Every tieback anchor should be tested. Some anchors are performance tested to at least 133% of the design load. PTI says to performance test the first 2 or 3 anchors and a minimum of 2% of the remaining anchors. The remaining anchors are proof tested to at least 120% of the design load and sometimes to 133% of the design load. Sometimes, anchors are also given long term creep tests. Refer to Recommendations for Prestressed Rock and Soil Anchors by the Post-Tensioning Institute (PTI).
Obviously it depends on your bond stress to the soil. FHWA recommends limiting bond lengths to less than 40'. I have gone as long as about 50' bond, but by then you really are not transferring much load into the soil(during jacking anyway).
In my area most of the contractor's prefer to keep their anchors to less than 150kips. This keeps the size or their rams somewhat reasonable and easy to move around in the field with a skid steer.
While it depends highly upon the soil or rock you have available to create your bond length in. In general I'd say a range of around 100-150 kips will prove to be the most economical, though the size and depth of your beams will be more of a factor in the price of your wall. Most of the cost of the tiebacks is in the installation, maybe only 10% or so is due to testing operations.
PEInc is spot on with the testing criteria.
dcarr82775 - what do you mean when you say that for long bond lengths "you really are not transferring much load into the soil?"
If the tieback load isn't being resisted by the soil its bonded to, how can it work?
Optimum design capacity of ground anchor is about 600 to 800 kN.
I suggest you to follow BS 8081: British Standard Code of Practice for Ground Anchorages.
The code describes in detail about ground anchor design, installation, and testing. This code request for acceptance test upto 125% of the design load, on any individual working anchor.
What I meant was that tests have shown diminishing returns with very long bonded lengths. The load simply doesn't get to the far end of the anchor when it is stressed (it may get down there later as load comes off the upper portion of the bonded length). Just saying that it is a bad idea of use very long bonded lengths.
Born2drill is correct. Take a look at the PTI recommendations for tiebacks. Our usual engineering assumption is uniform stress transfer, but in reality there is a much higher load near the front end of the grouted zone. Once the peak hits the soil to grout shear capacity, the load transfers down the tendon and redistributes...ie there is a load limit based on peak stress.
Distribution of bond stress along anchor bond length depends on the ratio of ground modulus Eg to anchor modulus Ea.
The smaller the ratio Ea/Eg (means anchor embedded in hard soils), the greater the stress concentration at the top end of anchor bond length. The above also implies that uniform bond stress will be developed if anchor embedded in soft soils.
(see "Foundation in Tension: Ground Anchors ", Chapter 4: Transfer of Load in Anchors and Associated Problems , by T.H. Hanna)
