Better quality in optics and geometric accuracy

When a material is processed with short light pulses, so-called smoke is produced, which is extracted as best as possible during processing. Due to the immense kinetic energy of the material particles, however, residual particles remain on the workpieces that cannot simply be removed using compressed air. However, the minimum tolerance requirements for geometry and roughness of the components can only be met if the manufacturing process is optimized and the actual surface is assessed. The residual dirt requirements that apply in modern manufacturing therefore demand a robust and repeatable cleaning technique for microcomponents manufactured using laser processes. "Even a small difference in geometric accuracy before and after cleaning can have an impact on the functionality of components that require low tolerances," Anton Pauli, managing director at laser micromachining expert GFH GmbH, explains the problem. In addition, such residues act as wear particles or - depending on the area of application - can cause damage by clogging throttles in injection systems, for example.

Foil tests provide information about ultrasonic effect
Since no cleaning process has offered a solution to this problem to date, the La- ser machine manufacturer and contract manufacturer GFH, which is continuously striving to improve the entire process chain, has taken up the cause. "If there is no specialist knowledge available on a topic, we develop it. Because in order to carry out successful laser micromachining, every single step must be optimally solved," says Pauli, explaining his company philosophy. Due to the fact that the ultrasonic cleaning process is best suited for the parts produced by laser micro machining, the following processes have been developed

The basis for the analysis is initially so-called foil tests. The holes torn in the aluminium foil placed in the tank allow conclusions to be drawn about the distribution and intensity of the ultrasound effect. "The evaluation revealed significant differences, which were important to know in order to carry out the subsequent study under consistent conditions, but also to achieve the best possible effect in daily use ", elaborates Barbara Schmid, who was responsible for the investigations at GFH. "In this way, we were able to check very fundamentally the functionality of the cleaning basin and place the element to be cleaned optimally in each case."

Algorithm detects minimal differences in cleanliness
For the subsequent parameter study, two component series of 200 pieces each were produced with the GL.compact laser micromachining machine developed by GFH: one made of stainless steel and one made of brass. These two materials were chosen because stainless steel is used very frequently and brass tends to discolour as well as causing a number of other problems during cleaning. The components with an edge length of 5 mm, small incisions and a hole all had the same geometry, so that the contamination was the same and the results were therefore comparable. Frequency, temperature, cleaning and rinsing medium, filling level, concentration of chemicals, duration of the actual cleaning as well as rinsing and drying were identified as relevant influencing factors, but accessories were also taken into account, such as various vessels in which small parts are placed that could otherwise be lost.

"All parameters were examined and evaluated individually, and we also validated the cleanliness under the microscope," reports Schmid. "In order to be able to show even the smallest differences, a special evaluation algorithm was developed and applied. This was based on various image processing procedures: the microscope images were first converted into grey scales. In order to be able to distinguish areas with laser smear - recognizable by dark traces - from the rest of the component, a threshold value was selected, and the dark pixels were extracted and counted. "For better interpretation, we classified these values on a scale, whereby a component directly after processing, which had a corresponding number of dark image points, was classified as 0 percent clean. The theoretical value of 100 percent thus corresponded to a component without dark pixels, i.e. free of any contamination," adds Schmid.

Process optimization through evaluation of the settings
For optimum results, the cleaning medium should be degassed at least 10 minutes before starting the procedure. A slightly acidic medium is generally suitable for brass, whereas an alkaline medium is suitable for stainless steel. It should be noted that because of the high temperatures during cleaning, some of the liquid evaporates. "Therefore, the filling level should be checked regularly and adjusted if necessary, because both too low and too high a filling level reduces the cleaning performance," says Schmid. If very delicate parts are being cleaned, glass beakers or plastic nets are recommended as vessels. More robust parts should be cleaned in a stainless steel basket.

Temperatures between 45 and 65 °C produce the best results - depending on the cleaning time - since, due to ultrasonic cavitation, the temperatures also rise with increasing time. The improvements are greatest after 15 min cleaning and 5 min rinsing. A longer duration of up to 45 min cleaning and 15 min rinsing achieves only minor improvements in comparison. The optimum cleaning frequency is the "dual frequency", which changes every 30 s between 37 and 80 kHz. For large objects or if several parts are being cleaned, it is advantageous to switch on the "sweep" mode. If the contamination is very stubborn, the "pulse" mode can lead to an improvement. Pre-cleaning is only necessary if there is oil or other grease on the components. During rinsing, a corrosion protection additive as well as a wetting agent help to improve the subsequent drying.

Optimized cleaning process becomes standard
The implementation of the knowledge gained led to an improvement in cleanliness from 74 to over 95 percent for the stainless steel series. "Unlike other processing methods, lasers do not use oil, coolant or grease, which also has an effect on the resulting contamination. We found that the main problem up to now was that the cleaning process had only been adapted to the material, but not to the previous machining process," Schmid sums up the initial situation. In order to optimise the internal laser machining processes and also achieve a visible improvement in quality for customers as well as higher geometric accuracy, measures have been implemented step by step since June 2016 to implement the improved process as a standard at GFH. If other materials come into play in a new project, the cleaning process will also be adapted accordingly. The response from customers to the overall package of machining and cleaning has been very positive throughout.