[D.L]

Got that Ian, Scratch Built mini version of Fadal's VMC 15 - YouTube that's impressive stuff.
I've looked into FEA, mind boggling stuff! It wanted an algorithm in order to give a result, the stuff of scholarly people.
I like your veiled reference to the 150 x 150 x 8 tube column, I have alluded to such a design in a previous thread.
And my concerns were not for the rigidity, more for the iron base cracking given that 9mm wall section SHS has probably 3 times the load bearing capacity.
Anyway, that's a different thread I'm referring to.

Apples - could you please give a rough scribble on some paper and take a snap with your phone, and post please? i'm keen to see what you propose.

[arizonavideo]

You can do this with a IH mill no problem.

IH-facemill-bluechips115.MP4 - YouTube

This was before it was CNC'ed, just a power feed which I have to sell if anyone wants it. IH/RF-45 POWER FEED FOR SALE $100.

The real problem is when you want to hold .001" all day. Any of the larger knee mills will be more rigid but unless you want to CNC the knee the benchtops are a good tool.

If I had a good 2,500 pound knee mill I would just CNC it with standard parts and a nice big spring on the knee but if starting with nothing a true RF-45 or clone and you can be up and running for about $3000.

After the CNC job, this cut was .160" not .250" like I said.

IH mill .160" thou 100 IPM.MP4 - YouTube

[handlewanker]

Got that Ian, Scratch Built mini version of Fadal's VMC 15 - YouTube that's impressive stuff.
I've looked into FEA, mind boggling stuff! It wanted an algorithm in order to give a result, the stuff of scholarly people.
I like your veiled reference to the 150 x 150 x 8 tube column, I have alluded to such a design in a previous thread.
And my concerns were not for the rigidity, more for the iron base cracking given that 9mm wall section SHS has probably 3 times the load bearing capacity.


Anyway, that's a different thread I'm referring to.

Apples - could you please give a rough scribble on some paper and take a snap with your phone, and post please? i'm keen to see what you propose.

Hi, the square tube column in naked form in my opinion has no chance in a mill build......true it's steel and strong, but that is relative.....once you anchor it it's base plate it becomes an open ended box structure with no integrity to resist deflective loads or more importantly resonance factors.

It was just a starting point to work from in order to generate a thought pattern, food for thought etc, but in no way can it be a design feature even if it's relatively simple to acquire and work with.

The most important factor for a column design is the aspect ratio, that is the width/depth of the cross sectional area in relation to the height, and in this I would suggest a factor of 4:1 would be the minimum to work to.......IE a column 200mm wide X 300mm deep and 1 metre in height.....just for starters.......the 150mm square tube fails on all points from it's aspect ratio alone.

Welding would be the main course for construction and fears of locked in stresses are only valid if subsequent metal removal enables the locked in stresses to pull previously machined faces out of alignment, and this can happen when a structure is subject to a load that acts in the same direction as the stresses and enables the structure to move or act like a spring.

Cutting tool loading does not come into the equation due to the many factors that would vary the figures from speed, feed rate, material hardness and cutter geometry......the addition of coolant throws any figures totally out of the window.

When all else fails the best solution that is very hard to beat is to make it strong by overengineering wherever possible......somewhere in that equation lies the ideal build that would be just strong enough to not fail and yet economic enough to not cause heart failure by your accountant.....the ideal build also is derived at by expertise and so costs a bomb to recover the R & D costs.

By building right from the ground up to be a CNC orientated mill means that there are no throw away parts that normally occur when a retrofit is thought of.

BTW, no need to scribble a design on paper and then photograph it with your phone.......just sketch it with MS Paint and save it as a Jpeg file like a photo.
Ian.

[Apples]

Arazonavideo, yeah that machine of yours looks pretty cool taking those cuts. I did see those videos ages ago. I would be happy if I could cut that deep in mild steel (.160" is 4mm). I guess a lot of this just comes down to Hp right. If you put a 10Hp motor on it would you be able to cut twice as deep at the same feed rate? Regardless of you machines strength?

If I had a good 2,500 pound knee mill I would just CNC it with standard parts and a nice big spring on the knee but if starting with nothing a true RF-45 or clone and you can be up and running for about $3000.
They say you are better of finding a used cnc mill to fix/convert etc. That is the key here "if" you can find a good used machine. Let alone a cheap used machine that is decent.


4:1 ratio 200 wide, 300 deep, 1000 high.
In my head I was thinking the bigger the better ie over engineering. I think having the z axis column rails further apart should make the design stronger. So instead of being 200 wide, how about 600wide. Same as with the x and y axis, have the linear rails wider apart.


Regarding cast iron vs mild steel and damping, vibrations resonance etc. If you take a bench mill made from cast iron, and you want to have the same amount of damping, vibrations etc but made with mild steel. How much more material do we need to use. This is the over engineering part.
I don't know what the figure is, but if mild steel is 5 times less damping than cast iron, would that then mean if making it from mild steel you would need to make it 5 times greater in volume, thickness etc
I'd like to know how much bigger this would be. Lets imagine a bench mill (rf45, IH mill etc) is made from a length of 150 x 150 x 1000 SHS that is 8mm thick, from cast iron. Then what dimensions would that same structure have to be in mild steel to have the same damping and vibration characteristics?

Would it have to be 150 x 150 x 20mm thick or only 15mm thick? etc.

[handlewanker]

Hi, you also have to have proportions in the equation.......increase one area and it impinges on another.

Making a column 600 wide to have rails on it at 600 centres would probably not work due to the aspect ratio of slideway width versus the length.

Imagine how big the head would be if the column was 600mm wide.

Your first task is to draw your work envelope as a box and design the machine around it with slides and support infrastructure in proportion to carry the load.

When you design for CNC you are making something that deals a "death by many blows" as opposed to one big chop.

A Bridgeport mill may make fairly large cuts with big end mills, and it's massively built to do so, but this is not how CNC machining is envisaged.

This is very debateable, and if you want to design to Bridgeport proportions using steel sections then you have lost the plot completely.

First decide on the job size and then on the machine size.

My ideal desktop CNC mill size would be to suit a job with a work envelope of 200mm x 200mm X 200mm in steel.....anything bigger and you end up with a Bridgeport, and even that would struggle to significantly cut down a solid block of steel 200mm cubed.

It really means the work envelope can cover a job that is 200mm X 200mm x 200mm at max slideway travel in all axis, and that means the Z axis needs to be able to raise high enough to enable drill chucks to hold drills that could be 200mm long, so the column must be proportionately high enough with a base area big enough to carry the load without deflecting.

In most cases the slideway travel proportions get modified as the compromise of work envelope dictates.

One of the most common mistakes is to anticipate the table length that might be needed to cover all future possible job lengths, whereas the majority of the time the length cannot or is not used.

Many times the largest type of work is done in the machine vice and this is probably in the region of 160mm wide.

However, if a 4th axis is mounted on the table you do need to anticipate the table length to support the job type it will hold, and that could be a fairly long shaft needing tailstock support, but how many time would that occur as opposed to the number of short jobs only needing the 4th axis with a chuck.

Table length will make or break a build purely because it is difficult by any standards to make a Tee slot table without a heavy mill.
Ian.

[handlewanker]

Hi, the only reason mainly a machine is made from cast iron is because once you have made the patterns you can make hundreds of identical machines from them, and the cast iron was ideally suited to slideways by nature of the graphite in the casting.

I worked for a firm in the 80's that made machinery for the car industry and all our builds were for linear transfer machines, fabricated from welded steel sections and machined to accept hardened and ground slideway modules etc.

One machine, used to produce the Holden V8 cylinder head inlet manifold, had 21 stations that drilled, tapped and milled from the raw casting to the finished manifold in 4 minutes.....the casting moving down the line in 21 steps each 4 minutes.

You don't need cast iron if you are going to use linear slideways for the axes or are building just one machine.

Forget about the damping factor of cast iron.....for a steel build it won't become a problem if the sections are proportioned correctly.....you don't need 5 times more steel just to replace iron etc.
Ian.