Question
We resaw 4/4 lumber and want to calibrate the halves to .40" thickness with accuracy prior to cutting into small floor inlay pieces on a CNC router. We hold our planks by vacuum on the CNC and onion skin the parts. After CNCing, we run through widebelt face down to free pieces. Our problem is our 20 year old widebelt is very unreliable and needs replacement. I have looked at all of the Taiwanese and Chinese imports as well as Butfering and SCM entry level 2 head widebelts with a metal drum on head one and a combi head for head two. The Asian imports are cheap, but how is the quality? Butfering's machines that are less than $40k are made in mainland China. SCM is made in Italy. Does this mean better quality, better resale value, or just another option? My head is smoking with all the research, but I am still undecided.
Forum Responses
(Solid Wood Machining Forum)
From contributor J:
I am a little unclear. Is the problem that you cannot hold accuracy before the CNC cutter, or after the CNC cutter when you free the pieces? If the problem is before the cutter, then any quality moulder with the proper setup will produce .040 with virtually no variance. I hold .001 tolerance day in and day out. You can even get a furniture quality finish right out of the machine.
The majority of material that comes to our shop is already hit or miss through a stratoplane from the mill. I split it on a large vertical resaw to make two matched pieces. I want to then surface accurately before CNCing parts and leaving as little onion skin as possible to still hold plank with our vacuum table. Many of our parts are less than a few sq. in., so onion skinning is a necessity. I think I am comfortable with leaving only .015" for the skin. (I currently leave more like .1" for safety since my pieces are poorly thicknessed from one side to the other.) I then run a continuous piece of 6" wide masking tape over the entire cut face to hold parts from getting sucked up in the dust collection of our widebelt when sanding pieces free. With our very old sander this takes three passes or so with a 40 grit to free all of the pieces. Is 0.015" enough skin to be safe to not cut our phenolic table and still maintain vacuum? Is this amount removable in one pass through a two-head widebelt with, say, 40 grit on the first steel drum and 80 grit on the second combi head? All of our inlay gets machine sanded once installed in the floor, so we don't really need to sand to higher than 80 grit except on the rare occasion that we do something outside of the usual like build a set of cabinets or a piece of furniture.
These odd contracts for things other than wood floor inlays are really only about 10% of our total business. I am completely self-taught in the manufacturing of wood floor inlay. We started out 7 years ago as wood floor contractors and started hand cutting inlays because the quality of most major manufacturers' product was low in our opinion. One thing led to another and we built a shop and bought equipment with little or no guidance except reading old threads from this website. I am fearful of spending $30-40k on a planer and sander that do not meet my expectation or improve our process. Any input is greatly appreciated.
So, I am not sure that even the best of widebelts will always free the parts simultaneously (in the same pass), especially when the changes in depth are slight as you re-set the sander for the next pass. I would definitely suggest that you use a good planer for the initial thicknessing before cutting the inlays onion skin style with the CNC.
With a pair of digital calipers to measure, you should be able to plane to an accuracy of +/- .005" or less. I like to surface a face of a board on the jointer as a first step before resawing. I do this to flatten and straighten, but also to establish grain direction. I use a felt tip pen on the end grain of each board to document grain direction. This step ensures that the thinner pieces will plane nicely after resawing when fed to the planer according to grain direction.
Also, I advise using an auxiliary table above (on top of) the bed rolls on the planer which helps eliminate snipe when planing thinner material.
Your .3125" finished thickness might prevent it, but we would often create small parts on the CNC leaving a .010" thick web or onion skin and then the parts could simply be broken from the web by hand. The residual web was then removed quickly with a flush trim bit in a router table near the CNC station. With .3125" thickness, it may be difficult to find a flush trim bit that would work for trimming the remaining onion skin because of the distance between the top of the cutter and the bottom of the bearing. And of course, this trim routing step is much more work than sanding the inlay parts free with a wide belt. Just general information.
I developed another method that I call "perforation." I would program the CNC to first cut the perimeter of the parts to a depth that left a web or onion skin of .020" thick. Then, the tool path was programmed to have the bit step away from the part perimeter by .030" and then cut the program again, this time leaving an onion skin of only .010 thickness.
What this does is allow parts with difficult and delicate face surfaces like two-sided melamine or solid wood, in your case, to be broken from the web by hand with no damage to the finished perimeter of the parts. This is because the onion skin will always break at the .010" thickness point, which is .030" away from the actual part perimeter or edge.
The .020" x .030 nub left on the edges of the parts is then removed with a flush trim bit on a fenceless router table.
Sanding still seems like the best bet to me for freeing your parts from the onion skin. I suppose you could apply some sort of tape to the routed side and then free the parts with a thickness planer. Might work, but you stand a chance of having some parts come loose and get destroyed by the planer, not to mention the added step of applying tape and peeling the finished parts free.
If you can find one used with the Quiet-Cut head, so much the better. We used to demo at the machinery shows, feeding a narrow drawer side cross grain through a Whitney with Quiet-Cut heads, an act that would suck the part into the head and up the dust pipe on any competitive production machine. The drawer side would machine flawlessly cross grain.
I don't know anything about the lighter machines, the Deltas, etc. - there may be some that will fit your needs. With your thin material, get a performance guarantee if you buy new.
Sanding is a very expensive way to remove wood. Get actual tooling costs numbers from someone using carbide knives and equivalent numbers from someone running a sander. HP is also a big factor. Removing the same amount of wood at the same feed rate will require approximately three times the HP when using a sander compared to a knife planer.
If you are planing large quantities, there is the dust/fire/explosion danger as well. I am not selling anything. My purpose here is to provide food for thought.
My advice would be to upgrade the planer first and take the material to a closer finish thickness at the router. For 10K to 25K you should be able to get a decent insert head single planer. Less if you buy used. I say do the planer first because you might find your sander requirements will be minimal with a good finish planer.
We have a Martin Tersa head single planer that is accurate to .001 in the 25” width and can be set to an accuracy of .005 digitally without checking with calipers. Rubber feed rolls and no bed rollers that give a snipe free surface. Boards coming out of our planer or S4S machine only require .015 to .018 over for finish sanding. The Tersa type vs. the spiral insert type planer head has been debated quite a bit here. The spiral insert might be better if you work with figured wood. The big advantage to the Tersa is the minimal upkeep and adjustment required. I would suggest trying your material through a few types of planers to see how it affects your final product. A little facing as contributor F suggests is a good idea, but at your thickness and minimal material removal, it is probably not possible or needed.
As for the sanding, if you only need 80 grit finishes, that should be easy. Our old Cemco combination head was very accurate for calibrating with only the roller once it was dressed and parallel to the bed. We now have a Kundig 3 head. For calibrating the first head is steel. We can remove .018 +- with 80 grit on that head. More with 60. I think the HP is 25 or 30. Be careful of high HP with the electric rates. For what you're doing, 15 to 20HP might be enough.
