What are the key factors to consider when choosing a CMM?
This is a big question. First, CMMs are very versatile, but you need to be clear about what the CMM’s function will be. For instance, will it be in a lab inspecting first-off parts or on the shop floor checking production parts 24/7? Will it be checking simple parts or parts with complex geometry? Will it be checking precision machined parts or castings or plastic moulded parts, etc. What are the tightest tolerances you’ll need to inspect and what types of tolerances (i.e. form, position and orientation)? Will the speed of the inspection or cycle time be an important factor? The difference between CMMs has a lot to do with the materials used rather than the basic design, which doesn’t seem to have changed drastically in decades. Modern CMMs are built with lighter materials and smaller, more powerful servo motors allowing them to accelerate to scary speeds (hint: never lean over the bed of your CMM while a program is running to have a closer look at the part. You could end up the victim of a 500mm/sec CMM head-butt, or worse, impaled by the styli.). I’ve personally experienced the head-butt, although not at 500mm/sec, but the styli impaling was a story told by a CMM calibration technician I know.
Materials such as aluminum and ceramic are used to lessen the thermal effect but for different reasons. Ceramic has a very low coefficient of expansion and is therefore very stable throughout temperature fluctuations. Aluminum, while having a much larger coefficient of thermal expansion, reacts to temperature changes rapidly and in a linear manner and therefore can be compensated for accurately and in real time. Having said all that, it still seems the biggest advancements in technology have been made in the area of probing systems and software.
As far as probing systems go, there are many to choose from so I’ll just touch on the main categories. There are two distinct methods, single point probing systems and scanning systems. Single point probing systems take individual “hits” on a given surface whereas scanning systems stay in contact with the surface being measured, gathering hundreds or thousands of points in single sweep. The big difference is point density and how much of the surface you can sample. Behind the probing system is the probe head, which can be fixed or articulating, and is one of the most important factors to consider.
A fixed head means you’ll be relying strictly on the probe build-up (you may need to get creative here) to reach the features to be measured whereas an articulating head gives you the flexibility of hundreds or thousands of probe positions, allowing you to reach many different features with a single probe. There are articulating heads that lock every 7.5° and 2.5° and more recently, a head with infinite possibilities. Another key factor to consider is whether to get a manual indexing head or a motorized head. If you’re only measuring 2D parts and a simple “down” probe is all you’ll need, then of course the motorized head may not be necessary, but if you’re measuring more complex 3D geometry, then I would absolutely go for a motorized head. Having a motorized head on a CNC or DCC machine allows you the freedom to run parts completely un-manned. Take the same CMM and stick a manual indexing head on it, now you’ll need to do some babysitting when running programs that have more than one probe position involved.
Equally as important as the probing system, is the software you choose. Off the top of my head, I can think of at least a dozen different CMM software packages on the market but it seems the lion’s share of the market consists of the big three OEM’s; Browne & Sharpe (Hexagon), Mitutoyo and Zeiss. At Quality Inspection Technologies, we use all three. Being a contract inspection lab we come across all sorts of applications and I like each software package for different reasons and wouldn’t dare pick a favourite (at least not here).
These days you can mix and match CMM manufacturers and software. For instance, my Mitutoyo CMM also runs with PCDMIS and could potentially run with Calypso if I so desired. My advice would be to take a couple of sample parts, typical of what you will want to inspect, and have each potential OEM create a short sample program in your presence, so you can witness how user friendly it is. Take a few features you consider to be your biggest inspection challenges and compare how each software package is able to handle them. While all packages do the simple stuff very well, each has their own special features the others may not have and could make all the difference for your application. This should help you conclude which software is the best fit for your application.
In the end it’s the combination of the machine, the probing system and the software that you’ll base your decision on but it’s a good idea to look at each factor individually as well.
Where is the best place to install your CMM?
In an ideal scenario, your CMM belongs in a separate room that is temperature controlled, free of dust (as much as possible), with a clean and dry compressed air supply, and as far away from the chips, coolant and vibrations of a typical manufacturing floor. Not all scenarios are ideal, which is why CMM manufacturers have answered some of these challenges with such features as real-time temperature compensation and air cushion vibration damping systems, which are extremely important in these less than ideal scenarios. Some pretty rugged shop floor CMM’s exist on the market with completely enclosed air bearings and guideways that can handle the dirt and dust of a manufacturing environment that would ruin a traditional CMM in no time. Some shop floor CMM’s have even done away with air bearings altogether. In either case, there’s nothing to prevent dirt and oil from collecting on the styli so it’s good practice to regularly clean the styli to ensure reliable measurements. There’s also a breed of ultra high accuracy CMMs which call for a clean-room type of environment to produce the sub-micron accuracies but these are the exception and not nearly as common.
Calibration: Is it a given that CMM’s are calibrated once per year?
While a one-year calibration cycle is pretty typical, there are periodic checks that should be done to maintain confidence throughout the year. It’s also a good idea to perform a check after a particularly hard collision; they do happen and it’s imperative to know whether it has a negative impact on the accuracy of the CMM. A periodic check can be as simple as running a program on a retained sample part or test block through a program or as involved as something like the Renishaw MCG (Machine Checking Gauge), which checks the volumetric accuracy of the CMM and is sometimes part of the annual calibration process. It’s best to run your periodic check directly following the annual calibration so you can establish a baseline from which you’ll compare the results from subsequent checks.
You could make the argument that running periodic checks would allow you to extend your CMM’s calibration frequency but there’s more to consider than just maintaining accuracy. In addition to calibrating your CMM, your calibration technician normally performs routine cleaning and preventive maintenance and believe me, I’ve seen CMMs that look like they are cleaned once a year at best, and that’s probably done by the CMM calibration technician. Given the extreme accuracy of modern CMMs and the tight tolerances we rely on them to inspect, choosing to extend your calibration cycle to save money, in my opinion, is a risky proposition.
What key advancements in CMM technology do you expect to see in the next 5 years?
Alright, so don’t get me wrong, I love CMMs and I don’t expect they’ll ever be obsolete, but there have been some extraordinary advancements in technologies such as laser scanning, white light scanning, camera systems and computed tomography (CT). Laser and white light scanners can rapidly scan complex geometry for comparison to nominal data or for reverse engineering and have accuracies reaching 0.0005 in. or better in ideal conditions. Camera technology overall has blown up over the past 10 years leading to advancements in photogrammetry (taking measurements from photographic images) and handheld probing systems that rely on camera tracking systems, soon to leave the portable arm in the dust.
Probably the most interesting emerging technology is computed tomography. Originally developed for the medical industry, it’s been further developed for use in metrology and has the ability to measure internal features, hidden passages, wallstock, defects, porosity and density.