[Strawbale] Re: Load Bearing Multi-storey Straw Bale Tower House Castle

[Rene Dalmeijer]

Chris,

I fully appreciate the fact that you want to push the envelope and are also 
aware of all the challenges you are facing by attempting new techniques and 
combinations thereof. As one of the respondents observed every new 
technique or combination introduces new risks. The initial mail you posted 
listed some of the possible obstacles but besides these there are many more 
that came to my mind you didn't list.

At 07:36 AM 7/31/03, you wrote:
> > 1) Build the two bottom stories with jumbo bales. They will easily take
>the
> > load. Build the other two either with normal bales or even  wood on the
>top
> > floor.
>We have considered using mini hesston bales (8' x 3' x 2') for all or part
>of
>the building and certainly a hybrid design has appeal, but my rough
>calculations suggest it shouldn't be necessary. I think the compressive
>strength
>of 'standard' straw bales, laid conventionally, should be sufficient to
>support

It is not bearing strength that I am considering but stability. It is the 
general opinion that a load bearing SB wall should not be more then 7 bales 
high. If you make it higher it will need some extra supporting structure ie 
350x7x4=9800mm high for normal 2 string bales which is half the height of 
the planned building. Using hesston jumbo bales easily overcomes this 
limitation. Another consideration is that it is far easier to stack a true 
and straight wall with jumbo bales then with standard bales. John Glassford 
will vouch for this fact based on experience.

>the building. The completed building would weigh approximately 100,000 kg
>(60,000 kg walls + 40,000 kg roof, floors etc) which will bear on 25 square
>metres of straw bale on the first course.

Are you sure about the wall weight I expect much more weight the 
stucco/plaster will weigh at least
20x2x800x0.03=960kg per meter wall length which means 48t alone for the 
plaster.

>  This represents a loading of
>approximately 4,000 kg per square metre, which, according to the Ghailene
>Bou-Ali research of 1993, is less than one tenth of their capability.
> >
> > 2) I suggest you make an architectural detail at each floor break to keep
> > rain off the walls. (like a 500-600mm skirt use thin slate to keep it
>light
> > but authentic). An alternative is to let successive floors jut out over
>the
> > previous. Both solutions offer easy access to the necessary jack lifting
> > structure.
>I like this idea provided we can make it look authentic. The planners might
>object if it's not authentic. Certainly tower castles usually had an
>projection
>just below the battlements. I'm not so sure about lower level skirts (or
>kilts!).
>We planned to make the conventional pitched roof project beyond the walls
>on all sides to provide shelter for the walls and disguise this with
>simulated
>battlements above the roof pitch. The trouble with this approach and any
>attempt at sheltering the wall is that we are subject to wind driven rain at
>this location and so any skirt or extended roof would fail to shelter the
>walls. If you frequently have driven rain is there any point in trying to
>shelter
>the wall? Would it not be better to concentrate our efforts on coating our
>straw bales with a material that resisted water penetration and allowed
>moisture to escape?

Even if you have driving rain. It is only a certain percentage that 
actually hits the wall. The fact that the wall is there creates a turbulent 
boundary layer effectively deflecting a lot of the rain. The 'kilts' keep a 
(large) percentage of the rain off the wall which helps because there is 
less wetting in total. Besides a lot of rain is not during driving 
conditions and this effectively is kept of the wall. Another effect with 
kilts is that not all rain hitting the top of the wall can trickle 
unobstructed down to the foot of the wall the kilts effectively shed this 
water at each break. Keeping water off the walls closest to the ground is 
the most crucial because they are less open to drying wind.


> >
> > 3) Don't discount the weight of plaster it is huge. We made a 2m^2 SB
> > plastered sample it weighed in including frame at almost 400kg. ie at 20m
> > high this means 4000kg per meter wall length without any other structure.
> > It is quite a weight but doable.
>The figure I have is 1 sq meter of straw weighs 75kg. It's not clear whether
>that is with or without plaster. If the weight is 200kg per sq m then I'll
>have
>to recalculate.

We count on about 65kg/m^2 bale weight minimum for building quality 2 
string bales. At 30mm clay/earth plaster both sides you could add another 
48 kg. I expect though that including penetration of the plaster into the 
bales the average cover would be about 50-60mm therefore nearer to 
96kg/^m2. Add a bit of  margin for error and higher density of the plaster 
and you get close to 200kg/m^2.

> >
> > 4) I suggest you build a story at a time and then jack it up, which can be
> > done in stages. As SB is much more forgiving then most other building
> > materials you don't need super precision.
>Certainly this is an option and our first inclination. It would certainly be
>quicker. We felt that perhaps it might be more dangerous because people
>would be working under a propped structure with no fail safe. Jacking it
>up a course at a time will take longer but equiment failure would be less
>likely to result in catastrophic collapse. There is also the added bonus
>that
>the weight of the completed structure can be used to compress the wall
>a course at a time, which should lead to greater stability.

You will need a fail safe anyway because the structure gets most of its 
stability form the plaster bale sandwich which is only available when the 
plaster has hardened. It is essential that the building and jacks are 
stayed during all stages.

> >
> > 5) I also suggest you use some method to keep the walls as flat as
>possible
> > thus reducing the amount of stucco needed. (you also need good tight bales
> > which also require less settling)
>Jacking the walls a course at a time would help, because when placing bales
>you would have two points of reference, above and below, to ensure the
>bales are vertically aligned.

One of the problems of SB is that bales are created unequal. This means 
that certain areas will experience more settling then others this only 
becomes obvious when a courses have been stacked. This includes out of 
plane bowing of the bale wall.

>Minimising stucco/render/plaster might also be
>an argument for laying bales on their side as the rendered surface would be
>flatter.

This could be an option for the top story only. The adhesion and cohesion 
of plaster on the cut folded surfaces of a tightly compacted bale are much 
better then to the other surfaces. A proper bond between the plaster and 
the bales is essential for the structural integrity of load bearing SB.


> > 6) Make sure all bearing portions are symmetrically placed otherwise the
> > building is sure to slowy become lopsided.
>This is a problem with our current design, because, although the tower is
>square, we plan to have a large part (50 percent) of the south wall curtain
>wall glazed. We planned to have internal straw bale walls to maintain the
>structural integrity. It would be a U shaped tower with glazing across
>the top of the U.

Again I repeat my initial remark make the openings as symmetrically 
distributed as possible not only bearing areas but also weight 
distribution. 20m is high for load bearing SB. I would also say check or 
balance your priorities.

>We might decided to build a square tower and fit the
>glazing retrospectively.

What difference will that make?


> > 7) I am very wary of the battlements. We all know what to what grief the
> > Sante Fe style leads to in SB construction it is asking for trouble. The
> > parapets are a sure source of leeks. If you have to make the battlements,
> > fake them ie not SB and with the roof going all the way out to and
> > including the previously suggested rain skirt.
>You'll notice I said simulated battlements. We quickly abandoned the idea of
>real battlements because of the problems of keeping moisture out. This is
>why we've gone for a conventional pitched roof (which is still authentic -
>most Scottish tower castles have a pitched roof garret, presumably also
>to keep the rain out). I see no reason why the simulated battlements
>couldn't be straw bale, so long as they are not structural. In fact their
>condition, given their lack of protection, would be a good forewarning
>about the condition of the rest of the walls.

Thats fine and as you suggest a good test for the condition of the rest of 
the SB structure.


> > Besides the structural measures I didn't do any calculations. I think
> > probably point 2 is the most important measure you should take. Keeping
> > rain off the walls is essential for longevity.
>With driven rain, I'm not sure about the practicality of keeping the rain
>off
>the walls. If anyone has experience of unprotected straw bale walls
>resisting
>moisture penetration we would be very interested. Lime render and lime wash
>seems the best bet so far. We might be better off looking at ways to apply
>regular coats of lime wash to seal any cracks.

Yes this is something you should take into account if not only just to 
inspect the exterior surface of the walls I see the 'kilts' playing a role 
here too like window washing balconies.



Rene Dalmeijer