Regarding the torsion box

Hey all, before I finally button up the top of my torsion box, I have a question:

Can/should I use a facing bit to level/square the torsion top to a desired depth?

I’d like to recess the top by 1/2" (the box is made with 2x4’s) so that the top will sit flush with the rails. The original plan I had was to also use the LR table as an outfeed - well when I mocked everything up, I forgot the spoil board… oops.

It wouldn’t be the end of the world if you think this is a bad idea - I can just use foam rather than MDF as a spoil board, easier to move.

Thanks for the help.

As always, you can do it if you want. Whenever you replace a spoil board you will want to surface that anyway, so it is just a one time thing to surface the top of the box before the spoil board. Not a bad idea, still, your call.

1 Like

I’m not sure gow flat the spoil board material will be. But if it’s pretty flat, it could save you time later.

I am thinking about a way to flatten the inner structure before putting on the top skin. That would be a lot less milling.

1 Like

What bit do you use for surfacing a spoil board? I am curious about this. My first cuts have been a bit deeper into my spoil board than I’d like. Figuring out how to enter the material thickness, depth of cut, and top of spoil board. I still am not quite adept at thinking in 3D.

I would surface my board some day, although the MDF isn’t that expensive to replace. I haven’t surface it yet but instead did a lot of work tramming and leveling the router and axes.

I have a 1" surfacing bit.

CMT is a good brand, they sell them local if you have a Menard’s. I found this surfacing CMT bit on Amazon that I was thinking of purchasing - not sure if the price is worth it compared to similar products at more than half the price.

If you really wanted to, a flat endmill bit as big as you can fit in the router/collet will work as well - I used an 8mm flat endmill bit on my MPCNC spoil board with great results… but since the LR2 is so freaking huge, I want a purpose-driven bit that can do it faster.

So I ordered the wrong shank size bit, doh. I was debating just using a 3/4 straight mortise bit - I have one that has 19mm cutting depth. Otherwise I could hold out another day for the replacement 1/4 shank hogging bit.

What depth of cut is the DW611 able to handle (I know, lots of variables to account for)? Could I have it take the full 12.7mm if I reduce the stepover enough? I was thinking maybe I could hog 3mm at a time at 12.7 DOC, 8mm/s. The alternative is for me to make a jig that rides on the outer rails and just use the plunge router instead of the CNC to do all this. Thoughts?

If you use a bit bit like that, you’re going to be limited to 1mm or so. How much do you need to remove? If you touch a bit at te high spot, how much do you need to drop Z in the lowest spot?

I need to remove 457,033,904 cubic millimeters of material (why cubic mm? because I like how large that number is when compared to the paltry 16.14ft3).

The estimated time to remove 12.7mm of material at 0.1DOC was something stupid like 12 hours.

While I could make a jig (I want 12.7mm taken from the entire torsion box wrt my rails) it would sag; and even though I have a plunge router, I want the depth to be consistent and not relative to each individual member/web.

I guess I am going to spend the day off hanging out in ye olde shoppe.

Oh, you’re trying to cut it down to size as well as surface it.

For the final pass, at least, I would make sure it’s not taking much off.

Bigger bits have a faster linear speed. The cutting edge of a 3/4" bit is moving 6x faster than a 1/8" bit. So if it tugs on some material, then material is going to push back. Probably 6x harder, although the material will break, so maybe not that much harder.

You’ll have to just spend some time testing it out. Trying different stepovers, speeds and depths. I hope the limit you’ll hit is just skipping steps and you can stay safe. Try it on a small area first. For the final 1mm or 0.7mm, use much less step over and only take off that little height.

Using the cnc to cut it has the advantage of not just flattening it, but making it parallel to the gantry. If the gantry consistently sags in the middle, the table will match that sag and when you put material down, it will conform to that shape. That’s better for the cnc tham using a router flattening jig.

Thanks for the reply Jeff. I calculated the feed rate for 1" cutting diameter, 3 flute, 0.1 cpt, 1/4" shank and the DeWalt 611 at setting 3 (20,400 rpm) and came up with 6120mm/m. I have no clue if I can even approach that speed. Thoughts on even attempting these speeds?

The other issue I forgot to mention is that we loose about 205mm of cutting distance - when I have my machine to the limit of Y, the center of the 611 plate is 205mm from the end of my table. How would I then approach precisely hogging 1/2" of material from that area my machine cannot reach?

I don’t even think I need to worry about getting it exactly right, since this is no man’s land for the cutter - I just have to get it close enough to the rest of the surfaced webs so that it does not cause any issues when the top skin on the torsion box is laying in it’s home. Besides, I will be surfacing 1mm from the top skin of the torsion box anyway - so any crowns or other fitment issue will be leveled wrt to the gantry anyway. Do you agree?

Nope. 100mm/s? Nope. I’ve never used a cut calculator, so maybe I’m not the person to ask, but there is no way I would run an 1/8" bit at 100mm/s. A 1" bit, no way.

More like 10mm/s. Or 15mm/s. But if possible, slower and deeper is easier on these machines than shallower and faster. I would probably start with something like 1mm depth of cut and 10mm/s, and then increase the depth until it sounded and looked like it was going to fail.

Can you make the top skin of your torsion box only the size of what you can mill? You could still attach other boards on the edges if you like.

If you’re doing the bulk of this milling on webbing, then don’t pocket the whole table, just the areas with webbing. (I’m going to assume 48" wide table for a moment). You could make a cad/cam to mill 48" wide and 1" long. Then just move to one piece of webbing and start that job. Then move to the next piece, etc. Then do the same for the perpendicular webs.

[quote="jeffeb3,post:13,topic:11568"] Can you make the top skin of your torsion box only the size of what you can mill? You could still attach other boards on the edges if you like.

If you’re doing the bulk of this milling on webbing, then don’t pocket the whole table, just the areas with webbing. (I’m going to assume 48″ wide table for a moment). You could make a cad/cam to mill 48″ wide and 1″ long. Then just move to one piece of webbing and start that job. Then move to the next piece, etc. Then do the same for the perpendicular webs.

[/quote]

In a way, yes. When designing the table I forgot to account for the 205mm lost at the back of the table - my rails should have extended a little over 8" in Y axis beyond what I wanted to machine. Which brings up the sore point of dimensional lumber - I forgot it applies to MDF... I should have built my table to actually handle the full sheet. Horseshoes and hand grenades though.

Since my fusion mockup is exactly the same dimensions as what my table turned out to be in real life, I simply did a push/pull command on the components of the torsion box to make them 12.7mm in height. I then used the face toolpath, selecting the model bottom geometry, set the bottom clearance to -0.1mm from model bottom, even stepdown at 1mm per pass, and used the feeds/speeds listed above. This generated a tool path requiring only 6 minutes (roughly) for each “slice.” Total machining time is reported as about an hour and 14 minutes - a far cry from the 12 hours before (it was not doing any rapids at all and running at 8mm/s for all moves).

Thanks for the input Jeff, it is appreciated.

  1. You need to remove 12.7mm off the last 205mm where the machine can’t reach.
  2. You need this to at least avoid conflicting with the full sheets you put on your low rider. Even after you surface the spoil board.

Right?

You could take something like an oscillating saw and just cut it, then sand it. A lille lower there is ok. But then if you put spoil board over it, you won’t be able to get the 1mm off of that part.

Is your spoil board thicker than 12.7mm? Can you just leave it uncovered?

You are correct regarding 1&2.

Every time I go to put my SD card into the machine and run the program I stop and second guess myself.

The easy thing to do here is just use my torsion top as my spoilboard - replacing this would be no different than if I put another spoilboard on top. In all the time I had MPCNC, I resurfaced once and never cut more than 1.5mm into the spoilboard. I could surface the torsion box top skin and forget about the webs. I would pull my skin forward 205mm and run a program to surface that area, push it back into final position and run code to surface the rest of the skin (rinse and repeat if I used a spoilboard in addition to the top torsion box skin). I will still lose 205mm. As for matching the height of the table saw, I could remove the spoil board or just throw 1/2in sheet under the saw to bring it up as needed… or do nothing if I use the torsion top skin as my spoilboard.

The right thing to do is take my rails off and extend them beyond the machine like I should have in the first place. I could also go unistrut, I think they make 12ft lengths. I’d have to buy another roll of belt. I should also take out the webs, remove a little less than 1/2" on my table saw and put them back in. This would be the most labor intensive option for not much gained.

The second best thing to do is exactly what we’ve been discussing. I would just sacrifice 205mm and that area would serve as a squaring frame of reference when I put sheets on. Losing 205mm of work area translates to 2320mm to play with - 4.6" lost or 1.55sqft.

Thoughts on using the top skin on the torsion box as the spoilboard?

Well, you’re already pretty close. If you just punch through and you decide it’s no good, you’re out what? A sheet of plywood? This is a kind of indecision I face a lot. A few choices, none of them are too bad or too good.

I think the top skin as spoilboard is fine. As long as you don’t glue it down.

I put a spoil board on top of the torsion box. Since it’s just a 4X8 sheet of MDF, and my table is bigger, I can surface the whole thing. I know where I can cut that way as well. You get most of your strength from having the top and bottom firmly attached to the webbing. Food for thought.

My torsion box is 3/4" plywood. I have a 3/4" MDF on top of that where I screw my corners down. Then my spoil board rides on top of that in between t-slots. I only have to replace the slats in between. I put the threaded inserts into the base mdf.

@jeffeb3, I have always been under the impression that one of the core properties of a torsion box is that the skins are glued to the webbing, since that’s what gives the box it’s strength. It’s the difference between a solid ‘I’ beam and three pieces of material stacked together. Without gluing the skins, you lose a lot of the stability. I suppose having only one side glued still provides some benefit, but the rule of thumb that a torsion box is as strong as a solid block of the same dimensions requires both skins to be glued down.

But then, I’m a computer scientist, not an engineer… :wink: