|
|
|
|
|
This image takes you to the menu page.
|
|
Page for Structural Engineers about to design a basement.
There are a 2 major ways for a workforce to CONSTRUCT a basement with waterproof reinforced concrete:
- Very badly - the usual, contractor way.
- Very well. I may be the only basement project manager to consistently achieve very high standards of workmanship but I do so with 100% regularity because I limit my customers to self-builders and others who require quality.
If you know that your client wants this basement for his own use, you can probably relax certain in the knowledge that he wants it well constructed.
There are 3 major ways to DESIGN a basement with waterproof reinforced concrete:
- Over-engineered to try to overcome poor workmanship.
- To BS 8110 (usually Part 1 only, not Part 2 as well).
- Properly.
The problem with designing a waterproof reinforced concrete basement to BS 8110 is that it does not attempt to cover the design of a waterproof structure neither does it cover the use of concretes with higher heat of hydration than a basic C35 concrete that would leak.
The Standard that addresses concrete with high heat of hydration is BS 8007 "Code of practice for Design of concrete structures for retaining aqueous liquids".
This Standard is expressly not to be used for basement design, probably because such structures are allowed limited leaks and vapour loss, but Appendix A of BS 8007 is very useful indeed in addressing the problems using concrete with a higher heat of hydration than C35. So, whilst the Standard cannot be used alone, the Appendix can be used in association with BS 8110.
Waterproof concrete cannot have cracks that leak, obviously.
BS 8007:1987 2.6.2.2 "Cracking arising from temperature and moisture changes in concrete structures can be controlled by reinforcement, by prestress, by movement joints, by temporary open sections closed with subsequent short infill strips, or by a combination of these methods."
I require reinforcement alone to control cracking.
I do not want any avoidable joints.
|
Cracks limited to 0.2mm will self heal autogenously so this has to be the crack limit. BS 8007 provides an equation to calculate crack widths but its use seems limited to a special concrete C35A that is not waterproof enough for a basement dwelling because it contains PFA.
|
BS 8110-2:1985 8.2
a) "The average coefficient of thermal expansion of normal-weight concrete is of the order of 10 x 10-6/°C and 8 x 10-6/°C for lightweight aggregate concrete (see Table 7.3); thus the difference in length of a concrete member 30 m long due to a 33 °C change in temperature could be approximately 10 mm. If this change in length were prevented by complete restraint of the member, it would cause a stress of the order of 7N/mm2 in an unreinforced concrete member made of concrete having a modulus of elasticity of 20 kN/mm2. If such stress were tensile and superimposed upon other already existing tensile stresses, cracking would occur. If, however, the concrete were to be reinforced, the distribution of the cracking would be controlled by the amount, form and distribution of the reinforcement which might even reduce the crack width and spacing to an extent such as to cause no harmful consequence."
|
BS 8007:1987 2.6.2.3 "the amount of reinforcement in each of two directions at right angles within each surface zone should be not less than 0.35% of the surface zone cross section, as defined in Figure A.1 and Figure A.2 for deformed grade 460 reinforcement and not less than 0.64 % for plain grade 250 reinforcement."
Figure A.1, walls: "assume each reinforcement face controls h/2 (half) the depth of concrete."
Figure A.2, slabs: The three examples of surface zone are
- under 300mm thick: top half only.
- 300mm to 500mm thick: top half and bottom 100mm only.
- over 500mm thick: top 250mm and bottom 100mm only.
|
But you must read more about controlling cracks below because these figures are based upon C35A concrete and not the fully waterproof concrete I use.
|
|
Designing to BS 8110, both Parts 1 and 2, and Appendix A of BS 8007 is a far more proper basis of design than either used alone.
Notes about wall design that enable me to achieve very high quality concrete and maintain my 100% success rate that all basements I supervise are fully waterproof from concrete alone.
The Importance of an Upstand.
|
|
It is always essential that a retaining wall extends to above outside finished ground level.
It is usually necessary for face brickwork to start below ground.
Unlike an ordinary house on footings, we need to prevent a horizontal ingress of water.
Please see the sketch at the top of the page and the bigger sketch below for ideas.
However, the upstand is not normally retaining so it just needs crack protection.
|
|
Room to properly place and compact concrete.
Further down the page you will see typical examples of engineers' designs that prevented good workmanship.
Below these articles you will see my diagram showing a selection of steel reinforcement that enables good workmanship.
The Basis Of Design in para 2.1.4 of BS8110-1:1997 states: "Design, including design for durability, construction and use in service should be considered as a whole." Your design has to allow us to achieve the required standard of workmanship.
|
Pouring concrete in walls.
Filling a wall bottom to top in one day results in poor concrete at the bottom. Therefore I form and pour 2m through a wide opening at the top and the remainder through the upstand another day.
|
Crack Protection.
Crack protection reinforcement 'catches' the concrete when it cracks restricting the crack to 0.2mm because that is the gap that good concrete can self-heal and seal. In fact, it allows concrete to shrink the same amount cracking minutely hundreds of times instead of one great gaping hole.
One engineer used A193 mesh in a 155mm thick concrete wall (this wall, against concrete piles, only had to be waterproof, it had no loads upon it but it was 17m long and restrained at both ends as well as against the slab). With this feeble 0.1241% steel the wall cracked vertically every 1200mm or so from about 150mm off the slab to near the top of the concrete pour.
There is no clue in either Standard as to how much steel is needed when using fully waterproof concrete that gets hotter during hydration though it hardens in a fraction of the time compared to the C35A with PFA example in BS 8007.
However, I have never had a wall leak, since using fully waterproof concrete, if it was 250mm or 300mm thick with just A393 mesh to both faces. This is 0.2618% with 10mm bars at 200mm centres in both directions in a 300mm wall. I think the reason less steel than 0.35% works for me is my mix design (P350, CEM I etc.) is hardening, rather than merely setting, within hours.
But if BS 8007 wants at least 0.35% in both directions and if BS 8110-1 wants 0.4% vertically as a 'simplified general rule', then how is best to achieve it?
12mm bars in both directions on 200mm centres to both faces is 0.377% so it might be the most efficient purchase of steel but it is terribly inefficient to tie it when welded mesh costs little more by weight.
A393 and A193 fixed to it, in both faces, is 0.3901% in both directions and much cheaper ..... but: ..... 2 buts. The first that the two sheets should be staggered but not by half the bar spacing else we lose the ability to put our hands through to connect the formwork.
The second is
My walls always have an upstand as well, also in fully waterproof concrete, but only 155mm thick. By the same standards the upstand needs A393 and A193 (but only 1 sheet of each) too. That is all fine and dandy until we think about laps because a lot of cut mesh tends to warp and be uncontrollable.
So the best solution is:
- mesh (200mm centres in both directions) to both faces 2.4m high
- wall poured first to 2m high
- vertical bars tied to every mesh upright on the inner face to within 40mm maximum of the top of the upstand
- distribution bars to be tied to the uprights maintaining the 200mm centres in the mesh below.
Complete the pour.
But in all truth A393 only to both faces has been sufficient in my experience.
I mentioned above how quickly fully waterproof concrete hardens. Concrete is generally wet (until the effect of added gypsum wears off after 2 hours or so); then it loses workability or sets as water is adsorbed into the cement grain surface; and after several reactions crystals form and when crystals from neighbouring grains interlock concrete gets strength, that is, it hardens.
The limited water content in a P350, CEM I, wcr 0.45 concrete means crystals from neighbouring grains of cement are interlocking as soon as any are formed and the extra heat means once crystal formation starts it is quick.
The preferred concrete mix design in BS 8007 is for P325, CEM II, wcr 0.5. Obviously more crystals have to develop to overcome the additional water and PFA slows the chemical reaction.
In my experience, fully waterproof concrete will have hardened to some extent in a quarter the time of the BS 8007 mix, and that is why it needs less steel, and not more, despite its higher peak temperature.
A393 only to both faces, as in the diagram below, will be fine. I guarantee it.
|
Starter Bars.
The workmanship standard of my fully waterproof concrete is to a number of BS codes including BS 8007:1987: 'Code of practice for design of concrete structures for retaining aqueous liquids'.
One of its requirements is that the joint between a slab and a wall upon it should be a proper, good construction joint and waterproof from the concrete alone, so it should not crack. It should not need any tapes, strips or bars or even repair.
This means that the starter bars have to resist the tension upon cooling mentioned above.
12mm starter bars to both faces seems to work. Somehow 16mm bars to both faces is more reassuring.
|
|
|
Cantilever, Not Propped Cantilever.
An apparent advantage of a concrete floor over timber would be the support a concrete floor gives to the top of the retaining walls by propping the walls one side against the walls the other side and the backfill beyond.
But this would mean that the floor had to be in place before the basement was backfilled and that causes problems.
It might also mean that internal basement walls had to be complete before the basement was backfilled as well.
The upstand to the waterproofing is essential. An engineered timber floor works with it far better than concrete.
The extra steel required in the retaining wall is minimal in comparison to the extra cost of a concrete floor.
|
Advantages of Permanent Polystyrene Formwork.
- Concrete Curing.
- No Kickers - which are notoriously hard to waterproof.
- No Threaded Bar Holes - which are also notoriously hard to waterproof.
|
The Floor Over Habitable Basement Space.
An engineered timber floor joist can be any width and any height and stronger than concrete (for the same overall floor construction depth). It can be supported with a web, lattice sides or plywood sides. Services can go through it. It can be top supported on a timber plate or sit in joist supports.
Clearly, engineered timber joists are more flexible than concrete and they are flat on top. With services through and not under, and without the need for a screed to overcome the curvature, timber floor construction depth is usually less than concrete.
|
Any Corbel.
There is a page about corbels. You can have anything.
But a corbel is only usually necessary with a concrete floor over the top because, by the time there is an upstand to contain the propping effect the brickwork needs a ledge outside to start below ground.
Please, therefore, look at the details above and below.
If you can specify engineered timber floor joists and a cantilever retaining wall you will save your client a lot of money versus a propped cantilever wall, a concrete floor, floor screed, a service void requiring the basement to be dug and constructed deeper, and so on.
|
Thickness of Wall Concrete.
The concrete wall thickness can be 115mm, 155mm, 201mm, 251mm or 301mm (and ties can be joined together to create wider walls where concrete pressure is not too great) - Only the 201, 251 and 301 would normally be suitable for a waterproof concrete retaining wall.
|
Bar Spacing.
The Standards state that thinner bars on closer centres is best for crack protection.
.
200mm x 200mm is the best for formwork strength and room for our hands through the steel.
So mesh wherever possible in the walls is the favourite choice.
|
No Fibre Reinforcement.
Fibre reinforcement dries the concrete so properly placing and properly compacting it is very difficult without adding water. But extra water cannot be waterproof, so we cannot have fibres.
In addition, we need more protection from cracking than fibres provide.
|
|
|
Concrete Specification.
Every brand of fully waterproof concrete has at least 350kgs of OPC cement per cubic metre and will always exceed 50N/mm²
I tend to use 20mm aggregate in the slab and the lower 2m of wall then 10mm for the top of the wall where I am filling 800mm high or so through two narrow gaps.
It is a fact that gravel aggreagate expands 50% more than limestone on warming, with granite being in between.
I make an effort to buy limestone but in the South East it can be too expensive.
I also try to buy appropriate cement which usually means 42.5R in cold weather and 42.5N from Spring till Autumn.
|
Capping Bars.
Some engineers like to see capping bars all along the top of a wall and, sometimes, even a central distribution bar.
Concrete cannot be poured through a tube to the bottom or properly compacted with a vibrating poker with this amount of obstruction.
In contrast, I like to have the same bar Shape 21 on 1m centres both horizontally and vertically throughout the wall to control the steel gap throughout, not just along the top which would leave the middle to curl and wave without control.
|
Cover.
Different Standards have slightly different requirements.
BS8110-1 would say 40mm
Table 5.1 in BS 8110-2 says that in mild conditions the nominal cover for a concrete with either 350kgs of cement or a wcr of only 0.45 needs just 20mm cover; whereas those figures increase to 40mm in severe conditions.
BS 8110-1, table 3.2, describes Severe as exposed to severe rain, alternate wetting and drying or occasional freezing or severe condensation.
BS 8007 (at 2.7.6) states that the minimum cover for all steel should not be less than 40mm. But it is probably reasonable that BS 8007, concerned as it is with structures to retain water, has to allow for a vessel to filled and emptied repeatedly.
But a waterproof concrete basement, covered on the inside with permanent insulating formwork and plasterboard, is always in mild conditions inside even if the outside might be considered more severe as water tables rise and fall.
But it could be useful, if reducing your retaining wall from 300mm to 250mm is beneficial) to increase your lever arm length by reducing the cover inside to 20mm.
|
Corner bars.
Half of you claim corners are a weakness and specify corner bars in 3s on 200mm centres.
The other half claim corners are the strongest part and don't specify any corner bars - and I never had one crack.
But we should have crack protection without creating a problem of access for the formwork ties. I would like to see one corner bar to each face instead of 3s. These two bars have an effective height in the wall of just 12mm instead of 24mm which makes life difficult.
|
|
Slab wall and other joints.
BS 8007 is quite clear:
"It is not necessary to incorporate waterstops in properly constructed construction joints."
From BS 8007:
Joints that are sufficiently rough for interlock, perfectly clean and reinforced to prevent cracking will not leak. Concrete needs to be properly placed and compacted.
My joints do not leak.
|
Conclusion.
You can see a sketch of the steel that best suits my high standard of workmanship below.
Everything is on 200mm centres and we will tie the bars in each face opposite each other so that we can get our hands through everywhere.
The space between faces of wall steel is carefully maintained by U bars so that no steel obstructs the concrete and poker getting down to the bottom properly.
We will tie the first mesh to the starter bars so it is 2.4m high then form and pour 2m high through a wide opening with easy access for concrete and pokers to the bottom. Then we will add 12mm bars and form the rest of the wall including the upstand.
|
|
|
Please note:
As well as fully waterproof concrete I always use a particular brand of polystyrene formwork that stays in place because it enables me to use workmen without extensive experience of concreting. This is absolutely necessary because workmen with extensive experience have bad habits (adding water, letting the concrete free-fall too far, vibrating parts too much and other parts not at all, not cleaning first, and so on) and without great care placing concrete and compacting concrete it will leak later.
I will also try very hard to persuade your client to have an engineered timber floor over the basement because it is so easy to prevent a horizontal ingress of water with the ledge board being part of the concrete formwork rather than a very narrow upstand with a concrete floor, having to cast insitu or create a corbel for brickwork.
|
