TSS Membership Community

 View Only
  • 1.  Porosity evaluation

    Posted 28 days ago
    Hallo everybody,

    Can you briefly share what is your favourite coating porosity evaluation method and why? Please state your field/typical application if relevant. I am also interested in "how people from the industry see it".

    There are generally many ways how to evaluate porosity in TS coatings - Archimedean, gas pycnometry, Hg intrusion, image analysis, X-Ray CT, SAXS, .... Each has some significant drawbacks.

    In our academic lab, we mostly rely on image analysis (IA) of cross-sections observed by SEM.

    Pros
    • we do routine microstructural observations anyway :)
    • SEM provides in back-scattered imaging mode nice materials/porosity contrast (we do not use light microscopy as it is not reliable for semi-transparent ceramics)
    • SEM is able to provide even high-mag images -> porosity distribution may be acquired even for suspension plasma-sprayed materials with ultrafine porosity
    • works for metals/ceramics/cermets sprayed from powders or liquids by different TS methods
    • repeatible once the polishing method is standardized
    • easy to provide statistics from various spots within one cross-section
    • I understand the data (unlike other fancy but "black-box" methods)
    Cons
    • need to be sure that there are no pull-outs/sample preparion artifacts
    • preparation method & image processing may be user-dependent
    • relatively slow as vacuum embedding with low viscosity epoxy takes couple of hours -> usually not suitable for quick on-site porosity screening.


    ------------------------------
    Radek Musalek
    Institute of Plasma Physics CAS, Prague, Czechia
    Prague
    +420266053077
    ------------------------------


  • 2.  RE: Porosity evaluation

    Posted 27 days ago

    Hi Radek

    Generally we measure porosity of coating with the help of optical microscope in Industry for coated sample. We used cross-sections images to check at 50X magnification in microscopic evaluation with std. software measurement analysis with the availability reference of coating porosity and structure. 

    I hope this information will be helpful to your question. 


    SEM is very much useful for research and development organisation to study in depth of coating structure, matrix, oxide and phase analysis. 

    Thanks and Regards 



    ------------------------------
    Chirag Raval - M.Engg. in Met. & Mate. Engg.
    Thermal Spray Applications & Know How Technology Transfer
    chirag.ase@gmail.com
    ------------------------------



  • 3.  RE: Porosity evaluation

    TSS Board Vice President
    Posted 27 days ago
      |   view attached
    Dear Radek and Chirag,

    Thanks for posting and start commenting on this very important topic.

    Based on what I have seen and heard, I would say that image analysis (IA) of polished coating cross-sections is typically the "preferred" method of porosity evaluation as a quality control (QC) method; mainly in production environments of our TS community. It is because you need to find a compromise between having the porosity data and in a short processing time. I hypothesize that for many people, the mercury intrusion porosimetry (MIP) would be their "favourite" method, but it is not practical as QC.

    About 20 years ago I participated in a round-robin initiative of TS coating porosity measured via IA (polished coating cross-sections). Several institutions participated in this initiative, but some porosity results were significantly different depending on the institution that evaluated the porosity.

    That said, right now, 20 years later, it is amazing how these IA technologies and supporting equipment have evolved in the positive way, for example, (i) digital cameras, (ii) automated polishing machines and (iii) mainly the quality of the IA software packages available today.

    As Radek stated in his message, a standardized polishing procedure is paramount. Today, with automated polishing machines, you can add the different polishing recipes to polish different coatings into the computer of the machine (via its touch screen). It includes the polishing paper type, load, RPM, time and cooling. Therefore, even if the polishing machine is operated by different users, as long as they follow all the polishing steps to the letter (which are uploaded into the machine), the human error/input will be drastically reduced. Of course, we cannot forget the initial steps of sample prepping, which includes the impregnation and cutting; which are equally important.

    In my opinion, the most important supporting lab of a TS institution is the metallography lab. What I mean is; you may use a state-of-the-art computer-controlled console, a modern TS torch, measure Tp & Vp with great accuracy/precision and use an infra-red camera to measure the temperature of the substrate during coating deposition. This is fantastic. But if you employ a "non-optimized" metallography procedure to evaluate your coating, you will not have a representative microstructure and you will be led to wrong conclusions. Consequently, you will lose the effort you put in trying to make your coating as best as possible in the first place.

    For those who are beginning to standardize polishing procedures of TS coatings, the ASM-TSS book Thermal Spray Technology: Accepted Practices; which is open access via the web, can provide very useful sets of info to our TS community.

    The link is here: Thermal Spray Technology: Accepted Practices | Handbooks | ASM Digital Library (asminternational.org)

    Of course, there are still important challenges remaining. For example, some people use SEM. Other people use OM. Moreover, there are issues about the magnification of the pic. Some people use 50X, others 100X, 200X or higher. All these issues will influence the measurements. When reporting the data in a report, I suggest (if possible) adding the following supporting info:

    - For SEM: (i) magnification, (ii) working distance and (iii) kV
    - For OM: (i) magnification and (ii) the exposure time
    - For all: (i) the threshold value, (ii) the original picture and (iii) the threshold-marked pic (in this way, if the coating porosity value is 11%, people will see what it is meant by that)

    Therefore, although there are still important challenges to be overcome, I guess that IA is becoming more and more acceptable as a QC method to measure the porosity of TS coatings in production environments and even labs, or any time you need to measure coating porosity on a regular basis.

    However, for sure it is not a panacea. For a more fundamental understanding of the porosity nature of TS coatings, Archimedes and MIP are paramount. As an example, I am attaching a very interesting paper from Dr. Robert Vassen's group. Free-standing YSZ TBCs were removed from substrates using different methods, and their porosity values were measured via MIP. Depending on the way the YSZ TBC was removed from the substrate, the porosity values could be distinct. A very-very good paper.

    I hope it helps!

    ------------------------------
    Rogerio Lima, PhD
    Senior Research Officer
    National Research Council of Canada
    ASM Thermal Spray Society (TSS) Vice President
    ------------------------------

    Attachment(s)



  • 4.  RE: Porosity evaluation

    Posted 26 days ago

    Dear Rogerio 

    Thank you so much. 

    You've described in a very well in brief information.  

    Yes that's in industry now best method is to check the porosity via OM IA with automatic measurements software in polished sample cross-section image. 

    Thank you 😊 



    ------------------------------
    Chirag Raval - M.Engg. in Met. & Mate. Engg.
    Thermal Spray Applications & Know How Technology Transfer
    chirag.ase@gmail.com
    ------------------------------



  • 5.  RE: Porosity evaluation

    Posted 26 days ago
    Hello All, 

    In my previous industrial experience at a TS materials manufacturer, we relied on auto-thresholding Image analysis of coating cross sections utilizing SEM images for porosity evaluation. On that point  I would concur with Radek regarding the Pro's and Con's of the method. I would say though that this method is suitable for large scale evaluations on a regular basis if performed correctly.
    In our case we had to process a significant number of cross sections each week depending on the coating trials that were performed. IA from SEM images was the only method that was rapid enough and accurate enough to manage such a workload. For this of course we had fixed preparation procedures as well as standard operating procedures used to perform the SEM imaging and image analysis itself. With this you may minimize the potential for error. 
    Its important to point out that this evaluation procedure covered a number of different coating types, from highly porous TBC coatings to very dense ceramics and metallic coatings. The only difference in this was that the higher the coating porosity is, the greater the number of Images required to get a stable average porosity value. for dense coatings 10 images was often sufficient, with additional images bringing only small returns in accuracy.  for TBC's where the porosity may be 15%-30%; then 20 images was the minimum required for a stable evaluation. The same is true for SPS thermal barrier coatings.

    We had on a number of occasions, compared the SEM IA results to those made on multiple free standing coatings (Archimedes evaluation). While the bulk measurement of porosity  (and coating density) was more accurate with a lower standard deviation. There was a good match for the absolute values when comparing with SEM (IA). For information, the bulk values were measured on samples prepared for thermal diffusivity measurement (LFA) and it was typical to prepare 4-8 free standing samples for measurement. For that analysis you require quite accurate measurements of coating density. 

    For optical images, I also started out with that years ago. As it was pointed out, it becomes inaccurate for coatings that are optically translucent, like YSZ. It can be accurate enough for other coatings like metals or carbides. That said, as industry has moved onto more refined (fine scale) coatings like HVAF and suspension sprayed coatings, Optical imaging hits its limitations. In a number of cases I've seen OM under or overestimate the porosity compared to SEM (IA) and bulk porosity measurement methods. For Suspension plasma spray coatings, we tried to use OM initially but concluded that the image magnification and resolution wasn't up to the challenge of detecting such fine scale porosity accurately. 

    I think a factor that now works in the favor of SEM analysis is the proliferation of the 'desktop SEM', for instance the TM3000-TM4000 series from Hitachi. These machines now approach the cost level of a high quality optical microscope set-up. They no longer require dedicated operators to run and can readily produce high quality images for porosity analysis. For the kind of rapid evaluations needed in coating and materials development, these smaller SEM's can do the bulk of the workload. 

    Overall its an important question to ask...."How do you measure your porosity?" 
    When we follow different procedures and methods, we are given different answers. There are some coatings out there in industry that are today metallographically prepared 'incorrectly' on purpose. this is simply due to the fact that if they are polished as we now know is correct to see the real microstructure, they fail the specification requirements in quality control. The coating itself has not changed in the 20 plus years its been produced, but our understanding of how to prepare them has. Coating specifications on the other had, are not updated as often.

    ------------------------------
    Nicholas Nicholas Curry
    Consultant
    Thermal Spray Innovations
    Bruck an der Großglocknerstraße
    ------------------------------