No. Because of low accuracy.

Yes, but you need to get numbers for it's accuracy and precision

Totally agree with Mr. Vikas R., and in addition you should check whether it (accuracy and precision) will sufficient with your experimental work.

I think, yes, with regard to accuracy as mentioned by Mr. Vikas

Have a look at this interesting article of how an iPhone (hardware) and Mathematica (software) were combined for a vibration analysis.

http://blog.wolfram.com/2017/03/02/how-to-use-your-smartphone-for-vibration-analysis-part-1-the-wolfram-language/

So the tentative answer is "Yes, an iPhone can be used for some vibration experimental work but caution should be applied due to the uncertainty in its accuracy levels in the absence of further verified metrological (measurement) information".

The tricky part would be in quantifying its experimental accuracy levels for a selection of vibration problems as first mentioned by Vikas Ramachandra.

There could be two possible ways in which the accuracy level could be quantified. The first would be the traditional "forward" or direct problem where the experimental results of a calibrated vibration sensor (say an accelerometer) with a known uncertainty (accuracy) are compared with the experimental results from the iPhone with an unknown uncertainty (accuracy). Then the results are compared for the En (normalized error values) and tentative estimates for the iPhone acceleration measurements are calculated in order for the results to be experimentally statistically consistent. This is physically more difficult due to the need for some vibration equipment/instrument standards but mathematically easier.

The other second approach would be the "reverse" or inverse problem approach where the iPhone results are analyzed "backwards" to try and figure out what the likely experimental accuracy specifications are but this can be more mathematically tricky.

Dear all thanks 

Then how we can refer to it on our papers ?can be acceptabel by reviewrs?

No. Because of low accuracy.