Liken this to a compactor running back and forth. In time some added density as well as increase in pressure against the wall takes place. Could get up to passive near the top, but likely significantly less effect with depth. As to this last sentence, I once monitored pressure against a high wall as fill was placed in layers and compacted. The increase in pressure was still detected down about 10 feet from compactors working next to the wall. However, compactors out more than 4 feet had little deep depth effect at the wall like I noted here. I know some soils have a springing effect, but don't depend on it..

For cantilevered, braced, and tiedback, non-gravity walls, both temporary and permanent; the active earth pressure coefficient is usually used to get the lateral earth load and the lateral surcharge load. The vehicular surcharge is usually computed by using Ps = (q x Ka) or by using a Boussinesq analysis which is independent of soil properties.

www.PeirceEngineering.com

Using at rest pressue over active results in an increase of approximately 30-40% in the applied pressure on the back of the wall, which in turn increases over turning moments. If you have an embedded wall then you will need deeper embedment to satisfy your FoS. Similarly if you have a concrete cantilever you may need to increase heal width or add a key.

The topic of active or at rest earth pressure does not depend on the type of loading i.e. vehicle surcharge or soil pressure. It depends, among other things, on the type of wall and the effects of movement/rotation. I think you may be trying to design for dynamic effects, which would not be needed for (most) walls with a vehicle surcharge.

For example if you had a masonry basement wall you would design for at rest pressures because you want to minimise movement.

If you had a garden landscape wall then you would design for active pressures as the effect of rotation/movement would probably not be noticed. There is no point in beefing up your design when the risk are not warranted.

Ok, on a bit of a tangent now... Say I was designing a wall for product storage, where the product is loaded against the wall and then removed, and then new product is loaded against the wall again. In this case, would you recommend using at rest instead of active pressures (no concern about wall rotation)?

In re-reading your post with a question apparently assuming the wall is elastic. That is an unlikely situation for most walls. Once you apply the traffic and get an increase in pressure, some permanent and some rebound within the soil only,you can't assume no change after the traffic.

I use elastic theory to determine the horizontal contribution of line-, and point- and limited areal loads.

f-d

ípapß gordo ainÆt no madre flaca!

I assume active pressure, since the tiny amount a retaining wall needs to move for the active assumption is easily acceptable.

DaveAtkins

We typically assume a distributed load of 10 kPa (about 200 psf) acting as a surcharge load to account for vehicular traffic. So, the lateral pressure distribution is rectangular as PEinc mentioned above.

For vehicular traffic, I think that since the location of the applied load is constantly varying, the distributed load of 10 kPa accounts for this also. If there is a permanent point load, I would use f-d approach, but I think that vehicular load may not be considered as a permanent point load.

Active pressure for traffic loading is unconservative.

ATSE, why?

I have designed a few thousand (and built many) non-gravity, cantilevered, braced, and tiedback retaining walls for contractors, owners, engineers, and leading US design-build specialty contractors. I can't remember ever using or being required to use at-rest pressure for traffic surcharges.

www.PeirceEngineering.com

Well, I don't want to be on the wrong side of PEInc, for whom I have great respect.
I do not believe a simplified active wedge accurately captures the actual strain behavior of the soil mass behind a partially restrained wall with repetitive high magnitude wheel loads. Lateral load is dependent on the wall system stiffness. Not all cantilever walls behave the same.
For a simplified Ka calc using phi = 34 degrees, Ka = 0.28.
For moist soil wt = 120 pcf, the p_H = lateral earth press = 34 psf / ft + Ka*pv. That's too low, and I'll stand by that claim.
For comparison, consider two cases.
a. Caltrans p_H for culverts = 100 psf / ft. [Section 6]
Caltrans, albeit a politically flawed institution and known for conservative provisions, does have a few smart engineers, and their engineers have made many edits to their Bridge Design Manual for several decades based on lessons learned.
b. Pressures for compactive equipment. See Terzaghi Peck Mezri Article 44 and 45. I do realize that compactive equipment is not the same as repetitive wheel loads. However, it does provide some insight on the wide spectrum of actual lateral pressures behind walls, based on non-static loading.
Do you want to choose the very lowest static pressure (Ka < 0.3) for a wall that will have compacted backfill, then dynamically loaded for the next 50 years?

Thank you, ATSE. Some additional thoughts: canwestengr really did not indicate the type of wall that he referred to. I assumed, rightly or wrongly, that canwestengr was talking about a conventional cantilevered concrete wall or a cantilevered, braced, or tiedback non-gravity wall which is usually a top-down, cut wall - not a compacted, fill wall. If you are trying to build a tiedback or braced, non-gravity wall to hold a new compacted fill; then, yes, the earth load could/should be higher than active earth pressure. You also are probably building the wrong type of wall. Correct me if I am wrong, but I don't think the original question mentioned a wall with compacted soil behind it. I believe that his question related to a some type of cantilevered wall and its movement under cyclic traffic loading. I have never seen this topic addressed in any books or manuals. I have never seen any books or manuals mixing active and at-rest pressures in the same design example. AASHTO 3.11.6.1, Uniform Surcharge Loads (ES) says that for active earth pressure conditions, Ks shall be taken as Ka, and for at-rest conditions, Ks shall be taken as Ko. AASHTO 3.11.6.4, Live Load Surcharge (LS) says that the value of the coefficient of lateral earth pressure K is taken as Ko for walls that do not deflect or move or Ka for walls that deflect or move sufficiently to reach minimum active conditions. So, pick and use the coefficient you think is correct and then you can argue about it later with the reviewer who will always disagree with your choice.

www.PeirceEngineering.com

It's not a cut wall but it's not compacted fill - planning on plopping down precast concrete and dumping in fill to form a ramp for vehicles.

Precast concrete what????? Jersey barriers, solid concrete blocks from the concrete plant, T-Wall, Double-Wall? Redi-Rock? Sounds pretty flexible to me; expect movement. Probably use Ka.

www.PeirceEngineering.com

Solid interlocking blocks. I'd agree it should be flexible

PEInc - follow up questions to argue about. In a friendly way, of course.
Consider four walls that are all 10 ft tall:
1. Cantilever retaining wall (inverted tee) with 10" thick wall and 5' wide footing - typical minimum design
2. Cantilever retaining wall (inverted tee) with 12" thick wall and 7' wide footing
3. Counterfort retaining wall (inverted tee) with perpendicular stiffener ribs tying back of wall to foundation heel, with 7' wide footing
4. Top supported wall or culvert box ("rigid")
- All backfilled, all with traffic loading directly behind
I would argue that as the wall system's rotational stiffness increases, the wall element must be designed for higher effective K values. Usually, structural engineers think of either active or at-rest, but items 2 and 3 above will definitively deliver more than Ka*pv, and less than Ko*pv.
Agree?
Another, somewhat separate item: As the backfill soil wedge behind the wall develops, there is a slight relaxation (from the wall's perspective). However, as the soil is vertically re-loaded with repetitive loading every hour, the wall either keeps rotating out(bad) or the wall pushes back harder (higher horiz load) because it has higher rotational stiffness - which partially depends on the foundation configuration. This is the modulus / strain part of the analysis that engineers seem to disregard.
I've seen many leaning walls, even with apparently good drainage. Seems like the best design method to address this re-loading and induced strain problem is via higher K.
Agree?

I don't see much difference between Walls 1, 2, and 3. Wall 4 would use Ko while the others would have Ka, in my opinion. Walls 1, 2, and 3 would have similar overturning and sliding behavior except for Wall 1's narrower footing. Wall 3's counterforts help reduce the stem and footing thickness and reinforcing. Due to its narrower footing, Wall 1 would have higher bearing pressure and possibly more rotation. Its 10" stem would need more resteel than Wall 2.

As for your last item, I don't agree. if it were true all walls with cyclic loading would eventually fall over. Using a higher earth pressure coefficient does not necessarily meant there is higher earth pressure. It just assures that the wall is designed heavier. You may be overthinking this. If it isn't broke, don't fix it. Nobody has more retaining walls with cyclic loading than DOT's. No one worries more or is more conservative than DOT's. Their cantilevered walls use Ka.

www.PeirceEngineering.com

PEinc - you make a good case. Thank you. I agree with just about everything you said.
If you have the patience and time, consider the walls 1 thru 4 noted above.
The pressure from the soil is dependent on the rotational stiffness and horiz translation of the structure. The soil does not know when we as engineers designate the wall as flexible or unyielding. Most cantilever walls are can be categorized as yielding, but not all.

A counterfort wall (Wall 3) and a 4-sided culvert (Wall 4) or even 3-sided box culvert (inverted U, say Wall 3.5) do not develop a classical phi/2+45 active wedge behind the wall. My argument is that there is a continuum of K values between Ka to Ko, with wall system rotational stiffness as an independent variable, and Ka is the lower bound applicable to Walls 1 and 2 only. K is not a step function.

Terzaghi Peck Mesri make a similar case in Article 45. If you have the 3rd edition of Soil Mechanics in Engineering Practice, pages 332-334. I bet you have this text, but if not I can scan and upload. Here they make the differentiation between soil pressures used for structural elements (higher than Ka), vs soil pressures used for stability (Ka).

On a lesser note, the added horiz pressures (in addition to the lateral pressures induced during compaction behind the wall) depend on the level of compaction after wall construction (and before "service life" begins). Again, Ka is the lower bound.
Smack me down. I'd rather get smacked here on the eng-tips forum than during a high profile peer review.

IMHO, you are letting theory get in the way of history and standard practice. The things to consider are how are most walls designed and how have they performed for many years under many different soil conditions. Earth pressure starts the wall movement; the pressure doesn't result from the wall movement. While I have the utmost respect for TPM, maybe this is all a bit too theoretical? Why fix an age old design method that doesn't seem to be broken? Leave the fancy, theoretical stuff up to professors who have little else to do in the real world. If, all of a sudden, walls start failing under the standard design method, then maybe the method needs to change.

www.PeirceEngineering.com

this thread has me confused. . .

Step 1: Evaluate the earth pressures irrespective of surcharge loading. Use either active or at-rest as the design warrants.

Step 2: Establish the location and magnitudes of the line or point loads that represent the vehicle.

Step 3: Use Bousinessq (sp) elastic solutions (Poulous and Davis) to obtain delta sigma H values for locations on the wall (i.e., locations with depth).

Step 4: Double the answer (bearing in mind that the elastic solutions are not for an elastic half-space, which the boundary condition of the wall.

Graph the change in loads and make your design.

Not sure how the line- or point-loading is processed by either active or at-rest soil friction or lack thereof. I'd use elastic theory.

Now if it's an areal load, maybe I'd think differently?

f-d

ípapß gordo ainÆt no madre flaca!

Hello fattdad,

We're heading back to the discussions about Bousinesseq, 2 X Bousinesseq, and trial wedge surcharge loads. There are lots of conflicting theory on the treatment of surcharges, how, when and where.

Bowles discusses this relative to Spangler's testing as I recall. AASHTO differentiates between traffic surcharge loads (uniform loa dapproximations) parallel to a wall and perpendicular to a wall. All answers can be right depending on the situation and assumptions made.

Not sure I have ever totally resolved this in my mind and I have studied it for many years.

Hello Doctormo!

As do I. . . wonder how to build a box around this conundrum?

That said, whatever I seem to do seems to work!

f-d

ípapß gordo ainÆt no madre flaca!

On a tangent...I have a lane servicing a loading bay that does a 90 degree turn to get around a building above a retaining wall. It is my understanding the vehicle surcharge loading is for traffic parallel to the retaining wall face. What about the situation where the vehicle is accelerating away from the wall or decelerating towards the wall? Would the lateral force from acceleration/deceleration be entirely addressed within the pavement structure with no loading on the wall?

A rule of thumb commonly used is to assume active vehicle surcharge loads = 2 extra feet of retained soil. Design the wall for the new height and at the last step cut the top 2 feet off the wall and draft it up.

@darthsoilsguy2 - interesting rule of thumb, i have never heard of it...and i am not sure if i understand it? Should it be to assume Ko add the extra 2 feet of soil? as that would increase the pressure like Ko would do compared to Ka. Also, there must be a limiting height that that applies to?

We do a fair amount of cantilevered retaining walls and we often get two design parameters from the geotechs.

They will provide an active earth coefficient (k_active) for the general soil load. This value is usually around 0.3.

They will also provide a second coefficient that only applies to surcharge loads. Often this coefficient will be 0.5 (k_surcharge)

Every now and then we will get a wall where we are concerned about excessive cracking or deflection. In those instances, we will have the geotechs provide an in-situ coefficient (k_naught).

I think the rule of thumb comes from a common construction equipment surcharge load that many people like to apply. They just want a blanket 250psf construction surcharge behind a wall. If you are designing for a 1ft wide section, then 250psf becomes equivalent to 250pcf. In Texas, our clays soils usually weigh 110-125 pcf. So...250psf surcharge is roughly equivalent to 2 extra feet of soil. I just realized that the extra 2ft would be getting multiplied by the lower k_active coefficient though versus the higher k_surcharge of 0.5.

The 2' of roadway surcharge for retaining walls comes from AASHTO whereas building codes and design guides typically use 250 psf. The 2' of soil surcharge would only be accurate if the load was from trucks carrying the backfill material, haha. Seriously, the weight of backfill varies from 100 pcf to 145 pcf thus the 2' soil surcharge is not a very good way to approach vehicular live loads.

Current AASHTO provides for additional feet of soil surcharge loads close to back of wall as I recall (within 1 ft.?). In AUS, they typically use 20 kN/m2 for highway surcharges which is around 400 psf. I suppose there is no right answer to this but walls rarely fail due to traffic surcharges so most highway departments do not spend a lot of time on it.

Another way to minimize the effects is to specify high quality backfill material behind walls like many highway departments do. The difference in strength, drainage, and creep properties make it a prudent investment for heavy load areas.

Quote (A rule of thumb commonly used is to assume active vehicle surcharge loads = 2 extra feet of retained soil. Design the wall for the new height and at the last step cut the top 2 feet off the wall and draft it up. )

That's how I learned to do it as well.