Using ASTM E1920 for Accurate Porosity Measurement in Thermal Spray Coatings

Throughout the quality assurance fields within both scientific research and industry, many companies endeavor to continually improve the performance of their products. Some of these improvements are accomplished through great discoveries which provide breakthrough technologies, however most improvements are made by small and cumulative progression. Improved understanding of the key components and materials drives this improvement process, overall driving the development of key features and minimizing the associated error.

There is particularly a requirement for increased capability and effectiveness within the industries of aerospace and power generation. Coatings play a very important role in advocating progress within performance. Inconclusive results, however, can affect the central measurements of coating. During several analysis studies within various laboratories, it has been revealed that there is a fluctuation in results within repeated studies. The preparation process in coatings analysis can usually be repeated, but it is very difficult to reproduce results and measurements can often depend upon the individual carrying out the analysis rather than the material itself.

Background

It has been over 100 years since thermal spray coatings were first introduced [1]. The use of this coating is very basic – a feed stock material (usually wire or powder) is firstly heated until it is melted, before the particle is then accelerated at a very high velocity towards the component, which is deformed upon impact. Once the particle solidifies, it anneals to the surface of the component through mechanical bonds.

A vast array of materials can be annealed through this process, with the majority focusing on resistance to temperature, wear and corrosion. There are also several different mechanisms which can be used to administer this spray, each with unique traits [2, 3].

Thermal spay creates porous coatings, which can be distinguished by metallographic preparation. However, one complication can be identifying a preparation technique which is reproducible [4, 5, 6, 7, 8, 9, 10].

Recently, variation was shown between different laboratories and showed that the same results could not be obtained from identical polishing consumables. Analogous results also could not reveal a superior approach to metallographic preparation. This study finally showed that the high levels of variability were introduced during the measurement itself.

There was also shown to be a high level of variation within porosity standards between different laboratories, even when they were all accomplished within both metallographic preparation and sample analysis. Another analysis has revealed all laboratories can obtain acceptable repeatability within their results. This therefore suggests that the main causal factor within both the preparation and analysis of these thermal coatings is from the variation in approach.

This article will assess the various processes and identify the vital points of each stage. Existing papers also analyze other factors which may contribute, such as microscopes and other imaging systems [12].

Experimental Approach

Many different assessments have revealed that it is the method of preparation which has a profound effect on the produced results [4-10]. Examples which create issues include overly aggressive sectioning, grinding and inadequate mounting techniques. Many dispute the correct method of preparation and it has become a trend to list a specific method as the only viable option.

The objective of this work was to analyze the hypothesis that deviations within measurements of porosity were less correlated with the precise preparation procedure used, and more dependent upon the regulation of equipment, consumables and competency of the individual. This analysis was achieved using samples acquired from the differing T800 (HVOF) coating and a WC-Co (Plasma Spray) coating. The experimental approach outlined by ASTM E1920-03 (2014) [13], was used. Methods were also identified by assessing which approaches were the most commonly found within industry and are acknowledged as standard for preparation.

The first methods utilized an array of SiC papers for short periods of time, succeeded by a preparation step on a Trident cloth and ultimately polished using Colloidal Siliac on a Microcloth. The second method utilizes a singular SiC grinding stage which was succeeded by a pair of diamond polishing stages on no-nap surfaces and finally polished with Colloidal Silica.

The analysis was devised in order to scrutinize the effects of these two differing methods on the possible quantified porosity.

Effect of Sectioning and Grinding

The least damaging methods are favored within industry, however, defining the effect of sectioning damage upon a sample can be very complex. This analysis therefore selected a low damage method for sectioning and compared the damage in samples which were mounted before sectioning, with samples which were sectioned prior to mounting. The sectioning was carried out with a Isomet High Speed precision wafering saw, which contains a high-powered motor and therefore produced invariable cutting. A minimal cutting load was maintained on the sample by using diamond or Cubic Boron Nitride (CBN) wafering blades. The coating was also maintained within a compressed state and the blade was dressed regularly, as recommended in common practice.

Table 1. Preparation based on ASTM E1920 Method I

Surface Abrasive Lubricant/
Extender
Force (Per Specimen) Time (min:sec) Platen Speed Head Speed (rpm) Relative Rotation
1 CarbiMet 180 [P180] Water 25N Until plane 250 60 >>
2 CarbiMet 240 [P220] Water 25N 0:30 250 60 >>
3 CarbiMet 320 [P500] Water 25N 0:30 250 60 >>
4 CarbiMet 600 [P1200] Water 25N 0:30 250 60 >>
5 CarbiMet 800 [P2400] Water 25N 0:30 250 60 >>
6 Trident 3 µm Metadi
Supreme
Metadi Fluid 25N 4:00 150 60 >>
7 Microcloth 0.06 µm
MasterMet**
25N 2:00 150 60 ><

** last 15-20 second use water only >>Complimentary ><Contra

Table 2. Preparation based on ASTM E1920 Method II

Surface Abrasive Lubricant/
Extender
Force
(Per Specimen)
Time
(min:sec)
Platen Speed Head Speed
(rpm)
Relative
Rotation
1 CarbiMet 180 [P180] Water 25N Until plane 250 60 >>
2 UltraPad 9 µm Metadi Supreme Metadi Fluid 25N 6:00 250 60 >>
3 Trident 3 µm Metadi
Supreme
Metadi Fluid 25N 3:00 150 60 >>
4 Microcloth 0.06 µm
MasterMet**
25N 2:00 150 60 ><

** last 15-20 second use water only >>Complimentary ><Contra

Isomet High Speed saw – used at 4000 rpm and 3 mm/min cut speed with an Isocut HC blade

Figure 1. Isomet High Speed saw – used at 4000 rpm and 3 mm/min cut speed with an Isocut HC blade

Two sets of specimens were prepared utilizing the second method before assessment of the porosity was carried out. These specimens were then re-ground by imitating the first stage 6 times for 1 minute in each repetition, to make sure that all surplus sectioning damage was eliminated. The method was then carried out for a second time and the samples were investigated again.

Within the samples which were sectioned before mounting, the level of porosity was notably higher. This porosity reduced dramatically once it was reground. Comparatively, the samples which were mounted and then sectioned had the same porosity within both measurements. Therefore, enclosing the specimen before cutting prevented damages, which could also be reduced by grinding and reanalyzing the result. This study also revealed that the ASTM procedure also did not prevent damage which occurs due to sectioning.

Porosity analysis on samples prepared using Method II and different levels of sectioning damage

Figure 2. Porosity analysis on samples prepared using Method II and different levels of sectioning damage

Effect of Encapsulation Material

The least damaging methods are favored within industry, however, defining the effect of sectioning damage upon a sample can be very complex. This analysis therefore selected a low damage method for sectioning and compared the damage in samples which were mounted and then sectioned with samples which were sectioned prior to mounting. The sectioning was carried out with a Isomet High Speed precision wafering saw, which contains a high-powered motor and therefore produces invariable cutting. A minimal cutting load is maintained on the sample by using diamond or Cubic Boron Nitride (CBN) wafering blades. The coating was also maintained with a compressed state and the blade was dressed regularly, as recommended in common practice.

Table 3. Mount materials used in encapsulation experiment

Material Viscosity Mount Hardness Shrinkage
EpoThin 2 Very low 78 Low
EpoThin 2 (wet sample) Very low 78 Low
EpoKwick FC Very low 82 None
EpoKwick FC + ceramic bead Very low >90 None
Epoxicure Medium 82 None
SamplKwick (acrylic) High 78 Medium

Each of the samples were placed into a sole specimen holder which ensured that they were all uniformly prepared. They were then processed using the first and second methods previously described.

The second method proved to be more efficient than the first method, particularly on the harder specimens. The addition of ceramic bead to the mount particularly highlighted this difference. This addition also greatly reduced the rate of grinding and polishing removal, which was affected by the minimal wear rate of the ceramic. Ceramic also has a dampening effect on SiC, meaning that the effect is more pronounced in the first method.

Final polish on each specimen using ASTM Route I

Figure 3. Final polish on each specimen using ASTM Route I

Final polish on each specimen using ASTM Route II

Figure 4. Final polish on each specimen using ASTM Route II

Porosity analysis from all samples with varied mount materials after, Method I and Method II

Figure 5. Porosity analysis from all samples with varied mount materials after, Method I and Method II

Lower values and minimal deviation was observed in low viscosity epoxies, whereas SamplKwick acrylic demonstrated comparatively poor results. Crudely dried samples were easily damaged during incompetent preparation, but similar results were also observed with the second method.

Grinding through the effect can be achieved as the samples have minimal porosity, which provides limited connectivity. It is expected that upon incompetent cleaning and drying of samples, there would be a more significant effect in the higher porosity specimens. This would theoretically allow the identification of the superior preparation method.

Preparation Method

To analyze the consequence of the chosen preparation method, the previously identified method was used and two sample arrays were mounted. The method is shown below:

  • Firstly, the samples were cleaned/degreased.
  • Secondly, the samples were rinsed with water and soaked in ethanol for approximately 10 minutes in order to remove water from the pores.
  • Next the samples were thoroughly dried, whilst minimizing skin contact to prevent contamination with oils.
  • The samples were then mounted in EpoThin with vacuum impregnation using the Cast N Vac 1000 technology.
  • Isomet High Speed Saw was then used to section the samples. They were then re-mounted to the desired orientation.
  • Finally, they were prepared within Central Force holder to ensure maximal flatness and reproducibility.

It was identified that the porosity following preparation was comparable to the samples which were mounted before they were sectioned. Porosity was also measured throughout the preparation process. This revealed that apparent porosity is not altered within first method, which suggests that adequate damage was not removed within the stages preceding the true porosity measurement.

Development of apparent porosity during preparation of WC-Co coating

Figure 6. Development of apparent porosity during preparation of WC-Co coating

This was further investigated by extending each step within the preparation method and assessing the level of porosity at regular intermissions, until stabilization was achieved. This process highlights two factors: how much time preparation should take and which stages generate the most damage. Some materials showed highly significant damage, with particular steps more damaging than with other materials.

This assessment revealed that the observed levels of porosity of 12-15% were normal following the 9 µm stage. It also showed that utilization of the standard ASTM route could not fully minimize the porosity. Increasing the 9 µm/800 grit stage also did not alter the level of porosity much. Although, extending the 3 µm period led to a significant decrease in the apparent porosity. Further addition of a 1 µm stage on a TexMet C cloth also saw a large reduction. Therefore, it can be concluded that lengthening the 3 µm step and introducing a 1um step leads to a large improvement in the surface of the specimen.

Change in apparent porosity through preparation stages, modified route based on ASTM Method II

Figure 7. Change in apparent porosity through preparation stages, modified route based on ASTM Method II

Change in porosity through preparation, modified Method II

Figure 8. Change in porosity through preparation, modified Method II

Areas A (3 µm stage, 5 min) and B (1 µm stage, 10 min) from Figure 8 - digitally magnified

Figure 9. Areas A (3 µm stage, 5 min) and B (1 µm stage, 10 min) from Figure 8 - digitally magnified

Table 3. Optimized preparation for this sample, based on ASTM E1920 Method II

Surface Abrasive Lubricant/
Extender
Force
(Per Specimen)
Time
(min:sec)
Platen Speed Head Speed
(rpm)
Relative
Rotation
1 CarbiMet 180 [P180] Water 25N Until plane 250 60 >>
2 UltraPad 9 µm Metadi Supreme Metadi Fluid 25N 5:00 250 60 >>
3 Trident 3 µm Metadi
Supreme
Metadi Fluid 25N 15:00 150 60 >>
4 TexMet C 1 µm Metadi
Supreme
Metadi Fluid 25N 15:00 150 60 >>
5 Microcloth 0.06 µm
MasterMet**
25N 2:00 150 60 ><

** last 15-20 second use water only >>Complimentary ><Contra

[Note: the sample used in this work has relatively high porosity and therefore is more susceptible to grinding damage – typical preparation method for WC-Co coatings would typically use shorter steps]

Conclusions

High quality thermal spray coating preparation requires each preparation stage to be followed to a high standard. This process can also be improved by analyzing the sample at each step and fully understanding the process, which has proved more reliable than just following the recommended approach.

Accurately following a rigid method can produce repeatable results but they may not be accurate, as shown by the comparison of the two differing methods. Identical porosity was revealed once the processes of encapsulation and sectioning were normalized. However, a modified method saw a significant decrease in porosity, of 5-6% down from 12-15%.

Altering conditions can also affect the sample porosity, meaning that uncontrollable factors may also alter the outcome. These factors can include variability within the sample, consumable items and equipment. Therefore by using the equipment produced by Buehler equipment, a high standards of quality control can be ensured, enabling the production of repeatable and reproducible results.

Acknowledgements

Data originally presented (unpublished) to the Thermal Spray Society at “Thermal Spray Characterization: Materials, Coatings and Processes”  April 11-12, 2017 in Charleston, South Carolina Grateful appreciation to Richard Bajan of Curtiss-Wright Surface Technologies Division for the invitation to speak at this event and for direction in selecting samples for study.

References

  1. Siegman, S., Abert, C. “100 years of thermal spray: About the inventor Max Ulrich Schoop”, Surface and Coatings Technology Vol 220 pp3-13
  2. https://www.oerlikon.com/metco/en/products-services/coating-equipment/thermal-spray
  3. http://www.praxairsurfacetechnologies.com/components-materials-and-equipment/coating-equipment/thermalspray-coating-systems
  4. Elssner, Wellner “Preparation problems and phase identification in plasma sprayed coatings for high temperature applications” 1981 Microstructural Science 9 p289
  5. Blann, Diaz and Nelson “Raising the Standard for Coating Analysis”  1989 AMP Vol 136 6
  6. Geary “Metallographic Evaluation of Thermal Spray Coatings” 1991: Technical Meeting of the 24th Annual Convention, IMS, Monterey, CA. p 63
  7. Geary, Leonhart “Improved impregnation and fluorescence techniques in metallography for coatings” 1996 IMS meeting Pittsburgh, Pennsylvania
  8. Sauer, Wonderen “Standardization in Quality Control of Thermal Spray Coatings” 1997 - 2003, European Airline Committee for Materials & Technology Sub-committee
  9. Puerta “Advances in the Metallographic Preparation of Thermal Spray Coatings” 2005 Proceedings of the Combustion Turbine Coatings Symposium, Houston, TX
  10. Puerta, Anderson “The metallographic characterization of thermal spray coating microstructures” 2008 International Thermal Spray Conference & Exposition p 791
  11. Schnarr, Motl Ergebnisse der Auswertung „Thermische Spritzschichten“ (Results of Evaluation “Thermal Spray Coatings“) 2013
  12. Keeble, M “Error and Uncertainty in Metallographic Measurement”, ASTM E04 100 year symposium (to be published)
  13. ASTM E1920-03(2014), “Standard Guide for Metallographic Preparation of Thermal Sprayed Coatings”, ASTM

International, West Conshohocken, PA, 2014, www.astm.org

This information has been sourced, reviewed and adapted from materials provided by Buehler.

For more information on this source, please visit Buehler.

Ask A Question

Do you have a question you'd like to ask regarding this article?

Leave your feedback
Submit