Manual TIG welding of Fe10Cr4Al+RE: overlay on an austenitic substrate and butt weld using alumina forming austenite consumable

(1)

Manual TIG welding of Fe10Cr4Al+RE: overlay on an austenitic substrate and butt weld using

alumina forming austenite consumable

HENRIK FRÖLUND

Master’s Thesis in Nuclear Energy Engineering Technologies Supervisor: Peter Szakálos

Examiner: XXX

TRITA xxx yyyy-nn

(2)

(3)

Abstract

This project focuses on the manual weldability of an alumina forming ferritic stainless steel (Fe10Cr4AL+RE). The theoretical weldability of the material was researched and found to be poor without the use of austenitic consumables which would improve mechanical properties such as ductility. As such, the use of the Fe10Cr4Al+RE as an overlay material were seen to theoretically difficult, if not impossible, for use cases where ductility is demanded. The material standards for the nuclear power generation industry demands ductility for overlay welds, but this is not the case in other energy production industries such as with concentrated solar power (CSP).

Fe10Cr4Al+RE was used as an overlay material on top of 304L austenitic stainless steel substrate using manual TIG welding to apply it. The overlay weld was then tested for ductility using a bend test. The test showed that the overlay weld had a low ductility which is due the large ferritic grains; the reason it had any ductility at all was assumed to come from mixing with the substrate which also could lessen the corrosion resistance if the amount of Ni was too high.

The presence of Ni in the overlay weld was later confirmed through the use of a scanning electron microscope (SEM). However, the levels were low enough that they should not threaten the integrity of the corrosion resistance.

In another experiment two pieces of the Fe10Cr4Al+RE was joined together with Nikrothal^® PM58 consumables; they were joined with a manual TIG butt weld. The welding of Fe10Cr4Al+RE together with Nikrothal^® PM58 consumables had good ductility and managed an angle of bending close to 180^◦. This marks the first time that two pieces of any FeCrAl-alloy has been welded together without pre- and post-heat treatment while still remaining ductile.

A final weld test was also done with an experimental alumina forming austenite with significantly lower Ni-content then the PM58 as consumable for a butt weld; this weld cracked during cooling and could not be tested further.

(4)

Sammanfattning

Manuell TIG svetsning av Fe10Cr4Al+RE: påläggsvetsning ovanpå austenitiskt substrat samt stumsvets med aluminiumoxid bildande austenitiskt tillsatsmaterial

Det här projektet fokuserade på den manuella svetsbarheten hos ett aluminiumoxidbildande ferritiskt rostfritt stål (Fe10Cr4Al+RE). Den teoretiska svetsbarheten av materialet undersöktes och visade sig vara dålig utan användan- de av austenitiska tillsatsmaterial som skulle öka de mekaniska egenskaper- na hos svetsen, såsom duktilitet. Av den anledningen ansågs användandet av Fe10Cr4Al+RE som ett påläggsmaterial vara teoretiskt svårt om inte omöj- ligt, för användade i processer där duktilitet är ett krav. Materialstandarderna inom kärnkraft kräver duktilitet hos påläggsvetsar, men så är inte fallet inom vissa andra energiproduktioner såsom exempelvis koncentrerad solkraft (CSP).

Fe10Cr4Al+RE testades som påläggssvetsmaterial ovanpå ett substrat av 304L austenitiskt rostfritt stål via manuel TIG svetsning. Duktiliteten hos påläggsvet- sen testades sedan i ett böjprov. Böjprovet visade att påläggssvetsen hade låg duktilitet vilket tillskrivs de stora ferritiska kornen; anledningen till att den hade någon duktilitet alls tillskrevs uppblandning med substratet vilket även kunde sänka korrosionsbeständigheten om Ni-halten var för hög. Närvaron av Ni i påläggsvetsen bekräftades senare med ett svepande elektronmikroskop (SEM), men nivåerna var tillräckligt låga för att inte påverka korrosionsbeständigheten.

I ett annat experiment fogades två bitar av Fe10Cr4Al+RE med Nikrothal^® PM58 som tillsatsmaterial; de sammanfogades i en stumsvets utförd genom manuell TIG svetsning. Böjprovet av stumsvetsen med Nikrothal^® PM58 som tillsatsmaterial visade på god duktilitet och klarade böjning nära 180^◦. Det här markerar den första gången någonsin som två bitar av någon FeCrAl-legering sammanfogats utan för- och efter värmebehandling med bibehållen duktilitet.

Ett sista svetsprov utfördes även där en stumsvets svetsades med ett experi- mentelt nytt aluminiumoxidbildande austenitiskt tillsatsmaterial som har be- tydligt lägre Ni-halt än PM58; svetsen sprack under svalning och inte kunde testas ytterligare.

iv

(5)

Preface

I have received a tremendous amount of help with my work leading up to, and during, this project, without which I would’ve not succeeded. It is not possible to name them all here, but I feel it’s important to name those who had a direct hand in the success of this project.

My sincerest gratitute to the two welders Johan Frölund and Andreas Bergdahl who dedicated both time and expertise to this project. Andreas is a third year student at the Fredrika Bremer high school in Stockholm and has showed both skill and responsibility in his work on the overlay welding of this project. His teacher, and my contact with the high school, Joel Andersson was also vital in bringing together the needs of this project with the skill of Andreas. Johan Frölund is a highly decorated welder with almost 30 years of experience, working the last 10 years on oil rigs around the world. Johan’s practical expertise was focused on the welding of Fe10Cr4Al+RE with Nikrothal^® PM58 but he has also been a source of theoretical and practical knowledge throughout the project.

Kanthal AB has been kind enough to provide both insight and materials for this project and they were thus invaluable. Most of the handling of the materials, as well as the bend tests were done with the help and supervision of Rolf Helg, Bosse Barksäter and others at the workshop of Albanova.

The support and guidance that I have received from my supervisor Peter Sza- kalos as well as his graduate student Peter Dömstedt has above and beyound ex- pectations. I thank them for this and for trusting their project in my hands.

Henrik Frölund Stockholm, maj, 2020

(6)

Introduction

This chapter will introduce the project and thus explain background, purpose, scope and method of it.

1.1 Background

The climate is changing and although the causes are many and the relationship between them and the effects are complicated one of the biggest causes are the green- house gas emissions (GHGs).[1] According to the European Environment Agency two thirds of the GHGs comes from the burning of fossil fuels for energy.[2] The generation of nuclear energy produces no GHGs, however it does produce some at the pre-generation stage of mining for fuel, constructing the plant, etc.. Public has not been supportive of nuclear energy and it is the nuclear waste which is the prominent cause of mistrust.[3]

A step towards using up existing nuclear waste and producing less of it is the Generation IV (GIV) nuclear reactors. [4] There are different kinds of GIV reactor designs; one of them is the Pb cooled fast reactor design.

The IAEA have recognized that one of the drawbacks with Pb reactor coolants is the damaging "[...] of components and fuel elements by the high corrosivity of heavy metals.", and that the elimination of this drawback can be "[...]obtained by selecting/developing appropriate structural materials[...]". [5]

A new such structural material has been developed by researchers at KTH in cooperation with Sandvik AB/Kanthal. The material, Fe10Cr4Al+RE, is resistant to such high temperature corrosion as can be expected in future energy plants which make use of liquid Pb as its reactor coolant.[6][8][27]

The resistance to corrosion of the Fe10Cr4Al had been shown, however investigation concerning the weldability of the material had not been done and was one of the most urgent future topics to be researched. [6]

(12)

CHAPTER 1. INTRODUCTION

1.2 Purpose

The purpose of this thesis was to examine the welding properties of a new corrosion resistant alumina-forming alloy called Fe10Cr4Al+RE; both as a base material and as cladding.[10]

1.3 Scope

This master thesis is primarily focused on the possibility of manual welding with the new alloy. There is already ongoing research into the welding of this alloy at Walter Tossto in Italy, using welding robots and with full control over pre- and post weld heat treatments. A material that is easily welded (e.g. manual welding) can be viewed as more valuable than a material which can only be welded through expensive equipment or complicated processes. As such this thesis will look at manual welding without pre- and post-weld heat treatment, mechanical treatment and other higher degrees of control beyond the manual capability of a welder.

The research in Italy is mainly focused on the possibility of using the new alloy as a cladding on top of another stainless steel substrate. Part of this project will look into this as well and another part will focus on the joining of two pieces of this alloy through manual TIG welding.

The as-cast ductility of the FeCrAl is very low [11] and it needs to be hot- rolled before it can be used as a construction steel with sufficient ductility. If one were to remelt the steel, the microstructure would return to the brittle cast structure. Melting the steel is necessary in order to weld with it and it is likely that the new weld structure of the steel will be similar to the cast structure; in theory the weldability of this alloy should therefore be low. This thesis will try and experimentally ascertain if this is true in practice or not.

1.4 Method

The method for this project can be divided into theoretical and experimental. The theoretical part of the project will search for sources of knowledge regarding weldability for similar materials and through these try to ascertain the weldability of the Fe10Cr4Al+RE as well as to find a basis from which to do the experiments.

The experimental part of the project will be of an explorative nature where the focus is not to establish a standardized welding procedure, instead it will try to narrow the possibilities and find the way forward for future research.

2

(13)

Chapter 2

Theory

This section will focus on the theoretical knowledge regarding both the materials used and the method of which they have been welded and then tested. The goal here is not to give an in-depth knowledge of either but to provide the necessary information for the reader to understand the experiments, testing and results.

2.1 Materials

In this section the different materials that were used are introduced with a descrip- tions of their properties.

2.1.1 Fe10Cr4Al+RE

The main material studied in this project is an alumina forming ferritic stainless steel with composition Fe10Cr4Al+RE. The material was developed in cooperation between KTH and Sandvik Heating Technology AB as a part of the Doctoral Thesis of Jesper Ejenstam, PhD.[6] The material was developed to be able to withstand the highly corrosive liquid Pb in high temperature Pb-cooled nuclear reactors, however it has since also been thought of as a part of future concentrated solar power plants, CSP, which also use liquid Pb as a coolant.[27]

Alloy Fe Cr Al C RE-mix

Fe10Cr4Al+RE Balance 10 4 < 0.02 0.2

Table 2.1. Chemical compostion of the Fe10Cr4Al+RE alloy used in the thesis.

Values in wt. %.

Alumina forming ferritic stainless steels are not new. Hans von Kantzow patented the FeCrAl alloy in 1929[7] and the company AB Kanthal was formed around it.

The alloy could endure very high temperatures with little degradation and were therefore excellent as a resistance wire which is the product produced and sold by Kanthal.[6]

(14)

CHAPTER 2. THEORY

The reason for the alloys high resistance to corrosion is the forming of a thin and stable protective alumina on the surface with self healing properties. This thin alumina has a high chemical resistance[6] and the formation of it is what gives the alloy its resistance to corrosion. The formation of this protective alumina is due to the precise balance of Al, Cr and Reactive Elements (RE) such as Ti, Zr and Y.

The exact chemical composition of the alloy can not be disclosed in this report but a general composition can be seen in Table 2.1.

The chemical composition of the alumina is Al2O3 and a type of alumina layer will also form on a binary alloy such as FeAl. Such an alumina, however, would not have the same properties and it would require up to 16 wt. % of Al to produce it;

this would make manufacturing of the alloy more difficult[6]. By adding Cr to the alloy, one can drastically reduce the amount of Al to levels as low as 3 wt. %; this is known as the "Third Element Effect".

There is no consensus regarding how this effect exactly works but there are different theories, all of which are beyond the scope of this thesis.[12] However, the fact is that initially a chromium rich oxide always develops on the surface and the most accepted theory is that it is this initial oxide which allows for the creation of the aluminum oxide underneath.

FeCrAl-alloys typically have a chromium content of around 20 wt. % in order to form the Al2O3 from which the high resistance to corrosion comes. However, there is a known embrittlement of these alloys at temperatures around 475 ^◦C which makes them unsuitable for nuclear applications. What happens around this temperature is known as spinodal decomposition, or phase separation, where "the microstructure decomposes into a Fe-rich phase (α) and Cr-rich phase (α⁰) in a particular temperature range [...] which in turn leads to loss of ductility"[6], see Figure 2.1.

This new FeCrAl with a lower chromium content of 10 wt. % and additional RE does not show any signs of embrittlement in the usual 475 - 500^◦C zone or even in higher temperatures up to 800^◦C in tests of up to 10000 hours.[6][13]

A problem with ferritic stainless steels in general is the low ductility mostly due to the formation of large ferritic crystals. While cooling the slabs often crack and the phenomena has been named "clinking" due to the accompanying sound.[11]

This brittleness would make the FeCrAl unfit for most use and Sandvik/Kanthal has therefor solved it by hot-rolling the FeCrAl. The hot rolling process deforms it which gives rise to recristallisation of smaller and equiaxed crystals after which the material can be bent 180 degrees without cracking.

2.1.2 Nikrothal^® PM58

This is a alumina forming Ni-based material with proven good corrosion resistance at high temperatures in air. [14]

4

(15)

2.1. MATERIALS

Figure 2.1. FeCr phase diagram [6]

Alloy Ni Cr Fe Si Al RE

Nikrothal^® PM58 Balance 19 18 0.4 5 added

Table 2.2. Chemical compostion of Nikrothal^®PM58. Values in wt. %

2.1.3 Austenitic stainless steel; 304L

Austenitic stainless steels are the most commonly used stainless steels because of their good corrosion resistance as well as good formability and weldability.[15]

Austenitic stainless steels have generally been alloyed with 12-27 wt. % Cr and 7-30 wt. % Ni[17] but different amounts and additions of other elements create austenitic alloys, or "grades" with specific qualities; the addition of Mo increases corrosion resistance for example while the addition of Nb will increase mechanical properties at higher temperatures.[15]

One such "grade" are the "304" alloys which is one of the most widely used and versatile of the austenitic stainless steels; it is also known as "18/8" because of its composition of 18 wt. % Cr and 8 wt. % Ni.[24] The "304L" (EN 1.4307) is a version of this steel with lower C; the 304 (EN 1.4301) has a maximum of 0.08 wt. % C and the 304L has maximum of 0.03 wt. % C. The lower amount of carbon makes the steel structurally weaker, but it increases weldability and corrosion resistance

(16)

CHAPTER 2. THEORY

by minimizing or even eliminating carbide precipitation[25].

Carbide precipitation is when C binds to the Cr in the boundaries between the crystallites of the material which lowers the Cr content in the boundaries. This can then result in intergranular corrosion (IGC) since the boundaries with lower Cr content have a lower corrosion resistance. The phenomena where these carbides form during welding is known as sensitization.

2.1.4 Alumina-forming austenitic stainless steel (AFA-A5-8)

The mechanical properties of austenitic stainless steels combined with the high corrosion resistance of the Al2O3surface layer is of interest for use in high temperature environments. Such steels are known as Alumina-forming austenitic stainless steels (AFA).

Alloy Fe Ni Cr Al Nb C

AFA-A5-8 Balance 23 16 3.14 1.19 0.01

Table 2.3. Chemical composition of AFA-A5-8, an experimental alloy within the EU project GEMMA[28]. Values in wt. %.

The properties of AFAs are interesting for use in Gen. IV nuclear power plants.

The solubility of Ni in Pb is greater than Fe[19] and it is therefore crucial that the chemical composition is such that the protective alumina-layer is formed. The formation of an alumina-layer on a austenitic crystal structure is harder to achieve than on a ferritic crystal structure because of the low diffusivity of aluminum in the structure. To combat the low diffusivity it has been shown that the addition of Nb increases the solubility of Cr and Al in the austenitic phase. This leads to a positive effect on the oxidation resistance by enabling alumina to form at lower concentrations.[6].

One such composition that has shown promise is shown in Table 2.3.

2.2 Welding

An overview of overlay welding with specifics related to arc welding and TIG welding is being presented here; the difficulties in regards to welding of ferritic stainless steels will also be discussed.

2.2.1 Arc welding based overlay

There are many different ways of getting a coating on top of a substrate. In regards to corrosion protection of metals, the most common ways of doing it are through welding or spraying. This project focuses on welding, and more specifically manual arc welding using a TIG weld.

6

(17)

2.2. WELDING

In all arc welding based overlay, the dilution of the overlay material is of utmost concern.

Dilution= As

A_s+ Ao

·100 [%] (2.1)

As can be seen in equation 2.1, the dilution of the overlay weld is a measure of the percentage of the substrate that has been mixed with the overlay material to form the weld. In this project, the dilution will be estimated using scanning electron microscope(SEM) to measure the amount of Ni at different points in the weld.

There are many different variables in welding but the ISO standards are most concerned with arc energy (AE) when qualifying an overlay welding procedure.[16]

The calculation for the arc energy is quite simple and can be written as

AE = 60UI 1000v

kJ mm

. (2.2)

Looking at equation 2.2, the most interesting variables in TIG welding are I and v and it is obvious that the relationship between these can be changed without affecting the resulting arc energy. Another term for arc energy is "heat input" and it is used in all areas of arc welding, not just overlay.

TIG

TIG-, or "Tungsten Inert Gas"-, welding is a type of welding technique where the welding arc is established between a tungsten electrode and the work piece. The arc is shielded by an inert gas, usually argon, which flows from the gas nozzle, within which the electrode is mounted centrally. [17]

Figure 2.2. TIG welding schematic [20]

(18)

CHAPTER 2. THEORY

The technique was developed during the second world war for aluminum welding and is now used withing the process industry where high quality welding of pipes with regards to homogenity, quality, purity and finish of the weld is demanded.

It is used in many other industries when welding thinner goods (up to 3 mm) of stainless steel, aluminum-, copper-, and magnesium alloys.

Manual TIG welding can be done with or without consumables and when they are used the are added manually by the welder in contrast to MIG/MAG welding when consumables are added automatically during the welding. [17]

In Figure 2.2 there is a "Copper shoe" placed underneath the weld. This is a root support which helps form the root, or back side, of the weld by supporting the melt and thus helps ensure complete joint penetration. Welding steel against at Cu-plate can be problematic since it might lead to liquid metal embrittlement (LME). LME is a loss of ductility in a material when it is stressed and in contact with a liquid metal. If LME is a possibility then a ceramic root support is a good alternative. There was no root support used during the welding in this thesis.

2.2.2 Testing

The initial question about the weldability of the Fe10Cr4Al is in regards to the ductility of the finished weld. In order to determine the ductility of these experimental welds it was decided that a bend test would be appropriate.

Figure 2.3. Schematic of bend test using a mandrel. [16]

The bend test employed in this project is based on the one known as "Test with mandrel" which is described in ISO 5173:2009[18] and a schematic of this type of test can be seen in Figure 2.3. The test piece (rectangular in Figure 2.3) is bent by moving the top mandrel down on it while keeping the bottom two fixed; the angle α is then measured. One can either perform the test until failure by closely examining the test piece during bending, or one can perform multiple test to certain values of α and the analyze the test pieces for any failures.

8

(19)

2.2. WELDING

Other than the actual test, ISO 5173:2009 also describes in detail how to take test pieces from a larger welding sample as well as how to prepare them before the bend test. It also gives suggestions on how to produce a report of the test.

(20)

(21)

Chapter 3

Practice

This section will provide information of the practical aspects in regards to preparation, execution and testing of the welds. It will be separated into two subsections based on the type of welds that will be done.

3.1 Theoretical weldability of Fe10Cr4Al+RE

This section will present what theory says about the weldability of Fe10Cr4Al+RE both as an overlay material and with regular welding.

3.1.1 Overlay "weldability"

A successful overlay weld would have as little dilution as possible in order to have a chemical composition close to that of the cladding material. The welder had to investigate for himself at what energy input(kJ/mm) he would achieve sufficient melting of the Fe10Cr4Al+RE without melting the substrate to much.

Knowing that the Fe10Cr4Al+RE needs to be hot-rolled after casting in order to achieve ductility, there was little hope of succeeding with a ductile non-diluted overlay weld.

3.1.2 Weldability of Fe10Cr4Al+re using consumables

The Fe10Cr4Al+RE is a new and unique material with regards to its chemical composition, therefore one can not look at recommendations or investigations into the weldability of similar alloys.

What can be done is to look at general guidelines regarding welding of ferritic stainless steels as well as to try and estimate the chemical compostion of the finished weld in order to pre-assess the mechanical properties of it.

(22)

CHAPTER 3. PRACTICE

General guidelines of ferritic stainless steel welding

There are several problems when it comes to welding of ferritic steels. Small amounts of martensite can form and high temperature can also result in very rapid grain growth affecting ductility negatively. Both of these problems are generally seen when welding with these types of materials but the effects can be mitigated through keeping the heat input low and using austenitic consumables. There are also other alternatives such as post-weld heat treatment, mechanical treatment, control of temperature between different weld runs, etc., but these are beyond the scope of this thesis.

Using the correct austenitic consumables when welding will produce a tougher weld metal, but the low heat input will be what affect the heat affected zone (HAZ).

The HAZ is the non melted part of the parent metal right next to the welding which has been heated to such a degree as to have underwent changes in material properties.

Pre heat treatment is not seen as required when welding thin sections up to 10 mm. [23]

Estimating the chemical composition of the finished welds using a modified Schaeffler diagram

Published by Schaeffler in 1949[21] the Schaeffler diagram has since been used as a good way of estimating the composition of a finished weld.

Figure 3.1. A Schaeffler diagram [22]

12

(23)

3.2. PRACTICAL WELDABILITY OF FE10CR4AL+RE

As can be seen in Figure 3.1 the diagram is made up of phase fields and isofer- rite lines. Using it, one calculates the Chrome equivalent (X-axis in Figure 3.1) and Nickel equivalent (Y-axis in Figure 3.1) for the two different materials and plots them as two points in the diagram; drawing a line between the points, the composition of the weld is found on that line. Where on the line is dependent on the percentage distribution of the two materials in the weld.

All of the materials in this study has aluminum in its chemical composition which is not included in the diagram seen in Figure 3.1. A different Scheffler diagram has been made which takes aluminum content into the calculations[29] and this is the one that has been used here. However, it should be noted that the author of this report has not seen any work done which checks the accuracy of this diagram to any of the new materials (i.e. Fe10Cr4Al+RE, AFA-A5-8 under investigation in this project. As such the results of the plots are not reliable for use to determine the properties of the finished weld. Furthermore, there has been no study of the microstructure in this thesis in order to see the phases of the weld metal.

The calculations and plots in the Aluminum-Schaeffler diagram can be seen in Appendix A.

3.2 Practical weldability of Fe10Cr4Al+RE

This section will present how the weldability of the Fe10Cr4Al+RE was tested practically with explorative testing.

3.2.1 Overlay welding of Fe10Cr4Al on AISI 304L austenitic stainless steel

A large piece of the austenitic stainless steel was cut into smaller sections of approximately 100x45x15 mm. This was done in order for the welder to be able to do several different tests which, done upon a single piece, could have otherwise affected each other.

Knowing in theory that it would be hard to achieve an overlay which was not diluted and yet kept the ductility of the welding wire, the welder was instructed to try different settings and see what worked best. However, the welder was informed that the best end result would theoretically be achieved if the substrate and overlay were to be melted just enough for them to bind; to that end, a low heat input was advised.

The Fe10Cr4Al welding wire was delivered from Kanthal AB in 50 one meter lengts of wire with a diameter of 2.5 mm. No cleaning or other preparations were needed before using it. Argon was used as shielding gas.

The welding procedure with which the welder achieved the best results with respects to coverage as well as weldability is described with the parameters in Table 3.1; welds were performed at similar heat input with varying results and can be seen in Appendix B. At lower heat inputs, the welder reported that the overlay material tended to only partially melt and stayed in globular forms on the surface of the

(24)

CHAPTER 3. PRACTICE

Current [A] Voltage [V] Speed [mm/min]

110 10.5 200

Table 3.1. Welding parameters of overlay welding test

substrate with resulting hardships in achieving even coverage. With an increase in heat input this problem was mitigated and the welder reported that the Fe10Cr4Al was easy to work with as a TIG welding wire.

An occular inspection of the finished overlay weld showed no defects.

Following the occular inspection, two 5 mm pieces of the welded bar was cut out using cold cutting so as not to let heat from the cutting affect the test piece. Since this experiment was interested in only the overlay layer and where it connected to the substrate the decision was made to remove a large part of the substrate from one of the pieces. This made the bend test easier to perform.

Figure 3.2. Part of welded test piece with most of the substrate removed so that approximately 4 mm of overlay and 4 mm of substrate remains.

The part that was left, see 3.2, was then put into a manual hydraulic press. The pressure was gradually increased while the two experimenters closely observed the test piece for any cracking. The other 5 mm piece were put into the SEM in order to look closer at the chemical composition of the overlay weld.

14

(25)

3.2.2 Welding Fe10Cr4al using Nikrothal^®PM58

The pieces of Fe10Cr4Al which were used in this experiment were not of exact geometry but were almost rectangle in shape and of an approximate size of 200x20 mm; the irregularity of the shape is reflected in the way that the weld curves in Figure 4.1 and 4.2. The Nikrothal^®PM58 consumable wires wire 50 cm in lengt and 2.5 mm in diameter. The two pieces of hot rolled Fe10Cr4Al were fixed, without separation, on a vice with clamps . A V-groove joint was then prepared manually with a grinder by the welder.

Figure 3.3. Prepared v-groove before second weld of Fe10Cr4al using Nikrothal^®PM58

The two pieces were welded together on both sides; after the first weld the pieces were turned over and a new V-groove joint was prepared before the second weld;

making the final joint a double V-joint. The preparation before the second weld can be seen in Figure 3.3 and the welding parameters for the two different welds can be read in Table 3.2.

Weld # \Parameter Current [A] Voltage [V] Speed [mm/min]

Weld 1 110 11 42,25

Weld 2 110 11 49,62

Table 3.2. Welding parameters for welding of Fe10Cr4al using Nikrothal^®PM58

After the ocular inspection the weld test was cut up in smaller pieces across the weld which were numbered and photographed for future reference.

(26)

CHAPTER 3. PRACTICE

Figure 3.4. The finished weld of Fe10Cr4al using Nikrothal^®PM58 cut up into smaller pieces

The bend test of the smaller pieces were executed in the manual hydraulic press.

The pressure, and thus angle of bend, was gradually increased while observing the test piece for any cracking. A successful test would be a bend to a 180 degree angle, however the bend test setup was in such a way that this was not possible. The test was therefore stopped before failure or 180 degrees.

16

(27)

Figure 3.5. Bend test of Fe10Cr4Al welded together with Nikrothal^®PM58

The bended piece of the weld was moved from the hydraulic press to a vise in order to further test the ductility. The result of which can be seen in section 4.

To test the material properties of the weld using a make shift bend test in a vice is not ideal. In this instance it was the only available way of further exploring the ductility of the weld but future tests should be prepared to utilize testing rigs with a higher degree of control.

(28)

(29)

Chapter 4

Results

The results of the three different weld experiments will be presented in this chapter.

4.1 Welding Fe10Cr4Al+RE using Nikrothal

^®

PM58

The welding of the two pieces of Fe10Cr4Al+RE with Nikrothal^®PM58 consumables according to the welding parameters seen in Table 3.2 was performed without noticeable difficulties according to the welder.

During the ocular inspection after the weld had cooled down, only one defect could be observed. That defect was that a part of the second weld had sunk down.

This is most likely due to the welders technique and not the materials.

The finished weld are seen in Figures 4.1 & 4.2 where the sunken portion is between the 250 and 300 marks on the ruler in Figure 4.2 (Not easily seen in the picture).

The curvature of the weld pass is due to the geometry of the pieces of Fe10Cr4Al.

Another defect was identified later when the weld piece was cut into smaller pieces (see Figure 3.4); a root cavity which seemed to follow most of the weld root. The root cavity is clearly seen in Figure 4.3.

(30)

CHAPTER 4. RESULTS

Figure 4.1. First weld of Fe10Cr4Al welded together with Nikrothal^®

20

(31)

4.1. WELDING FE10CR4AL+RE USING NIKROTHAL^®PM58

Figure 4.2. Second weld of Fe10Cr4Al welded together with Nikrothal^®

(32)

CHAPTER 4. RESULTS

4.1.1 Bend test

The bend test was carried out in a manual hydraulic press around a round work piece with a radius of 15 mm. The pressure, and thus angle of bend, was gradually increased while searching for signs of cracking in the test piece. Due to the relative small size of the test piece, the test could not be done until 180 degrees of bending.

The bend test was abandoned at an α-angle of 34 degrees and the piece can be seen in Figure 4.3.

Figure 4.3. The Fe10Cr4Al welded with Nikrothal^® test piece bent to an angle of 34 degrees around 15 mm radius

Following SS-EN ISO 5173:2010 and approximating that the extension of the material to be higher than 20 % would mean that the radius of the mandrel should be

22

(33)

4.1. WELDING FE10CR4AL+RE USING NIKROTHAL^®PM58

10 mm. The test piece was put into a vice and further bending was accomplished to approximately 180 degrees over a radius of 9 mm. The result can be seen in Figure 4.4.It is worth noting that this is the first time ever that Fe10Cr4Al has been welded without loosing ductility and without any pre- or post-heat treatment of the weld.

Figure 4.4. Extra bend of the Fe10Cr4Al welded with Nikrothal^® test piece to approximately 180 degrees in a vise with curvature corresponding to a radius of 9 mm

(34)

CHAPTER 4. RESULTS

4.2 Overlay welding of Fe10Cr4Al+RE on austenitic stainless substrate

The overlay welds did not crack during cooling as was expected to happen; this was attributed to a mixing between the ferritic overlay layer and the austenitic substrate. The corrosion resistance of the Fe10Cr4Al+RE should be intact with levels of the Ni kept below 3 or 4 wt. %. [26]

Looking at the results from the SEM, seen in Appendix C, the chemical composition of the overlay weld held around 1 wt. % of Ni. 1 wt. % of Ni in the overlay layer indicates a dilution of approximately 2

Figure 4.5. Overlay weld of Fe10Cr4Al+RE put into the hydraulic press for bend testing. Circled in red is a crack that formed after only a few degrees of bend.

During the bend test the test piece cracked from the top within a few degrees of bend, see Figure 4.5. Since both theory and practice showed that the Fe10Cr4Al+RE could not be used as an overlay without loosing almost all ductility, further testing was aborted.

24

(35)

4.3. WELDING FE10CR4AL+RE USING AFA-A5-8

4.3 Welding Fe10Cr4Al+RE using AFA-A5-8

The weld cracked on multiple places during the first 30 minutes of cooling after it was finished. There was therefore no possibility or need for further testing; there was also a limited amount of the AFA-A5-8 which made extensive welding impossible.

(36)

(37)

Chapter 5

Findings and Discussion

It was known theoretically from the start that it is hard to weld a ferritic stainless steel without adding Ni as well as pre- and post heat treatment. More specifically, the Fe10Cr4Al is so brittle that after casting it usually cracks during cooling without any outside forces acting on it. One could therefore not realistically expect to achieve an overlay weld, using this material, with the ductility that is needed to withstand a bend test until 180 degrees of bend; thus not being able to use it in nuclear energy applications.

The starting point for this thesis was, however, to experiment with manual TIG overlay welding to see if there was any indication that it could work. This thesis has found no such indication and concludes that it is not possible to perform an overlay weld of Fe10Cr4Al on top of an austenitic stainless steel which could then pass a nuclear application standard bend test.

Even if this overlay weld still is without sufficient ductility for the nuclear power industry, there are other areas which does not have the same demands.

One such example is the Nextower[27] project which is developing CSP with thermal energy storage. The liquid lead is used for storing thermal energy generated by the CSP. With the CSP use case in mind this thesis has found it possible to manually clad austenitic stainless steel (i.e. AISI 304L) with Fe10Cr4Al+RE with low dilution and above zero ductility.

There are also reports from an experiment in Italy with mechanized overlay MIG welding using both pre- and post heat treatment.[26] They had succeeded in getting a good overlay weld which did not spontaneously crack.

The experimental welding of Fe10Cr4Al using Nikrothal^® PM58 as a consumable were also of a highly exploratory character and would in a best case scenario give some indication for future work. As such, the hot-rolled Fe10Cr4Al pieces used were not specifically manufactured for this work and we did not focus on standardizing the welding tests or bending tests; indications for future work was the goal.

It is worth noting that the final bending of the test piece was achieved to almost

(38)

CHAPTER 5. FINDINGS AND DISCUSSION

180 degrees with the presence of a root cavity. This is suggestive of a very high ductility in the weld. This is interesting for many applications both within nuclear power generation and in other fields.

28

(39)

Chapter 6

Future Work

In order to build on top of this work and come closer to increased usability for this new material I would suggest the following:

• Continue to look at the weldability of Fe10Cr4Al+RE together with austenitic consumables with different Ni-content.

• Aging test at different temperatures of the above mentioned welds.

• Run appropriate corrosion studies on such welds.

• Further experiments of manual overlay TIG welding of Fe10Cr4Al+RE on top of austenitic stainless steels.

The welds with Fe10Cr4Al+RE with Nikrothal^® PM58 showed good promise but a more thorough investigation needs to be done in order to establish a procedure through standardized testing. To do this one would need to try varying welding parameters and test the results. If there are other potential candidates for consumables, such as AFA materials, these could be done with similar experimental setups as for the Nikrothal^® PM58.

The Fe10Cr4Al+RE was produced in order to handle the extreme corrosive environment of liquid Pb at higher temperatures. With this in mind, it is crucial to keep testing successful welds in such environments.

This thesis was geared towards nuclear energy applications, however there are other applications for the Fe10Cr4Al as cladding material. It could be interesting to further explore the usability of this alloy withing such applications which does not have the same high demands for ductility.

(40)

(41)

Bibliography

[1] https://climate.nasa.gov/causes/

[2] https://www.eea.europa.eu/signals/signals-2017/articles/energy-and-climate- change

[3] https://new.engineering.com/story/is-nuclear-power-a-solution-for-climate- change

[4] M. Joyce. Nuclear Engineering, A Conceptual Introduction to Nuclear Power.

Butterworth-Heinemann, 2017: Chapter 11, Advanced reactors and future con- cepts.

[5] International Atomic Energy Agency. LIQUID METAL COOLANTS FOR FAST REACTORS COOLED BY SODIUM, LEAD, AND LEAD-BISMUTH EUTECTIC. Austria, 2012 (p. 29).

[6] J. Ejenstam. Corrosion resistant alumina-forming alloys for lead-cooled fast re- actors. Doctoral Thesis in Chemistry, Royal Institute of Technology, Stockholm, Sweden 2015.

[7] H. von Kantzow Fire-resistant alloy with high electric resistance US patent no.

1717284, 1929

[8] J. Wallenius, et al.. Design of SEALER, a very small lead-cooled reactor for commercial power production in off-grid applications. Nuclear Engineering and Design 338, 2018 (pp. 23?33).

[9] http://www.h2020-nextower.eu/

[10] https://www.praxairdirect.com/Industrial-Gas-and-Welding-Information- Center/Technical-Data/technical-terms-glossary.html

[11] Siu Wah Wai, et al.. A study of high temperature cracking in ferritic stainless steels. Materials Science and Engineering A158, 1992 (pp. 21-30).

[12] F. H. Stott, et al.. The Influence of Alloying Elements on the Development and Maintenance of Protective Scales. Oxidation of Metals 44, 1995 (pp. 113-145).

(42)

BIBLIOGRAPHY

[13] J. Ejenstam, et al.. Microstructural stability of Fe-Cr-Al alloys at 450-550^◦C. Journal of Nuclear Materials 457, 2015 (pp. 291-297).

[14] B. Jönsson, A. Westerlund. Oxidation Comparison of Alumina-Forming and Chromia-Forming Commercial Alloys at 1100 and 1200^◦C. Oxidation of Metals 88, 2017 (pp. 315-326).

[15] Outokumpu. Handbook of Stainless Steel. Outokumpu Oyj, 2013.

[16] ISO 15614-7:2016. Specification and qualification of welding procedures for metallic materials ? Welding procedure test ? Part 7: Overlay welding

[17] K. Weman. Karlebo Svetshandbok. Liber AB, Stockjolm, 2013.

[18] ISO 5173:2009. Destructive tests on welds in metallic materials - Bend tests [19] J. R. Davis. ASM Speciality Handbook Heat-Resistant Materials. ASM Inter-

national, 1997: Properties of Superalloys.

[20] https://sv.wikipedia.org/wiki/Gasvolframsvetsning

[21] A. L. Schaeffler. Constitution diagram for stainless steel weld metal.. Metal Progress 56, 1949 (p. 680)

[22] https://www.welding-alloys.com/uploads/pdf/tool-box/schaeffler- diagram.pdf

[23] https://www.twi-global.com/technical-knowledge/job-knowledge/welding-of- ferritic-martensitic-stainless-steels-101

[24] https://www.thyssenkrupp-materials.co.uk/stainless-steel-304l-14307.html [25] https://www.marlinwire.com/blog/difference-between-grade-304-and-304l-

stainless-steel

[26] Personal communications with P. Szakalos [27] https://www.h2020-nextower.eu/

[28] https://www.gemma-project.eu/

[29] A. Weisenburger et. al.. IAEA Conference 2019: Technical Meeting on Structural Materials for Heavy Liquid Metal Cooled Fast Reactors. https :

//conf erences.iaea.org/event/205/contributions/15867/attachments/8275/10922/ID−

21 − W EISENBURGER − Developmentof_alumina_forming_weisenburger.pdf

32

(43)

Appendix A

Schaeffler

A.1 Calculations

Cr.eq(wt.%) = %Cr + 3 · %Al + %Mo + 0.5 · %Nb (A.1) N i.eq(wt.%) = %Ni + 30 · %C + 0.87 · %Mn (A.2)

Fe10Cr4Al+RE(A.1): Cr.eq(wt.%)F e10Cr4Al+RE = 9.99 + 3 · 4 = 21.99 Fe10Cr4Al+RE(A.2): Ni.eq(wt.%)F e10Cr4Al+RE = 30 · 0.08 + 0.87 · 0.12 = 2.5 Nikrothal^®PM58(A.1): Cr.eq(wt.%)N ikrothal®P M 58 = 19 + 3 · 5 = 34

Nikrothal^®PM58(A.2): Ni.eq(wt.%)N ikrothal®P M 58 = 57.6

AFA-A5-8(A.1): Cr.eq(wt.%)AF A−A5−8 = 16 + 3 · 3.14 + 0.5 · 1.19 = 26.02 AFA-A5-8(A.2): Ni.eq(wt.%)AF A−A5−8= 23 + 30 · 0.01 = 23.3

(44)

APPENDIX A. SCHAEFFLER

A.2 Plots

Figure A.1. Schaeffler with aluminum plot of 50 % dilution weld between Fe10Cr4Al+RE and AFA-A5-8

Figure A.2. Schaeffler with aluminum plot of 50 % dilution weld between Fe10Cr4Al+RE and Nikrothal^®PM58

34

(45)

Appendix B

Overlay welds

Test piece # \Parameter Current [A] Voltage [V] Speed [mm/min]

1 90 9.5 82.2

2 75 9 85.7

3 100 10.5 150.0

4 110 10.5 200.0

Table B.1. Welding parameters of the first overlay welds as seen in Figure B.1.

Voltage is an estimate since it fluctuates during welding and the welder could not properly register it during welding.

(46)

APPENDIX B. OVERLAY WELDS

Figure B.1. Overlay welding with different parameters using Fe10Cr4Al+RE wire on top of austenitic stainless steel. Test pieces number 1 to 4 from left to right;

number 4 was chosen for further testing with a secondary layer made before bending.

36

(47)

Appendix C

SEM analysis

The results of the SEM analysis is presented in the following pages. The first sample "BOCKAD" refers the bended test piece taken from the weld of Fe10Cr4Al+RE together with Nikrothal^®PM58 which can be seen in Figure 4.4.

After that is the second sample "EJ BOCKAD" wich referns to the sample piece taken from the overlay weld of Fe10Cr4AL+RE on top of 304L. Another sample from the same weld can be seen in Figure 4.5.

(48)

SVETS BOCKAD

5/15/2020 3:07:01 PM

Processing option : All elements analysed (Normalised)

Spectrum In stats. Al Si Cr Fe Ni Total

Spectrum 1 Yes 4.32 0.55 10.66 84.48 100.00

Spectrum 2 Yes 3.46 5.67 12.65 58.22 20.00 100.00 Spectrum 3 Yes 5.38 14.91 52.04 27.66 100.00

Spectrum 4 Yes 4.35 0.44 10.82 84.38 100.00

Spectrum 5 Yes 4.34 13.84 60.44 21.38 100.00

Max. 5.38 5.67 14.91 84.48 27.66

Min. 3.46 0.44 10.66 52.04 20.00

All results in weight%

(49)

SVETS

5/15/2020 3:11:27 PM

Spectrum In stats. Al Si Cr Fe Ni Total

Line Spectrum(1) Yes 4.24 14.09 59.45 22.22 100.00 Line Spectrum(2) Yes 5.48 16.96 35.31 42.24 100.00

Line Spectrum(3) Yes 4.43 10.73 84.84 100.00

Line Spectrum(4) Yes 4.82 14.15 56.28 24.74 100.00 Line Spectrum(5) Yes 4.72 0.78 14.48 54.32 25.70 100.00

Max. 5.48 0.78 16.96 84.84 42.24

Min. 4.24 0.78 10.73 35.31 22.22

(50)

SVETS

5/15/2020 3:12:00 PM

Project: SVETS Owner: Operator Site: Site of Interest 2

Sample: BOCKAD Type: Default ID:

(51)

SVETS

5/15/2020 3:19:09 PM

Spectrum In stats. Al Ti Cr Fe Ni Total

Line Spectrum(2) Yes 3.55 0.68 11.67 81.62 2.48 100.00 Line Spectrum(3) Yes 3.99 14.58 57.48 23.96 100.00 Line Spectrum(4) Yes 4.75 14.11 55.05 26.09 100.00 Line Spectrum(5) Yes 4.57 14.39 55.68 25.35 100.00

Max. 4.75 0.68 14.58 84.95 26.09

Min. 3.55 0.68 10.94 55.05 2.48

(52)

SVETS

5/15/2020 3:19:13 PM

(53)

SVETS

5/15/2020 3:26:16 PM

Too many elements to fit on page ( 9 max. ) Please consider using the copy to clipboard function.

(54)

SVETS 5/15/2020 3:27:08 PM

Spectrum In stats. N O Mg Al Si Ti Cr Fe Ni Rb Zr Hf Total

Spectrum 1 Yes 4.71 12.26 73.52 9.51 100.00

Spectrum 2 Yes 3.99 0.52 14.47 57.11 23.92 100.00

Spectrum 3 Yes 25.90 59.39 0.62 2.01 12.08 100.00

Spectrum 4 Yes 46.82 39.49 11.49 0.17 0.64 1.39 100.00

Spectrum 5 Yes 8.93 1.54 2.98 7.86 26.36 11.50 3.24 21.60 15.98 100.00

Spectrum 6 Yes 1.55 22.64 12.46 46.95 16.40 100.00

Spectrum 7 Yes 5.57 2.59 2.06 4.84 9.24 34.26 14.16 24.93 2.34 100.00

Max. 25.90 46.82 39.49 4.71 22.64 59.39 14.47 73.52 23.92 3.24 24.93 15.98

Min. 5.57 2.59 39.49 1.54 0.52 2.98 0.17 0.64 9.51 3.24 1.39 2.34

(55)

SVETS

5/15/2020 3:40:14 PM

Too many elements to fit on page ( 9 max. ) Please consider using the copy to clipboard function.

(56)

SVETS 5/15/2020 3:40:20 PM

Spectrum In stats. O Al Si Ti Cr Fe Ni Zr Nb Hf Ta Total

Spectrum 1 Yes 5.74 15.65 42.98 35.64 100.00

Spectrum 2 Yes 4.27 0.35 15.75 47.09 32.53 100.00

Spectrum 3 Yes 3.93 12.67 75.38 8.02 100.00

Spectrum 4 Yes 1.33 3.94 3.49 10.35 80.90 100.00

Spectrum 5 Yes 2.70 1.88 39.22 10.43 28.03 17.74 100.00

Spectrum 6 Yes 3.30 1.17 3.26 4.75 12.24 8.93 10.47 29.85 7.99 18.05 100.00

Spectrum 7 Yes 4.67 3.16 21.84 8.85 60.83 0.65 100.00

Spectrum 8 Yes 1.09 2.14 14.20 10.91 68.22 3.43 100.00

Max. 4.67 5.74 39.22 3.26 15.75 80.90 35.64 10.47 29.85 7.99 18.05

Min. 1.09 1.17 0.35 3.26 4.75 12.24 0.65 10.47 29.85 7.99 18.05

(57)

SVETS EJ BOCKAT

5/15/2020 2:36:51 PM

Spectrum In stats. Al Si Cr Mn Fe Ni Total

Line Spectrum(1) Yes 3.62 1.01 11.99 81.86 1.51 100.00 Line Spectrum(2) Yes 3.64 0.47 11.82 83.01 1.07 100.00 Line Spectrum(3) Yes 3.75 0.47 11.97 82.34 1.47 100.00 Line Spectrum(4) Yes 1.28 15.76 0.89 75.68 6.40 100.00

Line Spectrum(5) Yes 1.26 17.28 75.36 6.10 100.00

Max. 3.75 1.01 19.02 0.89 83.01 9.13

Min. 1.26 0.47 11.82 0.89 71.86 1.07

(58)

SVETS EJ BOCKAT

5/15/2020 2:40:21 PM

Sample: EJ BOCKAD Type: Default ID:

(59)

SVETS

5/15/2020 2:43:57 PM

Spectrum In stats. Al Cr Fe Ni Total

Spectrum 1 Yes 3.67 12.09 83.04 1.21 100.00

Mean 3.67 12.09 83.04 1.21 100.00

Std. deviation 0.00 0.00 0.00 0.00

Max. 3.67 12.09 83.04 1.21

Min. 3.67 12.09 83.04 1.21

(60)

SVETS

5/15/2020 2:46:01 PM

Spectrum In stats. Al Si Ti Cr Fe Ni Total

Spectrum 1 Yes 3.67 12.09 83.04 1.21 100.00

Line Spectrum(1) Yes 3.70 1.48 11.91 81.85 1.06 100.00 Line Spectrum(2) Yes 3.53 0.47 12.19 82.68 1.12 100.00 Line Spectrum(3) Yes 3.82 0.57 11.91 82.60 1.10 100.00

Max. 3.82 1.48 0.57 12.19 83.04 1.21

Min. 3.53 0.47 0.57 11.91 81.85 1.06

(61)

SVETS

5/15/2020 2:46:30 PM

Sample: EJ BOCKAD Type: Default ID:

(62)

Manual TIG welding of Fe10Cr4Al+RE: overlay on an austenitic substrate and butt weld using alumina forming austenite consumable