Thread: Best way to attach desk top?

Originally Posted by zoovegroover

Hey all

Posted a similar question here a while back, but need some different advice this time. I'm making a sort of mini-desk/ bench seat that is essentially two side panels, a support beam, and the top. For the support beam, I'm routing some sliding dovetails for the join, but am unsure about attaching the top. Is it ok to just glue it down across the top of the legs and support beam, as seen in the top view in the first attached image? Or do I need to use dowels or some other form of joinery, like figure 8 clips or something? Would cross grain expansion cause a problem?

The image shows the top view of how it'll look before the top is joined, and then there's a quick sketch of how it's going to look finished.

Thanks, guys!

There are many many options, and the one you choose depends on taste and the look you are trying to achieve.

I have a desk like that... I bought it as a student some 30 years ago and now my daughter uses it. The desk is made from 25 mm thick pine - sides top and what you called the beam. The sides finish up ~5 mm above the surface of desktop, and the distance between the top of the sides and the and the top of the beam is is 5mm more than the thickness of the top. The beam is the same length as the desktop. The beam and the desktop are secured by means of large button shaped hex screws that fix the sides to the ends of the desktop and beam - I guess you could call them coach bolts with a decorative head. Strong as anything, has never shifted or loosened or cracked and looks quite beautiful.

If you prefer for the top to be bigger and have a bit of "overhang" all around than you will need one of the methods you suggested. You could secure several blocks or a strip to the inside of the sides and secure them to both the sides and the top with screws in several places. Along the same vein, there are metal l shaped brackets that you can buy about 12 mm x 12 mm with a single screw hole on each leaf of the L.

Finally you could try some cam fixings like the one below

images.jpg

Hey thanks for the reply, mate. I guess I wasn't all that clear in my original post now that I think of it! Although your answer covered a lot of bases anyway

I would like the top to be flush with the sides, and the joins to be seamless, if possible. That's why I was thinking maybe dowels might work (put a couple thru the end grain of the sides and attach the top that way), but with the glue being so strong was wondering if just gluing the top down would do the trick in itself? I guess I could always supplement the glue by running some screws up thru the support beam into the top? Or would using metal brackets allow more for the movement of the wood?

My understanding of wood movement is when there is more of a frame involved in the base is when allowances need to be considered.

If the sides are made from the same wood as the top then I would expect the sides and top to move together. The center support is not going to move in the same lateral plain as the top and sides.

That being said I don't really see this as a problem in regards to movement.

I am really sure if just using glue will be sufficient to hold the top in place. Where the side meets the top, it will be a butt joint where end grain meets long grain. Where the top meets the center support it will be long grain meets long grain.

If it was me I would add buttons or figure eights where the sides meet the top and not use glue for the top. The thoughts are I can at least keep an eye on it and resolve any issues in the future.

Thanks for the advice, Christos...I was leaning towards using figure 8s for the section where the end of the top meets the sides. Maybe one at each end...would you use them along the beam, too? Or just screws, or nothing at all?

What material are you wanting to make the desk out of ?

Originally Posted by zoovegroover

Hey thanks for the reply, mate. I guess I wasn't all that clear in my original post now that I think of it! Although your answer covered a lot of bases anyway

I would like the top to be flush with the sides, and the joins to be seamless, if possible. That's why I was thinking maybe dowels might work (put a couple thru the end grain of the sides and attach the top that way), but with the glue being so strong was wondering if just gluing the top down would do the trick in itself? I guess I could always supplement the glue by running some screws up thru the support beam into the top? Or would using metal brackets allow more for the movement of the wood?

Figure 8's would be OK, but they are not really "seamless". The cam fixings that I showed you can be bought with caps... or you could use a fostner bit to drill a shallow recess around the hole that takes the cam, and use a shallow plug of veneer to hide the cam. I guess you could hide an L bracket in the same way if you installed it into routed channels under the top and on the inside of the sides and then cover the bracket with say a 2 mm thick veneer.

But if you are going to go to that much trouble and insist on an invisible joint, then I think that your idea of dowels is probably the best. Fluted hardwood dowels are as hard as nails and they are very cheap! And if you think about it, that joint will not really be stressed as the weight of the table top and anything on it will always be pushing the two surfaces together, not apart - so that the main purpose of fixings is not to stop the table top from lifting, but rather to stop it from sliding around. Depending on how heavy the top is, you may even get away with having 4-6 dowels on the sides and back and using them as locating pins WITHOUT glue. - or you can use just a little on the dowels only. The only problem with this strategy is that without something to pull the top down, there may be an unsightly gap between the top and the sides... these gaps is are not usually visible if there is an overhang, but you want a top that's flush with the sides (there is always putty or clear silicone )

I don't share the concern over "movement". Movement can occur due to thermal expansion and or due to humidity changes. You can look it up, but the thermal expansion coefficient of timber is a poofteenth of a millimeter per degree celcius change in temperature. Humidity can warp timber dramatically, but this stops once you provide a water proof seal...so get that timber finished and sealed as soon as possible.

I hope that these suggestions are of help

I have built 6 desks, from 60" to 96" length. Frames and leg space were the first considerations. I do nothing more than to depend upon the Law of Gravity to keep the tops from floating away. In the past 20+ years, nobody has made an appeal to a higher court with any success. Float the top. Decide on attachment in the next few years.

Originally Posted by zoovegroover

Thanks for the advice, Christos...I was leaning towards using figure 8s for the section where the end of the top meets the sides. Maybe one at each end...would you use them along the beam, too? Or just screws, or nothing at all?

I would.

Just one other question how thick are the sides going to be?

Originally Posted by GarciaJ

Finally you could try some cam fixings like the one below

images.jpg

These are called Zipbolts and you can get them from Elraco .
I built a large desk (2900mm x 800mm with a 1400mm x 600m return) and used Zipbolts to attach the top to the two end drawer stacks as well as attaching the brace along the back. Strong and easy to use but does require careful measurement. Removable in case you need to move the desk

Thanks again all for the advice. To answer a few questions from some of you. The material is recycled tassie black butt. Beautiful stuff I picked up from urban salvage in spotswood Melbourne. Can't recommend these guys highly enough. It's about 19mm thick.

These are all great suggestions. I think the reason I'm a little confused is because people are always saying how strong wood glue is. Stronger than the wood itself? So If that's the case and there is no strain on these joints as gravity and strain on the top all work for, not against the join, I just thought at first that gluing and clamping it in place would be enough. Why use dowels if the glue is strong enough?

That may be an amateur response, I don't mean it to sound that way, I'm just genuinely confused as to why it would need screws or dowels or anything else when the glue is so strong supposedly?

Yes glue is strong . When things move at different rates the glue holds and the wood shatters .
Things glued side to side ,the grain direction, hold one huge amount better than side grain onto end grain . or end grain to end grain.
In your case the top and base grain will be expanding and contracting in roughly the same direction , The boards though will have come from different parts of the tree or even different trees possibly ? so they wont all react the same .

I don't know if it was mentioned be the other guys above , a few more ways .
You could through peg it down with the pegs skewed a little , I use that method a lot and like it for its quick as a flash , it suits an antique sort of look .
Through dovetail it . a lovely quality look .
If the ends were in a little under the top you can do sliding dovetails like the base . The Spanish used to do this with tables . sometimes the dovetail is wider at the end and narrower at the start , they go on easier till it get to the other side.

From the ones above I read , the loose to with gravity holding it down is a good one . another one that I do a lot with dining table tops and locating pegs underneath . a cleat screwed under the top on each side would be a good thing as well, for locating and helping keep the top straight .

good luck .

Rob

Originally Posted by zoovegroover

Thanks again all for the advice. To answer a few questions from some of you. The material is recycled tassie black butt. Beautiful stuff I picked up from urban salvage in spotswood Melbourne. Can't recommend these guys highly enough. It's about 19mm thick.

These are all great suggestions. I think the reason I'm a little confused is because people are always saying how strong wood glue is. Stronger than the wood itself? So If that's the case and there is no strain on these joints as gravity and strain on the top all work for, not against the join, I just thought at first that gluing and clamping it in place would be enough. Why use dowels if the glue is strong enough?

That may be an amateur response, I don't mean it to sound that way, I'm just genuinely confused as to why it would need screws or dowels or anything else when the glue is so strong supposedly?

I think that this may have already been answered elsewhere times, but here is my version:

The notion that glue joint is stronger than the timber depends on the type of timber, the glue and the interfaces...

Glue on end grain to end grain is very weak, because the end grain is porous and therefore the surface area of contact with the glue is low, so you need to increase the surface area somehow. I've seen boards end-grain joined in laminates but the joins were in the pattern of a saws teeth effectively quadrupling the surface area, and the individual pieces were supported by side grain butt joints onto adjoining pieces in the laminate.

Side grain on end grain is weak too, that's why people bother with strategies like mortice and tenon joints, dowels, half laps, dovetails etc. To increase the surface area for glue and to provide some side grain onto side grain gluing area.

Side grain on side grain is very strong, even when the surfaces being joined are long and narrow, or the grain in the two pieces are perpendicular to one another or even if you simply "butt join" the pieces. I've made up wide boards from narrow material for furniture this way countless of times, and I have never had one of these split. Even so increasing the surface area by - say - tongue and groove or even a loose tongue will increase the strength of side grain joints as well.

And imagine the relative strength of the wood to that of the glue when joining balsa compared to joining any hard wood (to use extreme examples). The quality and the type of the glue will obviously make a big difference - PVA is a good, durable, general purpose glue... but there are stronger and weaker alternatives.

Finally your original question was not really about glueing, but on how best to fix a table top to a base... Some of us don't even bother fixing the top to the base at all. we just rest the top on its place on the base. When dowels are used on unglued tops, they are just there as locating pins to stop the top from sliding around. If you want a strong dowel joint that does not come apart, then you have to use glue on the dowels. Dowels are generally used to join end grain to side grain, which is generally weak if only glue is used - In these situations the dowels serve two purposes - They increase the gluing area a little - but most importantly they provide some side grain on side grain gluing area which would otherwise be non-existent; and this makes a huge difference to the strength of the joint.

For example: I would definitely recommend using dowels, mortices, screws or some other strategy to join the sides to the cross brace of your desk (you called it a beam) - using glue alone is unlikely to produce a strong enough joint in that situation.

I hope that this helps

Warm wishes

Jorge

Originally Posted by snowyskiesau

These are called Zipbolts and you can get them from Elraco .

I don't mean to argue, but I typed the words "cam fixings" into the google search box to get the picture that posted... I guess that they are known by several names. I agree that they can be a bit tricky to use, and the 15+ mm forstner bits required for the cam could be tricky to find at the major stores.

Measuring twice and cutting once (and careful templating) is important when using this type of product...

Regards

Jorge

Originally Posted by zoovegroover

Hey thanks for the reply, mate. I guess I wasn't all that clear in my original post now that I think of it! Although your answer covered a lot of bases anyway

I would like the top to be flush with the sides, and the joins to be seamless, if possible. That's why I was thinking maybe dowels might work (put a couple thru the end grain of the sides and attach the top that way), but with the glue being so strong was wondering if just gluing the top down would do the trick in itself? I guess I could always supplement the glue by running some screws up thru the support beam into the top? Or would using metal brackets allow more for the movement of the wood?

Uhm... I think I may have confused "seamless" with "invisible" and "gap-less"

For a gap-less joint you will have to pull that top onto the sides at several places along the side of the desk... and figure 8's, zip bolts (cam fixings), etc, would all be good alternatives, as 19 mm is not very thick at all. These fixings are all strong enough to get rid of any gaps caused by cupping of a 19mm thick top.

With laminate kitchen counter tops, a gap-less invisible joint is achieved by putting a bead of clear silicone on the edge of the joint near the top surface. special bolts are used on the underside of the counter to pull the two surfaces together really tightly... the excess silicone is wiped or cut with a razor blade... no gap and the sliver of silicone takes on the colour of the joined surfaces. No glue is used in these joints as the substrate is too porous for the glue to be effective - the silicone is there to hide the gap and to make the join water-tight.