Problem cutting 1 inch poplar

[yobot]

Hi, folks.

I am trying to get into production on a job with a lot of small parts made from 1 inch thick poplar. We have used the wood a lot, and even with its particular fondness for fuzz, we (normally) achieve very good results with it.

This particular job has me using 1/8 inch bits
http://www1.mscdirect.com/CGI/GSDRVSM?PACACHE=000000066531385 and I am having a lot of difficulty achieving any kind of finish that even resembles acceptable (including a little fuzz) I have seen much smoother cuts from a coping saw.

.125 inch bit at 10k - 13k rpm at 1.2 - 1.7 ips, both climb and conventional. The cuts are ramped and I have tried little bites (.125 steps) and bigger ones. The bit handles the cut without any sign of bogging, it is producing chips, not dust, and the cuts are within a few thousands (raggedness excluded).

This is my first time to cut poplar with such a small bit. I don't normally have this issue. Any ideas on what I am missing? I don't mind being stupid on this one, I just need to get it fixed so I can get the work done.

I thought about a full plunge cleanup pass with a negative offset of a couple thousands at a slow ips (the cutting surface of the bit is over 1 inch).

It may simply be that the bit is not a good match for the job, but finding .125 inch bits long enough has been a bear already.

I need a really good fit on this part straight off the bot as it needs to immediately go into a template in a laser for some detail etching before cleanup and sanding.

I am obviously missing something simple. Please help if you have any suggestions. This is the first time I have had problems achieving a good finish with the bot, but I haven't used it much in the last 2 years due to other business concerns, so I am not a newbie, but definitely currently a rusty.

Machine is a PRTAlpha

Thanks,
David

[ljdm]

Is it possible the vectors aren't clean enough - is it that kind of jagged cutting? Just looking for the obvious, easy answer.

[yobot]

There are only 242 nodes. It is very clean.

The first import of the file into Partworks was a DXF gone wrong, it ended up with 222,000+ vectors. I caught that, dumped it, and used a clean .eps for the import, and had a good look at it before generating the paths.

There are no "artifacts" of a bad file as no two pieces have similar faults, other than the finish. Like I said, it literally looks like it was cut with a coarse coping saw. The same design cut with a scroll saw is silky smooth. Same thing with the laser. I would be cutting them on the laser, but the 1 inch thickness causes problems with dissipating heat in the center of the design, the rest cuts fine.

I am stuck getting it to work on the SB, or losing money on the job, not counting the extra man hours to manually cut part of each piece. I have a bunch of these pieces to do, they also have to be assembled with other pieces, painted and top coated. The job is due in less than three weeks.

I knew better, but it is going to be bulk repeat business, sold in catalogs, hard to turn down even with the steep startup curve.

Thanks,
David

[oddcoach]

Have you tried a tapered bit? i would give the 1 degree carving bit a try.
John

[beacon14]

Your extra long 1/8 inch bit may be part of the problem, you may be getting a lot of bit flexing. Can you try a test cut with a normal length bit of the same type to see if that gives better results?

Also do you have a router or spindle? My solid wood cuts improved significantly when I upgraded to a spindle. The 7.2:1 motor upgrade might also help, although not in time for this project if it's due in three weeks.

[knight_toolworks]

I would cut at 1ps atleast on the final pass with that long of a small bit. but poplar can be a strange wood it is stringy.

[yobot]

I think I have it almost ironed out. I knew there would be some flexing issues at the start, but I thought I had a work-around.

I don't have a spindle, although I will be purchasing one in the next few months, but the problem isn't runout or chatter, but the wood.

The work-around is a ramped multi-pass conventional profile cut with a slight offset, then a nice slow full depth climbing cut with a -0.001 offset. If the results remain consistent, then I will be in good shape.

The problem with the wood is that it is the driest wood I have personally ever come across. I have tried two different moisture meters on it, and it all reads between 0.4 and 0.6 % moisture content. The slabs are all 10 - 14 inches wide, and I would swear that they have to be decades old. It is also the hardest poplar I have ever encountered. I had a real challenge getting the pins all the way in.

I tried the meters on other hardwood, including poplar from another source, and got average 6 - 8% moisture readings.

I didn't think it was possible, guess I learned something new, or maybe it was some hallucinogenic long dried fungi left on the wood from farmer Joe's lumber stash, circa 1920...

David

[yobot]

Replying to myself here just in case anyone else comes across a similar challenge.

Here is what I found.

1) No matter what I did, a conventional cut would cause the bit to bite into the wood enough to distort the bit,
2) All cuts had to be "climb" due to problem number 1
3) Use a collet that allows you to leave only the cutting surface and about 3/8 inch clearance between the spindle/router collet and your work. (Some collets will bind with the bit higher than the top of the "neck", and not properly seat)
4) Use multiple passes.
I made one offset pass, one 0.0 pass, then two full depth -0.004 passes, all at 1.0 ips.

After much wasted wood, dozens of combinations, the above is what worked. Using poplar, expect to have to spend some time sanding tool marks, largely dependent on the grain of the wood and the position of the bit at any particular point in the cut. I could not achieve a uniformly smooth cut. There were tool marks in random places.

David