Structural Design with Aluminum Extrusions

[jheidt2543]

I’m designing some aluminum extrusions for use as the structural framework for large horizontal sliding doors. As most of you know, trying to find the correct allowable stresses to use for aluminum extrusions is no simple task because the allowable stresses are dependent upon the Alloy, the slenderness ratio and the member configuration.

I’m revising a spreadsheet I developed a few years ago to include a more detailed analysis of the extruded member I’m using for the vertical edge to the door panel.
So the first thing is to compute the section properties of the member, which in my case is made up of two extrusions, one a rectangular tube, the other an odd shaped, non-symmetrical “channel”. The rectangular tube is inserted into the “channel” to reinforce the section. Once I have the combined section properties, I have to compute the allowable stresses which are dependent upon the slenderness ratio, which is based on the unbraced length of the member. I have braces, horizontal girts at 21” o.c. (Lb = 21”), therefore S = Lb/ry = 21.0/.6092 = 34.5.

Now there are a whole slew of choices as seen on the attached sheet for computing the allowable stresses. In working through the equations for S1, S2, and >S2 (see the circled area of my spreadsheet), the allowable stresses are really small, 3.8 ksi and even -2.9 ksi! The allowable stresses just seem unreasonably small considering my past design experiences and I don’t understand the negative number. I can space the girts closer which would reduce S, but it does not help much.

This is where I’m stuck and would like some input from the group.

 

[MikeHalloran]

That's not a spreadsheet; that's an image of a spreadsheet.
We can't see the formulae, so we can't check them.

I'd guess that the negative number you worry about derives its sign from the negative number used for wind suction near the top of the page, but we can't check your work from here.

3-ish ksi is a reasonable design stress for soft aluminum.
You didn't mention an alloy or a temper, so it's conservative, unless you're buying a strong alloy, heat treated, facts not in evidence.




Mike Halloran
Pembroke Pines, FL, USA

 

[jheidt2543]

Thanks Mike,
The alloy is noted on the first page, 6063-T5 and the equations are in the first column, but you are correct, you wouldn't can't see them or manipulate them. The S1, S2, and >S2 are strictly a function of the "base" allowable stress - a constant * b/t.
The negative numbers pop out thus: S2 = 14.7 - .370*34.5 = -2.9 in the second column. In the first column I used S2 = 16.6 - .70*34.5 = 3.8 just to see the difference the two equations make. The come from different specification sections, see 3.4.8 and 3.4.11 on the first page.

My problem really is deciding on which equations to use and if the chosen one results in a negative number, what are my options.

 

[IFRs]

You use the stress that is appropriate for the slenderness. S1, S2 and S3 are slenderness ratio thresholds for the allowable stress calculations. For instance, if your slenderness falls between S1 and S2, you use the allowable stress calculated for S2. For a given slenderness ratio, oddball results for calculations you don't need to do are simply superfluous and have no meaning.

 

[IFRs]

Your equation in line 65 is wrong. You don't use the minimum of the three allowable stresses, you use the one appropriate for your slenderness ratio. Also, this only checks the outstanding flange whose measurements are b and t. Have you checked the gross section for compression ( tension, bending, etc )? Also, since this is a wind case, have you used the correct allowable stress usually increased for transitory loading?

 

[IFRs]

On page 3 line 60 the yield strength of 6063-T5 is shown as 30ksi. Is this correct? 6063-T6 is 30ksi but I think 6063-T5 is 25ksi. Also note that the yield stress for 6063-T5 is different in compression and tension, you have used 30ksi which is the tensile yield strength for 6063-T6 but this calculation is for a compression load case.

On page 3 line 62 you show the design stress to be 8.7 for S1. Why is this not 13.5 ksi as shown in the red box on page 1? Why calculate it at all?

For 16.6 - 0.37 x (b/t) to come out to -2.9, (b/t) must equal 52.7. Using 52.7 to calculate the stress for S > S2 I get 3.19 ksi but your result is 4.9. Is there something I am doing wrong?

 

[jheidt2543]

IFRs:

Thank you, you have clarified a number of my problems, not the least of which was thinking in terms of single value properties like steel, when an aluminum alloy has multiple values depending on thickness, tension, compression, shear, or bearing. So, the Fy = 30,000 ksi has no place in my calcs. The 3063-T5 Alloy has the following minimum properties:
Ftu = 22 ksi Tensile Ultimate
Fty = 16 ksi Tensile Yield
Fcy = 16 ksi Compressive Yield
Fsu = 13 ksi Shear Ultimate

Secondly, I was using out of date references, I finally found and downloaded the "Aluminum Design Manual - Specifications & Guidelines for Aluminum Structures", the June 2005 edition, by the Aluminum Association, Inc. Much clearer than the references I was using.

And, yes you are correct, that the section has to be as a whole and then each element of the section has to be reviewed and a weighted average of the allowable stress used.

Thank you for your comments!

 

[Ron]

You seem to have some confusion as to the alloy you are using. Please clarify. Is it 6063T5 or other? Makes a big difference in aluminum. You also reference a 3000 series. These alloys are not typically used for structural extrusions.

Is your structure welded?

 

[jheidt2543]

The alloy is 6063-T5, and the member it is used in is a extruded, not welded. I think Mike has sent me on the right track, thanks for all the input! Once I get it worked out, I'll do a follow up post here.

 

[IFRs]

If you are designing in aluminum, it would be best to pay for the 2015 version of the Aluminum Design Manual. Worth it in my opinion, for any professional.

Why are you extruding in 6063-T5 when -T6 gets you much higher strengths, or even 6061-T6 for more?

I don't understand why you are using a "weighted average of the allowable stress"

Are you looking at combined bending and compression?