Printer Friendly

High end CMMs: how much precision do you need.

The coordinate measuring machine has finally arrived. Everybody needs one, but the big question is: How fancy a CMM do we need to measure the parts we make? While most go for $50,000 to $75,000, the best machines and best accuracies cost a lot more. Obviously, everyone would like to have the most accurate machine in the world, and the CMM builders, if there were no cost constraints, would love to build you one. But in the real world, compromises must be made and questions answered about what level of micron splitting your machining capabilities and tolerances really require.

It's a tough competition area, the high end. "There's not a bad machine out there," admits one builder. "They're all very, very good." But which one makes the most sense for your shop?

Arms versus bridges

Although the bridge design is the dominant design in high-precision measurement and inherently more stable than a cantilever, the horizontal-arm machines are gaining acceptance. They are better for automated shop-floor applications, such as reaching out to check parts as they pass on a production line. The latest high-end cantilevered CMMs are now superior to more conventional bridge CMMs, and can bring 2-micron accuracies to the shop-floor environment. Still, for the highest possible precision-splitting microns-the high-end bridge machines in controlled lab environments retain their edge over the horizontal-arm approach.

The concept of moving the measurement machine out of its temperature-controlled closet and onto the shop floor is clearly gaining popularity, says Brown & Sharpe's David Genest. "The Leitz Sirio gives manufacturing people their own gage as opposed to relying on a separate quality organization within the company. This allows them to control their own processes and, in essence, their own destiny."

Challenging the Sirio is the Ferranti Midas, a horizontal-arm machine with an aluminum bridge and ceramic quill. Once they bring acceptable accuracies to the shop floor, the dueling arm machines battle over speed, as we will see later.

Crossed bridges

Among the leaders in breaking the micron barrier -0.000 040"-with their high-end lab-precision machines are two German builders, Leitz and Zeiss; specifically the Leitz PMM bridge machines, and the Zeiss CARAT line (named for the high-thermal-conductivity aluminum alloy used in the carriage and Z ram). The best of the CARATs are the UPMC (Universal Precision Measuring Center) machines featuring a granite table that is thermally insulated around its sides and below to stabilize its temperature. Inside the table are four temperature sensors to detect any thermal gradients that could cause it to bend.

A recent challenger to these German machines is the Sheffield Summit machine. Sheffield's ring-bridge idea-linking the bottoms of the bridge side supports under the table to stiffen the carriage would seem to weaken the table's ability to support loads. Not so, says Sheffield's John Everhart. "We have a three-point supporting system, two legs in front and one in back. All the table load goes directly to the floor with no distortion in the bearing-way support mechanism. The table is sized for the weight capacity of the machine, ranging from 2200 lb for the RL-30 to 3000 lb for the larger machines. If we need to hold more weight than with our standard tables, we simply make the table thicker."

Which material?

Like the Zeiss CARATs, the key structural elements of the Sheffield Summit are made of a proprietary aluminum alloy that is very stiff, lightweight, and stabilizes quickly when subject to thermal change. Ceramic is challenging aluminum with its own advantages of stiffness, heat insensitivity, and low mass, but its application has been limited by its material cost, difficulty in machining, and the limited availability of parts large enough for machine structures. Although some builders mix ceramic, aluminum, and other materials in their designs, most shy from creating dissimilar-material situations.

Reducing moving-carriage mass is vital and every little bit helps. Says Ferranti application engineer, Stephen Edwards, "We were very surprised at how much weight we could save by switching to carbon-fiber air bearings. Conventional air bearings were almost half the weight of the chassis."

Scale choices

For axis positioning, most CMMs rely on glass scales, often of their own manufacture. Some try to match the expansion rate of the scale to the material used for that axis; i.e., steel scales on steel structures, glass scales on granite. Others, such as Zeiss, try for a low-temperature-coefficient glass that will remain independent of temperature effects in the axis structure or drive elements.

Sheffield has a new approach on the Summit: laser interferometry. Explains Everhart: "For ultra-accuracy, what does everyone use? The laser interferometer. So, it only seems logical to use that in the machine itself."

Two key steps in successfully applying the laser, he says, were integrating the laser into the design of the machine, not tacking it on afterwards, and using fiberoptic beam delivery, not mirrors. "Unlike glass scales that mount on the machine and can distort with thermal effects, the laser is independent of machine mechanics and unaffected."

There is a need, though, to compensate laser power for "weather" variations in the environment it must transmit through-temperature, barometric pressure, and humidity-and this is done automatically.

Comparing arm speed

Although the high-end lab machines do not compromise accuracy for speed or make an issue of it, the horizontal-arm machines on the shop floor consider speed a key competitive issue. The Ferranti Midas reaches a maximum velocity of 400 mm/sec max (15.7 ips); the Leitz Sirio, 500 mm/sec (19.6 ips).

Speed used to mean velocity, but no longer, Ferranti's Edwards explains. "Unless you're measuring very large parts, maximum velocity should not be a major consideration because the machine spends most of its time in the acceleration/deceleration mode. The machine rarely reaches full speed."

The Ferranti Midas offers maximum acceleration rates of 6000 mm/[sec.sup.2], compared to the Leitz Sirio's 2000 mm/[sec.sup.2]. But even with acceleration, Leitz' Genest points out, the key evaluation criteria is whether those high speeds detract from repeatability. "Is the machine stable, once you reach the point to be measured? If you get there quickly, but have to wait for the machine to damp out, then the speed benefit is lost."

A fast, ceramic gantry

Fanamation uses a gantry design based on ceramic. All its machines have a ceramic box-beam bridge, and the high-accuracy Comero also has ceramic box-beam legs. Combined with this is the unique use of linear motors for all axes, including both sides of the gantry Y axis, making it, in effect, a four-axis machine.

The result is remarkable acceleration: 1 G or 9800 mm/[sec.sup.2]. Traverse speed is 20 ips per axis for a net 35 ips, and the ability to take up to 120 data points per minute, or two per second. There's no need to insert a delay when the probe reaches its destination to damp out vibration, says Fanamation's Steve Hickerson. "Servodriven machines need to do that, but linear drives do not-they will stop on a dime.

"The problem with linear motors in the past is that they would generate heat that would affect measuring accuracy. Fanamation uses a special air-cooled motor, with coolant constantly across each motor. The heat-resistance and rigidity of ceramic also help."

Flashback to cast iron

The SIP CMM 5 is a fixed-bridge machine: the table provides X movement, and the cross carriage Y and Z. Mechanical bearings are used in the horizontal carriage, and a hydrostatic bearing in the vertical quill movement.

The key to its construction is the rigidity of cast iron, the same philosophy used in machine-tool bases. Says American SIP's CMM Manager, Brian Purcell, "The base, columns, linking bridge, and moving table are all cast-iron, a material we've been using for ages at SIP because it's stable, proven, and supports the sub-micron accuracies we achieve."

The quill is steel. Scales are steel to match the machine composition. All bearings are mechanical, more traditional machine bearings, not air bearings. Because most of the machine bulk doesn't move, SIP feels it makes no speed concession for accuracy. "As far as measuring accuracy and scanning speeds, we are as fast and accurate as the competition," says Purcell, who adds that SIP does not try to compete with the lower-accuracy air-bearing machines or those making concessions to get out on the shop floor.

A machine's material is immaterial, concludes SIP's Purcell. "People today want a machine that's accurate and demonstrates long-term stability. They can be sold on any material. Years ago, material was a selling point. Now it's a matter of demonstrating repeatability and accuracy."

Compensating for temperature

In micron-range measurement, new levels of temperature-compensation complexity are benefiting from faster computing capability. More CMM builders are using volumetric compensation or detailed error mapping of the machine's entire workspace to electronically correct for minute deviations in the machine's mechanics. For a true high-end machine with excellent mechanical stability, this is the frosting on the cake, adding the last iota of accuracy to an otherwise accurate machine.

Temperature is not exactly easy to measure. The questions are how many thermocouples and where to place them to catch all the possible thermal gradients. Sheffield lists a total of 17 on the Summit, including three on each axis, two in the table, and one to four on the workpiece. The Leitz Sirio has 14. SIP measures temperature at one point on each steel scale, plus two points on the part.

However, Zeiss, with confidence in its low-thermal glass scale, does not measure scale temperature or temperature anywhere in the machine structure-only the table and the part. "We don't need temperature sensors in the axes," says Zeiss' Gary Robison. "The scales stay the same, the guideways expand uniformly, and, even though the machine grows, it grows uniformly. Positioning is the same because it's based on scale accuracy. The more accurate the math model, the more thermal error can be taken out."

In the final analysis, says Ferranti's Edwards, "If temperature changes by 10 to 15 deg over the day, you will get errors, but the more accurate the machine is to start with, the better it will perform."

The DEA Omega is the only high-end machine with an air-cooling option. Says DEA's Otto Geiseman, "We can achieve our accuracy numbers without adding air climatization. Cooling is only for those who could not otherwise maintain thermal stability in a hostile environment."

Because the climate-controlled option achieves a steady state, DEA does not presently monitor temperature on either the part or the machine, but they soon will be adding a thermocouple system with nine temperature-measurement locations, including one for the part.

Part temperature

Although the machine may be held at 20 C, parts often come in at a higher temperature, and high-end machines must compensate for this. Part temperature can be monitored automatically and incorporated into the measurement, but a lot depends on the shape of the part, placing of thermocouples, and the time intervals chosen by the programmer to make this thermal correction.

Sheffield's Summit averages up to four part-temperature readings, multiplies this times the part's coefficient of expansion, and linearly corrects all dimensions back to 20 C. It's all automatic, but the user must decide where to place those sensors. "We continue to monitor the sensors during the measurement period and warn the operator of any rapid change," says Sheffield's Everhart. "Also, if we see any drastic change in the probe area, we will recommend that the operator recalibrate the probes."

Vibration monitoring

Part of any initial CMM-installation study is evaluating the normal vibrational environment in which it will operate to determine what vibration-isolation measures are required. Some CMM vendors go further and monitor vibration during CMM operation to alert the operator to transient factors that could affect measurement accuracies. Although SIP offers an isolation system, none of their 22 machines installed here use it, says Purcell, thanks to the vibration absorption of their cast-iron base.

Accuracy issues

The price question-how much to pay for a CMM-hinges on the more basic question of how accurately you need to make parts. Do you have the confidence to live with a 2:1 gage-to-part accuracy ratio, or, must this be increased to 5:1, 7:1 or even 10:1? As CMMs catch up to the latest advances in machining accuracy, the latest trend is a return to the traditional 10:1 gaging ratio.

Also, beyond the accuracy issue is repeatability. While most CMM builders are quick to offer an accuracy statement, some are less forward about discussing repeatability error-the bandwidth you would get at any table location when you measured a length artifact 100 times. Thus, repeatability can resolve a standoff between two machines with comparable accuracies.

If your manufacturing process is incapable of holding fine tolerances, you may argue, why spend big money on a machine to show how widely it varies? The answer, say CMM experts, is that if you have a process that's not capable, and you put in a very good measuring machine, it will tell you a lot quicker what that process is capable of doing than a machine that's as incapable as the process.

With an accurate gage of how the process is doing, you can start fixing that process. Eventually, you'll upgrade the process to the point where you really need a high-caliber gage to differentiate new levels of quality being produced. Without a precision instrument, you can never improve your process quality beyond elementary gage capability-gages that in today's competition are clearly obsolete. Thus, your gage multiplier is a clear indication of your emphasis on quality.

The need for teamwork

One overriding industry problem, says Les Ottinger, VP and general manager, Zeiss IMT Div, is that CMM vendors have been too enamored with their own technology and not concerned enough with how users can effectively apply it. "This industry is just now becoming user oriented. Up until now, as with most technologies, the focus has been on equipment and spec sheets.

"When you team up with the user, versus just being a vendor, you can help them get the most performance out of your machine and best incorporate it into their production system. The CMM isn't making any money if it's sitting idle waiting for part fixturing."
COPYRIGHT 1991 Nelson Publishing
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1991 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:coordinate measuring machine
Author:Sprow, Eugene
Publication:Tooling & Production
Date:Sep 1, 1991
Previous Article:Precision machine reduces grinding time.
Next Article:Specialization is key to success.

Related Articles
The need for interim checks for CMMs (Special Supplement)(Coordinate Measuring Machines)
Focusing on vision.
Selecting CMMs - manual or automatic.
Drawing a bead on CMMs.
How to buy a portable CMM.
Renaissance of the CMM.
Good-bye gages, hello CMM.
What to look for in your next CMM.
IMTS 2000.
An array of measuring options: when quality and speed are key concerns, then measuring equipment characteristics matter more than ever.

Terms of use | Copyright © 2016 Farlex, Inc. | Feedback | For webmasters