Router table flatness

[Kicked Back]

Hi all

I just received a UJK cast iron router table top when it was 15% off a few days ago with the intention of building it into a workbench.

It's an absolute unit at 815x615mm but I wasn't too surprised to see it wasn't dead flat. Across one of the diagonals I can comfortably fit a 0.25mm ruler under a straight edge up to about 150mm from the corners. Ie. the plate cutout is humped.

I'm not one to obsess over perfection. I have a Festool Kapex that isn't anywhere close to flat either, but it cuts dead square.

So, how important is table flatness for routing, really? Some crude maths is saying that even if I manage to wobble the workpiece to its limits (+/-0.25mm), a 50mm cutter is going to be moving ~0.03mm out of square at the top of the cut. Assuming correct, is there something else important I'm missing?

Cheers

 

[Sporky McGuffin]

I seem to remember Fine Woodworking saying that router tables should have a very slight crown. I forget the reasoning, but I'll see if I can dig it out.

 

[PerryGunn]

On the Axminster site the description for the UKJ cast iron top says

The heavy duty cast iron table top is flex free and completely flat

If yours isn't 'completely flat' I'd contact them and as for a replacement

 

[Kicked Back]

I seem to remember Fine Woodworking saying that router tables should have a very slight crown. I forget the reasoning, but I'll see if I can dig it out.

I came across something similar on another forum. The idea was that the weight of the router flattens the table a bit rather than causing it to sag. But I don't know if that'd even be a factor with cast iron?

Alternatively, maybe it's easier to slide the workpiece around when there's a crown rather than a hollow...

 

[Kicked Back]

On the Axminster site the description for the UKJ cast iron top says

The heavy duty cast iron table top is flex free and completely flat

If yours isn't 'completely flat' I'd contact them and as for a replacement

An Axminster rep on their forum answered someone's question that the tolerance was 0.5mm diagonally. Could end up with something that's worse than what I have now but still within spec, and it's a bit of a nuisance item to return

But will contact them nevertheless and see what the options are.

 

[Recky33]

0.25mm over 815+mm will not make a blind bit of difference plus all pressure when cutting will be over the router plate so as long as that's flat your quids in

 

[DBT85]

a quarter of a mm over a metre is not something I'd be worried about.

Have you hung the router on it yet? have you used the other edge of the straight edge?

When you say the plate cutout is humped, are you saying the edge of the cutout is higher than the area between the cutout and the edge of the table? Since the thing you are routing is referencing off of the router plate anyway I doubt it'll make a blind bit of difference.

 

[TheTiddles]

Completely flat… Nobel prize incoming, or perhaps a little of an exaggeration on their part.

Mine is (by eye) within 0.2mm over the diagonal, there’s a crown over the insert plate by less than a mm, none of this appears to make a difference to my work.

As soon as anyone starts saying, flat, rigid, silent… they’re using theoretical terms for real-world situations and you can probably disregard whatever they say

 

[Kicked Back]

When you say the plate cutout is humped, are you saying the edge of the cutout is higher than the area between the cutout and the edge of the table? Since the thing you are routing is referencing off of the router plate anyway I doubt it'll make a blind bit of difference.

Not the edge, just a gradual increase in height from the corners of the table as you move towards the centre.

 

[DBT85]

Not the edge, just a gradual increase in height from the corners of the table as you move towards the centre.

Nothing to worry about then imo. Your router plate is the reference surface and it is flat and .25mm above anything else.

 

[Fergie 307]

The only way something like this could ever be really flat would be for it to be absolutely massive, like the tables on a metal planer. You are bound to get some variation on a relatively thin casting. It will also vary with temperature. If you measure it at an ambient temperature of minus 2, and again at 20 degrees I bet it will be different. One reason is that castings like this should really be aged after casting. Colchester used to leave the raw castings for lathe beds and so on outside in the rain etc for several months before they were brought in to be machined, this relieved the stresses in the casting. I very much doubt yours has been through this process, and this will be at least partly the cause of the inaccuracy. It has probably been machined very soon after casting, and then "aged" on the machine, and has warped slightly as a result. On a metal planer or similar this would be a serious problem, on a woodworking machine I doubt it's going to matter atall. And if they do indeed quote 0.5 as an acceptable tolerance, then you are well within that, so they probably won't replace it.

 

[TheTiddles]

The leaving castings to “season” is an old practice not based on any science, unless you left them to season under a giant magnifying glass the temperature require to stress relieve it would never be anywhere close to the ambient temperature, hence why it’s not done anymore, it just doesn’t do anything.

 

[Kicked Back]

Sitrep

I discovered that huge chunks of cast iron are surprisingly flexible. Lifting the corner that's low makes the table flat... more or less. It's a little more complicated than that since it seems to chase the error around the table a few times, but with a few shims scattered about, I've been able to get it flat.

 

[DBT85]

The leaving castings to “season” is an old practice not based on any science, unless you left them to season under a giant magnifying glass the temperature require to stress relieve it would never be anywhere close to the ambient temperature, hence why it’s not done anymore, it just doesn’t do anything.

I've no doubt you are right. If it did they would have found a better faster consistent way of doing it.

I'm sure I heard that BMW used to do it with their old turbo engine blocks at the time when they'd rank the boost on an engine for Quali, hit 1500hp and, maybe, last 2 laps.