Reassessment of Thin-Layer Drying Models for Foods: A Critical Short Communication
Round 1
Reviewer 1 Report
It is unclear what is the novelty of the work. The analyzed mathematical models are very simple. More complex dynamics should be analyzed before reaching a conclusion. The quality of the figures should be improved. Most references are very old, newer ones should be included and discussed in order to justify the relevance of the study.
Author Response
I would like to thank Reviewer 1 for his/her suggestions and comments.
- The novelty of the work is explained in lines 53-59: There are common mistakes of using thin-layer models and these may be observed in almost all studies. Hence, the objective (and the novelty) is to show these mistakes and avoid to arbitrary usage of thin-layer models for drying data. Furthermore, this study provides a list of eliminated thin layer modes and a list of recommended models to help users make the right decision when selecting a model for their food drying process. It also aims to reduce the time spent on primary models and therefore, pay more attention on secondary models (dependence of the primary model parameters on environmental conditions such as temperature, air velocity and relative humidity) (see line 305 – conclusion).
- I agree that most of the models are simple but this is the nature of thin-layer modeling i.e., drying data are fitted by simple mathematical models and parameters are estimated through regression (linear and most of the time non-linear regression). Although the models are simple, many mistakes are still being made while using or comparing the models and this is the main reason of writing a such review. The reason of using simple (empirical) equations for thin-layer modeling is explained in line 35.
-All figures have the resolution of 600 DPI. The quality of figures may seem bad due to conversion of Word file into a PDF file.
- Since this article tends to review the application of thin-layer drying models, it is not surprising that some of the references are old (years are as such: 1921, 1949, 1961, 1968, 1973, 1974). New articles were added: In its present form manuscript has 51 references and more than a quarter of these references (14 out of 51) are from the last 5 years (2016-2021).
Reviewer 2 Report
The manuscript presents a review of the different mathematical models used in thin-layer drying of foods. The author gives some suggestions in order to choose the more suitable model to fit the obtained data.
Despite the possible interest the suggestions given in the review could have, one of the statements should be justified in a proper way, that is, the use of p-value, leading to the insignificance of the model parameters.
Moreover, some of the conclusions are too naive, others are too simple. What valuable information can this review provide for the research community?
Some comments to improve the manuscript:
- The use of p-value to infer some conclusions should be justified.
- It would be interesting to include how to determine or compute the different parameters used in the models.
- Please, do not use acronyms.
- English review of the manuscript is mandatory (punctuation marks, above all).
Author Response
I would like to thank Reviewer 1 for his/her suggestions and comments.
I agree that conclusions are simple [this was explained in line 59 – one of the objectives of this study give “some beneficial but simple suggestions” which would improve the analysis and gain time to researchers in the field. This would, in turn, reduce the time spent on primary models and therefore, pay more attention on secondary models (dependence of the primary model parameters on environmental conditions such as temperature, air velocity and relative humidity) (see line 307 – conclusion)]. Nevertheless, I have modified the conclusions a bit – please see lines 287-290, 293-294, 300 and 302.
The use and definition of p value are given in lines 174-177.
The procedure to obtain the parameter estimates of the models are explained in lines 220-224.
English of the manuscript has been checked and necessary changes were made throughout the manuscript.
Reviewer 3 Report
This paper is more a critical mini-review or short communication than a review. I recommend changing the title.
The paper focuses on the use of complex thin layer drying models with more than two parameters. The study provides a list of eliminated thin layer modes and a list of recommended models to help users make the right decision when selecting a model for their food drying process. Hower, in my opinion, all so called thin layer models are nothing else than simple fitting of data which is now widely available in commercial software such as MS Excel. These models are not sensible to changes of experimental conditions and for each experimental run another set of parameters are required. Dependence of model parameters on experimental conditions or models sensible to experimental conditions were not the subject of the presented study. This is the reason why the importance of the work is not high.
References in the text are not in the same order as in the list of references
In Figure1 and Figure 6, you use time (arbitrary scale), but the value of parameter delta depends on the time unit. This issue should be better described in the manuscript.
Author Response
I would like to thank Reviewer 3 for his/her suggestions and comments.
The title has been modified as “Reassessment of thin-layer drying models for foods: A critical mini-review”
That is true that the models are simple; however, this is not surprising because thin-layer modeling is based on the use of the equations presented in Tables 5 and 6 in the manuscript – please also see line 35. The aim is to describe the drying data with a simple mathematical equation or equations. However, the use of Excel to describe the drying data may not be that simple and may require the use of “SOLVER Add-In” of Excel. Suppose that any drying data can be described by MR = a∙exp(-k∙t) [Henderson-Pabis model in the manuscript] where MR is the moisture ratio (dependent variable), t is the time (independent variable), a and k are the parameters of the model. This model is available in the Excel library as the exponential model; however, if this equation is fitted to the exponential data, Excel will linearize the equation i.e., lnMR=a-k∙t and then, obtain the parameters by using linear regression (not non-linear regression!). Then, by using these parameter estimates the model is fitted as if these are the results of non-linear regression. (Unfortunately, this is an unknown property of Excel and missed by most of the Excel users). One can find the correct parameter estimates by using SOLVER Add-In of Excel and non-linear form of the equation. These two methods may produce close but different parameter estimates (a and k). This is just an example of the use of Excel for thin-layer drying models. Many models given in Tables 5 and 6 in the manuscript require non-linear regression and hence, SOLVER Add-In tool in Excel or another software such as SigmaPlot, Origin etc. should be used (This is, of course, less straightforward than using linear regression.) – please also see the discussion on transformation of the models (lines 123-155).
This statement is confusing because it is the nature of the models that parameter values are changing for each experimental run and this not specific to thin-layer models. However, it may be possible to fix one or two parameters in a model or it may be possible to use a fixed (average) value for the parameters, so that parameter(s) is/are only a weak function of temperature (or any other environmental condition such as air velocity and relative humidity) or not dependent on temperature. One of the objectives of this review is to reduce the time spent on primary models and therefore, pay more attention on secondary models (dependence of the primary model parameters on environmental conditions such as temperature) (see line 305 – conclusion). Too much attention has been paid to thin-layer modeling (primary modeling) with unnecessary details such as comparing the models just by looking at R2 and the use of the models with insignificant parameters.
All references, including the new ones, are arranged according to the Journal’s format.
Reviewer 3 has a point. Explanations were added – please see line 266 and 274.
Round 2
Reviewer 1 Report
The paper can now be accepted in its present form.
Author Response
I would like to thank Reviewer 1 for his insightful comments and suggestions.
Reviewer 2 Report
Although the author answered all the questions in my review, the manuscript does not provide any significant research results for the food processing community.
Therefore, the paper is not suitable to be published in Processes.
Author Response
I would like to thank Reviewer 2 for his/her contributions and as the Reviewer 2 has explained all necessary changes have been made. I believe that the manuscript in its latest version will contribute the researchers in the field.
