Need technical expertise regarding the REBUILD of HAAS 5th Axis Trunnion TR160-2

[imasen deshita]

Quick background:

a) Working for a company that has several HAAS VF3's of different age with TR160-2 5th-axis trunnions (the ones with two rotary platens on the trunnion)
b) Documented and regulated minimum spec's for accuracy/repeatability via ballbar and/or laser testing for all machine axes has been established by my company, and these requirements are about 6x less than HAAS' factory spec's (these machines have ran for many years without ever failing periodic verification of these requirements)
c) During a recent annual PM, the 5th-axis trunnion on one VF3 was found to be failing the established spec (I can dig for more details on what exactly failed if/when it becomes relevant)
d) After two "rebuilds" by HAAS, the TR160-2 unit still cannot meet our minimum internal spec. Further, a peer engineer spec'd out a 3rd party replacement from KME and this also failed our minimum spec. After a big fight, it was finally returned for a refund (this happened before I joined the company... the story I heard is that it was far less accurate/repeatable than KME promised, which I find a little curious).

Questions for you:

1) What does HAAS actually do to 'rebuild' one of these units? Parts replaced? Tests performed? Success criteria?
2) Can I rebuild it myself with the correct parts and tools? Is there a way to get what I would need, including procedure?
3) Is it possible that this is a problem with the mill, and not the trunnion, given that the brand-new KME also failed ballbar tests?
4) If it were your machine, and you already wasted a ton of time and money trying to get it back into production, what would you do next? (We are about to convert it into a 3-axis mill for material prep, but really want it to be a 5-axis workhorse again)

Thanks so much for your wisdom!

 

[imasen deshita]

Quick background:

a) Working for a company that has several HAAS VF3's of different age with TR160-2 5th-axis trunnions (the ones with two rotary platens on the trunnion)
b) Documented and regulated minimum spec's for accuracy/repeatability via ballbar and/or laser testing for all machine axes has been established by my company, and these requirements are about 6x less than HAAS' factory spec's (these machines have ran for many years without ever failing periodic verification of these requirements)
c) During a recent annual PM, the 5th-axis trunnion on one VF3 was found to be failing the established spec (I can dig for more details on what exactly failed if/when it becomes relevant)
d) After two "rebuilds" by HAAS, the TR160-2 unit still cannot meet our minimum internal spec. Further, a peer engineer spec'd out a 3rd party replacement from KME and this also failed our minimum spec. After a big fight, it was finally returned for a refund (this happened before I joined the company... the story I heard is that it was far less accurate/repeatable than KME promised, which I find a little curious).

Questions for you:

1) What does HAAS actually do to 'rebuild' one of these units? Parts replaced? Tests performed? Success criteria?
2) Can I rebuild it myself with the correct parts and tools? Is there a way to get what I would need, including procedure?
3) Is it possible that this is a problem with the mill, and not the trunnion, given that the brand-new KME also failed ballbar tests?
4) If it were your machine, and you already wasted a ton of time and money trying to get it back into production, what would you do next? (We are about to convert it into a 3-axis mill for material prep, but really want it to be a 5-axis workhorse again)

Thanks so much for your wisdom!

Sorry... I forgot to mention... the machine is specifically a VF-3B mfg'd in 10/09.
By the way, I would like to buy a new TR160-2 and hopefully be done with all this, but it's looking like we need to really exhaust this rebuild option before we bite that $50k bullet.

 

[dstryr]

You need to specify the tests you are doing, tolerances and failure details.
How big of a work envelope do you need?

 

[imasen deshita]

I'll get more specifics for you tomorrow.
I know they aren't doing anything unusual, I believe there is a ballbar test as well as a laser test being performed.
Work envelop is about 6" x 4" cylinder centered on each platen.
IIRC, it failed for repeatability on the A-axis rotation. HAAS' new TR160-2 spec's repeatability of 5 arc-sec, and we are looking for 30.
After two rebuilds, we still can't it back under 30. The KME failed our specs out of the box, not sure why or where. I will get more details on that as well.

Thanks.

 

[mhajicek]

Well, testing for #3 should be easy enough. You say you have several of these machines; test the failing trunnion on another machine, preferably the one that's currently testing the best.

Do you laser calibrate the linear axis' of your machines?

If it were me, and it turns out not to be the machine, I might consider buying a new trunnion. How much money are you losing by not having this running to your spec?

A 2009 economy class machine is getting a bit long in the tooth, and there should have been a plan in place to replace it when it inevitably loosens up. As you mentioned, it can still have useful life doing less demanding work.

 

[mhajicek]

I'll get more specifics for you tomorrow.
I know they aren't doing anything unusual, I believe there is a ballbar test as well as a laser test being performed.
Work envelop is about 6" x 4" cylinder centered on each platen.
IIRC, it failed for repeatability on the A-axis rotation. HAAS' new TR160-2 spec's repeatability of 5 arc-sec, and we are looking for 30.
After two rebuilds, we still can't it back under 30. The KME failed our specs out of the box, not sure why or where. I will get more details on that as well.

Thanks.

Oh! When you said your requirements were "6x less", I thought you meant six times tighter, and you just got lucky that the Haas rotaries made your spec when new. Yeah, if they can't meet 6x larger than their spec on a rebuild I'd be having some words with them.

 

[Billy_C]

3) Is it possible that this is a problem with the mill, and not the trunnion, given that the brand-new KME also failed ballbar tests?

I'm not sure how to ballbar test a rotary axis specifically. I'm not saying this is the wrong way of doing it, just looking for more information on what this kind of procedure looks like tailored specifically to a rotary axis and what the resulting data reveals.

 

[imasen deshita]

Well, testing for #3 should be easy enough. You say you have several of these machines; test the failing trunnion on another machine, preferably the one that's currently testing the best.

Do you laser calibrate the linear axis' of your machines?

If it were me, and it turns out not to be the machine, I might consider buying a new trunnion. How much money are you losing by not having this running to your spec?

A 2009 economy class machine is getting a bit long in the tooth, and there should have been a plan in place to replace it when it inevitably loosens up. As you mentioned, it can still have useful life doing less demanding work.

You bring up a few key points I'll address with more info.
This machine is part of a small fleet of three, currently scheduled for obsolescence in 2024 (replacing with new Kern's).
This timing aligns with some major NPI projects.
In the meantime, the other two machines are keeping up with current demand, but barely as I understand. It's too risky and disruptive to take a second machine down for troubleshooting the trunnion (especially since that second machine would then need to be requalified again before production).
So it's a funny situation... We don't plan to use these machines longer, and there's some competing opinions about buying a new trunnion.
I'll find out more about the testing performed today and post again.
I'm thinking I'm at a dead end on the rebuild topic.
I see some newer used ones on ebay. Might be an option...

 

[Richard King]

I would start from the beginning. I have seen CNC VMC's out of Level and the machine doesn't cut square because the base is twisted. Have your maintenance tech or you set 2 precision levels (.0005" / 12") like a Starrett 199 or finer at right angles on top of the table and run X and Y and the bubbles should no move or 1 or 2 lines. Move slowly. Once you know those results. Set a cylindrical square on the table and indicate Z on both sides. I suspect the machine is worn and that's the main issue.

 

[Vancbiker]

…….This machine is part of a small fleet of three, currently scheduled for obsolescence in 2024 (replacing with new Kern's)……

Kind of like moving from the basement to the penthouse.

 

[imasen deshita]

You need to specify the tests you are doing, tolerances and failure details.
How big of a work envelope do you need?

My numbers above on HAAS and our specs were a little off... fixed that below.
All ballbar tests for x-y-z passed pretty well. It also has the leveling checked on a PM schedule and is good and level.
And yes, the maintenance manager has been to hell and back with the vendor (I'll spare them the name drop) and that's partly why I've gotten involved. It ended up on the back burner after 3x failed attempts (two HAAS rebuilds and a new KME... still curious about how/why that new KME failed, it's a sticking point for me in all this).


Here are some details:
HAAS spec's for the A and B axes, new:


We have our internal spec set at 60 arc-sec for acc and repeat from -60 deg to +60 deg on A axis, and same 60/60 for the full 360 deg on B.

The test performed is Renishaw Planarity Laser Test. This is performed by a legit calibration service, by the way. Annually on every machine.
Plot of results AFTER best attempts to comp for backlash and pitch error.

So it is the A-axis that failed...
The numbers are:


That seems like a big jump from 2021 results which were 54.6 and 21.3, respectively. It was nearly failing our 60-spec at that time as well, though.
I don't have subsequent test results after the two rebuilds yet. I think I know who to ask tomorrow.

Do you think it's possible that this can still be brought into spec again? In other words, is it possible that it just didn't get rebuilt thoroughly enough to address the root cause? I know these things are not terribly complicated mechanically... I find it hard to believe that it can't be fixed.

If it CAN be fixed, and the documentation and parts are available to me, I can fix it. I think the latter may be the less likely.