Semi-solid die casting method with magnesium alloy damping tower center feeding sub-zone mold temperature coordination

Abstract

The application discloses a semi-solid die-casting method of magnesium alloy shock-absorbing tower center feeding partition mold temperature coordination, relates to the high-pressure die-casting control technical field, and aims at the problems that far-end cold isolation, gas trapping and hot spot shrinkage hole are prone to occur in the bottom or side edge feeding, uniform mold temperature and experience exhaust of thin-thick coupling parts and are difficult to be once shaped, and proposes a four-step coordination method: establishing center feeding or multi-point center neighborhood feeding, and completing the consistent design of main runner, runner, fan-shaped gate and injection curve; according to the filling path, the fixed mold and the movable mold are configured with partition mold temperature, and are linked with the injection timing; along the gas trapping trace, the open-close exhaust valve and the end overflow cavity are arranged, the total equivalent exhaust area and the valve timing are calculated; the process column is arranged at the hot spot, and the neck communication structure is arranged, and the CT, air tightness and cavity temperature and pressure curve backfilling are combined to form a parameter package. The scheme realizes the coordinated pressure drop of cold isolation, gas trapping and shrinkage hole, shortens the trial and repair cycle, improves the first pass rate and has the migratory property.

Description

Technical Field

[0001] This invention relates to the field of high-pressure die casting control technology, specifically a semi-solid die casting method with coordinated mold temperature control in the central feeding zone of a magnesium alloy damping tower. Background Technology

[0002] The trend towards lightweighting in automobiles is driving the application of magnesium alloys in chassis and body structural components. Thin-thickness coupling components such as shock absorber towers, shock absorber seats, and longitudinal beam end reinforcements are shifting from gravity casting to high-pressure die casting and semi-solid injection molding. Current mainstream processes typically involve bottom or side feeding, with gate speeds focused on cycle time (often falling within an engineering experience window). Mold temperatures are often set uniformly across the entire mold (e.g., a uniform 200–250℃ range). Venting channels and overflow cavities are strategically placed at the high points and ends of the parting line based on experience. Multiple trial moldings are then used to fine-tune the gate cross-section, runner, injection curve, and spraying rhythm to achieve consistent appearance and dimensions. This approach, under the influence of geometrically complex components (ribs, flanges, reinforcing steps) and operational disturbances (protective gas, vacuum extraction, segmented injection), is prone to three types of coupling defects: cold shuts due to premature cooling in the thinner distal areas; air entrapment induced by the free surface folding back at rib intersections; and shrinkage cavities caused by delayed solidification of thick-walled hot sections and obstructed feeding channels. More importantly, due to the lack of an integrated design that uses filling temperature margin, air entrapment risk, and shrinkage risk as core metrics, traditional methods often rely on trial and error based on experience, resulting in high trial mold costs, long cycles, and poor reusability across projects.

[0003] In the high-pressure die-casting scenario of magnesium alloy shock-absorbing towers with thin-thickness coupling and interlacing ribs and flanges, there is a lack of an engineering method that starts with central feeding and connects the entire process of zoned mold temperature, injection timing, valve-controlled venting / overflow, and thick-walled feeding. This results in difficulty in maintaining the filling temperature of the thin area at the far end, easy air entrapment at the intersection of ribs on the free surface, and the feeding channel of the thick-walled hot section is often cut off by the early solidification neck, thus forming a chain superposition of cold shuts, air entrapment, and shrinkage cavities. This leads to a high number of mold iterations and large fluctuations in air tightness and appearance.

[0004] This problem arises from the currently prevalent bottom or side-feed + uniform mold temperature + empirical venting / overflow process: the gate and runner are not consistently constrained by volumetric flow rate and local gate velocity; the critical filling path is too long, causing thin areas to cool down first; the mold temperature is not calculated backwards based on path-zone-safety margin, resulting in uncontrollable temperature difference between the inlet and outlet; the total venting area and opening sequence are not quantitatively calculated backwards based on pressure difference and target volume, causing trapped air to be forced into the outlet later; the overflow volume and process column heat capacity are not matched with the heat required by the hot spot, and excessively rapid valve switching tears the oxide film, contaminating the feed liquid source. This problem is further exacerbated by disturbances such as protective gas mixing, vacuum venting fluctuations, and segmented injection.

[0005] Early high-speed free surface fluctuations cannot be offset by subsequent temperature increases. Asynchronous positions of the exhaust window and punch lead to abrupt changes in intracavity pressure differential. The neck geometry's discontinuous solidification time window design shortens the feeding window. Furthermore, the lack of a closed-loop system that uniformly maps CT, leakage rate, and cavity temperature-pressure curves to parameter updates makes it difficult to precipitate test data into standardized parameter packages. The direct consequences are: decreased first-pass yield for airtightness, increased porosity and secondary shrinkage cavities, increased mold thermal fatigue and repair costs, and difficulty in reusing parameters across similar vehicle models. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a semi-solid die-casting method for magnesium alloy shock absorber towers with synergistic temperature control in the central feeding zone. This method achieves synergistic pressure reduction in the areas of cold shuts, air entrapment, and shrinkage cavities, shortens the trial and repair cycle, improves the first-pass yield, and is transferable. It solves the problems of cold shuts at the far end, air entrapment at the ribs, and shrinkage cavities at hot joints that are prone to occur in thin and thick coupling parts when fed from the bottom or side edge, with uniform mold temperature and empirical venting, and is difficult to shape in one go.

[0008] (II) Technical Solution

[0009] To achieve the above objectives, the present invention provides the following technical solution:

[0010] A semi-solid die-casting method for magnesium alloy vibration damping towers with center-feed and zoned mold temperature coordination includes: determining center-feed or multi-point distributed center-neighbor area feeding; completing the matching design of the main runner, branch runner, fan-shaped gate, and injection curve to ensure that the gate cross-sectional area is consistent with the injection volume flow rate; and outputting the feed-gate-branch runner scheme for subsequent use. Based on the filling path of the feed-gate-branch runner scheme, the fixed mold and moving mold are divided into an entrance area, rib area, end area, and flange area. The mold temperature of each zone is set and linked with the injection curve timing to form a zoned mold temperature setting and injection timing.

[0011] An openable and closable exhaust valve, exhaust groove, and vacuum interface are arranged along the air duct trace. An overflow cavity is set at the end and hot spot and connected to the process column. The total equivalent exhaust area and overflow volume are calculated, and the valve timing is determined and linked with the punch position and cavity pressure. A process column that can be machined and removed is set at the hot spot and connected to the neck structure, forming a sequential solidification and feeding channel with the overflow cavity. Trial production and testing are carried out, and the gate, zone mold temperature, valve timing, and overflow cavity are refilled according to the test data and solidified into a parameter package.

[0012] Furthermore, the equivalent cross-sectional area of ​​the gate is determined based on the flow reduction coefficient and is set synchronously with the segmented volumetric flow rate of the injection curve and the average speed of the gate; the cross-sectional ratio, transition fillet, and incident angle of the main runner and branch runner are determined according to the filling path constraints; the inlet fillet, throat thickness, and opening position of the fan-shaped gate are located at the geometric center of the part or its neighborhood; the injection curve is sent to the injection mechanism as a control parameter and is time-aligned with the punch position encoder.

[0013] Furthermore, a comprehensive criterion is established based on the filling temperature margin, air entrapment risk, and shrinkage risk, and these criteria are set according to project weights. Based on the comprehensive criterion, either center feeding or multi-point distributed center neighborhood feeding is selected, and the initial geometric combination of the gate and runner is simultaneously screened.

[0014] The stage speed and switching time of the injection curve are set to correspond to the segment boundaries of the comprehensive criterion and recorded as the feed-gate-runner scheme.

[0015] Furthermore, the partition boundaries of the inlet area, rib area, end area, and flange area are determined based on the filling path and path allocation coefficient. In each partition, at least two combinations of hot rods, oil temperature or water cooling, and heating plates are selected to configure temperature control channels and temperature settings.

[0016] The injection curve is divided into three segments: the initial segment, the middle segment, and the final segment, and each segment corresponds to a specific mold temperature setting table. The settings are then sent to the mold temperature controller and the injection controller.

[0017] Furthermore, the average velocity of the gate is convolved with the partition weight and the time kernel function to obtain the predicted end temperature, and the injection curve is iteratively updated with the predicted value as a constraint.

[0018] Each partition is set with a safe temperature limit and associated with the feasible range of the partition's heat dissipation capacity, serving as a linkage constraint for subsequent exhaust and overflow; the iterative update is executed in time slices within the control cycle, and the temporal mapping relationship between injection segments and partition numbers is preserved.

[0019] Furthermore, based on the target exhaust time, the time function of the gas volume and the cavity-vacuum pressure difference to be derived is used to calculate the total equivalent exhaust area, which is then used to determine the diameter of the openable and closable exhaust valve and the cross-section of the exhaust groove.

[0020] The openable and closable exhaust valve is arranged in parallel with the exhaust groove. The valve opening sequence is linked to the punch position, cavity pressure and target cavity pressure trajectory for control. The vacuum interface is connected to the vacuum box and connected to the pumping device through a pipeline.

[0021] Furthermore, the lower limit of the overflow volume is set according to the spatial integral of the liquid phase fraction and the volume shrinkage coefficient at the critical moment of filling, and the position and size of the overflow cavity are located at the end of the filling path.

[0022] The process column is connected to the overflow cavity via a neck, and the neck has a narrow section; the valve switching curve satisfies the constraint of the product of the opening change rate and the differential pressure, and the upper limit of the opening change rate and the differential pressure sampling are used for curve generation.

[0023] Furthermore, a heat capacity matching ratio threshold is set, and the volume, position, and material thermal properties of the process column and the overflow cavity are used for equivalent heat capacity calculation to balance the heat required by the hot spot; the minimum cross-section and length of the neck are calculated according to the solidification time window in sequence, and the feeding channel is arranged along the centroid direction of the hot spot; the process column is a machined and removable structure with a breakable machining allowance, and the end face of the process column is flush with or lower than the parting surface.

[0024] Furthermore, CT pore volume fraction, leakage rate, cavity temperature and pressure curve and end temperature data are collected, and a closed-loop loss and safety consistency regularization containing the above quantities is constructed. The parameters of the gate equivalent cross-sectional area, zone mold temperature, valve opening function, overflow volume and injection control function are backfilled and versioned for storage.

[0025] The parameter reinjection is performed under the constraints of the zone's safe temperature limit, valve smoothness, and minimum exhaust area. The parameter package includes geometry, timing, and thresholds and is associated with the part number.

[0026] (III) Beneficial Effects

[0027] This invention provides a semi-solid die-casting method for magnesium alloy shock absorber towers with coordinated central feeding zone mold temperature control, which has the following beneficial effects:

[0028] By using central feeding or multi-point distributed central neighborhood feeding and matching it with the main runner, branch runners, fan gates, and injection curves, the critical filling path is shortened, reducing the probability of premature cooling in thin areas and free surface backflow at path intersections. The consistent design of the feeding, gate, and branch runner maps the equipment-side volumetric flow rate to the local velocity boundary of the gate, providing a unified input for zoned mold temperature and valve-controlled venting, and avoiding nozzle drift in subsequent stages.

[0029] The mold is divided into entry zone, rib zone, end zone, and flange zone based on the filling path, and zoned mold temperatures are configured accordingly. Combined with segmented control of the injection sequence, the thermal boundaries are coordinated in space and time, suppressing sudden drops in local temperature and the generation of secondary free surfaces. The zoned mold temperature setting table and the injection sequence are established with a numbering and time alignment relationship, allowing for integrated programming of venting windows, overflow execution, and punch positions, reducing conflicts between processes.

[0030] The temperature rise is constrained by a zoned safety margin and a heat load matching function to prevent heat accumulation in the inlet and flange areas from exceeding the heat dissipation capacity, and to provide a reliable differential pressure baseline for the valve opening function and the target cavity pressure trajectory. Openable and closable exhaust valves, exhaust channels, and vacuum interfaces are arranged along the gas entrainment trajectory, and the opening timing is driven by the target cavity pressure trajectory to achieve synchronous scheduling of exhaust capacity and free surface stability.

[0031] The total equivalent exhaust area is calculated by considering pressure difference, exhaust time, and the volume to be derived, ensuring that the exhaust unit diameter matches the vacuum system capacity and reducing uncertainties caused by arbitrary slotting. An overflow chamber is set at the end near the hot spot and connected to the process column. The overflow volume is set with a lower limit based on the liquid phase fraction and volume shrinkage requirements. The valve switching curve is constrained by the opening change rate and pressure difference, forming a clean and connectable liquid storage end.

[0032] A machined-removable process column and neck connection structure are designed at the hot spot. The dimensions and positions are checked according to the heat capacity matching ratio and sequential solidification time window to ensure that the feeding channel remains accessible at the end of solidification. A joint acquisition process for trial production, CT inspection, metallographic determination, airtightness testing, and cavity temperature and pressure curves is established. The detection quantities are mapped to closed-loop loss and safety consistency regularization to drive the backfilling of parameters such as feeding, zoned mold temperature, valve opening, overflow volume, and injection curve. Attached Figure Description

[0033] Figure 1 This is a schematic diagram of the semi-solid die-casting method for the magnesium alloy shock absorber tower with coordinated central feeding zone mold temperature according to the present invention. Detailed Implementation

[0034] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0035] Please see Figure 1 This invention provides a semi-solid die-casting method for magnesium alloy shock absorber towers with coordinated central feeding zone mold temperature, comprising:

[0036] Step 1: Based on the part geometry and melt properties, construct a feeding topology criterion that prioritizes center feed and solve it in a consistent manner with the main runner-sub-sprue-fan gate-injection curve. This will output the boundary conditions and initial process solutions that can directly drive the partitioned mold temperature back-calculation and the venting-overflow organization.

[0037] The damping tower is a thin-thickness coupled component. The ribs and flanges make it easy for free surface folding and backflow to occur. If bottom feeding is still used, the chain problem of the thin area at the far end cooling first and the hot section of the thick wall being restricted in the feeding path will be aggravated. Therefore, it is necessary to make a calculable decision on the priority of center feeding and multi-point distributed center neighborhood feeding through a unified criterion.

[0038] First, the part feature length, thin area expansion path and rib bifurcation point are abstracted into a filling path diagram. Then, the melt outlet temperature, thixotropic flow behavior and injection execution curve are projected into a comparable comprehensive criterion. Finally, the feed topology is locked with the optimization result of this criterion, and a filling path sequence is generated simultaneously for subsequent mold temperature zoning reference.

[0039] When constructing the feed topology, the comprehensive criteria are used directly. The feasible balance weights of filling insulation-air entrapment suppression-feeding are characterized, and this criterion is used to select the optimal feed center or multi-point distributed center neighborhood feed center.

[0040] To avoid non-quantifiable empirical evaluation, the geometrically guided filling path, free surface folding risk, and hot spot shrinkage potential are mapped to three measurable components, and weighted uniformly to a single-value judgment, thereby achieving parameter integration with subsequent processes. Therefore, a comprehensive criterion is established. :

[0041]

[0042] Among them: Comprehensive criteria : A single-value index used to rank the quality of feed topologies; the larger the value, the better; a real number.

[0043] Temperature margin The positive temperature difference of the distal thin region relative to the critical temperature at the end of the filling process is used to measure the buffer against cold shut risk; it is a positive real number.

[0044]

[0045] Among them: terminal temperature prediction For endpoint prediction; critical temperature Take the critical filling temperature of the alloy (a safe temperature difference point above the liquidus line can be taken, and the process calibration range is a positive real number). It is obtained by path-weighted and temporal kernel convolution; The alloy properties and appearance / cold shut threshold specimens were calibrated together.

[0046] Gas risk : A negative index based on free surface return and vorticity path accumulation. The larger the value, the more severe the air entrainment. It is a non-negative real number.

[0047]

[0048] Among them: free region The set of surfaces containing instantaneous free surfaces, in actual calculations, is a volume fraction between... The discrete surface element set obtained by reconstructing the shape of the transition unit; velocity field Flow field obtained from simulation or calculation;

[0049] vorticity ;Measure the local vortex intensity, Its model; To measure the rate of volume expansion / compression, take Only the compressed converged portion is retained; weighting Positive coefficients defined by the project for dimensional balancing and engineering trade-offs; filling completion time. The upper bound of the time from the metal front entering the mold to the end of the filling process; the simulation end outputs free surface tracking and vorticity / divergence field, and the prototype end uses high-speed visual / pressure fluctuations as a proxy for verification; time / area integration can be numerically realized.

[0050] Shrinkage risk : A negative index based on solidification sequence and feeding connectivity, a non-negative real number;

[0051]

[0052] Wherein: temperature gradient mode : at the moment of local evaluation The magnitude of the temperature gradient at that location, and the hot spot region : Volume subdomains with later curing times in the casting; evaluation time It is recommended to use the time of equal solid fraction at the end of solidification. (like (as defined by the project), or in the final time window Internal use:

[0053]

[0054] To capture the worst-case scenario; cooling rate : Cooling rate at the same evaluation time, taken as non-negative; Niyama threshold Material-process calibration constants were determined by fitting CT / metallographic results from a single trial production; temperature weighting. , spherical weight Hole shrinkage weight Used to express the project's engineering focus on the three types of defects, satisfying... and .

[0055] In application, the three elements of geometry, melt flow, and execution can be unified into a comparable scalar, preventing the optimization of a single index from amplifying other defects in the center feed. This provides a clear optimization direction for subsequent zoned mold temperature back-calculation, focusing on improving the temperature margin. and reduce the risk of cyclone Improvements in the same direction; the trade-offs between different performance objectives (air tightness, appearance, fatigue) can be made transparent and traceable through the project-based setting of weights.

[0056] After the feed topology is locked by the comprehensive criteria, the volumetric inflow of the injection is consistent with the geometric transport capacity of the gate-runner, so that the injection curve-gate speed-runner pressure drop-free surface fluctuation link is closed; otherwise, although the center feed is good, the local acceleration and shear roll caused by the mismatch between the cross section and the constriction will still trigger air entrapment and cold shut.

[0057] Therefore, the volumetric flow rate on the injection execution side is equivalently correlated with the effective flow capacity of the gate area, and an equivalent cross-section solution is constructed to clarify the coupling boundary between geometry and execution:

[0058]

[0059] Where: equivalent cross-sectional area of ​​the gate : Effective flow area of ​​the fan-shaped gate after flow reduction correction, a positive real number; volumetric flow rate. : The volumetric transport rate of the injection process at a given time slice, determined by the injection profile, a positive real number; the shrinkage coefficient. Considering the effective area reduction factor caused by the boundary layer and rapid contraction, Average gate speed The average flow velocity in the local area of ​​the fan-shaped gate is determined by the engineering window based on the thin wall of the part and the mold material's tolerance; it is a positive real number.

[0060] In application, the four elements of the main runner, branch runner, fan gate, and injection curve can be quantitatively closed, allowing any adjustments at the injection end to be instantly absorbed by the gate cross-section; it can provide a true inlet thermal-dynamic boundary for partitioned mold temperature back-calculation, i.e., the average gate velocity. With volumetric flow rate The temporal distribution; can be determined by the flow reduction coefficient. By explicitly filtering out geometric combinations that could induce strong turbulence on the free surface, the burden on subsequent exhaust systems is reduced.

[0061] The thixotropic flow and oxidation sensitivity of magnesium alloys mean that geometry-execution matching still needs to cross the physical property-cleanliness threshold. Otherwise, even if the feed topology and cross-section are properly matched, insufficient cleanliness of the melt or improper control of supercooling will induce secondary free surface reversal and potential slag inclusions at the intersection of ribs.

[0062] A reproducible flow criterion needs to be defined within the combined space of outlet temperature, liquidus gap, and inclusion area ratio, and integrated with the prior matching conditions of gate heat load and zone heat capacity to ensure that the mold temperature back-calculation in step two has a clear upper limit of heat load and response capacity table.

[0063] Considering the fragility of the oxide film on magnesium alloys at high temperatures and the destructive effect of fine inclusions on the stability of free surfaces, a single outlet temperature is insufficient to guarantee the far-end insulation effect under center-feed conditions. Therefore, a joint criterion that considers both temperature difference-driven and cleanliness inhibition factors needs to be proposed as a material-side gate before the implementation of the geometry-execution scheme. Define the flowability index. for:

[0064]

[0065] Among them: Liquidity Index The contribution of overheating and inclusion inhibition to fillability is comprehensively characterized by positive real numbers; the outlet temperature is... : Temperature of the molten metal entering the casting system, a positive real number; liquidus temperature : The critical temperature at which the alloy begins to fully liquidate, serving as a zero-point reference for overheating; a positive real number; inclusion area ratio. The proportion of surface area of ​​oxide inclusions per unit cross section. Exponential coefficient , , Process calibration coefficient ; The overheat sensitivity index has a suitable range. ; is the inclusion suppression intensity coefficient, a positive real number.

[0066] In application, the flowability and cleanability of the material can be coupled into a single threshold, allowing the geometry-execution scheme to only apply to... When the target is met, it is released to the trial production; through the dual design of exponential gain and exponential suppression, the risk of inclusion is not masked by simply increasing the temperature due to overheating; the qualified soup discharge range and cleanliness range derived from this criterion can be directly solidified as the input parameter constraint for the mold temperature back-deduction in step two, ensuring that the thermal management strategy does not sacrifice cleanliness.

[0067] Center feed has a natural advantage in heat preservation at the far end, but if the heat flux per unit area at the gate exceeds the transient heat dissipation capacity of the mold zone, it will cause thermal fatigue and tempering color defects in the entrance area. Therefore, it is necessary to complete the prior matching of the gate heat load and the zone heat capacity in step one, so that step two no longer involves unconstrained temperature rise and fall search.

[0068] First, the heat load density of the gate is approximated using the law of energy conservation. :

[0069]

[0070] And define the heat load matching function. This is used to measure the norm of the difference between the distribution of the heat load in each temperature-controlled zone and its inherent heat dissipation capacity.

[0071]

[0072] Wherein: heat load density Instantaneous sensible heat input per unit area at the gate, a positive real number; melt density. Density of liquid magnesium alloy, positive real number; specific heat at constant pressure Sensible heat capacity per unit mass of melt, positive real number; tapping temperature. As defined above, a positive real number; mold initial temperature. Initial surface temperature of the temperature-controlled zone, a positive real number; volumetric flow rate. As defined above, positive real numbers;

[0073] equivalent cross-sectional area of ​​the gate As defined above, positive real numbers;

[0074] Partition set : A set of temperature control partition indices for fixed / dynamic molds, a finite discrete set;

[0075] Path allocation coefficient The filling path is in the partition. The influence weights satisfy and ;

[0076] make Mainly filled streamline, length Indicator function Indicates streamlines in the partition If the value is 1, then 0; otherwise:

[0077]

[0078] On-site, the main channel was approximated by inverting the displacement of the punch and the temperature-pressure curve, and then an approximate distribution was made. This was done at a representative moment during the filling process. (e.g., during the later stages of filling or at a volume fraction of approximately 0.5), take the velocity field. .make:

[0079]

[0080] in This is the seed point at the center of the gate section or at the point of maximum throughput. Integrating this ordinary differential equation yields the result. If multiple central neighborhood feeds are used, calculations can be performed separately for each gate, and the one with the largest throughput can be selected as the "main" flow line, or the average mainstream flow line can be calculated by weighting the throughput.

[0081] Partitioned heat dissipation capability Partition Effective heat dissipation capacity per unit area over a given time scale, a non-negative real number; heat load matching function. : A measure of the mismatch between heat load and heat dissipation capacity, a non-negative real number, with the objective of minimizing it.

[0082] When applying this method, exclusionary conditions for infeasible heat load zones can be given in advance to avoid unnecessary searching in high-risk areas in step two; the filling path distribution can be determined using path allocation coefficients. Explicitly projecting onto the mold temperature zone ensures tight coupling of the geometry-thermal-execution three-domain parameters in step one; when the thermal load matching function... Once minimized, step two only requires considering the upper bound of feasible heat dissipation capacity. Fine-tuning can be performed internally to shorten the parameter tuning time and reduce the risk of mold thermal fatigue.

[0083] Step 2: Based on the coordinated setting of zoned mold temperature and injection timing, construct a calculable end-temperature prediction, injection curve constraint, and zoned temperature control update law to ensure that the equivalent cross-sectional area of ​​the gate from Step 1 is... Volumetric flow rate Average gate speed Path allocation coefficient Partitioned heat dissipation capability Matching function with heat load Forming zone temperature settings With injection control function The one-to-one mapping ultimately outputs the partitioned mold temperature setting table and injection timing sequence.

[0084] While center-feed technology has shortened the critical flow path geometrically and in terms of flow channels, sudden temperature drops may still occur in the thinner areas at the end of the injection process. Furthermore, the reversal of the free surface at the rib intersection zone, if combined with higher injection velocities, can increase the risk of air entrapment. rise.

[0085] Therefore, firstly, a predictable model for terminal temperature is established based on the zoning-path mapping, and the zoning temperature is set. Average speed of gate The temporal effects are quantified within the same predictor; subsequently, this predictor is used as a constraint and feedback to modify the injection control function. This ensures that the filling temperature margin is improved. With the risk of suppressing air vortex .

[0086] The process unfolds as a single-chain prediction-driven control: first, the influence of each zone's temperature on the end point is superimposed using path weighting; then, an exponential kernel is used to describe the time decay effect of the thermal-dynamic input in the gating zone, thus forming an executable end-point temperature target; thereby, under the target constraint, the... Gradient updates are applied to achieve equilibrium between free surface stability and terminal temperature without exceeding the heat load limit given in step one; finally, the resulting zone temperature settings are applied. With injection control function As input for the next level of partition updates and security checks.

[0087] To avoid amplifying thermal fatigue by simply increasing the inlet temperature, a predictable representation of the terminal temperature is constructed using a partition-path weighting-time kernel convolution approach, ensuring a unified criterion for determining where and when the temperature and acceleration will occur. This is based on the path allocation coefficients given in step one. Average speed of gate Define the terminal temperature prediction quantity :

[0088]

[0089] Among them: terminal temperature prediction : The predicted temperature at the end of the thin zone at the end of the filling process, used to constrain sudden temperature drops and the risk of cold shut-off; a positive real number; a set of partitions. : The set of temperature control zone indexes for fixed / moving models, a finite discrete set; path allocation coefficients The filling path is in the partition. The influence weights satisfy and Project the geometric path information onto the partition;

[0090] Zone temperature settings Partition The set temperature, a positive real number, is used to control the local thermal boundary; weight transfer. Partition The thermal impact weight of the terminal temperature is a positive real number, which is used to quantify the importance of the heat conduction / convection channels from the zone to the terminal.

[0091] Average gate speed The local average flow velocity at the gate over time is derived from the injection control function and is a positive real number; the time kernel... The time-response kernel of the thermal-dynamic input in the gating zone to the end temperature is a non-negative function, reflecting the physical property that the speed input is more effective closer to the end of the filling process; filling end time. : The time point at which the filling process ends, a positive real number.

[0092] To ensure full transparency, time verification Using exponential form:

[0093]

[0094] Among them: kernel amplitude value : Amplitude coefficient of the time kernel, a positive real number, representing the overall weight of the gate speed input on the final temperature; Attenuation coefficient : Time decay rate of the time kernel, a positive real number, representing the lag in heat / momentum transfer from the gate to the end; other symbols , , , , , , The meaning and scope are the same as above.

[0095] When using it, firstly, the terminal temperature prediction quantity. By merging zoning settings, path effects, and injection timing into a single controllable quantity, the combination of temperature rise and rate rise has verifiable endpoint significance; secondly, the exponential kernel suppresses the excessive influence of early high velocity on the terminal phase, thereby encouraging moderate rate rise in the later stages to maintain the temperature of thin zones; thirdly, The explicit product locks the geometric path information and partition weights onto the same channel, facilitating subsequent processing. As a common constraint for injection and partitioning.

[0096] In obtaining the predicted terminal temperature Afterwards, the injection control function needs to be enabled. A dual-objective evolution aimed at warming without gas entrainment. To this end, a system is constructed based on the risk of gas entrainment. With filling temperature margin The core objective functional is used, and gradient iterative correction is applied within the heat load upper limit given in step one. The renewal law is defined as:

[0097]

[0098] Where: injection control function The control variables on the injection side of the die-casting machine (such as the punch speed baseline) are a set of positive real functions that drive the volumetric flow rate. Average speed of gate The time sequence; iterative index : Number of iterations for offline / online updates, rounded to a non-negative integer; learning rate Update step size, a positive real number, controls the magnitude of each update; vortex weight. Temperature weighting The comprehensive weight from step one satisfies... , and with The sum is 1;

[0099] Gas risk : Injection control function Zone temperature settings The varying risk function for air entrainment, a non-negative real number, serves to measure free surface reflection and vortex intensity; molding temperature margin. : Injection control function Zone temperature settings The changing temperature margin function, a positive real number, is used to measure the insulation degree of the thin end region.

[0100] Regularity coefficient : The injection control smoothing term weights are non-negative real numbers, and their function is to suppress excessive jitter;

[0101] Reference Injection Curve The reference curve provided by the equipment or process is a set of real functions. In the die-casting machine control layer, it serves as the baseline / default injection control function—the time curve. It is typically used to provide the setpoints for the target speed or equivalent volumetric flow rate of the punch at each stage, for use in the actual injection curve. Alignment or as a regularization reference in algorithms;

[0102] In application, the objective functional transforms temperature rise and gas suppression into the same differentiable objective, and the update law can be implemented within the sampling frequency allowed by the device. of The regularization term constrains excessive approximation to the equipment's limits, balancing equipment lifespan and stability; because and Simultaneously affected Impact, updated injection control function It can be absorbed by the partition update law, thus forming a closed-loop link of first temperature prediction - then injection coordination - then partition refinement.

[0103] If end-stage heat preservation is addressed solely by adjusting the injection timing, the conflict between free surface stability and mold thermal fatigue may still surface in the path intersection area. Therefore, it is necessary to set the temperature for each zone within the constraints of the feasible thermal load zone. Differential back calculation is used to improve the filling temperature margin. Improvement and Reduce simultaneous occurrence and utilize zoned heat dissipation capacity Set a safety upper limit in advance.

[0104] First, the zone temperature setting is solved using a sensitivity-differential update method. The minimum correction amount is used to align the effects on the terminal temperature and the air intake; then, a safety threshold is used to hard-constrain the heat load-zone heat dissipation capacity to prevent heat load accumulation in the inlet area; finally, the updated zone temperature setting is... With injection control function The data is distributed to the mold temperature controller and injection controller, and the measured closed-loop data of the cavity temperature and pressure curve and vacuum flow rate are fed back into the sensitivity estimation.

[0105] To ensure that every update of the zone temperature settings is accurate Make positive contributions, With a negative contribution, a coupled differential update law is used to perform a single-step verifiable advancement on the temperature-gas entrainment two-dimensional index surface. The update law is:

[0106]

[0107] Among them: zone temperature setting : No. Partitioning during the next iteration Temperature setting, positive real number; learning rate : Partition update step size, a positive real number, used to control the update speed; Temperature gain coefficient :emphasize Gain weights, positive real numbers; air-entrainment suppression coefficients :emphasize The suppression weights are positive real numbers;

[0108] Partial derivatives , : Sensitivity of zone temperature to two indicators, real numbers, which can be obtained by comparing simulation / measurement data with small step size excitation; safety suppression coefficient Safety step size, a non-negative real number; upper limit of safe temperature. : The upper limit of the zone temperature determined by the boundary between the mold material / coating and thermal fatigue; a positive real number; maximum value operator. : Take the standard operator for the non-negative part to ensure that exceeding the limit is suppressed.

[0109] In application, on the one hand, the explicit definition of positive and negative signs ensures that each update improves performance. With reduction A definite direction is formed in the space; on the other hand, the combination of hardware and software in safety items is close to The update law automatically weakens and updates in real time, taking into account the mold life; in addition, since the partial derivative terms are directly derived from the local excitation-response of the corresponding partition, the update law has the engineering properties of being easy to implement and easy to trace.

[0110] To prevent heat buildup in the inlet or flange area from exceeding heat dissipation capacity due to temperature rise, the heat load density given in step one should be used as a guideline. Path allocation coefficient With partitioned heat dissipation capability Define the partition safety margin and perform hard threshold verification accordingly:

[0111]

[0112] Among them: safety margin Partition The difference between the heat dissipation capacity and the allocated heat load, a real number, representing the safety margin. Larger is safer; zoned heat dissipation capability : The effective heat dissipation capacity per unit area defined in Step 1, a non-negative real number;

[0113] Among them: partitioning In the time window Inner and surface areas heat flux density per unit area Calculate the area-time average to obtain the heat dissipation capacity of each zone. Partition set : Set of temperature control partition indices for fixed / moving molds (finite discrete set); partition area Partition Cavity surface area; time window (s) Start time (s): Evaluation window. The moment when this partition is first wetted by the molten metal is taken as... , Continue until the filling process is complete or until the zoned temperature control enters a stable phase;

[0114] Heat flux density per unit area Instantaneous heat flux on the mold surface; zoned heat dissipation capacity. : Average heat dissipation capacity of the zoned area over time; heat load density at the gate , Path allocation coefficient Main-charge streamline in partition The proportion, Mismatch Global heat load - heat dissipation Measurement, non-negative.

[0115] Heat load density Step 1: Sensible heat input per unit area at the gate, a positive real number;

[0116]

[0117] Wherein: melt density Alloy grade designation is fixed or determined by temperature; specific heat. The alloy temperature is a function of temperature, and the average value over the filling temperature range can be taken; the boiling point temperature... The melt temperature near the main runner inlet; mold temperature. The mold temperature at the mold surface where the gate is located or at a position 1–2 mm below it;

[0118] Volumetric flow rate The equivalent cross-sectional area is obtained by converting the injection curve or the displacement / velocity of the punch. , where is the effective flow area after flow reduction correction, and the safety margin threshold. : for partitioning The minimum allowable safety margin, a non-negative real number, used to set the exclusive condition for infeasible regions.

[0119] When applying it, firstly, the safety margin. The single-factor verification provides a direct criterion for whether the temperature of each partition can continue to be raised, avoiding blind searching in infeasible intervals; secondly, the path allocation coefficients are... Repurposed for safety determination, achieving secondary coupling of path-thermal-safety; again, safety margin threshold. The existence of this allows different partitions to be set with differentiated margins based on the importance of lifespan and appearance, thereby achieving engineered safety allocation.

[0120] Step 3: Risk of air pollution The spatiotemporal distribution is coupled with the thermal flow field baseline of step two to form the structure and timing of exhaust-overflow, thus completing the total equivalent exhaust area. With overflow volume Quantitative solution, and through the valve opening function The coordinated control with the target cavity pressure trajectory enables the gas suppression-heat preservation-connection process to be calculable, verifiable, and solidified throughout the entire filling process.

[0121] Although center feeding has shortened the critical filling path, the intersection of ribs and the flange retraction can still create local free surface rollback areas. If quantitative venting and timing coordination are not achieved on this path, the risk of air entrapment remains. This will be superimposed during the later stages of molding, increasing the probability of airtightness failure. Therefore, based on the information given in step two... , and First, map the peak periods and locations of vortex formations to exhaust windows, and then use this to deduce the total equivalent exhaust area. Then, using the valve opening function Linked with the cavity pressure trajectory, the exhaust volume and free surface stability are jointly optimized on the same target plane, thereby fixing the control law of when to open, how much to open, and how long to open into a repeatable timing curve.

[0122] A single-chain model following quantitative size-time update-stability tracking: first, calculations are performed using volume-time-pressure difference constraints. , and then The target cavity pressure trajectory is tracked as the sole execution variable, and within a safety margin... Minimize the risk of cyclone formation without weakening it. .

[0123] To avoid relying on guesswork and experience, a fixed volume of gas will be discharged within a limited time and under a limited pressure difference as a hard constraint. The total equivalent exhaust area will be calculated by back-calculating the equivalent orifice plate model of a compressible fluid. And the time-varying pressure difference obtained in step two This serves as the actual driver, ensuring a one-to-one correspondence between the required area and the device's capabilities. Definition:

[0124]

[0125] Among them: total equivalent exhaust area : Equivalent flow area of ​​all exhaust units in parallel, a positive real number used to guide the configuration of exhaust slots / valve inlets; average flow coefficient The overall flow coefficient of the exhaust path, considering inlet constriction and local losses, has a range of values. ;

[0126] gas phase density : Density of the mixed gas phase inside the mold cavity (including air / protective gas / water vapor), a positive real number; gas volume needs to be derived. : The volume of gas transferred to the vacuum side within the exhaust window, a non-negative real number; the exhaust time of the target. : The time window from valve opening to reaching the target residual gas threshold, a positive real number; pressure difference. The difference between cavity pressure and vacuum side pressure. , a non-negative real number;

[0127] Cavity pressure The time-varying pressure inside the mold cavity is derived from a mold cavity pressure sensor and is a non-negative real number; the vacuum side pressure... Vacuum system end pressure, a non-negative real number.

[0128] When applying, specify the volume of gas to be exported. Target exhaust time Pressure difference The engineering constraints of the three factors are mapped to a single area index. Eliminate the arbitrariness of placing slots based on experience; Integral core The time-varying pressure difference is converted into an equivalent exhaust capacity, which can be reused under different vacuum pump / piping combinations; The explicit introduction of this provides a unified calculation path for subsequent valve type replacements (pneumatic / electromagnetic / diaphragm valves) and fine-tuning of exhaust groove shape.

[0129] Even assuming the area meets the requirements, if the valve timing and opening degree are not coordinated with the evolution of the free surface, slag inclusion or secondary air entrapment may still be triggered. (Based on the injection control function) With zone temperature setting Given the free surface baseline as input, construct the valve opening function. The gradient-based update law enables it to track the target cavity pressure trajectory. And simultaneously reduce the risk of air entrapment. The definition is as follows:

[0130]

[0131] Where: Valve opening function The valve's opening degree over time, and its value range. 1 indicates fully open; iterative index : Number of steps updated online / offline, rounded to a non-negative integer; learning rate : Opening step size, a positive real number; weight , : Engineering trade-off coefficient between gas suppression and pressure tracking, a positive real number; gas entrapment risk : The risk function of air entrapment, which is affected by injection, zone mold temperature, and valve opening; a non-negative real number; cavity pressure. Same as above;

[0132] Target cavity pressure trajectory According to step two and The pressure curve required for a stable free surface, obtained by reverse calculation, is a set of positive real functions.

[0133]

[0134] Among them: baseline item The instantaneous pressure (Pa) at the vacuum end, obtained from the vacuum line sensor, is the baseline of the target curve.

[0135] Target pressure : The target curve sent to the exhaust valve controller; used to obtain... The update law is calculated Vacuum end pressure : Actual measurement by pump-side or vacuum chamber pressure sensor; gate heat load density : Defined as ;in The measuring aperture is the same as that mentioned above.

[0136] Smoothing time constant : Determines how far back in historical heat load to look; The smaller the value, the closer it is to the present moment; the larger the value, the more emphasis is placed on the cumulative input over a recent period; mapping coefficient : The proportionality factor for converting heat load density into pressure offset; terminal temperature prediction :according to Calculated; (K) sets the target for the process; positive part operator Guarantee only in A bias is generated when the temperature is below the target to prevent a reverse effect when the temperature is too high.

[0137] norm Absolute value integration over time ensures sensitivity to spike bias; pruning operator : Truncate the updated opening to Interval operators to avoid physical boundary overflows.

[0138] In application, firstly, the suppression of gas and pressure trajectory tracking should be integrated into the same differentiable target to avoid inducing backflow by drastically reducing pressure to suppress gas; secondly, the operator should be pruned. and The combination of norms suppresses high-frequency jitter and extreme opening, which is beneficial to valve seat life and sealing reliability; furthermore, it reduces the risk of air entrapment. Simultaneously subject to injection control function Zone temperature settings Impact, after the update This forms a stable complement to step two, thereby closing the three-domain control of heat-flow-exhaust within the same time coordinate.

[0139] If only the reduction of entrained air in the exhaust is considered while ignoring the solidification accessibility of the overflow reservoir, the feeding channel at thick-walled hot spots may be blocked by the early-condensing zone in the later stages. Simultaneously, if the overflow period coincides with a sudden change in valve opening, oxide films and inclusions from the free surface can easily be entrained into the overflow cavity, contaminating the feeding fluid. Therefore, the overflow volume... The geometric constraint coupling with the process column is a condition for liquid phase fraction-connectivity, ensuring the accessibility of sequential solidification and controlled feeding. Subsequently, the introduction of inclusions is suppressed by functional minimization of opening speed-pressure difference-inclusion work, making the overflow cavity a clean and reusable liquid storage end, providing a usable liquid source for the feeding geometry in step four.

[0140] First, use the liquid phase fractional field to... A lower bound is given, and the valve switching rate is constrained by a risk functional coupled with the opening change rate and pressure difference, ultimately forming a clean and connected overflow execution sequence.

[0141] To ensure that the overflow cavity can both store liquid and form isolation before the hot spot, the overflow volume is defined by the time integral of the liquid phase volume. The lower bound is defined, and the result is coupled with the minimum cross section of the process neck according to connectivity constraints, defined as follows:

[0142]

[0143] Where: magnification factor Liquid phase fraction Take the critical moment of replenishment Field value; effective volume shrinkage coefficient Mapping density mutations to volume requirements; solid / liquid density .

[0144] Overflow volume : Total volume of the end overflow cavity, a positive real number; amplification factor Process redundancy factor, taking into account heat transfer hysteresis and interfacial thermal resistance. Hot spot region The spatial subdomain of the thick-walled thermal node, determined by the part geometry and temperature field, is a bounded volume; liquid phase fraction. :point The proportion of liquid phase at that location, It takes a small value at the end of solidification as it evolves over time; Notation for spatial coordinate variables.

[0145] In application, the overflow volume is directly constrained by a liquid phase fractional field. The minimum requirements allow the overflow size to be determined based on actual hot spot volume rather than experience; it can be determined according to the hot spot area. The joint determination from the geometry and thermal field allows for rapid recalculation during engineering changes. When connected with the minimum flow section of the process column neck, it can ensure that the feeding connection during the sequential solidification stage is not interrupted by the early solidification zone.

[0146] If overflow is superimposed on a sudden change in valve opening, the oxide film on the free surface can easily be torn and drawn into the overflow chamber. The work done by multiplying the rate of change in opening by the differential pressure is used as the proxy for the risk of inclusions, and the valve switching speed is constrained by minimizing this functional to achieve preventative control of inclusions. Definition:

[0147]

[0148] Among them: slag inclusion risk functional : Measures the risk of membrane rupture and entrainment driven by pressure differential caused by changes in membrane opening; a non-negative real number; rate of change in membrane opening. : The first derivative of valve opening with respect to time, a real number; differential pressure As defined above, non-negative real numbers; time intervals. : Valve switching range that overlaps with overflow execution, satisfying .

[0149] In application, firstly, the risk of inclusion is quantified using functional measurements in the energy dimension, thus quantifying the engineering intuition of slow opening / slow closing; secondly, The square term suppresses frequent small oscillations and avoids high-frequency shear layer rupture; furthermore, due to pressure difference... It can be determined by cavity pressure With vacuum end pressure Real-time measurement of the functional of slag inclusion risk The upper limit of the valve actuator's acceleration can be evaluated and constrained online to ensure the cleanliness of the overflow chamber.

[0150] Step 4: Within the unified coordinate system of feeding-zoned mold temperature-venting / overflow established in Steps 1 to 3, give the geometric and energy balancing rules for the connection between the process column and the neck, and feed back the quantitative results of the trial production test to the parameter vector through a calculable loss function and a regularized update law, thereby solidifying the linkage of the three objectives of shrinkage compensation-cleaning-life into a transferable standardized parameter package.

[0151] The solidification lag of thick-walled hot spots necessitates feeding, but whether feeding is achievable depends on two things: whether the effective heat capacity of the overflow cavity and process column is sufficient to maintain the available liquid volume; and whether the neck connection structure can maintain the solidification sequence of opening first and then blocking in the time dimension.

[0152] Therefore, a criterion for the ratio of heat capacity to hot spot heat is first established in the energy dimension to infer the volume and location of the process column and overflow cavity; then, the solidification end time difference between the neck and the hot spot is constructed in the time dimension to coordinate with the zone temperature setting in step two. and the overflow volume in step three Coupling in the same direction ultimately unifies the three-domain design of geometry, heat, and time under verifiable inequality constraints.

[0153] In engineering implementation, if the equivalent heat capacity of the overflow cavity and process column is insufficient, even if the feedstock is connected, it will fail due to early cooling. To avoid inefficient parameter tuning where the overflow cavity is enlarged but still cannot feed the hot spot, an energy ratio criterion is proposed to uniformly measure the heat support capacity of the process column and overflow cavity for the hot spot:

[0154]

[0155] Wherein: heat capacity matching ratio : A dimensionless ratio, a positive real number, measuring the energy supply capacity of the feed end (process column and overflow cavity) to the heat exchanger; equivalent heat capacity of the process column. The equivalent heat capacity of the process column before solidification, consisting of the total sensible and latent heat, is a positive real number; the equivalent heat capacity of the overflow cavity is... The equivalent heat capacity of the liquid stored in the overflow cavity, representing the total sensible and latent heat, is a positive real number; the total heat required for the thermal block is... : The total sensible and latent heat released in the hot spot region from the start of solidification to complete solidification, a positive real number; minimum matching threshold. : The lower limit threshold that guarantees the achievable compensation, a positive real number.

[0156] In application, firstly, the judgment of whether there is enough feed is elevated from empirical geometry to energy balancing, so that the overflow volume... The setting is tied to the actual heat demand of the thermal section; secondly, it is achieved through the heat capacity matching ratio. The lower limit control can be related to the overflow volume in step three. And the zone temperature setting in step two This creates a unidirectional adjustment, reducing ineffective mold trials; furthermore, this criterion provides a unified benchmark for adjusting the volume of process columns and optimizing the position of overflow cavities, facilitating layout trade-offs on a limited mold surface.

[0157] Even if the heat capacity meets the requirements, premature solidification of the neck will cut off the feeding channels; conversely, prolonged liquidation of the neck will lead to collapse or surface tempering defects. Therefore, a time window constraint is established based on the time difference between the end of solidification of the neck and the hot spot, quantifying the solidification sequence from the hot spot to the neck.

[0158]

[0159] Among them: the time of completion of neck coagulation : The moment when solidification ends at the narrowest section connecting the neck, a positive real number; the moment when solidification ends at the hot spot. : The moment when solidification of a representative section of the hot spot ends, a positive real number; the safe time window for feeding. : The minimum time interval that ensures the replenishing fluid can still be accessed at the end of the hot spot solidification period, a positive real number.

[0160] In application, on the one hand, a time window is used to combine geometry (neck slenderness ratio) and thermal ( )-flow( Locking them to the same constraint ensures that both temperature increase and speed decrease actions revolve around it. Precise implementation; on the other hand, the constraints naturally exclude extreme morphologies such as excessively wide / thin necks, reducing the risk of residual stress during subsequent machining; furthermore, it relates to the valve opening function in step three. After being connected in series, a gentle pressure differential can be maintained during the overflow phase, avoiding the need for a safe compensation time window. Internally induced free surface tearing.

[0161] The accessibility of the shrinkage geometry needs to be verified through real samples; indicators such as CT pore volume fraction, helium leakage rate, cavity temperature and pressure curve and metallographic shrinkage grade can provide feedback from three dimensions: bulk phase, air tightness and process.

[0162] Therefore, we first construct a loss function that unifies the multi-source quantization quantities of shrinkage risk, insulation margin, and end temperature into the same loss surface; then, under the constraints of safety and consistency regularization, we use regularized gradient updates to backfeed and correct the parameter vector, so as to achieve a single convergence with a small number of samples and high determinism.

[0163] To avoid substituting a single indicator (such as CT scan) for the actual density and airtightness, a closed-loop loss function is constructed, unifying the key detection quantities and the core criteria of steps one through three into a differentiable objective:

[0164]

[0165] Among them: closed-loop loss The objective function for collecting information on phase, airtightness, and thermal flow processes is a real number, and the objective is to minimize it; CT pore volume fraction. : Pore volume fraction obtained from CT segmentation of the sample Leakage rate Leakage rate obtained by helium detection or differential pressure method, a non-negative real number; risk of shrinkage cavities. Risk measure based on solidification sequence and feeding connectivity, non-negative real number; filling temperature margin. : Positive temperature difference of the thin end region relative to the critical temperature, a positive real number; predicted end temperature. The predicted terminal temperature, defined in step two, is a positive real number; weights. : Engineering trade-off coefficient, a positive real number.

[0166] In application, multiple feedbacks from shrinkage, porosity, and airtightness are pressed into the same loss surface, ensuring a unique optimization direction and predictable convergence; a filling temperature margin is introduced. With terminal temperature prediction Incorporating process-level information into the evaluation avoids misjudgments that focus solely on results without considering the process; loss patterns can directly drive parameter updates, reducing the subjectivity of human judgment.

[0167] To ensure that parameter adjustments do not exceed safety and consistency boundaries, regularization constraints are applied to the closed-loop loss, and a gradient-based update law is used for backfeeding correction.

[0168]

[0169] Where: parameter vector The set of process-structural parameters to be corrected by reinjection, including the equivalent cross-sectional area of ​​the gating gate. Zone temperature settings Valve opening function Overflow volume Injection control function Its range is the direct product space of the feasible regions of their respective projects; iteration step size : Parameter update learning rate, a positive real number; gradient operator The Fréchet gradient over the parameter vector serves to determine the steepest descent direction; regularization weights. : Adjusts the weight of the regularization term relative to the loss term; a non-negative real number; regularization function. : Differentiable penalty term for safety and consistency constraints, a non-negative real number.

[0170] To ensure full transparency, an engineering implementation of the regularization function is provided:

[0171]

[0172] Where: weight : The project weights of each regularization term, non-negative real numbers; partition set Temperature control partition index set, finite discrete set; upper limit of safe temperature. : The upper limit of the zone temperature given by the mold material / coating and thermal fatigue constraints, a positive real number;

[0173] Filling end time : The time of completion of filling, a positive real number; the rate of change of opening degree. : The first derivative of valve opening with respect to time, a real number; minimum equivalent exhaust area Step 3 , , The lower limit of the area obtained by reverse calculation is a positive real number; the total equivalent exhaust area : Equivalent flow area in the context of parallel exhaust units, a positive real number.

[0174] In application, firstly, the update law treats safe temperature, valve smoothness, and venting capacity as softened expressions of hard boundaries, avoiding the pursuit of purely theoretical goals. Minimize the impact on mold life or vacuum capability; secondly, parameter vectors. The components correspond one-to-one with the key variables in steps one through three, ensuring that the closed-loop update can be seamlessly absorbed by the original process; finally, the continuous... Regular explicit suppression of valve acceleration, superimposed with the slag inclusion risk functional of step three. This can further reduce the probability of overflow pollution.

[0175] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0176] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0177] In the several embodiments provided in this application, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0178] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0179] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the scope of the technology disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A semi-solid die-casting method for magnesium alloy vibration damping tower with central feeding zone and coordinated mold temperature, characterized in that: include, Determine whether to use a central feed or a multi-point distributed central neighborhood feed, complete the matching design of the main runner, branch runners, fan gates and injection curves, so that the gate cross-sectional area is consistent with the injection volume flow rate, and output the feed-gate-branch runner scheme for subsequent use; based on the filling path of the feed-gate-branch runner scheme, divide the fixed mold and moving mold into the entrance area, rib area, end area and flange area, set the mold temperature of each area, and link it with the injection curve timing to form the partition mold temperature setting and injection timing; An openable and closable exhaust valve, exhaust groove and vacuum interface are arranged along the air swirling line. An overflow cavity is set at the end and hot spot and connected to the process column. The total equivalent exhaust area and overflow volume are calculated, the valve sequence is determined and linked with the punch position and cavity pressure. A process column that can be machined and removed is set at the hot spot and connected to the neck structure, and forms a sequential solidification feeding channel with the overflow cavity. Trial production and testing were carried out, and the gate, zone mold temperature, valve timing and overflow cavity were refilled according to the test data and solidified into a parameter package; The equivalent cross-sectional area of ​​the gate is determined based on the shrinkage coefficient and is set synchronously with the segmented volumetric flow rate of the injection curve and the average velocity of the gate; the cross-sectional ratio, transition fillet and incident angle of the main runner and branch runner are determined according to the filling path constraints. The fan-shaped gate inlet fillet, throat thickness, and opening position are located at the geometric center of the part or its neighborhood; the injection curve is sent to the injection mechanism as a control parameter and is time-aligned with the punch position encoder. A comprehensive criterion is established based on the filling temperature margin, air entrapment risk, and shrinkage risk, and these are set according to project weights. Based on the comprehensive criterion, either center feeding or multi-point distributed center neighborhood feeding is selected, and the initial geometric combination of the gate and runner is screened simultaneously. The stage speed and switching time of the injection curve are set to correspond to the segment boundaries of the comprehensive criterion and recorded as the feeding-gate-runner scheme. The total equivalent exhaust area is calculated based on the target exhaust time, the time function of the gas volume and the cavity-vacuum pressure difference to be derived, and is used to determine the diameter of the openable and closable exhaust valve and the cross-section of the exhaust groove; the openable and closable exhaust valve and the exhaust groove are arranged in parallel, and the valve opening sequence is linked with the punch position, cavity pressure and target cavity pressure trajectory for control; the vacuum interface is connected to the vacuum box and connected to the pumping device through the pipeline. The lower limit of the overflow volume is set by the spatial integral of the liquid phase fraction of the hot spot and the volume shrinkage coefficient at the critical moment of feeding, and the position and size of the overflow cavity are located by the end of the filling path. The process column is connected to the overflow chamber via a neck, and the neck has a narrow section; the valve switching curve satisfies the constraint of the product of the opening change rate and the differential pressure, and the upper limit of the opening change rate and the differential pressure sampling are used for curve generation; Collect CT pore volume fraction, leakage rate, cavity temperature and pressure curve and end temperature data, construct closed-loop loss and safety consistency regularization including the above quantities, and perform parameter backfilling and version storage for gate equivalent cross-sectional area, zone mold temperature, valve opening function, overflow volume and injection control function. The parameter reinjection is performed under the constraints of the zone's safe temperature limit, valve smoothness, and minimum exhaust area. The parameter package includes geometry, timing, and thresholds and is associated with the part number.

2. The semi-solid die-casting method for the magnesium alloy shock absorber tower with central feeding zone and coordinated mold temperature as described in claim 1, characterized in that: The boundaries of the inlet area, rib area, end area, and flange area are determined based on the filling path and path allocation coefficient. In each area, at least two combinations of hot rod, oil temperature or water cooling, and heating plate are selected to configure the temperature control channel and temperature setting. The injection curve is divided into three segments: the initial segment, the middle segment, and the final segment, and each segment corresponds to a specific mold temperature setting table. The settings are then sent to the mold temperature controller and the injection controller.

3. The semi-solid die-casting method for the magnesium alloy shock absorber tower with coordinated central feeding zone mold temperature as described in claim 1, characterized in that: Time convolution is performed on the average gate velocity characterized by partition weight, partition temperature setting and time kernel function to determine the end temperature prediction, and the injection curve is iteratively updated with the end temperature prediction as a constraint. Each partition is assigned a safe temperature limit and associated with the feasible range of its heat dissipation capacity, serving as a constraint for subsequent exhaust and overflow. The iterative update is executed in time slices within the control cycle, while preserving the temporal mapping relationship between the injection segment and the partition number.

4. The semi-solid die-casting method for the magnesium alloy shock absorber tower with coordinated central feeding zone mold temperature as described in claim 3, characterized in that: A heat capacity matching ratio threshold is set, and the volume, position, and material thermal properties of the process column and the overflow cavity are used for equivalent heat capacity calculation to balance the heat required by the hot spot; the minimum cross section and length of the neck are calculated according to the solidification time window in sequence, and the feeding channel is arranged along the centroid direction of the hot spot; the process column is a machined and removable structure and is provided with a breakable machining allowance, and the end face of the process column is flush with or lower than the parting surface.

Images