The monthly "I'm new to HSM" thread

[Tom Szelag]

Not sure if HSM would be a solution to the parts we run here. I've been to mmsonline.com and some other sites to get a feel for the basic concept.

 

So let me see if I have this right -

 

HSM is a process of CNC machining designed for specific applications where conventional "big cut" machining would not work well. For example, parts that require a very high degree of accuracy and surface finish, intricate parts with thin walls and floors, or very hard materials.

 

Using very light axial and radial depths of cuts and uniform tool loading, you get longer tool life and low cutter deflection.

 

But how do I know if HSM is right for us? Let em fill you in a bit on the story so far. From '95 to '04 we had a senior design engineer who kept his parts simple and machine-friendly. Good thing because for a while we didn't even have a true CNC center, we had a 2-axis EzTrak and that was it. In '99 we got a Fadal VMC3016. We ran all our parts fairly conventionally. Majority of them are 6061 or 7075 AL, with the occasional stainless, brass, or engineering plastic.

 

Since he left, our new senior engineer has had a lust for either small, complicated, or small AND complicated designs. Right now I'd say we machine Ultem 1000 and AL6061 in about equal amounts. A lot of the plastic parts especially have small features and tight tolerances. We question the necessity of some of this complicated stuff...

 

Our production numbers are also changing. Used to be a lot of prototype and small-time batches, less than 10 items. But now we're getting into runs of 100-400. Also have some tight deadlines and budget restraints. So a good way of machining this stuff accurately and quickly would be great.

 

I'd like to try re-running the toolpaths on some of these parts we've made during some downtime, but I have no idea where to begin. I'm a student, fairly new to this stuff, and our senior engineer has 40 years experience but is fairly old school about doing things.

 

So where do I start on appropriate feeds, speeds, and depths of cut? I'm used to just running handbook values, bout 400sfm for Aluminum, cut depth of say 30% the tool dia, 50% stepovers. What are typical changes to these for HSM? And why in every demo video I've seen is there only air blast, no coolant at all?

 

Then again I'm not sure we're really capable of HSM. Our Fadal, spindle only gets up to 7500RPM and the travel rate either maxes out at 200 or 500IPM. But I think we can do better than where we're at, I've never seen any tool run past 5000RPM or 60ipm.

[Jimmy Wakeford from Barefoot CNC]

Tom

You can follow the link below and download a 30 day eval of HSM.

[?Mark]

Hi Tom.

quote:

---

So where do I start on appropriate feeds, speeds, and depths of cut? I'm used to just running handbook values, bout 400sfm for Aluminum,

---

It's a complicated subject Tom.

HSM is generally applied in mold making industry in hard materials. The key word is "generally" as you can use this technique in other materials as well.

The idea behind it is to take a light axial cut with a big chip load (go light but fast). For hardened materials you use air blast instead of coolant to prevent thermal cracking of a tool.

You would need a rigid machine, fairly fast control and a very good tool.

For example there are tools for alum that will handle up to 9800 SFM and up to .020 inch/tooth.

There are similar kick a$$ tools avail. for almost any material.

Might want to talk to your local tool rep about a demo vcd to see how it is being done.

We don't do molds here but we do hard milling in 60rc tool steels fairly often.

 

Hope it helps

Kind regards, Mark

[IntegrexMan]

Anyone ever tried HSM in Cobalt Chrome?

 

Craig

[Joepcb1018]

Tom,

 

I look at HSM this way..More like high performance machining. The largest material removal rate in the shortest amount of time to achive the desired finish and accuracy. That being said it will take you some time to develop your process and speed things up. I try to push my feeds and speed until you just start to lose something...finish, accuracy cutter breakage then back off 10% for safety. The tooling manufactures will usually give you a good starting point but your machine may not be able to keep up. There are many things that are involved. Good Luck

[betts]

Tom-

+1 to the comment that its a complex subject. Your understanding and Mark's description are the place to start. The link above may help too. You won't be doing "true" HSM in its purist definition, but you may attack your jobs with a different approach.

 

My thought was........How does someone quote and design parts at competitive pricing who is thinking "old school" and make any money? True you have to start with the equipment you have, but you won't get very high feed rates and maintain accuracy and suface finish with a Fadal. They are very good machines for the money, and if you have one of the latest controls hanging on it its better, but you will have to work within the ability of your machine. I think a '99 machine, well which control do you have? 32MP? Anyway how you attack the job is where you will save money, even working within the limitations of your machine.

 

Do some reading and testing, it will help.

[John T.]

We have recently installed a couple of high speed machines, we are still experimenting with them though, but results are very good. 99% of our stuff is in aluminum, 6061 and 6013. We use hi-helix cutters for fast stock removal (roughing), and carbide ball mills where necessary for finishing. Most of our machines are Fadals, and we have seen some tremendous improvements in finish quality when applying some HSM techniques to these also. We have added the HSM C-hook to our MC, and that has been one of the biggest helps in finish quality recently. It takes awhile to process some toolpaths with it, and has some quirks, but it is good.

 

John

[Tom Szelag]

quote:

---

How does someone quote and design parts at competitive pricing who is thinking "old school" and make any money?

---

Well we do very little outside work. Small shop for a small NASA research facility that does payload product development. Don't get me wrong. He's got loads of experience, which is very helpful for some of the crazy designs these guys throw our way. But on the other hand I think he's a bit set in his ways.

 

quote:

---

I think a '99 machine, well which control do you have? 32MP?

---

Marked on the controller is "CNC 88HS". If that means anything to you. Definately nothing too modern.

[MetalMarvels]

Just an FYI with a Fadal 3016 (1999 to 2003 vintage). If you are going to try pushing the envelope on a 3016 you will need to selectively turn up the servo gain with an M92 (Intermediate) in order to reduce overshoot. However, you should not leave it there and should return the gain to "normal" with an M90 (default) or an M91 (normal). Using an M93 (high) is typically reserved for rigid tapping. If you leave the gain at "intermediate" it will beat up the axis - you will hear it "bang" as the lead screws change direction.

 

Having said that, if you set up your tool paths so that there are no 90-degree changes in direction, the higher gain setting will not hurt anything. Sweeping moves are better for the Fadal. If your Fadal 3016 is anything like mine, the fastest you will be able to program for a G1 type move will be about 180 IPM (purportedly 200 IPM - but I have never been able to actually get it going that quick) - the G0 rapids will be about 400-500 IPM.

[betts]

I agree with Gary on turning up the gain, but we have noticed that using the "highfeed" and any smooth options possible that it significantly reduces the Fadal "banging" that you can get.

 

Actually your machine with 88HS control probably has DC motors and you can do a lot with dialing them in. Maybe someone else knows, I could be wrong about when they started with the AC drives.

 

There are other options as well. The Fadal Analyzer will basically "rewrite" a program but you have to have the right part shapes for it to be effective.

 

Anyone out there have a Creative Evolution control on a Fadal?

[Chris Rizzo]

I've also been wondering about utilizing HSM for applications other than hardmilling/mold work. I'm really curious as to trying it in small production runs in soft materials.

 

Tom-

I've come to rely heavily on highfeed, (which is not the same as HSM) for cycle time reduction, esp on "flimsy-er" machines here (Haas ) It's free, and does'nt take too much time to become familar with. The general idea is constantly changing the feedrate to maintain a consistent material removal rate. i.e light cuts and air, speeds up the feed. deep corners and high stepover, slow feed down. Now I don't use it for finishing ops, just roughing. It's a good tool I've found for hogging material out quickly.

 

John-

How does HSM improve your surface finish? Does anyone use HSM with highfeed? Highfeed would be nicer with tangential arcs and other smoothing out moves...but I suppose that's HSM's role?

 

Jimmy- I went to a demo of HSM at my dealer a few years ago..Should I call him for a demo/trial disk?

 

thx

[Guest CNC Apps Guy 1]

HSM is more a methodology than a specific technique. It means different things to different people. To a guy used to machining Rc60+ stuff, 500 SPM and .005 ipt may be high speed machining, while 10,000 SPM and .020 ipt may be HSM to another. In a nutshell it is a different approach where heat and stress are put into the chip rather than the part being machined. But one thing that reigns supreme in HSM is pushing the envelope. Like was stated earlier run it until sometheing breaks and back it off 10%.

 

I'm sure other htings will come to mind as I think of it.

 

JM2C

[Chris Rizzo]

Huh, that's for the background James. I was thinking of HSM as a means to smooooothing out machine motion. Lighter cuts, more of them, and at higher speed. Increased cutting distances sure, but IPM's go through the roof. To keep IPM's running smoothly, and not beating the hell out of yer axis, HSM would create "fluidity" in machine motion. ?

 

At least that's what I took away from the demo a few years back. Thus my reference to tangential arcs and such.