A System for Automatic Air Void Analysis

The fully automated appliance automatically detect the air void content of a properly prepared sample.

The Method

By taking pictures with light coming from different directions and stacking them together the result is a picture as shown below. This is the same detail as on the first page. One can very well see the air voids  with the (mostly) triangular shade in the middle.

Now the camera follows a virtual line across the whole sample and records the detected air voids. Just like you do in the manual proceeding.

The blue colour of the sample as you can see first page is the result of a pH-indicator. A valid air void must reside in a mostly blue neighbourhood.

Also the roundness of a possible air void is taken into consideration.

Additionally the shade can also be used for detecting the depth of a air void. Especially below approx.  <50µm this is used to reduce the influence of the surface roughness.

The whole procedure works with approximately 0,6 m/h.

The method is patent-registered in Austria under A 1750/2009 .


Basically every sample that can be used for manual analysis can also be used for Atlas. However Atlas relies on the same requirements on a usable surface.
It is very important that the borders of the air voids are intact. However this is a hardly to specify criteria.  So this depends mostly on the practical experience of the testing engineer.
A possible waviness or obliquity is no problem.
In contrast to the manual proceeding it is very important to remove any residues from the preparation process.
The so prepared sample has to be dried before getting sprayed with the ph-indicator which is responsible for the blue colour.


After placing the sample on the xy-stage, Atlas starts detecting the size and height of the sample. The profile of the surface is very important to focus the cameras during the measuring process.
During the measuring process 2 cameras are used. One for a section of 5×7 mm and a second one for a section of about 0,4×0,6 mm.
Each camera is responsible for a different size of air voids. Of course the distance between the 2 Cameras is exactly known in order to avoid doubled detections.
After taking 3 pictures with different light sources of every position the xy-stage moves on to the next position.


The .csv log files containing the parameters of each found air void can be imported into Excel for creating the report. Additionally you have the choice of directly transferring the results into the Baula/LabMan application online.
The results can also be used to detect the variability of the air voids. This is done by evaluating 9 virtual lines and finding a mean value as well as min/max value. Placing the lines in different places can cause a variability of up to +/- 1% of air void content.
For each test Atlas uses approximately 7000 pictures per meter.
All the pictures can be stored and archived for documentation or later “offline tests” if needed.


  • Unchangingly objective performance.
  • Repeatable results.
  • Almost no manpower requirements.
  • No human influence