COMPARISON OF FRF DYNAMIC FLEXIBILITY BASED ON INTEGRATION RESULTS IN TIME DOMAIN AND FREQUENCY DOMAIN
FRF (Frequency Response Function) dynamic flexibility can be used to determine the cause of engine problems. However, FRF testing is generally carried out using a velocity transducer sensor, so the results of the FRF test are FRF mobility. To get FRF dynamic flexibility, integration process is neede...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Subjects: | |
| Online Access: | https://digilib.itb.ac.id/gdl/view/70372 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | FRF (Frequency Response Function) dynamic flexibility can be used to determine the cause of engine problems. However, FRF testing is generally carried out using a velocity transducer sensor, so the results of the FRF test are FRF mobility. To get FRF dynamic flexibility, integration process is needed. In numerical integration there are two approaches that are commonly used, namely integration process in time domain and frequency domain. Therefore, this study aims to compare FRF dynamic flexibility integration results in time domain and frequency domain.
This research begins by investigating errors in integrated signal due to lack of data in one signal period. This research was continued with numerical simulation of one and two degrees of freedom vibration systems to obtain the response of drift and velocity. The parameters used in numerical simulations have been selected to avoid large errors due to lack of data in one signal period. This research is continued by performing integration in time domain and frequency domain on numerical simulation results of speed signal. FRF dynamic flexibility resulting from numerical integration is then compared to analytical FRF to obtain an error graph of the integration results. This integration process is also carried out for cases where the speed signal is polluted by noise.
Results of integration in time domain and frequency domain are then compared. In this study, results of comparison show that the integration results error in frequency domain and the integration results error in time domain with the Simpsons 3/8 method are smaller than the integration results errors in time domain with the Trapezoidal method. The results also show that addition of noise causes large errors at low frequencies. That error value at low frequency can be reduced by using the integration function with the limit frequency. |
|---|