Efficient and effective process control can be performed by installing coordinate measuring machines (CMMs) on the shop floor, close to production equipment. By using CMMs for process control, machining anomalies can be quickly identified in real time, allowing operators to take corrective action before out-of-tolerance parts are produced.
Process control does not replace traditional quality control; it complements it. Quality control procedures aim at determining whether the part conforms to nominal dimensions and tolerances. This kind of inspection procedure occurs under ideal conditions. The measuring system is usually installed in an environmentally controlled room or quality lab, the part is washed, degreased and thermally stable. The lengthy period between physical production and inspection of the part is problematic. The production line continues to work as usual. If anomalies are found in the final inspection operation, a number of defective parts already may have been manufactured.
The difficult environmental conditions on the shop floor, including temperature, atomized oils, dust and vibrations, make it difficult to measure parts accurately. However, the significant improvement in inspection cycle time, possible using a CMM on the shop floor, makes the problem worth tackling.
The theoretical dimensions of parts are defined at the reference temperature of 68 F, or 20 C. During in-process inspection, parts generally are not at this temperature because of the effects of ambient temperature, machining and washing, and type of material. The results of measurements performed on the parts at a temperature other than the reference temperature can be "corrected" by applying the expansion coefficients of the material. However, residual errors can be caused by:
• Uncertainty of the expansion coefficients of materials, that is, variations between the measured expansions and the calculated expansions.
• Uncertainty of the temperature sensor used for reading the temperature on the part.
• The temperature-dimension instability of the part during measurement. If the part is not in thermal balance, heat propagation generates variations in the part geometry. The extent of these variations depends on temperature, material and mass.
• Errors also can be caused by dirt on the part, such as oil film and chips, and dirt on the CMM's probe caused by repeated contact with
the dirty part.
Influence of machining
Part temperature also is a function of the process. Different types of manufacturing operations have different effects on the thermal properties of the part, and consequently, on its inspection.
With traditional machining using individual machine tools, an operator is responsible for collecting, cleaning and certifying the part. This procedure makes inspection similar to quality control rather than real process control. Often there is a significant period of time between the end of the machining cycle and the beginning of the inspection cycle. This time period generally stabilizes the part temperature at the ambient temperature.
Flexible manufacturing cells (FMCs) are made up of several machine tools and one measuring machine integrated into the process, often fed automatically. Whatever the sampling rate, which can be quite high in these applications, parts are measured at regular intervals and the time between the end of the processing cycle and the beginning of the measurement cycle is quite short. Part temperature appears to be a function of ambient temperature and machining and washing. Since a washer is seldom present in the FMC, parts often are covered with an oil film a few microns thick prior to inspection. Because of its viscosity, oil often sticks to the probe, causing a variation of the probe-tip radius, making it difficult for the probe to move on the workpiece.
A transfer line consists of dedicated machine tools performing sequential operations with automatic transfer of the parts between the stations. In a normal flow, parts are measured midway or at the end of the processing cycle and occasionally at the beginning. If anomalies are found, it is possible to determine, by measuring in an intermediate stage, whether the problem is upstream or downstream of the operation.
Subsequent measurements between the operations will locate the source of the anomaly. Like FMCs, the part to be measured is at an unknown temperature, different from the ambient temperature.
In processing with flexible manufacturing systems, part inspection is performed by one or more measuring machines integrated into the line.
The type of material being machined is an important factor affecting thermal stability. Materials most commonly machined are aluminum, cast iron and steel.
The form and structure of aluminum makes it a material of choice for thin castings, with limited thickness, and stiffened by ribs for more resistance. Aluminum is characterized by high thermal conductivity and tends to distort quickly as temperature varies. It also quickly reaches thermal stability. Aluminum is generally machined under a coolant that reduces part temperature. The liquid settles on the surface creating a thin film of oil that might alter part dimensions by a few microns if not removed.
Unlike aluminum, cast iron is characterized by high thermal inertia and is both dry processed and machined using jets of coolant, particularly in finishing operations. Cast-iron machined parts are characterized by high masses of material and are subject to long states of thermal instability.
In terms of thermal conductivity, steel is rated midway between aluminum and cast iron. It is generally machined under a jet of water. At the end of the machining cycle, a steel part is at ambient temperature, but, like aluminum, it is covered with an oil film.
A part inspected on the shop floor is in one of two thermal conditions:
• A state of balance, being the same as the ambient temperature, but typically not 68 F.
• A transient state, different from the ambient temperature, but not 68 F.
In both cases, measurements must be corrected and the values brought back to the reference temperature of 68 F.
CMMs designed for operation on the shop floor are generally equipped with a temperature sensor for the detection of part temperature. Part temperature can be measured both manually and automatically. For manual detection, the operator has to bring the sensor into contact with the part and keep it in position for several seconds. This is tiring for the operator and is subject to operator influence because even a slight variation in the length of the temperature detection procedure can influence the accuracy of the measurement result.
As an alternative to the rod-shaped sensor, a magnetic sensor can be used. It is applied to the part. While the sensor detects the temperature, the operator is free to unload the inspected part. These sensors, however, have long detection times-approximately 10 minutes-and their use might be incompatible with the need for high inspection throughput.
Automatic detection is performed by integrating one or more temperature sensors in the supporting and clamping fixture. This approach makes operator involvement unnecessary for temperature detection.
In both the manual and automatic approaches to temperature detection, the temperature taken by the sensor is communicated to software controlling the measuring machine. The software performs the necessary compensation of the inspection results.
If the part is thermally stabilized, temperature reading, dimensional inspection and compensation are reliable. A method used to check whether the part is actually thermally stabilized consists of taking temperature readings at intervals of a few seconds. If the readings of the temperature sensor are consistent, the part is thermally stabilized.
If the part is in a temperature transient state, it is difficult to read the temperature correctly and obtain correct measurement results, even though part temperature compensation is applied.
It may happen that the temperature is not constant everywhere on the part because machining might have unevenly affected temperature. Heat propagation in a body depends on material properties, as well as on its shape and dimensions.
If the part is not thermally stable, it is best to perform only short inspection cycles on the most critical part features, generally those with the tightest tolerances.
A reliable solution to protecting the measuring machine from the shop floor environment is to place it in a cell. If the temperature in the cell is the same as the ambient temperature, as is the case when a ventilated cell is used, there will be no further thermal effect on the part. If the cell is air-conditioned instead, the difference between part temperature and cell temperature might generate dimensional effects on the part. In this case, inspection cycles should be kept as short as possible.
Compensating for the effects of temperature and thermal gradients on a measuring machine is a challenging task. Just as complex is to effectively and correctly apply compensation factors to the dimensional data from inspected parts. Measurement accuracy and reliability are prerequisites for effective process control. As a result, if it is not possible to measure parts in thermal balance, it is important to inspect them under ambient conditions, similar to the way they were machined. In this way, the dimensional variations detected can be attributed to the process. Q
Anna Maria Izzi is a product manager at DEA (Torino, Italy), a Hexagon Metrology company. She can be reached at email@example.com or + 39 011 4025380.
• By using CMMs for process control, machining anomalies can be quickly
identified in real time, allowing operators to take corrective action before out-of-tolerance parts are produced.
• The difficult environmental conditions on the shop floor make it difficult to measure parts accurately.
• The type of material being machined is an important factor affecting thermal stability. Materials most commonly machined are aluminum, cast iron and steel.
• CMMs designed for operation on the shop floor are generally equipped with a temperature sensor for the detection of part temperature.