Small Chunks Sticking to the Edgebander Roller

Trying to get to the bottom of a performance issue with an edgebander: small particleboard core chunks sticking to the glue roller are starving parts of the joint for glue. November 15, 2010

We are continually having a problem with our bander. When parts come off the CNC the edges are blown off and then wiped off immediately before going through the bander. This is all particle board core melamine. Even with this attention to getting the edges clean our glue pot still gets backed up with small pieces of the core. This leads to starved lines on the roller where no glue is applied to the panel. We reverse the roller to back the chunks out and this works for a while, but they come back around shortly. This has turned the edgebanding process into a nightmare. Does anyone else have this problem, or any suggestions? We used to have a HolzHer bander and it seems like the cartridge under pressure would eliminate this problem. We are thinking of selling and going back to a HolzHer.

Forum Responses
(Cabinetmaking Forum)
From contributor D:
Are you perhaps running your glue roller against the direction of the panel? Our glue pot bander has this option, but we never use it, and we have never had the problem you're having. My thinking is that maybe the roller grinding against the panel is pulling chips off.

From the original questioner:
We have the roller running with the panel. We have tried it the other way with no noticeable difference. The guide shoe is set so that the panel does not actually touch the glue roller itself, it is set approximately .3 to .4 mm away from the panel edge to prevent wear on the roller.

From contributor K:
I have a granular glue pot machine also and have never had this problem. I know my machine is set up so the board presses against the glue roller. It actually moves the gluing section a little bit when the board hits the roller. The gluing section is mounted on a pivot to do this. This is my second machine that did this so I think it is supposed to do that. I would try changing the setup so that the board pressed against the roller and kicks the gluing station on its pivot a little bit.

From contributor Z:
Contributor K may have a good point. If there is pressure against the roller, then the chips cannot come off. My bander is also set up to touch the roller. As you said, the cartridge system applies pressure. So should the roller.

From contributor D:
How does the glue get to the panel? Are you saying you have a roller going around with .3-.4mm of glue on it and pass it with the panel so that the panel picks some up? That could, if true, be most of the problem. That thickness of roller coating could be a little too cool on the surface and be "sticky" before you want it to be.

Most glue pot set ups are:

1. Absolute minimum amount on the roller required to adhere. Too much glue is the enemy. The glue roller's purpose is to spread the hot adhesive on the panel edge. You don't want it cooling yet. Correct operating temperature for the glue being used is very important.

2. 1 to 2 mm movement on the pot and roller as the panel passes. You should not reach the full travel of the pot assembly.

3. Pinch the glued panel to the tape at the pressure roller. The glue should be cool enough to begin adhering at this point.

I would guess that you would be better served to set it up normally and not worry about the roller wear. If you wear it out, you can be proud of that production fact and it won't owe you a thing. I also read a long way back about the board flaking condition having a lot to do with bit selection, feed rates and cut speeds on the CNC given different cores and core densities.

From the original questioner:
The panel does push the gluepot back 1 to 2 mills as it passes, there is a guide shoe immediately in front of the roller that the panels contacts. The only time the panel touches the roller is the last 1/2'' as it comes off of the shoe and the pot returns forward. The gap between roller and board is set correctly according to Stiles. I will try moving the shoe back to achieve contact along the full panel length and see if there is any change.

From contributor L:
The questionerís pot is set correctly; the roller should not contact the panel except at the very end of travel. We recently banded some panels from another shop that had been CNC'd. They had a terrible edge. I asked the owner how often he changed router bits, every four units or so. That was the problem. The dull bit was pulling the fibers out of the core. We change bits every 40 sheets of melamine and get a really clean cut. We've run a glue pot bander for many years and never had a problem with chunks getting on the glue roller or into the recirculation tubes.

Check your roller and pot temperatures with an IR sensor. They cost less than $100. The problem with letting the glue roller hit the panel is you will get a worn area for the most commonly used panel thickness. When you need to run a different thickness you won't get an even spread. Check your glue coverage by running some clear tape. Too much glue makes a mess as it gets on the face of the panel. I've seen charred glue cause problems by getting chunks caught on the doctor blade and then wiping the glue off the roller. Keeping the pot and especially the tower and recirculation tubes clean prevents the charred glue problem. Does the automatic temperature reduction take place after a few minutes of non-use?

From the original questioner:
Yes the temp does come down after a short time of non-use. We change bits probably too often on the CNC for just this reason, I would say every 30 sheets. We have used the clear tape to set up the glue coverage. It is not a matter of getting enough glue, the problem is the pieces of wood starving the roller, in the short term the easiest fix is to open the glue meter more which lets the pieces of wood flow through and not get hung up, however obviously this puts too much glue on which fights you all the way down the line. The gluepot and tower are torn down and cleaned, and I mean cleaned every three to four months. The particles are not burnt glue which we have had before keeping a strict cleaning schedule.

We are using Panolam board, perhaps it is time to start trying some different mfg. and see if there is any difference. Even with the constant changing of bits it is remarkable the amount of crumbs that are in the panel edge after the CNC. We used to run everything through the slider and had no problems as the cuts were left perfectly clean. Stiles has said that this is the nature of the beast of using a CNC, and that we probably are not getting the pot clean enough during our cleaning. What I'm fairly certain the pot does not need is a 2000.00+ baking and cleaning.

From contributor S:
I just put a HolzHer Sprint into service and I am very happy with the glue system. I have a glue banding unit for curves and such, and it does not do very well. My shop is small and relatively low volume. The HolzHer comes up to temp faster than I can set up the trimmers and load the edgeband. The glue system is completely closed, there is no way for dust, or chips, or old glue to get back into the pot (there is no pot) or into the glue nozzle.

I suspect that nozzle holes will clog easily due to their small size. But there is an automatic flushing system that keeps everything clean (but it can waste a lot of glue if you leave the machine on standby a lot).

I suspect you will find a way to make your bander work correctly. There are thousands of shops using gluepot banders and CNCís. However, regardless of what the CNC sales people will tell you there is a big difference between the panel edge quality from a panel saw and a CNC. I have never seen a CNC produce the perfect "no-line" joints that a panel saw will. I am not just talking about the edgebanding, but also the butt joints when dowel construction. I know that others will disagree, but I see this over and over. The truth is it just doesnít matter, the clients can tell the difference anyway.

Finally, I always hear Edgeband sales guys say that pre-milling is necessary on a CNC edges for consistent perfect quality. Granted they might be motivated by the extra commission on the 4K price difference, but I think they are also speaking from experience. I think you will have this figured out in no time.

From contributor D:
The closer your settings are to recommended, the more it sounds like you may wish also to investigate the tooling and settings for the type (and density) of board you use. A different combination of tool, spindle speed and cut rate may prove to be of assistance. Others have had this issue and have at least minimized the problem.

From contributor C:
Try increasing fpm on speed to eliminate chunks. Try different CNC bits. I also agree with the comment on saw vs. CNC edge. Pre-mill is essential for good clean edges. I would highly recommend posting question in CNC for bits, speed and fpm tips.

From contributor M:
I use both edgebanders it sounds like to me you need to run some glue pot cleaner through your edge bander or have a tech come in and take out your pot and clean it. The black particles are burnt glue that comes from inside the pot. I have run p.b. melamine parts through the edgebander for years and never had to blow off the boards.

From the original questioner:
Thanks for all the responses. I will look into the feed/spindle rates on the CNC. These are not burnt glue pieces. If you stop the roller, you can pick them off the roller with a dental tool, and they are without question pieces of particle board core. The gluepot/tower was just taken apart and cleaned 1 1/2 weeks ago.

From contributor L:
We are running 600"/min at 18,000rpm, two wing compression from Integra tooling. We sometimes run Panolam. The only time we got lousy edges was when we tried single line cutting. The bit was turning the wrong way on 1/2 the cuts because of using the same cut for parts on both sides of the line. Has anyone been successful with single line cutting on melamine? I'd love to have pre-mill but am doing fine without.

From contributor R:
The guys are right about the roller not touching the panel. The SCM tech told us that it should be a fraction clear of the panel. We cut on the router and donít have many problems with glue not sticking. Just set up the shoe so it allows the panel to clear the roller, then open up the glue till you get good coverage on the piece. We have an SCM K208 ER if that helps.

From contributor P:
Several guys on this post are right on target. It sounds like you are floating the glue on the panel rather than pressing it into the panel edge. Some machine Mfg's say this is the best so you don't wear parts out in the glue system but I disagree. The fact is that when you float the glue on (do not contact the roller), number one you use more glue product, and number two you fail to get as deep of glue penetration into the edge at the roller where it's 200C and most liquid, and number three it creates a mess with squeeze out.

The shoe that is just prior to the roller should be used to pre-load the position of the glue system so there is no bump when the panel comes in contact with the roller. The rollers on the machines we sell are hardened and will accept years of contact against the panel before signs of wear appear. If the feed rate of the drag chain and the speed of the roller are the same, there should be very little friction to create wear or pull loose particles from the edge. You did not say what machine you have but run your roller the same direction as the panel and contact the roller. This should take care of the problem.

From contributor L:
I've have to disagree with all three points above. Having run without contact for many years Iíll take on each assumption one by one. "The fact is that when you float the glue on (do not contact the roller), number one you use more glue product, number two you fail to get as deep of glue penetration into the edge at the roller where it's 200C and most liquid, number three it creates a mess with squeeze out."

1. The amount of glue you use is controlled by the doctor blade and your clearances.

2. Glue "penetration" is controlled by: temperature, glue amount and pressure section. The panel is traveling at about 1' a second and only has to go about 1' to the first pressure roller how much temperature change do you think there is? The heat is carried by the glue so if cooling is thought to be a problem more glue is better! If you want to wipe more glue into the board run your application roller the other way, as you do for solid wood. Low end banders have fewer pressure rollers but I don't think that's an issue with PVC.

3. Is true if you have not properly adjusted your machine.

4. Having many years experience running without glue roller contact and not having excess glue on the face of the panels and it being the recommendation of the machine manufacturers why would I want to do otherwise? Iím currently running an IDM58.

From contributor P:
I read your thoughts and I'm glad you are getting satisfactory results with your method of applying glue. I have setup, ran and repaired banders for almost two decades and since the glue application is critical to everything else, I have spent much of my time there. Most of what you are saying I agree with but it goes further.

1. The doctor blade does control the amount of glue that remains on the roller and it controls how much transfers to the panel edge if you come in contact with the roller. I disagree with you when floating the glue on because you are dealing with contact of the glue without using force or pressure.

2. You again are correct when you say that penetration is controlled by temperature, amount of glue and pressure. Where we disagree is where the pressure is best applied. When the roller contacts the panel edge, the glue is forced into the edge at 200C or an ideal viscosity. It's most liquid when it is on the roller. As it leaves the roller the glue temperature begins to drop and thus viscosity increases. If your panel is arriving at the pressure station in 1 second at a 12" distance then your machine is traveling at 60fpm. Most cabinet shops have machines that operate between 30fpm and 40fpm. This increased time required to get to the pressure station also equates into a glue product that is thicker and allows for less penetration. Now I know you are running a hotmelt that is designed to skin more quickly than the hotmelts that are used in slower machines. In normal cabinet type edgebanders the panel is out of the pressure station in another second and the glue must be set enough to allow end trimming. So we are talking a quick process regardless of the speed. What I'm saying is that contact of the roller with the panel allows for the bond to the substraight to be made at the premium point and that is the glue roller.

3. You stated that if glue temperature is a problem then the more glue the better. And more glue can cause more squeeze out.

From contributor C:
Are you having problems with saw cut materials in the bander? If not, it is the CNC and perhaps the feed rate, bit, or material. Have you figured this out yet?

From the original questioner:
It is tough to say at this point if saw cut materials are a problem as we donít run much of anything through the saw anymore, and as the pieces of core material are already in the pot, you would not be able to tell if they were from the saw or CNC. As I said the saw produced a much cleaner cut, so we are currently working on rpm and feed speed to see if we can get the same cut results on the CNC as we do on the saw.