Retaining wall deflection 3

[LR11]

Does anyone have experience with allowable deflection of retaining walls.
Online research suggests one of the following: H/100, between H/200 to H/100, no criteria as it's a visual issue.
CRSI Design Handbook mentions H/240 without rotation at base, and neglecting creep which I assume at least doubles the immediate deflection. See attachment.

The other factor is the pressure distribution. Rankine pressure increases as depth increases and net force is 1/3 from the bottom. Trial wedge methods seem to have a net force 2/3 from the base. There's actually more uncertainty with this assumption than the criteria for allowable deflection.

What are your thought on this.

 

[kostast88]

I assume you are talking about concrete cantilever walls. Short answer: H/240

Long answer: It's ok to deflect as long as it doesn't fail. Deflection on retaining walls is more of a stability criteria, rather than serviceability.

I have always based my calcs on the following assumption:
If a cantilever wall is rigid, it deflects too little. Therefore earth pressures should be at rest (K0).
If a cantilever wall is flexible, it deflects too much. Active wedge develops and use active earth pressures (Ka). Cracked section properties are more suitable to justify the soil activation.

EN 1997-1 sets the limit between which scenario you got:

Depending on your scenario you would either calculate a stiff cantilever with a big force, or a flexible/cracked cantilever with a smaller force. You probably want to be safe - design for the worst. It's ok to deflect as long as it doesn't fail.

The point of application of this force is 1/3 from the base because of triangular distribution. This is a design assumption. If you have earthquakes, follow the relevant earthquake code. EN 1998-5 has some additional rules. I have seen earthquake forces at either 1/3 or 1/2 the height from the base in different cases.

Creep: I have never considered it for a retaining wall.

So all in all, depends on the force, which depends on the soil, which depends on an awful lot of things. So I always advise to design at-rest to be safe.

 

[Tomfh]

No one’s going to complain about H/100 unless it’s a critical application. It’s common for walls to deflect around 1degree, and that’s more than H/100.

 

[LR11]

Great, thanks for the comments kostast88 and Tomfh.
This gives me a bit more confidence.