Hello everyone,
I welded 4 aluminium (Al 5754-H111) sheets with Friction Stir Spot Welding technique (pin-less tool). I got a strong bond between the sheets, However, in joining line it can be seen thin layer made of particles (it is not compact) (Figure 1). I suppose it is a trapped alumina (al2O3). However, I read in book Corosion of Aluminium: "Certain elements will strengthen the protective properties of the oxide film by forming mixed oxides, if their structures are compatible. This is the case of magnesium. For this reason, alloys of the 5000 series have excellent corrosion resistance."
Unfortunately, it has no more details. So now I would like to know which type of mixed oxide will form during or after rolling on top of the 5754 alloy and its approximate thickness. Actually, any information (or publication) about this oxide would be very helpful.
It is also very interesting that when I etch sample with Barker's reagent (Figure 2) somehow this line become more visible (thicker) . When I looked at SEM (Figure 3) I saw that surrounding material is etched so it looks like thicker line on LM. I suppose that some chemical reaction occurred between Barker's reagent and this oxide that made this to happen. But I can't find anything in the literature (or I am searching in the wrong place). I would appreciate any help.
The possible oxides could be Al2O3, MgAl2O4 (spinel), and MgO. Fluoroboric acid could react with these oxides (especially when you anodize at high potential) and may form a wide range of products including fluoroaluminate, magnesium oxyfluoride, magnesium fluoride etc. These are the possibilities. In order to know the exact products, you may want to carry out Raman microscopy on these particles.
From SEM image it looks like on the interface we have etching-pits, not the precipitates... a combination of SE/BEI SEM images of the same area would provide some more details if these are particles or not... I see the structure is about 5 um in size, so if this are particles, a simple EDS would give you a reasonably good indication what is happening here.
I think if you carry out a surface analysis on the continuous surface oxide layer, I doubt if you will detect any mixed oxide on the rolled plate. Mostly the oxide will be of Al2O3 doped with Mg - not a spinel structure. This layer may be of 25-160 nm thick, as per a published paper by M. Fishkis and JC Lin, Wear 206 (1997) 156-170. Beneath this Al2O3 surface layer, oxide particles of MgAl2O4 and MgO could be present.
Thank you both very much. Janez Zavašnik I tried to do EDS but it is too small. SEM figure shows etching-pits, but oxide particles are smaller then that. And the problem is that layer is broken into small pieces during FSW so I think that with EDS I grab a lot of surrounding material. So for example, in one EDS I have 3%Mg, 23%, 71% Al and 3%Si (I have no idea how I got Si here) so I can´t know what is that exactly. I also got the results of XRD but I am first time using it and not sure what I got there. For me it is strange that when I measured that layer I got 1-3 micron thickness. And I found in literature that alumina is 2-10 nm or as Krishnan Raja says 25-160 nm. In my case is bigger. So I am not sure what this layer is.
Krishnan Raja thank you very much for the paper.
Dear Dr. Danka Labus Zlatanovic,
5754 aluminium alloy offers extremely good resistance to both seawater corrosion and industrially polluted atmospheres; for these reasons is used in applications in market sectors such as marine, oil & gas, chemical and nuclear. In general it is a material that should not give you big problems of corrosion but, I suggest you take a look at the following note/papers where you could find some information related to the joining method:
- Corrosion Behaviour of Friction Stir Welded AA5xxx Aluminium Alloys
A thesis submitted to The University of Manchester for the degree of Doctor of philosophy (2012)
The School of MaterialsCorrosion and Protection Centre
Ramadan.R.Abuaisha
Available at: https://pdfs.semanticscholar.org/3fda/2db79acf4f2c9b70ddb89e6dab7c92d14c8b.pdf
- The Effect of Vibration during Friction Stir Weldingon Corrosion Behavior, Mechanical Properties, and Machining Characteristics of Stir Zone
Sajad Fouladi, Amir H. Ghasemi, Mahmoud Abbasi, Morteza Abedini, Amir Mahyar Khorasani and Ian Gibson
Metals, 7(10), 421; https://doi.org/10.3390/met7100421 (2017)
- Corrosion behavior in AA5754 friction Stir Selded Alloys
Conference: Workshop on Material 2013 Second
Filipe Nascimento, Pedro Vilaca, R.M. Miranda, J.C.S. Fernandes, Francisco M. Andrade Pires
Available directly on ResearchGate
Best regards, Pierluigi Traverso
Hi Danka,
https://d-nb.info/945477562/04 It's a phd-Thesis in German about "Diffusion Bonding of high strength Aluminium" - alloys and it has a chapter abouth the growth and structure of the oxide films, even under the influence of additional elements such as Zn and Cu. And you can also use the cited literature. If the link doesn't work, please send me an email.
About your question for the Barker's reagent - ist no etching solution, so what you see aren't classic etching pits. Barker is an anodization of the aluminium that causes a thin epitaxial oxide-layer, which influences the optical reflection of the polarized light.
For the pits - the anodization causes a typical surface structure as in http://geb.uni-giessen.de/geb/volltexte/2009/7353/pdf/LotzAndreas-2009-11-24.pdf. It may be that the effect to your pre-Barker-oxide layer is different, because the oxide layer doesn't grow on top but spreads into the aluminium.
That's why I also wouldn't recommend any analysis of the oxide layer on post-Barker samples.
Many greetings to you and your family, Franzi!
Hi Franziska Petzoldt,
How I didn´t remember to ask you first. :) Thank you very much for your update it is very helpful. I will send you an E-mail because in the first link it is only a beginning of the thesis and also to ask you for some more details.
Many greetings to you and your family also.
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size