Drones & Robots
A growing class of devices need no, or only minimal human interference for their operation. When these devices, like drones and robots, operate around humans it becomes important that their noise emission is controlled to be quiet, pleasant, or functional. No one would like a nursing robot to rattle, to roar, to clunk, but you would still like to hear it when it is near.
Excessive vibration could also inhibit accurate operation, or even cause damage.
The thrust and functions of the device require electric drive, high speed rotation, mechanisms, transmissions, flow, etc. Unfortunately these generate impulses, vibration and turbulence. On top of this almost all mobile autonomous devices have to be light, or even extreme light weight. This means there is work to be done to control the sound emission: balancing, shifting modes, controlling impedance at interfaces, and isolating. Simulation models, digital twins, developed and updated. Models require checks, correlation with tests, missing parameters to be estimated and non-stationary loads need to be identified.
A well developed device gains value and reputation in the market.
Some applications
Inverse load identification
To be able to simulate and compare variants and concepts using numerical models, realistic dynamic load information is required. These may require experimental data, for example from a test-bed with the drive units or from first prototype hardware. In most cases the loads cannot be measured directly with force sensors. In that case an indirect estimation of the dynamic loads is possible using inverse force identification. This technique relies on the combination of an operational measurement with artificial excitation.
Accelerations, surface velocities, pressures, or strains are measured near the suspected forces as response signals in natural operation of the device or system on the test-bed. In the next step shakers excite the load interface for each possible force and direction and the same responses are measured and processed into an FRF matrix. Various approaches are possible, but basically the FRF matrix is inverted and multiplied with the vector of operational spectra to obtain force spectra estimates.
On the measurement side a main challenge is controlled artificial excitation at the force interfaces. Instrumented hammer impacts tend to produce an inaccurate FRF matrix and access can be difficult or even impossible.
Externally supported shakers with a force cell, when possible, can cause significant mass loading on light weight structures and stinger modes limit the frequency range.
Qsources developed its dynamically decoupled and compact shakers with load identification and transfer path applications in mind. Smaller than anything else, working well and aligning in any inclination, very low coupled mass in the working frequency range has allowed to extend the scope of inverse load identification limit significantly.
Our partners Siemens and Head Acoustics can advice further and provide software for this process.
Some shakers for structural analysis from small and higher frequency to higher force level and lower frequencies:
Experimental modal analysis supporting FE analysis
Components and frames are designed to withstand the loads and provide sufficiently high stiffness for accurate operation. The assessment on whether a sufficient level is achieved, within costs, and maintaining light weight is based on FE based models. The most reliable way to verify these FE models, especially when assemblies with multiple interfaces are included, is EMA experimental modal analysis. The knowledge from the correlation between the test results and the numerical models can be used to optimize and assure the effectiveness of the designs. Qsources provides highly accurate and efficient excitation devices for EMA applications.
Some shakers for structural analysis from small and higher frequency to higher force level and lower frequencies:
Sound sources with real time volume signal can be used for scaled EMA analysis for vibro-acoustics or pure airborne enclosed volumes. These sound sources can, for some difficult applications, also be used as indirect non-contact excitation of structures to determine their structural modes.
Sound power determination
To communicate the noise emission created by any machine, device, the sound power level is often used. Less intuitive than sound pressure, but much more relevant because sound pressure is very dependent on the nature of the environment and the distance to the device. Sound power allows combination with models on sound propagation and sound isolation.
More complex descriptions of noise emission are also possible like equivalent volume acceleration spectra, or the distribution of volume acceleration over the surface of machines, devices. (See automotive applications, ASQ)
For the practical sound power, the measurement in natural operating conditions is possible in dedicated test-chambers, by detailed sound intensity scanning, and by using reference sound sources.
The latter, using sound sources, is the most practical, because it can almost always be performed on-site, in basic labs, or in the natural operating environment. Beside the reference sound source, at least one microphone is needed. The procedure is well described in ISO 3747.
Qsources reference sound sources satisfying ISO 6926 for sound power measurement:
Because of the small size, the hemi-omnidirectionality, and the possibility for placement at any inclination, the Qmir and the Qref sources also allow sound power identification in more difficult sound fields than those covered in ISO 3747. For example; smaller rooms, nearby supporting equipment, nearby floor, low absorption in the environment, etc. In that case the reference source is placed against multiple surfaces of the machine and the calculation includes averaging over those multiple locations or power based inversion. Those applications are not possible with other, traditional, ventilator based power reference sound sources.
Vibro-acoustic transfer
On the example of an self-navigating vacuum cleaner, the noise from the electro-motor is subject to optimization. A soft purring sound is desired, no too tonal, not too sharp. Some of the noise is airborne, radiated from the motor. Given the light plastic construction, a significant part is radiation of structurally induced. Forces at the interfaces of the electro-motor, due to non-uniform rotation and fluctuating torque. To allow the manufacturer to use common electro-motors the interface structure and outward radiating shells can be optimized. The way to measure the effectiveness is vibro-acoustic transfer from the interfaces outward. It is a tricky task because there is hardly any space to accommodate shakers. Even with the Qhsh and Qlws, the smallest and lowest coupled mass shakers on the market the interfaces are not accessible. Partial disassembly or modifications to make space the vibro-acoustic transfer is significantly affected and no longer reliable to compare solutions. So what is the trick? Reciprocity. Structural excitation with sound pressure response is replaced by airborne excitation and structural response measurements. Even the tiniest accelerometers easily pick up the vibration response of the plastic support structure due to volume sound source excitation, resulting in well reproducing vibro-acoustic transfer functions. The video and photo show that the direct and reciprocal measurement produce the same vibro-acoustic transfer. The demo video uses a combination of:
All described measurements also require software and sensors, of which several are available in the market. Depending on the exact application our partners,
Head Acoustics, Siemens, Polytec
All propose leading full chain solution including the Qsources excitation sources and shakers
