Inkjet straightness sensing

Inkjet printing employs thousands of tiny actuators in communication with ink to eject droplets from a nozzle. Images are printed by dividing them into pixel lines, activating the corresponding actuators for each line at the right moment while the print product is moved under the printhead. The digital process eliminates setup costs and the non-contact nature allows printing on non-flat surfaces and 3D textures. This technology also reduces material waste by printing only what is needed. These advantages make inkjet printing an ideal choice for the project partners' main field of operation.

The inkjet process often suffers from reliability issues, such as deteriorated print quality due to individual nozzles failing to fire or firing at reduced performance. These issues are a result of the microscopic size of the nozzles and the challenge of maintaining high ink cleanliness levels. When a problematic state is detected, cleaning and maintenance procedures are required to resolve the issue, and in some cases, a printhead replacement may be necessary.

Some issues, such as missing parts of an image, can be easily detected by machine vision. However, other issues, like nozzles firing at reduced performance, can subtly impact print quality and are difficult to detect with machine vision. In these cases, experienced operators must inspect the prints after production. Unfortunately, failures are often only discovered during the final quality control stage, resulting in the waste of all products in the pipeline.

iPrint, a research institute, collaborated with Steinemann DPE and Polytype AG to develop a system that detects poorly performing nozzles during printing. The system analyses the acoustic response of the actuator in communication with the ink and determines deviations from an ideal acoustic response (or "fingerprint") of the fluidic part. This enables the quantification of ejection fitness in real-time, allowing for online detection of bad-performing nozzles during printing. At most one product is lost if a failure occurs.

A prototype was developed and thoroughly tested with an industrial printer, establishing statistics on detected and undetected bad-performing nozzles during normal operation. Analysis of many acoustic responses of the actuators provided insight into changes in response during malfunctions or bad performance, enabling targeted countermeasures for specific failure mechanisms. The system operates during the printing process and reports the status of all nozzles at short intervals between product prints, with a detection rate exceeding 98%.