In-Situ Monitoring: Process monitoring in real time
Melt pool monitoring and optical tomography on the way to freedom from errors.
1. Introduction: Flying blind was yesterday
In the early years of metal 3D printing, the printing process was like a black box: you started the laser, waited 40 hours, and prayed that removing the powder would reveal a defect-free component. If an error happened on shift 2,000, you would only find out about it days later during the X-ray or CT scan.
This “trial and error” is unacceptable in aviation or medical technology. The solution is called In-Situ Monitoring (ISM): permanent, high-resolution optical and thermal monitoring of the printing process in absolute real time.
What is the goal of in-situ monitoring?
The ultimate goal of ISM is real-time quality assurance. The machine should detect defects (such as pores, lack of fusion or warping) exactly in the millisecond in which they occur, map the defect in 3D software and - in the future - even repair it independently.
2. Melt Pool Monitoring
The heart of 3D printing is the melt pool, the tiny liquid metal area under the laser focus. During melt pool monitoring, ultra-fast diodes or high-speed cameras (coaxial, i.e. through the same optical path as the laser) look directly at the melt pool.
- The sensors measure the size, shape and infrared intensity (temperature) of the melt pool with over 10,000 images per second.
- If the melt pool suddenly becomes too small (e.g. due to a lack of powder or soot on the lens), there is a risk of a binding error. If it is too large (on corners or edges), there is a risk of heat build-up. The system recognizes these deviations immediately.
3. Optical tomography (OT) and powder bed cameras
In addition to the selective view of the laser, the entire shift is monitored.
- Powder Bed Monitoring: A high-resolution camera on the ceiling of the build space takes a picture after each squeegee process. The system analyzes the image using AI algorithms and immediately detects grooves, uneven coatings or bulging component edges (protrusions) that could damage the squeegee next time.
- Optical tomography (OT): Special infrared cameras (e.g. EOSTATE from EOS) photograph the heat distribution of the entire layer directly after exposure. They "scan" the layer for thermal anomalies (hot spots or cold spots), which are clear indicators of microscopic pores.
4. The flood of data: The big data problem
The biggest challenge in in-situ monitoring is not the sensors, but the amount of data.
A melt pool monitor generates terabytes of data over a multi-day print job. If the machine has four or eight lasers, the volume multiplies. It is impossible for a human to evaluate this amount of data. It requires massive edge computing power and trained machine learning algorithms directly on the machine to filter out exactly the five pores that are critical for the component from billions of data points.
5. Conclusion: The path to a self-repairing machine
In-situ monitoring usually only documents errors in order to break off faulty components early and save powder (scrap reduction). However, the holy grail of the AM industry is “closed-loop control”: When the system detects a pore, it stops, the laser moves back and simply melts the tiny spot again with changed parameters. When this technology is ready for the market, the CT scan will become a waste of time.