AISTech 2019 — Proceedings of the Iron & Steel Technology Conference

6–9 May 2019, Pittsburgh, Pa., USA

DOI 10.1000.377.110

Hydrogen Control of Large Bottom Poured Forging Ingots at Ellwood Quality Steels

Bjorn Gabrielsson1, Brendan Connolly2, Steve Lubinski2, Sean Cowden2, Hongliang Yang3
1
Ellwood Group, Inc.
New Castle, PA, USA
2
Ellwood Quality Steels
New Castle, PA, USA
3
ABB R&D Metallurgy
Vasteras, Vastmanland, Sweden

Keywords: Hydrogen, Ingot, Bottom Pour, Vacuum Treatment

1. INTRODUCTION
The Ellwood Group operates two separate Electric Arc Furnace (EAF) melt shops, each with a maximum heat size of 47
Mton:
• Ellwood Quality Steels (EQS) in New Castle, Pennsylvania, with a capacity of 390,000 Mton of plain carbon, low
and medium alloy steel, tool steels and martensitic stainless. Maximum ingot weight at EQS is 170 Mton.
• Ellwood National Steels (ENS) in Irvine, Pennsylvania, with a capacity of 80,000 Mton of high alloyed, low carbon
stainless, Ni-based and other sophisticated alloys[1]. Maximum ingot weight at ENS is 90 Mton.
The EQS melt shop was commissioned in late 1985 with a maximum heat size of 40 Mton[2]. The heat size increased
gradually to 47 Mton with the following modifications:
• Original 1985 runner tap EAF replaced with enlarged EBT
• Lengthening of the original ASEA-SKF ladles
• Optimization of ladle refractory thickness
With these heat size improvements the maximum as-cast bottom poured forging ingot weight increased to 47 Mton from a
single heat. Production flow at the EQS melt shop is shown in Figure 1.

Figure 1: Production flow at EQS melt shop

The EQS melt shop has produced roughly 8,300,000 Mton of forging and ring rolling ingots since start up in December 1985.
There are two teeming bays at EQS:

© 2019 by the Association for Iron & Steel Technology. 1067

 • West Teeming Bay for 24 – 47 Mton ingots teemed by overhead crane
• East Teeming Bay for 2 – 24 and 60 – 170 Mton ingots
The EAF, Ladle Furnace No. 1 (LF1), Vacuum Station, Ladle Furnace No. 2 (LF2) and chemical laboratory (located between
EAF and LF1) are very compact with a total steel making platform length of only 85 m. Both ladle furnaces are equipped
with EMS in order to allow for short-arc reheating under a fully liquid reducing slag cover. This makes it possible for EQS
to consistently produce extremely clean steel with very tight composition control while maintaining high productivity[3].
Table 1 lists some key performance indicators (KPI) for 2017.
The sandwich pouring process was implemented at EQS in 2015 in order to produce up to 170 Mton ingots using four ladles
of liquid steel. Sandwich pouring is shown schematically in Figure 2. The development work and excellent quality results of
sandwich poured, large cross-section ingots at EQS are well documented[4].

Table 1: Key performance indicators, 2017

KPI Value
EAF Gross T-T-T 52.7 minutes/heat
EAF Power On Time 36.1 minutes/heat
EAF Power Off Time 16.6 minutes/heat
Productivity 27.3 heat/day

Figure 2: Schematic of sandwich pouring process

2. HYDROGEN IN STEEL
th
It was recognized in the early 20 century that certain internal hairline cracks in large steel forgings were related to
hydrogen[5]. These cracks have been termed “hydrogen flakes” and extensive research on their formation and prevention has
been performed by both academia and industry.
Hydrogen is present in steel as a monatomic species with high diffusivity and low solubility in low temperature
transformation products. The mechanism for hydrogen flake formation remains controversial, however calculations have
been performed[6] to show that the pressure buildup due to hydrogen within a steel matrix is easily high enough to exceed that
which even a high-strength steel is able to withstand.
The atomic fraction of hydrogen in equilibrium with H2 gas at pressure P (atm) is given as[7]:

3440

C 0  0.00185 Pe T

where C0 is the atom fraction of hydrogen and T is in Kelvin. P must be replaced by fugacity at the pressures being
considered. An approximation of the Taylor expansion can be used to estimate fugacity, f:

P2V
f 
RT

1068 © 2019 by the Association for Iron & Steel Technology.

where V is the molar volume and R is the gas constant. In order to determine the molar volume, the van der Waals equation
of state for one mole of gas can be used:

 a
 P  2  (V  b)  RT
V
27 R 2Tcrit 2
a
64 Pcrit

RTcrit
b
8 Pcrit

After determining the fugacity, the pressure and molar volume can be simultaneously solved. Figure 3 shows the internal
pressure buildup versus various amounts of hydrogen in the steel matrix at different temperatures. Figure 4 shows the
difference in relative volumes of steel, hydrogen gas and water at standard temperature and pressure.

Figure 3: Internal pressure buildup from hydrogen Figure 4: Steel, hydrogen and water volume
The important point from Figures 3 and 4 is that even at a hydrogen content of 1 ppm, coming from a very small relative
volume of water, the matrix will be unable to withstand the high internal pressure buildup at room temperature. The
hydrogen present within the steel must be accommodated in some fashion. Hydrogen accumulates at voids and interfaces
within the steel, thereby lowering the hydrogen dissolved within the matrix. Grain boundaries, dislocations, microporosity
and inclusions are all potential trapping sites where hydrogen is able to diffuse out of the matrix and remain in these traps
without detrimental flakes occurring[7]. Fully dense forgings with low inclusion content are more susceptible to hydrogen
flaking due to the reduced availability of trapping sites.
Hydrogen can be removed from steel forgings by sub-critical diffusion annealing in order to prevent hydrogen flaking.
However, the diffusion annealing practice is both time consuming and expensive. Figure 5 shows the required diffusion
annealing time versus the forging diameter for removal of 50% of the original hydrogen content at 650°C according to
Thelning’s calculation[8]. The diffusion annealing time required for hydrogen removal in large cross-section forgings is
prohibitively long.

© 2019 by the Association for Iron & Steel Technology. 1069

 Figure 5: Diffusion time for 50% hydrogen removal

3. EQS VACUUM STATION

The secondary steelmaking operation at EQS utilizes a vacuum hood degassing station with combined argon gas and EMS
stirring for fast and efficient hydrogen removal from the liquid steel. The EMS stirrer was upgraded from 1000A max
current to 1350A max current in 1998. The EQS vacuum station is shown schematically in Figure 6.

Figure 6: Schematic of vacuum hood degassing station at EQS

Typical operating pressure of less than 1 mbar above the liquid steel bath is consistently achieved using a 4-stage steam
ejector vacuum pump. The 4-stage steam ejector vacuum pump utilizes parallel ejectors in the first two pumping stages in
order to achieve the highest possible pumping capacity and reach operating pressure in less than four minutes.
It is common for steam ejector pumps used in the vacuum treatment of liquid steel to require periodic high-pressure water-jet
cleaning in order to remove the buildup of concrete-like deposits of process dust inside the pump itself. EQS installed an in-
line vacuum bag filter in 2004 between the vacuum station and the vacuum pump in order to separate the process dust that is
generated by evaporation and condensation of low vapor pressure elements such as magnesium, zinc and manganese[9]. The
dust is highly pyrophoric and will combust readily if it is exposed to oxygen. The bags are cleaned by nitrogen pulse jet after
each vacuum treatment and the dust is collected in a nitrogen pneumatic ejection dispenser at the bottom of the bag filter. At
the end of the bag cleaning cycle, the dust is ejected into the ladle furnace offgas system where it is combusted and then
captured by the ladle furnace baghouse. The in-line vacuum bag filter has provided several advantages for the EQS
secondary steelmaking operation:
• No degradation of vacuum capacity or vacuum pressure due to dust buildup in the pump over time
• No cleaning (downtime) of the vacuum pump is required
• No manual handling of the pyrophoric dust is required

4. HYDROGEN SAMPLING
EQS performs hydrogen sampling a minimum of two times for every heat produced. The first sample is taken from the ladle
after vacuum treatment to ensure the process achieved the necessary low hydrogen content prior to bottom pouring. The final
sample, which is used for certification, is taken from the ingot mold when the mold is almost full to ensure that the hydrogen

1070 © 2019 by the Association for Iron & Steel Technology.

content reported to customers is accurate and includes any hydrogen pickup that occurs during post-vacuum treatment
secondary steelmaking operations and bottom pouring.
Standard hydrogen sampling and analysis at EQS has been previously described and proven to be an accurate method[3,10].
The standard method at EQS is summarized below:
• Liquid metal sampling by evacuated 6 mm diameter glass pin tube
• Immediate, rapid quenching of the pin with cold, clean water
• Storage of the pin in dry ice until analysis (analyzed within 24 hours)
• 3 – 5 gram samples broken from the pin while cold
• Hot extraction analysis at 1100°C using LECO hydrogen analyzer

5. HYDROGEN REMOVAL
The amount of hydrogen dissolved in liquid steel is proportional to the square root of the partial pressure of hydrogen in
contact with the liquid steel according to Sievert’s Law. Removal of hydrogen from the liquid steel is achieved by exposing
the liquid steel to an atmosphere of reduced hydrogen partial pressure. The amount of hydrogen removed from the liquid
steel depends on the following factors[10]:
• Mass transfer coefficient of hydrogen, kH, which depends on flow conditions in the liquid steel
• Ratio between the free metal surface area and the volume of liquid steel
• Vacuum treatment time
• Hydrogen concentration difference relative to equilibrium with the partial pressure above the liquid steel according
to Sievert’s Law
• Stirring gas flow rate and hydrogen concentration difference relative to equilibrium with the partial pressure in the
stirring gas according to Sievert’s Law
• Concentration of surface active elements such as sulfur and oxygen in the liquid steel

5.1 CFD Simulations

Stirring power density due to Ar-gas and EMS stirring during vacuum treatment affects several of the factors listed above.
Simulations were performed by ABB R&D Metallurgy to evaluate how different stirring parameters will influence the
important factors for hydrogen removal. Three cases were simulated as shown in Table 2.

Table 2: CFD modeling parameters

PARAMETER CASE 1 CASE 2 CASE 3
Heat size, Mton 45 45 45
EMS type None ORT34 ORT34
EMS Current, A 0 1000 1350
EMS stir direction Up Up Up
Vacuum pressure, mbar 1.0 1.0 1.0
Ar flow rate, Nl/min 80 80 80
Slag amount, kg 600 600 600

Figure 7 shows the argon gas plug location, ladle geometry and location of the ORT34 EMS stirrer. The Ar-gas plug is
located on the EMS side, ½ radius of the working lining inside diameter.

© 2019 by the Association for Iron & Steel Technology. 1071

 Figure 7: Ladle, argon plug and EMS configuration
Figure 8 shows the results of the CFD simulation for the top surface of the liquid slag and steel for Cases 1 - 3. Figure 9
shows the liquid metal velocity vectors through the ladle cross section for Cases 1 - 3. The benefits of combined argon gas
and EMS stirring with the maximum EMS current (1350A) are as follows:
• Free metal surface exposure to the low pressure atmosphere is increased
• Liquid metal velocity at the free metal surface is increased
• Bulk liquid metal flow (mixing) is increased

Ar‐only, no EMS    Ar + 1000A EMS    Ar + 1350A EMS   

Figure 8: CFD modeling results for top surface liquid metal exposure and velocity (m/s)
 

Ar‐only, no EMS  Ar + 1000A EMS  Ar + 1350A EMS   

Figure 9: CFD modeling results for liquid metal velocity

The increased liquid metal velocity for Cases 2 and 3 also aids in dispersing argon gas bubbles throughout the liquid metal.
Figure 10 shows the argon bubble distribution within the liquid steel, with the color of each bubble indicating its velocity.
Dispersion of the argon bubbles throughout the melt is beneficial for the hydrogen removal process for the following reasons:

1072 © 2019 by the Association for Iron & Steel Technology.

 • Maximization of the bubble retention time to achieve, as close as possible, equilibrium hydrogen content in each
bubble according to Sievert’s Law
• Exposure of bulk liquid to the argon bubbles rather than only in a narrow stream of argon

Ar-only, no EMS Ar + 1000A EMS Ar + 1350A EMS

Figure 10: CFD results for Ar bubble dispersion and velocity (m/s)
The results of ABB’s CFD modeling are summarized in Table 3. The combined argon-gas and EMS stirring at 1350A
provides superior conditions for hydrogen removal during vacuum treatment.

Table 3: Summary of CFD modeling results from ABB

RESULT CASE 1 CASE 2 CASE 3
(Ar-gas only) (Ar-gas + 1000A EMS) (Ar-gas + 1350A EMS)
Stirring power density, W/ton 65 600 700
Free metal surface, % 7.4 22.5 27.9
Average surface velocity, m/s 0.14 0.41 0.53
Average bulk metal velocity, m/s 0.11 0.48 0.71

5.2 CFD Model Verification

Several verification tests were performed in order to confirm the results of the CFD modeling:

5.2.1 Mixing Time Comparison

A mixing time comparison of combined argon and EMS stirring versus argon-only stirring was performed by Gabrielsson
and Lubinski in 1988[10] using a 1000-amp EMS stirrer. The comparison is extended in this work to include the1350-amp
EMS stirrer currently installed at EQS. The mixing time study was performed with Cu as a tracer element. 5 kg Cu / t steel
was added to the ladle followed by chemical sampling every 10 seconds. Due to the practical limitations of obtaining
chemical samples during vacuum treatment, the mixing time study must be performed at atmospheric pressure. The argon
gas flow rate was adjusted to compensate for the loss of mechanical stirring power of the Ar-gas expansion at a vacuum
pressure of 2 mbar for consistency with the previous work. Sundberg’s formula[11] for calculating power through isothermal
expansion of the gas:

 T  P
387QT1 1  0   ln 1 
Pmech   T1  P2 
ML

where Pmech is the mechanical stirring power density (W/t), Q is the argon gas flow rate (Nm3/sec), T1 is the steel temperature
(K), T0 is room temperature (K), P1 is the pressure at the ladle bottom (Pa), P2 is the pressure above the liquid surface (Pa)
and ML is the liquid metal weight (ton). Figure 11 shows the results of the mixing time trials. These experimental results
confirm superior liquid steel mixing with combined Ar-gas and 1350A EMS currently installed at EQS, as predicted by the
CFD modeling.

© 2019 by the Association for Iron & Steel Technology. 1073

 Figure 11: Copper tracer mixing time results

5.2.2 Free Metal Surface Comparison

The CFD modeling results for free metal surface were tested with trials at the EQS vacuum station using the stirring
parameters as indicated in Table 2, with the exception of argon flow rate. Due to the camera view angle at the EQS vacuum
station, it is not possible to see the entire liquid surface during vacuum treatment. At typical argon flow rate of 80 Nl/min the
visible area shows nearly 100% exposed liquid steel. In order to show the difference in surface area exposure, the argon flow
rate for this trial was reduced to 60 Nl/min. Images from the vacuum station for each case (and also without Ar-gas or EMS
stirring) are shown in Figure 12, where the increased free metal surface for combined induction and argon stirring was
recorded utilizing a CCTV mounted in the vacuum hood, see Figure 6.

No stirring  Ar‐gas only, no EMS  Ar + 1000A EMS  Ar + 1350A EMS 

Figure 12: Camera view images for different stirring at EQS vacuum station

5.2.3 Hydrogen Removal Comparison

85 heats were produced without the use of EMS during vacuum treatment. The trial parameters were as follows:
• Vacuum pressure < 1.0 mbar
• Vacuum treatment time minimum 20 minutes
Table 4 shows the results of this trial compared to heats using EMS that met the same parameter requirements above during
2017. A histogram of the hydrogen values after vacuum treatment is shown in Figure 13. The average vacuum pressure was
slightly lower for the case of combined Ar-gas + EMS stirring, however this would only account for about 0.02 ppm
difference in equilibrium hydrogen according to Sievert’s Law. Combined Ar-gas and EMS stirring significantly improves
the exposed liquid steel surface, bulk mixing and Ar-gas bubble retention time, giving consistently lower post-vacuum
hydrogen content.

Table 4: Results of Ar-gas only stirring trial vs. Combined Ar-gas + EMS 1350A
Parameter Ar-gas only (No EMS) Combined Ar-gas + EMS 1350A
Number of heats 85 3256
Average pressure, mbar 0.77 0.72
Average time, minutes 27 27
Average H after vacuum treatment, ppm 1.05 0.59

1074 © 2019 by the Association for Iron & Steel Technology.

 Figure 13: Post-vacuum treatment hydrogen for Ar-gas only stirring vs Combined Ar-gas + EMS

5.3 Hydrogen Removal Results

Vacuum treatment (typically at < 1 mbar pressure) is performed on every heat produced at EQS. Vacuum treatment time
ranges from 10 – 30 minutes depending on the steel grade and cross section being produced. The average hydrogen content
in the ladle after vacuum treatment with combined Ar-gas and EMS stirring at EQS is shown from 2014 – 2018 YTD in
Figure 14. Continuous improvement efforts by EQS Steelmaking and Maintenance departments have been effective in
maximizing the vacuum station efficiency.

Figure 14: Average hydrogen after vacuum treatment by year

6. HYDROGEN PICKUP
After vacuum treatment is complete the liquid steel must be protected from hydrogen sources such as moisture and
hydrocarbons. In the secondary steelmaking operation at EQS, the primary hydrogen sources after vacuum treatment are
atmospheric moisture and moisture present in alloying additions. Atmospheric moisture is controlled by maintaining
complete coverage of the bath with a liquid slag layer during post-vacuum secondary steelmaking operations. This is
possible even during reheating at the ladle furnace due to EMS stirring without breaking the slag cover. All potential sources
of hydrogen such as argon stirring, alloying additions and slag additions are strictly controlled after vacuum treatment.
Hydrogen is tested after vacuum treatment for every heat produced at EQS in order to ensure that the necessary hydrogen
removal was achieved. An additional hydrogen test was taken from the ladle just prior to bottom pouring on 190 heats and
compared to the test taken immediately after vacuum treatment in order to verify that the post-vacuum treatment secondary
steelmaking operations do not make a significant contribution to the hydrogen pickup. The results of this trial showed an
average of <0.1 ppm higher hydrogen in the test just prior to bottom pouring, which confirms that there is no significant
hydrogen pickup during the post-vacuum treatment secondary steelmaking processes at EQS.
During the teeming operation there are several sources of hydrogen pickup:

© 2019 by the Association for Iron & Steel Technology. 1075

 • Ladle slide gate refractory components
• Atmospheric moisture
• Bottom pour tile and mortar cover
• Teeming flux
The influence of these various hydrogen sources on hydrogen pickup during bottom pouring were investigated in detail using
data collected and analyzed for over 30,000 heats produced at EQS.

6.1 Ladle Slide Gate Refractory Components

The ladle slide gate refractory components include the ladle inner nozzle, slide gate plates and collector nozzle. Typical life
for various ladle and slide gate refractory components are listed in Table 5.

Table 5: Typical replacement life for ladle and slide gate refractory components
Component Component life (heats per replacement)
Ladle brick (barrel + slagline) 80
Inner nozzle 15-20
Slide gate plates 5
Collector nozzle 5

Figure 15 shows the average relative hydrogen pickup for the first five heats on a new ladle (all components from Table 5
replaced for 1st heat on ladle) and for heats where only the inner nozzle or slide gate plate / collector were replaced. Note that
the slide gate plates and collector nozzle are always replaced when the inner nozzle is replaced.
Hydrogen pickup for a brand new ladle compared to a new inner nozzle and slide gate plates is identical, which indicates that
the newly relined ladle has no impact beyond the new inner nozzle and slide gate components. This is due to vacuum
treatment after initial exposure of the barrel and slagline to the liquid steel. The largest contribution to hydrogen pickup from
these refractory components is due to changing the slide gate plates and collector nozzle, accounting for up to 17% of the
hydrogen pickup. EQS has specific ladle refractory requirements in place for hydrogen-critical heats as part of the “Ultra-
Low Hydrogen Practice” in order to ensure minimal hydrogen pickup from new refractory components.

Figure 15: Effect of slide gate refractory status on hydrogen pickup

6.2 Atmospheric Moisture

Atmospheric moisture can have a significant contribution to hydrogen pickup during bottom pouring if the liquid steel stream
is not protected from contact with the atmosphere. Argon shrouding is performed on all heats produced at EQS, with the
exception of specific grades where particular customer requirements will not allow this practice.
The benefits of argon shrouding are more pronounced during hot, humid weather due to the additional moisture in the
atmosphere. The relative hydrogen pickup for non-shrouded and shrouded heats produced since 2015 are shown by month
along with average monthly humidity in Figure 16. Relative hydrogen pickup is expressed as percentage of average (100%

1076 © 2019 by the Association for Iron & Steel Technology.

being the average of all heats). Argon shrouding results, on average, gave 35% lower hydrogen pickup compared to non-
shrouded heats.

Figure 16: Hydrogen pickup versus humidity by month

The EQS argon shroud design was improved significantly at the end of 2015. The new argon shroud design allows for 100%
sealing of the liquid metal stream from the atmosphere and positive pressure of argon within the shroud. Oxygen testing
within the argon shroud during actual teeming operations confirmed oxygen levels below 0.5%, indicating that atmospheric
exposure is reduced by more than 97% when the EQS shroud is utilized.
In the production of large ingots (> 47 Mton) where the sandwich pouring process is used to pour more than one ladle of
liquid steel into a single mold, the liquid stream from the top ladle(s) is completely protected from the atmosphere by a ladle-
to-ladle shroud. The shroud is submerged into the liquid bath of the lower ladle for the entire duration of pouring from the
upper ladle[4].

6.3 Bottom Pour Tile and Mortar Cover

Refractory bottom pour tile contains some small amount of moisture that is absorbed after production during packaging,
transport and storage. Moisture content testing of EQS bottom pour tile confirmed that the practice of sealed pallet, dry
storage prevents any significant moisture pickup in the bottom pour tile during storage.
One source of moisture to the bottom pour tile is the refractory mortar that is used for compression of the tiles once they are
set in the sprue plate. The refractory mortar is an olivine-based material that is mixed with 5 – 10% water in the bottom pour
plate preparation area and evenly applied on top of the runner brick channel after bricks have been set. When the refractory
mortar is cured by the hot sprue plate it turns very hard, locking all of the bottom pour tiles together and assisting in
compression.
A hygrometer was placed inside the bottom pour tile in order to measure the specific humidity (g H2O / kg air) within the
bottom pour setup. A separate hygrometer was placed in the open atmosphere for comparison. Figure 17 shows that the
humidity inside bottom pour tile initially is decreasing when the tiles are first set, likely due to the heating of the tiles by the
hot sprue plate driving off residual moisture. When the mortar is applied to the brick there is a significant increase in the
specific humidity.

© 2019 by the Association for Iron & Steel Technology. 1077

  
Figure 17: Specific humidity increase from application of compression mortar
This increase in specific humidity within the bottom pour system presented a potential source of hydrogen pickup. Ohmori et
al.[12] found that preheating the bottom pour tile with dry forced air can reduce the moisture content of the refractory and
reduce hydrogen pickup. EQS has developed and implemented a bottom pour tile preheating system that is used for the
production of all large (>47 Mton) ingots. The preheating system is capable to reach air temperatures of 500°C in order to
drive all moisture out of the refractory bottom pour system.

6.4 Teeming Flux

The teeming flux used during bottom pouring at EQS contains approximately 0.25% moisture by weight according to recent
analyses. At EQS the teeming flux addition is based on the ingot cross-section and total volume. Figure 18 shows the
average grams of hydrogen pickup versus the grams hydrogen present in the teeming flux moisture for various ingot sizes
produced at EQS. The dataset excludes heats with new slide gate or inner nozzle refractory components. Hydrogen pickup
correlates strongly with the hydrogen present in the teeming flux.

Figure 18: Hydrogen pickup versus hydrogen present in teeming flux

When using molds that have been adequately preheated, EQS has found that application of flux by pouring directly into the
hot mold allows removal of some residual moisture from the flux and results in ~8% reduction in the hydrogen pickup. This
flux application method has been incorporated into the Ultra-Low Hydrogen Practice.

1078 © 2019 by the Association for Iron & Steel Technology.

 7. RESULTS AND CONCLUSION
The EQS secondary steelmaking processes consistently deliver high-quality, low hydrogen liquid steel to the teeming bay.
Standard practices in the bottom pouring operation minimize hydrogen pickup during routine operation. In the production of
large ingots (> 47 Mton) where sensitivity to hydrogen becomes even more significant, special Ultra-Low Hydrogen
Practices are in place to ensure < 1.5 ppm H in the final ingot:
• Specific ladle scheduling rules to ensure previous liquid steel exposure to inner nozzle, slide gate and collector
nozzle
• Vacuum treatment at < 1 mbar pressure with combined EMS and Ar-gas stirring
• Strict limitations on alloying and slag additions after vacuum treatment
• Argon shrouding of the lower ladle stream
• Ladle-to-ladle shrouding of the upper ladle stream(s)
• Preheating of bottom pour refractory to 500°C
• Proper preheating of mold
• Flux addition by pouring directly into hot mold for residual moisture removal
The Ultra-Low Hydrogen Practices have resulted in an average of 35% reduction in hydrogen pickup during bottom pouring.
Figure 19 shows the hydrogen pickup for standard processing (including testing just prior to bottom pouring for comparison)
and the Ultra-Low Hydrogen Practice. Figure 20 shows the distribution of final hydrogen analyses from sandwich poured
ingots at EQS, as reported previously[4].

Figure 19: Hydrogen pickup for standard and ultra-low - Figure 20: Histogram of final hydrogen content in
hydrogen practice sandwich-poured ingots at EQS

REFERENCES
1. J.R. Paules, W.P. Edwards and B.E. Gabrielsson, VOD Processing of Forged Precipitation Hardening Stainless Steels,
IFM, Tokyo, 2014.
2. B.E. Gabrielsson and R.E. Rumcik, Start up of the ASEA-SKF Secondary Steelmaking Unit at EUS, AISE Annual
Conference, Chicago, 1986.
3. B.M. Connolly and B.E. Gabrielsson, Process Control and Quality Assurance During the Production of High Quality
Bottom Poured Ingots at Ellwood Quality Steels, 1st ICRF, Aachen, Germany, 2012.
4. B.E. Gabrielsson, R.O. Olivares, B.M. Connolly, D. Schwartz and W. Zaben, Development, Implementation and
Results of Large Bottom Poured Ingots at Ellwood Quality Steels, 20th International Forgemasters Meeting, Graz,
Austria, 2017.
5. R.J. Fruehan, A Review of Hydrogen Flaking and Its Prevention, ISS Transactions, August, 1997.

© 2019 by the Association for Iron & Steel Technology. 1079

6. Correspondence between J.P. Hirth (Washington State University) and R.E. Rumcik (Ellwood Quality Steels), 22
August, 1996.
7. J.P. Hirth, Effects of Hydrogen on the Properties of Iron and Steel, Metallurgical Transactions A, Volume 11A, June,
1980.
8. K. Thelning, Steel and its Heat Treatment, 2nd Edition, Butterworth, London, 1984.
9. B.E. Gabrielsson, Vacuum Treatment of Liquid Steel at Ellwood Quality Steels – A Function Description, EQS internal
publication, November, 2013.
10. B.E. Gabrielsson and S.F. Lubinski, The Control of Hydrogen During Production of Forging Ingots at Ellwood
Uddeholm Steel Company, 9th International Vacuum Metallurgy Conference, San Diego, 1988.
11. T. Hsiao, T. Lehner, B. Kjellberg, Fluid Flow in Ladles – Experimental Results, Scandinavian Journal of Metallurgy,
Vol. 9, No. 3, 1980.
12. H. Ohmori, et al., Making of 200 ton max Forging Ingot by Bottom Pouring with LD-RH Process, Mizushima Works,
Kawasaki Corp., 1981.

1080 © 2019 by the Association for Iron & Steel Technology.