// Showcase

The process window for laser soldering a vacuum cell

Most soldering processes have one temperature requirement: get hot enough.

01Starting point

Most soldering processes have one temperature requirement: get hot enough. This one has three numbers, and two of them contradict each other. The solder has to pass 309 °C. A glass bond in the same component must never exceed 250 °C at any point in the process - that is 60 % of its glass transition temperature of 450 °C, above which stresses build up in the part. And then came the third number, which decided everything: because of a surface roughness of 3 to 4 µm and the very small contact area between the solder ring and the metallisation, there is almost no thermal transition. Making things hotter does not help if the heat never reaches the place that needs it.

02Approach

First without a laser. Ceramic housing on a metal sheet, heated from below with a hot-air blower set to 500 °C: an ordinary lead-free solder with a 230 °C melting point flows immediately, while the solder rings stay unchanged. At 650 °C they soften - and still do not wet. Then we filled the inner volume of a ring with a flux-containing solder: wetting immediately visible. The flux solder melts first and substantially improves the thermal transition. That settled it - the problem was not the temperature, it was the contact. After that we measured instead of estimating: a matrix of three substrate temperatures against three laser powers, five-second exposure, four sensors, and the same runs compared against the simulation.

03Result

Hard numbers instead of assumptions. At 120 W the temperature rises about 100 K above the starting value - and by almost the same amount whether the substrate sits at 25 °C or 120 °C. The maximum occurs exactly at switch-off, and 22 seconds later 86 of those 104 K have already gone. There is no afterglow to rely on. The highest value measured anywhere in the campaign, 216.6 °C, is 93 K short of what the solder needs. And the simulation deviates by only 3.3 to 5.5 K at the hottest sensor, with a constant sign - a model that can be carried forward with a known correction.

What's inside

A process window with documented bounds at both ends, a diagnosis of the thermal contact from a bench trial without a laser, a measured temperature response across nine operating points, and a simulation verification with the deviation stated per measuring position.

And what happened next

The direction changed: not more power, but a warmer start and a longer, less dense exposure. The later runs in the archive are set up for a 150 °C substrate, 120 W and 16.5 s. The trial run at double the exposure had destroyed the metallisation - and the report attributes the damage explicitly to the high energy density, not to the amount of energy. That is precisely what makes the longer, gentler exposure the logical next step.

All showcases

Get in touch

We are ready for new challenges

You bring the challenge. We develop the solution. From the first idea through concept and development to the validated machine, we guide your project.