|
 |
Ingersoll-Rand produces vibrating soil and asphalt compactors sold internationally. The DD-158HFX compactor drum generates a sound power level of 114.8 dB(A). Environmental laws in Europe have mandated that the overall sound power level of the new machines must not exceed 113 dB(A). PSU Acoustics has been given the task to construct the scaled prototype of the DD-158HFX drum and run a baseline vibration analysis. Manufacturing of the prototype drums was performed by the Penn State Engineering Machine Shop and was assembled using Belzona 1121 Super XL-Metal. The two new scaled models are 1/7 the size of the full-size compactor drums in every dimension. The frequency scaling is proven to be inversely proportional to the scale for size. Testing was done using SigLab vibration network analysis. The models were suspended via a bungee through its center to ensure free-free vibration conditions. Three measurements were taken at each point of the grid using an uni-axial accelerometer and an impact hammer. By performing sound pressure tests, it was determined that the mass of the accelerometer has a negligible effect on the mass and stiffness of the drum. Upon completion of the testing and analysis of our data, PSU Acoustics used Star Modal to identify the natural frequencies and mode shapes. It was found that at least 16 natural frequencies lie below 5000 Hz for the scaled models. The majority of the vibratory response occurred in the overhanging shell. With this knowledge of the mode shapes, it will be possible to develop and test ways to reduce the sound power of the drums by working with cheaper and more easily manipulated scaled models.
PSU Acoustics is confident any modification applied to the model will scale to the full size drums.
|
The new standard for sound power levels of vibrating compactors contained in the EU's directive of 113 dB(A) must be met for Ingersoll-Rand's compactors. Ingersoll-Rand needs to modify the acoustic behavior of the DD-158HFX drums. PSU Acoustics needs to determine mode shapes and natural frequencies of the DD-158HFX drum.
PSU Acoustics' objectives include the manufacturing of two scaled models of the DD-158HFX drum that are a manageable size and have a constant scale factor. The lowest ten natural frequencies of the DD-158HFX drum will be determined and animated mode shapes will be produced with and without the drive lugs.
The relationship of the scaled model to the full scaled model was determined by two different methods. First, the Buckingham-Pi theory was used to show that the frequency of the scaled model was inversely related to the frequency of the full sized model by the geometric scale factor. The second method used to determine the frequency scale factor was Finite Element Analysis. Solid Works and Cosmos/Works were used to make a model that was analyzed with no boundary conditions to obtain free-free vibration. Convergence of the results was checked by running the model multiple times. The final and smallest mesh size used was 10mm and a picture of the meshed model is shown below. All of the results agreed with the chosen scale factor.
 |
 |
 |
- Two one-seventh scale, steel drums were manufactured.
- One drum had drive lugs attached and the other did not.
- Parts were assembled using Belzona Super XL-Metal epoxy.
PSU Acoustics used the following setup to perform all of the vibration testing:
- Uni-axial accelerometer measuring in the radial direction.
- 12 equally spaced circumferential points and 8 equally spaced longitudinal points for a total of 96 points.
- Roving accelerometer/stationary excitation method.
- Test stand designed to simulate free-free boundary conditions.
- Data acquired using SigLab.
After obtaining the data, Star Modal was used to identify the mode shapes and natural frequencies. A 3D model was constructed in this software and then the transfer function data was applied to it. After Star Modal applied a best fit function, PSU Acoustics was able to determine at least 16 individual mode shapes and natural frequencies in each of the two models. By examining these mode shapes, it is possible to identify the frequencies and regions of the drum that are most critical in sound production. The table below shows the obtained results.
Table 5 – Experimental Natural Frequencies
Without Drive Lugs
With Drive Lugs
Mode #
Freq. (Hz)
Full Size Freq. (Hz)
Nodes
Freq. (Hz)
Full Size Freq. (Hz)
Nodes
1
1720
245.7
4
1700
242.9
3
2
1740
248.6
4
1720
245.7
4
3
1830
261.4
3
1790
255.7
3
4
1850
264.3
3
1830
261.4
3
5
2190
312.9
5
1880
268.6
4
6
2250
321.4
5
2190
312.9
5
7
2490
355.7
2
2240
320.0
5
8
2780
397.1
4/4
2470
352.9
2
9
2850
407.1
4/4
2820
402.9
4
10
2960
422.9
5
2950
421.4
4
Sample mode shapes comparisons can be seen below.
 |
 |
 |
 |
 |
 |
PSU Acoustics feels that the scale models that were constructed provide an accurate vibratory model for their full scale counterparts. Ingersoll Rand can now test methods of sound reduction faster and at a lower cost. These models are much smaller, easier to handle and modify, and much cheaper to manufacture. Any stiffening additions made to the models will scale accurately when applied to the full drum.
The following is a list of suggested methods PSU Acoustics feels are a viable solution to attenuate the vibration of the DD-158HFX:
- Layer of visco-elastic material glued to the inside of the overhang.
- Would damp the vibrations produced by the drum, reducing the noise transmitted to the air.
- Would be the least expensive and easiest to apply.
- Fill the space between the bulkheads with a damping material.
- Material should be lightweight, easy to install, and have the ability to absorb sound.
- Additional bulkhead between the existing two.
- Would provide added stiffness to the center region of the drum and raise the natural frequencies.
- Would only attenuate the vibration in the center region.
- Integrated drive lugs/stiffening ring. An annular bracket attached along the entire circumference of the overhang and to the bulkhead.
- This option would combine the stiffening of the brackets and the circumferential ring.
