Bent lamination - kiln dried oak.

UKworkshop.co.uk

Help Support UKworkshop.co.uk:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

ManowarDave

Fighting the World
Joined
31 Jul 2018
Messages
98
Reaction score
38
Location
Worcester
Morning all.

I'm about to embark on a rather (for me) ambitious project. I have a bed frame to make for our new loft conversion.

The design authority found some inspiration on line and I have designed something less extravagant that meets approval. However it does still incorporate two curved posts to form the headboard frame.

The frame is to be made of kiln dried oak that I've already purchased from a local saw mill.

So, my question in relation to the curved posts. I will be forming these using bent laminations. The finished length of the lamination will be about 2.3m. The inner radius of the curve is approx 1.45m. To minimise wastage and resawing time, what is the maximum lamination thickness you think I can get away with on that bend radius?

My gut feeling is telling me 8 to 10mm maybe even up to half an inch if it were air dried and not kiln dried. Thoughts or prior experiences please.

TIA

Dave

Edited to add length of lamination.
 
Last edited:
That's quite a long radius, and assuming the 2.3m dimension refers to the arc length rather than the chord length I think your estimation of a laminate thickness somewhere between 8 mm - 12 mm is probably in the right area. The best thing would be to build your bending former to the arc determined by your the inner radius (1.45m), then clamp a sample of your wood of the right length and requisite oversized width at an intuitively chosen thickness, say 18 mm, at one end, and see if you can get it to conform approximately to the former's arc just by pulling or pushing it into place by hand. If not, plane off a couple of mm, and keep testing.

When you can get it somewhere near just by hand, that thickness should work for all your laminates, remembering, of course, that you'll have devised a set up where you can get lots of pressure (clamps perhaps?) in place to bring the glued lamination into its required arc, although you will of course need to consider springback. This link might help. Slainte.

PS. Afterthought. I'm assuming you realise that to create an arc length of 2.3m (or chord length of 2.3m) you'll need to bend material that is about 250- 300 mm, or more, longer than your finished length, so you'll need to bend pieces roughly 2.6m long, or longer, to get out a satisfactory arc length of 2.3m. If you're talking chord length, then you'll have to calculate that overlength bending length need yourself. Sorry, not doing those sums for you, ha ha.
 
Last edited:
My gut feeling is similar to yours but there may be other implications. Answers too these queries may throw more light on the subject:-
What's the overall width you're aiming for?
Is the whole 2.3m to be curved at 1.45m?
What size stock are you starting with?

Brian
 
Yes good questions, but there is another one as well, how thick is the finished piece going to be? It should ideally be a multiple of the thickness of laminations, plus a bit for cleaning up, which is my next point, the innermost lamination on the curve should be hand planed as you would wish it to finish, it’s an awful lot easier to plane it whilst it’s flat on the bench rather than in a curve, And just in case you weren’t aware, if you are down to strips that are two or three mil thick you need to have them just a little bit longer so that you can cramp them at the end at which you are starting to plane, fairly obviously if you try and plane them against a stop they will buckle up and snap.
It sounds like a nice project to get your teeth into to me, happy laminating. Ian
This is a piece I made a few years ago.
546AD84E-A953-4CC5-9B67-F7D72B22EEF1.jpeg
 
Thanks gents,

A little more background info to answer the questions and flesh it out.

The 2.3m is the total length of the final outer curve and includes a 375mm straight section at the one end.

The inner chord length of the curved section will be1.45m.

The final cross section of the laminations will be 75mm square or more probably slightly under at around 72mm square.

The laminations are being ripped from a piece approx 200 x 85 x 3200.

Here is what I'm hoping the lamination will end up as.

a2-f10-m5417-image001_2epng_4001D6B536_2e09411550.png


And if all goes well, the final bed will look like this.

a2-f10-m5416-image001_2epng_4001D6B535_2e50B658B0.png


Cheers

Dave
 
You mean something like this:-
Bed.jpg

This is my first attempt at laminating some 20 years ago. It's made in S. Yellow Pine. The laminae are (from memory) 6 x10mm.
The obvious solution for yours is 5 X 15mm but 15mm might be a bit too much. I think I would go for 8 X 9 and accept 72mm.
I would extend the curve at the top end by at least 150mm when making the formers as it's very difficult to get the ends of a curve to adhere correctly so extend them and cut off the waste. Do the same at the bottom end but 75mm will probably suffice. Don't forget to allow for the extra length of the outer laminar.
I note you are putting a little finial at the join. As you can see I did the same but mine covers a rather poor joint. The L & C are the initials of the bed occupants. I heard from them recently and they reported that the bed is still in daily (or nightly) use.
Brian
 
My limited experience suggests that there will be more waste than you expect, from both ends. Make the form longer than you need - I didn't and just hung the extra off the edge at either end, which limited my choice as there was no curve on the end bits. I also discovered that additional full length sacrificial strips inside and out are great because then you don't have to worry about clamping marks and faffing with blocks for each clamp. I lined mine with packing tape so they didn't stick (too much), and used polyurethane glue rather than the usual. It was still a hugely stressful, disorganised glue up.

This is in pine, with I think 7 10mm width strips. The curve radius is unknown, but about 2 metres, possibly more? (I used a factory produced example as a template, which was cheating but made it all much easier - I just copied what they did).

20200730_090440_copy_1328x747.jpg20200809_103427_copy_1328x996.jpg
 
Thanks Brian, TN,

How much (if any) spring back did you see once the lamination was released from the form?

Cheers

Dave
 
Hi
I cant advise on lamination but just a simple comment - make sure the finished shape/curve/size will fit around any corners, staircases and doorways on route from the construction area to loft extension destination.
cheers
 
Thanks Brian, TN,

How much (if any) spring back did you see once the lamination was released from the form?

Cheers

Dave
Spring back is always a problem in these designs. It can depend on a number of factors such as wood species, number of laminae, glue and, of course, bending radius. I think the reason I had trouble with the closing joint was that I hadn't allowed enough extra wastewood so I had to resort to a couple of big screws to pull the ends together. In this design I wouldn't worry too much about spring back, just live with the shape you end up with and shape the panel to match.
Remember to use a hard setting glue such as Aerolite One Shot, Cascamite having now fallen out of favour.
Brian

PS What's TN?
 
Thanks Brian,

Re Glue, I had planned on using a PVA. Gorilla wood glue is what I have to hand and I have to say I've never had any problems with it in normal glue ups. Alternatively I have some clear laminating epoxy. Thoughts?

Dave

TN is Trainee neophyte.
 
Re Glue, I had planned on using a PVA. Gorilla wood glue is what I have to hand and I have to say I've never had any problems with it in normal glue ups. Alternatively I have some clear laminating epoxy. Thoughts?
It's generally best to avoid aliphatic or PVA type resins: they suffer from glue creep, i.e., bent laminations tend to straighten because the adhesive retains a certain amount of elasticity.

It's best avoid polyurethane adhesives also: as these cure they expand into a foam, so if you're unable to get adequate pressure along the full length of the arc you might see foam filled voids here and there along the laminates length. If you can get adequate pressure as I described, then this shouldn't be a problem.

Best adhesives for this kind of work are non-creepers such as urea formaldehyde formulations (Aerolite One-Shot has been mentioned as has Polyvine Extramite (Cascamite), but there are other brands ). Another good one you've mentioned is epoxy resin, and there are slow set versions available, up to an hour's open time, well suited to this kind of work - the laminations of the table below were made with epoxy resin. You can also add fillers to this stuff if required, but you shouldn't need it for laminating if the bending former or jig and clamping is well planned and made so that no voids are possible.

As to springback, did you click on the link I provided earlier? Slainte.

AD-Table-1870-700px.jpg


Sq-Corn-Cab-800.jpg
 
How much (if any) spring back did you see once the lamination was released from the form?
Virtually none, but it wasn't mission critical so I didn't measure it. When I offered it up to the rod it was out by perhaps 2mm over the length - but I've slept since then and can't remember exactly. Sorry.
It's best avoid polyurethane adhesives also: as these cure they expand into a foam, so if you're unable to get adequate pressure along the full length of the arc you might see foam filled voids here and there along the laminates length.
Oh dear. Did I mention I don't know what I'm doing? Next time I'll get it right, hopefully.
 
You obviously know what you’re talking about Richard, I have been making laminated pieces for years using Pva and have never had a problem, perhaps I’ve just been lucky. I know it’s difficult to say but how much are we talking about?
 
You obviously know what you’re talking about Richard, I have been making laminated pieces for years using Pva and have never had a problem, perhaps I’ve just been lucky. I know it’s difficult to say but how much are we talking about?
Well, I remember back in the 1970s an acquaintance making a sofa with a curved laminated rear cum side rail using PVA adhesive. He ran a groove along the top edge to accept the metal tangs of the rubber seat webbing. A few years later he was called to say there was a problem. The problem was that those metal tangs no longer sat neatly in the groove because it was now morphing into rebate towards the inner edge. Essentially the people's weight as they sat was causing the glue lines to shear resulting in the laminations sliding past one another. I can't remember if or how he fixed the problem.

As to shear in the longitudinal direction, i.e., following the arc of the bend, shear slippage (glue creep) depends on the tightness of the bend. Gentle bends, as is proposed by ManoWarDave, and illustrated by Brian's bed, it's not likely to be a big issue, but it would be a bigger concern for the table I showed earlier in this thread. Generally, I prefer to avoid any likelihood of such problems through use of a non-creeper whatever the tightness or gentleness of the bend so, for example, the circular connections (legs) between top and foot in my table were glued together half a circle arc at a time with slow-set epoxy resin in a bag press over a former. Slainte.
 
Hello,
just resurrecting this thread as there seem to be some relative bent wood experts around.
I’m planning to make some curved oak steps in a quarter circle for the garden to match the oak sleeper walls I’m currently installing.
The structure will probably be galvanised steel but the treads I want to look like 100 x 200 (maybe x 300) sleepers laid flat.
The top step will have inner radius 600mm, the next 900mm, then 1200mm and the last(bottom) step 1500mm.

Is this possible/practical by bent laminating or better done by segmenting (with dominoes) and routing round? In either case will the glue stand up to many years of weather?
My instinct is that laminating will be much more effort but more likely to stand up to the weather.

Any advice?
Thanks
 
Hello,
just resurrecting this thread as there seem to be some relative bent wood experts around.
I’m planning to make some curved oak steps in a quarter circle for the garden to match the oak sleeper walls I’m currently installing.
The structure will probably be galvanised steel but the treads I want to look like 100 x 200 (maybe x 300) sleepers laid flat.
The top step will have inner radius 600mm, the next 900mm, then 1200mm and the last(bottom) step 1500mm.

Is this possible/practical by bent laminating or better done by segmenting (with dominoes) and routing round? In either case will the glue stand up to many years of weather?
My instinct is that laminating will be much more effort but more likely to stand up to the weather.

Any advice?
Thanks
A drawing/sketch would be helpful.

However, laminating may not be a good idea for external constructions: weather conditions may lead to failure of the glued joints over time, e.g., a year or so. Also laminating is time consuming and requires jig making plus lots of clamps.

You say that you want an appearance of sleepers 100(thick?) x 200 mm (maybe X 300) sleepers which leads me to ask: why not just use sleepers of those dimensions and cut the arcs/curves you need? And where you might need to join or butt the 100 mm? edge of one sleeper to another create structure beneath the step to hold the pieces together as well as perhaps edge joining with a domino/biscuit/mechanical fastener plus polyurethane adhesive (waterproof but not structurally gap filling).

Just a few first thoughts. Slainte.
 
Spring back is always a problem in these designs. It can depend on a number of factors such as wood species, number of laminae, glue and, of course, bending radius. I think the reason I had trouble with the closing joint was that I hadn't allowed enough extra wastewood so I had to resort to a couple of big screws to pull the ends together. In this design I wouldn't worry too much about spring back, just live with the shape you end up with and shape the panel to match.
Remember to use a hard setting glue such as Aerolite One Shot, Cascamite having now fallen out of favour.
Brian

PS What's TN?
Cascamite is now manufacrured by Eureka and is fine.
 

Latest posts

Back
Top