Geotechnical: Uplift/Heave in top-down construction 1

[ffinybryn]

I am an Engineer working on a soon to be (very) multi-storey building.

The contruction method is "top-down". Steel plunge columns are piled deep into the ground to allow steel erection to start from the ground floor upwards, whilst at the same time, excavation of the top-soil can continue downwards to create 3 basement floors. This saves a lot of time, not waiting for the basements to be concreted before starting steel erection.

The problem is: Steel erection in one area began before the ground excavation started, whilst later areas were erected during/after the excavation of thousands of tonnes of top soil.

Surveys now suggest that the initial areas of steel have risen by 10mm, whilst the later areas are unchanged.

I have suggested to my surprised client (who does not believe steel can go up) that it is as a result of heave ("A" level Geography).

I have Googled the subject briefly, but cannot find anything to back up my theory. As far as I am aware, water tables or soil types are not a factor.

Can anyone back-up my theory or put forward a lay-man explanation to suggest why this has happened. Luckily, another company had a "level datum" in the first area, which has also "uplifted" by the same amount.

Thanks in advance.

 

[Zambo]

I don't understand fully the different construction sequences between the 2 areas. Lets call B3 the lowest basement floor, then coming upwards we have B2, then B1 then 1F floor (US style) then continuing 2F, 3F etc.

Am I correct so say:

1. For area A the bored piles were constructed and the plunge columns (king posts) were cast in. Then the 1F floor steelwork was connected to the plunge columns and off went the steelwork 2F, 3F etc.

2. For area B the bored piles were constructed including the plunge columns. Then thousands of tonnes of soil were excavated. To what depth was this excavated B1, B2, B3?

But for area B there is no mention of a diaphram wall, secant piles, sheet piles or any method of excavation support. So presumably no great depth of excavation was made.

Or has the work progressed further. Is the basement excavation support in place abd has the B3 base slab been cast for either area A or B?

Usually I would only expect uplift if the slab has been cast and at that stage the uplift forces due to the water table level have exceeded the load from the steel structure and concrete floors above.

Where are you working?

 

[ishvaaag]

For floating piles, unloading might also bring up the built structure. It is surmised they (may) be competent by themselves not only for the present loads but for those of the complete building at a safety factor. Even if embedded somewhat at the bottom they may be extracted from it depending on the relative capacities in shear.

 

[ffinybryn]

Thanks for the prompt replies. Zambo, you are very nearly there...

1. The Secant wall piles and internal piles were piled across the site to below B3. Plunge columns established within the internal piles.

2. Cap plates welded to top of plunge columns and ground floor slab poured to sit on top of plunge column caps.

3. In Area A, prior to any excavation, Steel columns erected to level 5 (free standing without a concrete core, which will rise later from the basement later).

4. Excavation of basement initially to B2 level underneath Area A and most of Area B before erection starts in Area B.

5. Steel erection continued in Area B across rest of site.

6. Re-checking the levels of Area A now shows that the column bases on top of the concrete slab/plunge columns has risen by approximately 10mm, whilst the levels in Area B show no significant changes (either up or down).

Without giving away too much client confidentiality, the site is in London. Would the London clay play a part?. The entire site is 60mx60m. Area A is approx 15mx30m.

Hope this helps.

 

[aeoliantexan]

I have no experience with London clay, but carefully monitored heave on a project with about 25 feet of loess over hard clay till. I would guess the till extended to bedrock about 130 feet below the ground surface. A three-level basement was excavated through the loess and 10 feet or so into the till. We had installed heave points in boreholes below the excavation bottom before the work began and also monitored movement at the edge of the excavation and the face of a nearby building. The bottom heaved about 3/4ths of an inch (say 20mm) The edges heaved about 1/4 to 1/2 inch. If we had approached the job the way you described, the interior columns would have risen significantly. Heavily-loaded friction pile groups in the till settled about 5/8ths of an inch on other projects in the area, so I expect that the net movement of the fully-loaded columns would have been upwards.

 

[Zambo]

My feeling on this is that I, like your client, would be surprised by this and might ask why. (usually my initial thought in the case of 10mm errors is that it is a survey inaccuracy unless proved otherwise)

Of course without a construction stage analysis anything I suggest is mere conjecture. But the main uplift will come after the base slab has been cast and the groundwater makes its way through the clay under the secant wall and raises the pore water pressure at the underside of the slab. At this point you must have sufficient laod from the upper floors to withstand the uplift load. If the load is insufficient then the uplift would put the piles into tension and could in an extreme case cause failure.

As it is you will have had some heave, but I expect the piles have been designed for much greater tension than that, and in anycase you already had 5 floors of steelwork above.

aeoliantexan mentions heave of the bottom of the excavation on his project (he seems to have been using Soil Extensometers or other instruments)but I wouldn't have expected this to cause the piles to move upwards.

If any movement at this stage I would have expected minimal downward movement.

 

[ishvaaag]

Only to add that the presence of surrounding soil equilibrates any other chunk present; so any "elastic" rebound coming from suppression of pressure by removal of weight will be also affected by the fact of that the surrounding soil may find a new lower position of equilibrium that means as well some upwards movement in the excavated portion.

 

[TDAA]

I believe, as ish stated, that this is likely due to a rebound of the soil, caused by the loss of the overburdon pressure.

Look at a typical consolidation curve from a textbook, showing void ratio vs pressure. When pressure is released, the soil will rebound to a lower void ratio, resulting in a greater volume. Where dose the volume change happen? Up, along with your piles.

 

[fattdad]

Elastic behavoir in soil is well published. There is some hysterisis, but heave from unloading makes sense.

Not sure about your reference datum. If the survey datum is close to the edge of the 3-story excavation, you could have the datum going down and the steel showing some measure of elastic rebound.

You ruled out dewatering settlement. Is that because you know that there is no water table within 30 ft?

f-d

¡papá gordo ain’t no madre flaca!

 

[Zambo]

I believe it is very likely that there will be some heave as the excavation is carried out. But I expect the nored piles will be at least 1.5m in diameter and probably around 50m below the base slab of B3 I don't think the soil heave is going to be pulling them up.

As far as the water table goes it would be normal that the secant piles (which are into London Clay) will be deep enough and well enough constructed so that water seepage is not ocurring through or under the preimeter wall at anything other than a very slow rate.

Eventually the water will find a way through but by then the base slab should have been constructed and the superstructure loading the substructure from above.

In the meantime the dewatering is just emptying the excavation of water that has been trapped when the secant pile walll was constructed.