Companies that cut metal have been using measurement-based data to provide closed-loop control of the machined part's size since the 1960's. Early systems utilized a mechanically adjustable tool controlled by a pair of pushbuttons. Each press of the button would move the tool a pre-specified distance. One button would increase the part size, the other would decrease the part size. The machine operator would measure a machined part and use the buttons to adjust the tooling so that parts would be produced at blueprint nominal size. These systems were later automated by using an electronic gage with limit settings slightly above and below the nominal size to indicate that the measured feature was approaching a tolerance limit. When a measurement exceeded one of these limits, the gage would send a signal to the tool controller to mechanically adjust the tool. The benefit of this technique is the elimination of each machine operator's unique interpretation of gage readings and consistency in terms of when the compensation system responds.
The disadvantage to both of these methods was that tool adjustment decisions were being made with respect to the measurement of a single work piece that probably did not accurately reflect the true nature of the process with respect to tool wear.
The next advancement to be made in tool compensation feedback technology came in the way of systems that used absolute and statistical compensation. Absolute compensation is based upon sending the amount of actual offset that is required to adjust the process to the desired level, rather than making a fixed adjustment whenever the process shifts near the tolerance limits. Statistical compensation provides a mathematical analysis of several finished work pieces in order to make the decision when and how much to compensate. The benefit in using statistical analysis to make this determination is its predictive nature for tool wear and machine thermal growth factors. It also filters out random variations in the process (sometimes referred to as "flyers"), which are not attributable to these factors. Most current systems use both absolute and statistical compensation. Naturally, these systems have been ideal for use with CNC based machines that can create parts faster and more accurately than many of the older, manual machines.
The duty of the statistical algorithm is to determine when and how much tool compensation is required. A variety of different mathematical algorithms have been used over the years to make compensation decisions. All of them use size data from multiple work pieces (usually consecutively machined) to calculate the current process level which is then used to make the compensation decision. The current process level typically indicates the average probable size of the next part to be machined if no adjustments are made. Most compensation algorithms use a statistical running average to determine the current process level.
The rapid advance of electronics and personal computers (PC's) in the last decade has led to the development of many sophisticated, yet relatively inexpensive gaging technologies. Although one might expect that most of these systems would incorporate a PC, there are still quite a few that are based on a dedicated imbedded microprocessor design, either in the form of a CRT based product, or electronic column gages. This is because many manufacturers have yet to be convinced of the ability of a PC to withstand the harsh operating environment and extreme temperatures that the systems may encounter. There are also a wide variety of hand held electronic gages, such as calipers, micrometers, and bore gages, that incorporate both the ability to digitally display measurement values, and to transmit the readings to an other device.
Although there are many companies which supply measuring products (gages) to the metal cutting industry, there are only a handful that offer products which can be used to control the process. In addition, some of these suppliers are only willing to offer and support the gage side of the system. When it comes to connecting the gage to the machine for the transfer of offsets, most gage manufacturers and machine tool builders are reluctant to take the responsibility. Another consideration is that the price of such systems may not be cost effective for many manufacturers.
With the wide acceptance of the Microsoft Windows operating systems, there are a proliferation of vendors who supply great looking, powerful full 32 bit, real-time data collection and statistical process control (SPC) software packages. Most have of these programs have the ability to accept measurement-based data directly from gages, CMM's (coordinate measuring machines), and allow for manual data input from the operator. In addition to the traditional X-bar, R, and histogram charting many also include advanced capabilities, such as Individual X and Moving Range, suitable for job shops and short runs. They will also typically include a wide variety of attribute charts (c, u, p, np, PPM, etc.) to monitor defects or non-conforming items.
However, the key letter in SPC is the "C" and counting on manual intervention to control the process is not a guarantee that it will be done correctly, if at all.
It seems that if a company wants to purchase a system that will be used to measure parts and control the process, their choice of suppliers may be limited. Manufacturers may be faced with the dilemma of investing in a quality control system that either informs them that they have already produced parts which are statistically out of control, or that relies upon the operator to make decisions regarding when, and by how much to correct the process.
But not necessarily, because now manufacturers can have a choice. They can still use the gaging products and SPC packages that they prefer while obtaining the benefits of statistically based tool compensation. The EZ-Comp System can accept measurement data from a wide variety of electronic gages. From hand held micrometers and calipers to sophisticated electronic column gages. It can also acquire data from many of the popular SPC software packages available. In a similar manner, the EZ-Comp can be connected to most of the current CNC controls on the market.
Based on a small, single board microprocessor with integrated RAM, ROM, keypad, and LCD display, the EZ-Comp System allows the user to create the correlation between each of the measured features on a machined part and the offset on the CNC that creates the feature. Statistical subgroup sizes (trend group), upper and lower tool compensation limits, and upper and lower reject limits can all be programmed on a per feature basis. As the machine operator measures the parts, the device automatically collects the gage data and determines when and how much to correct for tool wear. This information is sent directly to the CNC with no intervention from the operator. The result is consistency in part size, the elimination of scrap, and minimization of down time due erroneously entered offset values.