Cricket table calculation

[mrspanton]

Jacob's post about the saw horse has brought to mind another unsolved problem that has kept me awake trying to visualise and solve with no sucess so far.
Has anyone a sound knowledge of how you set out the joints for a cricket table, bearing in mind theres only 3 legs, with splay's. How do you calculate the cross section of the legs, and how is that affected by varying splay angles? If the legs were true plumb, the mortices would be at 60 degree's to each other, but as the legs are splayed, does this angle change, if so how?The type I have in mind is the one with apron type rails at the top and stretcher rails about 1/3 of the way up from the floor, also with a drop in shelf rebated into the stretcher rails. I probably havent desribed it very well. Have you any tip's?
cheers Jonathan

[mrgrimsdale]

Spooky - I just happen to be doing that very thing today, but not splayed, to support a mahogany round table top which I've had for years. It'll be like this a bit


All at 60deg so relatively easy.
If you want a splay there are 3 approaches IMHO;
1. Bodge, make it up as you go
2. Trigonometry - work it out mathematically - which I could have done for the trestles. I say "could have" but I'm out of practice with school maths but am revising and have just bought a scientific calculator, but it's simple stuff about "O" level standard. Trying to keep the Altzheimers at bay
3. Do it graphically - draw out full size and project the sides, bevels etc with compasses. I've had to revise this from a textbook and it's not easy - but is probably easier than maths if maths aint your subject. Also you can use the drawing as a rod and lay your pieces on to mark up etc.
2 & 3 are the alternative ways for working out roof geometry so a roofing book would help, and perhaps a roofing square but I'm not sure about that. I learnt about it all those years ago but have forgotten it.

cheers
Jacob

[mrspanton]

Thanks for your reply Jacob. Thinking about it again, whatever the splay, the mortices wil always be at 60 degrees to each other. What will change though is the angle between the two faces on each leg into which the mortices are chopped. As the splay increases that angle will increase also. EG on yours (plumb up vertical legs) that angle willbe 120 degre's. On this example the angle is bigger
www.1stdibs.com/archives/upload/1stdibsA/061207_ZH/YewTreeNY/08/IMG_9265.jpg

I think the practical method would be to set up 3 dummy legs nailed at the desired splay with thin laths to hold them together, then use a sliding bevel to find the angle of that inside arris so it is at 90 degree's both ways relative to the other legs, then use that to plane up the leg profile's?
Maybe at one time once they'd found the correct profile they kept a template and used the same splay every time? ;D
Cheers Jnathan
PS my instinct is to do a rod every time but in this case (non square item) it seems limited, that mystery angle is what has foxed me thus far

[paulchapman]

This is all very interesting because I've been grappling with the same problem. I'm making an oak table with a round top and splayed legs (four, not three like yours) and the rails will be fitted to the legs with loose tenons.

Here's the top with the top and bottom of one of the legs marked



What I'm planning to do is temporarily fit the legs to the top with a couple of dowels like this (this is just a scrap piece of pine used to get the angle of the top of the leg right)



and then take measurements and angles from the top and legs. Hope it works

Got any better ideas

Cheers

Paul

[nickw]

Sorry to sound like a broken record, but you could model it in SketchUp, then measure the angles there, and/or print out templates gauges etc. to your heart's content.

[paulchapman]

You're probably right, Nick, but this Luddite doesn't use SketchUp. I do all my modelling on the back of a fag packet - and occasionally a larger sheet of paper

Cheers

Paul

[paulchapman]

Thinking about this a bit more......SketchUp would give you what it should be, but then you still have to make it, at which point you would be dealing with how it actually is rather than how it should be (these should, of course, be the same, but you never know ;D). So I'd probably still take measurements and angles from the actual piece as I went along.

But maybe that's just my cack-handed way of working...... ;D

Cheers

Paul

[mrspanton]

Sketch it up, etch a sketch etc are abstarct, rod's are real, because the actual size dimensions are right there and you can use the rod to obtain angles, lengths of rails, positions of mortice's etc. If you use the computer model, you still at some stage have to refer back to a tape or ruler etc to mark out the wood. Its a lot less easy to make mistakes if the component's are placeded directly above the rod in there constructional arrangement, and you transfer from the drawn rod onto the component's. BUT I havent figured out yet the way that splaying alters the cross sectional profile (of the legs that is), as jacob said on the horse post's, the section is diamond shaped not square, and finding it could be tricky ;D

[paulchapman]

mrspanton said:

BUT I havent figured out yet the way that splaying alters the cross sectional profile (of the legs that is), as jacob said on the horse post's, the section is diamond shaped not square, and finding it could be tricky ;D

This was the main problem I was having. I found that if I worked out the approximate angle of the top of the leg (in my case it was about 3 degrees) and cut a piece of scrap (as in the second picture above), it made it far easier to visualise.

However much I think about this, I keep coming back to the idea that it's best to cut the legs and secure them to the top with dowels so that they won't move while taking measurements, and then work from there.

Mind you, I still haven't finished mine, so there's plenty of scope for c*ck-ups ;D

Cheers

Paul

Edit. It's because of the complexity of the angles that I've decided to go with loose tenons to fit the rails to the legs. I reckon it will be easier to cut the rails and legs to the correct sizes and angles and then rout them for loose tenons, rather than to complicate matters by cutting conventional tenons. Mind you, Dom would probably use his Domino....... ;D

[mrspanton]

mrgrimsdale said:

2 & 3 are the alternative ways for working out roof geometry so a roofing book would help, and perhaps a roofing square but I'm not sure about that. I learnt about it all those years ago but have forgotten it.

In France they use a scribing method again where they draw the roof on a floor surface full size. This quickly and easily enables the charpentier to find plumb cut's seat cut's etc, he translers them from the drawing directly onto the log. Because the components are assembled in real relationship to set out the joints etc, its near impossible to make error's as happens if you try to do it by abstract measuring. And this method has the advatnage of still working with timber that is not perfectly square in section, is warped/bowed/twisted or even hewn with an axe (ie what you generally get with green oak). They set out one truss drawing then re use it to make 3, 4 how ever many they need . They dont make a truss and work from that as a pattern. The face edge is marked on the floor drawing and is always is to the outside of the building, discrepancies in timber dimension's get acomodated inside so they get good continuous flat planes for the roof purlins/rafters/laths whatever is incorperated in that particular one.

[nickw]

mrspanton said:

Sketch it up, etch a sketch etc are abstarct, rod's are real, because the actual size dimensions are right there and you can use the rod to obtain angles, lengths of rails, positions of mortice's etc. If you use the computer model, you still at some stage have to refer back to a tape or ruler etc to mark out the wood. Its a lot less easy to make mistakes if the component's are placeded directly above the rod in there constructional arrangement, and you transfer from the drawn rod onto the component's. BUT I havent figured out yet the way that splaying alters the cross sectional profile (of the legs that is), as jacob said on the horse post's, the section is diamond shaped not square, and finding it could be tricky ;D

Ah yes, but with SU you can print out plans of the model at full scale - how does that differ from a rod? You can even stick the plans to a piece of wood if you want. It's the way I work a lot of the time.

[engineerone]

like every body i have hoped to do something like this, and run away because of this problem.

i have therefore looked at this thread, and wondered how to consider it.

first thing is to remember that the circle only ever has 360 degrees in it, so if something is splayed, the basic angle still works out at 60 degrees from the internal centre line.

what changes is the angle of the shoulder of the tenon, not the way in which the leg is splayed.

lets assume that the legs are splayed at 10 degrees(no actually lets take paul's case with 3 degrees)

so the goes out by 3 degrees from the top to the bottom. the cross rail still has to be parallel with the floor. now normally you would cut the tenon at 90 degrees to the straight leg. but if the leg is at 3 degrees then the cheeks must be at 3 degrees off 90 degrees. as long as you keep the cut on the same faces relatively, then you can set the saw at 87 degrees or 93 degrees.

this would cut the shoulders. next you would need to set up the jig to cut the tenon cheeks so that they did not cut into the shoulders. but even if you use loose tenons, the same applies.
you drill with the cross rail at 3 degrees off the horizontal. you would probably need to cut both ends of the loose tenon at the same angle to ensure that it does not bottom out.

looking at the legs on paul's table they seem to be angled in toward the centre of the table, but all that means it that the legs are set in a jig for the centre rail so that when you cut the mortices they are at 45 degrees to the centre of the legs, whilst the jig also rises at 3 degrees so that the tenons line up with each other.

however i do think that you need to do this with a scrap before cutting the "real wood" because otherwise it is difficult to visualise.

as for the 3 legged table, it again depends on the angle at which the legs are splayed. however the mortices are all at 60 degrees to the centre line of the legs, be they square or round.

i would i think make the legs straight, cut the mortices at the angles, then shape the legs to allow for the angles. that way you do not have to consider that until after the mortices have been done. seem to remember watching norm make some legs. he cut the mortices first, then shaped them later.

i was taught to simplify things in my mind when i trained as an engineer. even race car designers do the same thing. get the overall shape you want then whittle away to make the item lightweight and yet still strong.

not sure how much this helps, but it might clarify a couple of things.

paul

[dom]

Sorry to bring this down to the gutter, but somewhere in the deep recceses of the mind I have an episode of Norm doing just this operation, so if you could find it............

[9fingers]

That would be series 3, episode 11 - I've got it somewhere on tape -purely for personal research naturally

www.newyankee.com/getproduct3.cgi?311

Bob

[mrspanton]

nickw said:

Ah yes, but with SU you can print out plans of the model at full scale - how does that differ from a rod? You can even stick the plans to a piece of wood if you want. It's the way I work a lot of the time.

Can you get a printer that will do a drawing in one piece, or do you have to selotape lots of A4's together ;D?

[Scrit]

SU is all very well, but what did people do before it? The answer is that any worthy joiner knows enough about geometry and trigononmetry to set-out a job without the need for a computer. After work tomorrow I'll take the time to consider the problem and come up with a joiner's solution rather than a computer hacker's one. You'd be amazed at hoe good joiners can be at setting out

At the moment it's late on Friday evening and I've been the pub so I'm in happy mode, hence the incomplete answer ;D

scrit

[mrgrimsdale]

No doubt you can do it with SU but there is still a steepish learning curve at the end of which you have learnt only how to operate SU which you can only do on a computer.
Whereas if you do the maths or the drawings, preferably both, you are learning some very basic and useful skills.
If you do graphic methods only, you don't need any maths at all - the drawings do the calculations for you; a sophisticated calculator - even works for the inumerate, for the illiterate too, but you'd still need to have your wits about you!

cheers
Jacob

[paulchapman]

scrit said:

After work tomorrow I'll take the time to consider the problem and come up with a joiner's solution rather than a computer hacker's one.

That will be very helpful, Scrit.

Cheers

Paul