Embedment Length: Short vs Long Piles / Drilled Shafts

[SaintPatrick]

I am designing a drilled shaft foundation for a high mast lighting tower. There are some differing opinions around the office about the required shaft length and I am seeking some more opinions.

The HML tower is 120' tall and the soil is sand with phi = 0.33. The applied service loads at the pile head are as follows: Axial = 5kip, Shear = 10kip, and Moment = 300kip-ft. The shaft diameter is 4.5' and controlled by the geometry of the anchor bolt connection to the tower.

I originally determined the depth using Broms' Method and came up with a length of 25'. I then input the information into LPile to see what the deflection curve looked like.

Here is my question. Is a short shaft adequate in this situation or is a long shaft required for stability? When a shaft length of 25' is input into LPile, the shaft is rigid and rotates in the soil like a "fence post". Is this acceptable as long as the deflection is within allowable limits (1/2")?

Some people I have discussed this with feel the shaft needs to be lengthened. They feel the shaft needs to have a minimum of two points of zero deflection for stability so the shaft does not behave like a "fence post" and tilt under load. In order to acheive this condition, the shaft has to be 48' long. Is doubling the shaft length to satisfy this requirement necessary or is a short shaft okay?

Any comments are apprectiated. Thanks.

 

[civilperson]

The large diameter of the foundation makes the model as rigid correct. This is NOT flexible so do not put in inflection points.

 

[SaintPatrick]

I realize the diameter of the shaft makes it rigid and I am not putting in any inflection points.

If the shaft is made long enough, then it will bend before the soil at the base of the shaft yeilds and there will be zero deflection at the base.

Is this a necessary requirement for a foundation of this type or is tilting of the foundation allowed?

 

[oldestguy]

Just for the heck of it, I input your data into a pole foundation nomograph that I have used for estimating tower foundations and I come up with a depth of embedment of 16 feet for "average soil" and 20 ft. for "weak" soil.

I'd do some altenative design procedures to get you in the ball park.

Does your computer program allow for submergence of the soil?

 

[STVU]

What makes the foundation rigid is to keep the length/diameter ratio equal to 10 or less. Teng did a paper in 1969 for granualr soils for supporting pole foundations. For N=10 & the other data provided, I am getting 16 feet embedment-very close to oldestguy. 20 ft should suffice.

That paper is titled "Tapered Steel Poles - Caisson Foundation Design", Prepared for USS Corp., Teng and Associates, July 1969

For the structural design portion, refer to "Resistance to Overturning of Single Short Piles" - by Eli Zerniak,
ASCE Structural Journal, Vol. 83, March 1957.

 

[SaintPatrick]

From the previous two post, it sounds like a short shaft is okay to use in this situation.

My concern about using a rigid, short shaft is that the foundation will rotate and the light pole would be tilted. Based on the LPile model, the foundation will rotate about a single point and tilt like a fence post. A 1/2 inch deflection at the top of the foundation would be about 1 foot deflection at the top of the light pole.

A tilted light pole could be perceived as a safety issue with the general public.

 

[DRC1]

Assuming most of your weight is at the lights on top of the mast, plus an allowance for wind, a one foot deflection will increase your overturning. How does that effect your analysis?

 

[tdunaway]

ACI 336.3R-93 "Design and Construction of Drilled Piers" has a section on Ultimate Capacity Method (Broms Method).

It states that capacity for "Short" piers are governed by soil failure and "Long" piers are governed by structural failure of the pier.

Prior to that, it also states that static methods (i.e. Broms Method) are primarily for small diameter piles (although it does not specifically define "small diameter piles"). And that static methods can be employed for preliminary design or as a check of the computer output in simple cases.

All of that considered, I would recommend using some type of computer analysis that considers p-y curve anlysis of the soil like L-Pile or something similar, and then using Broms Method as a comparison.

 

[SaintPatrick]

DRC1, the additional overturning moment is anoter reason why I am not sure it is a good idea to allow the light pole to tilt.

tdunaway, AASHTO also says to use Broms Method for preliminary analysis only. That is one reason why I used L-Pile to compare the results. However, when using L-Pile I am unsure of exactly how to determine required embedment length and none of the books I have reviewed or post on here give any requirements.

I have reviewed previous designs for HML lighting foundations and the drilled shaft lengths compare favorably to the lengths calculated using Broms method. I know this length is long enough to prevent an ultimate collapse, but is it long enough for all serviceability issues?

Still uncertain on whetere or not a long pile is required for this case.

 

[Ron]

In a wind strong enough to cause the maximum deflection at the lights, I doubt that public perception means much....who's gonna be watching?

If your deflection is permanent, then your system probably does not meet the stability requirements of the code (resisting moment > 1.5 times overturning moment)

It has been quite a while since I've used LPile, but check the point where the moment in the shaft goes to zero. That should be your point of fixity.

I haven't run the numbers, but the 20-25 foot depth seems about right.