top of page
Search

Understanding Secondary Effects in Post-Tensioning : A Physical Explanation

Adisorn O.



We mostly recall that M = M_primary + M2, but here is a simpler explanation that can be made, so we don't just memorize the math.



1. Moment Decomposition



For a prestressed member, at any section:

M = M_{primary} + M_{secondary}


  • Primary:

    M_{primary} = F*e

    (prestressing force × eccentricity, as if the member were determinate and free to bend)

  • Secondary:

    M_{secondary} = M - F*e

    This is the “extra” moment you only get because restraints prevent the tendon from producing its natural deflected shape.


What can cause this? Yes, the restraint reaction!


M_{secondary} = R_{restraint}*x

where R_{restraint} is the additional reaction caused by compatibility and x is the lever arm.

That’s the essence: secondary moments are just restrained reaction effects redistributed as moments.




2. Reaction Decomposition



Similarly, at supports:

R = R_{static} + R_{restraint}


  • Static part R_{static}: comes from external loads only (dead, live, wind, seismic, etc.).

  • Restraint part R_{restraint}: arises from prestressing when the structure can’t freely deform.


That's our conclusion


If restraint is absent (say, a simply supported single-span girder, free to take the tendon profile without continuity):


  • R_{restraint} = 0

  • So M_{secondary} = 0

  • Meaning: prestress only contributes through F*e.




3. The Unifying Interpretation



  • Secondary = effect of restraint = “compatibility correction”.

  • Without restraint, prestressing acts like a free body load (pure F*e).

  • With restraint, structure fights back → extra reactions → extra moments/shears.



  • If R_{restraint} = 0, then

    M_{secondary} = M - F*e = 0

  • If R_{restraint} ~= 0, all secondary actions are just “reaction forces from restraint”.



Why is post-tensioning a special case?


Restraint and secondary effects mostly occur in post-tensioning because of structural continuity. In pre-tensioning, the member is freely deformed, so that the restraints and hence secondary effects are not present.



How to use this in the ACI318-19 code?


In ACI, the secondary effects from post-tensioning must be included by 1.0P, i.e.

1.2D + 1.6L +1.0P

where P is not the total prestressing forces as we might be misunderstood by its notation P (it shall be replaced by R in my opinion), but only this secondary (restraint) effect.


The systematic way to get this P can be done as following:


  1. Analyse the structure under external load without considering prestressing effect, i.e. to get 1.2D + 1.6L

  2. Apply equivalent prestressing load (some software might allow you to input the physical tendon inside the member).

  3. From step 2, we get {R_restraint}, which only occurs if the structure is indeterminate.

  4. Compute member forces, usually M & V, due to this R_restraint. This can be easily done by hand calculation or a dedicated post-processing script.


    SAP2000 can perform these post-processing steps when you define the load pattern (of prestressing) as HYPERSTATIC. In fact, this can be applied to other similar load patterns like shrinkage and temperature.

ree

AGAIN, load case P is not the member forces directly caused by any equivalent loads from prestressing. They are caused by restraint reactions only!

 
 
 

Comments


bottom of page