Routing semicircles?

[Night Train]

I have a student who has been making a semicircular table of about 18" radius. He wanted to rout out the semicircle from solid timber he had glued up to form the table top. The straight side of the semicircle was along the grain. I first made sure that the end grain at the start and finish of the semicircle was well supported for cutting through. I set him up so that he could rout out in short sections leaving the waste still attached at points along the curve to support the router on its trammel and to protect the edges from splitting. Once he was all the way through the thickness he was then to rout out the remaining points of contact in several passes to release the waste.
That worked and he got his table top.

Unfortunately he made it too thin at 16mm and so glued up another.

This time, he set himself up and in one pass cut the full 20mm thickness on one pass 2mm away from his cut line. He then reset the trammel tried to cut to the cut line.

The problem was that he was now cutting against the grain for the first quarter circle and as he got close to the short straight grain he split the wood.

It is now glued back but what would be the beat way to recover this without splitting his work again?
He is adamant that he wants to rout it rather then finish it with hand tools.

 

[wizer]

I think the problem is trying to take the full depth of cut with too much material being removed at once.

I would set the trammel to the line and take full passes in 4-5mm depth increments. Nibble it away bit by bit.

 

[Night Train]

For the last pass I think he was only cutting about 5mm depth but as the cutter got round to the sraighter grain it just took a whole chunk off, full thickness.
The timber looks like Utili.

I would have put a compass plane to it myself but he's not into hand tools.

 

[Digit]

I think he's approaching it wrongly, no offense intended but I would have roughed it out either on a band saw or with a jig saw then routed.
Also he should consider the dia of the cutter, a small dia cutter leaves the cut at right angles to the timber, more or less, where as a larger dia cutter is more likely to leave the cut parallel to the direction of cut and thus less likely to 'lift' the timber along the grain.
Hope I've explained that ok.

Roy.

 

[Night Train]

Well the first pass at full depth was as good as it would have been with a band saw as there was timber on both sides of the 'kerf' to support the work as he cut against the grain. That bit seemed to work ok and was no rougher then if he had bandsawn it.

It was as he did his cleaning up pass that it all went wrong. He is using an 13mm cutter.

If he repeated the process I showed him the first time he would have been fine.

Anyway, I can see why he got in a mess, is there anything can he do to recover from it without recourse to hand planing or starting again with more timber?

 

[Digit]

The first thing that comes to mind NT is to simply make it smaller in diameter by re-cutting.

Roy.

 

[Night Train]

He is a bit stuck size wise, the piece is a folding card table with a circular top. He has already made the frame and legs for it and was only remaking the top as the first one was too thin for the hinges he could get. He can probably only loose a couple of mm or so before it starts to look too small.

Ultimately, whatever he does he will have to rout against the grain for the first quarter of the curve. He was using a new cutter so it should have been sharp enough, he was only taking off 1mm of the 2mm to the line, he was cutting 5mm depth of the 20mm thickness.

So, given we have glued the split back together again and the repair is good enough. What is the best/correct method that he should now use to finish trimming the remainder of the 1-2mm waste?



If I was making it I would have cut it by the method I showed him first off as that works when cutting from a large blank. Otherwise, in his current position, I would have made a template for the circle centre and cut the second quarter of the curve, going with the grain, and then flipped over to do the same from the other side. That way there is no problem with cutting against the grain.

 

[Digit]

I think he should accept perhaps that machines aren't always the answer to the Maiden's prayer and plane it.

Roy.

 

[Night Train]

Agreed.

The edge will also be reeded and again I suggested a scratch block and he though of a router cutter.

 

[Mr T]

Hi

Assuming he is supporting the router on the outside of the cut in some way I would make the perhaps heretical suggestion that he take the cut in the opposite direction where he would otherwise be working against the grain. If he is only taking 2mm off there should not be any problems with controlling the cut, providing he has the work held securely. I might add that I have never tried this myself, but I think it may be a way out.

One often has to think of novel solutions to the problems that students get themselves into!

Chris

 

[Digit]

I suspect that would be the answer Chris, but personally I would hesitate to suggest it to a student, and absolutely not a single pass!
I have this mental picture of the router running round and round the table and the student disappearing head first through a window! :lol:

Roy.

 

[Night Train]

I don't think routing backwards would help as the cutter is still going against the grain when cutting. It would still split the unsupported edge.

Maybe there is a case for a left handed router and cutter working anticlockwise.

 

[ivan]

The split came at the pointy bit of the "D", where curve meets straight? Could you glue on infeed/outfeed blocks with piece of newspaper in glue joint, so they knock off afterwards, then it's the block that splits, not the job? If the split is in the middle of the curve, a spiral, always cutting, would be kinder here. I'd be tempted to finish with some climb cutting if the router was in a table and the work jigged.

 

[Digit]

Ah, but there is a difference NT. If you route in the conventional direction the cut gets deeper and deeper as the router moves forward, this puts the maximum cut at the point where the cutter leaves the timber, thus encouraging break out.
If you rout 'tother way the cut gets shallower and shallower, thus minimising any chance of the wood lifting away.

Roy.

 

[Mr T]

Hi

Yes, I agree, one hesitates to suggest such practices to students. I have previously suggested back feeding to students and later thought "Oh dear what have I done", but fortunately they are sensible and only take a very shallow cut. I use it regularly if I am worried about break out at the top of the cut when rebating, I just take off about 1mm in the "wrong" direction then reset deeper to cut in the "right" direction.

Chris

 

[Mr T]

Alternatively take a number of plunging cuts at the point of weakness before taking the full cut, this would have the effect of severing the grain so that break out if minimised.

Chris

 

[Night Train]

The split came at the pointy bit of the "D", where curve meets straight? Could you glue on infeed/outfeed blocks with piece of newspaper in glue joint, so they knock off afterwards, then it's the block that splits, not the job? Take your point about bit diameter, but a spiral, always cutting, would be kind too. I'd be tempted to finish with some climb cutting if the router was in a table and the work jigged.

No, I already had that part covered the first time round though he forgot when he did it himself.

The split came as he reached the midway part of the curve. The start and end of the cuts are end grain. The straight part of the D is long grain. The split happened when he got to the part of the curve where the grain begins to go straight again. Even though his cut was only 5mm deep it ripped off about 20mm of the width of the D midway along the curve. As it was quick and clean I managed to get it glued back on for him lining up the fibres so it hardly shows.

 

[ivan]

Sorry, just realised that and edited my 1st post accordingly.

Maybe you've got a point, a full depth pass at rad + 2mm doesn't smack of undue care! Has he got strong arms? Try a spiral, since getting 1/2" up, down, and up/down to try in the Woodrat have been quite impressed; presumably because the cut is continuous and not 40,000 hits per min. They do cut fast too, should please the boy....

 

[Night Train]

Ah, but there is a difference NT. If you route in the conventional direction the cut gets deeper and deeper as the router moves forward, this puts the maximum cut at the point where the cutter leaves the timber, thus encouraging break out.
If you rout 'tother way the cut gets shallower and shallower, thus minimising any chance of the wood lifting away.

Roy.

OK, seems reasonable. I'm not a great hand held router user so I might have to try that see what the effects are like.

 

[Johnboy]

Route a template out of plywood and attach with double sided tape. Use a bearing guided cutter with a bearing at both ends then you can invert the work so you can rout downhill from the centre outwards but still not climb cut.

John