Adjusting the Glue Line on an Edgebander

A thorough description of how to systematically adjust an edgebander glue line. June 7, 2011

As with any diagnosis and/or adjustment to a machine, the person making the adjustment assumes all responsibility for damage to the machine and/or person(s) and/or any loss in production time. Please read the entire article before attempting any of the below procedures. If you are unsure or uncomfortable with the processes, Technicians are available to perform this work.

The following is a short article on how to diagnose, confirm and adjust a glue station to minimize the applied glue line. This is only not an end all of processes to remedy a glue line issue, just one of many. We will assume the side presser section and all of its components are operating normal and no adjustment or tuning is needed. In general what we are looking for is one of three conditions.

1. Consistently consistent (good).
a) Adjust nothing.

2. Consistently consistent (bad).
a) Adjust the machine being tested.

3) Consistently inconsistent.
a) Re-evaluate the material being tested for possible issues in a prior manufacturing process.

Evaluating the glueline is very straightforward. Just place the tool over the edge and look at the appropriate scale (Item 1493T14 from McMaster Carr. This is the tool I use.). What we would like to see is a glue line as small as possible. I strive for a measurement of 0.1mm when using 0.5mm PVC. This will change depending on the edge product, thickness, and other factors.

From there, it is just a matter of making the correct adjustments to the machine. Letís assume that you do not have a pre-mill unit on your edgebander. If you have to ask what a pre-mill, chances are good that you donít have one. If your machine has a premill station, the below can be used to test the panel edge for perpendicular, but not for actual adjustment to the glue station. Due to the complexity of the addition of a premill, we will have to cover that at another time. Diagnosis of a glue line always starts with the panel. (I would dare to say any diagnosis, not just for the glue line).

1. How many machines supply the edgebander with parts? Mixing parts from different sources is very difficult to diagnose. Several part samples must be obtained from each machine. Top of the panel must be identified on each cut, as well as the panelís relationship to the blade or bit. This must be done for each machine.

2. During the diagnostic period, do not re-cut an edge on a different machine. Always use the same machine and relationship to the cutting tool.

I try to stack the panels together to the maximum thickness that can be processed through the edgebander. I.e. stack, fasten and glue 3 pieces of 19mm panel approximately 300mm long (12Ē) together to create a 57mm thick panel. (2 1/8" for those not comfortable with metric) The better you prep these panels the more accurate the result. I prefer MDF, then Particle Board. When I was at the factory, we used an Aluminum block to confirm the squareness of the glue station. If you choose to use a metal block, you do so at your own risk. Improper use could seriously damage your machine.

The added thickness will multiply any error you have, using a good quality machinist square, not a combination square, test the edge. It should be no more than 0.2mm out of perpendicular from top to bottom. To confirm, test the opposite face. A mechanics feeler gauge can be used for measurement.

If the panel is within specifications, introduce the panel to edgebander, topside up, with only the glue station activated. No trimmers, no edge product, nothing. (All we want is to apply glue, so back out your side presser rollers.) What we want to see is if the glueline is consistent from top to bottom. Rotate the panel and re-introduce it, this time with the topside down. If the panel is within specification, and the glue station is correct, then the glueline should be consistent on both tests.

You will want to repeat this test several times to create a history, donít assume one time is correct - noting the relationship of the panel to the edge, and the edge in relationship to the blade or bit. (If this is not clear, I can clarify later). Also, this must be done with every machine that supplies panels to the edgebander. I try to run at least 10 panels per machine. Looking at this history, you want to generate an average for each panel and each source machine.

1. If the glue line from is not consistent on the edge from face to face, there is something not correct with panels being supplied.
a) If they come from multiple source machines, it could be that you are producing a different type of square from each machine.

2. If the error is consistent in relationship to the ďupĒ face of the panel, as it is processed within the edgebander, it is probable that the glue station is not perpendicular to chain. Each machine has its own means of correcting this. HolzHers are notorious for this if the machine is used and/or been moved without properly supporting the glue station. This is an issue with how the machine was prepped for moving, not an issue with the design, and it can be easily corrected.

Making the assumption that the above is complete and acceptable, next look at the glue.

1. Is it the correct glue for the application?
a) Different substrates and edge products require different glues. Also, there are different glues for different machines.

2) Does it have the correct viscosity?
a) Typically the viscosity will thinner when the glue is at its correct temperature. Depending on your location in the world, you will need to adjust the glue temperature throughout the year. The glue will apply different in Ecuador than it will in Northern Alaska.

b) Running the temperature to low will cause the glue to be too thick to correctly press.

On the flip side, running it too high will not allow the glue to cool quickly enough and in some cases the top and bottom of the glueline will open slightly just as the glue cools. Make whatever mechanical adjustments are needed to correct any misalignments (adjusting the saw, changing the router bit, or adjusting the glue station). Now, cut some test panels that are 19mm (3/4Ē) thick, around 300mm long (12Ē) square. Apply the banding as per normal operation. Using the magnifying tool adjust the glue line until it is a light as possible and still give you the correct wetting of the panel and edge product, and the correct pulling of the core substrate. Below are some considerations for the panel preparation.

It is not uncommon to have multiple glue quantity settings for different substrates, although we would like to have only one.

1. The panels that you use to test need to be at least 60 degrees Fahrenheit. The time that the glue is exposed from its application to the moment that the edge product is applied called the open time.

a) If your machine has multiple speeds, does the glueline quality improve or deteriorate when changing speeds? If the glueline deteriorates then the open time may be wrong in relation to the temperature of the glue when it is applied. The applied temperature could be too hot as well as too cold.

b) A cold panel will also affect how fast the glue begins to cool.

i) Extremely cold panels will cause ice crystals to form on the edge of the panel when it is cut. The ice crystals will liquefy when the glue is applied and not allow the glue to adhere.

ii) Material that is stored in sub-freezing temperatures need time to acclimate to room temperature. Material pulled from the center of a frozen bunk will remain frozen at the core long after the material has been place into a climate controlled area.

c) Edgeproduct that sits over a heat source (the glue station), can affect the glue line dramatically.

i) If the area of the glueline changes in relation to the edge products preposition over the glue station, modifications may be need to allow the edge product to remain at a controlled temperature prior to being applied.

2. The length of time that an edge sits after it has been cut will matter also. Letís assume for now the edge is new or at least less than 24 hours old. Also, letís assume we have no chipouts or scoring marks. If the edge is produced on a router process, we will assume the bit to be new. As the bit dull, you will notice a pulling effect on the core of particleboard, which will produce a core that is proud in relation to the face of the panels. This is due to the type of cutting action being preformed. This is one of the down sides to using a router to produce the panel edge. It can usually be eliminated with sharp tooling.

Forum Responses
(Cabinetmaking Forum)
From contributor D:
Thank you for taking the time to post. It will take some time to digest and run the tests but I really appreciate the information. I am ordering the 1493T14, pronto.

From contributor L:
A peel test is invaluable with a Holz-Her.

Click here for higher quality, full size image