Pile Bent Design 1

[umrce]

I recently designed a pile bent and had a very difficult time getting a design to work. Scour combined with the longitudinal temperature forces killed the pile design due to bending. My supervisor, a well respected bridge engineer in our state has since told me that in the good 'ole days they use to not consider temperature forces on steel pile bents due to the flexibility of the piles. Is this still common practice? Was my design grossly conservative?

 

[gwynn]

It depends. How did you determine the temperature forces and how did you apply them to the piles?

A typical approach I see for thermal loads on steel piles is to apply the maximum thermal movement as a displacement to the top of the bent. The piles are then designed based on the load effects generated by the movement of the top of the pile.

 

[graybeach]

Forces in your pile bents should only be a problem if your spans are very long and continuous and your piles are very short and socketed into rock.

 

[OSUCivlEng]

graybeach: I disagree. If you have 30 or 40 feet of scour, it's very difficult to get a pile bent design to work. The piles get huge, like say HP 14X117.

I you are using elastomeric pads then it is very easy to calculate the amount of force transmitted into the bent with temperature and contraction. I don't think you need to apply the total thermal displacement to the bent as the pad with deflect some before the pile does. This is assuming a continuous expansion/slotted bearing plate.

The force transmitted by the pad to the cap is the shear modulus of the elastomer multiplied by the pad area multiplied by the change in length divided by the thickness of the elastomer.

Pile bents work best with integral type abutments with a single row of vertical piles.

 

[graybeach]

Sorry for the confusion. I meant to say "longitudinal forces" meaning that if you have a tall slender pier, longitudinal thermal movements generate very little longitudinal force in the pier.

Scour is definitely an issue for vertical forces because of p-delta and the increased unsupported length.

 

[gwynn]

"I you are using elastomeric pads then it is very easy to calculate the amount of force transmitted into the bent with temperature and contraction. I don't think you need to apply the total thermal displacement to the bent as the pad with deflect some before the pile does. This is assuming a continuous expansion/slotted bearing plate."

Yes, if you have elastomeric pads as an expansion joint you need not apply the full deflection to the piers. At the same time, if you have 30-40 feet of scour, in that condition a fair amount of the thermal movement is going to be taken up by deflection of the piles and applying the force required to displace the bearings is conservative.

 

[snowmachine88]

I've done this both ways. When I worked for the DOT we considered all lateral forces (wind, temperature, braking, stream, ice, etc.) as being applied to the piles with fixed heads in the direction perpendicular to the beams (or slab spans) and with a free head in the direction of the spans.

Now as a consultant, and using RC-Pier instead of a spreadsheet, all of the force to the piles are assumed purely vertical. RC-Pier's explanation is "Please note the analysis is based on the assumption that the cap is infinitely rigid which results into only axial forces in the piles."

This seems completely absurd to me. The piles are C-I-P piles without a rebar cage. Only the steel shell will be effective in bending.

At the DOT we limited pile bents to a maximum pile length of about 25-30' since the bending will become so large from lateral forces. Now I am seeing a design with 100' long piles.

Does the flexibility of the piles and the stiffness of the cap really merit an infinitely rigid assumption?

 

[umrce]

Sounds like the consensus is to include it. To respond to the previous statement yes I did have elastomeric bearing pads. These pads transmitted the force to the piers but I included. A reduction in the force due to the reduced displacement of the pad as the piers deflected. It reduced the force significantly. thanks for all the help.

Snowmanmachine it's statements like that why our chief engineer refuses to buy RCpier. I've locally heard very mixed reviews about that program.

 

[OSUCivlEng]

Snowmachine, you are only looking at the Pile Reaction tab. You have to look in the Analysis tab for the moment in the pile.

RC Pier takes into account lateral loads on the piles and calculates the bending moments in the piles as well. I know because I helped another engineer who designed a pile bent in RC Pier. You have to select "piles integral with cap" under the Frame Analysis Method, which is under the Pier Config. button.

RC Pier will not check the bending strength of the piles, but that is easily done in a spreadsheet. You can export all of the shear & axial forces and the moments to an excel file from the analysis tab. You'll want to select "Load Combination" under Type. You can also filter by member number to just see the member forces in the pile. From there it is pretty easy to calculate the interaction factors for combined bending and axial loads.

If you select the Frame Analysis Method "cap as continuous beam", then yes it will ignore all out of plane forces.

I sound like some kind of Bentley sunshine pumper don't I? I will add that I'm an AutoCAD guy if that helps. Also RC Pier can be buggy at times, causing me to have to close what I'm working on and restart the program.

 

[snowmachine88]

I have seen the option of design the "piles integral with cap". The DOT here has provided tables that list the effective horizontal loads that different size piles can withstand assuming there are embedded in sand with a friction angle of about 30-32 degrees. This allows the use of RC Pier's pile horizontal capacity feature to also check the bending. Finally, since I had similar concerns that have been addressed here I dumped the pile loads into a biaxial concrete column design and modeled the steel shell as reinforcing steel of a similar area. The axial load kept the column in the compression controlled region of a column. This made me feel like a purely axially loaded assumption is not completely invalid.

Really the maximum moment should be experienced somewhere below the cap level in the soil, so the design needs to be done in a pile design program, whereas RC Pier stops at the footing level. I used to use Allpile, but that is shaky at best, especially for long piles.

Typically, from what I have seen most footings are modeled as a rigid plate with only axial loads being applied to the piles. That makes me think it could also be a valid design method for a pile cap as well.

It sounds like I had better require the additional step of designing the piles for their moment as well for anything I plan on signing.

Thanks!

 

[snowmachine88]

Just as a clarification I have attached the loads from the analysis tab. Are these the loads you are referring to when you say I need to get the pile loads from the analysis?

 

[OSUCivlEng]

There's no attachment.

 

[snowmachine88]

Forgot to insert after loading it

 

[OSUCivlEng]

Here is what I would do. You can export all the load combination results to excel and calculate the interaction factors. I have filtered these for just the members that are piles.