Box beams and wood laminates, to core or not to core

[Boston]

I'm about to head back down to the shop today, and I've a bunch of stuff I'm gluing up anyway. I thought I'd make a few groups of samples of both box beams and laminated beams just to see which ends up on average stronger

my two basic designs for test pieces are going to have similar sections of 2" x2" each and 3' lengths. Cedar core of 1" x 1" on the box with veneers of 1/2 and just veneers of 1/2 on the laminated beam. I changed those dimensions based on the size pieces I had in the shop, the cores are 1 x 1 1/2 and the veneers are 3/8 lengths are now 40" . I'm kinda curious as to which one might hold more weight. Both will have the same volume of each type of material, W oak and red cedar.

anyone want to make any predictions
I'm going to point load the center of each to destruction on a press and see what I come up with

cheers
B

almost forgot the key element of also trying it without the core on the box beam. Trying some drilled and some with spacing blocks instead of solid cores Also thinking I might try some compound boxs with and without cores. Takes pretty much no time at all to do cause I've got so much waste off this window job, got in another thousand feet of W oak just the other day and I've got some left over clear red cedar from an earlier project.

[Tad]

What are the beams for?

Without the core you'll have buckling problems and so add web blocking (bulkheads tying surfaces together) at various spacing.

[John Riddle]

Boston:

I'll take the box beam for $50. If I understand what you're doing, the oak content will be three times the cedar. See sketch.

[CatBuilder]

I'm going to bet on the box as well. Much stiffer.

[Petros]

It is not weather it is hollow or not, the soild core beam of the same dimension will be stronger. But the hollow one, if built correctly, will have the best strength to weight ratio, and make the most efficient use of materials (uses less). You can save weight with hollow beams, you can make it stronger by putting light weight blocks up the hollow core to prevent it from buckling. It will be lighter than the solid core beam and almost as strong, but it will require more parts to assemble (more labor to make).

If want better strength to weight ratio make it larger with thinner walls, but this configuration is more susceptible to damage from rough handling.

Strength of a hollow beam is a complex question, a truss would be the best strength to weight ratio, but also require the most parts and detailed assembly. It is also the least damage tolerant, and it can be optimized for only one axis of loading, or multiple axlis of loading.

If you want light, you need more smaller pieces arranged to be optimized to the design requirements. If you want fast and simple assembly with fewer parts to make it will save time, but at the cost of extra weight.

[Boston]

Hi Tad, I'm just playin around again since I"m so close to having the truck done. I've got a bunch of cut offs in the shop from all these 1/2" window laminates so I thought I'd glue up some stuff and see what gives me what in terms of bending under load and to destruction.

none of it really applies to the build except it gives me some excuses to buy more clamps

I've got three beams made today, not quite of the dimensions specified earlier but close, I went with the predominant scrap size on my left over cedar and then cut the oak pieces to it. About 1 x 1 3/8 core with 3/8 veneers of white oak. Went long on the vertical side and short on the horizontal tops and bottoms. Not sure if thats the prescribed method but its what I cut today. My apologies to John who went to all that trouble with the cad drawings. Its kinda what I had handy and what I can pound out in the largest convenient size, my theory being that the larger the sample the more accurate the findings. Assuming I get scientific about it.

Deal is I had 1/2 inch W oak strips but they were rejects from the windows, I planed them down to 3/8 to remove the various imperfections. I also had five or six 8/4 sticks of clear red cedar in the rough so I planed it down till it was spotless, ended up losing a lot of wood on that.

I'll weigh each piece and measure for bend at each x time period under load. I also don't think my idea of using the press is going to work out so I'll just build a rack and hang a load off each in turn. Press isn't exactly a fine tuned instrument.

the first three are about four feet long each, once again the length of the scrap, so I'll cut each in half ( 2" lengths ) and that will be my standard test subject length. Having build houses for so long the standard rule of thumb for cutting holes in floor joists also came to mind and I thought I'd make a few with the 1/3 perforations in the center 1/3 of the beam and see how they stack up against unperforated beams

I guess I was reading the Pine VS spruce thread and was kinda curious about combination's of materials but didn't want to screw up a perfectly good thread over there, thought I'd glue up some stuff and see what happens. Most of the engineering tables I have deal with simple beams rather than composite structures so, what the hell.

cheers
I'll post some pictures of the test pieces when I get a few more done.

oh
I'm thinking the perforated box beam with a core is going to be my best strength to weight but the hollow compound box beam might be a close second. Might depend on how crazy I get with the jointery on it. ;-)

I also thought of gluing up some sandwiches of cedar and w oak say 3/8 each in the same dimensions as the cored box beams so maybe four layers of cedar and five of Oak and see how it stacks up. Once again I'll weight each sample and calculate the percentage of oak to cedar in order to get a fair comparison of how the different configurations compare

[Boston]

cut enough stuff for 15 test pieces at 40". Yah I know I mentioned 20" pieces but the longer piece means I don't have to use as much weight, and I've got plenty of left overs I'm slicing and dicing things out of

doing three with solid cores of red cedar, three with hollow cores, three with 2" cedar blocks at 10" intervals and I'll probably do glue lam types with the rest of the stuff.

I did notice on some the wider stock that the glue makes it want to cup away from the core, specially on the ends, which is why I haven't cut anything to length till after its glued up. Gives me about 48" blank cut down to 40" once its all set up.

cheers
B

[Petros]

rule of thumb for spacing of web stiffeners is 2x height, so you might glue up a sample with your blocks at 2x spacing (would that be 4"?) rather than 10", to see if it makes any difference in strength.

[Boston]

so I made one drilled out like an engineered steel beam but it was a pain in the ass to do and no way would this lend itself to a compound curved structure
getting ready to glue up several as tad suggested
going to try a few like Petros suggested but again webbing doesn't seem like it lends itself to a compound curve as well as just block stiffeners every x inches
and I have several simple box beams with cedar cores
I'll make a few glue lams of just oak and a few of oak cedar layers

[Boston]

weighed some of the pieces today, all are 40 inches long with 3/8 veneers of W oak and Red Cedar core in various details

solid core weighs 6.75 oz
solid core drilled out weighs 5 oz
block core weighs 5.5 oz

going to glue up a few more block cores and drill them out as well

Petros
I might try that but the blocks wouldn't be 2" long if I did, more like 1/4 and I'd go C to C on the spacing. Which would be about 4". I'm thinking that it would weigh in at about 5.5 oz maybe a touch lighter

I'm thinking the best weight to strength ratio is going to be a block core, drilled out. But its also quite the pain in the ass to do and wouldn't lend itself to building in a compound curve form very well. Also sealing the inside of the box would be a pain. I'm expecting that this configuration would weigh in at about 4 oz.

cheers
B

[Yellowjacket]

Boston,

All of this can be (relatively) easily compared using simple beam theory and by calculating the the relative section properties of what you are trying to create. You can do all this by testing, but it's a lot easier (and more accurate) to do it by calculating the section properties of the beams that you want to compare.

The stress in a beam is simply M*C/I, where M is the moment in the beam times the max distance (C) from the netrual axis, divided by the moment of intertia of the beam (I). Cross section inertia is higher, and stress is lower (on a strength to weight basis) for beams that have reduced material in the web of the beam, and concentrate material in the beam caps. This is all basic beam theory and optimizing the strength to weight ratio means removing material in the web of the beam and putting more material further away from the netrual axis. Here is a link that explains it:

http://www.engineeringtoolbox.com/be...on-d_1312.html

You can relatively easily calculate the the I value and neutral axis locations for various other beam sections using the "parallel axis theorem", but you pretty quickly will find that I beams and box sections are by far the strongest beam sections.

The problem in working with wooden beams is that wood is very anisotropic. This means that wood has more strength in specific directions than it does in others. In wood, the shear strength of the material is substantially lower than the compressive or tensile capability along the grain. Wood basically splits along the grain and the strength in this direction isn't high. To put it another way, in an isotropic material (like most metals), has a shear strength that is (roughly) 1/3 of the tensile or compressive capability. Most wood, on the other hand is much lower in strength shear relative to it's tensile and compressive (along the grain) strength. This means that the web of a wooden beam is going to have to have more material in it than it would if it was made from an isotropic material like metal.

You can diddle around with a variety of cross sectional approaches, but the bottom line is that if the shear section in the middle of the beam is strong enough, the failure location moves out to the outermost fiber on the top and bottom of the beam. If you take too much material out of the web, the beam will fail when the shear strength of the web is exceeded.

If you do some serious analysis you will find that a plywood web gives you a web that can withstand the necessary shear stress better than a piece of solid wood. A box section with thin plywood shear caps is exceptionally strong and a lot lighter than the sections that you are working with. Also additional analysis shows that an I beam has the same strength as a box beam (it has a lower torsional stiffness, but we are just talking beam strength here, if you need higher torsional strength then a box section is superior). The advantage of an I beam is that it doesn't hold water inside and there are advantages to that in a boat.

There is a pdf titled "Design and fabrication of glued plywood-lumber beams". Take a look at that and it explains how strong this type of beam is.

Since this is going to be used in a boat you are likely going to want to seal the wood to prevent rot. If you really want to high strength to weight ratio, make up an I beam with a plywood web, then laminate a strip of fiberglass tape to the top and bottom of the beam and one layer of glass on each side of the web.

As a practial matter, the beams that you are constructing are relatively poor from a strength to weight ratio basis.

[redreuben]

An additional way to substantially increase strength without weight is to use unidirectional fibres like glass or carbon on the caps.
RR

[Boston]

interesting stuff Yellowjacket, I'm on hold while I get this window job buttoned up for the winter and the truck running finished. Just started it today after a 5 month or so retrofit. First time. Ran great. A few more bits and pieces and then its on to the cosmetics.

[PAR]

The basic point the Petros and again with Yellowjacket's reply is strength to weight for the application. It's fairly simple to spec a big ass hunk of wood, that more then covers the application. The problem with this is you have to pay for, carve up, and install this big chunk of lumber, plus the weight penalty that rides along just for giggles. Now, knowing how you feel about big hunks of timber, there's certanily an appeal, but unless you're building a Buehler or other rather antiquated design, that calls for massive timbers, so that 20 pound cannon balls might bounce off, you need to consider the strength to weight ratio. It's not enough information to know if one construction type is stronger then another. You need to access the application and apply an appropriate structural element that "fits" and carries the safety margin you've elected to employ.

[sabahcat]

Ply on edge is incredibly light and stiff if you can keep it in column.