Lateral Pile Analysis: Episode 3 - Brom’s Method – Ultimate Pile Resistance

One of the classic approaches we learn in lateral pile design is Broms’ Method (1964–1965).

It is elegant, closed-form, and widely taught because it gives quick estimates of the ultimate lateral capacity of piles, especially when you’re trying to get a first feel for the problem.

What Broms’ Method Does Well:

• Estimates ultimate lateral resistance (geotechnical ULS).

• Provides maximum bending moment and shear for given lateral loads.

• Works for both:

  • Short piles (rotation controlled -  soil tends to fail first)

  • Long piles (bending controlled - pile material tends to fail)

• Can be used for free-head piles (common in practice).

But… there are Key Assumptions:

• Uniform pile cross-section

• Soil is homogeneous with depth

• Ultimate soil pressures are fully mobilized

• No axial loads

• No variation of soil stiffness with depth

And these assumptions matter.

 Where You Need to Be Careful:

• Long Piles:
  FHWA does not recommend using Broms’ method for long piles because it is based on a simplified subgrade reaction assumption. It can significantly misrepresent real bending behavior.

• Fixed-Head Conditions:
  Also not recommended by FHWA. The head fixity cases in Broms’ solution don’t capture the real rotational restraint seen in practice or in numerical models

Push-Over Analysis ≈ The Practical Evolution of Broms’ Method

When we need a more realistic estimate of the ultimate lateral capacity — especially when soil stiffness varies with depth, the pile head restraint is partial, or loads are large — we move to push-over analysis in p–y curve-based software.

This approach does what Broms’ method intended, but:

• Accounts for nonlinear soil behavior

• Captures deformation patterns

• Shows capacity development as a function of displacement

• And reflects real boundary conditions

This is the approach I demonstrate step-by-step in my Lateral Pile Analysis Course — starting from Broms’ fundamentals and building into full push-over modeling so the why behind the numbers is always clear.