In the smelting process of steel, copper, aluminum and other metals,
melt spattering has always been the core pain point plaguing the
industry - the traditional process of up to 15% -25% of the metal
loss, the annual loss of more than 800 million U.S. dollars due to
the spattering of the loss of safety accidents caused by the line
downtime to clean up the loss of production capacity of up to
12%... ...Behind these figures, a huge technological gap in the
dynamic stability control of the smelting process is exposed.
With the breakthrough development of dynamic wind pressure
regulation and control technology, the milestone goal of smelting
process spattering rate below 0.5% has been realized for the first
time through the technology closed loop of real-time sensing-intelligent
decision-making-accurate execution, which pushes the metallurgical industry
towards the ultimate goal of “zero spattering, zero waste and zero accident”.
I. The nature of smelting spattering: the
cost of uncontrolled fluid dynamics
The spattering phenomenon originates from the intense interaction
of gas-liquid-solid three-phase flow inside the molten pool:
Gas injection destabilization: under the traditional fixed air pressure
mode, the oxygen/inert gas flow rate is disconnected from the reaction
state of the melt pool, resulting in a sudden rise in local pressure (fluctuation >30kPa)
Vortex energy build-up: random vortices (0.5-3m in diameter) formed
on the melt surface carry high temperature metal droplets (speeds up
to 12m/s) through the slag layer.
Interfacial tension imbalance: micro-jets (pressure peaks > 5 MPa)
generated at the moment of bubble bursting trigger the liquid metal film to break up.
The direct consequences of this dynamic imbalance are: loss of 18-25kg
of metal per ton of molten steel, 3 times faster erosion of the furnace
lining due to spatter, and a 40% increase in the load on the flue gas treatment system.
Second, the core of the technology: dynamic
wind pressure regulation and control of the
three intelligent closed-loop
1. Millisecond melting pool state sensing system
Multi-physical field sensing matrix: deploying high-frequency
pressure sensors (sampling rate of 10kHz), infrared thermal
imagers (accuracy ± 2 ℃), electromagnetic eddy current detection
module, real-time capture of the melt pool temperature gradient,
bubble distribution, slag layer thickness and other 20 + dimensional parameters
Fluid dynamics digital twin: based on Computational Fluid
Dynamics (CFD) to build a three-dimensional dynamic model
of the melt pool, and update the gas-liquid interface evolution
prediction every 0.1 seconds.
2. Adaptive wind pressure control algorithm
Deep Learning Prediction Engine: Analyze the historical splash
event data through LSTM neural network, and predict the risk
point of sudden pressure change 300ms in advance.
Non-linear Model Predictive Control (NMPC): Dynamically
optimizes the gas flow rate setting value, compressing the
wind pressure fluctuation from ±25% to within ±3%.
3. High-precision actuator
Intelligent regulator response time <50ms, realizing precise
adjustment of gas supply flow at 0.1m³/min level.
Multi-nozzle cooperative control system, according to the
melting pool partition state independent adjustment of
the wind pressure parameters of the 18 injection units
Technical breakthroughs: from theory to
industrial scenarios of the four major
innovations
1. Active regulation of bubble behavior
By adjusting the gas pulse frequency (0.5-5Hz), the bubble
diameter is controlled in the ideal range of 8-15mm,
reducing the kinetic energy of bubble rupture by 75%.
Inhibit malignant bubble aggregation, so that the melt
pool surface bubble distribution uniformity increased to 95%.
2. Vortex energy dissipation mechanism
Within 0.2 seconds of detecting the vortex formation, the
reverse airflow injection strategy is activated, which
accelerates the vortex energy attenuation speed by 8 times.
Combined with acoustic intervention technology
(frequency 20-40kHz) to further destabilize the vortex structure.
3. Dynamic compensation of interface tension
Real-time adjustment of gas components (O₂/N₂/Ar ratio)
according to slag composition (CaO/SiO₂ ratio) to optimize
slag-gold interfacial tension
Introducing micro-bubble buffer layer to raise the critical
pressure threshold for liquid film rupture to 6.8MPa.
4. Full-cycle process optimization
Constructed a synergistic pressure-temperature-composition
control model for the smelting process, and increased the
spatter risk warning accuracy to 99.3%.
Developed a self-learning control system, and realized
autonomous optimization of algorithm parameters within
200 smelting cycles.
Industrial Value: Reconstructing the
economy and safety of metallurgical production
Metal recovery rate jumped
Metal loss per ton of steel has been reduced from 18kg to
0.5kg, saving more than 36 million RMB per year for a
million-ton steel plant.
The recovery rate of precious metals (e.g. copper, nickel)
has increased to 99.8%, breaking the limit of resource
utilization in the industry.
Extended equipment life
Furnace lining erosion rate reduced by 60%, overhaul cycle
extended from 800 to 2500 furnaces
Flue gas treatment system load reduced by 35%, annual
maintenance cost reduced by 40%.
Upgraded safety
Accident rate due to sputtering reduced to 0.02 accidents/thousand
furnaces, 25% savings in insurance costs
Workshop dust concentration (PM10) reduced from 15mg/m³ to
3mg/m³, meeting occupational health and safety standards.
Green manufacturing breakthrough
Comprehensive energy consumption per ton of molten steel reduced by
12%, CO2 emissions reduced by 8.5 tons/thousand tons
Metal oxide dust emissions decreased by 90%, helping the enterprise
to achieve carbon neutrality goals
V. Future evolution: from zero spatter to
full intelligence of smelting process
Multi-objective synergistic optimization
Development of a joint decision-making model for spatter control,
energy consumption optimization, and composition regulation to
achieve global optimization of the smelting process.
Extension of cross-scale regulation
Integration of molecular dynamics simulation to regulate melt viscosity,
surface tension and other intrinsic properties at the microscopic level.
Autonomous decision-making system
Construct AI controller with causal reasoning capability to maintain
stable control under abnormal working conditions such as raw
material fluctuation and equipment aging.
Integration of industrial meta-universe
Realize the visualization and monitoring of melt pool fluid
movement through AR/VR technology to create an immersive
process control platform.
Conclusion
The successful application of dynamic wind pressure control technology
in the field of smelting zero spattering marks that metallurgical process
control has formally stepped into the era of “data intelligence” from
“experience-driven”. This technology not only solves the process
problems that have plagued the industry for a hundred years, but also
provides key technical support for the sustainable development of the
global metallurgical industry by reducing resource loss, improving
production safety, and reducing environmental load. With the penetration
and integration of 5G, digital twin, quantum computing and other
new-generation technologies, dynamic wind pressure regulation and
control will surely give rise to more breakthrough innovations, and
continue to promote the metallurgical industry in the direction of
high-end, intelligent and green evolution.