Data Acquisition Systems for Operational Modal Analysis

Operational Modal Analysis is today performed on a wide range of structures, under various operating conditions, in different test environments and often under extreme time and cost constraints. As a result, numerous requirements have to be taken into consideration when designing an optimal data acquisition system for Operational Modal Analysis. This includes the ability to make very accurate measurements in a fast, safe and easy manner regardless of whether the system is to be used in the field, in a lab, or in a centralised or distributed setup. In addition, modularity, scalability and Plug & Play are functionality expected from a modern data acquisition system. Many of these requirements are, however, very divergent. Consequently data acquisition systems for operational modal analysis have up until now either been optimised for dedicated use or been the result of a compromise. This paper describes the new LAN-XI concept combining several different technologies including Precision Time Protocol (PTP), Power over Ethernet (PoE), Dyn-X, REq-X and TEDS. With the introduction of the LAN-XI concept, a hitherto unseen combination of requirements can be fulfilled in the same data acquisition system thereby significantly increasing the versatility of the system and at the same time reducing the amount of measurement equipment and accessories needed. 216 IOMAC'09 – 3 International Operational Modal Analysis Conference 2.1 Nature of test object From being a novel technique mainly done on large civil engineering structures, OMA is today performed on a large variety of civil engineering as well as mechanical structures ranging from the smallest components to the largest structures. Measurements range from a few channels to hundreds of channels and a versatile data acquisition system must therefore be scalable, modular and easily reconfigurable. The same data acquisition system capable of doing a large multi-channel measurement using up to several rack systems one day, should be easily dividable into multiple systems the next day for doing many smaller measurements at different locations without compromising performance, ease-of-use or cost. In general, it is advantageous to place the data acquisition system as close as possible to the test object to shorten the length of the transducer cables. Apart from significant cost savings on expensive high-quality transducer cables and the lower risk of setup and measurement mistakes due to reduced cable “infrastructure”, short cables minimise the potential risk of adding noise to the measurement data. For large test objects, this requires that small front-ends – or individual modules can be freely distributed around – or inside the test object and the measurements can be made 100% sample synchronously between all measurement channels. Many traditional systems are, however, neither distributable nor capable of ensuring sample synchronisation between frontends. Newer systems have offered various cable-based synchronisation techniques between the individual front-ends, but all with the disadvantage of requiring extra cabling. 2.2 Test environment Historically most OMA measurements were performed as in-situ field measurements, e.g. for civil engineering structures like buildings, towers, bridges and off-shore structures. Many OMA measurements are still performed in-situ including in-operation measurements on mechanical structures like aircraft, vehicles, ships, trains and rotating machinery in general. However, many civil engineering structures are also measured as scaled models in labs and especially many mechanical structures are measured in labs and test cells. Field measurements are often performed in harsh environments placing high demands on the robustness of the data acquisition system. In addition, the data acquisition system must be light weight, small and truly portable, have low power consumption, have the option of battery operation and the possibility of distributing the front-ends. For fixed lab or test cell setups, a centralised data acquisition system with one or more rack systems has traditionally been used. Long transducer cables have been required with the significant disadvantages previously mentioned. In both test environments the use of distributed front-ends could significantly reduced the required cabling. Fig. 1 shows a traditional test cell setup using long transducer cables between connector panels in the various test cells and the operator room. Figure 1: Traditional test cell setup requiring extensive cabling.