New Machine Build- How to calculate cutting forces

[Swiss]

Hi all,

I would like to build a new CNC milling machine for my company, and I'm currently looking at the various options for axis motors and spindles. One thing I really need to know, in order to spec the axis motors correctly is:

How much force do the axis motors need to provide?

Is there any way I can calculate or estimate this? We are going to be cutting aluminium with an end mill, max diameter 6mm. According to the spec for the tool, the spindle speed should be 9400rpm, and the feed 0.025mm (.001") per tooth, giving a total feed rate (with 4 flutes) of 940mm/min.

Can anyone offer any suggestions?

Many thanks in advance.

Hugo

[fpworks]

Since you asked this question in the HSM forum AND considering that you will be machining aluminum, your largest consideration should be in the ability of the SYSTEM to maintain toolpath accuracy under high feedrates. Your cutting forces will be negligible compared to the force required to accelerate/decelerate the moving components of the machine.

In light of that, you should be considering the mass of the machinery components that you are moving, the required accuracy of your parts, and most importantly, the sophistication of the control...does it have the ability to "see" ahead in the toolpath and make velocity adjustments so that accuracy can be maintained? (and will the servos/drives be adequate to accelerate/decelerate the mass of the machine elements appropriately?)

Please don't be offended by this statement...but considering the magnitude of this project coupled with your knowledge, I HIGHLY recommend that you simply buy a machine. I really don't mean to sound rude, but I think you're setting yourself up for a lot of frustration and likely failure. If you really don't want to buy a machine, at least contact one of the OEM control system manufacturers (or retrofitter) to help you choose the right components for your mill.

[Swiss]

Hi fpworks,

No worries, no offence taken. You're right, making a milling machine, especially a higher speed one, is a daunting task. But isn't that what drives us? I'm only really happy when I'm solving an almost impossible problem.

I totally agree about the control being a critical part of the machine's ability to operate quickly, and I am looking for control software at the moment. I'm pretty sure Mach3 is unsuitable, but someone pointed me towards Softservo, which looks awesome, but will probably be expensive.

The alternative is to write my own. I know, I know. That's really hard too. But, actually that's the part I would be most comfortable with. I have twenty years' programming experience, 15 of which I was working at a robotics company.

Anyway, I'm not going to bother calculating the forces. I'm just going to glue a strain gauge to some aluminium and run it through our current milling machine and see what the forces are like.

Hugo

[fpworks]

No worries, no offence taken. You're right, making a milling machine, especially a higher speed one, is a daunting task. But isn't that what drives us? I'm only really happy when I'm solving an almost impossible problem.

I certainly agree with you, and I also take great satisfaction in solving difficult problems. However, you said this was company related project, which means money and time is involved. I fear the time and money you may spend to achieve an acceptably performing machine may outweigh personal satisfaction and experience you will gain. I highly suggest throwing money at this problem!

I wouldn't bother with the strain gage...like I said earlier, your cutting forces are the LEAST of your concerns. F=ma should be your primary concern.

For instance, even on my machines, cutting feed at over 12,000 mm/min, the servo loads are low...around 15-20%, but when accelerating/decelerating the axis, spindle loads often exceed 100% load. (often, the axis perpendicular to the feed axis will show the higher load)

[dertsap]

http://metalworking.com/shareware.html look for the "machine zip" file which is a cutting force calculator

as far as cutting ally you may need to concider two strong factors rigidity and how you program and prepare your work piece ,
for eg 2 weeks ago i was experimenting on my diy machine which is not the most rigid setup but i was cutting 4' pocket in some aluminum plate , i was taking only .01 in. depth of cut but with a spindle speed of 24000 rpm at 600 ipm with a 1/2" 4flt endmill there was barely any load on the system and the rate of material removal was fairly quick , of coarse it is not in the same calibre of a professional system but then i didnt pay the price of a professional system either , normally i would cut at .05 depth at 50 ipm ,
im sure you could create a solid setup quite easily if it is well thought out , but you will need to add some pricey gear to create a truely accurate and repetative machine

i do agree 100% with fpworks about time is money when it comes to tackling such a project , it may be worth looking at the price of a small used mill to see if you can justify the cost and to be able to create the same machining capabilities

[Swiss]

money and time is involved

Yes, the time is the biggest concern here. Also the worry that the build will hit a major problem somewhere, adding unexpected delays.

I did some calculations based on the Emco 105 machine we have now. It cost us £27,000 and had a delivery time of three months. So, assuming that's my budget, I added up the price of all my dream components, and it comes in at less than that.

Twin linear brushless motors per axis (about 100N total force per axis)
Really nice Hofer Precision spindle.
Twin ball slides per axis.
One Renishaw 1um optical encoder per axis.
Drive electronics.
Motion control software and hardware (maybe CamSoft and Galil).
PC etc.
Misc components (framework, coolant system, cables, sensors)
Three engineer months.

Comes to about £20,000. So, for less money, we're getting a much more capable machine than the Emco. I believe that, for a machine like that, we could end up spending more like £50,000 - £100,000.

Money is just something we have so little of (though we have so little time too). We get paid way less than our market value, and for years took home nothing at all. If we were a more functioning company with a better cashflow, I wouldn't consider making one at all. I'd definitely buy one on finance.

Hugo

[Swiss]

Actually, the other reason to make one is to have a machine designed specifically for makine the types of parts we want, and the way we want to work.

[Swiss]

dertsap,

thanks for the link. I'll try it out.

It's an interesting point about the acceleration forces being more important. I guess they're hard to quantify at the moment, because I don't have numbers for the mass of the components. Maybe I'll do some ballpark estimates and see if the drives are upto it.

I'm not sure if what we're doing really qualifies as high speed machining. Looking at the feet rates for the largest aluminium tool we'd ever use (6mm end mill), it works out at just under 1000mm/min (39in/min). Although for plastic we can get it up to 3000mm/min (118in/min). But nowhere near the 600in/min that other people are talking about.

Hearing other people talk about speeds like that is a bit like someone saying "well, I was driving my car at mach6, but I had to take it down to only mach 4 the other day because it was raining".

hugo

[camminc]

Basically forces of a cut can be estimated to determine Horsepower required in Milling. Depth of cut axially, depth of cut radially, feedrate, specific power. These parameters will give you the Chipload that will provide you the power required in a specific material being machined. So, take Chipload / specific power of material being machined to derive HP. Pretty much straight forward. It does not matter if your doing HSM or Conventional, it is the same, the only difference is that you clear the chips out of the cutting operation of the material to stabilize power, which HSM might provide more. Therefore, chip load / power is determined by chip removal of the operation. I can send you a program to determine forces, hp, etc if you desire. CAMM

[daedalus]

Hi Hugo,

What kind of positional accuracy and stiffness are you aiming for with this machine? Also what is your working envelope?

If you are new to building precision machines having a look at a copy of "precision machine design" by Alexander Slocum would be a good idea, it covers quite a bit of the level of engineering involved.

I have been collecting parts for a potential linear motor mill for some time now, but am doing so for my hobby, so for me part selection is more down to what i can scrounge then what is best. Personally im doing this project for the fun of it, rather then to meet a work deadline, and if the case were different i would probably buy a machine even if it cost lots more.