In Response spectrum analysis while defining spectrum we have to give multiplier fator, & genrally this would be 1 for X & Z while 0.667 for Y.

but in many files i have seen that they have taken 1000 & 667 inplace of 1 & 0.667, what is reason of that does this factor is dependent on unit file that we have used.

This is a straight multiplication factor - unitless. However, if the spectrum data is 1000 times too small, then you would need to use the factors you mentioned above.

sorry , but i didn't understood

as whether we have to use 1000, 667 or 1 .67

& in which cases it is used

pl explain in detail

Assuming your spectrum data is provided in "g's", or "length/sec^2", then your multipliers are 1.00 and 0.67.

You would have to have a reason to use something else.

& when to use 1000 & 667

It depends on the data, I simply can't answer this. I've never encountered a need to use anything other than 1.0 and 0.667 though.

you said that
"Assuming your spectrum data is provided in "g's", or "length/sec^2", then your multipliers are 1.00 and 0.67."

that mean can i consider if the spectrum data is in m/sec^2 ("length/sec^2")

then we have to use 1 & 0.667

& if spectrum data is in mm/sec^2 ("length/sec^2")
then we have to use 1000 & 667

is it currect

The data should match whatever your current units system is. So, if the spectrum isn't in that set of units, that is an instance where you would want to alter the multipliers.

The expected units are shown at the top of the dialog where you enter the data values.

"The expected units are shown at the top of the dialog where you enter the data values"

where

See this image.

thank you

Richard