Q: In our dimension mill for the center panels on doors, we present the inspector with staves which are 8 feet to 14 feet wide. The inspector marks the unacceptable knots, cracks and other defects with a UV "pen." The staves are passed to a saw operator who (using an UV lamp for ID) cuts out the defective areas and creates shorter staves for subsequent processing (along with some amount of scrap). We then glue-up these staves into a panel downstream in the manufacturing process. In a different plant we use the "glued-up stock" and make a panel for a door. When we stain the doors (or our customers stain the doors), we often find cracks that were missed. In the short term, I'd like to optimize the lighting in the stave and panel inspection area. Can you recommend a reference that would suggest inspection lighting alternatives (ambient lighting, alternative light sources, etc.) that could be evaluated? In the longer term, I'd like to consider some type of a more automated process for crack/check identification. This process would have to be quick and non-invasive to the wood. Have you ever seen this done in your travels?

A: The wood is too strong and the time is too short to create any new checks, so we are concerned about old cracks only. As you have discovered, the old checks are often impossible to detect, especially if the lumber has been exposed to a humidity over 40 percent RH. The only way that cracks will show up better at inspection is to make sure that the pieces are exposed to very dry air for a few hours before marking. This dry air will result in a little bit of shrinkage, which will in turn open any pre-existing cracks. Of course, good lighting and alert inspectors (give them a rest period every hour) are also keys.

Incidentally, cracks are often closed tightly on the surface and open underneath, so as you sand, you will find open cracks or checks. This "open underneath the surface" characteristic can be used to detect the cracks. That is, there have been claims about ultrasonics being used to detect cracks that are difficult to visually discern, but I have never seen the system work on tightly closed checks. One system I saw was able to find small checks, but it also found lots of checks that were not there"false positive" might be a good description. As the system finds fewer false positives, it starts to miss more actual checks. So, the answer is that there is no system yet developed that I am aware of that will work as you need it to work.

It is true that a plant with high RH will have a very poor likelihood of finding cracks, as they will swell tightly closed. Of course, the tightly closed checks or cracks will reopen when the piece is exposed to dry air in your plant or at the customer's home or office.