And now ... a *bad* idea?

[Greg Guarino]

This came out well in the end, but I have a feeling that this sort of thing is frowned upon, possibly with good reason.

I made the first set of curved rails this morning, using the jig I made yesterday. It's a very gentle curve, in this case deflected 1/2" in the middle over a 15" span.

The pieces are 3/4" oak. I used a jigsaw to cut out the first blank, reasonably close to the line. Then I routed out the finished curve. It worked well.

I momentarily forgot that I hadn't roughed out the other pieces and put a full blank in the template jig. I realized the mistake right away, before I even turned on the router. But I decided that I'd give it a try without sawing first; I had one extra blank anyway.

Here's how it went:

(for those of you who may not watch the video, I of course did *not* try to make the full-depth cut all at once)

The first attempt was especially clumsy, but even that piece turned out fine. I decided to do them all that way, or at least until I ruined one. It never happened. Oh there was plenty of ugly-looking tearout on the first pass, but all of them came out smooth in the end.

So was I just lucky, or is this borderline OK, at least for a shallow cut like this?

[Swingman]

Yep, I would say the gawds were smiling on you.

I can guarantee you that you will not always be so lucky when pattern routing.

A work piece will come along with the grain just right and it can get dangerously grabbed and thrown/jerked quickly, and at the very least, only ruin some valuable stock, if you're lucky.

When pattern routing like that it is much safer to rough cut to within

1/8" of your line, and while you still may ruin the occasional piece due to grain issues, much better to be safe than sorry in both regards.

A safe practice, for both you and your wood, is to take shallow cuts whenever possible with a router, for a number of reasons (including less lateral force on the bit, and bits do break) ... you can't take a shallow cut with a pattern bit that is not riding against the pattern with a SAFE degree of control ... you are entirely at the mercy of the direction/grain of the work piece.

Lack of total control by you when routing is when bad things happen!

Murphy will strike, so ALWAYS, and I mean ALWAYS, take the safest course of action with any routing and cutting you do.

[Greg Guarino]

I've had another idea.

I'll take your word for it on the safety issues; I was prepared to ruin some wood, but I'd rather keep my body parts in working order (or whatever facsimile of working order they are in currently).

Without a bandsaw, it was kind of a drag to make the first cut in tough wood. I have eight more pieces to rout (shorter, but otherwise similar). I'll need to remake the jig with the new template anyway. If I were to place the "fence" 3/8" further back, I could rout with the roller against the template, but cutting only a maximum of 1/8" in. I could then use spacers made of something of reliable dimensions (1/8" aluminum comes to mind) to move the workpiece forward in increments until the finished cut was made.

Or I could make the fence movable (forward and back), with fixed stops to ensure that the final position was accurate.

I don't know if I'll bother to try that in this case. Much as I won't enjoy jigsawing the oak 8 times it will probably take less time than making a more complicated jig. But I may store the idea away for a time when I have a more intricate curve to make.

Problem solving really is half the fun. :)

[Leon]

One thing to consider. The less cutting you do with a router bit the longer that bit will last. While you can resharpened some bits a pattern bit no longer cuts true to the bearing diameter after being sharpened. Jig saw blades are cheap by comparison.

[dadiOH]

I don't think of 1/2" as shallow but what you did is OK...just keep in mind that routers are basically for trimming and that they chop as much as cut. That means there is a LOT of force being applied to the wood and - if the cut is deep (or even if it's not) - it is easy to spit the wood. You mitigated that by turning the work piece end for end. Had you not - and if you made deep cuts - the workpiece would have probably split out.

It is not always easy to limit how deep you are cutting; one way you could have done so would be to climb cut the left side a bit (rather than reverse it) until you had cut in a ways, then finish it normally.

[Swingman]

Only on the first one, and certainly not on the second one, IIRC.

The first one only made me only marginally uncomfortable because he did turn the piece 180 before finishing, but, and having done this a few hundred, if not thousands of times, the second one, starting around

2:17, made me cringe.

There is another excellent reason for NOT encouraging the method of turning the work piece 180 degrees to finish the cut when pattern routing:

You can only get away with that if the arc faired in the pattern is _perfectly_ symmetrical, AND it is a two point arc ... the first is rarely the case when the pattern is faired; and anything but a two point arc would obviously not allow a 180 degree flip.

Best Advice: when pattern routing, you always want to use the same REFERENCE EDGE (against the jig fence) and REFERENCE FACE (against the jig table) throughout the operation, otherwise you risk a badly skewed, and unusable representation of your desired curve.

[Greg Guarino]

That's just editing; I did indeed turn them all around.

Let me ask about that then. I turned the work around *not* because I was trying to get away with a dangerous procedure, but because I read it somewhere; keeping the cut "with the grain" to avoid tearout. Not necessary?

I thought about that. I was convinced that I had gotten pretty damned close to that line when I made the template; closer than anyone would notice anyway. I even wondered if the proper way to ensure the piece would be symmetrical would be to turn them around and only rout half (plus a little) in each direction.

In the end I decided I had made the template well enough to pass inspection and went all the way across after the flip. The bit did indeed cut something on the second half, but it was very very shallow, maybe a 64th.

I guess I'm not familiar with the term "two-point arc", at least in terms of it being a special mathematical shape. CAD programs and Sketchup have various ways of drawing them, but they are all parts of a circle and are thus all inherently symmetrical. If you mean that the arc can't be "tilted", or that it must indeed be an "arc" rather than some other curve, then I understand.

So again, should I take that to mean that routing "with the grain" is unnecessary, at least if I'm removing a sufficiently small amount of material?

[Swingman]

Ouch, possibly dangerous trap for the inexperienced watching a video? Something to think about when producing and posting one.

Yep, mostly necessary to route 'with', and not 'against' the grain to avoid tearout.

Just keep in mind by flipping the board 180 along its length after the first half pass (and assuming the grain direction is even along the board, and doesn't change midway, which is not always the case), you are in effect changing the "direction" of the grain for that last half pass. ;)

Yep, If you don't have symmetry, in this case a smooth line joining two points on the curve of a circle, you won't be able to flip the board halfway, then go all the way through the operation on the flipped pass without the possibility of some grief.

And it gets worse the sloppier your fairing is. DAMIKT ;)

The tool that you use to draw most of your arcs in SketchUp is called the "Two Point Arc" tool, and the reason why I used that specific terminology.

Note that it actually takes three points to draw the arc, but it is often referred to as a "two point arc" ... IOW the length of the chord, measured from the two ends, specifies the limits of the height of the arc, otherwise it won't be the arc of a circle.

Again, for best results, and on most wood, you indeed want to route 'with the grain', not against the grain, as long as you are routing in the proper direction for the rotation of the bit (against the direction of the bit), i.e, not 'climb cutting'.

[Greg Guarino]

I wasn't thinking of the video as "instruction"; more a record of what I did to share with some virtual and non-virtual acquaintances. In fact, I generally expect the "instruction" to flow in the opposite direction, as in this thread.

But I see your point. Some people take "authoritative advice" from places they ought not. Perhaps I should put up a disclaimer: "Novice trying stuff for the first time", or "Half of what you see here is wrong; we don't know which half".

[Greg Guarino]

I had actually taken some time to select which side was "up" and had marked each piece accordingly. This project, by the way, is probably the first time I have ever paid attention to that. I selected the grain pattern mostly for what I thought would look better (where the was a curve to the grain I selected it to "follow" the curve to be cut rather than the opposite), but this also meant that "with the grain" would always be "from the corner toward the middle".

you are

I knew that. There was such a thin sliver being removed that I doubted there'd be any tear-out. There wasn't.

But I'm still curious; if routing "with the grain" is important, and keeping a single reference face is important, and avoiding climb-cutting is important, I think we have a quandary.

[Greg Guarino]

Well, (putting on my geometry hat) you can of course draw an arc of a circle of any height through two points. To be sure, the greater the height, the closer the arc will come to being a full circle, and it's "width" at its widest point will be much greater than the distance between the two points.

What you mean is that you can only draw an arc of a certain height *if* you want its widest point to be where the two points are. Exceeding that height (which should be 1/2 of the distance between the points, producing a half-circle) would require a different sort of curve, a parabola, for example.

:)

[Swingman]

More about the above, below.

Indeed, you are correct that there is a quandary.

All rules have exceptions. Knowing when, where, and how far those exceptions can be applied is called "experience" and is the key to resolving said "quandary".

Experience comes from making mistakes, often by bending the rules a bit too far. ;)

The problem is, mistakes when routing can injure you, often beyond repair.

So, when routing in particular, learn the rules (and you have already exhibited, by your first quotes above, an innate understanding of the why the rules are rules, and what to do to incorporate them in your thinking).

And exercise the maximum caution to minimize the injurious results when the inevitable shit happens.

Take it for what it is, and why I brought it up ... I've got a lot of experience doing this particular operation, and it's a fact I was made uncomfortable by watching method/procedure in your video.

You are welcome to temper that personal observation with the fact that I approach every routing job, handheld or table mounted, with, if not trepidation, at least a healthy caution, having _experience_ numerous applications of Murphy down through the years when routing.

IOW, learned long ago to NOT think of any routing task as merely a "routine" operation, but one that must be thoroughly thought out, taking into account type of wood, grain, direction of travel, depth of cut, desired results, etc ... and I always breath a sigh of relief when any routing tasks are successfully finished, with me and the wood intact and I can put the damned thing away.

[Swingman]

You missed my point, but mainly because I was lacking in clarity:

IOW the length of the chord, measured from the two ends,specifies the limits of the height of the arc, otherwise it won't be the arc of a circle WITH THE RADIUS YOU WANT.

[Greg Guarino]

I should have mentioned that you had me convinced me not to try my "method" again from your first post. I was just clarifying what I had done. [The instrument I play is difficult to negotiate with less than the full complement of fingertips. One of the guys in our band, who is missing one "tip" courtesy of a table saw, is lucky that his chosen instrument is the drums. I am similarly fond of my eyes and chiseled countenance :)]

You may have noticed one detail that I thought of as a safety/confidence feature when I added them: the "handles". I've seen pictures and/or videos of pattern routing that made it look too easy for fingertips to slip into harm's way. I like that I was able to keep a firm grip with my hands far from the action. This also may be a common practice, but for me it was another personal "discovery".

[Swingman]

Yep, I duly noted that. And by your first above, and me being a stringed instrument musician myself, we can both appreciate each others concern in doing everything possible to keep our fingers attached and working properly. ;)

Funny thing is, I hate when people play the "safety card" posts, mainly because in most posts on the interwebz it is just more filler to hide the fact they really don't fully understand what they are blathering about.