Introduction

In the production of complex systems, consisting of many modules, any production deficiencies are increasingly difficult to track down as the production process progresses. Therefore, it is essential that the individual modules are ‘qualified’ before they are integrated into the system. The qualification process includes tests, e.g. to check electrical connectivity, proper functioning of sensors and actuators, and hydraulic and pneumatic circuits for leakage or blockages.

To discover any mechanical deficiencies such as loose bolts or glue errors, the transfer function of the device under test (DUT) can be measured to determine non-rigid-body dynamics. Specific (anti-) resonant frequencies can be used to identify any production defect as well as give an indication for end-product performance.

Ideally, the sensors to measure this transfer function are mounted on a fixed, separate metrology frame (MF) such that the sensors themselves do not move along with the device, as this would disturb the measurements. This static MF contains sensors that measure the displacements of the DUT with high accuracy.

The traditional solution is to separate the MF from the force frame and fixate it to the machine floor. What can we do, however, if this is not possible due to practical limitations? In this article, we discuss how a simple 1D mass-spring-damper model was used to develop a concept support frame, from which the design specifications of the eventual system were derived.

Furthermore, we present the realisation of the solution and compare it to the simplified initial model to underline the effectiveness of the model-based approach for this system. Figure 1 shows the simple ‘model’ representation of the measurement system. The red block designates the base frame (BF) on which the DUT is placed for qualification. The DUT itself is designated by the cyan and green blocks and corresponding springs and dampers. The DUT support is indicated with a purple block. The MF is represented in yellow.

Hence, the main question is how to connect the yellow block to the red block in such a way that the actuator forces F_act and the external vibrations do not excite the MF. In the following section, we elaborate the requirements and the corresponding challenges in more detail, followed by a description of the evolution of the concept design. Thereafter, the resulting mechanical design and corresponding analysis results are discussed.