The use of Large Volume Metrology equipment – such as Laser trackers - is becoming evermore prevalent in industry where its ability to inspect and build large components directly to CAD nominal data offers significant benefits both in terms of the accuracies achieved and time savings in the manufacturing cycle.
As the number and variety of this type equipment grows, users are facing the challenge of ensuring that their operators are proficiently trained on the various software packages that are provided with these systems. In addition, as the reliance on this type of system grows – users are asking “What is the uncertainty on my measurements” rather than just accepting the numbers off the screen displayed to the last micron.
Spatial Analyzer (SA) software has enabled users world-wide to solve both these problems. Spatial Analyzer is a traceable metrology software platform that offers users the opportunity to utilise a common software platform for all their instruments. For example, the Laser Tracker interface is identical regardless of which Laser Tracker you are using , such as API, Faro, Leica, SMX - see Fig 1. Other equipment, such as total stations and Theodolites, can also be incorporated with equal ease. This enables operators to run any of these instruments with minimum training and makes software updating easy to manage down the line.
Much work has been done to understand and quantify the performance of the various portable large volume metrology systems. The manufacturers of the measurement devices publish performance specifications. These ‘laboratory’ specifications are not, however, representative of the actual performance of the instrument on a users application. Many other significant effects are prevalent and ignored including operator contributions to the uncertainty and the variability of real-world shop floor measurement environments. The manufacturers uncertainty statements often do not address the geometric nature of the co-ordinate uncertainty but instead provide a volumetric statement based on the spherical uncertainty at each point.
Many large volume metrology processes require more than a single instrument in one location. Examples include commercial airplane manufacture and shipbuilding. These measurement applications necessitate either a combination of various measurement devices or the relocation of a single instrument in many positions to acquire the necessary data. Users often combine measurement systems by tying individual measurement systems together based on common reference points and then assume that they are working within the instruments published uncertainty. Alternatively, many users apply heuristics to determine the uncertainty as they progress along the chain of measurements. Both these methods provide poor approximations of the uncertainty in all but the most simplistic of applications. Even in cases where a single instrument is used and from one location its measurements are typically tied in to a reference coordinate frame. The uncertainty from this tie in process is often ignored.
The ISO (International Organization for Standardization) standards focusing on Global Product Specification require that part measurements be described by two numbers. The first is the result of the measurement and the second is the stated uncertainty. This uncertainty statement represents the estimated variability in the result. This specification mandates uncertainty statements in order to provide traceability for measurement results. In addition, it is recommended that measurement systems provide uncertainty statements in order to be considered accredited systems (Forbes and Harris, 2000). NIST states that a measurement result is complete only when accompanied by a statement of its uncertainty.
These requirements and recommendations reflect good practice in the production environment. Expensive part rework and costly delays can result from decisions that are made based on unreliable coordinate measurement data. These decisions should be backed by a rigorous knowledge of the uncertainty in the measurements that are used to pass or fail parts.
|Point ID||Total Station||SMX||Leica|
The USMN Process
The goal of the Unified Spatial Metrology Network (USMN) is to properly combine nominal data and measurements from multiple coordinate acquisition systems to produce a measurement network complete with a realistic statement of the measurement uncertainty. The simple example below shows the basic concept of how USMN works. There are three instruments in the job.
A Total Station – which has measured a point group called MeasS1 with points p0, p1, p2 p3,p4 and p5.
A SMX Laser Tracker – which has a measured point group MeasS2 with points p3, p4, p5, p6, p7, and p8·
A Leica Laser Tracker – which has a measured point group MeasS3 with points p2, p5, p6, p7, p9, p10 , p11 and p12
The figure below shows the network of instruments and points along with the measurements which networks them together.
After running through the USMN process see Fig.2 . A USMN composite point group is produced. This group contains the optimal point co-ordinates after solving the network. Inputs to USMN include the instruments and their measurements. USMN manipulates instrument positions to get minimal measurement closures on the common points. The optimisation process uses the estimates of the instrument uncertainties and the range of each observation to weight the individual contributions for each measurement.
There are three individual processes in USMN which are used for three distinct purposes on the measurement network. The first of these functions solves the network. It optimises the complex multi-station measurement network to solve for the best station and target locations and it produces the best point co-ordinates from the measurements.
The second component analyzes instrument performance within the network or, as USMN refers to it, uncertainty. With the optimal network (Stations and target locations), USMN computes instrument uncertainty estimates based on the measured network of targets. This analysis characterises and produces reports for the instrument and individual station performance estimates from the actual measurement network.
The third piece estimates target uncertainty. With inputs for instrument uncertainty and the actual geometric network of measurements, USMN computes estimated target field uncertainties. These are represented in SA by point clouds which clearly show the shape and density of the uncertainty fields. l
Brunson Instrument Co. is an ISO 9001:2000 certified manufacturer of large volume metrology instruments and accessories and is the master distributor for Spatial Analyzer software.
Spatial Analyzer is developed by New River Kinematics