What's new
What's new

New CMM decision time!

Mr.Bronze

Cast Iron
Joined
Oct 20, 2014
Location
Duluth
Hey, I got a new question for everyone!

I am going to pull the trigger on a new portable type CMM.
We have demoed a Faro arm in the past, it seemed nice, but I bet there isn't that much difference between brands, the brains of the machines were probably all made in the same chip fab anyways.

So, my question is -
Anyone have a good dealer in the MN region that provides CMMs and related products and service?

I have a full version of mastercam, and would prefer to keep my metrology and Q/C integrated with that, rather than having additional stand-alone programs/file systems. (this is not a discussion on the merits of MC vs other CAM programs, so keep it to yourself if that's all that you have to offer).

The problem is, my VAR for Mastercam, only offers a 4' Romer arm. I feel that this is too small, we commonly make parts in the 24"x 48" range (large bushings) and those are traditionally the most pain in the ass to measure, particularly while in the machine.
The Faro arm that was demoed at the shop was a 9' arm, and it seemed it could cover all of our medium parts without being obtuse, and could measure all of the large parts without stretching.

I know I can just plop the CMM arm on a mount inside the machine, inside the part even; but I don't think a 4' arm can do everything it needs to on a 36" dia x 20" tall bushing, without getting crafty.

And needing to get crafty is a big what I am trying to put a stop, by getting a CMM.
Mostly it's to put an end to garbage parts going out, and to end arguments by customers about whether or not a part was in spec when it left. Trace-ability! hooray! Thin wall bronze bushings are hilariously floppy.

Any thoughts on resellers/brands/sizes?

I'll be putting up a "New machine day" post as soon as I can get the details hashed out, (and my chips sold)

Thanks,
Adam
 
Sounds like you have a plan. If I were looking at the application and the choices, I would likely do some rudimentary working envelope checks to make sure the 9' arm will be operable in all the locations you are considering, especially the "in-machine" places. I tend to agree, the 4' arm sounds too small for the part scale you're talking about.

It's surprising to me that your MC VAR wouldn't offer to try to get you the larger version of their offering. Hexagon certainly has them listed in 8+ and 9+ foot versions.

Faro and Romer seem to be the "big names" in the US. I note that there are some other offerings here and there, such as Kreon (Italy/France) and some others. The Kreon unit looks pretty decent, but all I've seen is the web pages.

Whatever you end up buying, look VERY carefully at the tolerances for repeatability and volumetric accuracy, and make an informed decision about those parameters vs. your part measurement requirements. I think there is greater statistical likelihood of user disappointment due to unrealistic expectations when buying a portable arm-type system than in a fixed-location CMM. Bear in mind that inconsistent probing technique seems to be a contributing factor in inconsistent results, in comparison to standard CMM also.
 
Sounds like you have a plan. If I were looking at the application and the choices, I would likely do some rudimentary working envelope checks to make sure the 9' arm will be operable in all the locations you are considering, especially the "in-machine" places. I tend to agree, the 4' arm sounds too small for the part scale you're talking about.

I don't think machine envelope will really come into play, smaller parts have legitimately tighter tolerances, and things that are <0.001" tolerance under 10" long would be better suited to more accurate measurement methods anyways. Its a lot easier to wiggle that arm in there than it would be for a 40" micrometer.

It's surprising to me that your MC VAR wouldn't offer to try to get you the larger version of their offering. Hexagon certainly has them listed in 8+ and 9+ foot versions.

I really tried to twist my VARs arm over this, asking if hes SURE he cant get another arm. Its not like the carbon fiber is the expensive part in these things, its all the precision feedback and compensation, which are probably identical in all sizes of arms. My rep was pretty staunch about not wanting to take my money though. :nutter:

Faro and Romer seem to be the "big names" in the US. I note that there are some other offerings here and there, such as Kreon (Italy/France) and some others. The Kreon unit looks pretty decent, but all I've seen is the web pages.

Like I mentioned above, I'm not really partial to any brand of arm within the quality mfrs of these. It's more about trying to get it integrated smoothly with the software investments that I already have. Service and support in MN is ideal, although IL is probably were I will have to snoop around to get what I want.

Whatever you end up buying, look VERY carefully at the tolerances for repeatability and volumetric accuracy, and make an informed decision about those parameters vs. your part measurement requirements. I think there is greater statistical likelihood of user disappointment due to unrealistic expectations when buying a portable arm-type system than in a fixed-location CMM. Bear in mind that inconsistent probing technique seems to be a contributing factor in inconsistent results, in comparison to standard CMM also.

Agreed, but I have to assume that at least 'some' of the words I say and the training that my guys will get, lands and sticks about proper operation of these things. There are a ton of things that can go wrong with these kind of measuring devices, poor hand/eye coordination, not understanding sine errors with small angles, haste.
The plan is to have it as our redundant backup for Q/A purposes, and to catch feature errors that cant really be "measured".

The FARO 9' arm specifies an a volumetric accuracy of ±0.0016 over the reach of the arm. and a single point repeatability of 0.0011" (not sure if that results in a ±0.00055"?)

From a metrology standpoint, yeah, sure, many of our parts cant be "measured" by this, since they have tolerance bands in the ±0.002" (IIRC, you want 10x the accuracy as the tolerance band you are looking at?) But at the end of the day, more measurements will be accurate than not accurate, and it is a second look at the parts.

Additionally, our nuanced discussion about tape measure precision vs what size you are measuring is a concept lost on 'most' customers and inspectors of incoming goods.
Our hand-written inspection sheet done by micrometers and skilled hands may be more reliable than the CMM printout; but the unfortunate truth is a fancy computer printed report will hold much more authority for 99% of situations. Not ideal, but at least I'd have formality on my side.

Most of our parts and a vast majority of the features have tolerance bands larger than that, and it allows for feature detection, such as hole placement, grease grooves existing, as the operator follows the procedure for inspecting the part.
And for the most part, the big ass bushings are ±0.005 on a 3' diameter, which should be legitimately verifiable by an arm of that accuracy/repeatability.
 
I'm surprised, and a bit pleasantly so, that they are backing off selling you what you think you want.
At 1.6 arm on a 5.0 tolerance it sounds like a solution the end user won't be real happy with down the line.
At the bottom level inside these things measure angles not position or linear size. For this reason they don't scale nicely so small ones do much better.
Still, It's weird not to see take the money and run.
Ask them "why".
You do realize that your +/-.005 part print is now a +/- .003 inspection print and you have only two ticks to work with.
On a 2 part this gauge becomes worthless.
Bob
 
My rep's answer to "why" had nothing to do with measuring capability, but more of "we only offer the 1 arm"
Which, after hearing some hesitant feedback here.... is kind of concerning that he didn't have anything to say about the accuracy/repeatability relative to our parts.

Yes, I realize that it is no good for stuff under a ±0.002 window, but as I said above. A big part of this is for feature verification and measurement with electronic documentation, in addition to the precision measurements that have to be done with a micrometer. Plus, the ±0.002 or smaller tolerances really only happen on OD/ID, where it is reasonably easy to get a mic in there.

I will inquire about their more precise arms, and grill the rep about the capabilities of the arm, with respect to the parts that it will be used on.

At the end of the day, the amount of rework that was done this last year(and every year), plus the pattern making we do for reverse engineering, would have paid for this machine already. EASILY. And most of the re-work isn't because we missed a tolerance by a tenth, it's usually because someone just plain forgot to put a hole, the buttress threads are facing the wrong direction, stuff that would easily be picked up by a second inspection on an arm.

We spent at least 150 hrs last year dicking around with making patterns from customers samples, if I could probe the part (they are weird shaped bronze or aluminum doodads), generate a file that can be converted into a 3D model in a few minutes, rather than trying to produce these shapes in by hand in CAD, or doing it super old school and making casting, then bondo, then casting, then bondo, to get a pattern; well.. that alone would probably justify the $50k for the arm and accessories.

The software that comes with the arm also provides an incredibly valuable QA function. Having an inspection protocol in place prior to the part going onto the machine, even for double check and verifying that all the features are in the correct place is what may take precedence over the arm's inability to reliably measure our high precision parts.

I normally say, "you wouldn't believe how many times that hole that's supposed to be on the left hand, is on the right hand"... but I am sure many of the people reading this know how often it happens... TOO OFTEN. Or 0.125"R grease groove when it should be 0.25"R groove, stupid shit like that, that causes me massive headaches. It is definitely a symptom of other deficiencies my shop has for its QA program, but we make such widely varying parts... like 12000 different parts over the past 15 years, I need a good integral process to facilitate part verification with flexible solutions.
 
If you are just wanting this for on machine inspection and verification I'd take a renishaw OMP probe over a romer arm any day of the week. Now it won't be that nice to use for reverse engineering but if you do a CPK study of your machine it will be a lot more accurate than your arm...
 
If you are just wanting this for on machine inspection and verification I'd take a renishaw OMP probe over a romer arm any day of the week. Now it won't be that nice to use for reverse engineering but if you do a CPK study of your machine it will be a lot more accurate than your arm...

If we did runs of more than 1-10 parts... a CPK would be valuable... but we don't even get into the 50 part runs most of the time.

Again, the arm was never intended to replace the precision measuring tools.
Half the cost of the project is the software behind the arm, and integration into mastercam, and solidworks, all of which have extraodinary value, not even considering the arm and its precision. A quarter of the cost can easily be covered by pattern work and reverse engineering. So that leaves about 12k to be recovered over 3 years by the arms measuring capabilities - which, as stated above, would have been paid off in 1 year, easily.

You guys would not believe the disconnect between the part/process I quote, and the part/process that is used for manufacturing. Any tool that helps me integrate the entire process, quote -> sales order -> work order (casting dimmensions, CNC program, tooling, MTRs, etc) -> part inspection -> shipping-> and back to quote, is going to allow me to move the business forward.

My company is not profitable/large enough yet to hire someone purely for quality control, so that's not an option. Not to mention, many of our inspections are done on a big VTL, and I find it really hard to believe that an arm would be invaluable for our big, high dollar parts, with reasonable tolerances, and hard to measure geometry.
There is also something to be said for the constant re-iteration of inspection and quality(same with safety), rather than it being a side note for production, it being core to production.

To play devils advocate here, and an arm is not the best investment for improving the process where do you guys think the money would be better spent?

The root of my problem here is poorly laid out quality assurance protocol - which results in rework and angry customers.

Instead of poking holes in my bucket.. lets hear some constructive suggestions!
 
Sounds to me like you are expecting (hoping?) that a new tool will solve what I think might be systemic attitude issues. That may sound harsher than it's meant to, but if there isn't a core of highly competent guys in the shop who WANT to make sure the right stuff goes out the door, no amount of new gadgets is going to help.

Especially with a portable arm system, if you don't have the same couple of guys running that show pretty continuously, you may get results that vary enough to screw up your existing system, whatever it is. I'm all for better tools, sometimes to my own detriment, but if you take this particular step, it would be a good idea to have a companion system in mind for more closely controlling what goes out. When a shop relies on the machinist making the part to check his own work, there are inevitable conflicts of interest, especially where delivery schedules are part of the pressure.

"The root of my problem here is poorly laid out quality assurance protocol - which results in rework and angry customers"...
This is what needs to be addressed, in terms of how stuff is checked for compliance. The use of the portable CMM has to be part of the larger system with which EVERYONE agrees, unless the head guy is going to lean on it all the time, every time. Conversely, maybe the purchase would get everybody excited about participating in a revamped QC/QA system. Long story short, though, the machine has to fit into an integrated functional QA structure.
 
One has to ass-u-me the machine is not running while you use the arm.
So a in the spindle probe can't do the measuring/part verification with a star probe tip?
Or is your machine itself sloppier than the thou plus from the arm?
Bob
 
It is harsh, but it is sort of the truth! No offense taken.
I wouldn't pin it so much as my guys don't care, as much as just being a huge uphill battle right now, because there has been such piss poor QA and workflow feedback for like.... 120 years... I commonly say I am trying to bring the company into the 20th Century; it usually lands as a joke, but it's not that funny.
I've really made a lot of improvements to our process already, and have put a huge emphasis on the importance of continually trying to improve. Up until July of last year I was really able to focus on that, but the mining/oil tank of the second half took my attention away from throughput to getting orders. Had all of our customers go into hibernation for the last 6 months of 2015, we would have already had the CMM, plus several other improvements.

But, business is back up, for varying reasons, so I have been looking at investments in capability, process and tooling.
Just last week started requiring 2 separate inspections for a large majority of jobs. 1 on the machine and 1 prior to shipping, by 2 different people. Soon I will put out second forms that are a redacted print with cells to fill in the blanks, and highlighted features - so no just writing "fun numbers that work".(I'm sure that is rare, but I guarantee it's happened)

I think I have been dissuaded, from a metrology standpoint, from getting a CMM arm. But as I mentioned previously, there is at least 1 other use that probably justifies the investment - pattern-making and reverse engineering. Plus castings need to be measured routinely at some point in the future. Plus verifying difficult to measure features and "loose" tolerance parts are a breeze.

So I still am probably going to throw money at one. :dopeslap:
 
Conversely, maybe the purchase would get everybody excited about participating in a revamped QC/QA system. Long story short, though, the machine has to fit into an integrated functional QA structure.

Getting them excited about quality parts, and upping our game as a whole business is also huge part of it. Before I started causing a ruckus about processes and tools that we were missing out on, both from a "running the business" standpoint and production related stuff; it was pretty classic short sighted, small business operations. Too much sticker shock about both dollar costs, and effort costs, rather than looking at the value of that improvement - and it's subsequent increase in profitability.
Sure when they were a foundry in the 50's paperwork was for idiots, but running a business without the best tools available these days (computers and document are tools too!) is a dead end road.
 
One has to ass-u-me the machine is not running while you use the arm.
So a in the spindle probe can't do the measuring/part verification with a star probe tip?
I think in machine probes will be looked at in depth before I make a decision. It seems like that would be a smart thing to do.


Or is your machine itself sloppier than the thou plus from the arm?
Bob

I would hope so! We do ±0.0005 on 10" dia every day for some of our big repeat jobs. (those would not be measured by a CMM.)
 
Getting them excited about quality parts...
Attitude is 150% of the quality control equation.

If they WANT to make quality parts, they WILL make quality parts.

If they don't, they won't.
=====
Present any QC process as a tool that supports their efforts, not a hammer that finds fault.

Present quality awards every month, with real $ to folks who have zero defects.
Their attitudes will change rather rapidly.

From a business standpoint it costs much less to reward good work than to redo bad work.
=====
Whenever you have a QC failure, IMMEDIATELY investigate to determine the cause.
Is there a problem with a machine, or with a particular gage, or with ???

The last thing you want to do is blame an operator for something that's not his fault.
That will kill morale faster than anything.


- Leigh
 








 
Back
Top