UKC

Ok, not a H-beam but a laminate composite, but the same physics apply.

If you were to have a third flange on a beam forming a slight arch, which is offset towards one of the 'sides', does it make a difference with regards stiffness and/or strength when loaded from above whether the 3rd flange is offset towards the top or the bottom?

Are the compressive loads on the beam the same top and bottom when shaped like an arch?

Pic for illustration https://i.imgur.com/ZPMYl9y.jpeg

If it was a straight beam supported at each end with the load in the middle, the top flange would be in compression and the bottom in tension, so where to place the third flange would depend on whether the composite material is stronger in tension or compression.

Not sure about this case as it's curved and I'm not quite following how it's supported and loaded

Did anyone else expect this thread to be about cyclotrons?

Would have been simpler with a straight beam as the material performs better in compression, so would have been doubled up at the top. Is the fact it's arched that throws me off as well.

If it performs better in compression and it was a straight beam you'd be best doubling up at the bottom as the flange in tension is the weak link so to speak.

How are the ends fixed/supported? I'm assuming it's an arch between two points with the load applied centrally or evenly along the length? Can the ends spread apart as the load is applied, i.e. can the beam start to flatten out.

I'm not sure I can give an answer though, haven't done any structural calcs since my first year at university which was about 2005 so my memory is a bit hazy.

As bit more detail and context would be needed to assess it properly, but I'll assume the beam is arched so that it is loaded only in compression (like an arch) and will be well restrained at the ends. In that case, the failure mode would be buckling of the web. A third flange could be used to stiffen the web. I can't support this with calcs but it feels like placing the third flange centrally would be best. Better still would be to add several vertical plates perpendicular to the web, or a different shape like a rectangular section. 

I watched H-beams glitter in the dark near the Tannhäuser Gate...

No, that would be H+ beams. You can’t make cyclotrons with H beams.

H+ (i.e. a proton beam) is the 'default' but people also do H- beams. Which is how I read the title

Looks suspiciously like a deck beam on a boat? From the stuff I model up in steel/alu and the few bits I've ran simple FEA on, adding the extra material to thicken up the bottom flange and then add in tripping brackets along the side of the web to stop the flange from twisting out of plane is the standard answer. The knees at the corner are generally the highest stressed areas if you're having a leg on the arch to bring the load down to a foundation.

Most composites (what material is it?) perform better in tension. 

Whilst a straight beam would be in tension on the bottom flange, and arched unit would be in compression. I believe the bottom flange would be under more compression load than the top flange.

So since compression is the failure mode (ignoring torsion failure) your additional flange would be best placed where the compression is highest. Which is at the bottom [caveat].

[caveat, I haven't got any modelling software to hand and my civils degree didn't focus much on arches, I think compression will always be highest on the bottom flange of an arched I-beam when loaded from above]

I thought perhaps C-beams glitter .. ..

Out of curiosity, What's the third flange actually doing? I guess you have a panel that's not stiff enough so you're adding a thin core and and a new skin?

I think that's going to depend how the load is applied and how the ends are constrained.

Simulate or prototype and test it if it matters.

jk

That makes sense. I was thinking more on it of reducing the tension in the bottom by adding stiffness, rather than it beimg better used to reinforce the weakest point. Guess that's where we get into the stiffness vs strength 

Loads applied both at points and everywhere as it'll ultimately be a roof, so possibly a person standing on it as well as snow 

Thanks 

OK, to give a bit more general detail as there's a bunch of people asking.

Its and epoxy/fibreglas/foam core sandwich composite making the superstructure of a 12ft wide boat. The arched section makes the roof. 

The 3rd beam is due to this section needing a slightly thicker sandwich, so will add in a thinner layer of foam core so the end result is GRE>foam>gre(heavy in epoxy light on glass) >foam>gre

So with that being said, web twisting isn't an issue. It comes down purely to whether the extra layer of epoxy/glass is mor beneficial towards the top or the bottom.

Thanks all

Instinctively I'd either keep the sandwich balanced or put a bit of extra glass in the top (greater ding and buckling resistance) and just bond the cores together with thickened epoxy, having a layer of glass in the core isn't doing much for you unless it's already there and you're just thickening an existing sandwich.

jk

Need to bulk it up by 7mm, and the extra core foam is 5mm. I want the core to be aligned up top to prevent stress points as the thicker core does not end at a fillet,  so can't bulk it up there. At the bottom those extra 2 mm might cause bubbles or delmination due to sain (using slow hardener), so the obvious place to bulk up is at the joint of the cores (~0.5mm thikened exposy in each side and a 450gsm biaxial lay up).

I'm halfway through doing ias we speak anyway. It's going up top 'just because'. Haha

* due to sagging

Beam stiffness is the cube of the depth or something like that isn't it? So if you have the choice you'd want to make things deeper.