Aluminum Mullion Reinforcing!

[pranayvarma]

I have an aluminum curtain wall mullion which needs to be re-inforced using steel tube. I would be inseting the tube into the mullion hollow space. Now how do i analyse this combination for strength & deflection. Do i need to connect the steel stiffener to the aluminum member at regular intervals and check for shear just like in composite structures. What about Bending Moment, how do i analyse for strength. Can i distribute the Bending Moment among the stiffener & the mullion as per ratios of their respective rigidities (EI/L).

Thanks in advance

 

[LowLax]

I don't consider them composite. I apply a proportion of the load to the aluminum and the rest to the reinforcing based on their EI ratio. They should both come out with the same deflection.

 

[pranayvarma]

thanks, but what is the theory behind this method...?

 

[lkjh345]

Relative Rigidity.

LowLax is correct. As long as their deflection are the same (as they must be) then the amount of the wind force they take is porportional to their EI

 

[pranayvarma]

thanks all. so do i need to connect the stiffener to the mullion along it's length or just sliding it inside will work (I need to keep a gap of 2~3mm allaround the steel for PVC seperator)?

more importantly i have to prove this system works using some theory in textbooks...what do u all suggest?

 

[LowLax]

Just let the stiffener slide inside the mullion. I'd call it relative stiffness theory or something like that. I'm not sure what the text books call it. You're just making the assumption that deflections are equal. So you have two equations and two variables (w1 and w2). I'm thinking of this in terms of simple spans and uniform loads.

w1/(E1*I1)=w2/(E2*I2)
w1+w2=wtotal

 

[pba]

If there is a gap all the way around the steel insert then any applied load will need to be applied to the sluminium extrusion first. This will bend until it contacts the steel at which point the steel will start to contribute to the overall strength. If the gap is packed with insulation (?) then it must be capable of transfering the load from the aluminium to the steel.

Potentially with a 3mm gap the middle of a simply supported member would deflect by 6mm before the two ends and the middle were all in contact with the steel.

The principle of relative stiffness does still hold true but the aluminium would be 'pre-loaded'.

 

[pranayvarma]

thanks all for u'r inputs...now i have proposed this theory of relative stiffness, the proof cheking engineer doesn't seem to agree with this...as per his comments,this is against theory of elasticity...(i don't know how). He is in no mood to explain anything more!!. SO what else should i provide to him so as to prove this theory correct...

 

[UcfSE]

You should get out some mechanics of materials books and study. We cannot teach you over the internet. I suggest you look up flexural rigidities and put your thinking cap on. You may also lookup "transformation of sections" or "transformation of materials".

If you cannot connect these two members enough to make a composite section, you'll have to add their strengths together, properly proportioned for the difference in stiffness of the two. Keep in mind that you will need to connect them somehow at least enough to insure that load is transferred into the steel whether it is composite or not. If your steel section is not connected to supports then you will have a nonprismatic beam and will have some additional work to do (different from what is being discussed so far), as well as additional thinking to understand how the two members will interact.

 

[GNight]

If you can, make the steel section strong enough to take all the load.

 

[apsix]

My view;
Forget composite action.
Assume the sections are locked together, say with a series of self tapping screws along the length; you then have 'compatiblity of deflection'.
As deflections are the same (for UDL, simply supported);
w1*L[sup]4[/sup]/(76.8*E1*I1)=w2*L[sup]4[/sup]/(76.8*E2*I2)

Cancel out like terms and you end up with;
w1/(E1*I1)=w2/(E2*I2) as posted by LowLax
You also know w1 = w2;
2 unknowns and 2 equations; what could be simpler?

How can a checking engineer argue against that?

 

[apsix]

Of course it's not that simply; w1 does not = w2

It should have been w1+w2 = wtotal


Note to self; always double check if there is no edit function

 

[pranayvarma]

thanks all for your views...so infact the section is not supposed to be composite...but the screws need to be checked for shear, is'nt it??
what if the sections are only connected at their ends by means of a thru bolt?
and there is the other problem of thermal variation...what happens then?

 

[LowLax]

If the aluminum mullion can deflect a 1/4 inch before it hits the reinforcing steel I think you'd have:

defl1=defl2+1/4

which would lead to:
w1/(E1*I1)-1/4 = w2/(E2*I2)

and
w1+w2=wtotal

Just check that you don't exceed the yield point of both members and your deflections are acceptable and you should be good. And provide a vertically slotted hole if you're worried about thermal elongation.

 

[apsix]

Correct me if I'm wrong but it should be;

w1/(E1*I1)-1/4*(76.8/L[sup]4[/sup]) = w2/(E2*I2)

where w1 is the load on the aluminium mullion.