Structural design of aluminium panels 2

[canadiancastor]

I recently got a job for an exterior siding installer. They want to get into bending of aluminum panels for use as exterior cladding. The architect requires engineering calculations for these panels. I stumbled upon this thread and am hoping to get the attention of berkshire who said he had done something similar in the past:

https://www.eng-tips.com/viewthread.cfm?qid=317129

I'm wondering if this can be done 100% with regular structural engineering calculations or if I'm going to have to do some lab testing. Some things that come to mind:
[ol 1]
[li]Fatigue of aluminium: I think I can find the fatigue curves of the aluminum we will be using. I'm not sure where to look for number of cycles under wind loading for 50 years[/li]
[li]Residual stress after bending. Our panels will have 90 degree bends on all sides, 90 degrees is way past Euler beam theory, not sure if this is problem[/li]
[li]FEA modeling: I'm thinking of using shell elements to check bending stress. I'm thinking of modeling screws as pined supports[/li]
[li]Stiffness: Competitors seem to glue stiffeners in the back of large panels, probably to stop them from fluttering in the wind. I think the stiffeners are simple aluminum angle extrusions and they are glued with epoxy, but I have not confirmed this yet. As with all aluminium, I suspect deflexion might control design. [/li]
[/ol]

Any advice or tips as to where to start would be appreciated!

 

[drawoh]

canadiancastor,

Are you using commercial off the shelf aluminium siding, or are you going to fabricate your own?

Are you proposing to use aluminium siding as structure?

Here in Canada, most sheet metal is fabricated from aluminium 5052[‑]H32, a work harded, non[‑]heat[‑]treatable grade. The 7075 proposed in your link is an aircraft grade, which is usually machined. The regular shops do not stock it or fabricate it. You can ask them about ordering 6061[‑]0, and heat treating that. I imagine this will be expensive. The high strength heat treated aluminium grades are brittle. I have never tried to verify this, but I would not be surprised to learn that 5052[‑]H32 has the same fracture energy as 7075[‑]T6. 5052 also is very corrosion resistant.

There are all sorts of handbooks on the internet that show recommended minimum bend radii for various grades of aluminium. Your fabricators will want bend radii that conform to their tooling, and you should accommodate them if you want to control cost and quality.

Before you glue stiffeners or anything else to your aluminium siding, remember the Grenfell[ ]Tower.

--
JHG

 

[Tmoose]

Aluminum paneling , like aluminum siding ?

For site forming, like those continuous rain gutter?

Smooth, or some interesting wood-like surface texture , and a clapboard like Z shape?

Unique dimensions, or similar to what is already commercially available ?
Or maybe for replacing sections of NLA vintage siding?

What about "painting" ?

 

[canadiancastor]

We are fabricating and painting our own panels from aluminium sheets. We currently have external suppliers who bends and paint our pannels, but we will start doing it in house this summer (we have an automatic panel bender and an automated paint line that is being installed).

We have two types of panels, both come in 2mm or 3.2mm thickness. The grade that is presently used is 3003H14, from what I gather because it is cheap.

The first type is simple two-bends ("Z" shaped) on the whole permieter which gives us a lip to screw to (usually) omega bars. The screws remain visible for this type of panel.
The second type uses welded studs to fix aluminium extrusions ("Z" section) which is then used to fix the panel. The screws are hidden for this type of panel.

All dimensions are unique, they are measured on site once the windows are installed. The panels are rectangular or square, and can go from 1' x 1' to 5' x 10'. I ran some quick calculations and I don't think there is any way to justify the 5' x 10' without reinforcing (I'm getting 6" of deflexion at mid point).

@drawoh, I have been looking at canadian code requirements for exterior siding. If the siding is considered "non-combustible", I don't think anything is required by code. I see the stifeners as being only there to aid with stiffness, I'm not sure it would cause collapse if they "unglued".

 

[EdStainless]

The comment about Grenfell is related to panels that use a foam backing for both insulation and stiffness.
This works great, until it doesn't. And burning is just one issue.
I recall years ago seeing panels similar to #2 that had transverse stiffeners mounted to the same studs as the edge strips. The stiffeners were top hat shaped and there was a glue to prevent rattling.

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, consulting work welcomed

 

[drawoh]

canadiancastor,

I have never specified 3000 grade aluminium. It is not anodizable. If you are painting your panels, you have no problem with that. You siding is non[‑]combustible if it cannot catch fire. If stiffness is required, your stiffeners cannot fail. Your structure's failure mode almost certainly is buckling.

If you have a steel frame supported by structural aluminium panels, what is thermal expansion and contraction going to do? My understanding is that siding is normally not structural. In their book Why Buildings Fall Down, authors Matthys Levy and Mario Salvadori describe some elaborate schemes for managing thermal expansion. These work fine on non[‑]structural panels. Read their chapter on Structural Dermatology.

--
JHG

 

[MintJulep]

Gluing and painting aluminum for long term durability is non-trivial. What is your warranty period?

You'll need to manage thermal movement.

If what is under your panels is combustible and starts burning then you have made a very nice chimney to help spread the fire. Again, see Grenfell.

 

[dvd]

Try to call someone in engineering at A Zahner Co. they are a standout in the architectural sheet metal industry. Might be able to get a few pointers by being honest and transparent with them.

 

[canadiancastor]

To make this a little more clear:
I am not in charge of setting up the painting or the bending process, but I will have to attest that the panels are structurally sound and will not fall off the building. If I am not able to do so, I will have to find someone who can. The panels are not "structural", meaning to they hold anything else than their own weight and resist lateral loads (they are simply hung on the building as a rain screen). However, I assume structural calculations can be made for the aluminum itself, the screws, omega bars, "Z" bars, and connection to steel studs. The steel studs have to be checked by the engineer over at interior systems, so that's where my responsibility ends.

For the warranty, from what I understood, the folks that are selling us paint powder will attest our process once things are up and running and they can do accelerated weathering on the panels.

Thermal movement I will look into. Maybe slotted connections can work?

 

[dvd]

I have dealt with many aspects of thermal expansion. I think that rigid fixing of these panels with buckling as the means of expansion could be a valid approach, if the fastening system is able to resist the expansion/buckling forces. I have made calculations to determine what constrained thermal growth forces are, and, typically, buckling forces are much less. I have never incorporated buckling into a design, but I have never designed architectural panels ...

 

[m_struct]

For expansion, the panel can buckle assuming the fixing have sufficient capacity. How is contraction being dealt with?

dvd - how are you calculating the thermal expansion force?

 

[dvd]

This gets the point across -

https://www.engineeringtoolbox.com/stress-restricting-thermal-expansion-d_1756.html

 

[rb1957]

I would try to avoid both thermal effects ... expansion and contraction. I would use an oversize hole and a washer. To be fancy (and tidy) use a CSK head and washer.

another day in paradise, or is paradise one day closer ?

 

[CWB1]

My major concern would be controlling wind flutter, so I would be taking a good hard look at modal analysis vs lab testing to ensure these don't fail mechanically. The other concerns mentioned above are simple junior engineering paper studies by comparison.

 

[berkshire]

canadiancastor,
Most of the information I can give you now is contained in the post and discussion with Kenat. I was responsible for fabricating 1/8" (3.2mm) siding panels with welded corners, erected on frames made of perforated channel. We also made 1/4" ( 6.5mm) laminated (Reynobond ) panels. Although unlike Grenfell Towers, anything we used more than 10'-00" above grade had a self extinguishing core. I have been retired for 9 years now and am rapidly losing touch with the latest industry practices. Stay away from 7075 it is expensive and has poor corrosion qualities
B.E.

You are judged not by what you know, but by what you can do.

 

[canadiancastor]

After working on this a bit, I've come to the conclusion that the larger panels act much more as a membrane than as a rigid shell under perpendicular to the face loading. I'm now looking for some equations to determine panel stress for rectangular membranes.
@CWB1: I believe you are might be overthinking this, we have charts in the building code that give us static wind pressures to be used for building cladding. I believe this to be conservative and include all of the dynamics effects of wind.
@berkshire: Did you hapen to use studwelding to fix your 3.2mm aluminum panels to the frames? Or did you have apparent connectors? Also did you have any testing done for fatigue at connections? Some of my connections are pretty close to the edge of the lip, I'm wondering if this could fail under long term load reversal.

 

[berkshire]

canadiancastor,
@berkshire: Did you hapen to use studwelding to fix your 3.2mm aluminum panels to the frames? Or did you have apparent connectors? Also did you have any testing done for fatigue at connections? Some of my connections are pretty close to the edge of the lip, I'm wondering if this could fail under long term load reversal.
For the most part no. When we did jobs like this, the building was measured, and the framework was designed and drawn up in CAD ( Solidworks.), the panels were then designed so that where possible the edges of the panels matched the flanges on the panels. These were assembled in Solidworks to check for conflicts Adjustments made as needed, then sent for fabrication, Studs were only used when there was no other way and holes in flanges were kept at 2.5 hole diameters.Or on the centerline of the flange.

You are judged not by what you know, but by what you can do.

 

[CWB1]

@CWB1: I believe you are might be overthinking this, we have charts in the building code that give us static wind pressures to be used for building cladding. I believe this to be conservative and include all of the dynamics effects of wind.

Sure, if you want to build panels 10x heavier than necessary go for it - static paper study designing every panel with zero deflection in mind. A standard analysis for manufacturing tho would be a quick modal optimization study to understand the effect of flutter on fasteners and know where/how to dampen unsupported sections of the panel.

Your last statement is literally an engineering disaster in the making btw.

 

[canadiancastor]

Your last statement is literally an engineering disaster in the making btw.

I think I understand where you are coming from. However, unlike in manufacturing, each building project is unique in structural engineering and we aren't in the habit of going all the way down to flutter analysis for our designs. I don't see too many buildings falling over under wind loading, even though every geometry would in theory require a specific wind analysis. The line has to be drawn somewhere, and I'm quite comfortable with where I draw it myself. The types of panels we are recreating here already exist everywhere in north-america and have been installed for over 30 years, and more often than not it's 2mm thick instead of the 3mm we will be using.

More to your point, what do you mean by modal optimization? Does this have to do with natural frequency? I haven't done any analysis, but I assume whatever frequency in wind variation will be orders of magnitude less than that of an aluminum panel.

 

[canadiancastor]

From de CSA S6.1-14 Canadian Highway Bridge Design Code Commentary:

A typical aeroelastic effect is excitation due to vortex shedding. When the wind blows across a slender
prismatic or cylindrical body, vortices are shed alternately from one side and then the other giving rise to
fluctuating forces acting along the length of the body at right angles to the wind direction and the axis of
the body. There is, in addition, the tendency for the aerodynamic damping to become negative. The
critical wind speed, Vcr, when the frequency of vortex shedding equals the natural frequency, f0 , of the
structure or component is discussed in Clause A3.2.4. Negative aerodynamic damping characteristics are
also found at certain windspeeds in both lift and torsional motion of bridge decks. A particularly important
situation is produced by the negative aerodynamic damping forces set up at the critical wind speed at
which the vortex shedding frequency coincides with the natural frequency of the structure.
The forces are sometimes referred to as “locked-in” forces rather than negative aerodynamic damping
forces.

Other forms of instability can occur involving the coupling of several modes of vibration. These are
described as “flutter”. They are only likely to affect exceptionally light, flexible structures such as
cable-supported bridges (Ostenfeld 1992). In all of these instances, the problem should be given
special treatment and an expert in the field should be consulted.

I don't think there will be vortex shedding on the back on my panel which is 1 inch from the face of the building. I'm not sure what is meant by "flutter", but I also do not think it applies to my panels as they are only loading from one face the loading will always remain perpendicular to the face.