Costruction of FRF Matrix

I want to costruct a FRF [H]column matrix of size 3x1 from impact test i.e [H11 H21 H31]^T.On a cantilever,I have taken 4 points 1,2,3&4.First I mounted an accelerometer on 1& impacted at 4 with an impact hammer.I noted A1(t)&F1(t).Next I mounted Accelerometer at 2 & again imacted at 4 only.Noted A2(t)& F2(t).Lastly with acclorometer at 3,impacted again at 4.Noted A3(t)&F3(t).Now can we take H11=A1/F1;H21 =A2/F2 and H31=A3/F3 in frequency domain after taking FFT of each matrix?Is it that we have to mount 3 accelrometers simulteneously at 1,2 and 3, and impacting once only at 4? My problem is after finding this[H] matrix,I am going to mount accelerometer at some other point,say 5 and impact at 4.Using the response at 5,A5,I want to reconstruct foece F5 at point 4,assuming I dont know this.Later I will compare with recorded history. I want to find this using: F5(jw) = [H(jw)]^-1*A5(jw)at every frequency.Pl let me know is this possible? SLG

[Shrinivas Gombi]

I am surprised ,since one week I have put this question and waiting some expert to help me.Pl PL some body help me my next work depends on clearing this doubt.Positively waiting

[Rick Rosson]

Please provide the following information clearly and concisely:

1. What are you trying to do?

2. What have you tried so far?

3. What part of what you have tried is not working?

4. What specifically is your question?

I do not think you are likely to get much help unless you provide these four pieces of information as clearly and simply as you can.

Thanks!

[Walter Roberson]

Reformatting your question to be more readable would help.