Haas Deflection/out of tolerance

Blackmarkit · Nov 16, 2019

Hi I have a 1999 Haas VF0 that I'm fighting deflection.
Have to make multiple repeat passes to get close to my feature dimensions.
I have installed: all new ball screws in all axis, Z axis linear rails and a new spindle all within the last 6 months.
For example - 1/2 niagra 3 flute 2" loc in Haimer shrink fit .0002" runout
Cutting Aluminum
7500 rpm.
1.5" doc
.020" radial
50ipm
I have 2000 lbs drawbar pressure.
getting .007" off of programmed feature at the top and .013" at the bottom both X and Y are the same.
after running 2 more (spring passes) i was .001" over on top and .0025" on bottom.
Because of the taper it makes most sense that the problem is in Z.

Open to any suggestions.

D Nelson · Nov 16, 2019

How about that 2 inch cutter flexing on you I bet if you ran it about 20 more times it would be good
Don

Sent from my iPhone using Tapatalk Pro

Blackmarkit · Nov 16, 2019

According to what I've heard and and read it should be tenths taper. And with many tool company's grinding back taper i would guess .0002".0003" taper top to bottom

DavidScott · Nov 16, 2019

What are you cutting?

Blackmarkit · Nov 16, 2019

Sorry im machining Aluminum

mmurray70 · Nov 16, 2019

I bet its just deflection in the tool, maybe a little from machine too. Razor sharp aluminum tools and even more spring passes will help.

If its super critical to have no taper use a tool with a solid reduced shank and make multiple passes. This will work much better. If im doing a tight tolerance with long tooling sometimes I will even go a step further and grind away maybe 75% of the cutting edge of a reduced shank endmill. Making the engagement and deflection even lower still. Good luck with it.

Milland · Nov 16, 2019

All machines deflect, but isolating where it's happening can be time consuming. I'm a little confused that you'd spend all that money and effort on an old Haas, but IIWII...

First thing I'd check is are the machine support pads uniformly loaded, on a stable floor? If you didn't check for pad loading and that the machine is fundamentally square and level, that's something to confirm right away.

When you replaced the ball screws, did you also replace (and properly install) new thrust bearings? Did you assess the wear on the X and Y linear rails?

If you want to continue digging into this and the machine is properly resting on the floor, then get yourself a 50 to 100lb fish scale (like these: McMaster-Carr), and start by anchoring the slide to various parts of the machine and pulling uniformly, while having a friend either do the pulling or read off indicators that are placed to record relative movements between the "stable" and the pulled elements.

When you've found a weak link chase it down further to either confirm it's due to slop or wear, or find the true culprit if it's some adjacent element.

Handy tip: Be sure scale is anchored well so it can't slip off while under load. Hilarity may ensue, but it's usually accompanied by bruises or worse...

Blackmarkit · Nov 16, 2019

Taper is vertical. Measuring X and Y taper looks equal which makes me believe its the Z axis linear rails. but with them being so new it would be sad to say there bad already. the head was trammed within .0003" over 13" better than Haas spec but what the shims provided limited me to.

Blackmarkit · Nov 16, 2019

I got the machine cheap ($8500) and ive installed everything myself. ive gotten a pretty good deal on everything so ive put in another ($8000)I did reaplce all bearings and torqued every spanner nut and bolt according to haas spec.
Is 50 lbs of side load force all the machine gets under load?

Blackmarkit · Nov 16, 2019

If .013" off of the featured spec is acceptable ill roll this things out i the middle of the freeway. lol. i feel like it was pretty accurate the last 6 months until recently. of coarse i didn't do any test parts and document.

Blackmarkit · Nov 16, 2019

I hope the part is square. machining a raised square boss i get a smaller square at the top than the bottom. im not sure if its all on one side but im leaning towards some kind of Z axis deflection because measuring 2 different directions on the raised boss show equal taper.

Vancbiker · Nov 16, 2019

Blackmarkit said:
I got the machine cheap ($8500) and ive installed everything myself. ive gotten a pretty good deal on everything so ive put in another ($8000)....

OUCH! >$16k in a 20 year old VF0?

Where did you get the Z linear guides? If you got them from a source other than Haas, are you sure you got the correct preload grade?

Same question for the spindle rebuild. Any finish issue on floors or odd chatter that might indicate lack of preload on the spindle?

Milland · Nov 16, 2019

50lbs is enough to show a deflection that's greater in one direction than another. It just shows the trend. The actual "pushing force" of an endmill varies (of course), but under finish conditions (single pass) on Al it's not a lot.

You can up the force, but you want to be consistent for it to be useful. If you can do a 200lbs side force by hand then do so, but remind me not to arm wrestle you...

plastikdreams · Nov 16, 2019

If it's hss tooling there's a great chance it's pushing the tool off. If it's carbide there's still a chance of it deflecting but not that much.

As a test, run your full depth then run a spring pass using a ramp pattern, see if there's any taper there.

Blackmarkit · Nov 16, 2019

I was thinking $16k to buy a machine with all new ballscrews, Z axis linear rails and spindle was a good deal. Of coarse hoping that would make me accurate parts for a couple years. i received all parts except the spindle form my HFO. Spindle was NOS on ebay, came in a HAAS box. Im getting really good finishes the first pass.(Mirrors) They get hazzier with each spring pass. but still considered pretty good. My tool rep from western tool said if i want better finishes i should get a better machine. Im leaning toward Z axis linear rails again. If i put a 2x4 between a vise and a spindle and pry on it how much deflection should it get?

g-coder05 · Nov 16, 2019

getting .007" off of programmed feature at the top and .013" at the bottom both X and Y are the same.

Is the taper the same on the part at the X- and X+? Say X+ at the bottom the taper would be -.013 and when measuring X- is the taper +.013 or -.013?
If its a positive taper on one side and Negative on the other that would point to something in the Z axis being knocked out. But if its consistent pos or neg then it would be a case of deflecting the Z somehow, Could be a preload issue or the rails twisting.

1.5 LOC with .020 Radial at 50IPM should not even begin flexing the machine so I'm going to lien on the mechanics of the Z.

As far as having $16K in a 20 year old "O" model is a bit much but still doable. Only thing I see bad there is its not supported by Haas and if you go down electronically it's going to be ugly. The only way they will help you is if you buy the $20K retrofit kit. Then again, That little machine with modern drives feeding at 1200 IPM and a 30 HP spindle with the NG control would be a lot of fun.

Blackmarkit · Nov 17, 2019

Well I pulled the spindle cover off last night and started to pry on stuff while moving a tenths indicator around. The spindle column is definitely moving. I’ll be able to dig into it today.

Can someone else run that same tool recipe and tell me what a good machine should machine run?

Milland · Nov 17, 2019

IMO you shouldn't see over .0005" per side on a 2" depth with a sharp 1/2" carbide endmill doing a finish cut 2" deep in Al. But that can vary with tooth count, speed, etc.

If your Z is moving, that might indicate improper preload of the trucks, but you said they all came from Haas. Did you try sliding the trucks on the rails before installation, and were they free moving, or did you get a little friction (and not just from the seals). Did you rock the trucks about the axis of the rails? They shouldn't have moved.

Delw · Nov 17, 2019

Blackmarkit said:
Well I pulled the spindle cover off last night and started to pry on stuff while moving a tenths indicator around. The spindle column is definitely moving. I’ll be able to dig into it today.

Can someone else run that same tool recipe and tell me what a good machine should machine run?

how much is it moving???

Get a long straight gage pin, ground bar or make one on a lathe that has no taper. put it in the tool holder (3/4 or bigger is best)
then run an indicator on 4 different side and move the z axis up and down. if you get taper your z head, or spindle squareness is off. but before you determine that try a different holder to make sure its not your draw bar clamping wrong or bad tool holder. also try re clamping the tool a few times.

if that dials good then put pressure on the head in the x directions then try in the y directions. it shouldn't move but a few tenths if that.

I cut with 1/2" alumstars 1.5 depth and 2" depth. I get a few tenths taper at most, I also take 2 finish dry pass's on 2" or more and 1 on 1.5 or less
I run usually about 3500rpms on finish pass roughing1.5" or longer depths(always leaving .008 per side for finish pass) I run 5000-8000rpms for rough depending on side DOC.
its a vf2ss I run slower than most.

I am assuming you ran a new unused endmill and also checked the endmill to see if it was tapered its rare but it happens.
what happends if you run the same endmill 1" depth you get a big taper still? how about a short endmill at .500 length any taper at all?

CarbideBob · Nov 17, 2019

The big size difference between a normal run and 2 springs passes concerns me.

Haas Deflection/out of tolerance

Plastic

Stainless

Plastic

Diamond

Plastic

Stainless

Diamond

Plastic

Plastic

Plastic

Plastic

Diamond

Diamond

Diamond

Plastic

Titanium

Plastic

Diamond

Stainless

Diamond

Similar threads