Thread: Trouble Shooting Splined Bent Lamination

When I do the next piece the lamination will be longer - filling / using the whole former then the finished piece will be cut down to length.
I will make the inserts thickness sanded to my bandsaw kerf so that next time I can have accurate slots.

I have machined up the final pieces that will be used for my laminated curved stretchers at 30 x 68 x 500mm (x2).
Also the splines at 3.2mm thick

Plan

I have cut the spines at 3.2mm and then run through the thickness sander down to 1.5mm.

This means I make eight 1.5mm kerfs each spaced 2mm apart.

-|-|-|-|-|-|-|-|-

2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2 | 1.5 | 2

Reasoning

Thinner laminations tend to be easier to bend and 1.5mm is my bandsaw kerf simplifying the cutting of the slots


Thoughts?

IMG_1184.jpg IMG_1185.jpg

Best of luck

Originally Posted by GraemeCook

No, no, and no. I think that this is one of those "solutions" that is good in theory, but not so good in practice.

When you put the glue between the laminates you are effectively lubricating the laminates and they wish to slide everywhichway, usually where you do not want them to go. One solution is to put a tiny amount of sand or salt in the glue; this minimises the lubrication and makes them a lot more stable, but it is still a "slippery" operation... Adding a loose spacer just complicates an already difficult operation.

I frequently have to add a couple of clamps orientated vertically to resist any attempts at vertical sliding by errant laminates.

Well, I beg to differ and I have done so many hundreds of times with far more "slippery" glue ups making inlay banding blocks from hundreds of very small components! I think the quality of my glue ups stand testament to my methods.

Using sand or salt as you suggest may work but offers little opportunity to "adjust" during final assembly.

Yes glues do for a while act as a lubricant and will make for a slippery assembly. That is both a benefit and can be a hindrance. We want the laminates to be able to slip against each other as we form the bend, however once the desired shape is achieved we want them to stay in place. For that to occur we must control any potential movement whilst firming up the clamps. Having control is the key, however we only have a short assembly "open time" before other factors start to complicate matters and potentially risk losing the glue up.

The spacers are not "loose" they become a part of the glue up assembly and are treated as waste to be cut away or trimmed. Again preparation and control are the keys to a successful glue up. A traditional stacked bent lamination is relatively easy when sufficient excess length allows for slippage. The splined lamination requires one end to be "fixed" whilst the other end is "free." Ideally the fixed end will be tight and have no visible gaps or wide glue lines on completion.

So to control the "fixed" end we can spread the glue; assemble and "pin" the laminates; then place in the formers and clamp up hoping the pins hold.

Another method to control potential gaps at the fixed end is to cut the strips much longer than required so they protrude past the former blocks. Once the initial light clamping pressure is applied the protruding end of each strip can be lightly tapped with a small hammer (and suitable drifts from the strip waste) to close up the gaps on the fixed end. Start with the widest gap first as it will drag others with it. Then progressively but quickly work out the gaps. Clamp up tight to close the gaps but not to firm to squeeze out all the glue.

Or you can do a "dry fit" assembly first - no glue! Proceed as above and once happy mark a line square across the laminates longer than the required "final" length and trim. This creates a single square end that can accept one spacer block (machined to the desired final dimensions) that will become a temporary part of the glue up assembly. Proceed with a second dry fit assembly to check all is good. Then do the final glue up as above, however this time we only have one block to tap to close up the "fixed" ends of the laminate strips.

Preparation is key as the assembly glue open time means that we must be organized and have every thing at hand, ready to go. There are no half arsed ways to achieve good results with lamination because machining tolerance and or glue up assembly errors become quite noticeable and permanent. Preparation, quality control, organization, and controlled speed!

Good luck, and try a scrap assembly first, however it must be machined to tolerances as if it were the final piece. Pine dunnage is a wonderful thing.

Having the slots in the parent stock cut on the bandsaw and the lamination strips sanded will give you reasonable friction in clamping but the very thin strips are going to be difficult to tap up firm. I would suggest the dry fit method, cut to the bent length and use the one machined end block.

ps - with the dry fit method we can also steam and "pre-bend" the components making final assembly much faster and more accurate.

Good Morning Mobyturns

I really do not have a disagreement with what you are saying, except that I think that you may be under-valuing your own experience in laminating.

Thumbsucker is at the other end of the experience spectrum, a self-confessed novice at curvilinear lamination and he is jumping in the deep end:

• he is doing quite sharp bends, not gentle sections of large circles, and
• he is inserting splines in the middle of the curves.

These are quite advanced laminating jobs, not what I would chose to learn on.

Those sharp bends really load up the springiness in the laminates, those loose splines will be quite contankerous, just like herding cats, as BL would say.

I have been reading and thinking if I increase the lamination's and kerfs to 2mm thick each then we reduce this problem.

Originally Posted by Mobyturns

but the very thin strips are going to be difficult to tap up firm

So I am thinking doing it like this

2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 | 2 = 30 mm

2.5 | 3 | 2.5 | 3 | 2.5 | 3 | 2.5 | 3 | 2.5 | 3 | 2.5 = 30 mm

This sounds like the method I will try

Originally Posted by Mobyturns

Or you can do a "dry fit" assembly first - no glue! Proceed as above and once happy mark a line square across the laminates longer than the required "final" length and trim. This creates a single square end that can accept one spacer block (machined to the desired final dimensions) that will become a temporary part of the glue up assembly. Proceed with a second dry fit assembly to check all is good. Then do the final glue up as above, however this time we only have one block to tap to close up the "fixed" ends of the laminate strips.

I plan to do a practice run in some pine. Several if I need to until I pull it off.

Thought - would doing a draw bore method aid me? The idea is to drill two stacked blind 3mm holes (from the inside face of the stretcher) through the lamination were they start in the kerf and pin them with some bamboo skewers. This would fix them hard at one end but should allow them to slide past each other through the bend on the other end? It is adapting the idea borrowed from laminated plane bodies.


Screen Shot 2021-06-05 at 6.42.44 pm.jpg

Originally Posted by thumbsucker

Thought - would doing a draw bore method aid me? The idea is to drill two stacked blind 3mm holes (from the inside face of the stretcher) through the lamination were they start in the kerf and pin them with some bamboo skewers. This would fix them hard at one end but should allow them to slide past each other through the bend on the other end? It is adapting the idea borrowed from laminated plane bodies.


Screen Shot 2021-06-05 at 6.42.44 pm.jpg

To be honest, the KISS principle rules imo, keep it simple!

I would start with a scaled down version of your project - not so many leaves in the lamination and using a larger bending radius. Do a simple stacked lamination with say 5 leaves. Then when you have that sorted out move to say two kerfs in the main stock, which makes 3 leaves in the main and two veneers. That will give you an appreciation of the many tasks and the short open time you have to work with. Then tighten up the bending radius and practice some more.

To pull off what you are proposing is quite difficult to do without assistance. Spreading glue on that many surfaces then assembling requires at least two persons, three would be better. When we built 16' sailing racing skiff frames we had four people involved most of the time, so had extra hands when required.

Remember you only need one poor glue joint in that sort of glue up to create a failure.

I have a friend who can help me and I was considering using an epoxy. To take the pressure off and give me breathing room.

Originally Posted by thumbsucker

I have a friend who can help me and I was considering using an epoxy. To take the pressure off and give me breathing room.

Good idea. And consider using a slow hardener with your epoxy. For example, if using WEST System use hardener #206 rather than #205. I stopped buying #205 about 20 years ago!

I have never tried using a skewer to hold things together but I am wary; it will be visible on the finished faces and it may complex the clamping process - it its too long or an outside laminate "jumps off" the skewer. But I haven't tried it.

Adding to what Mobyturns said, the outside raduis of a bend is always a little longer than the inside radius (simple geometry). The splines will move a little during the bending process because of this factor. You may notice gaps openning at the ends of the splines. During your dry practice runs may I suggest that you rotate the bending piece 180 degree so that it bends in the opposite direction. You may notice that in one alignment the splines move together better than in the other.

This is a very difficult job and you just have to find what works adequately.

My experience is that after you have done 15 or 20 glue ups, after each you think "That was difficult, but we got there..." Then suddenly they start to come together relatively easily - you have passed some magical threshhold - but never rest on your laurels as the next batch may contain a really contankerous laminate that wants to fight.....