Thread: Mill spindle deflection: Acceptable amount?

Machine: Brand new Haas VF-2SS

Test: Out of curiosity, I put a piece of 1.500" aluminum round bar in an endmill holder sticking nearly 12" below the gauge line, then put a .0005" DTI on a vice touching the endmill holder roughly 1.500" down from the gauge line, oriented to the Y axis. I MDI'd a spindle orient so it wouldn't move and pushed / pulled on the round bar fairly hard. The result: .003" deflection one way, .003" the other. Also, though only a few tenths, the indicator didn't go back to zero after pulling or pushing on the bar.

That can NOT be right, can it?

Check that the tool is fully seated in the taper, by tapping upwards on the end of the tool with a soft hammer. I know sometimes, when I remove a bunch of tools from the carousel, that some of the holders 'pop' out of the spindle, and some seem to just drop out as though they were not really tight.

I suspect as soon as the tool begins to cut, that the vibrations help it settle into the taper. You might be seeing the same thing.

Might not be the spindle. Try putting the DTI base on the head and see if the toolholder moves in reference to the head, then put the DTI base back on the vise with the DTI touching the head, not the toolholder, and see if the head moves in reference to the table. I bet there's more movement in the head than there is in the spindle.

I'll go with mudflap. Before you say the spindle is loose, you should isolate to determine where the movement is coming from. I have even mounted an indicator on the ways and looked at the housing of a particular axis. Keep moving the mag base and the indicator, when you cross the loose item, the indicator will let you know. Three thou is a good bit and if all of that was coming out of the spindle bearings, I hafta think they would bea rattling.
Let us know.
Regards Walt..

How tight is that 1.5" aluminum in the toolholder?

Pulling "fairly hard" on the end of a 12" long 1.5" diameter Al bar is going to cause the bar to deflect somewhat, even at the 1.5" down location. I don't know if that's all of your measured deflection, but its definitely some of it.

Do you see the same movement if you just push on the bar up close to the gauge line?

Paul T.

You could try putting the DTI on the toolholder instead of the bar to eliminate the bar/toolholder fit as a possible source of movement.

QB

If you can wade through the math and calcualte the deflection, use:
http://www.mmsonline.com/articles/060105.html

with modulus of elasticity from:
http://screem.engr.scu.edu/artemis/s...h/al_data.html

Matt@RFR

Test: Out of curiosity, I put a piece of 1.500" aluminum round bar in an endmill holder sticking nearly 12" below the gauge line, then put a .0005" DTI on a vice touching the endmill holder

As Walt said.. but mount the mag base ON the spindle housing or spindle carrier... Point the indicator on the spindle flange.. NOT the tool holder.. There should be ZERO movement there or if there is it's play in the Spindle Bearings... NOT GOOD, and as Walt said... I think you'd hear that.

It's probably pertinent to mention that this is my first CNC machine and started checking things after trying to machine my very first part so I'm going to make this into a two part post if you guys don't mind. But first, the results (all in the Y axis):

1)DTI base on vice, DTI on head = .0015" total deflection

2)DTI base on vice, DTI on spindle = .0025"

3)DTI base on vice, DTI 1.5" down from gauge line, touching the toolholder = .006"

4)DTI base on head, DTI on spindle = .001"

5)DTI base on head, DTI 1.5" down from gauge line = .002"

And Hu, I tapped the toolholder into the spindle like you said, but made no difference in the original test.

Also, I'm not saying the spindle is loose. I wouldn't know a loose spindle from a perfect spindle. [img]smile.gif[/img] What I do know is, following the manufacturers recommended speeds/feeds, I get really bad finishes with ALL tools, lots of chatter, broken endmills, etc. on the mild steel part I'm working on. And yes, there's a definite possiblity that it's just me being new, but I can't imagine the tool manufacturers would have me running that far off.

Can anyone give me some sure-fire speeds/feeds/DOC for a 3" face mill with 5 SEHT43 inserts (both from TMX) that should leave a very good finish on 1018 mounted rigidly?

I thought my sensory-type tuning style would carry over from the manual mill. I thought wrong.

If it is new, one would think you could get a service man to come and inspect it to see what is going on.

How do the toolholder shanks look? Any sign of fretting (rustlike markings)?

Check that the bolts holding the spindle cartridge up under the head housing are tight. Also check that the 6 small allen capscrews around the bottom cap on the end of the spindle cartridge are tight. That cap holds the spindle bearings up in the housing.

However, that bottom cap should actually have a wee bit of a gap, maybe a very thin feeler gauge can slip in about 1/4" or so, before it hits an O-ring. This gap would be normal, and its existence would show that the cap is indeed capable of putting clamp pressure onto the bearing pack when the bolts are tight, and a narrow gap still exists.

Thanks Hu, I'll check those bolts in the morning. Also, a tech is scheduled for Wednesday morning (good response time) and I've been informed he'll have a new spindle with him. The tool holders are coming out of the spindle with no marks whatsoever on them.

Curtis (exkennemetalguy) had me run a couple test passes with an endmill (Imco Powrfeed) and heard this thing run over the phone. He can back me up when I say that it squeals like a stuck pig when he's got the same endmill running the same parameters in other VF-2's with great results. Below are pictures of the tests Curtis had me do with a 1/2" endmill and my 3" facemill.

Yep, something really wrong there. Keep us informed, I'd like to know what the tech finds. Your DTI tests make it look like there's excess movement between the head and the ways, but that doesn't fully explain the looks of your cuts.

The problem could also be due to an issue with the linear guides. I would have the service tech. put a granite or ceramic square on the table and look for "fish tail" in all of the axes.

Let us know what he finds.

RL

Lets see, tool retention, tool contact to spindle taper,loose gibs in any of the axies, lost motion in any of them axies.
Do as you wish, but isolate the problem before they just start replacing parts.
I don't know why they just "bring a new spindle" when their is a host of things that could be involved. Might even be loss of preload when the nut backed off or the bearings settled in, tighten the preload nut and make chips.
O yes, check bearing preload and or spindle end play.
Regards Walt..

I totally agree with Walt, and "isolate the problem first".

There are a lot of possibilities here!

Regards,

RL

Isolate the problem first.

Isn't that the tech's job? I'll say it again: This is my first CNC machine. Put another way, this is literally the first CNC machine I've ever physically touched. I'm not about to go digging into **** I've no idea what it is, what it does or even know if I'm looking at something normal or not. It may be much more simple than I'm thinking it is, but, especially being under warranty, I'll just watch over the shoulder of the tech and begin my troubleshooting and repair knowledge there.

Matt,

Yes it is the tech's job. We are all just giving you other things to have him look at, and to make sure you do not have any other issues.

Regards,

RL

Ahh, I misunderstood then. Sorry about that.

Also Walt, I doubt they are bringing a new spindle with them because they believe the opinion of some CNC newbie and his internet friends. [img]smile.gif[/img] I'm sure they are bringing it in case it's actually the problem. And hey, I appreciate the hell out of that.

Who else has a bet on bad leveling, perhaps a feet not touching, cracked casting somewhere? seems its been the cause for a few other machines that were causing similar finishes. Then again, could very well be the spindle.

I must say I know of another place running a VF2 and I haven't been very impressed with any of the finishes. Maybe some of it is because of tooling and programing, but well... still not impressed.