What's new
What's new

What's your opinions on the Haas UMC 750 Superspeed

implmex

Diamond
Joined
Jun 23, 2002
Location
Vancouver BC Canada
Hi All:
My business partner is looking very seriously at a UMC750 SS as the newest addition to our shop.
The Haas sales guy is talking up the virtues of the machine with great and florid enthusiasm, and I have to say, there are some nice convenient features that come with the machine, all of which have been declaimed about in loving detail.

Any of you have one?
What's the real deal with these; are they "good" machines, "OK" machines, or "crummy" machines?
The little bit I've found online on the subject has not been super positive, but so much of it sounds like the "Haas Haters" going to town.

On the other hand, I was at the local open house with demo, and the machine sounded great cutting 6061: then it was asked to cut a chunk of 4140 prehard.
I didn't much like the noises it was making chopping through that lump; I got the distinct impression it was not as tight or stiff as it needed to be to make decent time in that material without wiping out a lot of cutters.
The part was maybe 4" diameter and fixtured so the top was a foot or so from the table surface.
The cutter was an indexable, something like a Sandvik Coromant maybe 3/4" diameter.

I realize this is all hopelessly vague, but I'm looking for impressions as much as specifics.
Any comments will be welcome.

Cheers

Marcus
Implant Mechanix – Design & Innovation - home
Vancouver Wire EDM -- Wire EDM Machining
Clarus Microtech
 
I've never seen accuracy specs published for this machine. I would have to know that before seriously looking at one.

Also what is the price? For around 200k you could pickup a Mazak Variaxis J-500. Mazak also has a VCU series 5 axis machine for less than the Variaxis.
 
From my experience, not this model but Haas in general - noise comes more from the enclosure "amplifying" the vibrations. Surface finish and dimensions are the true tell.
 
So some years ago, I was surprized to discover that the various built-up haas 5-axis options where not particularly cheaper than other born as 5-axis machines.
I don't know about current relative pricing (has surely changed over nearly 10 years), but my key advice is to be sure there isn't some other machine like a Mazak or DMG-Mori or even Okuma or Makino that doesn't pencil out better for a price that's closer than you would expect. Now, if your partner has already done that, you are ahead of this question.
 
Mazak has been making the variaxis far longer than Haas. Certainly a benifit in mazaks favor.
If either company is serious about getting your business they would do some sample cutting on one of your parts. Only this way can you be sure head to head.
With TCP DFO options they should be able to run the same program for a true test.
Oops the tilt axis are different, huh?
 
Last edited:
I can't offer much help other than being interested in the machine as well. Supposedly (according to salesman - so not necessarily true), the UMC machines are flying off the shelves. There was a 6 month backlog shortly after they came out, which IIRC is now down to more like 2-3 months.

I think the SS version of the machine is ridiculously priced compared to the non-SS version, when you look at what (tiny) amount you get for the extra $$. It almost seemed to me like Haas priced the standard version too cheap, and decided to try to make it up with the SS version.

As for the machine itself... I have seen it in operation and only talked to a couple of owners. Heard mixed reviews... it is slow to move in it's 4th and 5th axis (and again, charging so much more for the SS seems crazy here). I am surprised there aren't more owners out there considering the machine has been out for 3+ years now.
 
...the machine sounded great cutting 6061: then it was asked to cut a chunk of 4140 prehard.
I didn't much like the noises it was making chopping through that lump; I got the distinct impression it was not as tight or stiff as it needed to be to make decent time in that material without wiping out a lot of cutters.
The part was maybe 4" diameter and fixtured so the top was a foot or so from the table surface.
The cutter was an indexable, something like a Sandvik Coromant maybe 3/4" diameter.

In general, if you intend to machine steel by burying the cutter, you need box ways and a lot of iron. A linear-bearing machine needs a different approach--high RPM, low DOC, and high feedrates. You have to use the advantage of the machine, which is all about quick movement. The machine features that contribute to high productivity on aluminum parts are pretty much the opposite of those for steel parts. It is possible to machine steel parts fast by utilizing cermets and pretending you're programming for aluminum, but it's going to sound like angry hornets. On the other hand, we have been milling steel parts with a 3/4 Ceratip endmill at 5000 RPM in the same VF-0 all day long since 1998, and while it's scary loud the parts are still dead on. The noise level is no worse than when it was new. Go figure.
 
Saw it at HaasTech last month. Very impressive. It was making parts in a 12 minute cycle time that would take multiple setup in a VMC. Very impressive. The operator did admit the 5 axis programming was complicated and requires powerful CAM and the know how.
 
Heard horrible things about thermal comp and center of rotation error. I wouldn't touch it with a 10ft pole. Just my opinion.
 
Heard horrible things about thermal comp and center of rotation error. I wouldn't touch it with a 10ft pole. Just my opinion.

I don't have an SS model, just the regular. As far as thermal comp in xyz, I have not seen so much. I can pick up a part in Y cold (well cold as in after a 10-15 minute warm up) and run all day and pick up again and only see .0005"-.001" difference. My Z offsets are normally within .0005", yes this is .001" or greater when you factor in both sides of a part (if you rotate about the C axis). This might be huge for some of you depending on your work. It is not a big deal for what i am doing. I will also note that the probe (IMO) is only good to about .0001-.0002" as far as picking up, or repeating as far as touching an edge or bore or whatever. Again, might be huge to you guys trying to hold +-.0002" on everything. For the record I do do some stuff that is less than .0005" tolerance, the machine will do it, but I doubt you could set it up and run all day like that without monitoring. I have not done any full 5 axis so I can't comment to that part yet. Without doing any full 5 axis I can't say about the center of rotation issues (if there are any), but it does come with the WIPS system standard which includes the spindle probe, ots disc touchoff, the 5 axis calibration sphere, and a master gage length tool to set the ots.
 
Edster,

X Pos, ±0.0002” Rep. ±0.0001"
Y Pos, ±0.0002” Rep. ±0.0001"
Z Pos, ±0.0002” Rep. ±0.0001"
B Pos, ±15 arc-sec Re. 10 arc-sec
C Pos, ±15 arc-sec Rep. 10 arc-sec
 
Edster,

X Pos, ±0.0002” Rep. ±0.0001"
Y Pos, ±0.0002” Rep. ±0.0001"
Z Pos, ±0.0002” Rep. ±0.0001"
B Pos, ±15 arc-sec Re. 10 arc-sec
C Pos, ±15 arc-sec Rep. 10 arc-sec

That sounds right. I think the Hurco we looked at was +-7 arc seconds on position for the rotaries. I don't remember what the repeatability was on them. So what does +-15 arc seconds translate into on say a 5" diameter part? Say you have a couple holes around the periphery and two features milled 90 degrees apart. Your 90 degree features would be then +-30 arc seconds, correct? Which is 89.9833 to 90.0167 degrees if I did it correctly. So a feature on a 5" circle would have a radial positional of +-.0004". Someone correct me if I am wrong.

I guess that is not great... :( I have not seen the thermal growth issues. But that is not to say I won't, add that to the rotary axis accuracy and I suppose you would have a hard time holding less then .002". Although I have heard even on higher end machines sometimes getting the rotaries all "true" is very difficult as well. I don't know, just heresay for me as I have never programmed one or tried to hold a 5 axis part 'dead nuts'.
 
So what does +-15 arc seconds translate into on say a 5" diameter part? Say you have a couple holes around the periphery and two features milled 90 degrees apart. Your 90 degree features would be then +-30 arc seconds, correct? Which is 89.9833 to 90.0167 degrees if I did it correctly. So a feature on a 5" circle would have a radial positional of +-.0004". Someone correct me if I am wrong.

Using the old rule of thumb that one degree is .016 per inch, 2.5 inches from the center would be:
2.5 x .016 = .040
.040 / 60 (minutes) = .000666
/ 4 (15 seconds or 1/4 minute) = .000166 circular distance

Seems pretty close for the money. Remember, that's the maximum absolute angular error, the maximum repeatability error is 2/3 of that. Quite an exercise even inspecting it, and closer than anything I would ever claim to be able to do...
 
I have a UMC 750 and for 3 plus 2 machining in aluminium its does what its says on the tin but and its a big but chips get sprayed around the enclosure and settle inside the spindle casing and falls on the Z axis proximity switch and throws up a Z axis limit switch error rendering this nice bit of kit useless!! not happy Mr Haas!!!!!
 








 
Back
Top