The paper about the prediction ot the fouling prospensity of orange juice suspended particles through their physical characteristics presents some interest, as it is a less common topic with less common approach.
The introduction is well written
The methods are appropriate
The results sound meaningful but some issues to be explained
Some questions were arisen that need clarifications

page 23 a PLS prediction approach was used based on 16 physical variables.
The prediction was improved (lines 23-38) choosing 5 variables. However, it is not clear why this prediction was based on 5 variables and not on 6 or 7 and why these specific variables were chosen. Please explain more (page 21 lines 52-60 as well).
Moreover the predictability was not improved radicaly and according to Table 2 only three variables were correlated to SRF in a linear way. Are there any other of correlations? e.g. exponential, why the log (SRF) was used?
In that point and considering also the PCA figure why all of the variables were used. e.g. which is the meaning of using five similar variables for particle size, D4,3, d3,2 d,10, d50, d90?
In Table 1 the n' , n'' values are almost constant, could they be excluded from the analysis and the PLS model? Moreover, if some variables are related to each other could be further excluded? e.g. A', A'', are they correlated with wach other?