Performance of Exterior Wood Coatings in Temperate Climates

Round 1

Reviewer 1 Report

This manuscript compares the performance of 12 wood coatings for exterior application, aiming to provide a guide for users to specify coatings for protecting the wood used in mass timber structures in a Mediterranean climate (Western Oregon). The performance of the selected exterior coatings was evaluated based the measurements of CIE Lab color parameters (L*, a* or b*) made on non-weathered and naturally weathered surfaces of several wood species. Using cluster analysis, the variation in color components measured on coated wood during 6-month exposure intervals up to 18 months was described. It was found that the selected coatings failed to protect the wood from biological and UV exposure over 18 months, although their performance varied during the exposure period and on different wood species.

Certainly the idea of evaluation and classification of commercial coatings based on their performance in exterior applications are of considerable interest to the users, as well as manufacturers for obvious reasons. This work, however, could not reach conclusive results and draw firm conclusions in this regard, hence fails to provide the users substantially new information on coatings performance, at which it was aimed.

Unfortunately, the experimental design is one-sided and excludes well-stablished additional/essential experiments for evaluating the outdoor performance of coatings.

The assessment of coating performance solely by analyzing variables obtained by color measurements is not fully reliable due to the measurement errors and inherent variation in wood surfaces. In addition, there is a lack of control samples to detect the color changes in uncoated wood as a reference. Moreover, using only k-means clustering for statistical analysis is not robust enough to make clear distinctions.

I cannot underline any significant contribution of this work to the existing knowledge; therefore, I regret to say that I cannot recommend the current paper for publication.

Author Response

Please see the attachment

Author Response File: Author Response.pdf

Reviewer 2 Report

The subject of the study is interesting and important because it is noted globally the increasingly frequent use of wood with exterior exposure in construction. The study has a number of shortcomings. The protection quality of the different coating systems was characterized by a single technique, namely the analysis of color changes. Discussions and conclusions also do not include comparative information on the performance of water-based and solvent-based coating systems. The conclusions of the study are valid only for a limited area in terms of climatic conditions.

Author Response

The subject of the study is interesting and important because it is noted globally the increasingly frequent use of wood with exterior exposure in construction. The study has a number of shortcomings. The protection quality of the different coating systems was characterized by a single technique, namely the analysis of color changes. Discussions and conclusions also do not include comparative information on the performance of water-based and solvent-based coating systems. The conclusions of the study are valid only for a limited area in terms of climatic conditions.

The technique addressed the central question of color change of coated wood surfaces exposed to natural weathering conditions. The technique used here has been used in other studies that are concerned with color change due to weathering. See section 2.3 for details.

 The experiment took place in a Mediterranean temperate climate in western Oregon. Similar climates exist in a significant proportion of North America and Europe. Therefore, the results could apply to a far greater geographical area.

Additionally, the results did not indicate striking differences between waterbourne and solvent based formulations for most combinations of different species/coatings, therefore, it was not addressed in the discussion.

Reviewer 3 Report

Review of the article: Performance of wood coatings in Western Oregon for exterior

   The topic of this research and output might be useful for the Coatings readers. The conducted work is worth to publish, however in order to improve the manuscript the following major suggestions should be considered.

 1.Title: I suggest to write: Performance of exterior wood coatings in Western Oregon 

2.Keywords: Please add eg.: aging or weathering; UV radiation. A larger number of keywords will make the article search easier. 

Abstract:

3.Line 20: Formulation of what? Please detail; coatings or wood? 

Introduction

4.Line 39: Should be hemicelluloses instead of hemicellulose. This is the matrix polysaccharides. 

5.Line 40: There are many publications on wood surfaces colour changes under UV radiation. Please provide some literature from other countries e.g:§  Laskowska A., Dobrowolska E., Boruszewski P. 2016: The impact of ultraviolet radiation on the colour and wettability of wood used for facades. Drewno 59(197): 99–111

• Pastore T.C.M., Santos K.O., Rubim J.C. 2004: A spectrocolorimetric study on the effect of ultraviolet irradiation of four tropical hardwoods. Bioresour. Technol. 93(1): 37–42
• Tolvaj L., Mitsui K. 2010: Correlation between hue angle and lightness of light irradiated wood. Polym Degrad Stab 95: 638–642

or influences of extractives:

• Chang T.C., Chang H.T., Wu C.L., Chang S.T. 2010: Influences of extractives on the photodegradation of wood. Polym Degrad Stab 95: 516–521
• Pandey K.K. 2005: A note on the influence of extractives on the photo-discoloration and photo-degradation of wood. Polym Degrad Stab 87(2): 375–379

6.Line 50: What "physical degradation" means? Maybe should be ...lead to lowering physical properties....?

7.Line 84: Please add L. ie. Pinus sylvestris L. 

2.8. Materials and MethodsThere is no basic information about the research material eg.: density, sections (radial, tangential).

9.Table 1: Do you have manufacturers' permission to use chemicals (listed in Table 1) and publish the results? 

10.Line 175: Why no attempt was made to evaluate differences between species? Maybe you should compare the coatings groups between species? 

11.Figures 4 and 5: It seems to me that Figures 4 and 5 can be smaller and can be put together next to each other. 

12.Figure 5: Cluster analysis of surface color parameters on groups of coatings applied to Douglas-fir test specimens only. Why are there no cluster analysis for other species? 

13.Table 2. It seems to me that you should give the ranges of ΔE and the groups of chemicals (coatings) that are in them. You wrote that the differences between them were sometimes slight. The reader or buyer should decide for themselves which one to choose. 

There are no references to literature data in the section Discussion. I am aware that this is an original and pioneering research. But there are no literary references. See literature in line 100.

Comments for author File: Comments.pdf

Author Response

Reviewer 3.

The topic of this research and output might be useful for the Coatings readers. The conducted work is worth to publish, however in order to improve the manuscript the following major suggestions should be considered.

1. Title: I suggest to write: Performance of exterior wood coatings in Western Oregon 

 Replaced with “Performance of exterior wood coatings in Western Oregon”

 2. Keywords: Please add eg.: aging or weathering; UV radiation. A larger number of keywords will make the article search easier. 

 Added several. Line 24 of the revised manuscript is now: wood coatings; exterior exposure; mass timber coatings; natural weathering; UV degradation; discoloration

 Abstract:

3. Line 20: Formulation of what? Please detail; coatings or wood? 

 Added “coating” before formulation. Line 20 is now, “Cluster analysis showed that coating performance varied widely with coating formulation and wood species.”

Introduction

4. Line 39: Should be hemicelluloses instead of hemicellulose. This is the matrix polysaccharides. 

 Line 45 now reads, “On uncoated surfaces or for surfaces where the coating has failed, the process of surface degradation generally begins when UV initiates photochemical reactions that decompose lignin, hemicelluloses [2-5], and extractives [6, 7] and which are then leached from the wood by water.”

 5. Line 40: There are many publications on wood surfaces colour changes under UV radiation. Please provide some literature from other countries e.g:§  Laskowska A., Dobrowolska E., Boruszewski P. 2016: The impact of ultraviolet radiation on the colour and wettability of wood used for facades. Drewno 59(197): 99–111

• Pastore T.C.M., Santos K.O., Rubim J.C. 2004: A spectrocolorimetric study on the effect of ultraviolet irradiation of four tropical hardwoods. Bioresour. Technol. 93(1): 37–42
• Tolvaj L., Mitsui K. 2010: Correlation between hue angle and lightness of light irradiated wood. Polym Degrad Stab 95: 638–642

or influences of extractives:

• Chang T.C., Chang H.T., Wu C.L., Chang S.T. 2010: Influences of extractives on the photodegradation of wood. Polym Degrad Stab 95: 516–521
• Pandey K.K. 2005: A note on the influence of extractives on the photo-discoloration and photo-degradation of wood. Polym Degrad Stab 87(2): 375–379

We thank the reviewer for these suggestions.  All of these suggested references were added in the appropriate places.

 6. Line 50: What "physical degradation" means? Maybe should be ...lead to lowering physical properties....?

Addressed to indicate that coatings are used to reduce the degradation of the visual attributes. Line 50 now reads, “Although surface degradation rarely results in significant losses in structural capacity, the unsightly surface appearance often requires refinishing the wood surface or replacement of the material.”

 7. Line 84: Please add L. ie. Pinus sylvestris L. 

 This was added.  This is now line 104 which reads, “For example, van Meel et al., [11] found that a waterborne non-film-forming alkyd stain reduced water uptake on Norway spruce but had little or no effect on Scots pine (Pinus sylvestris L.) or meranti (Shorea sp.).”

2. 8. Materials and MethodsThere is no basic information about the research material eg.: density, sections (radial, tangential).

 Added information about grain and density- lines 191-194 state: “The samples were representative of commercially available material and were primary flat-sawn, although the Douglas-fir samples contained nearly equal proportions of flat and vertical grain. The densities of the sample material were within the published ranges for the species [26]. The average density of the acetylated wood was 0.508 g/cm3.”

 9. Table 1: Do you have manufacturers' permission to use chemicals (listed in Table 1) and publish the results? 

 The policy at Oregon State University is that if the coatings are commercially available the material can be used for research that can be publicly published. We simply added that the material was publicly available-see line 197 – “All the coatings were publicly and commercially available at the time the study was initiated.”

10. Line 175: Why no attempt was made to evaluate differences between species? Maybe you should compare the coatings groups between species? 

 Such a comparison would not address the fundamental question(s) asked by users, i.e., “what coating(s) should I use for this species?” In most cases, the species is given, therefore there is only a choice of coatings involved. The comparison suggested above would, however, provide comparisons of coating performance on different species.    

 11. Figures 4 and 5: It seems to me that Figures 4 and 5 can be smaller and can be put together next to each other. 

 In the interest of consistency, the figures were kept separate here. Figures similar to Figure 5 are presented for the other species in the paper.

 12. Figure 5: Cluster analysis of surface color parameters on groups of coatings applied to Douglas-fir test specimens only. Why are there no cluster analysis for other species? 

 There are analyses for other species. The difference between Douglas-fir and the others is that additional clustering is presented for Douglas-for to compare color changes between intervals. It indicates that some coatings change a lot at first, then not much or that some coatings continue to change throughout the experiment. As Douglas-fir is probably the most important softwood species to be used for exterior applications in the study region, it was thought that this additional information may be interesting to some users. Further, the authors felt that providing the same level of analysis for all of the species would result in an excessively long manuscript.

 13. Table 2. It seems to me that you should give the ranges of ΔE and the groups of chemicals (coatings) that are in them. You wrote that the differences between them were sometimes slight. The reader or buyer should decide for themselves which one to choose. 

 Addressed. All ΔE values are presented in a separate section. Table 2, line 300 was added:

Table 2. Color change (∆E) over 18 months for each of the species and the corresponding coating used in this study. The lowest and highest values for each species are indicated in bold

14. There are no references to literature data in the section Discussion. I am aware that this is an original and pioneering research. But there are no literary references. See literature in line 100.

 Addressed. The discussion now has several comparisons between other studies and this study.

Round 2

Reviewer 1 Report

The revised version did not address the two main flaws:

1. Statistical analysis does not contribute anything to the study.

K-means clustering is appropriate for large datasets with a large number of variables, where other, more conclusive statistical methods cannot be applied in a simple way. This method is useful to group samples based on a multidimensional dataset, which cannot be represented in simple graphs. The authors here are evaluating a very limited number of variables (3), and the samples can perfectly be plotted in graphs and the grouping evaluated visually, which would not be per definition less good than the k-means clustering. K-means cannot show a significant difference between groups, but will split the samples/treatments in the number of groups that were arbitrarily chosen. If a group of samples is not different, but k=2 is set, the algorithm will result in two clusters. And the statement in the response to me, that k-means compares within cluster variability to between cluster variability, is wrong. This would be done by an ANOVA analysis, but k-means just creates the imposed number of clusters with minimal within-cluster variability. There is even no guarantee that the algorithm results in the best clustering (lowest within-cluster variability). Again, this method is thus only appropriate where other, robust analyses cannot be used.

The authors state that ANOVA did not show significant differences but “trends” were visible, after which they describe: “the samples first became darker then remained the same then became again a bit darker…”. The problem is, k-means cluster does not add anything to these observation, and ultimately, the conclusion of the manuscript is “the effects of different treatments on color change was not significantly different, but some looked like this and others looked like that”.

2. In the response to the comments, the authors state that the “The aim of the work was to evaluate the effectiveness of several coatings to prevent or delay color change of coated surfaces.” In that case, it is necessary to compare the color change of coated samples to color change of uncoated samples.

Author Response

Following the previous round of reviews, the authors consulted on several occasions with the statistician in our college and opted to redo the analysis based on Reviewer 1's comments.  Specifically, the analysis was performed using Welch’s ANOVA, Games-Howell pairwise comparisons tests, and a clustering procedure using distances between each pair of groups means for the 18 months ΔL*, Δa*, Δb* values.  As such, all of the analysis, results, discussion, and conclusion sections of the paper have been revised.

Regarding the reviewer's statement that uncoated controls were needed, we certainly agree that controls could have been informative.  However, because the aim of this study was to evaluate the ability of coatings (i.e., the decision to to use uncoated wood in exterior exposure in a temperate climate with high rainfaill, was not considered) to prevent color change during the experiment. In short, testing uncoated would not directly address the objectives of the study.

Reviewer 2 Report

The answers and completions of the authors of the manuscript are sufficient for the acceptance for publication of the paper in this form.

Author Response

We thank the reviewer for their time and effort reviewing this manuscript.  As another reviewer has suggested substantial changes to the analytical method used.  Specifically, the revised analysis was performed using Welch’s ANOVA, Games-Howell pairwise comparisons tests, and a clustering procedure using distances between each pair of groups means for the 18 months ΔL*, Δa*, Δb* values. As a result, the results, discussion, and conclusions of the paper have also been revised accordingly.

Reviewer 3 Report

The article has been significantly improved, but still requires some corrections.
Introduction
Line 143: I found that literature item 3 should be cited as: Laskowska et al.
Line 178: Please be consistent and use the past tense as in the sentence 174.
2. Materials and Methods
Line 192, 193: In the literature on the subject, we use the following terms: tangential, radial cross-section. Please verify this and make appropriate corrections.
Results
Table 2: These are the mean results for the tested variants? If so, standard deviations should be given. In Figure 3. Plots of mean and standard deviation of ΔE... you wrote.

Discussion
Line 541, 544: biological attack - Delete this stated as it has not been studied in the work.
Lines 257 - 260: No attempt was made to evaluate differences between species at each interval as this would not address the question often asked by practitioners, i.e., given a species, which coating will provide the longest color protection?
This sentence contradicts what is written below. Please verify the statements.
Lines 638 - 639: However, no single coating was uniformly protective on all the wood tested nor was any coating completely effective over 18 months of exposure.

Comments for author File: Comments.pdf

Author Response

We thank the reviewer for their time and effort reviewing this manuscript.  As another reviewer has suggested substantial changes to the analytical method used.  Specifically, the revised analysis was performed using Welch’s ANOVA, Games-Howell pairwise comparisons tests, and a clustering procedure using distances between each pair of groups means for the 18 months ΔL*, Δa*, Δb* values. As a result, the results, discussion, and conclusions of the paper have also been revised accordingly.

The reviewers comments on line 143, 178, 192, and 193 have been edited accordingly; use of the term biological was removed on lines 541, 544. 

For lines 257-260, the intent was to evaluate within species changes, if you will, and not between.  That is, the species is often a given - and with that choice now made, which coating works best? 

We do not understand the reviewer's confusion with the statement, but perhaps revising it as we've done "No single coating was uniformly protective within the range of species tested nor was any coating completely effective over 18 months of exposure."  In short, what we are saying is that color change occurred, most often in dark blotchy areas that were likely mildew, on all species/coatings combinations; for some this occurred within the first 6 months of exposure, for others it did not occur until the end of the study at 18 months.