I/O Error 103

03-Unexpected behaviour/errors
Post Reply
klntejaswi
Posts: 2
Joined: 17 Sep 2013, 22:27

I/O Error 103

Post by klntejaswi »

Hello,

I have been trying to save the matched and original time series records but I'm getting an error "I/O error 103". Please can you let me know how to save these records?

Thanks.
huffte
Posts: 978
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: I/O Error 103

Post by huffte »

Hi klntejaswi.

There are 2 workarounds I have found. I assume you are running Windows 7, in which there is an issue in removing Read-Only access to certain folders.

A. To save the accelerograms in your personal user folder, which is defaulted to be full access, not read only, simply change to that folder when you save the accelerograms. You may then move the files to any folder you choose. The personal user folder will be something similar to "C:\Users\YourName".

B. To save the accelerograms in just about any other folder of your choice, follow these steps once the matching process is complete:

1. Go to the "Time Series" tab.
2. Select the "Table" sub-tab.
3. Hover the cursor over the time series - original or matched - which you wish to save.
4. Right-click --> Select All
5. Right-click --> Copy Selection
6. Open Notepad or your favorite text editor.
7. Ctrl-V or Paste the clipboard into the blank Notepad file.
8. Save the file in the desired location with the desired name.

Obviously, Method A is more straightforward, but both work.

I have tried opening folders in Explorer and removing the Read-Only status, but it doesn't seem to work. Perhaps you know a way to remove Read-Only status of folders in Windows 7.

I hope this helps, klntejaswi. And best of luck with your work.
Tim Huff
klntejaswi
Posts: 2
Joined: 17 Sep 2013, 22:27

Re: I/O Error 103

Post by klntejaswi »

Hi huffte,

Thanks a lot for your advise! I can save the scaled time series data now. I would like to ask you to comment on the accuracy of this program with regards to scaling of time series records? I'm planning on using these records to test a bridge in SAP2000 and the scaling software provided by PEER in their website was a bit glitchy.
huffte
Posts: 978
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: I/O Error 103

Post by huffte »

In my experience, SeismoMatch has done an excellent job of matching accelerograms to a target spectrum. I used SeismoMatch extensively in my research on the isolation of bridges in the New Madrid Seismic Zone.

It is important to know the difference between the SeismoMatch process - spectral matching with wavelets - and that used as a tool for the PEER ground motion database - simple amplitude scaling.

SeismoMatch uses wavelets to actually alter the input time history. The procedure is not a simple scaling, as is used at PEER.

Modern codes generally permit both spectral matching with wavelets and amplitude scaling. It is important to note however, that if you are interested in estimating both the dispersion and the mean of structural response, amplitude scaling is generally accepted as the better option since spectral matching, by its very nature, reduces dispersion. If you are interested only in mean response, spectral matching should be fine.

It is possible to get an extremely tight fit over a wide range of periods using SeismoMatch. You can also obtain a reasonably tight fit and still preserve dispersion using amplitude scaling if the records are selected appropriately. If you decide to go the scaling route, a valuable reference for you would be SigmaSpectra, which permits one to match both spectral acceleration and variance by checking thousands of combinations in a short time.

Should you decide to go the spectral matching route, I recommend SeismoMatch. Note that if you are doing bi-directional analyses, you do not want to match both components to the design response spectrum. You should make an effort to preserve the relative spectral ratios at a key period, or periods, between the two components while keeping the geometric mean very close to the design spectral acceleration. If you are doing single component analysis, then just match the component to the design spectrum.

Finally, if you are studying near-field sites with pulse-type motions, you should carefully examine the unaltered records and the matched records to determine whether or not the pule-type character of the motion has been preserved through the spectral matching process.

I hope this helps klntejaswi. Best of luck.
Tim Huff
Post Reply

Return to “03-Unexpected behaviour/errors”