Experimental design application is supported on some principles.

according to Kuehl, Experimental design, those are

* Local Control

* Randomization and

* REPLICATION

if you have no replicates, you could not made some statistical calculations on effects, like ANOVA, Signal/Noise, etc.

When there are no replications, it is not possible to obtain a reliable estimation of error. (According to Box, Statistics for Experimenters, section 10.8, available at https://bit.ly/2IKB2m8 ) but still you can make some calculations.

From NIST, Engineering Statistics Handbook, Section 5.3.3.3.2) ( Available at https://bit.ly/2Iom8DH)

* Replication provides information on variability

* Running the entire design more than once makes for easier data analysis because, for each run (i.e., `corner of the design box') we obtain an average value of the response as well as some idea about the dispersion (variability, consistency) of the response at that setting.

According to Ross (Taguchi techniques for quality engineering, ISBN 0-07-053866-2, https://amzn.to/2Vs7hOV) , for SIgnal to noise analysis, at least two replications are required.

You may obtain more information from any good text on DoE as the afore mentioned.

Regards

i found a interesting article for you.

Article 10 simple rules for designing and analyzing field experiment...

Refer

https://www.amazon.in/Applied-Design-Experiments-Taguchi-Methods-ebook/dp/B00K7YKDIS/ref=sr_1_1?keywords=applied+design+of+experiments&qid=1567448724&s=digital-text&sr=1-1

best book to clear your doubts

The suggested

Article 10 simple rules for designing and analyzing field experiment...

IS ONLY waffling !!!!!!!!!!!!!!!!!!!!!!!!!

Article Optimization of the hydrodynamic performance of a vertical A...

Mohammad Hassan Khanjanpour the article Optimization of the hydrodynamic performance of a vertical A... is NOT DOWNLOADABLE!!!! Send it to me and I will read it carefully...

• I wrote the same thing in another thread