Scarf cutting file for plywood

[jehayes]

Can anyone help me with figuring out how to cut a scarf in plywood using V-Carve Pro? It can either be a step-scarf or a ramp. I just can't figure out how to set it up in V-Carve.

Thanks

Joe

[jehayes]

Further to my question above, I found this picture of a finger scarfing system used by a kayak kit maker. Does anyone have any thoughts on how to do this in V-Carve Pro? (I have been to the Vectric Forum and CNC but have found both less than helpful).

Thanks

Joe

[MetalHead]

Explain what your trying to do for us.

I got this on a quick search.

http://www.oneoceankayaks.com/stitchglue/plyshophtm/scarfjig2.htm

here is my search string

http://www.google.com/#hl=en&source=hp&q=scarfing+plywood&aq=2&aqi=g1g-s1g1g-s1g6&aql=&oq=scarfi&fp=18ec2db39eb50b9d

Mike

[jehayes]

The ocean Kayaks scarf jig is OK and will work, but I was hoping to be able to lay out the pieces for a stitch and glue boat with the scarfs (step, or finger or ...?) built into the design to be cut out on the MM so that I could then just assemble to pieces. I just found this web site (http://www.duckworksmagazine.com/06/howto/puzzle/index.htm) that has a really good idea but the question now is: how to implement the concept on the MM using VCarve. I assume it involves using the inlay function in VCarve but since I am a newbie on that software I can't get a handle on it.

Thanks

Joe

[sailfl]

Joe,

Most of the time when you talk about scarf joints you mean like:

http://www.duckworksmagazine.com/05/articles/scarfs/index.cfm

Your link took me to a puzzle joint.

ShopBot had a scarf joint that was more like the first one.

So I am not sure which you want help with the first or second. Then I went and looked at Ocean Kayak and it is the finger joints.

You need to create a profile of the finger joints and cut it using the create profile toolpath. Your profile will fit together in the two piece you want to join together. Most likely using an 1/8" bit for the gaps. But you do not need to inlay the cut.

Do you have a profile that you think you want to use? Post it.

IF you are designing a kayak, I am interested and would like to participate.

I have attached a exagerated example because the fingers are 2" wide. So you would make them 1/8".

[lumberjack_jeff]

We build boat kits using puzzle joints. The marine architect we work with believes (and I agree) that a scarf has no advantage, and several limitations compared to a puzzle joint.

See this (http://www.ptwatercraft.com/ptwatercraft/PT_SKIFF.html) for an example.

The main limitation with a scarf joint is that (if you use it to assemble precut panels) there's no good way to be really hyper-accurate on the lateral or longitudinal alignment; the scarf does not accurately position the parts. Second, it doesn't provide significantly more panel strength because although the scarf joint allows a greater glue area, the panel's weak point is at the transition, whether it's the scarf transition, or the imaginary line where plywood becomes the joint.

The finger joint above would be a decent alternative, (it provides some enforced lateral alignment if not longitudinal, and it minimizes the straight-line weak point) but it would be slow to cut, create a great deal of waste, and unless I miss my guess all that glue will create an unfair spot - a part of the panel that bends less than the rest.

[jehayes]

Jeff: That puzzle joint is fine for what I am trying to do. Topologically it is the equivalent of the finger joint Nils provided. So the question remains: once I have the design in DXF format (which I can do) how do I set up VCarve Pro to cut the two sides so they will meld when pressed together. Do I cut one on the "inside" and on the "outside" or both on the inside or what?

Thanks

Joe

[lumberjack_jeff]

For me, the design of the puzzle shape was the result of a fair amount of trial and error. The shape I came up with is dependent on the peculiarities of my setup. In general, I started with a set of parallel lines (or polylines for acad folks) spaced an appropriate distance - say .010". For the kits we build, maintaining strict positioning between the spliced panels is important, so the profile needs to interlock.

The two lines are then separated and used to border matching panels. Both panels are then cut on the outside.

Most people doing this choose symmetrical profiles, but using the example of dovetails, I don't know that they necessarily need to be.

PS: The shapes used by PT watercraft are not my work. I will post photos of ours when I can.

[Kobus_Joubert]

Joe, I think what you are asking how to cut some joint with V-Carve is the following.

Both pieces you cut on the OUTSIDE of the line.....just make sure that your cutter size matches the joint profile. If the cutter is too big it wil DISTORT your joint..

If your workpiece is big, then go with a big cutter, if it is small, go with a small cutter.

[jehayes]

Jeff: Thanks. That tip about the parallel lines is a big help. I did the attached DXF file in TurboCad as a sample and have done a VCarve file which I will test tomorrow. Please let me know if this is what you had in mind (the dovetail version, not the oblongs)

I'll let you know how it goes.

Joe

[sailfl]

Joe,

I had to bring the file into aspire to open it what version is the dxf created in?

[isladelobos]

This is the puzle method?
Dxf file.

8868

[sailfl]

Joe,

I took your dxf file made the pattern for one. Brought it into Aspire but it will be the same with V-Carve, created a 2D profile tool path, cutting with a 1/4" bit on the outside of the vector using my new favorite test material foam.

This is the results.

Let me know if you have any questions.

I missed the comment that you want the dove tails. Now it becomes more difficult because of the angle in the corner. You have to use a smaller bit.

[sailfl]

Joe,

Once again, I took your DoveTail drawing but I added a .125 radius to the corners. Used 1/4" bit cutting on the outside.

Hope these help.

[hennie]

Nils, make it more interresting and do it so that the actual dovetail is only cut halfway on both the male and female side that should give you a verry strong joint so that when you glue them together there is a bigger surface for the glue to sit onto.

[jehayes]

Nils: That is perfect! Exactly what I needed. Can you just confirm for me what the cutting settings were in Aspire (better yet, send me the .crv file so I can see them).

Many thank. Joe

[jehayes]

Nils, make it more interresting and do it so that the actual dovetail is only cut halfway on both the male and female side that should give you a verry strong joint so that when you glue them together there is a bigger surface for the glue to sit onto.

Henny: That is a good idea except I will be working with 1/4 inch ply and am not sure I can cut to those tight tolerances. I will try one and see. Thanks for the excellent suggestion.

Joe

[jehayes]

Joe,

I had to bring the file into aspire to open it what version is the dxf created in?

Nils:

It was created in TurboCad 14.2 I can resave in Version 12 if you still want it but it seems you have imported into Aspire which is the goal anyway. Thanks again.

Joe

[lumberjack_jeff]

In a boatbuilding application, where the panels will be primarily bent around a curve, the point of failure will be the edge of the joint transition. Strength is improved by making this transition as gradual as possible.

There is actually a step above "strongest"; the kayak splice shown upthread. Its problems with it have to do with alignment, fragility for assembly and material waste.

[sailfl]

Jeff,

What are the dimensions of the last tabs? Height and width should be the same?

[lumberjack_jeff]

Personally, I use a rule of thumb that for panels which are bent, (as opposed to bulkheads) the joint (height) should be at least 8x the material thickness.

How many dovetails to put per foot of panel width? More ambiguous... I make the width and the height about the same, the main objective is that the joint should minimize the straight line parallel to the bending axis (the seam on the green dotted line) as much as possible.

[Gerald D]

Jeff, I agree with your logic on what makes the stronger joint. But, isn't the natural conclusion that a triangular finger joint is the best, albeit a bit boring?:

8894

[lumberjack_jeff]

Yes, but... strength is not the only consideration. Boat hull panels are often only a 10" or so wide x 20' (or more) long. It is very easy to glue the sections together with a non-obvious (yet catastrophic) misalignment that only becomes apparent when you're trying to stitch the panels together into a boat shape. The joint not only needs to be strong, it needs to enforce alignment during glue-up.

Granted, a finger joint is better than a traditional (or stepped) scarf, but it still is less foolproof than interlocking joinery.

[Gerald D]

Years ago we did stepped scarfs (in z-direction) for a boatbuilder, but we also added 2 dowel holes through the joint, which made the alignment a piece of cake. See http://www.dixdesign.com/oneill2.htm

Also, you would only need interlocking at the outer edges, the middle area can be finger jointed.

[normand blais]

Example of s.b. wiggle step scarf joint

[lumberjack_jeff]

Nice Gerald! I like the dowel pins.

[Gerald D]

Jeff, what you might like even more, is that we supplied a big board with registration holes (you can see it in that link) and the dowels went right through into that board. The individual pieces had a few more dowel holes that lined them up to the big board.

[sailfl]

Jeff,

Then it seems that the picture in post #2 is a good way to go but with some changes as far as size and frequency.

[lumberjack_jeff]

Nils,
The more I look at that joint, the more I like it... it looks like the short pin does lock in place.

[sailfl]

Jeff,

It also give the advantage of not having a straight line all in the same place so I would guess it is distributing the load which means there would be less likely that there would be a failure.

[jehayes]

Gerald's pin method is excellent for half-lapped frame joints. If/when I build another framed boat I will definitely use it.

Attached is a modified version of the finger joint, using Jeff's shape design which I can now incorporate into the plans for the boat panels. However, I am having a little problem getting VCarve to create the cutter path for the Male side. For those interested I have attached both the DXF file and the VCarve file for this joint.

Any advice on the VCarve problem would be gratefully accepted.:)

(The VCarve file is zipped to comply with the board upload limits.)

Joe

[Kobus_Joubert]

8903

Hi Joe, is this what you want....

Problem 1. on the bottom vector you have a broken vector line.
To check this unselect everything..... right click and ask for ANY OPEN VECTORS....you will see your bottom vector go pink.
With this selected click on the left hand side on JOIN OPEN VECTORS...looks like a small racetrack. Select some tolerance here and click JOIN.
Now your vector is OK

Problem 2.... select both vectors and create a toolpath OUTSIDE / Right and you are on your way.

Good luck

[jehayes]

Kobus: Thanks. You hit is right on the head. Joe

[MetalHead]

Did I mention how awesome this group is. Ya'll are great!!!

[jehayes]

After testing the initial cut files on the MM this AM I found the joint was very loose (as in 1/4" loose) and I revisited the design.

I changed the cuts for the two vectors to:

Female: Inside/Left Climb cut with a Minus 0.125 allowance (-0.125) and
Male: Outside/Right Climb cut with a Plus 0.125 allowance.

This seems to work perfectly. Attached are the DXF and CRV files for those interested.

I am still working on a version of the joint from post #2.

[Kobus_Joubert]

Glad it is working for you this way. I would have thought -0.125 on the one piece and +0.125 on the other piece should bring you to the same midpoint.

But if this works, then it is maybe the way to do it. Thank's for the feedback.

[sailfl]

Joe,

There is no reason that the outside cut with out any allowance should not work. In reality you have done the same thing by having a + / - allowance on the matching sets. I cut your example + / - and I cut one with no allowance and they both worked for me. I am cutting in foam and that makes a difference because it is muc more forgiving. I am going to run test in mdf later. I would also suggest that you make your test the size of what you are going to use because I found them to be very large and I don't think you want that size on your boards.

I have also created a small sample of the post #2 design that I cut in foam. It turned out good. I will cut it in mdf also. I will post some things later.

[lumberjack_jeff]

In my experience, the fit is very material dependent. Marine ply fits differently than CDX which fits differently from OSB - all cut from the same file. In fact, when I measure a straight kerf cut in a sheet of OSB with a .250 bit, the measurement is consistently .240 or so.

The behavior of the wood fibers in the material you cut has significant ramifications for the design of the joint.

One other thing; fitting one dovetail is easy. Fitting an array of them (such as in post 32) is harder.