AASHTO LRFD Driven Pile Estimated Length

[Melzar]

We have a disagreement about how to estimate the pile length for driven piles using AASHTO LRFD.

The AASHTO LRFD says that static analysis should be used to estimate the pile length, even if a load test or other method will be used during production. So to estimate the length:

phi (static) x Rn (static) = phi (dynamic) x Rn (dynamic)

As an example, consider a 60 kip factored resistance pile. Let's say that we use static analysis with phi=0.4 to estimate the length. So we have to design for a nominal resistance of 60 kips/0.4 = 150 kips. Let's say that this gives us an estimated length of 40 ft. To get the same factored resistance using the dynamic testing, let's say phi=0.65, we need a lower nominal resistance - 60 kips/0.65 = 92.3 kips.

Does this mean that we can expect a dynamic nominal resistance of 92.3 kips for the 40 ft length?

It seems to me that the equation tells us that we have equal confidence in the piles on either side of the equation - not that they are the same pile! The static analysis will give us the average nominal capacity for the pile - the actual capacity is just as likely to be higher as it is lower. So shouldn't we target the dynamic nominal resistance with our static analysis when we are trying to estimate pile length?

In the example, we desire a 60 kip pile and we are using dynamic testing (phi = 0.65). So the required nominal resistance is 92.3 kips. Shouldn't we use static analysis to target 92.3 kips, then, for our estimated length. The estimated pile length would be just as likely to be short as long.

If we use the lower static resistance factor to estimate the length, wouldn't a "smart" contractor identify that the production piles would be shorter, resulting in a jacked bid process?

Am I off base? How do you all determine an estimated length - the AASHTO LRFD static method, or engineering judgment? If the static methods are used with the lower resistance factors, do the production piles consistently end up shorter?

Thank you,

Melzar

 

[DRC1]

LRFD took a simmple thing and made it much more complicaed than, in my opinion, than it had to be. But we are where we are.
Resistance factors can be corralated to the old factors of safety. If we could be 100% certian of the pile capacity, we would design for the factored capacity of the pile (60 kips). Since we can't, we reflect our uncertianty in our pile capacity with resitance factors. Static analysis is far more uncertian than capwap, so in order to achieve at least the factored load, it has a much higher factor than the capwap. The only thing that is equal between the two equations is the certianty that the pile has sufficent capacity.

AASHTO implies you should use these factors to develop estimated lengths. It is certianly conservative, especially if you are not planning on reducing the lenghts based on the test pile program. For example, say you have your example and figure you need 40 feet to give you 156 kips ultmate load by static design and 92 kips by dynamic. You drive 3 testpiles and tap them. Dynamic tests all come back in the 120 to 150 kip range. You can declare the piles good and drive all 40 footers. On the other hand you can analyze the tests, and decide you could drop 5 feet off the length. You could give an order length of 35 feet, or you could just order 3 more test piles, drive test those. say results are 100 to 120kips. This is still acceptable.
Since the static method is more varible than the PDA or capwap, I would say don't design the pile to the 92 kips. You may very well have have test piles that are too short. However, if you are going to test the piles and have some experience with the local soil, you may design for a smaller ultimate capacity.
I' sure this makes every hing clear as mud.
Any questions I wil do my best to answer.