The fate of a casting is often decided in the few seconds it takes for the mold to fill. If the speed is slightly too low during that brief moment, the filling will be incomplete; if the transition point shifts, cold shuts occur; if the compaction is weak, the porosity increases. The result is familiar: spatter, flash, excess swarf, scrap, downtime, delivery pressure.

That's why real-time shot control means more than just "monitoring" in foundries. The system measures data during the shot, compares it to the target profile, detects deviations, and makes corrections within the same cycle. In short, the process keeps itself more consistent.

In this article, we will clearly explain what real-time firing control actually controls, which parameters it monitors, how closed-loop logic works, what results are felt in the field, and how it is implemented at the facility. Finally, we will discuss how to measure success.

What exactly does real-time fire control control?

The goal of the shot is simple: fill the mold cavity each time with the correct speed, correct pressure, and correct timing. The difficult part is this: You look at the same settings screen and think you are making the "same shot," but the conditions change slightly with each cycle. Metal temperature, mold temperature, lubrication, cycle time, piston friction, even the way the ladle is filled all have an effect.

Real-time shot control makes this variability more manageable. The word "control" here does not simply mean drawing a graph. The system takes measurements and generates commands via the control unit to bring the shot profile closer to the target. This approach is very common in pressure casting, as the process proceeds with injection-like logic. The speed, pressure, and position curves, known as the "shot profile," determine the repeatable quality of the part.

 

Break down the firing cycle into parts: slow loading, rapid loading, consolidation, holding

Think of shooting like a car journey. You drive slowly in the city, then speed up when you get on the main road, and as you approach your target, you "settle" the vehicle and steady it. Shooting has similar phases:

• Slow filling (1st stage): Ensures the metal enters the nozzles more gently, helping to reduce air entrainment.
• Fast filling (2nd stage): Fills thin sections and remote areas in a timely manner, reducing the risk of freezing.
• Intensification: Suppresses shrinkage during solidification, reduces the risk of porosity.
• Hold: Maintains pressure for a short time, contributing to measurement stability and surface quality.

One of the most error-prone moments is the phase transition (switch-over). For example, if the transition occurs too early, the metal accelerates, turbulence increases, and the gas may close. If it is delayed, the metal cools inside the mold, increasing the risk of cold joints and incomplete filling.

The fundamental parameters the system looks at and their practical implications

Real-time fire control generally focuses on several key variables. Each of these variables is directly linked to a problem observed in the field:

• Speed (piston speed): Low speed can cause incomplete filling, while excessive speed can lead to gas lock and splashing.
• Pressure (hydraulic and metal pressure): Insufficient pressure weakens compaction, excessive pressure causes chipping and mold strain.
• Position (piston position): Clarifies the transition point. Incorrect positioning may trigger cold welding.
• Time (phase durations): The same setting does not produce the same result with a different cycle time. If the time is extended, the metal cools down.
• Metal temperature (if applicable): Low temperatures make filling difficult, while high temperatures affect surface and mold life.
• Mold temperature (if applicable): Low mold temperature accelerates freezing, while high mold temperature may increase the risk of flash and sticking.

It's not enough to make each of these parameters "good" individually. The real challenge is to keep them all consistent within the same shot.

How does the system work when the mold is filling, and which sensors and data does it use to make decisions?

What happens during mold filling occurs in the blink of an eye. That's why real-time shot control collects data quickly and responds rapidly. The basic flow generally proceeds in the following order: the sensor measures, the control unit processes the data, compares it with the target curve, calculates the deviation, and then corrects the movement with valve and driver commands.

The idea here is similar to a cruise control system in a car. When you hit an incline, you press the gas pedal a little harder. The system detects changes that appear like an "incline" during firing (heat, friction, viscosity) as deviations, then adjusts the profile to bring it closer to the target.

The sensors used vary depending on the machine, but common examples include: piston position sensor, high-resolution encoder for speed calculation, hydraulic pressure sensors, and in some applications, mold-internal pressure and temperature sensors. The control unit usually communicates with a PLC. Some facilities also keep records on the MES side, so that quality data and production data are stored in the same place.

 

What is the difference between open-loop and closed-loop systems, and why does real-time provide more consistent results?

It's easier to see this difference with a brief comparison. The table below summarizes the most common distinction in practice:

Subject	Open-loop	Closed-loop
Logic	The setting remains fixed; the system only applies it.	The system measures, finds the deviation, and corrects it immediately.
Variable conditions	When conditions change, the error may increase.	Even if conditions change, strive to reach the goal
Operator load	Frequent adjustments and testing may be required.	The time spent searching for settings decreases, standardization increases
Consistency	It is more dependent on the shift and the operator.	It gives more consistent results for the same piece.

The critical message at this point is: Each shot looks the same, but the conditions are not the same. The mold is slightly warmer, the lubrication is a bit different, the cycle is a few seconds longer. The closed-loop approach attempts to catch these small differences before they become significant.

How is the target profile (reference curve) determined and updated?

The target profile is not a copy of the "best part." Rather, it is the controlled limits of the process that produces the good part. Generally, the following method works: first, a stable production range is found, then this range is used as a reference, and finally the tolerance window is defined.

It is not right to do this once and then shelve it. The profile must be revalidated during product changes, mold revisions, travel changes, and even different alloy batches. Otherwise, the system will run towards the wrong target while trying to correct the error.

To make it practical in the field, the team can proceed with the following questions: Under what conditions was the reference part approved? What were the metal and mold temperature ranges? Where was the transition point located? What were the densification pressure and duration? Was there any scrap in that cycle? If these records are kept regularly, updating the profile ceases to be a "guess" and becomes a controlled process.

What concrete results can be seen from increased sensitivity and efficiency?

Real-time shot control is most evident in two areas: quality deviations decrease and production becomes more predictable. It would be inaccurate to cite specific percentages here, as the outcome depends on part geometry, mold design, and equipment condition. Nevertheless, the metrics you can track within the facility are clear.

The following metrics take improvement beyond being "noticeable" and make it measurable:

• Scrap rate (by cause breakdown)
• Re-processing and deburring time
• Cycle time variation (simple indicators such as standard deviation)
• Mold cleaning and mold opening frequency
• Setup and initial part approval time
• Downtime (especially process-related downtime)
• Energy consumption (per part or per cycle)

When you gather these metrics on the same dashboard, the phrase "the machine is working well" takes on concrete meaning.

Less scrap and less rework, the root cause of errors becomes apparent

Pores, cold joints, underfilling, and flash are often seen as "results." However, process data provides the opportunity to trace the chain of causes backwards. For example, if pores have increased, it is not enough to look only at the vacuum. Fluctuations in the speed profile, shift in the transition point, or a drop in the compaction pressure can also lead to the same result.

The advantage here is this: You don't deal with the symptom, you deal with the cause. When the quality team and the production team see the same curve, the discussion doesn't drag on. Instead of "Which shift did it happen on?", the question "Which phase did it deviate in?" comes to the fore. This also increases decision speed.

Once you start regularly reading the firing data, many errors cease to be "coincidental" and become a recurring pattern.

More stable cycles, better planning, and fewer stoppages

As repeatability increases, the operator's "search time" decreases. Quality fluctuates less during shift changes. Process memory is established for the same mold and the same product. This also makes planning easier, as cycle time and scrap become more predictable.

Additionally, firing control can indirectly indicate certain mechanical issues early on. For example, if the pressure requirement continuously increases while targeting the same profile, friction may have increased. Similarly, abnormal vibration or irregular speed response can be considered indicators for maintenance checks. The goal here is not to "predict failure," but to avoid missing small signals.

A practical roadmap for commissioning at the facility: where to start?

Implementing real-time firing control is not a one-time setup task. It is more of a discipline that starts small and expands with the right KPIs. The best place to start is on a single line, with a single part family. This allows the team to learn how to read the data and clarify the control strategy.

A practical roadmap proceeds as follows: first, a pilot is selected; then, the sensor and calibration are verified; next, a reference profile is created; tolerance windows are defined; training is provided; KPI tracking is initiated; and finally, a standard operating procedure (SOP) is written. Involving the maintenance team from the outset makes a significant difference, as measurement reliability is directly dependent on mechanical health.

A guide to selecting press machines can also be a resource that completes the decision-making process regarding the impact of machine and equipment selection on foundry performance.

Things to consider when selecting a pilot: the right part, the right mold, the right target

The "most problematic part" for the pilot is not always the right choice. A very chaotic part may indicate a problem with the mold design or metal preparation rather than the system. Instead, choose a job that has high variability but a measurable problem.

A good pilot generally has the following characteristics: production volume is sufficient, quality criteria are clear (e.g., porosity limit or flash target), the measurement method is defined, and the same mold remains in use for a long time. Define the target from the outset. For example, reduction in scrap causes, shorter first part approval time, and lower cycle variation.

Common mistakes and how to avoid them

The most challenging aspects in the field can be summarized under a few headings:

• Neglecting sensor calibration: If the measurement is incorrect, the control will not function properly. A periodic inspection plan is required.
• Assuming a one-time setting is standard: The profile, product, and mold must be revalidated as they are used.
• Setting the tolerance window too narrowly: The system generates unnecessary alarms, and the team becomes fatigued.
• Setting the tolerance window too wide: Errors increase, quality fluctuates again.
• Just look at the pressure: Location and transition point are decisive in most errors.
• Making data entries irregular: A prerequisite for root cause analysis.
• Not involving the maintenance team in the process: Mechanical deviations directly impair control performance.

These measures may seem simple, but when applied with discipline, they determine the fate of the project.

Measure during firing, see the deviation, correct it immediately

The essence of real-time fire control is this: Measure while firing, compare with the target, detect deviation, then correct within the same cycle. When this approach is properly implemented, three key benefits emerge:

• Consistency increases, results become more consistent from shift to shift.
• Scrap and recycling decreases, because it appears earlier.
• Production stability improves, making planning and maintenance easier.

Now the clearest step: Select a pilot part, collect the existing data regularly, determine the target profile and KPIs, then roll out the system step by step. Those few seconds in mold filling no longer have to be full of surprises.