See discussions, stats, and author profiles for this publication at: [Link]

net/publication/343271279

Dross Formation in Aluminum Melts During the Charging of Beverage Can

Scrap Bales with Different Densities Using Various Thermal Pretreatments

Article  in  JOM: the journal of the Minerals, Metals & Materials Society · July 2020
DOI: 10.1007/s11837-020-04268-4

CITATIONS READS

0 81

3 authors, including:

Jan Steglich Bernd Friedrich

TRIMET Aluminium SE RWTH Aachen University
12 PUBLICATIONS   6 CITATIONS    804 PUBLICATIONS   3,919 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Low-emission synthesis of titanium alloys View project

GetIn Cicero View project

All content following this page was uploaded by Jan Steglich on 11 October 2020.

The user has requested enhancement of the downloaded file.

Dross Formation in Aluminum Melts
During the Charging of Beverage Can
Scrap Bales with Different Densities Using
Various Thermal Pretreatments

Jan Steglich, Bernd Friedrich & Marcel

Rosefort

JOM
The Journal of The Minerals, Metals &
Materials Society (TMS)

ISSN 1047-4838
Volume 72
Number 10

JOM (2020) 72:3383-3392

DOI 10.1007/s11837-020-04268-4

1 23
Your article is protected by copyright and all
rights are held exclusively by The Minerals,
Metals & Materials Society. This e-offprint is
for personal use only and shall not be self-
archived in electronic repositories. If you wish
to self-archive your article, please use the
accepted manuscript version for posting on
your own website. You may further deposit
the accepted manuscript version in any
repository, provided it is only made publicly
available 12 months after official publication
or later and provided acknowledgement is
given to the original source of publication
and a link is inserted to the published article
on Springer's website. The link must be
accompanied by the following text: "The final
publication is available at [Link]”.

1 23
 Author's personal copy
JOM, Vol. 72, No. 10, 2020
[Link]
 2020 The Minerals, Metals & Materials Society

ALUMINUM: RECYCLING AND CARBON / ENVIRONMENTAL FOOTPRINT

Dross Formation in Aluminum Melts During the Charging

of Beverage Can Scrap Bales with Different Densities Using
Various Thermal Pretreatments

JAN STEGLICH ,1,3 BERND FRIEDRICH,2 and MARCEL ROSEFORT1

1.—TRIMET Aluminium SE, Aluminiumallee 1, 45356 Essen, Germany. 2.—IME Process

Metallurgy and Metal Recycling, RWTH Aachen University, Intzestr. 3, 52056 Aachen, Germany.
3.—e-mail: [Link]@[Link]

Herein, dross formation (D) in used beverage can (UBC) scrap bales was
investigated in laboratory-scale experiments to understand the reactions of
solid can coating residues with aluminum melts. The experimental conditions
were chosen according to a multi-chamber furnaces process, in which thermal
pretreatment and submerged melting of the scrap bales are combined. The
thermal pretreatment results show that the supply of oxygen should be con-
tinuously adjusted to meet the oxygen demand of UBC scrap bales for ther-
molysis. Moreover, it is necessary to maintain the UBC scrap bales at an
effective material temperature of 550–570C for at least 30 min. It is shown
that adhering carbon-containing pyrolysis char and oxidized can sheet sur-
faces decrease the coalescence of molten scrap, resulting in dross with a high
metal fraction. Further melting experiments with UBC scrap bales of different
densities (q) show that by lowering the q, the dross formation reduces.

INTRODUCTION AND MOTIVATION LITERATURE REVIEW

Aluminum beverage cans are widely used, and the Can Coating Removal and Residue
can market is continuously growing. After the use
Waterborne epoxy, polyester, and acrylic polymer
and disposal of these cans, the valuable aluminum
resins are widely used as beverage can coatings.
metal must be recycled to produce new cans to satisfy
Bisphenol-a-diglycidyl ether (BADGE)-based epoxy
the requirements of an eco-friendly circular economy.
resins are examples of the commonly used coatings
The metallurgical recycling of used beverage can
in the can industry. The curing temperature of
(UBC) scrap is a combined process of thermal pre-
waterborne resins is in the range of 150–200C
treatment used to remove the can coating and
depending on the resin-cure system and time.1,2
melting. The melting of scrap still poses a challenge
When the curing temperature of a thermosetting
to aluminum recyclers as it leads to metal loss and
polymer exceeds this range, the cross-linked poly-
undesirable dross formation (D). Open questions are
mer network starts to degrade at the crosslinking
how the organic content, coating residues, and UBC
bonds.3 The decomposition starts with degassing
scrap bale density (q) affect D.
and the formation of pores in the coating. Inorganic
As the dross formation reactions are quite com-
pigments and fillers are retained as they do not form
plex, this investigation focuses only on the can
volatile products. At elevated temperatures between
coating and its residues in UBC scrap bales after
300C and 400C, different reaction products are
pretreatment. The influence of oxygen in the com-
formed depending on the oxygen concentration in
mon furnace atmosphere was excluded by perform-
the pretreatment atmosphere.4 In industrial rever-
ing the laboratory melting experiments under an
beratory furnaces, the ratio of available oxygen and
inert atmosphere.
required oxygen for stoichiometric combustion is
described by the lambda ratio (k), which is calcu-
(Received April 14, 2020; accepted June 30, 2020; lated according to Eq. 1:5
published online July 28, 2020)

3383
 Author's personal copy
3384 Steglich, Friedrich, and Rosefort

available oxygen EXPERIMENTAL PROCEDURES

k¼ ð1Þ
stoichiometric required oxygen Material Description
The amount of oxygen calculated using k is Melting experiments were performed with com-
transferred to the thermal pretreatment system mercial UBC scrap bales to obtain D results using
and used to describe the total oxygen availability in industrially relevant materials. All UBC scrap bales
a reactor vessel with respect to the oxygen demand had the initial dimensions of approximately
of the can coating. The ratio, k, has significant 400 mm 3 300 mm and a variable height, depend-
effects on coating removal and residues, which are ing on their degree of compaction. A description of
summarized in Table I. the various properties and impurities of UBC scrap
Isothermal annealing experiments by Ref. 3 show bales is provided in Table II.
that thermolysis conditions, i.e., 500–600C temper- The average oxygen and carbon concentrations in
ature and an oxygen atmosphere, are necessary to the initial UBC scrap bales were measured by
successfully remove BADGE-based can coatings. drilling out 12-g chips out of each scrap type and
The pyrolysis of a similar coating leads to the analyzing them by the carrier gas method using the
formation of adhering carbon coating char on the LECO TC600 and LECO CS844 instruments,
can sheets.7 respectively. Herein, nine spot samples of each
scrap type were analyzed to estimate the inhomo-
Dross Formation D and Can Alloy Oxidation geneity. UBC C was compacted into three samples
D is attributed to the oxidation of metals and of different q, i.e., C1 dense (1110 kg/m3), C2
inclusion of liquid aluminum alloys in correspond- medium (808 kg/m3), and C3 low (685 kg/m3), in
ing oxides. The oxidized metal cannot be recovered an industrial stamping press.
in the current industrial processes and is defined as
metal loss, whereas the trapped alloy can be recov- Thermal Pretreatment
ered by dross processing.8 The metal oxidation of UBC A, B, and C were cut into sections with a
UBCs starts during thermal pretreatment after the thickness of 40–60 mm, height of ca. 300 mm, and
coating is partially removed. Aluminum beverage mass of 1095 ± 127 g. The thermal pretreatment of
cans are produced from EN AW-5182 can end sheets UBC sections was performed in an airtight steel
and EN AW-3104 can body sheets or similar alloys.9 cylinder (height: 445 mm; diameter: 260 mm) with
The major alloying elements are Al, Mg, and Mn. It purge gas and off-gas connections in a resistance-
is generally known that Mg increases the oxidation heated furnace. Argon, argon with 4 vol.% O2, and
rate of aluminum alloys because of the formation of air were used as purge gases at atmospheric
Al2MgO3 and MgO instead of just Al2O3 as in the pressure and a constant gas flow rate of 10 l/min.
case of pure aluminum.9,10 Annealing experiments The UBC slices were charged before starting the
with uncoated can end sheets show that the end procedure and removed after cooling to maintain a
sheets oxidize significantly faster (within 7–10 min controlled atmosphere during pretreatment. The
above 570C, but not at 550C) than the can body resistance heating resulted in an average heating
sheets.9 ramp of 5C/min and a plateau at the target
Once the metal surface is oxidized, the coales- temperature of 450 ± 5C or 550 ± 5C. The reactor
cence during melting is impeded, and dross is temperature was measured using an internal ther-
formed. Submerged melting is necessary to limit mocouple. The target temperature was held for
the exposure of the metal surface to the hot furnace 30 ± 5 min before cooling down the samples. Every
atmosphere.8,9 Experiments by Rossel11 show that experiment was repeated three times to generate
increasing the specific surface area of scrap by enough mass for subsequent melting.
reducing the sheet thickness directly leads to Pretreatment conditions were chosen to represent
increased D. the pyrolysis and thermolysis conditions with two
different target temperatures. The low thickness of
the UBC slices and the low heating rate led to a
homogeneous temperature distribution in the

Table I. Definition of k for different thermal pretreatments and the resulting reaction products

Idealized reaction products

Pretreatment condition Lambda ratio5 Gaseous6 Solid4

Combustion, exothermic k‡1 CO2 and H2O Metal oxides and inorganic pigments
Thermolysis, endothermic 0< k £ 1 CO, CxHy, H2O, and CO2 Inorganic pigments
Pyrolysis, endothermic k=0 CxHy Inorganic pigments and solid carbon residue
 Author's personal copy
Dross Formation in Aluminum Melts During the Charging of Beverage Can Scrap Bales with 3385
Different Densities Using Various Thermal Pretreatments

Table II. Properties and impurities of UBC scrap bales

Property Unit UBC A UBC B UBC C (1–3) Analysis method

Main impurity – Coating Coating, plastic, iron, copper Coating, plastic Visual
Density kg/m3 450 910 1110, 808, and 685 Weighing
Volatile organic wt.% 2.8 ± 0,3 8.4 ± 4.0 10.0 ± 2.6 Annealing at 450–550C
Moisture wt.% 0.9 ± 0,7 1.7 ± 1.1 0.9 ± 0.7 Drying at 105C
Porosity % 83 66 59, 70, and 75 Measuring

samples, assumed to be equal to the reactor

temperature. dross mass
Dross formation D in % ¼  100 ð2Þ
dry scrap mass
Melting of UBC Scrap Bales
After pretreatment, the UBC slices were carefully dross mass  metal mass in dross
NMP in % ¼  100 ð3Þ
cut into 100 ± 1 g cubes such that the original pore dross mass
structure of the initial bales was retained. Samples
were also cut out of the initial UBC bales for Metal yield g in % ¼
reference experiments without heat treatment. The  
metal mass in drossþrecovered metal mass ð4Þ
experiments were first repeated three times to 100
measure the dross and metal yield (g) and then a dry scrap mass
fourth time to analyze the solidified dross/metal Note that the g is the sum of the metal reclaimed
interface. by UBC melting and the metal fraction recovered of
The UBC samples were submerged into dross remelting (mM). Therefore, it specifies the
1000 ± 5 g pure aluminum heated in a SiO2/Al2O3 aluminum metal that has been recovered by both
crucible by an induction furnace to 750C under an processes. The metal yield is unequal to the metal
argon atmosphere to exclude reactions with oxygen. recovery, setting the recovered aluminum mass into
A steel cylinder with a removable lid and argon relation just with the metal input.
injection was placed on the crucible before starting
the melting process. The lid was designed to carry a RESULTS
UBC sample carrier to push the sample into the
melt or a dross skimming tool for a uniform Thermal Pretreatment
skimming procedure. The setup is schematically The main influence of thermal pretreatment was
shown in Ref. 7. After melting the scrap and an the formation of an adhering carbon-containing
additional 30-min reaction (35 min total including coating residue on the can sheets by pyrolysis when
melting), the dross was skimmed and cooled in argon was used (k = 0). This residue was visible on
argon. Additional experiments in ambient air were all scrap types at varying quantities depending on
performed without argon. the initial organic content. Thermolysis under argon
The UBC melting experiments were assessed by with 4 vol.% O2 atmosphere (0 < k < 1) leads to less
D, the formation of non-metallic products NMP, coating residue; however, the coating is still not
which can also be referred to as system metal loss, completely removed. Similar results were obtained
and the metal yield g. D is calculated using Eq. 2 by for the decoating experiments performed at 450C.
setting the dross mass into relation with the dry Therefore, all other decoating experiments were
scrap mass, which is also referred to as gross melt performed at a 550C material temperature. The
loss.13 As the exact amount of metal in a commercial incomplete coating removal was evaluated by com-
UBC scrap sample is unknown before any treat- paring the remaining carbon and oxygen concen-
ment, a true melt loss cannot be calculated. Even trations of the scrap with the respective initial
though the true melts loss would be more precise to concentrations.
describe the recovered aluminum metal, the D is The use of an air atmosphere (k ‡ 1) led to an
more practical and widely used by the aluminum almost complete decoating of UBC A with only very
recycling industry.13 The NMP was determined by little oxidation, whereas severe oxidation of UBC B
remelting one-third of the dross in an equimolar salt under comparable process conditions was observed.
flux of NaCl:KCl with 5 wt.% cryolite (Na3AlF6) and Further investigations showed that a higher organic
subsequently weighing the metal and non-metallic content (UBC A: 2.8 ± 0.3% versus UBC B
contents. The three criteria are calculated according 8.4 ± 4.0%) triggered an exothermic combustion
to Eqs. 2–4.
 Author's personal copy
3386 Steglich, Friedrich, and Rosefort

reaction in the pretreatment reactor, resulting in The comparison results of different scrap types
extensive metal oxidation and dross formation in with q < 750 kg/m3 in Fig. 1 obviously show that
subsequent melting experiments. After this obser- the stoichiometric thermolysis lowered the D to a
vation, the oxygen supply during de-coating was minimum value as compared to that of the initial
adjusted, and the experiments were repeated with scrap. Based on the previously measured carbon
UBC C. and oxygen concentrations in the scrap, a lower q
The oxygen demand of UBC C was roughly promotes the removal of organics and the effective-
calculated using the organic mass of the samples, ness of thermal pretreatment. As a result, low-
the molecular formula of epoxy, polyester, and PET density scrap has fewer contaminants and results in
monomers, and a simplified combustion calculation less D. A lower q also lowers D in the raw scrap
for thermolysis and combustion conditions. The without thermal pretreatment; however, the slope
calculation yielded the oxygen demand of 5–8 mol of the linear best-fit line is not very steep. As the
O2 per kg UBC scrap, which was supplied by argon atmosphere prevented the combustion of
ambient air into the reactor at 10 l/min over the organic contaminants during melting, the organics
reaction time of 50–60 min, following the ideal gas were removed as volatiles.
law. Herein, the balance of oxygen demand and
supply over the process time is defined as stoichio- g of UBC scrap bales
metric thermolysis. These conditions led to the best
The g of UBC A and B was not improved by the
decoating while simultaneously reducing the metal
applied thermal pretreatment. Unfortunately, the
oxidation.
melting of the initial scrap samples resulted in the
Note that the stoichiometric thermolysis condi-
highest g. Initially, the static process parameters
tion (k  1) was only fulfilled over the thermal
were not adjusted according to the scrap bales and
reaction time of 50–60 min. The limited but contin-
led to incomplete pretreatment or metal oxidation.
uous oxygen supply resulted in substoichiometric
Subsequently, the protective argon atmosphere
conditions in the reactor as most of the oxygen
used in the melting experiments prevented metal
reacted with organics. Supplying all the oxygen at
loss in the untreated scrap. This unexpected result
once into the heated reactor would result in an
is discussed in ‘‘Understanding the Organic Content
exothermic combustion reaction and metal loss.
as a Heating Value’’ section.
This approximation should be further developed in
For UBC C, stoichiometric thermolysis conditions
future research.
were applied. The g results are shown in Fig. 2.
It is summarized that industrial thermal pre-
After adjustment, the thermolysis pretreatment
treatment should not be operated under static
conditions simultaneously maximized the g of the
process conditions; however, the oxygen supply
scrap and minimized D (see Fig. 2). The theoretical
should be continuously adjusted to meet the oxygen
g of UBC C was 87.4% based on the measured
demand of the scrap. Therefore, the organic content
organic content of up to 12.6%. This g was achieved
of the scrap in the reactor should be determined.
in one of the samples with a medium q after
Additionally, the treatment time needs to be
thermolysis at 550C in air (thermo mid). Even
adjusted according to the q, gas flow, and degrada-
though the theoretical metal yield cannot be met in
tion kinetics of different organic contaminants for
industrial processes, the results show that thermol-
successful decoating.
ysis maximizes the performance of submerged scrap
melting and subsequent dross processing.
Impact of q on D
An increasing g trend with decreasing q is also
The D results of UBC C1–C3 are plotted versus observed in Fig. 2 for all pretreatment conditions. If
their q in Fig. 1. The results are displayed in groups UBCs have to be recycled as bales, the q should be as
based on the pretreatment. A linear best-fit line was low as possible to minimize metal loss and as high as
plotted through the results, and the correlation required for economic transportation and storage.
coefficient rxy used to rate how well the variables q
and D fit to the trend. Finally, the D results of UBC
A and B were added. DISCUSSION
The results in Fig. 1 highlight that pyrolysis leads Dross Formation Mechanisms
to a constantly high D over the entire range of q. As
previously shown, pyrolysis leads to a film of In an industrial multi-chamber furnace, scrap
adhering carbon-containing coating residue, lower- bales are pushed into the melt after thermal
ing the coalescence of molten sheets in the metal. pretreatment. All remaining coating residues are
Therefore, D is constant with respect to the can submerged with the scrap into the melt. In contrast
sheet surface area. Compared to the case of to a reverberatory hearth furnace, the scrap does
pyrolyzed scrap, D in the initial beverage can not melt in the hot burner atmosphere, but is
increases with q. submerged in the melt.
 Author's personal copy
Dross Formation in Aluminum Melts During the Charging of Beverage Can Scrap Bales with 3387
Different Densities Using Various Thermal Pretreatments

Fig. 1. Dross formation (D) results of UBC as well as comparison of the linear dependence of scrap density q with D and the corresponding
correlation coefficient rxy.
 Author's personal copy
3388 Steglich, Friedrich, and Rosefort

Fig. 2. Metal yield g of UBC C with the bale densities of 1110 kg/m3 (dense), 808 kg/m3 (mid), and 685 kg/m3 (low), without pretreatment (raw),
after pyrolysis in argon at 550C (pyro), and stoichiometric thermolysis in air at 550C (thermo).

The melting of untreated UBC scrap bales with D and maximized the g. Only fractured oxide films
coating led to large amounts of dross. This dross had of beverage can sheets were detectable as non-
a high metal content with honey-like viscosity. Even metallic inclusions in the dross layer, showing that
though there were no solid coating residues from coalescence was achieved. The dross formation
pretreatment in the initial scrap, the metal was mechanisms are summarized in Fig. 3 and
trapped in oxide films, generated by gaseous decom- explained in further detail in Ref. 7.
position products of the organic contamination.
Melting of scrap bales with pyrolysis pretreatment
increased the D compared to that of the untreated Dross Formation by Metal Inclusions Caused
material. This is caused by the adhering of pyr- by Solid Coating Residues
olyzed coating char residue to the metal sheets.
The influence of the above described solid coating
Consequently, pyrolysis leads to reinforced compos-
residues and oxide films on dross formation, caused
ite ‘‘bi-films’’,8 supporting the structure of scrap
by the thermal history of the scrap, is quantified in
bales. Melting of scrap bales after adjusting the
Fig. 4. The molten metal, trapped in the solidified
thermolysis pretreatment conditions minimized the

Fig. 3. Proposed dross formation mechanisms of coated can sheets with and without thermal pretreatment based on metallographic results.
Reprinted with permission from Ref. 7.
 Author's personal copy
Dross Formation in Aluminum Melts During the Charging of Beverage Can Scrap Bales with 3389
Different Densities Using Various Thermal Pretreatments

Fig. 4. Impact of trapped metal mM in oxide films on the dross formation D of UBC bale scrap compared to data of Ref. 11 determined via pilot-
scale experiments with sheet scrap.

dross, was recovered by remelting under salt flux. Understanding the Organic Content
This metal fraction mM is set with respect to the can as a Heating Value
sheet surface area and plotted on the x-axis. The
The melting of UBC A and B without pretreat-
according D results of UBC A and C are plotted over
ment resulted in an approximately 60% dross and
the y-axis as they have a low q, whose influence can
90% g although the organic contaminants in UBC B
be excluded here. The UBC masses and surface
were three times those in UBC A. Contrary to the
areas were comparable and constant in all the
case of industrial furnaces, the laboratory melting
melting experiments with mUBC = 0.1 kg and
experiments were conducted in argon. It is hypoth-
A = 0.5 m2. In addition, these values are compared
esized that an inert gas effectively prevents the
with the D results of Ref.11, which were conducted
reaction of organic content. To verify this hypothe-
on a pilot scale with 200 kg uncoated Al and
sis, melting experiments were repeated in air
AlMg4.5 sheets. The scrap had a total surface area
instead of argon with all three UBC types. The
of 973 m2 and was melted in a gas-fired reverber-
impact of organic content on D and g was then
atory furnace. The comparison in Fig. 4 shows a
compared with that assessed by previous experi-
correlation between the mM in oxide films and D. In
ments in argon.
these laboratory-scale experiments, the scrap bales
Figure 5 compares the D and g of UBC scrap bales
were molten under inert gas, limiting molten metal
plotted against their organic contents. The rxy was
oxidation. The D is therefore mainly influenced by
also used to evaluate the fit of the variables to a
trapped metal in the carbon or oxide-reinforced can
linear trend. It is confirmed that melting in air
sheet oxide films.
results in a functional correlation between D and
From Fig. 4 it is summarized that pyrolyzed
the organic content, whereas this is not the case for
UBC produces comparable or even more dross
argon. Melting in air resulted in visually observable
than the initial scrap without any decoating. The
combustion of the organic content and immediate
carbon-containing coating residue on the can
metal oxidation. Metal oxidation leads to increased
sheets was found in the solidified dross even after
D and reduced g.
a 35-min reaction in the liquid melt at 750C.
The g obtained by melting the UBC scrap bales in
Metallographic investigations of the dross metal
argon and air is also compared in Fig. 5. As
interface revealed that the carbon layer effectively
expected, melting in air leads to a lower g because
prevented metal coalescence, trapping 70–90% of
of the uncontrolled combustion of organics, creating
the can alloy in the dross layer. Note that the
temperature peaks at the can sheet and resulting in
laboratory-scale D results are not directly trans-
metal loss. The rxy shows a smaller difference than
ferable to the industrial scale. Upscaling effects,
previously as the argon atmosphere does not pre-
such as the metallostatic pressure of liquid alu-
vent reactions of the liquid melt with gaseous
minum in the dross layer of industrial furnaces,
decomposition products of the organic
led to less mM and therefore less dross.
contaminants.
 Author's personal copy
3390 Steglich, Friedrich, and Rosefort

Fig. 5. Comparison between D and g of UBC A, B, and C2 by submerged melting in argon and air with the correlation coefficient rxy.

Consequently, the relationship between the supply required oxygen during the process. This
organic content and D can be understood via the appears quite challenging for state-of-the-art pro-
calorific value (Hi) of organics under these experi- cess-control techniques, thus requiring further
mental conditions. Herein, the Hi in combination improvements.
with the oxygen-containing furnace atmosphere led
to the uncontrolled combustion of scrap bales.
CONCLUSION
Due to the different behaviors of scrap bales
under similar process conditions, the organic The D and g of UBC recycling are the result of the
removal should be distinguished in the cases shown two-step process of thermal decoating and sub-
in Fig. 6. During pretreatment, UBC B was burning merged melting. The results show that the thermal
in over stoichiometric atmosphere, while UBC A pretreatment of UBC scrap bales effectively reduces
had a too low carbon content (or Hi) to ignite. It is D if stoichiometric thermolysis is applied for a
decisive that the Hi of organic contaminants (mix- holding time long enough to achieve a material
ture of polyethylene/polyester) is similar to that of temperature of 550–570C for at least 30 min. The
fuel oil with Hi = 10 kWh/kg.12 The cases in Fig. 6 use of shredded or loose UBC is recommended as a
show the different behaviors of UBC scrap bales and larger specific surface area will promote heat trans-
conditions to avoid uncontrolled combustion. port to the material and thermolysis gas transport
Regarding thermal decoating, Hi should be out of the scrap. The laboratory results indicate that
removed without exothermic combustion, which is UBC scrap bales with a low q (< 450 kg/m3) should
achieved by limiting the oxygen availability to a be used because a higher q negatively influences
stoichiometric ratio over the pretreatment time. For mass loss and D. This result needs to be verified in
an industrial process, it is necessary to measure the industrial-scale furnaces with submerged scrap
organic load, calculate the oxygen demand, and melting.
 Author's personal copy
Dross Formation in Aluminum Melts During the Charging of Beverage Can Scrap Bales with 3391
Different Densities Using Various Thermal Pretreatments

Fig. 6. Conditions of organic removal from UBC scrap bales with different calorific values expressed as a simplified stationary process.

Finally, it is shown that the organic content of ACKNOWLEDGEMENTS

UBC scrap bales should be considered while calcu-
We would like to acknowledge TRIMET Alu-
lating the Hi, which leads to metal oxidation in
minium SE for supporting this work. We also ex-
combination with oxygen-containing atmospheres
press our gratitude to Anne Kvithyld for her
in industrial furnaces. Metal oxidation is equal to
inspiration and motivation for this publication.
metal loss and promotes D by trapping liquid
aluminum in the dross layer.
 Author's personal copy
3392 Steglich, Friedrich, and Rosefort

REFERENCES 8. J. Campbell, Complete Casting Handbook: Metal Casting

Processes, Techniques and Design (Oxford: Butterworth-
1. J.V. Koleske, Paint and Coating Testing Manual, 14th ed. of Heinemann, 2011), pp. 52–54.
the Gardner-Sward handbook (Philadelphia: American 9. J. Steglich, C. Matthies, M. Rosefort, and B. Friedrich, Light
Society for Testing & Materials,1995), pp. 87–92. Metals 2018, ed. O. Martin (Springer, 2018), pp. 1123–1129.
2. D. Stoye, Paints, Coatings, and Solvents, 1st ed. (New York: 10. F. Ostermann, Anwendungstechnologie Aluminium, 3rd ed.
Weinheim, 1993), p. 8. (Berlin: Springer, 2014), pp. 171–182.
3. A.P. Mouritz and A.G. Gibson, Fire Properties of Polymer 11. H. Rossel, Light Metals 1990, ed. C.M. Bickert (TMS, 1990),
Composite Materials (New York: Dordrecht, 2006), pp. 22– pp. 721–729.
57. 12. M. Schwalbe, World Metall. Erzmetall 64, 157 (2011).
4. A. Kvithyld, J. Kaczorowski, and T. Abel Engh, Light Metals 13. J.H.L. van Linden, H.G. Reavis, Essential Read. Light Met.
2004, ed. A.T. Tabereaux (TMS, 2004), pp. 151–156. 165 (2013).
5. M. Modigell, Grundlagen der Luftreinhaltung, 3rd ed. (Aa-
chen, 2006), pp. 10–29.
6. S. Rumpel, Die autotherme Wirbelschichtpyrolyse zur Publisher’s Note Springer Nature remains neutral with re-
Erzeugung heizwertreicher Stützbrennstoffe, Universität gard to jurisdictional claims in published maps and institutional
Karlsruhe (TH), Dissertation, 2000, p. 11. affiliations.
7. J. Steglich, R. Dittrich, G. Rombach, M. Rosefort, B. Frie-
drich, and A. Pichat, Light Metals 2017, ed. A.P. Ratvik
(Springer, 2017), pp. 1105–1112.

View publication stats