[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GSBN: Wind loads and racking



Andrew -

Some delayed input re: your e-mail of last week.

Derek Roff's suggestion that your engineer look at Bruce King's (et al) book
"The Design of Straw Bale Buildings" and Bruce's Ecological Building Network
website (ecobuildnetwork.org) is a good start.  In particular, section 4.3
of the book explores very well the complex behavior that plastered straw
bale walls exhibit when resisting out-of-plane loads.

Although your subject line includes the word racking (which implies
resistance to in-plane loads) I'll focus on out-of-plane loads, because it
is the only load direction for which I believe tying one skin to the other
really matters.  That is, through-tying is not important for carrying
vertical and/or in-plane lateral loads (both being axial loads), unless
those loads are extreme enough and/or the wall tall/thin enough to induce
global buckling.

But I do agree it is important that the two sides of render on a straw bale
wall be tied together in order for the wall to effectively resist
out-of-plane loads (wind or seismic), especially if they are severe.  (This
assumes that any posts and/or exterior pinning do not themselves provide
sufficient resistance to out-of-plane loads)

However, in most circumstances the bales themselves adequately tie the two
sides of render together.  We've all experienced and have heard of (with
rare exception) the extreme difficulty of separating bales from plaster once
the plaster has cured/dried.  So across each bale, the skins are tied
together very well.  BUT, there are the joints between bales.  On page 99 of
DSBB - discussing tension from skin to skin across bales . . .

"It may well be that there is no slippage between bales, . . ."

But it continues . . .

"This discussion also suggests the purpose, and perhaps need, for wire or
string ties that connect the mesh through the wall at bale courses.  A tie
would in theory carry the tension force, relieving both the bale and the
plaster skin from having to do so.  ("in theory" because the test results do
not necessarily confirm this; . . . )  Wall ties will certainly strengthen
and stiffen a wall against out-of-plane lading, and should be considered
essential in conditions of very high seismic risk and/or very tall walls."
(Or cyclonic loads?)

(note: thanks Bruce King and Kevin Donahue for the above quoted text.)

The question for me is always - How much is enough?  Without knowing the
height to width ratio of the wall, the magnitude of the cyclone loads, and
certain specifics of the wall assembly, I can't say whether through ties are
necessary.  As the above paragraph states, through ties would make it
stronger, but maybe much stronger than it needs to be.  I witnessed most of
the out-of-plane EBNet tests, and those walls (described in DSBB) withstood
extremely large loads without through ties.  It's your engineer's call if
the through ties are necessary for your situation.

What is more important to me than through ties, and is not mentioned in your
e-mail, is plaster reinforcement.  Reinforcement that allows each side of
the wall to act in tension as needed, with the opposite side acting in
compression through the plaster itself.

All of the above assumes there is no sheet barrier between the plaster and
the bales.  (If there were a barrier, tying mesh from one side to the other
would be necessary).  It also assumes that the bond between the plaster and
the straw is, and ALWAYS WILL BE strong (enough).  There are three parts to
this:  1. the surface of the bales have enough tooth for the plaster to key
into.  2. the plaster is applied in such a way as to adequately engage the
straw (including the fact that any mesh used is not one that inhibits such a
bond).  3.  the straw in the bales in general and, more susceptibly, at
their exterior surface do not degrade because of sustained exposure to
moisture.

A plastered straw bale wall is a COMPOSITE WALL SYSTEM.  It acts as a UNIT.
This seems to be the case in spite of the joints between the bales, and
without special through-ties.  The friction between bales and/or the ability
for reinforced mesh to span those joints appears to be more than sufficient
to allow the assembly to act as one and resist out-of-plane loads in common
(and even uncommon) circumstances.

I hope this helps, and others feel free to comment if you feel I've
misstated anything.

Martin Hammer

PS -  In your last line you ask about tests for the racking resistance of SB
walls.  Maybe that's why the word racking is in your subject line.  See
Bruce's book and website for those tests.


> I am working with an structural engineer who has designed several SB
> buildings and is insisting on an unusual detail that I have not seen
> before.  Because of the potential cyclonic conditions here (sub-tropical
> but cyclones can occur), he wants to ensure the two sides of render are
> tied together to reliably ensure the wall acts as a sandwich panel to
> resist lateral loads.  He proposes to put a grid of 8 gauge wires at 450
> centres through the bales with the ends bent or looped to key into the
> render, essentially like brick ties.  In so doing he assumes no lateral
> resistance whatsoever from the bales, regardless of compression system
> or pinning.  This detail also requires that the top and bottom plates
> are fixed to the post and beam and floor structure, so that in itself is
> not enough lateral restrain.
>
> My concern is primarily practicality because:
> 1) it would be difficult to render, especially by hand, with these wires
> sticking out.  Frequent puncture wounds
> 2) it would be hard to keep the wires from pushing in and out and moving
> around while rendering and therefore to ensure they get embedded
> properly in the render.
> 3) local cracking and ineffectiveness of the system
> 4) time consuming
>
> One slight modification to improve 1 and 2 above would be to insert the
> wires and bend them down after the brown coat is on, on both sides, and
> before it dries.  3 would not be a big concern with cement render
> (because it would adhere well to the steel), but I don't want to be
> limited to that.
>
> The extreme wind loading here is not that common in most of the SB
> world; nor even in other parts of Australia where more SB buildings are
> built.  I want to maintain flexibility of compression systems and
> renders, basically clients' choice based on who they have trained with
> (I'm an architect, not builder).  And, as I said, the engineer's concern
> is regardless of compression system.
>
> I am confident that this engineer has put a lot of thought into this and
> his concerns are justified.  I also appreciate that he is trying to
> achieve it as inexpensively as possible.  However, if we can reliably do
> without the wires, with proof or come up with a better detail, I'd
> prefer that.
>
> Also, have there been any tests done on the racking resistance of SB walls?
>
> Any thoughts on this would be appreciated.
>
> All the best,
> Andrew