Test Setup

4.1.2 Stand Based

Exceeding the clamping force of the chuck during rotation results in large reaction forces that need to be compensated by the human operator. This might easily lead to dangerous situations. Therefore, another test setup has been developed, in which the machine is held by a stand and the displacement sensors for the core bit are placed on the ground. A picture of the setup is shown in Figure 4.4. The stand holds the machine at the screw connections for the side handle, which are available on both sides since the side handle can be mounted either for left or right-handers. Rubber-bonded-to-metal components are used for the connection. At the rear end, the machine is supported by a mount of elastic material, which has been formed by the rear handle in a casting operation. The challenge is to support the machine rigidly enough to ensure safe operation, while, at the same time, keeping the rigid body resonances of the machine in the support low enough in frequency to allow a clear separation between support and structural resonances. The displacement of the core bit is again tracked by triangulation lasers – this time fixed to the stand, therefore measuring in an absolute coordinate system. This is in contrast to the hand-held measurement setup in Chapter 4.1.1, in which the lasers are attached to the machine via the additional rig, and therefore reference a local, machine-based coordinate system.

Figure 4.4: Measurement setup in which the machine is supported by a stand

4.2 Test Setup

While in the previous section the experimental setup was described with respect to the applied sensor systems, here the focus lies on how the actual measurements are performed and how the data is post-processed to aggregate the obtained information.

4 Experimental Setup

4.2.1 Operational Measurements

One of the main objectives of the underlying research is to study the dynamic system behavior when crossing a bending resonance during run-up and run-down. The tests are performed without a drilling operation, which means that the core bit can rotate freely without any contact with the underground. As a matter of protection, a kind of bunker contains the core bit when the tool is running. Although a person of average physical strength is able to safely hold the tool including the extra weight of rig and sensors, additional elastic cords support the weight of the system in order to relieve the operator during hand-held measurements.

Figure 4.5 shows the corresponding test setup. For stand-based measurements, the setup corresponds to that in Figure 4.4.

Figure 4.5: Test setup for hand-held measurements

During tests, the rotational speed is indirectly controlled through the voltage supply of the electric drive. To run up the tool, voltage is ramped-up in a linear manner. The same applies accordingly for a run-down. However, even if there is no external load on the core bit, the rotational speed cannot be perfectly controlled through the voltage supply.

When crossing the critical speed of the resonance, part of the energy that is fed to the electric drive is transfered into the lateral bending motion, instead of further accelerating the spinning motion. This phenomena is known as the Sommerfeld-Effect [28], named after Arnold Sommerfeld (1868 - 1951), and will be discussed in Chapter 7.4.3.

4.2 Test Setup

4.2.2 Experimental Modal Analysis (EMA)

Experimental modal analysis (EMA) is a powerful tool to study the dynamic behavior of a vibrating system and to validate and update theoretical models. Nowadays, it can be considered a standard approach in research and in the industry. The procedure is described in many textbooks and will not be repeated here. The works of Ewins [31] and of Maia and Silva [89] are well accepted and frequently cited. In the underlying research, EMA is used to analyze single parts of the rotor system, certain assembly groups, as well as the complete test setup as a whole. This chapter describes, how the FRFs have been measured and how they have been analyzed. Chapter 6 discusses the results of the modal analysis.

EMA generally consists of two steps: First, FRFs need to be measured in decent quality at sufficient positions of the structure to geometrically resolve the mode shapes within the frequency range of interest. In a second step, the modal parameters are extracted by using the appropriate identification methods. In the underlying thesis, the single components are measured in free boundary conditions as described in Chapter 4.2.2. The so-called PolyMAX algorithm is then used to estimate the modal parameters. PolyMAX is a polyreference least-squares complex frequency-domain method that was introduced in [50] and can be considered a benchmark in modal parameter estimation [108, 109]. One of the advantages of the algorithm is the fact that it allows one to separate closely spaced modes. This fact comes in handy with rotordynamic problems, in which, due to the more or less symmetric geometry of typical rotors, double modes are the rule and not the exception.

Single components are measured with free boundary conditions, which is also the most common approach [31, 110]. Figure 4.6 shows, as an example, the test setup for FRF mea-surements on the shaft. The roving hammer technique is applied, using two accelerometers (two references) to identify the double modes of an (almost) axial symmetric object.

Figure 4.6: Test setup for FRF measurements on the shaft in free boundary conditions Impact testing is also applied to measure FRFs on the system as a whole. The focus of interest lies on the first bending mode, which is a global mode, as will be shown later. The

4 Experimental Setup

roving-hammer method is used to identify the bending mode, as well as to study the dynamic behavior of the housing and the hand-grips of the machine in detail. In total, 156 points are measured. The measurements are performed at a standstill with free boundaries conditions, when the system is simply supported by the elastic cords mentioned before.

In addition, measurements are performed while a human operator holds the tool to catch the influence of the user on the modal parameters. Therefore, an operator grips the tool at first just at the rear handle, then just at the side handle, and, finally, applies a grip at both handles. The measurements are repeated for different amounts of gripping force. To avoid repeating the FRF measurements for all cases at all 156 points, the results of the detailed EMA are first evaluated: The system shows no other modes in the frequency range of interest other than the bending mode mentioned before. That means that the previously described sensor setup for operational measurements allows one to identify the bending mode with a single excitation point. Good excitation points turned out to be the front end of the core bit and the top side of the rear handle.