Used Spectrum from accelerogram

02-Analytical capabilities
Post Reply
NICOLA CHIEFFO
Posts: 2
Joined: 16 Jul 2018, 13:20

Used Spectrum from accelerogram

Post by NICOLA CHIEFFO » 16 Jul 2018, 13:34

Dear All,

I have a problem with the use of SEISMOARTIF. Pratically, I have a .txt file with real accelerogram recorder in Romania. I need to generate accelerogram in order to perform Time History Analysis. The units are second for the period and about acceleration is g/10.

I converted the units in order to obtained the acceleration expressed in "g" (compatible with the software) and I imported these values in another .txt file. Subsequently, in SeismoArtif 2016, I selected "Used Spectrum from accelerogram" and when I insert my .txt file (9 lines) expressed in "g", SeismoArtif generate a Spectrum, but the values seems esxpress in other units such as "ms-2", but I am not sure.

For example, for T= 0, Sa=3.21 g, for T= 0.08, Sa=9.34 g, it is impossible, because I imported the .txt file in "g" units.

I checked also the default units format of SeismoArtif, and they were in "g".

Can someone explain me how can I do in order to obtain a good results?

Thank in advance,

Nicola

User avatar
seismosoft
Posts: 565
Joined: 06 Jul 2007, 04:55

Re: Used Spectrum from accelerogram

Post by seismosoft » 16 Jul 2018, 17:05

In SeismoArtif the units are only needed for the transformation of acceleration to velocity and displacement.
The unit of the spectrum record are the units in which the accelerogram is created, it is as simple as that.
However, it seems that you are not loading correctly the spectrum. When you use "Spectrum from loaded accelerogram" and load a record of just 9 lines this is not correct (acceleration records are much more than that, 500, 1000, or more values). Can it be that you are loading a spectrum and you should probably need to use the 'Load spectrum from file' option?

Seismosoft Support

NICOLA CHIEFFO
Posts: 2
Joined: 16 Jul 2018, 13:20

Re: Used Spectrum from accelerogram

Post by NICOLA CHIEFFO » 17 Jul 2018, 11:33

Thank you for your kindly reply.

I will try to be more precise in the description of the problem. I recived a .txt file, in which there were a recorded accelerograms in all direcion. In that case, the units of the accelerograms were in g/10 and second.

I converted that units in "g" in order to import them in the software SeismoArtif. So, if I use "Load Spectrum from file" the software needs only two columns, one for the period and other one for the acceleration, but, as I wrote before, I have 9 columns with a lot of data for each columns and, in this case it is impossible to use "Load Spectrum from file".

For this reason, I use to generate the spectrum derived from the real accelerogram and then, I can obtain the accelerogram to use in excel file and after that in the software in order to perform time history.

As I explained in the previous message, I have the problem concerning the units it means I have a spectrum in the other format.


How can I solve the problem?

Thank you in advance

huffte
Posts: 827
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Used Spectrum from accelerogram

Post by huffte » 17 Jul 2018, 13:41

Nicola, perhaps you can use the original file. If I understand correctly, you have 9 columns, with three of those columns being accelerations is the three perpendicular directions. It is not clear regarding units. If the units are acceleration divided by 10 in g, then the appropriate scale factor to specify upon loading the accelerograms would be 10. If the units are "tenths of a g" then the scale factor would be 0.10.

1. When loading the accelerograms, change the input format to "Multiple Acceleration Values per line".
2. Determine which of the nine columns contain the acceleration data you wish to load. Call this column "n".
3. Specify "n-1" for the "number of initial values to skip".
4. Specify "9" for the frequency.

This will skip to the desired column initially, and skip 8 values upon each subsequent read, thus retaining only the data in column "n".

You may be well advised to try this in SeismoSignal first to make certain the data is being read correctly. In SeismoSignal you will be able to view the resulting acceleration, velocity, and displacement histories to ensure that a properly filtered and corrected accelerograms is being used.

This should work if I have correctly understood your situation there. If not, please clarify or send the file to thuff@tntech.edu if you wish.
Tim Huff

huffte
Posts: 827
Joined: 22 Jul 2011, 10:19
Location: Cookeville, Tennessee, USA
Contact:

Re: Used Spectrum from accelerogram

Post by huffte » 17 Jul 2018, 18:27

Nicola, the problem is not with SeismoArtif, but with the raw data you are using. If you will look at the actual values in the file 'Real-Accelerogram-g-10-Units.txt' you will find that the maximum is 32.835. Indeed, this is even provided in the 18th line of the original file. I can only conclude, thus, that the units cannot be g/10 as stated. Otherwise, the record would have a PGA of 3.2835g, which we know cannot be the case for the July 12, 1991 earthquake in Romania.

To rectify the dilemma, you will need to contact the individual who provided this record to get the correct units and accelerograms if this is the record you are bound to use.

From all appearances to me, SeismoArtif is doing exactly what you have asked it to do. The problem is in the raw data.
Tim Huff

Post Reply