Almost all types of industrial machinery contain rotating parts subject to impacts, sliding, contact, and imbalance, which translate into excitation forces. These forces propagate throughout the machine structure, generating surface vibrations and consequently sound waves, which are always present and almost always detectable, being inherent to the machine’s operation. When properly processed, these signals can reveal important information about the health or wear condition of the machine’s internal components.
PEI VM Srl, specialized in consultancy and instrumentation in the experimental NVH (Noise, Vibrations, Harshness), has been conducting extensive experimental activities for years with the support of several customers, developing specific analysis modules for different machines (gearboxes, axles, transmissions, electric motors, combustion engines, pumps, etc.). These modules are currently available through dedicated software licenses within a diagnostic system called VMGears.
Based on custom-developed algorithms, VMGears identifies defects in internal components, which generate abnormal noise or vibrations.
Different types of defects from different components are displayed to the operator through a simple and intuitive interface, in the form of numerical indicators that are easy to compare, making it immediate to assess the machine’s health condition.
VMGears is successfully implemented by leading companies in the power transmission, automotive, motorcycle, off-highway, and industrial sectors. It is currently available in three distinct versions for different applications, one of which is VMGears DM (Durability Monitoring), dedicated to continuous monitoring of endurance tests.
The analysis presented here concerns the monitoring the wear condition of a gearbox by means of VMGears DM system during an endurance cycle that simulates the transmission’s entire lifecycle.
VMGears DM is configured to automatically perform several operations, including:
Synchronous acquisition of signals from accelerometers, electric motor tacho / encoder and other sensors (thermocouples, torquemeter, pressure sensors, etc..)
- Extraction of dedicated indicators from raw signals using the analysis algorithms available in the software modules
Thanks to VMGears specific indicators, each anomalous vibration is directly associated with a specific mechanical cause, for instance: vibrations generated by abnormal gear meshing, shaft unbalance, shafts misalignment, dented tooth, defects on individual bearing raceways.
VMGears DM allows the acquisition of other types of signals, in addition to vibrations: during this test, two thermocouples, positioned on the gearbox structure near bearings, were measured to monitor any abnormal temperature variation.
VMGears DM performs the self-learning of acceptability thresholds in function of initial conditions: these thresholds can then be adjusted by the operator manually or automatically during the test. By adding relay boards or digital output modules to VMGears DM hardware configuration, it’s possible to send a signal to the test bench to stop the test if one of the acceptability thresholds is exceeded.
1) TEST CONDITIONS
The tested gearbox consists of a single reduction stage with helical gears and four commercial bearings supporting input and output shafts.
Using the dedicated kinematic model construction tool, the transmission was modeled within VMGears into a simple tree structure, including a cascade sequence of shafts, bearings, and reduction stages.
Figure 1 – Gearbox test setup on endurance test bench.
The test cycle adopted by the customer involves steady-state speed and braking torque: in this case the software performs data acquisition and analysis at programmed time intervals, displaying the trend lines of vibration indicators.
Generally speaking VMGears DM allows to map steady or variable conditions.
2) INSTRUMENTATION
VMGears DM is a complete system including hardware and software.
The supply includes acquisition modules and IEPE sensors (IEPE = Integrated Electronic Piezoelectric Excitation). It’s possible to use monoaxial or triaxial accelerometers and optionally microphones (only if the environment is not noisy).
For this test, a high-frequency acquisition board (51.2 kHz) and two monoaxial accelerometers were used, mounted on the transmission housing using magnetic bases, in positions theoretically sensitive to the direction of the force exchanged between gears.
Figure 2 – VMGears system in portable shockproof case configuration, including laptop, acquisition boards, and sensors
3) RESULTS
The fixed-point endurance cycle was automatically monitored by the VMGears DM system until the parameter trend lines showed clear and irreversible increases compared to the values recorded during the initial self-learning phase.
The following growth trends were observed:
+100% increase in RMS overall (overall Root Mean Square vibration at the measurement point, m/s²). Despite the significant percentage increase, this indicator reveals a substantial change in gearbox health but does not identify the critical component, unlike other indices.
- +500% increase in TQV [Tooth Quality Vibration, m/s²]: this indicator represents the vibration related to gear meshing. This strong increase, together with the essentially constant trend of the RTD (Relative Tooth Defect), reveals that the damage is widespread across the entire gear, and not limited to one single tooth.
- Stability of RTD (Relative Tooth Defect), for both Pinion and Wheel: this indicator rises up in case of not uniform vibration among the teeth of the same gear. A nearly constant RTD for both gears, far from its threshold, indicates a generalized issue rather than a localized one, as confirmed by the widespread pitting found on the gears.
Figure 3 – Trend of RTD indicator for PINION (top) and GEAR (bottom). RTD relates to a localized defect on a single tooth: values remain nearly constant throughout the entire test.
Figure 5 – Gear teeth affected by widespread pitting, as found at the moment of +500% TQV index: teeth are visibly damaged but still far from breakage.
4) CONCLUSIONS
VMGears DM proved to be an accurate and easy-to-use tool for monitoring transmission’s health condition, allowing to stop the test before more severe damage occurred.
This information is strategically important to plan an inspection stop and avoid a severe and sudden failure, from which it would be impossible to find the root cause.
VMGears DM can be flexibly configured to monitor:
- Other types of sensors in addition to accelerometers, such as thermocouples, torque sensors, and oil contamination sensors
- Complex endurance tests involving multiple operating points of speed and load
- Complex kinematic chains (gearboxes with multiple stages, axles, electric motors, complete transmissions)
VMGears DM completes the VMGears “portfolio”, which includes:
VMGears RD (Research & Development): a portable tool for performing a vibro-acoustic fingerprint during development phase (prototype comparison, benchmarking respect to competitor products, evaluation of potential integration issues in systems/vehicles), as well as for on-site investigations.
VMGears QC (Quality Control): a tool for performing a product / assembly quality control check at the End-Of-Line.
Regardless of the application field, the analysis algorithms implemented in VMGears allow discrimination between defective/damaged components and healthy ones, which is a key factor when analyzing complex kinematics.
The results are displayed in a simple and intuitive way, making the system usable even, and especially, by non-NVH specialists.
