Circular Concrete Water Retaining Tanks Designed to EC2-3 1

[patswfc]

I am currently looking at the design of 3 similar circular tanks with pinned connection at the base of the wall to existing slabs. They are approx. 7m tall with a working water level of 6m. The difference in the 3 tanks is the diameters which are 9.4m,12m and 14.5m.

According to the code, the allowable crack width at the base is 0.125mm as the walls are 300thk.

For early thermal effects to Ciria C660 (better known as drying shrinkage in the US) I calculate that the walls require 16s@125crs for the 0.125mm crack limit (includes use of GGBS in the mix, basalt aggregate, steel forms, etc to help reduce the thermal effects from hydration).

For ring tension the rebar needs to 20s@125crs for the 0.125mm crack limit.

Note, the tensile capacity of the concrete exceeds the max SLS ring tension by a factor of 3, however as cracks will have already formed due to early thermal effects then once the tanks are loaded, these cracks will get bigger due to the ring tension.

As both types of cracks will be vertical and through the full width of the concrete, I would think that both forms of cracking should be added together. If I do this then the required horizontal rebar increases significantly to 32s@125!

I find it difficult to believe that this much horizontal rebar is actually required, nor that others designing such tanks are providing this quantity of rebar.

I've found an example in a text, however this only checks the cracks widths for early thermal and loading separately, and doesnt combine these effects. Surely this isnt correct.

How do you consider crack widths in circular thanks?

Thanks

 

[ukbridge]

In the UK at least it is not common practice to add crack widths due to Early Thermal and Serviceability (i.e. during the design life) at least for bridge abutments and pilecaps and things like that. I think CIRIA C660 talks about it somewhere.

I'm not that surprised that 32s @ 125 are needed for this type of application, in the UK ETC requirements tend to govern the reinforcement for things like abutments and pilecaps. I'm not really familiar with tanks but as allowable crack widths are very small its not particularly surprising that you need such a large amount of reinforcement.

It used to be a great deal simpler than this (the original UK document was a 14pg DMRB standard). However C660 is now the de-facto standard in the UK at least for better or for worse, and in lieu of anything else is the best we generally have.

 

[patswfc]

Thanks ukbridge, agreed that ETC used to be much simpler when i was using BS8007 and Ciria 91, compared to Eurocodes and Ciria 660.

I understand that the it is not common practice to combine the crack widths.
For rectangular tanks assuming cantilevered walls this is how i would usually approach the design, as the thermal cracks at the base of the wall are generally in the vertical direction whereas the flexural cracks at the base are horizontal, thus no need to add these. Also flexural cracks are not as much of a concern for leakage as the crack doesnt pass through the full width of the wall.

This is the first time designing a circular tank. The issue is that the early thermal cracks will be vertical, as will be the load induced direct tension cracks. Direct tension cracks will also be through the full width of the section. In this circumstance i think that i need to combine the crack widths, resulting in a lot of rebar.
Note, the Concrete Centre guide on basement design states that there are cases where ETC and load induced cracking should be combined, I think that this is such a case.

 

[ukbridge]

Thinking about it in further detail, I think its sensible to combine the crack widths as you said for a circular tank because of their nature. Learned a few things there! Also think if the concrete centre suggests it for some cases, thats good enough for me.

All I can really add is "if it needs more steel, it needs more steel", gl

 

[Klitor]

Out of curiosity, what is the length of wall section concrete pours you are using?

 

[patswfc]

Kilter,
Exact pour sequence still to be agreed with contractor, will probably pour in 4 quarters, thus approx 7.5-11.5m lengths.