Vibration analysis with ViRA

[PitchFella]

Hi folks,
I would like to present you my module for vibration analysis today.
Since I had to spend some flight hours searching for the source of vibrations, I stepped deeper into it, finding it increasingly amazing
what software could read out of the vibrations. So I sat down and tinkered a bit and programmed a lot.

I think that needs a good working title I've decided to name it ViRA: Vibration Reporting & Analysis

Design of the module: With an acceleration sensor and a microcontroller vibrations from Heli are recorded and analysed.
The recorded vibration signal is converted by a Fourier-transform and visualized by the software.

As a small example, I want to show some diagrams of the analysis of my 450s. In flight it showed an increasingly poor performance.
The FBL system was probably already affected by vibrations. The last maintenance was back more than 300 flights and the need for some work was more than obvious.

The analysis I started by disassembling the main rotor blades since I already had a suspicion where the problem come from.
Shortly the diagram:

- The x-axis represents the rpm range of the analysis
- The y-axis is the magnitude of the vibration, I selected unit [Vir], the higher the stronger the vibration
- 1Mvir = 1000kvir = 1000000vir
- Shown in green is the vibration signal over frequency/rpm
- The strongest vibration is top left output (peak) in diagram header
- Right above a small Diagram comment is inserted each

Note: this was an pre-release version of the software so the unit [Vir] was used to indicate strength.
Release version using [mg] as the correct unit of acceleration/vibration.


Now the diagrams of the test run:





Test Run 1 - Clearly visible vibration in the Tailrotor (TR) range. TR rpm due to test run without major load (main rotor blades removed). In the area of motor rpm (about 43000rpm) a clear peak is present.
The vibrations are, by the way, higher than my Nitro at about 1.2MVir. Unacceptable for such a small heli, isn’t it ?



Test Run 2 - Without mass on the tail shaft ( blade holder with hub removed), the vibrations are considerably lower. Even the peak in the area of engine rpm (about 40000rpm) has decreased significantly.




Test Run 3 – after replacing shaft and the two bearings.

After the rear shaft was replaced and the two ball bearings replaced in the rear housing, with Bladeholder and Hub refitted, only a very slight increase of Vibes. => Source of the vibes eliminated !


Test Run 4 - The tail blades from the first test run were remounted. The vibration at the tail
1470000r (1.47Mvir) dropped to 936000 (936kvir). The motor peak is clearly seen here (about 39000rpm), probably due to the slightly reduced rpm as compared to 1st Test Run.
A slight variation in the course of the measurement (+/- 30kvir) may also occur.



Test run 5 - The Gorilla Tailblades were interchanged in the tail blade grips. With an unchanged rpm to test run 4, although the two peaks now considerably lower. Reduction of vibration in the area at the rear of 936kvir on 578kvir.
The vibrations are lower than for test run 3 without blades! This can be explained by the rear unit (shaft, hub, blade grips) tolerances in terms of mass and geometry subject. In other Words : from Vibes point of view, one blade may fit better to only one of the two Blade holders, generating less vibes there.


Test Run 6 - Test manner: other blades were mounted but have significant vibration. Also striking is the broad spectrum, i.e., the bandwidth of the two peaks at the rear and the engine rpm. Also, the vibration level is again increased significantly to 1.09Mvir.



Test Run 7 - After the blades were finely balanced both by weight and center of gravity
(0.002g, thanks again GARTH) is already a lot better. The vibrations are even lower than for test run 3 without (!) tail blades.



Test Run 8 - The exchange of fitted blades to each other bring a small reduction of vibration to 130kVir. The vibration level is now in a more than acceptable range.


Conclusion:
After replacing the tail shaft and the worn-out bearing vibrations are significantly reduced.
Another improvement is the exchange of the tail blades in the blade holder.
Since I have no way of balancing the rear unit effort, I am content with this approach.
The 450s now flying as it should again

Later I will post some pictures of displayed data via GRAUPNER-SJ HoTT Telemetry and also how to use the various axes of the accelerometer to locate and limit the vibration causes.

Further information in english: http://www.pean-engineering.com/vira_en/


Have fun,
Andi