New version page

Lecture notes

This preview shows page 1-2-22-23 out of 23 pages.

View Full Document
View Full Document

End of preview. Want to read all 23 pages?

Upload your study docs or become a GradeBuddy member to access this document.

View Full Document
Unformatted text preview:

176Chapter 5 Modifications to the VSE-355G2 —The VSE-355G35.1 IntroductionIn Chapter 3 the VSE-355G2 instrument was introduced, and the major problem with alow clipping level was exposed. In Chapter 4, this problem was observed by the sameinstrument in the field during the M8.3 Tokachi-Oki earthquake. This Chapter summarisesthe modifications made to the instrument by the manufacturer, and provides laboratory andstation data which indicates that the instrument is now performing to original specifications.During the modification and re-testing process, low clip levels were found in the verticalchannel once the horizontal problem seemed to have been corrected. The cart test was runin a elevator in Japan, as the set-up at Caltech did not easily facilitate this type of test.Cart test data now shows impressive strong motion performance from the VSE-355G3,with motions exceeding 200cm/s being recorded linearly by the instrument. The dynamicrange of the instrument has not been compromised by the modifications, and retains similarperformance to the VSE-355G2.The equivalent SDOF determined from the calibration testing indicates the instrumenthas a different response from the VSE-355G2, with a longer equivalent free period of about105s now observed. This is well in excess of the range specified by the manufacturer ofonly 80s. Damping is slightly lower at 65% of critical. There much better correlation withthe observed SDOF response and the expected theoretical Transfer Function response thanis shown for the VSE-355G2.The instrument is currently deployed within the CISN alongside a CMG-1T at CRP, theCaltech Robinson Pit.1775.2 ModificationsThe instrument has been modified internally once, and fully overhauled twice. Both over-hauls involved an engineer from Tokyo Sokushin visiting Caltech, and removing the exist-ing suspension and feedback for all 3 components, replacing them with new sensors andfeedback circuitry.The first internal modification occurred in 18 June 2002, after Sokushin had identifiedthe source of the low saturation velocities as a fault with the power regulator in the elec-tronic feedback system which prevented the final stage amplifier from operating correctly.This attempt to increase the clip level to the expected level of 200cm/s only succeeding inmoving the observed clip from 15cm/s to 40cm/s (see Figs 3.21 and 3.24).A full overhaul took place on 11-12 November 2002. This was again in response tothe low clipping levels. Cart tests subsequent to this showed the low clip level problemhad been resolved for the two horizontal components, but not for the vertical component(Figure A.3(b)). Also, there was significant cross-coupling between all three components(Figure A.3(a)).After investigation, Tokyo Sokushin reported that both problems were caused by a me-chanical problem — the suspension spring in the Z-axis being unable to adequately resistthe large velocities. This was fixed, as evidenced by elevator tests performed in Japan,where the vertical component was subjected to ∼ 1m/s velocities without any problems(Figure A.4).On 15-16 Sept 2003, this same modification was incorporated into the Caltech sensor,which again involved replacing all 3 sensors. Cart tests subsequently confirmed the cliplevel was acceptable for the horizontal components, recording linearly beyond 2m/s. With-out the ability to perform an elevator test at Caltech, the vertical component could not beexcited up to 2m/s. Velocities up to about 10cm/s were successfully recorded. Cross-axissensitivity seemed to have been eliminated.At this stage, with the apparent success of the instrument in strong motion, calibra-tion tests were performed. These tests produced large permanent offsets in raw velocityoutput from simple step-function excitations that should not result in permanent offsets178(Figures A.7 and A.8, the expected response shown in Figure 3.5).Once again the manufacturer visited Caltech, where it was discovered from carefulanalysis of the response of a similar VSE-355G3 instrument in Japan, that the errors wereproduced by incorrect pin connections between the sensor and the datalogger. When thiswas repaired, the instrument performed well in the calibration tests. At this stage, a fullseries of tests were repeated to confirm compliance with the sensor specifications. Anin-depth chronological list of site visits by Sokushin, including results, and important mile-stones in the testing regime, are presented in Appendix A.5.3 Test Analysis5.3.1 Instrument ClippingAfter all the repairs had been completed, by September 2003, the VSE-355G3 was ob-served to perform well in the cart test. A typical result is seen in Figure 5.1(a). The‘raw’ output only is plotted, with nominal station gains removed, and acceleration dataintegrated to velocity. All three channels, the in-plane motion in the direction of the cartdisplacement, the out-of-plane, and the vertical, have very similar timeseries for both sen-sors. Figure 5.1(b) presents the large in-plane motions only, in acceleration, velocity anddisplacement. Also included in this plot is the time-domain deconvolution solution for theVSE-355G3. The deconvolved motion maps the EpiSensor’s motions extremely well. Inthe velocity time-series, the maximum recorded deconvolved velocity is 211cm/s, whichexceeds specifications. After the motion has ended, the deconvolved VSE-355G3 velocityrecords a near-zero velocity, reflecting the true physical position of the sensor, and indicateslittle tilting has occurred. The similarities in the displacement timeseries are also obvious,and the lack of tilt in the VSE-355G3 is reflected by a very steady final displacement ofabout −16.4m being recorded. This is about the same distance traversed in the test (itwas not precisely measured). This Figure can be compared to the theoretical solution of asimilar cart test shown in Figure 3.10. This example has the least tilt of all the runs per-formed, in other tests, as seen in the subsequent plots, tilt seriously distorts both sensors’179translational response.In the suite of tests which included the above data, clipping was observed only wellbeyond the 2m/s level, near 2.5m/s. No cross coupling was observed, though some largeoffsets were observed in the ‘zero’ level for the horizontal channels. A plot of the entiresuite of tests is in Figure A.6, showing a timeseries which consists of numerous cart tests,forwards and backwards along the hallway, and is interspersed with


Loading Unlocking...
Login

Join to view Lecture notes and access 3M+ class-specific study document.

or
We will never post anything without your permission.
Don't have an account?
Sign Up

Join to view Lecture notes and access 3M+ class-specific study document.

or

By creating an account you agree to our Privacy Policy and Terms Of Use

Already a member?