Does anyone know any references about design of composite FRP I-beams?

[JMiller1357]

Does anyone know any references about the design of composite fiberglass reinforced plastic (FRP) I-beams? For example, the flanges are aluminum and the web is FRP. Particularly, what ratio of web height to thickness of the web should be used to resist local web buckling? 25? 20?

Thanks for any input.

 

[40818]

You could first try using a method called (amongst other things) "method of transformed sections". Its pretty basic but will allow you to get a decent ball park set of results which you can build on further.

Assuming that there is no sliding between your aluminium caps and your FRP web during bending, then the normal theories still hold true. So the elongations and contractions of longitudinal fibres are proportional to the distance from the neutral axis. Due to differences between the modulus of elasticity of your beam components, the part with the lower modulus will be trasnformed to an equivalent thickness in the ratio of the modulus'. Obviously you have to recover actual stresses within the web and by using the ratio applied to the stresses (which will be calculated as either of your materials. You will also have to ensure that the joint between your web and caps can maintain integrity due to the applied shear between them.

As for the ratio of the web height to thickness (i'm aero so we do it all by hand without set guideline), then i would obtain the web shear stresses and combine them with the induced compression stresses from bending and compare them to your calculated your combined bending/comptression buckling allowable. Depending upon the loading set up, your compression will probably be biaxial due to direct compression from the loading. You could also check the web for buckling by assuming its a simply supported plate. And dont forget crippling of the cap flanges.

 

[JMiller1357]

Thanks for detail input.

 

[40818]

By the way, why have you got 2 handles, and please dont double post in different forums.

 

[Jonmiller]

I can't login the first one anyway.

 

[wes616]

How large is your beam anyway?

Shearweb = 45 deg plies (unless you subscribe to that silly 10%rule) only!



Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[Jonmiller]

3" (depth) x 0.5" (width). The alum flanges is approximately 0.080" and the FRP web is 0.090". The loading about 40 lbf/ft uniformly acting at the top flange of I-beams.

 

[JMiller1357]

Anyone can input a fresh idea about the connection?

I think there should be other better way to design the connection, such as snap-in-lock method. However, the tolerances of an aluminum profile and a pultruded shape are much different. Therefore, if the pultruded shape snap into an aluminum profile. There will be a very small movement and not good enough. By the way, a die cost of a pultruded custom shape is too expensive and then purchase price also too high comparable to an available FRP sheet. As you can see, just buy the flat FRP sheet and cut it to a custom web you want, easy!

 

[wes616]

Hi John,

One of the problems with cross posting, is many of us don't follow the posts into another forum. Keep that in mind while asking questions on eng-tips.

If anyone is interested here is the tread to the post in the structural engineering fora: thread404-197304

A couple of things to point out. You will probably never get a "perfect" composite I Beam.

Concerning the buckling of the web:

Roark's (7th edition) pg 182 (paraphrased)

S= V/Aweb = unit load carried by a thin strip modeled as a Euler Column.

Bruhn (sec C5.8) defines formula for critical buckling of a flat plate in bending (elastic buckling) as:

Scr= ((Pi^2 * Kb * E)/12(1-nue^2))*(t/b)^2

Where Kb is the buckling coefficient (a graph is given) for various rations a/b (values range from 24 to 48)and edge restraint. (Glass fiber has a fairly high strain to failure ration). I'd give you a definitive answer if I knew the length of the beam.

Keep in mind that as I mentioned in a previous post, if your shear web is oriented at +/- 45 deg, the fibers are in alignment with the shear loads, therefore the fibers are in tension/compression. My experience with pultrusions are uni-fiber and that makes them poor shear members.

I don't really follow your attachment methods. We use adhesive bonding of shear webs to spar caps (usually composite as well in our case) using glass shear tapes. This method may or may not work for your application. Generally though, you can get a good metal-fiber bond with glass.

I recommend performing a test of your beam, if there is any concern.

Also, keep in mind that hand calcs with composites are difficult and generally need to be conservative. This should be somthing that could easily be fem'd.


Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.

 

[Jonmiller]

Wes C:

My I-beam is 8 ft long with 2 ft O.C. lateral supports to the web. Linear loading is 40 lbf/ft applying at the top flange. I figured out the ratio of web height to web thickness is about 18-22 for web thickness <=0.100" and 22-25 for web thickness >0.100". So what's your opinion about the ratio to size the web?

Thanks for your advice as well as the suggestion about the cross posting.

John

 

[wes616]

Jonmiller,

You should do the caluclations, in lieu of trying to define a ratio. Based on the bruhn calucuations, the a/b is off the charts.

a/b=(8'*12"/1')/(3-(2*.080))= 33.8

You should try to perform the calcs according to Roarks.



Wes C.
------------------------------
No trees were killed in the sending of this message, but a large number of electrons were terribly inconvenienced.