[peteeng]

So if you design your current machine with a tall Z to accommodate a future rotary you have a less rigid machine for your current work. Better to use a pit so the machine is stiffer. Better again is to build the rotary now and get the benefit now. If you do not have an immediate need for the rotary then don't include it as it compromises the current machine (unless you include a pit) then you get through the CNC learning curve and build a future machine with your current machine with all the experience you have gathered in the mean time... After building 8 machines in 6 years (plus helping others along the way) I can say I've covered a lot of territory. Building two more at the moment...Peter

edit Re: plywood and flood coolant
I have considered plasticising the plywood with epoxy so its coolant proof. Lots of ply boats have been in the water for decades and they are OK. Plywood has proven to be a very good Maker level build material. I am concerned with change in size due to moisture ingress/egress but the important parts I stabilise using sheet aluminium to create ply/AL laminates. But metal is the go. Steel is the stiffest and most economical material to use in production. But design development for me has to be Al or ply... as I can't cut steel. I've started playing with ethanol mist coolant. Datron and others don't flood they use alcohol mist and air coolers vs flood.

[peteeng]

Hi Simon - Datron have a lot of info on their site about aluminium, they are AL experts. They mention steel in several of their documents and vids but they do not specify cutting data in their general docs. See attached, really good data. To achieve some of this performance you require very small runout tools and very high stiffness however. The runout on hobby machines can be as big as the required cut so it in fact does not cut or just rubs etc. Since steel is 3x stiffer than aluminium the machine needs to be 3x stiffer to hold the same tool deflection as the AL cut. Since most steels are much stronger than AL you need more torque to make the cut. Then steel work hardens alot more than al so during the cut you need more torque again to overcome the hardening and the material push back... It goes on, steel is quite an animal to deal with. Then there's stainless steel its even worse... The main hurdle with AL on a router is it gauls to the tool, dulls the tool edge picks up even more material, then chokes, then dies... very quickly. Recently I have started using DLC coated tools and they are brilliant on Al.

Technically if you know the chip thickness for aluminium it should be the same as for steel. This is because the max chip thickness is a function of the tool gullet size and its ability to clear the chip. But the machine has to be stiff enough (3x min) and true enough to hold the cut... Typically I start at 0.050mm chip thickness for an AL job...

Peter

https://shop.datron.com/ in resources download the cutting guide.

[joeavaerage]

Hi,

Then there's stainless steel its even worse...

Well, yes it is tougher but not hugely so. My target surface speed for uncoated carbide in low carbon steel is 100m/min. For austenitic stainless, namely the 300 series the target is 70m/min.
One thing that stainless does that gives it its reputation is that it work hardens like 40 b********s. If you are going to cut it...bore right in there, under no circumstances just 'give it a rub' with a machine
that lacks rigidity because thereafter you've just made it doubly hard.

High rigidity and plenty of torque is the ticket for steel. The rigidity goes up a little for stainless, another step again for titanium and even more again for Inconel.
Surface speed for titanium and Inconel is in the region of 25-37 m/min.

Craig.

[joeavaerage]

Hi,
I know you have a lot of things to think about already, but I'm going to add to the list.

Making way covers, or buying ready made way covers takes some thinking about. In particular the gap between the Y axis and the column is very important.
Firstly you wish to make this distance large so that any part on the table does not foul the column when the Y axis is at its closest to the column. The downside is that
the greater the distance the bigger and thicker the section needs to be in order to maintain rigidity. The rigidity is inversely proportional to the cube of the length,
so its far from trivial.

In addition to having a gap to allow for part clearance the way covers have to fit in there. If you have a small gap but a large travel you may require a stack or six or eight narrow
telescoping way covers. If however you have a larger gap then fewer wider sections are required. I gave this a great deal of thought, and I decided on quite a large gap. This in turn
meant that I have only three section of way covers. One section is permanently attached to the Y axis, one is attached to the column and is therefore stationary and that leaves just
one section that floats and actually telescopes.

I designed my own way covers (my sheet metal guy did the bending) and I did the rest, so a simple and reliable design was required, and has worked out pretty well in practice.
The gap is larger than I would like for rigidities sake......it's all a balance and compromise, and this is what I did. I'm not sure I would do the same again, but what would I do?....I don't know???.

I suggest you give it some thought also because I struggled with my first machine with way covers, and in the event is was all the swarf and crap that I could not protect the ballscrew
and rails from that wore them out. To be fair they were well used when I even got them...but not being able to exclude swarf and coolant effectively shortened their life.

Craig

[peteeng]

Hi Simon -Thats why I recommend a high rail design. Keep those things out of the muck... Peter

[peteeng]

Hi All - I looked at the Suttons cutting data and extracted a few things. If you have an ER20 collet so 13mm is the biggest tool, then at 6000rpm you have a Vc=226m/min and at 12000 you have Vc=453m/min surface speed. Surface speed is the speed at which the metals are rubbing. The more speed the more rub, the hotter it gets and the closer to surface failure you get, even with flooding. The tool manufacturers publish data like this. An Er16 collet is Dia10mm max and I like these... at 6000rpm Dia10 is 188m/min and at 12000 its 376m/min I run a lot of jobs at 12000rpm even in timber/plastic.

So the max surface speed for a coated tool in general steels is 220m/min ie the 12mm tool at 6000rpm. Don't think you'd cut as it would have poor torque. Run faster than 6000rpm and you will cook the tool.

In AL you can run to 600m/min so a 6mm tool which is quite common can run at 24000rpm and get to 450m/min which is quite good.

So in steel you have to run small tools (with a high speed spindle) to keep under 200m/min so your material removal rate (MRR) is tiny... Thats why we need big tools at slow RPM to get the MRR up in steel... So comes down to what you want to machine in steel, how many days you want to run the machine to wear away the job and the tools (vs hogging) or if your just engraving then Happy Days... Peter


Catalogues | Sutton Tools

FSWizard: Speeds and Feeds

Last time I went thru this exercise I decide to get a 6000rpm AC servo so I'd get 1000-6000rpm with a small overload overhead as a direct drive or a 3000rpm with a 2x pulley so I could play with the pulley up to 9000rpm if needed. I also looked at using a high speed spindle with a speed reduction but it started to get silly...Peter

[peteeng]

something like this with a pulley to a servo. I was going to use a 750W or 1.1kW servo....

[Azalin]

err, why this spindle? They sell BT30 cardridges for the almost same price. I got one last month. The quality is very decent.

[joeavaerage]

Hi,
the real problem with spindle cartridges (built in motor) is that they have so little torque, unless you are talking a huge spindle.

I have just pulled the trigger on a spindle off Ebay. It's a new Dake 1040. It is 3.5kW continuous, 6kW peak, 10,000rpm (rated) , 40,000rpm (max) withe an HSK32 ATC tool interface.
I paid $2000USD for it, and the guy selling it put in ten HSK32 to ER20 toolholders and that included shipping to New Zealand. I've had notification from Fedex that its on its way and they expect I'll
have it by 28/12....and I only bought it on the 22/12. Must have cost the seller a fortune! I would have been more than happy even if it had taken six weeks to get here (from Vietnam), I mean its
going to take me that long to save up for a VFD to run the damn thing!!!

But as good as it is, or at least as good as I hope it is, given that I do not have it yet, its rated torque is 3.34Nm. Don't get me wrong that is ten-fold more than my current 800W 24000rpm spindle,so
a big step in the right direction, but 3.34Nm is nothing to write home about. I expect it will do 6mm and 8mm comfortably in low carbon steel, but 12mm and 16mm only by taking very light cuts.
I have a servo driven spindle I made some years ago now that I swap to when I'm doing steel and stainless. It is 6.1Nm continuous and 18Nm peak, 1.8kW (rated) an 3500rpm. So it is (by comparison)
a low speed high torque spindle. It does 16mm in steel OK, but with care, ie you can stall the spindle by taking too agressive a cut. The bottom line is that you need huge torque to drive moderate sized tools in steel.

Ideally you'd want about 2000rpm and 20Nm (4.2kW) to drive a 16mm uncoated carbide tool in low carbon steel with aggressive cuts. That's a lot of power and most hobbyists cannot afford such large and heavy
spindles and you are well into three phase territory.

The spindle that peteeng linked to is a price compelling alternative. You can power it with a servo, or with a regular induction motor. An induction motor is likewise very price compelling.
I would guess such a spindle and a 2 or 3 hp three phase motor driven by a VFD is just about the best low speed high torque combination you can get ....on a budget.

Craig

[Azalin]

The one I got has no built-in motor. It's just a BT30 atc cartridge. I use it with a Schneider 1.3kw 4500RPM servo. A little slow but machine is a mill. I built it for steel so no need crazy speeds. For aluminum or wood I mostly use my 2.2kw 24k RPM router.

Sent from my MI 5s Plus using Tapatalk

[joeavaerage]

Hi,

So the max surface speed for a coated tool in general steels is 220m/min ie the 12mm tool at 6000rpm

Uncoated carbide in steel at 220m/min????? Thats too high by a factor of two. Uncoated carbide in mild steel is 100m/min. Coated carbide a bit more, say 125m/min. Even then in both situations
you will require flood cooling, otherwise your carbide tool will turn red hot in a few seconds and thereafter it's stuffed.

I run a 1/8th tool (3.125mm), a good quality (Destiny Tools) tool with a good coating at 15000rpm in mild and med tensile steel. This is 148m/min. Without flood cooling the tool is wrecked within
10 seconds, with plenty of flood cooling I can get two-hours continuous use. These tools are $20.00USD each at discount rates. At 220m/min they would last a few minutes at most.

Have you actually tried cutting steel with uncoated carbide at that speed?? I have and I can't get within a bull's roar of it, 100m/min; yes, 150m/min; yes, but with reduced tool life, 200m/min; yes but
tool life measured in minutes.

Craig

[joeavaerage]

Hi, Azalin,

The one I got has no built-in motor. It's just a BT30 atc cartridge. I use it with a Schneider 1.3kw 4500RPM servo.

Oh yeah, that sounds like a good combination. I've used a few Schneider servos on customer machines, and they are very nice indeed. They are quite expensive in New Zealand so
I don't buy them for my own use, but a sure do like them. I used a second hand 1.8kW Allen Bradley servo for my 'steel' spindle. I built in myself and its manual change. I used a RegoFix ER25 toolholder,
cost a packet but it sure is nice.

Now that I have bought this new spindle, the impetus to complete another long running project of mine is likely delayed.

Years ago, I bought a remanufactured Vickers servo from circa '95. It's about 3kW, 12Nm continuous and 48Nm peak and 4000rpm (max). It was my intention to turn it into
an ATC spindle. It has a resolver rather than an encoder, and so finding an affordable servo drive for it is near impossible. As electronics is my thing I decided to design
and build my own Field Oriented Control drive for it. I had a rough prototype working as of six years ago. Since then I've sold my home, shifted, bought a business, built
a new machine....and so this project has been in abeyance for years.

I will probably use a BT30 or better a BT40 tool interface, and I'd now be looking for a gear or belt reduction to bring the max speed down to 2000rpm but increase the continuous torque
to 24Nm, and 100Nm peak. Should be just the ticket for face mills in steel and rigid tapping to maybe 16mm or even 20mm.

Once I get this new spindle up and running and explore its working envelope I might resurrect this high torque spindle project. For my business I really have need of a high speed spindle around 24000rpm,
but am hoping for major productivity gains by going to HSK32 tool interface. This is the priority, it is after all how I make my living.

Craig

[Siddvo]

I know the high rpm spindle isnt the best choice for steel but im planning on mostly cutting aluminium so im ok with using smaller endmills in steel as i dont need a high MMR.
Maby the goal will change in the future but i think im going to stick to a high rpm spindle for now.

Hi,
I know you have a lot of things to think about already, but I'm going to add to the list.
Craig

Im going to ,make bellow Way covers, or whatever their called. I already tried it out with a Plastic/rubber sheet I had laying arround, It works quire well.

[joeavaerage]

Hi,

Maby the goal will change in the future but i think im going to stick to a high rpm spindle for now.

Good idea. Most of my machining happens with my mall 800W spindle. I'd use the high torque spindle for an hour but use the little spindle for ten or more hours.
Additionally, they are pretty cheap and certainly good value. Bang for your buck they are superb. I would recommend a VFD at least one cut above the el-cheapo Chinese, if not two cuts above,
they really are pretty poor. You are much more likely to blow your spindle up with an el-cheapo VFD.

I use Delta VFD's and pretty much always have done. They are Taiwanese but probably made in China. They are about twice the price of el-cheapo, but they are worth it. I've been using a little 750W
one for 13 years. I've bought and supplied customers ten or a dozen or so over the years....never a fault with any of them.

If you've seen the pic I've already posted you'll see that I need another VFD. It needs be 460VAC capable, and I need 1333Hz to get to the top speed of the new spindle I've just bought.
The Delta MS series of VFDs have high speed models (to 1500Hz in V/F mode only). A 5.5kW or 13A model is about $500USD including shipping to New Zealand. There are plenty of the standard
models (600Hz) locally, but the high speed models are a little rarer.

Craig

[joeavaerage]

Hi,

More and more people are using HS spindles on steel as they get an understanding of HSM technique and the correct tool coatings

HSM stands for High Speed Machining....but what it DOES NOT mean is that you can spin the tool faster. HSM refers to the toolpath, not the spindle speed. In particular HSM uses trochoidal toolpaths
that present an optimum material load to the tool. A lot of short high-speed toolpaths in a continuous and flowing motion. The net result is often faster metal removal and therefore decreased
cycle times and increased profits.

The spindle speed is fixed by the material of the tool, be it HSS or carbide or ceramic or whatever, it coating, and the material which its cutting. It is independent of the toolpath and really represents a thermal limit
to the tool/material interaction.

Consequentially the torque/speed issue that we have been discussing is not affected by HSM, or to put it bluntly, HSM is not going to rescue you from a spindle which spins too fast with too little torque.

Craig

[peteeng]

Hi Craig - I agree with all the above except the bit about high spindle rpm. We are in a place where coatings play a huge role in metal removal. Normal HSS and carbide have very high friction co-efficients in the order of 0.7 to 1.0 dry... New coatings can be as low as 0.1 dry. Studies have shown that metal cuts so fast that flood lub is not in the cut so in fact we are cutting dry. Lubs require hydrodynamic lubrication to work best and at the cutting face this is not happening... So heat generated by the cut is a function of pressure (or force if you like ) and friction. Since the pressure is a function of the mechanics of the cut (shearing metal and creating a chip) we can't do much with that. But if the coating and edge sharpness can be improved to lower the friction then we can increase the speed yet maintain the tool surface temp.

Simply put if you have a good working chip thickness you can double the feed rate and double the rpm, you have changed nothing except you need more HP (power = force x velocity). But you have doubled the surface velocity therefore doubled the heat at the surface and since metals are relatively slow at heat transfer the temp spikes and the surface fails. But if the tool friction goes down a factor of two your in the same place.

https://www.corehog.com/resources/tool-coatings

So if this process is linear and we have a HSS or carbide with a Fc=0.8 then move to a DLC with a Fc=0.1 technically we can spin the tool 8x faster and be at the same thermal place.... even if we can speed up 4x this places us into the AC spindle world. Introduce minimum lub systems using ethanol (so we are using the evaporation of the liquid to remove heat ) and we are a bit better off as well. I'm surprised by the performance improvement of DLC tools on aluminium to "normal" tools. Dry cut, no pickup and good edges. I'll have to try some steel next....

Peter

this does not include discussing chip removal that's another story... high pressure air or "coolant" are the solutions for that.

[peteeng]

Heres a really good article on HSM - Peter

A Machinist’s Guide to Trochoidal and Peel Milling – Make It From Metal

What High Speed Machining Is And How You Can Do It – Make It From Metal

How to Machine Aluminum: All You Need to Know – Make It From Metal

[joeavaerage]

Hi peteeng,
have you actually tried it?

Friction is only a small part of the heat generated. The chip is being deformed until its ductility is exhausted and the tool applies sufficient force to exceed the ultimate tensile strength
of the material and the chip breaks. All that while the tool is doing work on the chip, and the chip gets hot.

You've got a spindle. Go and fit a small diameter tool, say 1/8th or 3mm. Try cutting a 1mm depth slot in steel, just take it gently but with the spindle rpm such that the surface speed is 100m/min, 10600rpm.
Then try the same path at 150m/min, 16000 rpm, and then if the tool has not fried up try it at 200m/min, 21000rpm. It a simple experiment, without good cooling it will wreck the tool, in fact even with good
cooling the tool life will be minutes at most at 200m/min.

PM me an address and I'll post you some 1/8th Kyocera four flute uncoated tools. I've still got half a dozen left, and these are fairly cheap, about $5USD each. You can experiment with those.
I have coated 1/8th Kyocera (SGS) tools, but they are about $10USD each and I have four Destiny Tools 1/8th four flute Raptors with a 30 thou radius. They go real well in steel but are $24USD a piece.
I use the uncoated 1/8th four flute tools in plastics, aluminum and brass....but unless you slow the spindle down they just don't last in steel.

I've been CNCing for quite a while, and I have tried every possible way I can think of to cut steel with high speed spindles. It is possible but there is a very distinct upper limit on surface speed, you'll know when you've exceeded
it when a 'little red light turns on' on the very tip of the tool....ie it goes cherry red and its stuffed. Leave it running for another ten seconds and it'll break. I've seen it happen dozens of times. If the coolant is interrupted
or misdirected for more than a few seconds at 150m/min you are in danger of yet another 'little red light failure'.

Craig

[peteeng]

Hi Craig - I did neglect to mention the internal heat generated via friction in deforming the chip. But that heat (should) be taken away by the chip and some of the coolant, ideally it stays in the chip. We are mainly concerned with the tool surface temp ie rubbing of tool on the surface. Thats why I didn't go for 8x speed only 4x speed. I will try eventually. I have DLC 6mm and 8mm tools here to try... My main aim at the moment is cutting composites and aluminium effectively on my router Scoot. I did a small job today using the Suttons tool figures and a 6mm DLC tool quite aggressive plus dry and it flew through the material with no pick up. Doing this with a HSS tool would have killed it in milliseconds... The worksheet is in the shed but I think the mm/tooth was 0.061mm (18000rpm/1100mm/min and surface speed 340m/min) If I want steel machined I send it to a machinist. My stuff is development work for myself. If its production I send it out to a big shop with big machines.

I've run mills as an operator commercially and have seen your red glows before... I'll digest what you have written here. No need to post me any tools I can get those if needed. I think my next development project is more along the lines of a large scale (5x5x5m) 5 axis router for composite trimming... I have to learn up on diamond & PCD tooling. The makers of the diamond tools just say run it slow, not helpful.. Peter

Hi Simon - hopefully we are not highjacking your thread!!

Hi Craig - I looked thru Suttons data and free cutting steels max out around 220m/min. Mild steels are around 150 as you say (none coated tools) coated up to 200-220 so that's the limit... Unfortunately they don't have DLC in their range... I'll try to find out...

[Azalin]

Hi, Azalin,



Oh yeah, that sounds like a good combination. I've used a few Schneider servos on customer machines, and they are very nice indeed. They are quite expensive in New Zealand so
I don't buy them for my own use, but a sure do like them. I used a second hand 1.8kW Allen Bradley servo for my 'steel' spindle. I built in myself and its manual change. I used a RegoFix ER25 toolholder,
cost a packet but it sure is nice.

Now that I have bought this new spindle, the impetus to complete another long running project of mine is likely delayed.

Years ago, I bought a remanufactured Vickers servo from circa '95. It's about 3kW, 12Nm continuous and 48Nm peak and 4000rpm (max). It was my intention to turn it into
an ATC spindle. It has a resolver rather than an encoder, and so finding an affordable servo drive for it is near impossible. As electronics is my thing I decided to design
and build my own Field Oriented Control drive for it. I had a rough prototype working as of six years ago. Since then I've sold my home, shifted, bought a business, built
a new machine....and so this project has been in abeyance for years.

I will probably use a BT30 or better a BT40 tool interface, and I'd now be looking for a gear or belt reduction to bring the max speed down to 2000rpm but increase the continuous torque
to 24Nm, and 100Nm peak. Should be just the ticket for face mills in steel and rigid tapping to maybe 16mm or even 20mm.

Once I get this new spindle up and running and explore its working envelope I might resurrect this high torque spindle project. For my business I really have need of a high speed spindle around 24000rpm,
but am hoping for major productivity gains by going to HSK32 tool interface. This is the priority, it is after all how I make my living.

Craig

The person who sold me the Schneider servo knew nothing about servos. He tried to turn the shaft but it was stuck then he said "I can't guarantee this servo works or not". So asked something like $50 for it.

The shaft didn't turn because the brake. :d

However the servo came with an advanced digital encoder with it and my ABB servo drive didn't have the function to read this encoder. I don't remember what it was. Endat maybe? So I replaced the encoder with a digital magnetic incremental encoder. I also removed the brake from the motor. Works perfect.

If I remember correctly Siemens made some nice high speed 230vac servos some decades ago. Hard to find but if you do they are perfect for mill spindles.

Sent from my MI 5s Plus using Tapatalk